Structural and biophysical analysis of the proteasomal ...

Purdue UniversityPurdue e-Pubs

Open Access Dissertations Theses and Dissertations

Spring 2015

Structural and biophysical analysis of theproteasomal deubiquitinase, UCH37Marie Elizabeth MorrowPurdue University

Follow this and additional works at: https://docs.lib.purdue.edu/open_access_dissertations

Part of the Biochemistry Commons, and the Biophysics Commons

This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact [email protected] foradditional information.

Recommended CitationMorrow, Marie Elizabeth, "Structural and biophysical analysis of the proteasomal deubiquitinase, UCH37" (2015). Open AccessDissertations. 522.https://docs.lib.purdue.edu/open_access_dissertations/522

https://docs.lib.purdue.edu?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F522&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/open_access_dissertations?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F522&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/etd?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F522&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/open_access_dissertations?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F522&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/2?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F522&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/4?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F522&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/open_access_dissertations/522?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F522&utm_medium=PDF&utm_campaign=PDFCoverPages

PURDUE UNIVERSITY GRADUATE SCHOOL

Thesis/Dissertation Acceptance

To the best of my knowledge and as understood by the student in the Thesis/Dissertation Agreement, Publication Delay, and Certification/Disclaimer (Graduate School Form 32), this thesis/dissertation adheres to the provisions of Purdue University’s “Policy on Integrity in Research” and the use of copyrighted material.

Marie Elizabeth Morrow

STRUCTURAL AND BIOPHYSICAL ANALYSIS OF THE PROTEASOMALDEUBIQUITINASE, UCH37

Doctor of Philosophy

Chittaranjan Das

Andrew D. Mesecar

Christine A. Hrycyna

Chittaranjan Das

Kavita Shah

R. E. Wild 04/15/2015

STRUCTURAL AND BIOPHYSICAL ANALYSIS OF THE PROTEASOMAL

DEUBIQUITINASE, UCH37

A Dissertation

Submitted to the Faculty

of

Purdue University

by

Marie Elizabeth Morrow

In Partial Fulfillment of the

Requirements for the Degree

of

Doctor of Philosophy

May 2015

Purdue University

West Lafayette, Indiana

ii

ACKNOWLEDGEMENTS

To my parents, Jamie and Christine, thank you for everything and for

supporting my dreams. To my siblings, Thomas and K-Kolbz, you kids are the

best, I love you.

To Missy Ray (Kennedy) who made me love chemistry first. I want to be

as intimidating as you were back then.

Thanks to everyone at the University of Florida (GO GATORS!), especially

Carrie Haskell-Luevano who gave me the opportunity to do research in her lab,

and Erica Haslach-Meckes and Anamika Singh who introduced me to research

and inspired me to continue doing this, I’m so lucky to have had the both of you

as mentors.

Thanks to everyone at Purdue who helped me in classes, trained me or

showed me a new technique, let me borrow chemicals and instruments, or just

made me laugh on bad days. First, to my advisor, Chitta Das, who convinced me

to pursue protein crystallography even though I swore I never would get near it. I

have been lucky to be able to learn from you and I appreciate every fight we’ve

had. Seriously. Second, to my committee members who have provided me with

tons of guidance and support, amazing letters of recommendation, and a few

memorable parties: Andy Mesecar, Chris Hrycyna, and Kavita Shah. Thanks to

iii

the entire Hrycyna group for all of their help over the years, especially the great

Karen Olsen, Patty Wiley, and Kelsey Bohn. To Lake Paul for the biophysics help

and the abuse. To my Purdue friends for all our adventures outside of lab: Sarah

St. John, Chris Clark, Chris Collins, Renata Everett, Chad Keyes, Ian Klein,

Vanessa Klein, Ruth Campbell, Karen Olsen and Walter Bosley, Patty Wiley,

Kelsey Bohn, Judy Ronau, and Mike Sheedlo. Special thanks of course to all

former and current members of the Das lab: David Boudreaux, Joe Chaney,

Chris Davies, Myung-Il Kim, Judith Ronau, Mike Sheedlo, Rashmi Shrestha, Amy

Bueno, Raj Babar, and Cameron Wade. Judy, you are the craziest person I’ve

ever met, but grad school has been so much better because of that. I appreciate

all your help in my project and your awesome friendship. Rashmi, you don’t

deserve all the teasing we do, but you’ve grown up a lot since the first time we

met and I’m so glad we’ve been labmates and friends. Mike, we’ve done this

whole thing together, luckily for both of us. Now we finally get to have separate

projects! Thanks for all of the help, BFF.

The last acknowledgment goes to Alex Riehm: I love you!!! The past 5

years haven’t been easy, and I couldn’t have done this without your support,

every day.

iv

TABLE OF CONTENTS

Page LIST OF TABLES ................................................................................................ vii LIST OF FIGURES ............................................................................................. viii ABSTRACT ......................................................................................................... xi CHAPTER 1: INTRODUCTION ............................................................................ 1

1.1 Ubiquitination .............................................................................................. 1 1.2 Deubiquitination .......................................................................................... 4

1.2.1 UCH Family ..................................................................................... 7 1.3 The 26S Proteasome .................................................................................. 8

1.3.1 Deubiquitination at the 26S Proteasome ....................................... 13 1.4 References ............................................................................................... 17

CHAPTER 2: STRUCTURE OF TSUCH37cat-UBVME ....................................... 26

2.1 Introduction ............................................................................................... 26 2.2 Materials and Methods ............................................................................. 27

2.2.1 Synthesis of Ubiquitin Vinyl Methyl Ester ...................................... 27 2.2.2 Cloning, Expression, and Protein Purification of TsUCH37cat ........ 30 2.2.3 Complexation of TsUCH37cat with UbVME .................................... 31 2.2.4 Selenomethionine-labeled Protein Purification .............................. 32 2.2.5 Crystallization and Structure Solution ............................................ 33

2.3 Results ..................................................................................................... 37 2.3.1 Structure of TsUCH37cat-UbVME ................................................... 37 2.3.2 Active Site Binding ......................................................................... 39 2.3.3 Distal Site Binding ......................................................................... 40 2.3.4 Crossover Loop ............................................................................. 41

2.4 Discussion ................................................................................................ 43 2.5 References ............................................................................................... 45

v

Page

CHAPTER 3: KINETIC AND BIOPHYSICAL CHARACTERIZATION OF TSUCH37 ........................................................................................................... 47


3.2.1 Cloning, Expression, and Protein Purification ................................ 50 3.2.2 Analytical Ultracentrifugation ......................................................... 51 3.2.3 Ubiquitin-AMC Hydrolysis .............................................................. 51 3.2.4 Isothermal Titration Calorimetry ..................................................... 52

3.3 Results ..................................................................................................... 53 3.3.1 Analysis of Crystallographic Dimerization of TsUCH37cat-UbVME 53 3.3.2 Kinetic Characterization of TsUCH37cat and TsUCH37FL .............. 54 3.3.3 Analysis of UCH37 Oligomeric State ............................................. 57 3.3.4 Analysis of UCH37 Binding to Rpn13 ............................................ 58


CHAPTER 4: ANALYSIS OF THE TSUCH37∆C46–UBVME STRUCTURE AND THE ROLE OF THE ULD ................................................................................... 64


4.2.1 Molecular Dynamics Simulations ................................................... 67 4.2.2 Site-directed Mutagenesis and Protein Purification ....................... 67 4.2.3 Ubiquitin-AMC Hydrolysis Assays ................................................. 68 4.2.4 Synthesis of Asymmetric Triubiquitin Substrate and Assays ......... 69

4.3 Results ..................................................................................................... 71 4.3.1 Analysis of Molecular Dynamics Simulations ................................. 71 4.3.2 Ubiquitin-AMC Hydrolysis by UCH37 ULD Mutants ....................... 73 4.3.3 Triubiquitin Cleavage by ULD Mutants .......................................... 75


CHAPTER 5: RECENT FINDINGS ON THE STRUCTURE AND ACTIVATION OF UCH37 AND OTHER DEUBIQUITINASES .................................................. 82

5.1 Introduction ............................................................................................... 82 5.2 Analysis of UCH37-Rpn13-Ub and UCH37-NFRKB-Ub structures ........... 82

5.2.1 Crossover loop .............................................................................. 84 5.2.2 NFRKB Mode of Inhibition ............................................................. 85

5.3 Analysis of Kinetic Findings in Vanderlinden et. al and Sahtoe et. al ....... 87 5.4 Small Structures Effect Large Changes: A Review of Deubiquitinases .... 87

5.4.1 Unproductive Active Sites .............................................................. 88

vi

Page 5.4.2 Insertions and the JAMM Domain .................................................. 91 5.4.3 Substrate Filtering Loops ............................................................... 93 5.4.4 Substrate-occluding Loops ............................................................ 95

5.5 Conclusions .............................................................................................. 96 5.6 References ............................................................................................... 98

APPENDIX ....................................................................................................... 102 VITA ................................................................................................................. 117 PUBLICATIONS ............................................................................................... 118

vii

LIST OF TABLES

Page Table 2.1 Table of crystallographic statistics ................................................................ 38 3.1 Dissociation constants of proteasome-associated DUBs and their binding

partners ........................................................................................................ 60 4.1 Lys48-Glu51 distances in ubiquitin-bound PDB structures ........................... 66 4.2 ULD mutations introduced into human UCH37 ............................................. 73

viii

LIST OF FIGURES

Figure Page 1.1 Scheme of mechanisms of ubiquitination and deubiquitination ...................... 2 1.2 Scheme of the types of ubiquitination ............................................................. 3 1.3 Scheme of directionally specific deubiquitination ............................................ 5 1.4 Domain diagrams of UCH family deubiquitinases ........................................... 7 1.5 Structure of the 26S proteasome .................................................................... 9 1.6 Structure of the 20S and AAA ATPases ....................................................... 10 1.7 Scheme of deubiquitination/degradation at the 26S proteasome ................. 14 2.1 Synthesis of UbVME ..................................................................................... 29 2.2 Generation of the TsUCH37cat-UbVME complex .......................................... 32 2.3 Crystals of the TsUCH37cat-UbVME complex ............................................... 34 2.4 Diffraction patterns ....................................................................................... 35 2.5 Active site of TsUCH37 ................................................................................ 39 2.6 Distal site of TsUCH37 ................................................................................. 41 2.7 Crossover loop of TsUCH37cat-UbVME structure ......................................... 42 3.1 Domain diagram of UCH37 and Rpn13 ........................................................ 48

ix

Figure Page 3.2 Structure of human UCH37 .......................................................................... 49 3.3 Analytical ultracentrifugation of the TsUCH37cat-UbVME complex ............... 53 3.4 Michaelis Menten kinetics for TsUCH37cat UbAMC hydrolysis ..................... 55 3.5 Analytical ultracentrifugation of human UCH37 and Rpn13 .......................... 56 3.6 Isothermal titration calorimetry of UCH37 and Rpn13 binding ...................... 58 4.1 Structure of TsUCH37∆C46-UbVME ............................................................... 65 4.2 Mutant triubiquitin synthesis ......................................................................... 70 4.3 Molecular dynamics simulations of TsUCH37-UbVME ................................. 72 4.4 UbAMC hydrolysis by UCH37 ULD mutants ................................................. 74 4.5 HPLC/MS separations of mutant ubiquitin monomers .................................. 76 4.6 Cleavage of mutant triubiquitin by UCH37 in the presence of Rpn13 ........... 78 4.7 Cleavage of varying length K48 polyubiquitin chains by UCH37 .................. 79 5.1 Conformational changes in the ULD domain ................................................ 83 5.2 Activation of UCH37 by crossover loop binding to Rpn13 ............................ 84 5.3 Mode of NFRKB inhibition of UCH37 ............................................................ 86 5.4 Misaligned active sites of deubiquitinating enzymes .................................... 90 5.5 Insertion domains of JAMM metalloproteases .............................................. 92 5.6 Crossover loops for ubiquitin-bound UCH DUBs .......................................... 93 5.7 Occluding loops of USP14 ............................................................................ 95

x

Appendix Figure Page A 6.1 Scheme of Ubl-UIM purification method of endogenous proteasomes ... 103 A 6.2 Polyubiquitination of GFP-titin-cyclinPY substrate .................................... 107 A 6.3 Silver-stained SDS PAGE gel of rabbit 26S proteasome ........................ 109 A 6.4 Activity of UbVME-treated proteasome ................................................... 110 A 6.5 Activity of UbVME- and MG132-treated proteasome .............................. 112

xi

ABSTRACT

Morrow, Marie Elizabeth. Ph.D., Purdue University, May 2015. Structural and Biophysical Analysis of the Proteasomal Deubiquitinase, UCH37. Major Professor: Chittaranjan Das.

Ubiquitin carboxyl-terminal hydrolase 37, or UCH37, is a deubiquitinating

enzyme associated with the 26S proteasome, the primary protein degradation

machinery in eukaryotic cells. UCH37 is responsible for the disassembly of

polymeric ubiquitin chains, or polyubiquitin, which have been ligated onto

proteins in order to target them for degradation. The 26S utilizes two associated

deubiquitinating enzymes, UCH37 and USP14, and one intrinsic, Rpn11, to

remove polyubiquitin chains from substrate proteins as they are unfolded and

translocated into the proteolytic core of the proteasome, where proteins are

cleaved into small peptides and then released for recycling by the cell. UCH37

associates with the proteasome via binding of its C-terminal KEKE motif to the C-

terminus of Rpn13, a proteasomal ubiquitin receptor which ensnares

polyubiquitinated prey for degradation. UCH37 is known to be catalytically

activated upon binding to Rpn13, allowing cleavage of Lys48-linked polyubiquitin

chains from their distal end, an exo-specific deubiquitination. However, free

UCH37 cleaves polyubiquitin poorly and is believed to be autoinhibited by its C-

xii

terminal UCHL5-like domain, or ULD, which may also be responsible for its

oligomerization in solution. This work examines the structural, biophysical, and

catalytic characteristics of UCH37 in order to elucidate its mechanism of

activation by Rpn13, assess its biophysical assembly with Rpn13 within the

greater proteasomal context, and ascertain its mechanism of exo-specificity

despite the proteasome’s processing of a variety of polyubiquitinated substrates.

To this end, a 1.7 Å resolution X-ray crystal structure was solved of the

catalytic domain of a UCH37 homolog from Trichinella spiralis in complex with

ubiquitin vinyl methyl ester (UbVME), a suicide inhibitor substrate. Our structure,

in combination with another solved of a longer construct of TsUCH37 in complex

with UbVME, provided structural insights into the ability of UCH37 to process

polyubiquitin, namely that its C-terminal UCHL5-like domain (ULD) is responsible

for its exo-specific activity due to a network of interactions with ubiquitin’s Lys48.

Through biophysical and kinetic characterization, we have affirmed the

poor activity of UCH37 alone, but do not ascribe it to autoinhibition because it

does not oligomerize as previously thought, rather we find that it sediments in a

monomer-dimer equilibrium in analytical ultracentrifugation experiments. We

have characterized its binding and activation by Rpn13, finding that UCH37 binds

to Rpn13 with a 22 nM dissociation constant and that mutations to UCH37’s ULD

render it unable to be activated by Rpn13. Interestingly, we have found that while

Rpn13 activates UCH37 for ubiquitin-AMC cleavage, a monoubiquitin fluorogenic

substrate, it appears to slow the enzyme’s processing of Lys48-linked

polyubiquitin chains in our assays.

xiii

Altogether, we have confirmed that UCH37 exists primarily as a monomer

which binds tightly to its proteasomal subunit, Rpn13, and can exo-specifically

cleave Lys48-linked polyubiquitin chains. However, UCH37 may not be activated

as was previously thought, by Rpn13 alone, and likely requires full association

with the 26S proteasome.

1

CHAPTER 1: INTRODUCTION

1.1 Ubiquitination

Ubiquitination occurs through a coordinated enzymatic cascade ending in

the attachment of ubiquitin’s C-terminal glycine (Gly76) to an acceptor lysine

residue via an isopeptide bond. This is achieved through sequential ubiquitin

activation (E1 enzymes), conjugation (E2 enzymes), and ligation (E3 enzymes).

The E1 enzyme, of which there are only two in humans, binds both ubiquitin and

ATP-Mg2+, forms an adenylated ubiquitin intermediate, and then its catalytic

cysteine attacks this adenylated ubiquitin to form a ubiquitin-charged E1,

connected by a high energy thioester bond 1. Ubiquitin is then passed on to one

of about 40 E2 enzymes by attack of their catalytic cysteine to form a charged E2

2. Subsequently, the charged E2 binds to one of hundreds of E3 enzymes, which

then permits ubiquitin ligation onto a target protein either through direct transfer

from the E2 onto the substrate, or by E2 hand-off to the E3 enzyme, which itself

ligates the ubiquitin onto an acceptor lysine (Fig. 1.1) 2. The determinant of either

of these two mechanisms is inherent in the E3 enzyme; RING/U-box ligases

mediate direct E2 transfer, while HECT ligases form a thioester with ubiquitin and

transfer it themselves. RBR ligases (RING in-between RING) act by combining

both

F

m

b

th

ta

re

o

c

b

Figure 1.1: S

mechanisms

ut utilizing

hemselves

andem to b

esponsible

thers are c

hain extens

y the elong

Scheme of

s, binding t

a catalytic

3. After d

build a poly

for chain in

chain elong

sion using t

gating E2.

mechanism

he E2~Ub

cysteine to

deposition

yubiquitin c

nitiation an

gators that

the initiator

ms of ubiqu

complex in

o form a thio

of an initia

hain onto a

d emphasi

t recognize

as a handl

itination an

n the same

oester and

ator ubiquit

a substrate

ze recognit

e the initiat

le 2. Chain

nd deubiquit

manner as

then trans

tin, E2s an

e protein. S

tion of sub

tor ubiquiti

n specificity

tination.

s RING liga

fer the ubiq

nd E3s wo

Specific E2s

strate, whe

n and faci

is generall

2

ases,

quitin

ork in

s are

ereas

litate

y set

F

K

p

h

m

p

a

p

a

Figure 1.2: S

As ub

K63), its mo

olyubiquitin

omotypic c

most well-st

rotein degr

protein de

rogression

ttributed to

Scheme of

biquitin con

onomers ca

n (Fig 1.2).

chain linkag

tudied and

radation at

gradation s

through de

lysosomal

the types o

ntains seve

an be ligate

This host

ges, all res

classical p

the 26S pr

signal, but h

egradation

trafficking

of ubiquitina

en lysines i

ed together

of accepto

sponsible f

polyubiquiti

roteasome.

have also b

of cell cycle

through the

ation.

tself (K6, K

r to form po

or lysines p

for a variet

n chain is

K11-linked

been found

e regulators

e ESCRT p

K11, K27, K

olymeric ch

provides a

ty of cellul

K48-linked

d chains ar

to be cruc

s4-9. K63-lin

pathway, N

K29, K33,

hains, know

wide varie

lar events.

d, the signa

re also prim

cial for cell c

nked chains

F-κB activa

3

K48,

wn as

ety of

The

al for

marily

cycle

s are

ation,

4

and DNA damage repair. Ubiquitin can also be linked through its start methionine

to form linear ubiquitin chains, which are involved in NF-κB activation as well as

cell death 10. Additionally, monoubiquitination serves as a signal for a variety of

cellular events, notably transcriptional regulation and degradation of membrane

proteins 11-13. Currently, little is known about the biological function of chains

linked through K6, K27, K29, and K33 14. Adding further complexity to the

system, ubiquitin chains can be heterotypic, either through mixed ubiquitin chain

linkages that may be “branched” (mixed chain type) or “forked” (two ubiquitin

chains stemming from one monomer) chains, or as mixed ubiquitin-SUMO

chains, all of which are in their early stages of biological characterization 9,15-17.

The mechanisms by which E2s and E3s recognize, bind, initiate, and elongate

ubiquitin chains of varying topologies is still under investigation, as well as

identification of their specific substrates.

1.2 Deubiquitination

In opposition to ubiquitination lies deubiquitination, the hydrolysis of the

isopeptide bond (or Met1-linked amide bonds in linear polyubiquitin) and

subsequent release of ubiquitin from its substrate (Fig. 1.1). This is achieved by a

~100-membered group of enzymes called deubiquitinases, or DUBs. They are

further broken down into mechanistic families, the cysteine proteases and the

metallo-proteases. Cysteine DUBs hydrolyze isopeptide bonds utilizing catalytic

Cys, His, and Asp triads, as well as an oxyanion-stabilizing Gln residue. Metallo-

D

m

w

T

h

p

z

m

u

m

ty

h

(a

F

DUBs have

molecule. T

with the cat

The cystein

ydrolases

roteases (O

inc metall

metalloenzy

biquitin an

monomers.

ypes, direc

ighly spec

associated

igure 1.3: Sc

a zinc ion

his Zn is st

talytic wate

ne proteas

(UCHs), u

OTUs), and

oproteases

ymes (JAMM

d a substr

DUBs hav

ctional chai

cific for on

molecule o

cheme of dir

n responsib

tabilized by

er stabilized

ses include

ubiquitin-sp

d the Mach

s have o

Ms). DUBs

rate protein

ve additiona

n cleavage

ne chain

of the SH3 d

rectionally s

ble for cata

y coordinat

d by hydrog

e four sub

pecific prot

hado Josep

only one

can hydro

n’s lysine re

al specificit

e, and sub

type, for

domain of S

pecific deub

alysis by a

tion to two

gen bonds

bfamilies:

teases (US

phin domai

subfamily,

olyze both i

esidue, as

ties for par

bstrate pref

example

STAM) can

biquitination.

n active si

His and on

to a nearb

the ubiqu

SPs), the

in protease

the JAB

isopeptide

well as be

rticular poly

ference. So

the metal

only cleav

te-bound w

ne Asp res

by Glu res

uitin C-term

ovarian tu

es (MJDs).

B1/MPN/MO

bonds betw

etween two

yubiquitin c

ome DUBs

lo-DUB AM

e Lys63-lin

5

water

idue,

idue.

minal

umor

The

OV34

ween

o Ub

chain

s are

MSH

ked

6

polyubiquitin chains. Other DUBs have specificity for the substrate which has

been ubiquitinated. DUBs that are responsible for chain cleavage have further

specificity for the directionality of their cleavage activity: some remove whole

chains from the site of attachment to a substrate, called en bloc cleavage; some

cleave in the middle of a chain, or endo specificity; and the third group cleaves

from the furthest end of the chain (distal monomer) and removes monomers

sequentially, exo-specific cleavage (Fig. 1.3) 18.

In addition to their DUB domains, many deubiquitinases contain ubiquitin

binding domains (UBDs) which either provide additional stabilization to ubiquitin

binding or confer specificity. Typically, these domains bind monoubiquitin,

sometimes polyubiquitin, with weak affinity in the high micromolar range. They

are most efficient at improving ubiquitin binding when multiple UBDs are found in

one DUB, or if a DUB within a larger complex binds to other proteins containing

UBDs 18,19. Examples of some of the most frequently-occurring UBDs are UBAs

(ubiquitin associated domains), UIMs (ubiquitin interacting motifs), and ZnFs

(zinc finger ubiquitin binding domains) 19. UBDs are crucial for the activity of

many deubiquitinases and are also critical regulators of ubiquitin binding across

the entire proteome.

m

U

p

fa

k

p

re

c

N

F

The

members in

UCHL3, of

roteasome

amily memb

nown to

olyubiquitin

eferred to

atalytic clef

N-terminal U

igure 1.4: D

smallest fa

n humans:

unknown f

; and BAP

bers, UCHL

cleave sm

n chains. T

as the cr

fts 29,30. The

UCH domai

omain diagr

1.

amily of D

UCHL1, k

function; U

P1, which d

L1 and UC

mall substr

his is believ

rossover lo

e other two

n as well as

rams of UCH

2.1 UCH F

DUBs is th

known to b

CH37, invo

deubiquitina

HL3, are si

rates from

ved to be d

oop, which

family mem

s extra C-te

H family deub

amily

he UCHs,

be involved

olved in de

ates histon

ingle-doma

the C-te

due to the p

sterically

mbers, UCH

erminal dom

biquitinases

composed

d in Parkin

eubiquitinat

e H2A 20-2

ain proteins

erminus of

presence o

controls a

H37 and BA

mains (Fig

.

of four fa

nson’s dise

tion at the

8. The first

s which are

ubiquitin,

of a gating

access to

AP1, conta

1.4). UCH3

7

amily

ease;

26S

t two

only

not

loop,

their

in an

37’s

8

C-terminal extension was named the ULD, or UCHL5-like domain, which is

believed to autoinhibit the enzyme’s catalytic activity. At the end lies its KEKE

motif, a region responsible for its binding to the 26S proteasome through the

proteasomal subunit Rpn13, which has a complementary KEKE motif of its own

21,22,31,32. BAP1 has a putative ULD domain, by sequence similarity, which has yet

to be characterized 24,33. BAP1 additionally has a nuclear localization signal at its

far C-terminal end responsible for its cellular localization 24. Both UCH37 and

BAP1 are known to process larger substrates than UCHL1 and UCHL3; UCH37

disassembles polyubiquitin chains at the 26S proteasome, while BAP1

deubiquitinates histone H2A as part of the Polycomb repressor DUB complex

(PR-DUB) 25,26,34,35. UCH37 has been found within the assembly of another

macromolecular complex, the Ino80 chromatin remodeling complex, where it

exists in a generally inactive form, the role of which has yet to be explored 36.

This study focuses on the activity of UCH37, especially as it relates to its role at

the 26S proteasome.

1.3 The 26S Proteasome

The 26S proteasome is a 2.5 MDa proteolytic machine responsible for

degrading the majority of cellular proteins 37-39. It consists of a 20S core particle

composed of proteolytic enzymes and a 19S regulatory particle responsible for

capturing and feeding ubiquitinated proteins into the mouth of the 20S. The 20S

is made up of 4 stacked heptameric rings of structurally similar, but not identical,

subunits 39-42. The external rings contain seven α subunits while the internal rings

c

th

lik

o

th

e

b

FhdA

ontain seve

he proteoly

ke (β2), an

f sequence

hreonine fo

nsures clea

y the α sub

Figure 1.5: ighlighted ieubiquitina

Adapted from

en β subun

tic chambe

d chymotry

e sites 39.

or catalytic

avage into

bunits; the “

Structure in blue, 20

ase in red,m PDB ID 4

nits. The ex

er, where β

ypsin-like (β

The three

activity 39,

small pepti

“closed” for

of the 26SS core par, and ubiq4CR2.

xternal α su

subunits u

β5) activitie

β-subunit

,43. Passag

des, provid

m of the ch

S proteasorticle in yellquitin recep

ubunits act

utilize their

es to cleave

proteases

ge of prote

ded they ca

hannel is on

ome. The 1low, AAA Aptors Rpn1

as the gat

caspase-lik

e proteins a

all rely on

eins through

an enter. En

nly 9 Å wide

19S regulaATPases in10 and Rp

ed channe

ke (β1), try

at a wide va

n an N-term

h this cham

ntry is regu

e, only allow

atory particn purple, Rppn13 in gr

9

l into

psin-

ariety

minal

mber

lated

wing

cle is pn11 reen.

e

to

p

te

th

m

w

m

FcA1

ntry by sma

o the “open

olypeptide

ermini of R

he 20S α su

In ord

must be un

whether un

mounting to

Figure 1.6: rystallized

ATPases, RRYP.

aller peptid

n” form at ~

chain (Fig

pts 2, 3, an

ubunits 42,45

der to trans

folded into

nfolding an

support the

Structure in an ope

Rpts 1-6, are

es, rather t

~20 Å wide

. 1.6) 44. T

nd 5 of the

5-51.

slocate thro

o a linear p

nd transloc

e theory tha

of the 20en and close shown at

than entire

e, the core

This openin

19S AAA

ough the po

polypeptide

cation are

at unfolding

0S and Ased gate (t right. Ada

globular pr

particle ac

ng is facilita

ATPases i

ore of the 2

e chain; the

sequentia

g is coupled

AAA ATPas(left) and tpted from P

roteins 44. W

ccommodat

ated by do

nto the bin

20S core p

ere is som

al, howeve

d to transloc

ses. The 2the heterohPDB ID 4C

When conve

tes an unfo

cking of th

ding pocke

particle, pro

e debate a

er evidenc

cation. This

20S has bhexameric

CR2, 1G0U,

10

erted

olded

he C-

ets of

oteins

as to

ce is

s

been AAA

, and

11

event is achieved by the 19S regulatory particle’s AAA ATPase subunits, Rpts 1-

6, a heterohexameric motor which utilizes ATP hydrolysis to pull polypeptide

chains into the 20S (Fig. 1.6). These Rpts dock to the outer α rings of the 20S

and serve as the base of the 19S RP. Studies of other AAA unfoldases,

especially ClpXP, a bacterial unfoldase, has suggested that translocation and

unfolding are simultaneously achieved through bursts of mechanical force 52,53.

Both ClpXP and the φ29 DNA packaging motor have been shown to exist 90% of

the time in a dwell state, with only 10% of its time spent in a burst of activity 53,54.

This has yet to be confirmed in the 26S Rpts, but cryoEM structures of the Rpts

engaged and disengaged with substrate suggest this may be the case 48-50,55,56.

In addition to Rpts 1-6, the base of the 19S regulatory particle contains

two scaffolding proteins, Rpn1 and Rpn2, as well as the two constitutive ubiquitin

receptors, Rpn10 and Rpn13 39,57. Rpn1 and Rpn2 act to recruit associated

proteins and shuttle factors to the 19S. Through interactions with Ubl (ubiquitin-

like) domains, Rpn1 acts as a docking site for shuttle factors which bring

polyubiquitinated proteins to the proteasome, such as Rad23B and Dsk2 58-61.

Rpn1 is also responsible for recruitment of one of the proteasome’s associated

deubiquitinating enzymes, USP14, discussed below. Thus far, Rpn2 is only

known to anchor one of the intrinsic proteasomal ubiquitin receptors, Rpn13, to

the proteasome, no other shuttle factors or associating ubiquitin receptors 57,61-64.

Rpn10 and Rpn13 are the intrinsic ubiquitin receptors at the proteasome,

although shuttle factors and some temporarily-associating ubiquitin receptors

(Rad23B, Dsk2, Dss1, Ddi1, AIRAP) also bind polyubiquitin and transport it to the

12

proteasome 58,65,66. Interestingly, deletion of these receptors and shuttle factors

(currently known ones) does not impair growth of yeast 60,64. Rpn10 and Rpn13

bind tightly to the proteasome, whereas the other shuttle factors bind weakly and

transiently 61,63. It is possible that there are even more shuttle factors or receptors

to be discovered that may rescue protein degradation upon deletion of this set.

Rpn10, or S5a in humans, utilizes two ubiquitin interacting motifs (UIMs) to bind

polyubiquitin avidly and can also recruit the shuttle factor Rad23B 57,58,67-72. It has

an additional N-terminal von Willebrand A (VWA) domain of unknown function.

Rpn13 contains an N-terminal pleckstrin homology domain referred to as the

pleckstrin-like receptor for ubiquitin (Pru) domain, which binds ubiquitin in a novel

mode compared to other ubiquitin binding domains 21-23,62-64. Rpn13’s C-terminal

domain is responsible for binding UCH37, the second proteasome-associated

deubiquitinase. Rpn10 and Rpn13 lie on the outer edge of the 19S, at opposite

ends, affording polyubiquitin chains a broad surface area for binding as well as

the flexibility of multiple conformations and chain branching (Fig. 1.5) 50,56,57.

Wrapping around and above the 19S base complex lies its lid complex,

one of the least understood components of the 26S. Functions have not been

assigned for its 9 subunits except Rpn11, the proteasome’s constitutive

deubiquitinase. Rpns 3, 5, 6, 7, 9, and 12 contain a proteasome cyclosome

initiation factor (PCI) domain, but the function of these proteins is currently

unknown, aside from acting as scaffolds for other components 73. Rpn11, a

JAMM metallo-DUB, requires dimerization with Rpn8, which contains an inactive

MPN domain, to form its active deubiquitinating module 55,74-80. The lid sits above

13

and around the opening pore of the AAA ATPases, with Rpn11 poised

immediately adjacent to the access point of polyubiquitinated substrates 50,55.

1.3.1 Deubiquitination at the 26S Proteasome

After polyubiquitinated proteins are brought to the 26S proteasome via

shuttle factors and transient ubiquitin receptors, they bind to the proteasome’s

intrinsic ubiquitin receptors, Rpn13 and Rpn1021,23,57,58,62,81. As substrates are

unfolded and translocated into the interior of the core particle, the metallo-

deubiquitinase, Rpn11, cleaves off whole ubiquitin chains from the substrate

protein, releasing them back into the cellular pool of ubiquitin 32,75-77. Rpn11

utilizes a catalytic zinc ion bound by two histidines and an aspartate to cleave

polyubiquitin chains in an en bloc fashion, that is, the entire chain is removed

from its acceptor lysine on a substrate protein74,79. From cryo-EM structures of

the 26S engaged and free of ubiquitinated substrates, it is known that Rpn11

initially exists in an occluded state that is misaligned with the central pore and

ATPases, which subsequently undergoes a dramatic conformational change

upon substrate binding and engagement 50,55. This conformational change aligns

the active site of Rpn11 immediately above the central pore and ATPase ring

opening, which then allows it to cleave entire polyubiquitin chains from an

engaged substrate protein 50,55. Rpn11 is a highly promiscuous DUB capable of

cleaving many different chain types and possibly having endopeptidase activity

a

p

Ftohd(gfr

u

to

s well as

olyubiquitin

Figure 1.7: Sop half of ighlighted eubiquitinagrey with yrom PDB ID

Eithe

nknown), th

o trim polyu

en bloc; th

nated subst

Scheme of the 26S in blue, 2

ases in red,yellow Ub cD 4CR2.

r simultane

he proteaso

ubiquitin ch

his promis

trates that m

f deubiquitinproteasome20S core , and ubiquchain) is s

eous with th

ome’s asso

hains that h

cuity is ne

must feed in

nation/degre is showparticle in

uitin recepthown bind

his activity,

ociated deu

have bound

ecessary g

nto the 26S

radation at n, with the yellow, A

tors in greeing to Ub

or prior to e

biquitinase

d to Rpn13

iven the b

S 55,79,80.

the 26S pre 19S regAAA ATPaen. A ubiqureceptor R

engagemen

s, UCH37 a

3 or Rpn10

broad varie

roteasome.gulatory paases in puuitinated prRpn10. Ada

nt (still curr

and USP14

0 32,34,82-86.

14

ty of

The rticle

urple, otein

apted

rently

4, act

Both

15

UCH37 and USP14 are cysteine protease DUBs which cleave polyubiquitin

chains exo-specifically, that is from the furthest monomer (distal) from the

substrate protein and working their way inwards. USP14 associates with the

proteasome through its Ubl domain, which binds to Rpn1, a known docking point

for other Ubl domain-containing proteins. UCH37, however, binds to an ubiquitin

receptor, Rpn13, through matching KEKE motifs within both of their C-terminal

domains. USP14 and UCH37 have poor basal levels of deubiquitinase activity

alone, but become significantly activated upon recruitment to the 26S

proteasome21,78,84,87,88. They are generally thought to be present in

substoichiometric amounts at the 26S, especially USP14 due to the fact that its

binding partner, Rpn1, is known to bind to multiple proteins at that same site.

Currently it is believed that UCH37 is specific for Lys48-linked chains and that

USP14 may process other chain types, however, the variety of ubiquitinated

species brought to the proteasome indicates that these DUBs are probably more

promiscuous than first thought.

A few theories exist as to what role these associated DUBs play in

proteasome degradation: (1) they recycle monoubiquitin, for further use by the

cell 89,90; (2) they may allow dissociation of chains prior to substrate commitment

for degradation, and in turn rescue a small portion of proteins slated for

degradation that may be inappropriately labeled 34,35; or (3) after a polyubiquitin

chain has been freed from its substrate by Rpn11, the two associated DUBs

sequentially remove ubiquitin monomers until the affinity of the polyubiquitin

chain for Rpn10 or Rpn13 is poor enough to dissociate from the 26S, allowing

16

“resetting” of the proteasome for another round of degradation 39,91. Their

inhibition has been shown to accelerate proteasomal degradation, however,

further work is needed to clarify the biological role of proteasome-associated

deubiquitination 86,92.

Within this work, we present the X-ray crystal structure of a UCH37

homolog bound to ubiquitin, as well as biophysical and kinetic data, which

provides a better structural understanding of the specificity and activation of this

proteasome-bound DUB. Despite the broad spectrum of ubiquitinated

proteasomal substrates, this DUB appears to maintain a limited specificity. We

hope that these studies of UCH37 will obtain a better picture of how

deubiquitinating enzymes in general balance a need for specificity in the face of a

plethora of ubiquitinated proteins, as well as how these enzymes are kept in

inactive/active states by cellular protein partners.

17

1.4 References

1 Schulman, B. A. & Harper, J. W. Ubiquitin-like protein activation by E1 enzymes: the apex for downstream signalling pathways. Nature reviews. Molecular cell biology 10, 319-331, doi:10.1038/nrm2673 (2009).

2 Meyer-Schwesinger, C. et al. A new role for the neuronal ubiquitin C-

terminal hydrolase-L1 (UCH-L1) in podocyte process formation and podocyte injury in human glomerulopathies. J Pathol 217, 452-464 (2009).

3 Smit, J. J. & Sixma, T. K. RBR E3-ligases at work. EMBO reports 15, 142-

154, doi:10.1002/embr.201338166 (2014). 4 Kirkpatrick, D. S. et al. Quantitative analysis of in vitro ubiquitinated cyclin

B1 reveals complex chain topology. Nature cell biology 8, 700-710, doi:10.1038/ncb1436 (2006).

5 Jin, L., Williamson, A., Banerjee, S., Philipp, I. & Rape, M. Mechanism of

ubiquitin-chain formation by the human anaphase-promoting complex. Cell 133, 653-665, doi:10.1016/j.cell.2008.04.012 (2008).

6 Matsumoto, M. L. et al. K11-linked polyubiquitination in cell cycle control

revealed by a K11 linkage-specific antibody. Molecular cell 39, 477-484, doi:10.1016/j.molcel.2010.07.001 (2010).

7 Rape, M. Assembly of k11-linked ubiquitin chains by the anaphase-

promoting complex. Sub-cellular biochemistry 54, 107-115, doi:10.1007/978-1-4419-6676-6_9 (2010).

8 Wickliffe, K. E., Williamson, A., Meyer, H. J., Kelly, A. & Rape, M. K11-

linked ubiquitin chains as novel regulators of cell division. Trends in cell biology 21, 656-663, doi:10.1016/j.tcb.2011.08.008 (2011).

9 Meyer, H. J. & Rape, M. Enhanced protein degradation by branched

ubiquitin chains. Cell 157, 910-921, doi:10.1016/j.cell.2014.03.037 (2014). 10 Iwai, K., Fujita, H. & Sasaki, Y. Linear ubiquitin chains: NF-kappaB

signalling, cell death and beyond. Nature reviews. Molecular cell biology 15, 503-508, doi:10.1038/nrm3836 (2014).

11 Mukhopadhyay, D. & Riezman, H. Proteasome-independent functions of

ubiquitin in endocytosis and signaling. Science 315, 201-205, doi:10.1126/science.1127085 (2007).

18

12 Henry, K. W. et al. Transcriptional activation via sequential histone H2B ubiquitylation and deubiquitylation, mediated by SAGA-associated Ubp8. Genes & development 17, 2648-2663, doi:10.1101/gad.1144003 (2003).

13 Zhang, Y. Transcriptional regulation by histone ubiquitination and

deubiquitination. Genes & development 17, 2733-2740, doi:10.1101/gad.1156403 (2003).

14 Komander, D. & Rape, M. The ubiquitin code. Annu Rev Biochem 81, 203-

229 (2012). 15 Kim, H. T. et al. Certain pairs of ubiquitin-conjugating enzymes (E2s) and

ubiquitin-protein ligases (E3s) synthesize nondegradable forked ubiquitin chains containing all possible isopeptide linkages. The Journal of biological chemistry 282, 17375-17386, doi:10.1074/jbc.M609659200 (2007).

16 Kim, H. T., Kim, K. P., Uchiki, T., Gygi, S. P. & Goldberg, A. L. S5a

promotes protein degradation by blocking synthesis of nondegradable forked ubiquitin chains. The EMBO journal 28, 1867-1877, doi:10.1038/emboj.2009.115 (2009).

17 Guzzo, C. M. & Matunis, M. J. Expanding SUMO and ubiquitin-mediated

signaling through hybrid SUMO-ubiquitin chains and their receptors. Cell cycle 12, 1015-1017, doi:10.4161/cc.24332 (2013).

18 Komander, D., Clague, M. J. & Urbe, S. Breaking the chains: structure and

function of the deubiquitinases. Nature reviews. Molecular cell biology 10, 550-563, doi:10.1038/nrm2731 (2009).

19 Husnjak, K. & Dikic, I. Ubiquitin-binding proteins: decoders of ubiquitin-

mediated cellular functions. Annual review of biochemistry 81, 291-322, doi:10.1146/annurev-biochem-051810-094654 (2012).

20 Maraganore, D. M. et al. UCHL1 is a Parkinson's disease susceptibility

gene. Ann Neurol 55, 512-521 (2004). 21 Yao, T. et al. Proteasome recruitment and activation of the Uch37

deubiquitinating enzyme by Adrm1. Nature cell biology 8, 994-1002 (2006).

22 Hamazaki, J. et al. A novel proteasome interacting protein recruits the

deubiquitinating enzyme UCH37 to 26S proteasomes. The EMBO journal 25, 4524-4536 (2006).

19

23 Qiu, X. B. et al. hRpn13/ADRM1/GP110 is a novel proteasome subunit that binds the deubiquitinating enzyme, UCH37. The EMBO journal 25, 5742-5753, doi:10.1038/sj.emboj.7601450 (2006).

24 Ventii, K. H. et al. BRCA1-associated protein-1 is a tumor suppressor that

requires deubiquitinating activity and nuclear localization. Cancer Res 68, 6953-6962 (2008).

25 Stone, M. et al. Uch2/Uch37 is the major deubiquitinating enzyme

associated with the 26S proteasome in fission yeast. Journal of molecular biology 344, 697-706 (2004).

26 Scheuermann, J. C. et al. Histone H2A deubiquitinase activity of the

Polycomb repressive complex PR-DUB. Nature 465, 243-247 (2010). 27 Kurihara, L. J., Semenova, E., Levorse, J. M. & Tilghman, S. M.

Expression and functional analysis of Uch-L3 during mouse development. Molecular and cellular biology 20, 2498-2504 (2000).

28 Dennissen, F. J. et al. Mutant ubiquitin (UBB+1) associated with

neurodegenerative disorders is hydrolyzed by ubiquitin C-terminal hydrolase L3 (UCH-L3). FEBS letters 585, 2568-2574, doi:10.1016/j.febslet.2011.06.037 (2011).

29 Misaghi, S. et al. Structure of the ubiquitin hydrolase UCH-L3 complexed

with a suicide substrate. The Journal of biological chemistry 280, 1512-1520 (2005).

30 Boudreaux, D. A., Maiti, T. K., Davies, C. W. & Das, C. Ubiquitin vinyl

methyl ester binding orients the misaligned active site of the ubiquitin hydrolase UCHL1 into productive conformation. Proceedings of the National Academy of Sciences of the United States of America 107, 9117-9122 (2010).

31 Qiu, X. B. et al. hRpn13/ADRM1/GP110 is a novel proteasome subunit

that binds the deubiquitinating enzyme, UCH37. Embo J 25, 5742-5753 (2006).

32 Koulich, E., Li, X. & DeMartino, G. N. Relative structural and functional

roles of multiple deubiquitylating proteins associated with mammalian 26S proteasome. Mol Biol Cell 19, 1072-1082 (2008).

33 Sanchez-Pulido, L., Kong, L. & Ponting, C. P. A common ancestry for

BAP1 and Uch37 regulators. Bioinformatics 28, 1953-1956 (2012).

20

34 Lam, Y. A., Xu, W., DeMartino, G. N. & Cohen, R. E. Editing of ubiquitin conjugates by an isopeptidase in the 26S proteasome. Nature 385, 737-740 (1997).

35 Lam, Y. A., DeMartino, G. N., Pickart, C. M. & Cohen, R. E. Specificity of

the ubiquitin isopeptidase in the PA700 regulatory complex of 26 S proteasomes. The Journal of biological chemistry 272, 28438-28446 (1997).

36 Yao, T. et al. Distinct modes of regulation of the Uch37 deubiquitinating

enzyme in the proteasome and in the Ino80 chromatin-remodeling complex. Molecular cell 31, 909-917 (2008).

37 Voges, D., Zwickl, P. & Baumeister, W. The 26S proteasome: a molecular

machine designed for controlled proteolysis. Annual review of biochemistry 68, 1015-1068 (1999).

38 Coux, O., Tanaka, K. & Goldberg, A. L. Structure and functions of the 20S

and 26S proteasomes. Annual review of biochemistry 65, 801-847 (1996). 39 Finley, D. Recognition and processing of ubiquitin-protein conjugates by

the proteasome. Annual review of biochemistry 78, 477-513 (2009). 40 Lupas, A., Koster, A. J. & Baumeister, W. Structural features of 26S and

20S proteasomes. Enzyme & protein 47, 252-273 (1993). 41 Groll, M. et al. A gated channel into the proteasome core particle. Nature

structural biology 7, 1062-1067, doi:10.1038/80992 (2000). 42 Smith, D. M. et al. Docking of the proteasomal ATPases' carboxyl termini

in the 20S proteasome's alpha ring opens the gate for substrate entry. Mol Cell 27, 731-744, doi:10.1016/j.molcel.2007.06.033 (2007).

43 Seemuller, E. et al. Proteasome from Thermoplasma acidophilum: a

threonine protease. Science 268, 579-582 (1995). 44 Rabl, J. et al. Mechanism of gate opening in the 20S proteasome by the

proteasomal ATPases. Mol Cell 30, 360-368, doi:10.1016/j.molcel.2008.03.004 (2008).

45 Gillette, T. G., Kumar, B., Thompson, D., Slaughter, C. A. & DeMartino, G.

N. Differential roles of the COOH termini of AAA subunits of PA700 (19 S regulator) in asymmetric assembly and activation of the 26 S proteasome. J Biol Chem 283, 31813-31822, doi:10.1074/jbc.M805935200 (2008).

21

46 da Fonseca, P. C. & Morris, E. P. Structure of the human 26S proteasome: subunit radial displacements open the gate into the proteolytic core. J Biol Chem 283, 23305-23314, doi:10.1074/jbc.M802716200 (2008).

47 Unverdorben, P. et al. Deep classification of a large cryo-EM dataset

defines the conformational landscape of the 26S proteasome. Proc Natl Acad Sci U S A 111, 5544-5549, doi:10.1073/pnas.1403409111 (2014).

48 Sledz, P. et al. Structure of the 26S proteasome with ATP-gammaS bound

provides insights into the mechanism of nucleotide-dependent substrate translocation. Proc Natl Acad Sci U S A 110, 7264-7269, doi:10.1073/pnas.1305782110 (2013).

49 Beck, F. et al. Near-atomic resolution structural model of the yeast 26S

proteasome. Proc Natl Acad Sci U S A 109, 14870-14875, doi:10.1073/pnas.1213333109 (2012).

50 Lander, G. C. et al. Complete subunit architecture of the proteasome

regulatory particle. Nature 482, 186-191, doi:10.1038/nature10774 (2012). 51 Tian, G. et al. An asymmetric interface between the regulatory and core

particles of the proteasome. Nature structural & molecular biology 18, 1259-1267, doi:10.1038/nsmb.2147 (2011).

52 Maillard, R. A. et al. ClpX(P) generates mechanical force to unfold and

translocate its protein substrates. Cell 145, 459-469, doi:10.1016/j.cell.2011.04.010 (2011).

53 Sen, M. et al. The ClpXP protease unfolds substrates using a constant

rate of pulling but different gears. Cell 155, 636-646, doi:10.1016/j.cell.2013.09.022 (2013).

54 Nyquist, K. & Martin, A. Marching to the beat of the ring: polypeptide

translocation by AAA+ proteases. Trends in biochemical sciences 39, 53-60, doi:10.1016/j.tibs.2013.11.003 (2014).

55 Matyskiela, M. E., Lander, G. C. & Martin, A. Conformational switching of

the 26S proteasome enables substrate degradation. Nature structural & molecular biology 20, 781-788, doi:10.1038/nsmb.2616 (2013).

56 Lasker, K. et al. Molecular architecture of the 26S proteasome

holocomplex determined by an integrative approach. Proceedings of the National Academy of Sciences of the United States of America 109, 1380-1387 (2012).

22

57 Sakata, E. et al. Localization of the proteasomal ubiquitin receptors Rpn10 and Rpn13 by electron cryomicroscopy. Proc Natl Acad Sci U S A 109, 1479-1484, doi:10.1073/pnas.1119394109 (2012).

58 Elsasser, S., Chandler-Militello, D., Muller, B., Hanna, J. & Finley, D.

Rad23 and Rpn10 serve as alternative ubiquitin receptors for the proteasome. The Journal of biological chemistry 279, 26817-26822, doi:10.1074/jbc.M404020200 (2004).

59 Elsasser, S. et al. Proteasome subunit Rpn1 binds ubiquitin-like protein

domains. Nature cell biology 4, 725-730, doi:10.1038/ncb845 (2002). 60 Elsasser, S. & Finley, D. Delivery of ubiquitinated substrates to protein-

unfolding machines. Nature cell biology 7, 742-749, doi:10.1038/ncb0805-742 (2005).

61 Rosenzweig, R., Bronner, V., Zhang, D., Fushman, D. & Glickman, M. H.

Rpn1 and Rpn2 coordinate ubiquitin processing factors at proteasome. The Journal of biological chemistry 287, 14659-14671, doi:10.1074/jbc.M111.316323 (2012).

62 Schreiner, P. et al. Ubiquitin docking at the proteasome through a novel

pleckstrin-homology domain interaction. Nature 453, 548-552 (2008). 63 Chen, X., Lee, B. H., Finley, D. & Walters, K. J. Structure of proteasome

ubiquitin receptor hRpn13 and its activation by the scaffolding protein hRpn2. Molecular cell 38, 404-415 (2010).

64 Husnjak, K. et al. Proteasome subunit Rpn13 is a novel ubiquitin receptor.

Nature 453, 481-488, doi:10.1038/nature06926 (2008). 65 Wilkinson, C. R. et al. Proteins containing the UBA domain are able to

bind to multi-ubiquitin chains. Nature cell biology 3, 939-943, doi:10.1038/ncb1001-939 (2001).

66 Paraskevopoulos, K. et al. Dss1 is a 26S proteasome ubiquitin receptor.

Molecular cell 56, 453-461, doi:10.1016/j.molcel.2014.09.008 (2014). 67 Wilkinson, C. R. et al. Analysis of a gene encoding Rpn10 of the fission

yeast proteasome reveals that the polyubiquitin-binding site of this subunit is essential when Rpn12/Mts3 activity is compromised. The Journal of biological chemistry 275, 15182-15192 (2000).

23

68 Zhang, D. et al. Together, Rpn10 and Dsk2 can serve as a polyubiquitin chain-length sensor. Molecular cell 36, 1018-1033, doi:10.1016/j.molcel.2009.11.012 (2009).

69 Riedinger, C. et al. Structure of Rpn10 and its interactions with

polyubiquitin chains and the proteasome subunit Rpn12. The Journal of biological chemistry 285, 33992-34003, doi:10.1074/jbc.M110.134510 (2010).

70 Walters, K. J., Kleijnen, M. F., Goh, A. M., Wagner, G. & Howley, P. M.

Structural studies of the interaction between ubiquitin family proteins and proteasome subunit S5a. Biochemistry 41, 1767-1777 (2002).

71 Wang, Q., Young, P. & Walters, K. J. Structure of S5a bound to

monoubiquitin provides a model for polyubiquitin recognition. Journal of molecular biology 348, 727-739, doi:10.1016/j.jmb.2005.03.007 (2005).

72 Zhang, N. et al. Structure of the s5a:k48-linked diubiquitin complex and its

interactions with rpn13. Molecular cell 35, 280-290, doi:10.1016/j.molcel.2009.06.010 (2009).

73 Ciechanover, A. & Stanhill, A. The complexity of recognition of

ubiquitinated substrates by the 26S proteasome. Biochimica et biophysica acta 1843, 86-96, doi:10.1016/j.bbamcr.2013.07.007 (2014).

74 Verma, R. et al. Role of Rpn11 metalloprotease in deubiquitination and

degradation by the 26S proteasome. Science 298, 611-615 (2002). 75 Maytal-Kivity, V., Reis, N., Hofmann, K. & Glickman, M. H. MPN+, a

putative catalytic motif found in a subset of MPN domain proteins from eukaryotes and prokaryotes, is critical for Rpn11 function. BMC biochemistry 3, 28 (2002).

76 Yao, T. & Cohen, R. E. A cryptic protease couples deubiquitination and

degradation by the proteasome. Nature 419, 403-407 (2002). 77 Guterman, A. & Glickman, M. H. Complementary roles for Rpn11 and

Ubp6 in deubiquitination and proteolysis by the proteasome. J Biol Chem 279, 1729-1738, doi:10.1074/jbc.M307050200 (2004).


roles of multiple deubiquitylating proteins associated with mammalian 26S proteasome. Molecular biology of the cell 19, 1072-1082 (2008).

24

79 Worden, E. J., Padovani, C. & Martin, A. Structure of the Rpn11-Rpn8 dimer reveals mechanisms of substrate deubiquitination during proteasomal degradation. Nature structural & molecular biology 21, 220-227, doi:10.1038/nsmb.2771 (2014).

80 Mansour, W. et al. Disassembly of lys11 and mixed linkage polyubiquitin

conjugates provides insights into function of proteasomal deubiquitinases rpn11 and ubp6. The Journal of biological chemistry 290, 4688-4704, doi:10.1074/jbc.M114.568295 (2015).

81 Janse, D. M., Crosas, B., Finley, D. & Church, G. M. Localization to the

proteasome is sufficient for degradation. The Journal of biological chemistry 279, 21415-21420, doi:10.1074/jbc.M402954200 (2004).

82 Lee, M. J., Lee, B. H., Hanna, J., King, R. W. & Finley, D. Trimming of

ubiquitin chains by proteasome-associated deubiquitinating enzymes. Molecular & cellular proteomics : MCP 10, R110 003871, doi:10.1074/mcp.R110.003871 (2011).

83 Crosas, B. et al. Ubiquitin chains are remodeled at the proteasome by

opposing ubiquitin ligase and deubiquitinating activities. Cell 127, 1401-1413, doi:10.1016/j.cell.2006.09.051 (2006).

84 Leggett, D. S. et al. Multiple associated proteins regulate proteasome

structure and function. Molecular cell 10, 495-507 (2002). 85 Hanna, J. et al. Deubiquitinating enzyme Ubp6 functions noncatalytically

to delay proteasomal degradation. Cell 127, 99-111 (2006). 86 Lee, B. H. et al. Enhancement of proteasome activity by a small-molecule

inhibitor of USP14. Nature 467, 179-184 (2010). 87 Hu, M. et al. Structure and mechanisms of the proteasome-associated

deubiquitinating enzyme USP14. The EMBO journal 24, 3747-3756 (2005).

88 Jiao, L. et al. Mechanism of the Rpn13-induced activation of Uch37.

Protein & cell 5, 616-630, doi:10.1007/s13238-014-0046-z (2014). 89 Hanna, J., Meides, A., Zhang, D. P. & Finley, D. A ubiquitin stress

response induces altered proteasome composition. Cell 129, 747-759, doi:10.1016/j.cell.2007.03.042 (2007).

25

90 Chernova, T. A. et al. Pleiotropic effects of Ubp6 loss on drug sensitivities and yeast prion are due to depletion of the free ubiquitin pool. The Journal of biological chemistry 278, 52102-52115, doi:10.1074/jbc.M310283200 (2003).

91 Thrower, J. S., Hoffman, L., Rechsteiner, M. & Pickart, C. M. Recognition

of the polyubiquitin proteolytic signal. The EMBO journal 19, 94-102, doi:10.1093/emboj/19.1.94 (2000).

92 Lee, M. J., Lee, B. H., Hanna, J., King, R. W. & Finley, D. Trimming of

ubiquitin chains by proteasome-associated deubiquitinating enzymes. Molecular & cellular proteomics : MCP (2010).

26

CHAPTER 2: STRUCTURE OF TSUCH37CAT-UBVME

2.1 Introduction

Structural approaches to studying ubiquitination/deubiquitination

machinery has yielded extensive information about its detailed mechanisms,

providing vital understanding of these proteins’ ability to recognize either highly

specific chain types or to be grossly promiscuous for any ubiquitinated molecule

available. This approach has given the field incredible insight into the biological

significance of ubiquitination. The structures of many deubiquitinases have been

solved alone and in complex with ubiquitin or a ubiquitin variant. Generally, DUBs

bind monoubiquitin quite poorly, especially if they act as polyubiquitin chain

trimmers in cells. Therefore, in order to capture a DUB-ubiquitin bound state,

covalent linkage of monoubiquitin is required, to prevent dissociation during

crystallography. For this end, a handful of suicide inhibitor ubiquitin variants are

used in structural biology. One of these is ubiquitin vinyl methyl ester (UbVME),

used in this study, which seems to have the highest reactivity with UCH family

DUBs. Here, I have solved the X-ray crystal structure of a UCH37 homolog from

Trichinella spiralis in complex with ubiquitin vinyl methyl ester. This structure

highlights the similarities of UCH-family DUB binding to ubiquitin, as many

contacts are conserved with UCHL1 and UCHL3. However, the active site

27

crossover loop, a structural feature common to UCH enzymes, is not resolved in

the TsUCH37cat-UbVME structure due to a high amount of flexibility that is not

abrogated upon ubiquitin binding, an unexpected result that hints at UCH37’s

mechanism of activation.

2.2 Materials and Methods

2.2.1 Synthesis of Ubiquitin Vinyl Methyl Ester

The synthesis of glycine vinyl methyl ester (GlyVME) has been previously

published, but was modified in our hands (Fig 2.1) 1-3. For the Boc protection

reaction, 8 grams (88 mmol) of 3-amino-1,2-propanediol was dissolved in 150 mL

water, then cooled on ice in order to add 23 grams (105 mmol) of di-tert-butyl

dicarbonate (Boc anhydride), after which the reaction was returned to room

temperature. Then the reaction was brought to pH 10.5 by addition of sodium

hydroxide and the reaction was allowed to run overnight at room temperature.

The reaction was diluted with 100 mL of ethyl acetate, cooled on ice, and then

brought to pH 2.5 with hydrochloric acid. The product was then extracted out with

8 x 50 mL ethyl acetate. The organic layer was washed with NaHSO4 and brine,

dried over sodium sulfate, and then rotovapped down and stored at -20 C. For

the oxidation reaction, 7-8 g of Boc-propanediol was dissolved in 125 mL water,

to which 1.4 molar equivalents of NaIO4 were added. The reaction was stirred for

2-12 hours. The product was extracted out with 3 x 100 mL ethyl acetate, dried

over sodium sulfate, and then rotovapped down. The aldehyde product was used

28

within a day and stored at -20 C. For the Horner Wadsworth Emmons reaction, 1

equivalent of sodium hydride (60% suspension in mineral oil) was added to a

flame-dried round bottom and immediately suspended in 40 mL dry THF, then

purged with N2. The sodium hydride was washed 3 x 30 mL dry THF and then 1

equivalent of trimethyl phosphonoacetate was added over 1 hr on ice. Additional

THF was added as needed to keep the reaction in solution. The Boc-aldehyde

was dissolved in minimal THF and added to the reaction over 1 hr on ice. After

addition, the reaction was allowed to warm to room temperature and run from 6-

12 hrs. The reaction was quenched with 200 mL water and then THF was

removed by rotovapping. The product, Boc-GlyVME, was extracted out with 3 x

50 mL chloroform and the organic layer was washed once with 50 mL of 2%

hydrochloric acid and once with 50 mL saturated sodium carbonate. The organic

layer was dried over sodium sulfate and then rotovapped down and stored until

purification. Boc-GlyVME was purified by silica flash chromatography using a

gradient of 0-20% ethyl acetate in hexanes, pooling only fractions containing the

E isomer. Solvent was rotovapped off, the product was washed 2 x with DCM,

and then rotovapped down again. For Boc deprotection and crystallization of the

final product, 2 molar equivalents of p-toluenesulfonic acid was dissolved in 100-

200 mL diethyl ether, dried over sodium sulfate, and decanted off. Boc-GlyVME

was dissolved in minimal ether and added to the pTSA solution. GlyVME tosyl

salt crystallized out overnight, was filtered out, and stored at -20 C until reaction

with UbMESNa.

in

e

a

in

w

FUCaasa

Ubiqu

n E. coli R

xpression o

t 100,000

ncubated, w

with column

igure 2.1: SUbVME fromCys of TsUC

lcohol, ii) Oldehyde to palt.

uitin1-75 inte

Rosetta ce

overnight a

x g for 1 h

with rocking

buffer and

Synthesis of Ub1-75, then

CH37. B) SyOxidation of produce an E

ein fused to

ells to an

at 18°C. Ce

hour. The

g, at 4°C fo

d then 2-me

UbVME. A) n its formationthetic reacthe alcoho

E-alkene, iv)

o a chitin-bi

O.D. of 0

ells were lys

supernatan

or 2-4 hour

ercaptoetha

Scheme of on of a covations to genl, iii) Horne) deprotectio

inding dom

0.8 and ce

sed by Fre

nt was app

rs. Unboun

ane sulfona

UbVME reaalent thioethnerate GlyVMer Wadsworton of GlyVM

main (CBD)

ells were h

nch press

plied to a c

nd protein w

ate (MESNa

activity showher linkage wME, i) Boc pth Emmons

ME and forma

was expre

harvested

and spun d

chitin resin

was washe

a) was adde

wing formatiowith the cataprotection o

reaction ofation of the

29

ssed

after

down

and

ed off

ed to

on ofalyticf thef thetosyl

30

the column to displace Ub1-75 from the intein group by incubating overnight at

37°C. The eluate was collected and concentrated down to 1.5 mL.

In order to generate UbVME, UbMESNa was incubated with 200 mg

GlyVME and 125 mg NHS dissolved in 1 M NaHCO3 at pH 8 overnight at room

temperature. After incubation, UbVME was dialyzed into 50 mM NaOAc pH 4.5

for 4 hours, then applied to a Mono S cation exchange column for purification

from UbMESNa or hydrolyzed Ub1-75. Fractions were tested for reactivity with

UCHL3 and the most reactive fractions were pooled, concentrated down, and

flash frozen and stored at -80°C.

2.2.2 Cloning, Expression, and Protein Purification of TsUCH37cat

Full-length Trichinella spiralis (Ts) UCH37 in the pET28a vector was sent

from the lab of Katerina Artavanis-Tsakonas, who had previously identified the

enzyme as a UCH family member and confirmed it to be UCH37 by co-

immunoprecipitations and pull-downs of proteasomal subunits 4. Following

standard cloning protocols, the catalytic domain of TsUCH37, residues 1-226,

was subcloned into the pGEX 6P1 vector between BamHI and XhoI digestion

sites. The protein was expressed in E. coli Rosetta DE3 cells to an O.D. of 1.0

and the cells were harvested after expression overnight at 18°C. Cells were lysed

by French press and spun down at 100,000 x g for 1 hour. The supernatant was

applied to glutathione sepharose beads and unbound proteins were washed off

with column buffer (1 x PBS, 400 mM KCl). GST-fused TsUCH37cat was eluted

with reduced glutathione and incubated with PreScission Protease (GE

31

Biosciences) overnight at 4°C. TsUCH37cat was run back over the glutathione

beads to capture GST, and then the pure protein was concentrated down and run

on a HiLoad Superdex 75 for further purification. Pure fractions were

concentrated down, flash frozen, and stored at -80°C.

2.2.3 Complexation of TsUCH37cat with UbVME

Test reactions to complex TsUCH37cat with UbVME were set up in 12 uL

scale to determine the ideal concentration to push complexation to completion.

Three tests were done at 37°C for 3 hours at 29 mg/mL, 14.4 mg/mL, and 9.6

mg/mL TsUCH37cat (Fig. 2.2). For the final scale up reaction, 14.4 mg/mL was

chosen. The scale-up reaction was composed of 600 uL of 14.4 mg/mL

TsUCH37cat, 600 uL UbVME, and 70 uL 1M Tris pH 8.0, for a total volume of 1.9

mL (Fig. 2.2). After 3 hours at 37°C, the reaction was diluted to 4 mL and run on

a Superdex 75 for further purification, but an unexpected higher molecular weight

species was not purified, so all fractions from this step were pooled, buffer

exchanged, and run on a MonoQ anion exchange column in 0-40% 50 mM Tris

pH 7.6, 1 M NaCl, 1 mM DTT over 45 column volumes. The pure complex eluted

at 17% 1 M NaCl (Fig. 2.2). Pure fractions were pooled, concentrated down to 3-

5 mg/mL, flash frozen, and stored at -80°C.

FtePTMa

fo

T

n

w

a

Figure 2.2: Gest titration

PAGE gel TsUCH37cat

MonoQ aniot 280 nm.

In ord

or experim

TsUCH37cat

ative prote

with amino a

n O.D. of

Generation with variedof final r

, reacted fon exchang

2.2.4 Se

der to intro

mental phas

protein wa

ein. TsUCH

acids (Lys,

0.6. The p

of the TsUd concentrareaction offor 3 hrs age column.

elenomethio

oduce heav

sing of pr

as purified

37cat was e

Phe, Thr,

rotein was

UCH37cat-Ubations of Tsf TsUCH3at 37°C. CThe chrom

onine-label

vy atoms in

rotein crys

and compl

expressed

Ile, Leu) an

purified ex

bVME comsUCH37cat

37cat with C) SDS PAmatogram s

ed Protein

nto the TsU

stals, a se

lexed with

in M9 min

nd selenom

xactly as n

mplex. A) SDat 37°C forUbVME a

AGE gel ofshown abov

Purification

UCH37cat-U

elenomethio

UbVME in

imal media

methionine a

native TsUC

DS PAGE gr 3 hrs. B)

at 14.4 mgf fractions ve is monit

n

bVME com

onine enri

addition to

a suppleme

after growin

CH37cat, ex

32

gel of SDS g/mL from

tored

mplex

ched

o the

ented

ng to

xcept

33

that each buffer was supplemented with 2-5 mM DTT to keep selenomethionine

in a reducing environment. Mass spectrometry of SeMet TsUCH37cat by protein

MALDI confirmed that all four methionines in the protein were enriched with

SeMet, an M+1 molecular weight of 26238.6 Da and M+2 of 13117.1 Da, with a

calculated molecular weight of 26238 Da. SeMet TsUCH37cat was complexed

with UbVME and purified by MonoQ anion exchange chromatography. Pure

fractions were pooled, concentrated down, and flash frozen and stored at -80°C.

Yields for the SeMet protein were reduced; therefore the SeMet complex was

lower concentration than the original complex.

2.2.5 Crystallization and Structure Solution

Native TsUCH37cat-UbVME was screened at 3 mg/mL in ~700

crystallographic conditions by sitting drop vapor diffusion. A hit was identified in

the Hampton Research Ammonium Sulfate grid screen, composed of 3 M

ammonium sulfate, 0.1 M bicine pH 9 at room temperature after 2 days by

hanging drop vapor diffusion. However, rather than single, 3D crystals, the initial

hit appeared to be stacks of 2D plate crystals. In anticipation of poor data due to

multi-latticed crystals, the initial hit was optimized by additive screening. Single

3D crystals appeared with the addition of 2 mM glutathione (mixture of oxidized

and reduced). Crystallization attempts with the SeMet complex in the same

mother liquor composition as the native hit did not yield any crystals, therefore

microseeding with native crystals was done to induce SeMet complex

c

m

fr

o

g

w

c

Fin2cc

rystallizatio

mother liquo

ragments. A

f 1 µL SeM

rew over 2

were looped

ryoprotecta

Figure 2.3: Cn 3 M ammD plates. rystals of omplex cry

on. Microse

or, vortexin

A volume o

Met complex

days at ro

d and flas

ant 5.

Crystals of monium sulf

B) Optimizthe SeMet

ystal looped

eeds were o

ng for 5 m

of 0.2 µL of

x, 1 µL mot

oom temper

sh frozen

the TsUCHfate, 0.1 M zed crystat TsUCH37d and moun

obtained by

minutes, an

f microcryst

ther liquor,

rature. Both

using 2.5

H37cat-UbVMbicine pH

ls with 2 7cat-UbVMEted at the b

y crushing

nd spinning

tals were a

0.4 µL glu

h native an

M sodium

ME comple9.0. CrystamM glutat

E complex. beamline at

native cry

g down an

added to a d

utathione ad

nd SeMet co

m malonate

ex. A) Unopals are genthione add D) An opt APS GM/C

ystals in ~4

ny large cr

drop comp

dditive. Cry

omplex cry

e pH 7.0 a

ptimized crynerally layeditive. C) Sptimized SeCA CAT.

34

40 µL

rystal

osed

ystals

ystals

as a

ystals ers of Small eMet

L

6.

c

d

w

a

w

Fca

Data

aboratory o

. Native cry

ollected at

ispersion (

was collecte

dditive. Th

which was u

Figure 2.4: rystals anddditives. Co

was colle

on a Mar30

ystal data w

t the sele

SAD) phas

ed on som

e diffractio

unable to be

Diffraction d B) the Seollected at

ected at

00 CCD det

was collect

nium peak

sing with an

me of the

n pattern in

e indexed (

patterns. eMet TsUCthe ID-B be

the 23-ID-

tector (Mar

ed up to 1

k, 0.979 Å

n f’ of -8.74

unoptimize

ndicated a

Fig. 2.4).

From A) thCH37cat-UbVeamline of G

-B beamlin

r USA) and

.9 Å at 1.0

Å for singl

4 and an f

ed crystals

multi-lattic

he unoptimVME crystaGM/CA CA

ne at Arg

d processed

033 Å and

le-waveleng

f’’ of 7.12.

s, lacking

ce or multip

mized TsUCals after op

AT at Argon

gonne Nat

d with HKL2

1.7 Å data

gth anoma

Diffraction

the glutath

ple-crystal d

CH37cat-UbVptimization ne’s APS.

35

tional

2000

was

alous

data

hione

data,

VME with

36

The initial model was obtained from the Phenix AutoSol wizard using

selenium SAD phases with an input of 8 Se sites (from Matthews coefficient,

determined to be a dimer in the asymmetric unit). The initial model was given a

FOM (figure of merit) of 0.338, and initial Rwork of 0.3695 and Rfree of 0.3884. Its

sequence was built in using the Phenix AutoBuild wizard with additional manual

model building in Coot 7,8. Two copies of the complex were found in the

asymmetric unit, having a space group of C2. Refinement of the structure was

done in Phenix using some TLS refinement (entire asymmetric unit considered to

be one TLS group) and optimized weighting for stereochemical restraints 7.

Overall completeness of the data was poor, at 88.5%, but this can be credited to

poor completeness in the highest resolution shells (42%), which did not prevent

structure solution or refinement. The final model had an R factor of 17.4% and an

Rfree of 21% with <0.2% of residues in the disallowed region of the

Ramachandran plot and scoring a 98% in assessment by Molprobity 9. The

structure was deposited in the Protein Databank (PDB) under the entry 4I6N 10.

37

2.3 Results

2.3.1 Structure of TsUCH37cat-UbVME

TsUCH37cat-UbVME crystallized in the C2 space group with two copies of

the complex in the asymmetric unit. The final model had an R factor of 17.4%

and an Rfree of 21% (Table 2.1).

The first structural element to come to our attention was the presence of

electron density for a disulfide bond between Cys71 of each TsUCH37cat

monomer, leading to disulfide-mediated dimerization in the asymmetric unit.

Human UCH37 was previously thought to oligomerize in solution through its C-

terminal domain, therefore this result was unexpected. In order to determine if

this dimerization has biological relevance, the TsUCH37cat-UbVME complex and

TsUCH37cat alone were both subjected to analytical ultracentrifugation, the

results of which are discussed in Section 3.3. We concluded that this disulfide

bond formation was a crystallographic artifact rather than a biologically significant

event. It likely arose as a result of the introduction of glutathione as an additive

and may have assisted crystal packing into a better form than the initial 2D plate

crystals. The two copies of the complex have an RMSD of 0.39, indicating very

few differences between them. Analysis of the complex, for the purpose of this

document, will focus on Chains A and B in the PDB file rather than the copy,

Chains C and D.

T

aNbRincRdRreeOa

Table 2.1: Ta

Numbers in Rmerge =Σ|Ih - ntensity. Rwork = Σ ||Fo

Rfree is the saeflections thaOrdered resind Gln152 to

able of crysta

parentheses<Ih>|/Σ Ih , w

obs|-k|Fcal||/ Σame as Robs at was not inidues: Pro-3o Gln225 in

allographic s

s refer to datwhere Ih is th

Σ|Fobs| for a selecte

ncluded in pr3 to Gly141 a

Chain A.

statistics.

ta in the highhe observed

ed subset (5rior refinemeand Lys153 t

hest resolutiintensity an

5% and 9%, ent calculatioto Asp224 in

ion shell. nd <Ih> is the

respectivelyons. n Chain C; P

e average

y) of the

Pro-3 to Gly1

38

141

fo

(P

P

c

te

h

b

FcoCsu

The s

our other U

PDB ID

Plasmodium

atalytic clef

erminal tail

ydrogen b

etween the

igure 2.5: Aompared to

C-terminal tauperposed i

structure o

UCH family

1CMX), U

m falciparum

ft (Fig. 2.5)

of ubiquitin

bond netwo

e backbone

Active site human UCHil of UbVMEn purple. All

2.

f TsUCH37

ubiquitin-bo

UCHL3-UbV

m UCHL3-

) bind ubiqu

n, residues

ork within

of TsUCH3H37 (purple)E (orange) a contacts ar

3.2 Active

7cat-UbVME

ound struct

VME (1XD

-UbVME (2

uitin in a hi

70-75, is t

the cataly

37. A) Activ) upon ubiquand TsUCH3re 2.7 – 3.2 Å

Site Bindin

E shows ve

tures in the

D3), UCHL

2WDT). Its

ighly conse

thoroughly

ytic cleft,

ve site archuitin binding.37cat (teal), wÅ.

g

ery similar

e PDB: Yuh

L1-UbVME

s active sit

erved mann

stabilized b

notably th

hitecture of . B) Interactiwith human

features to

h1-Ub alde

(3KW5),

te and gen

ner 11-13. Th

by an exten

rough con

TsUCH37 ions betweeUCH37 res

39

o the

hyde

and

neral

he C-

nsive

tacts

(teal) en the idues

40

of the tail as well as Arg72 and Arg74’s side chains. The active site tetrad,

composed of Cys85, Asp176, His161, and the oxyanion-stabilizing residue

Gln79, is arranged in a canonical orientation for the UCH family, which is seen in

papain-like cysteine proteases as well. The catalytic Cys85 of TsUCH37 has

flipped about 90° compared to Cys88 of the unbound human enzyme (PDB

3IHR) upon binding to GlyVME, a mimic of the acyl-enzyme intermediate during

catalysis. This phenomenon is also seen in the ubiquitin-bound and unbound

structures of UCHL1, and is believed to be a conformational switch from an

unproductive form of the enzyme that may exist as a protective mechanism 12.

Some deubiquitinases operate within an oxidative environment, and this

conformational change may protect the enzyme against cysteine oxidation 14.

2.3.3 Distal Site Binding

Stabilization of ubiquitin’s C-terminal tail is the primary mode of ubiquitin

binding by UCH family enzymes, with the second-most important being its distal

site interactions with ubiquitin’s Leu8, Thr7, and Thr9 as well as ubiquitin’s Ile44

patch. Ubiquitin-interacting residues from the Ts to human UCH37 are not highly

conserved compared to its catalytic cleft residues. The distal site of TsUCH37

utilizes different hydrophobic groups than the human enzyme for ubiquitin

binding, such as replacement of Ser37 with Leu36 and substitution of the large

Trp36 with Val34 and Val35 (Fig 2.6). Additionally, comparing the distal pockets

of the human unbound enzyme versus bound TsUCH37, there appears to be a

c

L

FTusu

w

c

st

fa

a

st

b

onformation

8-T9 hairpi

igure 2.6: DTsUCH37 (te

biquitin’s Ileuperposed i

The c

which varies

left of the U

tabilizing it

amily memb

ll make co

tructures o

elieved tha

nal change

n turn.

Distal site oeal) comparee44 patch (on purple.

crossover l

s in length

UCH doma

ts C-termin

bers crysta

ontacts with

f the enzym

at the UCH

e that enclos

of UCH37. Aed to humanorange) and

2

oop is a st

among the

ain and bind

nal tail furth

llized in com

h ubiquitin

mes, are no

crossover

ses the dis

A) Ubiquitinn UCH37 (pud TsUCH37

2.3.4 Cross

tructural fe

e family me

ds to ubiqu

her for cata

mplex with

via their

ot resolved

loop is lock

stal residues

n (orange) burple) unbou

7cat (teal), w

sover Loop

ature comm

embers 13.

uitin when i

alysis 11,12.

an ubiquiti

crossover

d due to hig

ked into a s

s tighter aro

binding to tund. B) Inter

with human

mon to all

It lies acro

t is bound

The singl

n variant fo

loops, whi

gh flexibility

specific con

ound ubiqu

the distal siractions betwUCH37 res

UCH enzy

oss the cata

to the enzy

e domain

ound in the

ich in unbo

y (Fig. 2.7).

nformation u

41

uitin’s

ite of ween idues

mes,

alytic

yme,

UCH

PDB

ound

. It is

upon

u

fo

k

a

u

FTinPa

c

fl

re

biquitin bin

or cleavage

nown that

ccommoda

biquitinated

igure 2.7: CTsUCH37 shn red (PDB IPfUCHL3-Ub

symmetric u

Intere

rossover l

exibility, de

esidues 14

nding, which

e of small

the large

ating larger

d histone H

Crossover lown in teal ID 3KW5), U

bVME in yellunit of the Ts

estingly, th

oop, desp

espite crys

2 – 151 in

h likely con

moieties fr

r family m

substrates

H2A for BAP

loop of TsUcompared t

UCHL3-UbVow (2WDT)

sUCH37cat-U

he structure

ite being

stal packin

either cop

ntributes to

rom the C-

members, U

, polyubiqu

P1 15-17.

UCH37cat-Ubto other UCHME in green. B) Crosso

UbVME struc

e of TsUC

bound to

g, that no

py in the as

the selecti

-terminus o

UCH37 an

uitin chains

bVME strucH-Ub boundn (1XD3), Y

over loop (picture.

CH37cat-UbV

ubiquitin.

o electron

symmetric

vity of this

of ubiquitin.

d BAP1, a

in the case

cture. A) Crd structures,UH1-UbAl innk) of each

VME has

It has re

density ca

unit (Fig. 2

family of D

. However,

are capabl

e of UCH37

rossover loo, UCHL1-Ubn purple (1C monomer i

an unreso

etained eno

an be seen

2.7). It does

42

DUBs

it is

le of

7 and

op of bVME CMX), n the

olved

ough

n for

s not

43

make any contacts with ubiquitin in the situation of this complex: UCH37 –

monoubiquitin. This leads one to believe that an additional protein binding event

would be required to stabilize the crossover loop, that it may require a different

minimal substrate (diubiquitin, triubiquitin, etc) or that the crossover loop binds to

another protein regulator. We speculate that this other protein may be Rpn13,

and that this binding event may be the source of activation of UCH37’s catalytic

activity.

2.4 Discussion

Here we have presented the structure of TsUCH37cat bound to ubiquitin

vinyl methyl ester, which has provided some valuable insights into the

mechanism of this UCH family deubiquitinase. The enzyme relies on a complex

network of interactions around ubiquitin’s C-terminal tail for substrate

stabilization, which is highly conserved between TsUCH37 and the other yeast

and human homologs of UCH enzymes. Additionally, TsUCH37 utilizes distal site

binding to recognize ubiquitin’s Ile44 patch and Leu8-Thr9 motif. However, the

residues responsible for distal site binding are not as conserved as those in the

catalytic cleft, compared to human UCH37. This lack of conservation may impact

the affinity of ubiquitin binding, which will be explored in Part 2 through

comparison of the enzyme’s catalytic activity compared to the human enzyme.

This region of the enzyme may confer selectivity among UCH family enzymes,

distinguishing each from one another, as their catalytic clefts are nearly identical.

44

The most significant structural difference between TsUCH37cat and the

other UCH family structures is that its crossover loop has not gained sufficient

stabilization upon ubiquitin binding to be visualized in its X-ray crystal structure.

The crossover loop is a structural element of UCH enzymes which is responsible

for substrate filtering and binding, which appears to not play a role in ubiquitin

binding for TsUCH37, and likely human UCH37 as well. We speculate that the

crossover loop would be resolved in the structure if it was satisfying all its

necessary contacts, which probably requires binding to an additional partner. We

further hypothesize that this binding partner may be Rpn13, the proteasomal

subunit which anchors UCH37 to the 26S proteasome. It seems therefore that

the crossover loop in UCH37 may be a key element in the regulation of UCH37’s

catalytic activity through protein-protein contacts. Further examination of the

crossover loop in binding studies should confirm our hypothesis.

45

2.5 References

1 Borodovsky, A. et al. Chemistry-based functional proteomics reveals novel members of the deubiquitinating enzyme family. Chemistry & biology 9, 1149-1159 (2002).

2 Borodovsky, A. et al. A novel active site-directed probe specific for

deubiquitylating enzymes reveals proteasome association of USP14. The EMBO journal 20, 5187-5196 (2001).

3 Liu, S. & Hanzlik, R. P. Structure-activity relationships for inhibition of

papain by peptide Michael acceptors. Journal of medicinal chemistry 35, 1067-1075 (1992).

4 White, R. R. et al. Characterisation of the Trichinella spiralis

deubiquitinating enzyme, TsUCH37, an evolutionarily conserved proteasome interaction partner. PLoS Negl Trop Dis 5, e1340 (2011).

5 Holyoak, T. et al. Malonate: a versatile cryoprotectant and stabilizing

solution for salt-grown macromolecular crystals. Acta crystallographica. Section D, Biological crystallography 59, 2356-2358 (2003).

6 Otwinowski, Z. a. M., W. Processing of X-ray Diffraction Data Collected in

Oscillation Mode. Vol. 276 (Academic Press (New York), 1997). 7 Adams, P. D. et al. PHENIX: a comprehensive Python-based system for

macromolecular structure solution. Acta Crystallographica Section D: Biological Crystallography 66, 213-221 (2010).

8 Emsley, P. & Cowtan, K. Coot: model-building tools for molecular

graphics. Acta crystallographica. Section D, Biological crystallography 60, 2126-2132 (2004).

9 Chen, V. B. et al. MolProbity: all-atom structure validation for

macromolecular crystallography. Acta crystallographica. Section D, Biological crystallography 66, 12-21, doi:10.1107/S0907444909042073 (2010).

10 Morrow, M. E. et al. Stabilization of an Unusual Salt Bridge in Ubiquitin by

the Extra C-Terminal Domain of the Proteasome-Associated Deubiquitinase UCH37 as a Mechanism of Its Exo Specificity. Biochemistry, doi:10.1021/bi4003106 (2013).

46

11 Misaghi, S. et al. Structure of the ubiquitin hydrolase UCH-L3 complexed with a suicide substrate. The Journal of biological chemistry 280, 1512-1520 (2005).



13 Johnston, S. C., Riddle, S. M., Cohen, R. E. & Hill, C. P. Structural basis

for the specificity of ubiquitin C-terminal hydrolases. The EMBO journal 18, 3877-3887 (1999).

14 Kulathu, Y. et al. Regulation of A20 and other OTU deubiquitinases by

reversible oxidation. Nature communications 4, 1569, doi:10.1038/ncomms2567 (2013).

15 Lam, Y. A., Xu, W., DeMartino, G. N. & Cohen, R. E. Editing of ubiquitin

conjugates by an isopeptidase in the 26S proteasome. Nature 385, 737-740 (1997).




Polycomb repressive complex PR-DUB. Nature 465, 243-247 (2010).

47

CHAPTER 3: KINETIC AND BIOPHYSICAL CHARACTERIZATION OF TSUCH37

3.1 Introduction

The biophysical characteristics of UCH37 keenly regulate its kinetic

activity as well as biological association with its proteasomal binding partner,

Rpn13. Its ULD, or UCHL5-like domain, has been shown to alter its activity and

ability to bind to the 26S proteasome. Within this ULD lies the KEKE motif, a

region spanning the final 20-30 amino acids of the protein, which is responsible

for its binding to the proteasomal subunit Rpn13. Rpn13 harbors a

complementary C-terminal KEKE motif, which binds to UCH37 (Fig 3.1).

Interestingly, the ULD of UCH37 is also thought to play two additional roles within

the enzyme: (1) regulation of its oligomeric state and (2) autoinhibition of the

enzyme’s catalysis1-5. The oligomerization of UCH37 was explained by

tetramerization of the human enzyme in its X-ray crystal structure (PDB ID 3IHR)

as well as in-solution higher order oligomers observed during size-exclusion

chromatography (Fig 3.2) 5. Autoinhibition has been seen by multiple groups in

the context of purified protein, by deletion of the ULD and comparison of its

activity versus that of the full-length enzyme against ubiquitin 7-amino-4-

methylcoumarin, a fluorogenic monoubiquitin substrate standard in the DUB field,

but limited in that it does not address the processing of polyubiquitin 1,2,5-9.

U

o

a

d

w

c

U

a

o

p

a

F

UCH37 has

wn, yet it

ppears rath

ismantling

when Rpn13

leavage, w

ULD autoinh

ctivated up

n the prote

resence of

t the time o

Figure 3.1: D

never bee

has good a

her advanc

its proteas

3 has been

which has le

hibition thro

pon binding

easome1. E

f UbAMC is

of these exp

Domain dia

en shown to

activity aga

ced, when t

somal subs

added to U

ed others t

ough KEKE

to the entir

xamination

s a limited a

periments.

gram of UC

o cleave po

ainst UbAM

he enzyme

strate, poly

UCH37, it i

to speculat

E motif bind

re 19S regu

n of the enz

approach,

CH37 and R

olyubiquitin

MC 3,4. The

e has not b

yubiquitin.

s slightly m

te that Rpn

ing1. Ultima

ulatory part

zyme witho

but was th

Rpn13.

chains ap

e concept o

een definiti

To add to

more active

n13 relieves

ately, UCH

ticle, where

ut the 19S

e system w

ppreciably o

of autoinhib

ively capab

the confu

in polyubiq

s the effec

37 is maxim

e Rpn13 res

and only in

within our r

48

on its

bition

ble of

sion,

quitin

cts of

mally

sides

n the

each

U

c

th

b

Fa

h

a

T

a

Our g

ULD: does i

onfer its au

his ultimate

e responsib

Figure 3.2: symmetric

Herei

uman UCH

utoinhibitio

Through kin

utoinhibited

goals for th

t mediate o

utoinhibitory

ely relate to

ble for all th

Structure unit. The c

in, the kine

H37 are pr

n, activatio

netic analy

d in its full-

is compone

oligomeriza

y effect, or

o binding to

hree activiti

of human atalytic Cys

etic and b

robed, in o

on, and pr

ysis of TsU

length form

ent of the p

ation in solu

r is this a s

the protea

es.

UCH37 (Ps residue is

iophysical

order to be

roteasomal

UCH37, w

m, but desp

project were

ution? If so

separate ev

asome, as t

PDB ID 3IHs shown in y

properties

etter unders

l associatio

we find tha

ite investig

e to dissec

, does this

vent? And h

this one reg

HR) as a tyellow sphe

of TsUCH

stand the

on still elu

at the enz

ations into

ct the role o

oligomeriz

how does a

gion appea

tetramer ineres.

H37 as we

mechanism

uding the

yme is ind

the biophy

49

of the

ation

all of

ars to

n the

ell as

ms of

field.

deed

ysical

50

characteristics of the event, we are still unclear about how the enzyme transitions

between its basal and activated states.


3.2.1 Cloning, Expression, and Protein Purification

TsUCH37cat and TsUCH37cat-UbVME were purified as described

previously in Part 2.2.2. TsUCH37FL with an N-terminal 6xHis tag in pET28a+

was expressed in E. coli by Dr. Myung-Il Kim as described in Morrow et. al, 2013

10. For isothermal titration calorimetry, human Rpn13 was expressed and purified

by Dr. Judith Ronau from E. coli on glutathione beads and subsequently by size

exclusion chromatography. Human UCH37 proteins used for isothermal titration

calorimetry were wild-type and an E284A mutant (discussed further in Section

4.3.2), both expressed from a pET28a+ plasmid in E. coli, purified on Ni-NTA

beads using a 50 – 500 mM imidazole gradient (purification buffer of 20 mM

sodium phosphate pH 7.0, 300 mM NaCl, ). Due to an engineered HRV 3C

protease site, Prescission Protease (GE Biosciences) was added to remove the

6xHis tag and linker, which was subsequently removed by incubation with

glutathione beads. Second step purification was done on both wild-type and

UCH37 E284A on a Sephadex S200 size exclusion column (GE Biosciences)

and pure fractions were pooled, concentrated down, and flash frozen as aliquots.

51

3.2.2 Analytical Ultracentrifugation

TsUCH37cat and TsUCH37cat-UbVME were both dialyzed extensively

against 50 mM Tris pH 7.4, 200 mM NaCl, and 1 mM DTT. Samples were run at

concentrations a range of concentrations: 8, 16, and 32 µM for TsUCH37cat and

10, 18, and 31 µM for TsUCH37cat-UbVME to determine oligomeric states at high

concentrations. Samples were run on a Beckman-Coulter XLA analytical

ultracentrifuge at 50,000 rpm and monitored at 280 nm for 150 scans.

Sedimentation coefficient distributions were analyzed by SEDFIT (v. 13.0b) 11.

For analysis of human UCH37 and Rpn13, proteins were both extensively

dialyzed against 50 mM Tris pH 7.5, 50 mM NaCl, 1 mM TCEP. For analysis of

individual oligomerization states, UCH37 was run at 8, 16, and 32 µM and Rpn13

was run at 13.5, 27, and 54 µM. Analysis of the UCH37-Rpn13 complex was run

at concentrations of 4 and 8 µM, 4 and 16 µM, and 4 and 32 µM of UCH37 and

Rpn13, respectively. Samples were run and analyzed by the same methods as

TsUCH37cat and TsUCH37cat-UbVME, above.

3.2.3 Ubiquitin-AMC Hydrolysis

Cleavage of 7-amino-4-methylcoumarin from the C-terminus of

monoubiquitin, or UbAMC cleavage, was monitored in a reaction buffer

containing 50 mM Tris pH 7.6, 0.5 mM EDTA, 0.1% bovine serum albumin, and 5

mM DTT. TsUCH37cat and TsUCH37FL were diluted in reaction buffer to 7 nM

final reaction concentration and preincubated at 30°C for 5 minutes prior to the

reaction. Reactions were initiated by addition of UbAMC (Boston Biochem) and

52

were measured on a Tecan fluorescence plate reader (Männedorf, Switzerland)

with 380 nm excitation wavelength and 465 nm emission wavelength at 30°C for

1 hr. Progress curves and Michaelis-Menten kinetics were plotted and fit in

SigmaPlot (Systat Software).

3.2.4 Isothermal Titration Calorimetry

For isothermal titration calorimetry, wild-type UCH37, UCH37 E284A, and

Rpn13 were dialyzed extensively together against 50 mM Tris pH 7.5, 50 mM

NaCl, 1 mM TCEP. ITC experiments were done using a MicroCal ITC200 (GE

Biosciences). For determination of the Kd of UCH37 wild-type and Rpn13

binding, two experiments were averaged together: 20 µM UCH37 in the cell with

228 µM Rpn13 injected, and 10 µM UCH37 in the cell with 100 µM Rpn13

injected. For UCH37 E284A, 10 µM E284A was in the cell and 100 µM Rpn13

was injected. The data was analyzed and fit to a single binding site model in

SEDPHAT 11.

c

C

d

c

b

Fc

fr

A

3.3.1 A

As d

rystallized

Cys71 of ea

imerization

rystallizatio

y itself was

Figure 3.3: Aompared to

rom the sam

AUC was ru

Analysis of

discussed

as a dimer

ach TsUCH

n was a b

on, the oligo

s determine

Analytical uo TsUCH37

me batch o

un on the co

Crystallogr

in Sectio

r, with dime

H37 protein

biologically

omeric state

ed by analyt

ultracentrifu7cat alone (A

f the TsUC

omplex as w

3.3 Resul

raphic Dime

on 2.3.1,

erization me

n. In order

y relevant

e of the com

tical ultrace

ugation of tA). C) Table

CH37cat-UbV

well as the

lts

erization of

the TsUC

ediated by

to determi

process o

mplex as w

entrifugation

the TsUCHe of sedime

VME compl

catalytic d

TsUCH37c

CH37cat-Ub

a disulfide

ine if TsUC

or merely

well as the c

n (AUC). Ta

H37cat-UbVMentation coe

lex used fo

omain of T

cat-UbVME

bVME com

e bond betw

CH37cat-UbV

an artifac

catalytic do

aking aliquo

ME complexefficients.

r crystalliza

TsUCH37 a

53

mplex

ween

VME

ct of

main

ots

x (B)

ation,

lone.

54

At concentrations higher than that in cells (8 – 32 µM), neither TsUCH37cat nor

the TsUCH37cat-UbVME complex were found to exist in solution as dimers. Both

are monomeric, with sedimentation coefficients (S20,w) of 3.3 for the complex and

2.8 for the catalytic domain (Fig. 3.3). Therefore, the dimerization event observed

in the crystal structure is an artifact of crystal packing, mediated by disulfide bond

formation resulting from oxidative conditions prevailing in the crystallization buffer

(glutathione additive).

3.3.2 Kinetic Characterization of TsUCH37cat and TsUCH37FL

In order to characterize the catalytic activity of TsUCH37, its activity

against a standard DUB substrate, a fluorogenic monoubiquitin derivative called

UbAMC, was assessed. The original goal of studying TsUCH37 previously was

for drug targeting12, therefore, it was of interest to examine its catalytic

mechanism compared to that of human UCH37 and the other UCH family DUBs.

Compared to the catalytic domain of human UCH37, TsUCH37cat has about a 20-

fold lower KM, indicating an improvement in substrate binding, however, the kcat

was 100-fold lower, yielding an overall 5-fold decrease in efficiency of the

enzyme (Fig 3.4) 9. It would appear that TsUCH37’s catalytic domain binds

substrate tighter, but that may also impair its ability to dissociate product for

another round of catalysis. Not surprisingly, TsUCH37cat’s KM is about 14-fold

higher than UCHL3 and 23-fold higher than that of UCHL1, both of which bind

m

s

FTpTeg

u

it

5

c

monoubiquit

ubstrates o

Figure 3.4: TsUCH37cat

arameters TsUCH37cat

t. al, 2012 enerated b

biquitin 9. S

s catalysis

0-fold wors

leavage, bu

tin well an

off the C-ter

Michaelis cleavage generatedcompared 9. Curve fity Sigma Pl

Since UCH

of a monou

se than tha

ut was 12-

nd are bel

rminus of

Menten kiof up to 1

d in Sigmato human

tting error iot.

37 is thoug

ubiquitinate

at of UCHL

fold better

lieved to o

inetics for 12 µM UbAa Plot. B) UCH37cat, Us listed for

ght to cleav

ed substrate

L3, one of

than that o

only be ca

TsUCH37c

AMC. CurvMichaelis

UCHL3, anr experimen

ve polyubiq

e is poor. T

the most e

of UCHL1.

apable of

cat UbAMCve fit to Mis Menten nd UCHL1 fntal values

quitin, it is r

The kcat for

efficient DU

The kcat/K

cleaving s

C hydrolysischaelis Meparametersfrom Boudrof TsUCH3

reasonable

TsUCH37c

UBs for UbA

KM for TsUC

55

small

s. A) enten s for eaux 37cat,

e that

cat is

AMC

CH37

a

s

d

d

p

U

p

FS–cc

nd UCHL1

imilar. As n

ifferences

istal bindin

lausible tha

UCHL1 and

rotein-prote

Figure 3.5:Sedimentati– Rpn13 coefficients omplex det

are only a

nearly all re

in KM and

g site resid

at having a

UCHL3, m

ein contacts

: Analyticaon of A) UCcomplex cand apparetermined by

about 2-fold

esidues in t

ubiquitin b

dues, which

relatively lo

makes UCH

s.

al ultracenCH37 aloneompared tent moleculy AUC.

d different,

he UCH fam

inding are

h may confe

ower KM co

H37 better

ntrifugation e, B) Rpn13to the indar weights

indicating

mily active

generally d

er some se

ompared to

suited for

of huma3 (ADRM1)dividual prof UCH37,

that their

site are hig

due to diffe

lectivity for

its single-d

regulatory

an UCH37) alone, androfiles. D) , Rpn13 (AD

efficiencies

ghly conser

erences in

this family

domain cou

control thro

7 and Rpd C) the UC

SedimentDRM1), and

56

s are

rved,

their

. It is

usins,

ough

pn13. CH37 ation d the

57

3.3.3 Analysis of UCH37 Oligomeric State

In order to probe whether the previously proposed model of UCH37

tetramerization or higher order oligomerization was possible, we examined its

oligomerization by analytical ultracentrifugation (Fig 3.5) 5. Additionally, the

stoichiometry of the binding of UCH37 to Rpn13, its proteasomal binding partner,

was determined. Analytical ultracentrifugation of UCH37 alone at 8, 16, and 32

µM yielded data indicating that the enzyme primarily exists as a monomer at

lower concentrations, but is capable of a concentration-dependent rapid

monomer-dimer equilibrium, which is seen most prominently in the 32 µM

concentration sample. Higher order oligomers were not detected at that those

concentrations, which does not rule out the possibility, but indicates that at

cellular concentrations, the enzyme is likely monomeric.

As for the UCH37-Rpn13 complex, first the solution state of Rpn13 was

determined alone at 13.5, 27, and 54 µM. Rpn13 primarily exists as a monomer

with a small population of higher order oligomers or aggregates, however this

proportion is quite small. The UCH37-Rpn13 complex was run at three different

concentrations of Rpn13 (8, 16, and 32 µM), but with UCH37 fixed at 4 µM. The

complex exists in a 1:1 stoichiometry, which is not a surprise given that Rpn13

only has one recognition motif for UCH37 to bind. These results do not appear to

support the theory that Rpn13 may relieve UCH37 of its autoinhibition through

binding its ULD to change the oligomeric state of the enzyme. Both proteins are

predominantly monomeric in solution and form a 1:1 complex at the proteasome.

k

th

cy

tw

re

p

h

w

3

FoR

Ultim

nowledge o

he binding

ysteine DU

wo ubiquitin

elationship

urified full-

ave looked

were studied

.6). The dis

Figure 3.6: f binding c

Rpn13. C) T

3.3.

ately, the

of UCH37

affinities o

UB that ass

n receptors

with Rpn1

-length hum

d at the ye

d for their b

ssociation c

Isothermal urves of A)

Table of the

.4 Analysis

most imp

is its bindi

of the pro

ociates wit

s, Rpn10 an

3 has bee

man UCH3

ast homolo

binding to R

constant wa

titration ca) wild-type rmodynam

of UCH37

portant biop

ng to Rpn1

teasome s

th the prote

nd Rpn13,

n less exp

7 and full-

ogs. For ou

Rpn13: wild

as determin

alorimetry oUCH37 anic paramete

Binding to

physical p

13. Previou

subunit Rp

easome, as

for ubiquiti

plored. For

-length hum

ur experime

-type and a

ned by aver

of UCH37 and Rpn13 aers.

Rpn13

arameter m

us studies

n1 for US

s well as th

in13-15. How

this experi

man Rpn13

ents, two U

a ULD muta

raging two e

and Rpn13and B) UCH

missing in

have exam

P14, the o

he affinity o

wever, UCH

iment, we

3; most stu

UCH37 pro

ant, E284A

experiment

binding. HH37 E284A

58

our

mined

other

of the

H37’s

used

udies

teins

A (Fig

ts for

Heats A and

59

wild-type, one with 20 µM UCH37 in the cell and 228 µM Rpn13 as the titrant,

and a second with 10 µM UCH37 in the cell and 100 µM Rpn13 as the titrant.

The average Kd was 22 ± 6 nM. UCH37 E284A was only run as a single

experiment, with 10 µM E284A in the cell and 100 µM Rpn13 injected, which

yielded a Kd of 18.5 ± 7 nM. Although UCH37 in cells is known to exist as a

population of free enzyme, not bound to the 26S proteasome and can associate

with the Ino80 chromatin remodeling complex, these dissociation constants

suggest very tight binding between this DUB and its proteasomal anchor, Rpn13

2. This interaction is known to be abolished upon deletion of UCH37’s KEKE

motif, and it is clear that even though the ULD mutation E284A impairs activation

of the enzyme (Section 4.3.2), the ULD region likely does not contribute

significantly to binding to Rpn13 1,6-8. Additionally, as the Kd of UCH37-Rpn13

binding is so low, it further disproves the possibility that UCH37 oligomerizes in

cells.

3.4 Discussion

Thorough characterization of UCH37’s kinetic and biophysical properties

is necessary to dissect its cellular association with the 26S proteasome and

potential autoinhibition. These studies have shed more light on the role of its ULD

in catalysis and binding, but more work is still needed to understand its

activation. We have confirmed that dimerization of the TsUCH37cat-UbVME

complex in its crystal structure is a crystallographic artifact of tight packing. Our

studies of the kinetic activities of TsUCH37cat and TsUCH37FL, and our analysis

o

p

c

c

U

(T

b

a

Tb

m

u

f the oligo

revious lite

ellular conc

omparing a

UCH37 bind

Table 3.1).

e a const

ssociating

Table 3.1: inding part

measured 1

ndoubtedly

omerization

erature, in th

centrations

activity to th

ds to Rpn1

This data s

itutive mem

subunit. Its

Dissociationers. Value

12 nM diss

y constituti

state of h

hat we do n

, but we sti

he catalytic

3 in a 1:1

seems to in

mber of th

s binding is

on constantes for * are

sociation c

ve membe

human UCH

not see dim

ll observe t

c domain al

ratio, and

ndicate tha

he 26S pro

within rang

ts of protefrom ref 13

constant fo

ers18. In c

H37, both

merization o

the autoinh

lone 5,9,16,17

with a 22

at UCH37, i

oteasome

ge of the pre

easome-asswhile those

or the Rpn

comparison

support an

of the full-le

hibition phe

7. We have

nM dissoc

n contrast

rather tha

eviously

sociated De under # ar

n2-Rpn13 i

n, the yea

nd conflict

ngth enzym

nomenon w

e confirmed

ciation con

to USP14,

n a transi

DUBs and re from ref

interaction,

ast homolo

60

with

me at

when

d that

stant

may

ently

their 18.

two

og of

61

USP14, Ubp6, has two possible binding sites on Rpn1, one a tighter 62 nM Kd

site, the other much weaker at nearly 2 µM13. Rpn1 is known to bind to shuttle

factors and other Ubl domain-containing proteins, therefore, USP14 is not always

bound to it. These numbers would suggest that UCH37 is more frequently found

in a proteasomal context than USP14 and may play a more significant biological

role.

However, these investigations still leave open the question of how UCH37

is activated at the proteasome, if it occurs merely through association with a

conformationally-accessible Rpn13, or if another binding partner is required. A

UCH37 mutant, E284A, which could not be activated by Rpn13 during ubiquitin-

AMC hydrolysis (Section 4.3.2) bound to Rpn13 with nearly the same Kd as the

wild-type enzyme. This mutation isolates Rpn13’s activation of UCH37 to an

event independent of simple binding. Further studies of this mutant in the

presence of di- or tri-ubiquitin, as well as in the presence of Rpn2, the

proteasomal subunit which binds Rpn13’s N-terminus, may provide the key to

UCH37’s mode of activation.

62

3.5 References

1 Yao, T. et al. Proteasome recruitment and activation of the Uch37 deubiquitinating enzyme by Adrm1. Nature cell biology 8, 994-1002 (2006).

2 Yao, T. et al. Distinct modes of regulation of the Uch37 deubiquitinating

enzyme in the proteasome and in the Ino80 chromatin-remodeling complex. Molecular cell 31, 909-917 (2008).





5 Burgie, S. E., Bingman, C. A., Soni, A. B. & Phillips, G. N., Jr. Structural

characterization of human Uch37. Proteins (2011). 6 Qiu, X. B. et al. hRpn13/ADRM1/GP110 is a novel proteasome subunit

that binds the deubiquitinating enzyme, UCH37. The EMBO journal 25, 5742-5753, doi:10.1038/sj.emboj.7601450 (2006).

7 Hamazaki, J. et al. A novel proteasome interacting protein recruits the

deubiquitinating enzyme UCH37 to 26S proteasomes. The EMBO journal 25, 4524-4536 (2006).



9 Boudreaux, D. A., Chaney, J., Maiti, T. K. & Das, C. Contribution of active

site glutamine to rate enhancement in ubiquitin C-terminal hydrolases. The FEBS journal 279, 1106-1118 (2012).



63

11 Schuck, P. Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and lamm equation modeling. Biophysical journal 78, 1606-1619 (2000).

12 White, R. R. et al. Characterisation of the Trichinella spiralis

deubiquitinating enzyme, TsUCH37, an evolutionarily conserved proteasome interaction partner. PLoS Negl Trop Dis 5, e1340 (2011).

13 Chen, X., Lee, B. H., Finley, D. & Walters, K. J. Structure of proteasome

ubiquitin receptor hRpn13 and its activation by the scaffolding protein hRpn2. Molecular cell 38, 404-415 (2010).

14 Riedinger, C. et al. Structure of Rpn10 and its interactions with

polyubiquitin chains and the proteasome subunit Rpn12. The Journal of biological chemistry 285, 33992-34003, doi:10.1074/jbc.M110.134510 (2010).

15 Husnjak, K. et al. Proteasome subunit Rpn13 is a novel ubiquitin receptor.

Nature 453, 481-488 (2008). 16 Zhou, Z. R., Zhang, Y. H., Liu, S., Song, A. X. & Hu, H. Y. Length of the

active-site crossover loop defines the substrate specificity of ubiquitin C-terminal hydrolases for ubiquitin chains. The Biochemical journal 441, 143-149 (2012).


Protein & cell 5, 616-630, doi:10.1007/s13238-014-0046-z (2014). 18 Rosenzweig, R., Bronner, V., Zhang, D., Fushman, D. & Glickman, M. H.

Rpn1 and Rpn2 coordinate ubiquitin processing factors at proteasome. The Journal of biological chemistry 287, 14659-14671, doi:10.1074/jbc.M111.316323 (2012).

64

CHAPTER 4: ANALYSIS OF THE TSUCH37∆C46-UBVME STRUCTURE AND THE ROLE OF THE ULD

4.1 Introduction

Although the TsUCH37cat-UbVME structure provided valuable insight into

the mechanism of UCH37 and its ability to recognize and bind monoubiquitin, we

are still lacking information about the role of the ULD in catalysis, binding, and

activation. However, another group member, Dr. Myung-Il Kim, was able to

crystallize and solve the structure of a longer construct of TsUCH37 in complex

with UbVME for us to glean information regarding the ULD, hereafter referred to

as TsUCH37∆C46-UbVME (Fig 4.1). Due to cleavage of the protein during

purification, only a portion of the ULD was shown in the structure, but it provided

important clues regarding ubiquitin recognition by the enzyme. Contacts between

the catalytic domain of TsUCH37∆C46 and ubiquitin are identical to that of

TsUCH37cat, including a lack of ordered density for the crossover loop residues.

However, contacts between ubiquitin and the ULD are seen, making this an

additional ubiquitin binding interface. These contacts involve hydrogen bonds

and salt bridge interactions between TsUCH37’s Arg261 and Tyr262 with

ubiquitin’s Gln49 and Lys48, which forces a salt bridge interaction between

Fw

L

b

U

w

b

s

L

w

T

p

Figure 4.1: with the ULD

ys48 and G

ound struc

UCH37 beha

works its wa

elieve that

pecificity; U

ys48 will n

with the ULD

This mecha

olyubiquitin

Structure oD.

Glu51 of ub

ctures in th

aves exosp

ay inward

the TsUC

UCH37 is o

ot be enga

D. This rec

anism expl

n, however

of TsUCH3

biquitin, an

he PDB (T

pecifically, t

to the prox

CH37∆C46-Ub

only capab

aged in an

cognition of

lains the e

r, it does n

37∆C46-UbV

event here

Table 4.1).

that it only c

ximal mono

bVME acco

ble of cleav

isopeptide

Lys48 wou

enzyme’s

not provide

VME with in

etofore unse

It has bee

cleaves fro

omer of a

ounts for th

ving the dis

bond and

uld only pe

specificity

e structural

nset of Lys

een in any

en previou

om the dista

polyubiquit

he structur

stal monom

will be free

ermit exosp

in cleavin

clues as

s48 interac

other ubiq

usly shown

al monomer

tin chain 1

ral basis of

mer becaus

e for intera

ecific cleav

ng Lys48-li

to its abili

65

ctions

uitin-

that

r and

. We

f this

se its

cting

vage.

nked

ty to

c

u

Te

leave other

tilizes addit

Table 4.1: Lntries indic

r chain top

tional intera

Lys48-Glu5cate multiple

pologies. Th

actions by t

51 distancee copies of

his chapter

the ULD to

es in ubiquUb in the c

r explores t

confer spe

uitin-bound crystallogra

the possibi

cificity and

PDB strucaphic asymm

lity that UC

bind ubiqu

ctures. Mumetric unit.

66

CH37

itin.

ultiple

67


4.2.1 Molecular Dynamics Simulations

A starting model of human UCH37-UbVME was made by modeling of full-

length TsUCH37 using the human UCH37 structure (PDB ID 3IHR) as the search

model in the SwissModel homology modeling server 2. The final 46 residues

missing in the TsUCH37∆C46-UbVME were appended from the homology model

in Coot, which then underwent one round of refinement in Phenix 3,4. Professor

Markus Lill (Purdue University) then utilized this model for molecular dynamics

simulations, methods described in Morrow et. al, 2013 5. From the 2 ns

simulation, snapshots were examined for specific residues’ proximities to

ubiquitin. The majority of potential interactions could be seen at the 1.3 ns

snapshot.

4.2.2 Site-directed Mutagenesis and Protein Purification

Based on interactions seen in the MD simulations described above, a list

of Ub-interacting ULD residues were generated from the Ts enzyme and

corresponding residues in the human enzymes were mutated to Ala. Site-

directed mutagenesis was performed using the AccuPower PCR PreMix

(Bioneer) and mutations were confirmed by sequencing. Proteins were

expressed in Rosetta2 DE3 E. coli expression cells and purified by Ni NTA

beads. After cleavage of the 6xHis tag by Prescission Protease (GE

Biosciences), proteins were passed over GSH beads to remove the tag and

68

protease. Proteins were further purified by size-exclusion chromatography on a

Superdex 200 HiLoad column (GE Biosciences). Pure fractions were pooled,

concentrated down, and flash frozen as aliquots. Concentrations were

determined by UV/Vis. Human Rpn13FL was provided by Dr. Judith Ronau.

4.2.3 Ubiquitin-AMC Hydrolysis Assays

Cleavage of 7-amino-4-methylcoumarin from the C-terminus of

monoubiquitin, or UbAMC cleavage, was monitored in a reaction buffer


mM DTT. UCH37 wild-type and mutants were pre-incubated with Rpn13 on ice

for 1 hr, and then diluted in reaction buffer to final reaction concentrations of 0.5

nM and 15 nM, respectively, and warmed to 30°C for 5 minutes prior to the

reaction. Reactions were initiated by addition of UbAMC (Boston Biochem) and

were measured on a Tecan fluorescence plate reader (Männedorf, Switzerland)

with 380 nm excitation wavelength and 465 nm emission wavelength at 30°C for

1 hr. Progress curves were plotted in Kaleidagraph.

69

4.2.4 Synthesis of Asymmetric Triubiquitin Substrate and Assays

Site-directed mutagenesis was used to introduce a Gly76Val mutation into

a UbW77 construct in the pGEX-6P-1, which was subsequently confirmed by

sequencing. The double mutant UbG76V W77 was expressed in Rosetta2 DE3 E.

coli cells and purified on glutathione beads. After treatment with Prescission

Protease (GE Biosciences) to remove its N-terminal GST tag, the protein was run

back over glutathione beads to remove the tag and protease. UbG76V W77 was

further purified by size exclusion chromatography on a Sephadex 75 HiLoad

column (GE Biosciences) and pure fractions were pooled, concentrated down,

and flash frozen as aliquots.

Untagged ubiquitin in the pRSET vector was expressed in Rosetta2 DE3

E. coli cells, spun down, resuspended in purification buffer A (50 mM sodium

acetate pH 4.5, 2 mM DTT), lysed by French press, heated to 80°C for 10

minutes, then spun down at 30,000xg for 30 minutes. The supernatant was

brought to pH 4.5 by 1N HCl and then was purified by cation exchange

chromatography on SP sepharose beads (GE Biosciences) by gradient elution

with purification buffer B (same as A, but with 1 M NaCl). Pure fractions were

pooled and concentrated down, then further buffer exchanged and purified by

size exclusion chromatography on a Sephadex 75 HiLoad column (GE

Biosciences) into 50 mM Tris pH 7.6, 50 mM NaCl, 1 mM DTT. Pure fractions

were pooled, concentrated down, and flash frozen as aliquots.

Fsa

In

b

N

c

fu

C

T

–

w

Figure 4.2: howing natnd SDS PA

n order to g

iosynthetic

Native diubi

olumn (GE

urther react

Cdc34, wild

Triubiquitin w

– 600 mM N

were pooled

Mutant triutive distal mAGE gel (rig

generate the

ally made

quitin was

E Bioscienc

tions. Muta

d-type diub

was purified

NaCl salt gr

d, concentra

ubiquitin symonomer aght) of Mon

e substrate

using hum

purified by

ces). Pure

ant triubiqui

biquitin, Ub

d from exce

radient ove

ated down,

ynthesis. And proximaoS purifica

e shown in F

an Uba1, C

cation exc

fractions w

itin was ge

bG76V W77, a

ess UbG76V

r 50 colum

and flash f

A) Scheme al UbG76V W7

tion of asym

Figure 4.2,

Cdc34, wild

change chro

were pooled

enerated by

and ATP-M

V W77 on a M

mn volumes

frozen as a

of asymm77. B) Chrommetric triU

first native

d-type Ub,

omatograp

d and save

y incubating

Mg2+ for 12

MonoS colu

. Pure triub

liquots.

etric triubiqomatogram Ub.

e diubiquitin

and ATP-M

hy on a Mo

ed at -80°C

g human U

2 hrs at 3

mn using a

biquitin frac

70

quitin (left)

n was

Mg2+.

onoS

C for

Uba1,

37°C.

a 200

ctions

71

For polyubiquitin cleavage assays, wild-type K48-linked di-, tri-, and tetra-

ubiquitin were generated biosynthetically in the same manner as mutant tri

UbG76V W77. 1.5 µM wild-type or E284A UCH37 was incubated for 1 hour on ice

with 50 µM GST-Rpn13 (for triUbG76V W77 assays) or 5 µM untagged Rpn13 (for

wild-type polyubiquitin assays) in buffer containing 50 mM Tris pH 7.6, 0.5 mM

EDTA, 0.1% bovine serum albumin, and 5 mM DTT. Reactions were started with

the addition of 15 µM triUbG76V W77, tetraubiquitin, wild-type triubiquitin, or

diubiquitin and time points were quenched with SDS PAGE buffer. Reactions

were run on 15% SDS PAGE gels and stained with Coomassie.

4.3 Results

4.3.1 Analysis of Molecular Dynamics Simulations

After Professor Markus Lill (Purdue University) generated a 2 ns molecular

dynamics simulation for full-length TsUCH37-UbVME, each frame was analyzed

for potential interactions between the ULD of TsUCH37, the mobile element, and

ubiquitin, which was held stationary. The majority of interactions were seen at a

1.3 ns snapshot, showing potential interactions between many ubiquitin-facing

ULD residues and ubiquitin (Fig. 4.3). Although some interactions are only within

van der Waals or salt bridge distances, some possible hydrogen bonds were also

observed. The majority of the residues at this interface are also highly conserved

from yeast up to human UCH37, therefore, the most conserved residues were

Fcyhh

d

in

d

m

Figure 4.3: yan, grey, ighlight intighlighted i

deemed to

ntroduced in

etermine if

more likely,

Molecular and orang

eractions fn red.

be candida

n a handful

f these resid

its exospec

dynamics sge). TsUCHfrom a 1.3

ates for solu

of these h

dues are re

cificity.

simulationsH37∆C46-Ub

ns snapsh

ution studie

ighly-conse

esponsible

s of full-lenbVME is shhot, with h

es. Single p

erved residu

for UCH37

ngth TsUCHhown in olhighly cons

point mutati

ues (Table

7’s activatio

H37-UbVMlive. (i) andserved resi

ons to Ala w

4.2) in orde

on by Rpn1

72

E (in d (ii) dues

were

er to

3 or,

Tw

a

sy

p

b

w

w

a

c

m

Table 4.2. Uwere unable

To as

nd cleave

ystem exa

resence an

inding and

would abrog

were pre-in

ssociation

omplexes

monitored a

ULD mutatioe to be purif

4.3.2 Ubiq

ssess the e

ubiquitin c

mining the

nd absence

activation

gate activa

cubated w

of the two

were warm

nd plotted a

ons introdufied, due to

uitin-AMC

effects of U

hains withi

e efficiency

e of Rpn13

of the enz

ation. For th

with 30-fold

o proteins.

med to 30°

as progress

uced into hu poor solub

Hydrolysis

ULD mutati

n the prote

y of these

3. If ULD in

zyme, pres

he UbAMC

d excess R

Immediate

°C in reac

s curves, sh

uman UCHbility or insta

by UCH37

ons on the

easome, w

mutants in

nteractions

sumably mu

C assay, U

Rpn13 at

ly prior to

ction buffer

hown in Fig

37. Mutatioability.

ULD Muta

e ability of

we utilized a

n cleaving

are neces

utations wi

LD mutatio

4°C for 1

addition of

r. UbAMC

gure 4.4.

ons listed w

nts

UCH37 to

a more min

UbAMC in

ssary for Rp

thin this re

ons (Table

hour to a

f substrate

hydrolysis

73

with *

bind

nimal

n the

pn13

egion

4.2)

allow

, the

was

FU

U

w

c

m

m

to

e

R

d

Figure 4.4: UCH37 with

All U

UbAMC hyd

with previou

hange6,7. T

mutation ha

merely catal

o human U

ither in bin

Rpn13, or b

issect its r

UbAMC h and withou

LD mutant

drolysis ap

usly publish

The lack of

as somehow

lytic activat

CH37, whic

nding and/o

binding and

ole within t

hydrolysis but Rpn13.

ts assayed

proximately

ed activatio

activation

w abolishe

ion of the e

ch would in

or catalysis

d/or deactiv

the context

by UCH37

were activ

y 2-fold, ex

on of wild-t

of E284A

ed either R

enzyme. E2

ndicate tha

s of polyub

vation to pa

t of Rpn13

ULD mut

vated in th

xcept E284

type UCH37

was intrigu

Rpn13’s ab

284 is highly

t it must ha

biquitin, bin

artners with

3 activation

tants. Prog

he presenc

4A. These

7 in the ran

uing, thoug

ility to bind

y conserve

ave functio

nding and/o

hin the Ino8

and bindin

gress curve

e of Rpn13

results co

nge of a 10

h, because

d to UCH3

d from yea

nal signific

or activatio

80 complex

ng, we pur

74

es of

3 for

onflict

0-fold

e this

37 or

st up

ance

on by

x. To

sued

75

isothermal titration calorimetry (ITC) to determine the binding affinity of E284A

with Rpn13 compared to wild-type UCH37. The results, shown in Section 3.3.4,

indicate that binding to Rpn13 is not impaired by this mutation as its Kd is close to

that of wild-type UCH37. Therefore, this mutation may specifically inhibit the

mechanism of activation of UCH37 by Rpn13, specifically.

Interestingly, ULD mutations near E284 do not impair activation within the

context of the UbAMC assay, such as R280 and Y281. These two residues are

not as conserved as E284; R280 is substituted with Met, Lys, or Leu, and Y281 is

replaced by Trp in lower organisms. Perhaps these residues are more important

for ubiquitin recognition, rather than activation.

4.3.3 Triubiquitin Cleavage by ULD Mutants

In order to assess directional cleavage by UCH37, an asymmetric

polyubiquitin substrate was needed. To this end, UbW77, a construct utilized for

studies of the activity of UCHL1, was given an additional mutation, Gly76Val, by

site-directed mutagenesis in order to render its Trp77 non-cleavable by UCH37 8.

This double mutant monomer can be detected by an HPLC/MS assay due

to changes in its biochemical properties: 1) increased hydrophobicity and 2)

increased molecular weight. If ULD mutations abrogated exospecificity, equal

proportions of Ubwt and UbG76V W77 would be cleaved from either end of the

FWMcc

tr

w

p

la

c

g

ty

Figure 4.5: Wirth groupMS/MS spechromatograalculated a

riubiquitin c

would only

roperties, H

ab, Purdue

ould be sep

After

el-based c

ype UCH37

HPLC/MS p, Purdue Uctra (right) am of a mnd observe

chain, how

detect the

HPLC/MS

University

parated (Fig

generation

leavage as

7 and the

separationUniversity. for UbG76V

mixture of ed.

ever, if the

e distal or

experiment

, demonstr

g 4.5).

of the triUb

ssay was d

E284A mu

ns of ubiquPanels shoW77 (A) anthe two.

ey maintain

r middle m

ts were do

rating that t

bG76V W77 su

done to sho

utant in the

uitin monomow HPLC d Ubwt (B) D) Table

ned exospe

monomers,

one by coll

the mutant

ubstrate (M

ow activity

e presence

mers done chromatogmonomersof Ub mo

ecificity, ini

Ubwt. To

aborators i

t and wild-t

ethods in S

on the sub

e and abse

by Zhen Wgrams (left)s, and C) Hnomer ma

itial time p

confirm t

in Mary W

type monom

Section 4.2.

bstrate by

ence of Rp

76

Wou, and

HPLC asses

points

hese

Wirth’s

mers

.4), a

wild-

pn13.

77

Interestingly, Rpn13 appears to slow processing of triUb for both the wild-type

and E284A UCH37 (Fig 4.6). In the presence of Rpn13, almost no monoubiquitin

is generated by the two hour timepoint but the monoubiquitin band at the two

hour timepoint for UCH37 alone is about eight times more intense. Additionally,

after two hours, UCH37 + Rpn13 still has a significant amount of triubiquitin to

cleave, whereas UCH37 alone has cleaved almost all of its triubiquitin down to

FSp

d

p

tr

c

w

s

R

w

Figure 4.6: CSDS PAGE

resence an

i- or mono

olyubiquitin

riubiquitin le

In ord

an be seen

was assess

ubstrates. T

Rpn13 to de

which may b

Cleavage ogel of triU

nd absence

o-Ub. The

n processin

eft over afte

der to deter

n using a v

sed against

Tetraubiqui

etermine if

bind better t

of mutant trUbG76V W77 ce of GST-Rp

UCH37 E2

ng compare

er two hours

rmine if thes

variety of p

t tetraubiqu

itin cleavag

f inhibition

to the ubiqu

riubiquitin bcleavage bpn13.

284A mutan

ed to the w

s. E284A is

se same kin

polyubiquitin

uitin in par

ge was also

persists wi

uitin-binding

by UCH37 iby wild-type

nt appears

wild-type en

s similarly in

nds of rates

n chain typ

rallel with t

o done in th

ith a longe

g PRU dom

in the prese or E284A

s to be mil

nzyme beca

nhibited by

s of polyub

pes, the ac

triubiquitin

he presence

er polyubiqu

main in Rpn

ence of RpA UCH37 in

dly impaire

ause it still

Rpn13.

iquitin cleav

tivity of UC

and diubiq

e or absenc

uitin chain,

n13’s N-

78

pn13. n the

ed at

l has

vage

CH37

quitin

ce of

one

FPp te

te

c

fo

w

c

th

h

ti

h

b

Figure 4.7: CPAGE gel o

resence an

erminus. Ho

etraubiquitin

leaves tetra

or ubiquitin

would also

apable of e

hat of diub

owever, the

me points.

owever and

y ImageJ q

Cleavage oof tetra-, trnd absence

owever, Rp

n substrate

a-, tri-, and

n chain len

appear th

endo-cleava

biquitin spe

ere is still a

These po

d results w

quantificatio

of varying leri-, and di-

e of Rpn13.

pn13 contin

e (Fig 4.7).

d di-ubiquiti

gth (Fig 4

at UCH37,

age. The ra

ecies (espe

a small pop

olyubiquitin

ill need to b

on.

ength K48 ubiquitin cl

ued to inhib

From this

n equally w

.7). From q

, while pri

ate of mono

ecially notic

pulation of d

cleavage

be verified

polyubiquitleavage by

bit polyubiq

experimen

well and do

qualitative

marily an

oubiquitin a

ceable with

diubiquitin t

assay res

by repeat e

tin chains by wild-type

quitin cleava

nt, it appea

oes not hav

analysis o

exo-specif

accumulatio

h tetraubiqu

that is build

ults are fro

experiment

by UCH37. UCH37 in

age even w

ars that UC

ve a prefer

of these ge

ic DUB, is

on is faster

uitin cleava

ding up in i

om single-

ts accompa

79

SDS n the

with a

CH37

rence

els, it

s still

than

age),

initial

-trials

anied

80

4.4 Discussion

We have analyzed the contribution of UCH37’s ULD and found it to

provide 1) exo-specificity through binding to ubiquitin’s Lys48 and 2) a means of

activation of the enzyme through interactions with Rpn13. After analysis of

ubiquitin-AMC cleavage by ULD mutants in the presence and absence of Rpn13,

we have identified Glu284 as a critical regulator of Rpn13’s activation, in that

when it is mutated to Ala, activation is lost.

We have generated a novel polyubiquitin for the study of directional-

specific cleavage, triUbG76V W77, which allows detection of a monoubiquitin variant

by differences in molecular weight, hydrophobicity, and molar absorptivity. We

have not utilized this triubiquitin for exo-specificity assays yet, but have analyzed

UCH37’s ability to process tetra-, tri-, and di-ubiquitin in the presence and

absence of Rpn13. It initially appears that Rpn13 slows polyubiquitin cleavage by

UCH37, which has been noted by others but not fully explored9. The investigation

into the mechanism of activation of UCH37 has yet to be completely exhausted,

especially as it pertains to polyubiquitin cleavage. However, our novel substrate

has broader uses for other directional-specific deubiquitinating enzymes that

have sufficient rates of polyubiquitin cleavage and may be a useful tool within this

field.

81

4.5 References

1 Lam, Y. A., DeMartino, G. N., Pickart, C. M. & Cohen, R. E. Specificity of the ubiquitin isopeptidase in the PA700 regulatory complex of 26 S proteasomes. The Journal of biological chemistry 272, 28438-28446 (1997).

2 Kiefer, F., Arnold, K., Kunzli, M., Bordoli, L. & Schwede, T. The SWISS-

MODEL Repository and associated resources. Nucleic acids research 37, D387-392 (2009).

3 Emsley, P. & Cowtan, K. Coot: model-building tools for molecular

graphics. Acta crystallographica. Section D, Biological crystallography 60, 2126-2132 (2004).

4 Adams, P. D. et al. PHENIX: a comprehensive Python-based system for

macromolecular structure solution. Acta Crystallographica Section D: Biological Crystallography 66, 213-221 (2010).



6 Yao, T. et al. Proteasome recruitment and activation of the Uch37



Protein & cell 5, 616-630, doi:10.1007/s13238-014-0046-z (2014). 8 Luchansky, S. J., Lansbury, P. T., Jr. & Stein, R. L. Substrate recognition

and catalysis by UCH-L1. Biochemistry 45, 14717-14725 (2006). 9 VanderLinden, R. T. et al. Structural Basis for the Activation and Inhibition

of the UCH37 Deubiquitylase. Molecular cell 57, 901-911, doi:10.1016/j.molcel.2015.01.016 (2015).

82

CHAPTER 5: RECENT FINDINGS ON THE STRUCTURE AND ACTIVATION OF UCH37 AND OTHER DEUBIQUITINASES

5.1 Introduction

As of the writing of this document, recent work by two groups has

uncovered two novel structures of UCH37: (1) bound to its activator, Rpn13, and

ubiquitin and (2) bound to a fragment of NFRKB, its deactivator and a component

of the Ino80 complex, both done by Vanderlinden et. al, 2015 and Sahtoe et. al,

2015 1,2. These papers confirm some of the findings of this work, as well as leave

some questions open still open about how UCH37 is regulated. This chapter will

encompass an analysis of the new structures of UCH37 followed by a review of

deubiquitinating enzymes whose specificity and activation rely on small structural

elements, such as loops, in the same manner as UCH37.

5.2 Analysis of UCH37-Rpn13-Ub and UCH37-NFRKB-Ub Structures

Both structures reveal dramatic conformational changes on the part of

UCH37’s ULD domain (Fig 5.1) The ULD of apo UCH37 involves a helix-turn-

helix (α9 and α10) followed by a shorter helix, α11, and ending with a short

unstructured loop, as the structure lacks density for the final 18 amino acids of

Fa

U

b

a

a

st

U

a

b

a

Figure 5.1: Cs labeled.

UCH37’s C

inding to b

bout 120°

pproximate

tructure wit

ULD, with he

bove the ca

inding to R

ctivation.

Conformati

-terminus.

both Rpn13

towards he

ely 90° tow

thout ubiqu

elix α10 kin

atalytic dom

Rpn13 and

onal chang

However,

3 and NFRK

elix α10 an

wards the ca

uitin, even g

nking at its c

main1. Thes

NFRKB as

ges in the U

this region

KB, in whic

nd a new h

atalytic dom

greater con

center, pos

se structure

s well as pr

ULD domai

n dramatic

ch helix α1

helix, α12,

main (Fig 5

nformationa

sitioning the

es all show

rovide a str

in. From se

cally remod

11 of UCH3

is resolved

5.1). In the

al changes

e rest of α10

a direct rol

ructural bas

even structu

dels itself u

37 rotates

d, which is

NFRKB-bo

are seen in

0, α11, and

le of the UL

sis for UCH

83

ures,

upon

back

bent

ound

n the

d α12

LD in

H37’s

re

w

a

e

u

in

R

Flo

Surpr

esolved in

was proven

ctivator an

nough flex

nresolved)

n these stru

Rpn13, a str

Figure 5.2: Aoop is in pin

risingly, ou

x-ray crysta

n incorrect

nd substrat

ibility that it

1,2. Howeve

uctures (Fig

ructural val

Activation onk, with key

5.2.

r prediction

al structure

by these

e, the cros

t could not

er, importa

g 5.2). Both

idation for R

of UCH37 by residues h

1 Crossove

n that the c

es upon UC

recent st

ssover loop

be entirely

nt interactio

h Met148 a

Rpn13’s ac

by crossovehighlighted.

er Loop

crossover

CH37 bindin

tructures.

p (residues

y resolved (

ons betwee

and Phe149

ctivation me

er loop bind. From PDB

loop would

ng to ubiqu

Despite th

s 143-163)

(residues 1

en it and R

9 make dire

echanism. R

ing to Rpn1B ID 4UEL.

d be compl

uitin and Rp

he presenc

) still main

154-159 are

Rpn13 are

ect contact

Rpn13’s

13. Crossov

84

etely

pn13

ce of

tains

e still

seen

t with

ver

85

stabilization of an open conformation of the crossover loop likely allows improved

substrate binding and catalysis, especially for the proximal ubiquitin monomer not

seen in these structures. Mutations to Met148 and Phe149 render the enzyme

unable to be activated by Rpn13 in UbAMC assays1,2. It can be predicted that the

rest of the crossover loop would be visualized in a UCH37-diubiquitin-Rpn13

structure, and that more of the residues in this region would contribute to

substrate stabilization at the active site, especially the isopeptide bond.

Alternatively, it is possible that the cross over loop maintains its dynamic

character throughout catalytic steps of the enzyme, independent of substrate

binding.

5.2.2 NFRKB Mode of Inhibition

The structures of UCH37 bound to the Ino80 component, NFRKB,

illuminate the way in which the DUB is inhibited catalytically through both its

active and distal sites. NFRKB hijacks the distal region of UCH37 that binds to

the Leu8-Thr9 hairpin of ubiquitin, a key motif within ubiquitin for binding. NFRKB

buries its own Phe100 and Arg101 within the distal pocket, occluding ubiquitin

binding (Fig 5.3). Additionally, the large helix of NFRKB that lies across the active

site face of UCH37 induces small conformational strains that lead to an

unproductive orientation of the active site. The catalytic His has rotated and now

lies an unproductive 6.3 Å from the catalytic cysteine. All of these small changes

can be utilized for small-molecule targeting of UCH37, as they directly occupy

binding sites of ubiquitin.

FBbU(Ps

Figure 5.3: MB) active si

inding siteUCH37 apo PDB ID 4Whown.

Mode of NFite rearrang. NFRKB iin purple (

WLP), and

FRKB inhibgement ans shown in(PDB ID 3IH UCH37 b

bition of UCnd C) occlun yellow (PHR), UCH3bound to R

CH37. NFRKusion of thPDB ID 4W37 bound toRpn13 and

KB inhibitiohe UCH37 WLP), ubiqo NFRKB ad Ub (4WL

on of UCH3distal ubiq

uitin in oraand Ub in gLR). Rpn13

86

37 by quitin ange, green 3 not

87

5.3 Analysis of Kinetic Findings in Vanderlinden et. al and Sahtoe et. al

Interestingly, these groups studied the activation of the enzyme in UbAMC

hydrolysis assays using one of the ULD mutants discussed earlier, E284A in my

studies, but numbered E283A in the isoform these groups used. They found that

Rpn13 lowers the KM of UCH37 for ubiquitin 5-fold, but that the KM of the E283A

mutant in the presence of Rpn13 is only 1.5-fold improved compared to UCH37

alone, indicating that this residue may be essential to its activation mechanism,

similar to the results presented previously (Section 4.3.2)2. These results validate

our earlier findings, that E284 is essential to the activation mechanism of UCH37.

5.4 Small Structures Effect Large Changes: A Review of Deubiquitinases

Among the ~100 deubiquitinating enzymes in the human genome, a little

less than half of these do not contain auxiliary ubiquitin binding domains or

ubiquitin-like domains beyond their active sites3. These deubiquitinases must rely

on conformational movement and binding pockets inherent in their active sites

alone, or utilize non-canonical interactions with ubiquitin to bind substrate. There

have been many thorough reviews of the various ubiquitin binding and ubiquitin

like domains; however, little focus has been drawn to the smaller dynamic

movements that significantly contribute to deubiquitinating enzyme catalysis4-8.

88

5.4.1 Unproductive Active Sites

One of the simplest, but integral, conformational changes within

deubiquitinases (DUBs) is the reorientation of catalytic residues from an

unproductive form in the apo enzyme to a productive conformation upon ubiquitin

binding. This has been seen in structures of the cysteine protease DUBs and

frequently involves misaligned catalytic cysteines or histidines within their papain-

like active sites, less often their catalytic aspartic acid or stabilizing oxyanion

glutamine residues (Fig 5.4).

Catalytic rearrangement occurs upon ubiquitin binding for the UCH family

members UCHL19,10 and UCH371,2,11-14. In the apo UCHL1 active site, the

catalytic histidine resides 8.2 Å away from the catalytic cysteine and is turned 90°

from the typical orientation of a papain-like active site, an unproductive distance

for catalysis9. Upon ubiquitin binding, the histidine swings 90° to lay in-plane with

the catalytic Cys and has moved to a productive 4 Å distance10. In the apo active

site of UCH37, the catalytic His is in a productive orientation, but its catalytic

cysteine is rotated inwards, toward the His residue rather than towards the rest of

the catalytic cleft where ubiquitin will bind11-13. Upon ubiquitin binding, the Cys

rotates 70° to face Gly76 of ubiquitin, an appropriate conformation1,2,14.

Unproductive active sites have been found in the active sites of OTU-

domain containing DUBs as well, both OTU1B, K48-specific, and OTULIN, a

linear polyubiquitin-cleaving DUB. Upon binding to ubiquitin, OTUB1 has a

similar conformational change to UCHL1 and UCH37; its His rotates down 90° to

lock in plane with the catalytic Cys, and its Cys flips inward to face Gly76 of

89

ubiquitin, altogether moving the two residues closer by 3 Å into a catalytically-

productive conformation15,16 (Fig 5.4). The structure of OTULIN’s active site has

partial occupancy for two distinct orientations: one in which the catalytic His and

Cys are in appropriate conformations, and one in which the catalytic His is flipped

to occupy the space which Gly76 resides in the linear diubiquitin-bound

structure17.

HAUSP/USP7, one of the most well-characterized USP-family DUBs also

contains a misaligned active site, wherein its catalytic Cys is positioned 10 Å

away from the catalytic His (Fig 5.4). Upon ubiquitin binding, its catalytic Cys,

His, and Asp move closer together, to a 3.7 Å distance between the Cys and His

and a 2.7 Å distance between the His and Asp.

The current theory as to why these DUBs prefer a catalytically-

unproductive active site orientation in absence of ubiquitin is that it may protect

against oxidation18. Some deubiquitinases have been found to be highly

susceptible to oxidation, which can lead to irreversible modification (sulphinic or

sulphonic acid) of their catalytic cysteines, rendering the enzyme catalytically

dead. In the seminal work describing DUB oxidation, the OTU A20 was capable

of an initial state of reversible oxidation, which would attain irreversibility upon

continued exposure to the oxidant18. A20 does not have a misaligned active site,

however it is believed that DUBs with misoriented cysteines may induce

protective interactions with nearby residues, keeping the cysteine shielded from

oxidants.

FuHy

Figure 5.4: biquitin-bou

HAUSP/USPeast OTU1

Misalignedund statesP7 unboun (green) an

d active sits for A) d (wheat)

nd Ub-boun

tes of deubUCHL1

and boundd OTUB1 (

biquitinating(yellow) a

d to Ub (da(cyan).

g enzymesand UCHark red), a

s. Compare37 (blue),nd C) unbo

90

ed to , B) ound

91

5.4.2 Insertions and the JAMM Domain

Regulation of JAMM domain containing deubiquitinating enzymes is

mainly held by their insertion domains, numbered Ins-1 and Ins-219. The

insertions act as substrate stabilizers and confer specificity, as seen in the

structure of AMSH-LP bound to diubiquitin19 (Fig 5.5). AMSH-LP uses the sheets

of Ins-1 to bind the distal ubiquitin monomer and stabilize the isopeptide bond for

cleavage, and Ins-2 for additional isopeptide stabilization and binding of the

proximal ubiquitin monomer. Superposition of the structures of human AMSH,

another JAMM family DUB, and the AMSH-like S. pombe orthologue Sst2 on the

AMSH-LP structure shows similar modes of binding and stabilization by their

highly-conserved insertion domains. These small domains provide significant

contribution to catalysis; when mutated, they can render the enzyme catalytically

impaired. However, mutations to Ins-2 do not contribute to substrate binding as

they only affect kcat, not KM20,21. The isopeptide contacts by the insertions also

maintain specificity, in that AMSH-LP, AMSH, and Sst2 are all highly specific for

K63-linked polyubiquitin chains.

In contrast, two structures were recently solved for Rpn11, the JAMM DUB

resident in the 26S proteasome responsible for en bloc cleavage of ubiquitin

chains from proteasomal substrates, in which only Ins-1 was utilized for substrate

recognition and catalysis22-24. Ins-2 does not contribute to substrate catalysis;

rather, it assists docking Rpn11 to the proteasomal subunit Rpn2, as predicted

Faaoin

b

c

s

c

p

Figure 5.5: nd Ins-2 tond the morthologue S

n pink, teal,

y the rece

atalytic eng

pecificity27.

omplete re

rotein22,28-3

Insertion doo ubiquitin bodels bounSst2, and D light grey,

ent cryoEM

gagement

It is know

moval of po

0.

omains of binding in And to diUbD) Rpn11. Jand green,

M maps of

of Ins-2 p

wn to cleav

olyubiquitin

JAMM metA) the strucb of B) AMJAMM dom, respective

the 26S p

rovides a

ve entire ch

n from the l

talloproteascture of AMMSH, C) t

mains are inely.

proteasome

structural b

hains in an

ysine of att

ses. ContribSH-LP bouthe AMSH-n dark grey

e24-26 (Fig

basis for R

n en bloc f

tachment o

butions of und to K63--like S. poy with inser

5.5). This

Rpn11’s lac

fashion, tha

onto a subs

92

Ins-1 -diUb ombe rtions

non-

ck of

at is,

strate

ty

m

te

a

d

Fin(r

Memb

ypes of ub

members (e

ermed the c

bsence of

ensity is no

Figure 5.6: Cn grey surfared), yeast

bers of the

biquitin sub

except BAP

crossover l

ubiquitin, th

ot resolved.

Crossover ace, ubiquiYUH1 (pur

5.4.3 Su

e UCH fam

strates the

P1) contain

oop, to filte

hese DUBs

. Upon ubiq

loops for utin in orangple), UCHL

ubstrate-filte

mily of DUB

ey act upon

n a flexible

er out large

s’ crossove

quitin bindin

biquitin-bouge, and croL3 (green),

ering Loops

Bs utilize g

n. Structur

loop span

er substrate

r loops are

ng, nearly e

und UCH Dossover looand UCH37

s

ating loops

res of all t

nning their

es for cleav

e flexible an

every family

DUBs. The ops highligh7 (teal).

s to restric

he UCH fa

catalytic c

vage31-33. In

nd their elec

y member

UCH domahted for UC

93

ct the

amily

clefts,

n the

ctron

ain is CHL1

94

displays an ordered crossover loop that contributes some interactions to

ubiquitin’s C-terminal tail, buried within the active site (Fig 5.6). Only UCH37 still

lacks an ordered crossover loop, even in the presence of ubiquitin and its

activator, the proteasomal subunit Rpn131,2.

Through diubiquitin cleavage assays and protein chimeras altering the

length of the crossover loop, it is apparent that this loop is responsible for

substrate specificity by steric filtering32,33. Generally, the UCH family is believed

to only cleave small moieties from the C-terminus of ubiquitin, not processing of

polyubiquitin. The smaller family members, human UCHL1, human UCHL3, and

yeast YUH1, which contain solely a UCH domain, have the shortest crossover

loops and are not capable of polyubiquitin cleavage, only cleavage of ubiquitin-

AMC, a fluorogenic ubiquitin substrate with only the small AMC fluorescent group

attached to its C-terminus. The two larger UCH family members, UCH37 and

BAP1, have C-terminal extensions beyond their catalytic domains and contain

longer crossover loops than the other family members, which is believed to allow

them to cleave polyubiquitin. A Drosophila homolog of BAP1, Calypso, has been

shown to deubiquitinate histone H2B34 and UCH37 is proteasome-bound, and

therefore must process the variety of polyubiquitinated prey captured by the 26S.

The substrate-filtering theory was proven by an elegant experiment in which the

crossover loop of UCH37 was expanded by a poly-glycine insertion32. This

chimeric DUB was capable of both K48- and K63-polyubiquitin chain cleavage32.

Within that same study, the crossover loop of wild-type UCHL3 was

biochemically nicked, but this damage did not inhibit the DUB’s cleavage of

U

c

st

lo

Fain

UbAMC32. T

ontacts wit

tabilization

In co

oop of the U

Figure 5.7:nd BL2 in

n views in B

This would

th ubiquitin’

and cataly

ntrast to th

UCH DUBs

Occludingthe apo for

B) Ub-bound

d indicate t

’s C-termin

sis.

5.4.4 Sub

he insertion

, some loop

g loops of rm (pink) ad and C) un

that althou

al tail, it do

bstrate-occl

domains o

ps occlude

USP14. USnd bound tnbound. Fro

ugh the cro

oes not sig

luding Loop

of the JAM

substrate f

SP14 substo ubiquitinom PDB ID

ossover lo

gnificantly c

ps

MM DUBs o

from active

strate-occlu-aldehyde

D 2AYN and

op does m

contribute to

or the cross

sites. The

ding loops (blue). Zoo

d 2AYO.

95

make

o tail

sover

best

BL1 omed

96

known case of this is seen in the structures of USP14 alone and bound to

ubiquitin-aldehyde35. In its apo form, USP14’s active site is blocked by two loops,

BL1 and BL2, which occupy a portion of the space where ubiquitin’s C-terminal

tail would bind in order to access its catalytic cysteine (Fig 5.7). Upon binding to

ubiquitin, the entirety of these loops, as well as individual side chains, open up to

allow ubiquitin binding. Oddly, many of the residues within these loops are also

responsible for ubiquitin binding, such as Tyr333 and Phe331. These loops are

attributed to USP14’s poor reactivity with ubiquitin probes, namely ubiquitin-vinyl

sulfone, but that it becomes activated upon association with the 26S proteasome,

potentially through conformational restrains of these loops into a more open

form35. Although many other USP family DUBs contain loops within this region,

such as HAUSP/USP7, they do not sterically block those USPs active sites. It is

believed that the length and conformation of these loops confer USP14 with

unique reactivity compared to other USP family DUBs.

5.5 Conclusions

Through examination of the recent UCH37 activating and deactivating x-

ray crystal structures as well as an assessment of other small conformational

contributions to the regulation of deubiquitinating enzymes, we have a newfound

appreciation for the layers of specificity and mechanisms of action of DUBs, a

class of enzymes which process an incredible variety of cellular substrates. Only

~100 DUBs specifically recognize monoubiquitinated substrates, 8 homotypic

97

chain types, and a startling number of mixed linkage chain possibilities. Even

more shocking is that only half of them require additional ubiquitin binding

domains beyond their catalytic core in order to attain specificity and improve

substrate binding. DUBs containing only a catalytic domain rely on small

structures within themselves to restrict absolute specificity, as in the case of the

JAMM DUBs AMSH and AMSH-LP, or to allow processing of a broad spectrum

of substrates at a highly specific location and under certain conditions, as in the

case of the proteasomal DUBs Rpn11, USP14, and UCH37.

Although we have uncovered significant information regarding the binding

and catalysis of ubiquitin, as well as the binding and activation of Rpn13, to

UCH37, many questions still remain regarding its ability to process polyubiquitin

at the 26S proteasome. Despite our structural findings regarding its exo-specific

recognition of Lys48-linked chains, it is highly implausible that UCH37 would

have such limited polyubiquitin processing skills at the proteasome, with

increasing reports of proteasome processing of non-canonical ubiquitin

signals36,37. Further structural and biochemical studies of UCH37 within the

context of the 26S proteasome are needed to understand its role in the coupling

of deubiquitination and degradation.

98

5.6 References

1 Sahtoe, D. D. et al. Mechanism of UCH-L5 Activation and Inhibition by DEUBAD Domains in RPN13 and INO80G. Molecular cell 57, 887-900, doi:10.1016/j.molcel.2014.12.039 (2015).

2 VanderLinden, R. T. et al. Structural Basis for the Activation and Inhibition

of the UCH37 Deubiquitylase. Molecular cell 57, 901-911, doi:10.1016/j.molcel.2015.01.016 (2015).

3 Komander, D., Clague, M. J. & Urbe, S. Breaking the chains: structure and

function of the deubiquitinases. Nature reviews. Molecular cell biology 10, 550-563, doi:10.1038/nrm2731 (2009).

4 Faesen, A. C., Luna-Vargas, M. P. & Sixma, T. K. The role of UBL

domains in ubiquitin-specific proteases. Biochemical Society transactions 40, 539-545, doi:10.1042/BST20120004 (2012).

5 Husnjak, K. & Dikic, I. Ubiquitin-binding proteins: decoders of ubiquitin-

mediated cellular functions. Annual review of biochemistry 81, 291-322, doi:10.1146/annurev-biochem-051810-094654 (2012).

6 Dikic, I., Wakatsuki, S. & Walters, K. J. Ubiquitin-binding domains - from

structures to functions. Nature reviews. Molecular cell biology 10, 659-671, doi:10.1038/nrm2767 (2009).

7 Grabbe, C. & Dikic, I. Functional roles of ubiquitin-like domain (ULD) and

ubiquitin-binding domain (UBD) containing proteins. Chemical reviews 109, 1481-1494, doi:10.1021/cr800413p (2009).

8 Harper, J. W. & Schulman, B. A. Structural complexity in ubiquitin

recognition. Cell 124, 1133-1136, doi:10.1016/j.cell.2006.03.009 (2006). 9 Das, C. et al. Structural basis for conformational plasticity of the

Parkinson's disease-associated ubiquitin hydrolase UCH-L1. Proceedings of the National Academy of Sciences of the United States of America 103, 4675-4680 (2006).



99

11 Burgie, S. E., Bingman, C. A., Soni, A. B. & Phillips, G. N., Jr. Structural characterization of human Uch37. Proteins (2011).

12 Nishio, K. et al. Crystal structure of the de-ubiquitinating enzyme UCH37

(human UCH-L5) catalytic domain. Biochemical and biophysical research communications 390, 855-860 (2009).

13 Maiti, T. K. et al. Crystal structure of the catalytic domain of UCHL5, a

proteasome-associated human deubiquitinating enzyme, reveals an unproductive form of the enzyme. The FEBS journal 278, 4917-4926 (2011).



15 Messick, T. E. et al. Structural basis for ubiquitin recognition by the Otu1

ovarian tumor domain protein. The Journal of biological chemistry 283, 11038-11049, doi:10.1074/jbc.M704398200 (2008).

16 Edelmann, M. J. et al. Structural basis and specificity of human otubain 1-

mediated deubiquitination. The Biochemical journal 418, 379-390, doi:10.1042/BJ20081318 (2009).

17 Keusekotten, K. et al. OTULIN antagonizes LUBAC signaling by

specifically hydrolyzing Met1-linked polyubiquitin. Cell 153, 1312-1326, doi:10.1016/j.cell.2013.05.014 (2013).

18 Kulathu, Y. et al. Regulation of A20 and other OTU deubiquitinases by

reversible oxidation. Nature communications 4, 1569, doi:10.1038/ncomms2567 (2013).

19 Sato, Y. et al. Structural basis for specific cleavage of Lys 63-linked

polyubiquitin chains. Nature 455, 358-362, doi:10.1038/nature07254 (2008).

20 Davies, C. W., Paul, L. N., Kim, M. I. & Das, C. Structural and

thermodynamic comparison of the catalytic domain of AMSH and AMSH-LP: nearly identical fold but different stability. Journal of molecular biology 413, 416-429, doi:10.1016/j.jmb.2011.08.029 (2011).

100

21 Shrestha, R. K. et al. Insights into the mechanism of deubiquitination by JAMM deubiquitinases from cocrystal structures of the enzyme with the substrate and product. Biochemistry 53, 3199-3217, doi:10.1021/bi5003162 (2014).


degradation by the 26S proteasome. Science 298, 611-615, doi:10.1126/science.1075898 (2002).

23 Pathare, G. R. et al. Crystal structure of the proteasomal deubiquitylation

module Rpn8-Rpn11. Proceedings of the National Academy of Sciences of the United States of America 111, 2984-2989, doi:10.1073/pnas.1400546111 (2014).

24 Worden, E. J., Padovani, C. & Martin, A. Structure of the Rpn11-Rpn8

dimer reveals mechanisms of substrate deubiquitination during proteasomal degradation. Nature structural & molecular biology 21, 220-227, doi:10.1038/nsmb.2771 (2014).

25 Matyskiela, M. E., Lander, G. C. & Martin, A. Conformational switching of

the 26S proteasome enables substrate degradation. Nature structural & molecular biology 20, 781-788, doi:10.1038/nsmb.2616 (2013).

26 Lander, G. C. et al. Complete subunit architecture of the proteasome

regulatory particle. Nature 482, 186-191 (2012). 27 Mansour, W. et al. Disassembly of Lys11- and mixed-linkage polyubiquitin

conjugates provide insights into function of proteasomal deubiquitinases Rpn11 and Ubp6. J Biol Chem, doi:10.1074/jbc.M114.568295 (2014).

28 Yao, T. & Cohen, R. E. A cryptic protease couples deubiquitination and

degradation by the proteasome. Nature 419, 403-407 (2002). 29 Cooper, E. M. et al. K63-specific deubiquitination by two JAMM/MPN+

complexes: BRISC-associated Brcc36 and proteasomal Poh1. The EMBO journal 28, 621-631, doi:10.1038/emboj.2009.27 (2009).


roles of multiple deubiquitylating proteins associated with mammalian 26S proteasome. Mol Biol Cell 19, 1072-1082, doi:10.1091/mbc.E07-10-1040 (2008).

31 Misaghi, S. et al. Structure of the ubiquitin hydrolase UCH-L3 complexed

with a suicide substrate. J Biol Chem 280, 1512-1520 (2005).

101

32 Popp, M. W., Artavanis-Tsakonas, K. & Ploegh, H. L. Substrate filtering by the active site crossover loop in UCHL3 revealed by sortagging and gain-of-function mutations. The Journal of biological chemistry 284, 3593-3602 (2009).

33 Zhou, Z. R., Zhang, Y. H., Liu, S., Song, A. X. & Hu, H. Y. Length of the

active-site crossover loop defines the substrate specificity of ubiquitin C-terminal hydrolases for ubiquitin chains. The Biochemical journal 441, 143-149 (2012).


Polycomb repressive complex PR-DUB. Nature 465, 243-247 (2010). 35 Hu, M. et al. Structure and mechanisms of the proteasome-associated

deubiquitinating enzyme USP14. The EMBO journal 24, 3747-3756 (2005).

36 Mansour, W. et al. Disassembly of lys11 and mixed linkage polyubiquitin

conjugates provides insights into function of proteasomal deubiquitinases rpn11 and ubp6. The Journal of biological chemistry 290, 4688-4704, doi:10.1074/jbc.M114.568295 (2015).

37 Meyer, H. J. & Rape, M. Enhanced protein degradation by branched

ubiquitin chains. Cell 157, 910-921, doi:10.1016/j.cell.2014.03.037 (2014).

APPENDIX

102

APPENDIX

Inhibition of Uch37 and Usp14 at the 26s Proteasome and Its Effects on Degradation

6.1 Introduction

Although UCH37 is capable of activation by Rpn13 alone, its

activity is further enhanced within the context of the entire 26S proteasome, a

mechanism of activation which remains a mystery1-3. Due to UCH37’s poor

cleavage of di- or polyubiquitin substrates alone and in the presence of Rpn13, it

became clear that activity assays would not succeed without the entire 26S

proteasome present as an activator. To this end, we pursued purification of

endogenous 26S from rabbit tissue using an affinity-tag method developed by the

Goldberg group4,5. Using this proteasome, we hoped to study the roles of the

deubiquitinases UCH37 and USP14 within overall protein degradation and

whether their activities are coupled to the rate of degradation of polyubiquitinated

substrates.

This purification method relies on the affinity of the Rpn1 and Rpn10

subunits of the 26S proteasome for the Ubl domain of Rad23B, a shuttle factor

Fp

c

b

re

G

c

a

w

re

m

Figure A roteasome

apable of b

inding dom

eceptor. Fo

GST-tag, a

aptured by

nd then the

which binds

emoved by

more gentle

6.1: Sches.

binding to e

main of Rp

or purificati

allowing im

y this GST-

e proteasom

s to the GS

a subtract

e purificatio

eme of U

either of th

n1 or to o

on of the

mmobilizatio

-Ubl, non-s

me is elute

ST-Ubl and

ion step ov

n method t

bl-UIM pu

ose subun

ne of the

proteasome

on on glu

specifically

ed by additi

d releases

ver Ni NTA

than the tra

urification

its6,7. Its U

UIM doma

e, the Ubl

utathione b

binding pr

on of a 6xH

the protea

beads. Th

aditional su

method o

bl binds to

ains of Rpn

of Rad23B

beads. Pro

roteins are

His-tagged

some. Exc

his method

ucrose or g

of endoge

either the

n10, a ubiq

B is fused

oteasomes

washed a

UIM of Rp

cess His-UI

is consider

lycerol gra

103

nous

Ubl-

quitin

to a

are

away,

pn10,

IM is

red a

dient

104

centrifugation method or the TAP- or FLAG-tagged Rpn11 method, allowing

association of transient factors and ensuring the presence of the associated

deubiquitinases, UCH37 and USP144,5. This purification method is outlined in

Figure 6.1.

The goal of this purification method is to investigate the contribution of

UCH37 and USP14 to proteasomal degradation. There is evidence for the role of

USP14 from purified S. cerevisiae proteasomes, however, these lack UCH378,9.

The goal of the following work is to understand how inhibition of UCH37 and

USP14, achieved by incubation with the suicide inhibitor UbVME, affects

substrate degradation and deubiquitination.

6.2 Methods

6.2.1 Purification of Rabbit 26S Proteasomes

Following the method established by Besche et. al, we purified

endogenous levels of 26S proteasome from rabbit tissue (Pel-freez)4,5. This

affinity-tag method relied on the affinity of the proteasomal shuttle factor,

Rad23B, specifically its Ubl domain, for either Rpn10’s UIM domain or Rpn1’s

Ubl-binding site6,7. 2-4 grams of rabbit muscle tissue were homogenized on ice in

proteasome purification buffer (25 mM HEPES pH 7.5, 50 mM NaCl, 5 mM

MgCl2, 10% glycerol), PB, to which 1 mM ATP and 1 mM DTT were added, and

then were spun down at 100,000xg for 1 hour. 2 mg of GST-tagged Rad23B Ubl,

which had been previously purified recombinantly from E. coli, was immobilized

105

on glutathione beads and any excess washed off with PB. Rabbit lysate was

rocked with immobilized GST-Ubl for 2 hours at 4°C. Beads were collected in an

empty glass column and unbound proteins were allowed to flow through. 20

column volumes of PB with added ATP and DTT were used to wash the beads. 2

mg of 6xHis-tagged Rpn10 UIM was added to the beads and was rocked

overnight at 4°C to induce elution of pure proteasomes. Proteasomes were then

run over Ni NTA beads to re-capture the His-UIM. Pure proteasomes were

concentrated down and flash frozen as aliquots.

6.2.2 20S Activity Assays

To confirm the presence of the 20S core particle and test its activity, the

hydrolysis of succinate-Leu-Leu-Val-Tyr-AMC, a known 20S substrate, was

measured in the presence of rabbit 26S proteasomes. Suc-LLVY-AMC was

dissolved in DMSO. 5 nM 26S proteasome was diluted in AMC assay buffer


mM DTT and allowed to reach 30°C. 100 µM Suc-LLVY-AMC was added to 5 nM

proteasome in assay buffer and AMC hydrolysis was measured on a Tecan plate

reader at 380 nm excitation wavelength and 465 nm emission wavelength at

30°C for 1 hr. Progress curves were plotted in Kaleidagraph.

106

6.2.3 Ubiquitin-AMC Hydrolysis Assays

To test the deubiquitinating activity of endogenous UCH37 and USP14,

hydrolysis of UbAMC was measured. Proteasomes were diluted to 5 nM in

reaction buffer, 50 mM Tris pH 7.6, 0.5 mM EDTA, 0.1% bovine serum albumin,

and 5 mM DTT. Reactions were initiated by addition of UbAMC (Boston

Biochem) and were measured on a Tecan fluorescence plate reader (Männedorf,

Switzerland) with 380 nm excitation wavelength and 465 nm emission

wavelength at 30°C for 1 hr. Progress curves were plotted in Kaleidagraph.

6.2.4 Synthesis and Degradation of GFP-Titin-CyclinPY Substrate

In order to measure rates of degradation by the rabbit 26S, a

polyubiquitinated proteasomal substrate was needed. In the literature, there are a

handful of substrates, however, each is limited in scope and/or synthetic

simplicity. We utilized a substrate developed by the Martin group at UC Berkeley,

a GFP-tagged unstructured protein, titin, fused to cyclin, a known proteasomal

substrate, with an engineered PY motif, a degron which signals E3 ligases for

polyubiquitination10,11. After expression and purification from E. coli, the GFP-

titin-cyclinPY was polyubiquitinated by the E3 ligase Rsp5, also bacterially

expressed.

Fle

(E

D

in

m

w

o

e

w

a

c

Figure A 6.2eft, SDS PA

A mix

E3 enzyme

DTT in buffe

ncubated fo

mixture was

which point

f 40 µM GF

For d

ither 1 µM

were run w

dded to the

ontained 10

2: PolyubiquAGE gel of u

xture of 2.3

e), 0.5 mM

er containin

or 3-6 hours

s aliquoted

the substra

FP-titin-cycl

degradation

UbVME o

ith one tim

e t=0 time

00 nM 26S

uitination ofubiquitinatio

3 µM E1 en

ubiquitin, 4

ng 50 mM H

s at 37°C (

out, flash f

ate was use

linPY-Ubn.

n reactions

r buffer co

me point pe

point to pr

S proteasom

f GFP-titin-on reaction

nzyme, 5 µ

40 µM GFP

HEPES pH

Fig 6.2). Af

frozen, and

ed directly w

s, rabbit 26

ntrol for 2

er tube, an

revent any

me (+/- UbV

-cyclinPY sun shown at

µM Ubc4 (E

P-titin-cyclin

7.5, 50 mM

fter the fina

d stored at

with an ass

6S proteas

hours on i

d with 5xS

proteasom

VME), 2 µM

bstrate. Scright.

E2 enzyme

nPY, 5 mM A

M NaCl, 10

al time poin

-80°C unti

sumed stoc

some was

ce. Degrad

SDS PAGE

mal degrada

M GFP-titin-

cheme show

e), 11 µM R

ATP, and 1

0% glycerol

nt, the subs

l further us

ck concentr

incubated

dation reac

E buffer alr

ation. Reac

-cyclinPY-Ub

107

wn at

Rsp5

mM

was

strate

se, at

ation

with

ctions

eady

ctions

bn, 1

108

mM ATP, 1 mM DTT, and an ATP recycling system (creatine phosphokinase,

inorganic pyrophosphate, creatine phosphate) in 50 mM HEPES pH 7.5, 50 mM

NaCl, 10% glycerol. The reactions were run at 25 °C and were quenched with 5x

SDS PAGE buffer.

6.3 Results

6.3.1 Impact of Deubiquitinase Inhibition on Proteasome Degradation

The 26S proteasome was purified from rabbit muscle tissue using the Ubl-

UIM method developed by the Goldberg group and can be seen in Fig 6.34,5.

This method did not purify a large amount of proteasomes and retained the GST-

Ubl and His-UIM proteins despite subtraction over Ni-NTA beads. However, there

was sufficient to study the effects of inhibition of the associated deubiquitinases,

UCH37 and USP14, by UbVME treatment. To this end, proteasomes were

treated with and without 1 µM UbVME for 2 hours on ice to catalytically inactivate

the endogenous associated deubiquitinases. Then each sample was assessed

for its deubiquitinating activity by UbAMC hydrolysis assays and for its 20S core

particle activity using a fluorogenic peptide substrate, succinate-Leu-Leu-Val-Tyr-

AMC, or SucLLVY-AMC. Not surprisingly, in the presence or absence of UbVME,

the proteasome’s 20S activity was not changed (Fig 6.4). However,

Fu

d

(F

d

b

cy

a

in

S

s

Figure A 6.3sing the Ub

eubiquitina

Fig 6.4). U

egrade a p

y Rsp5, a

yclin, a kno

largely uns

n the prese

SDS PAGE

ubstrate. U

3: Silver-stabl-UIM meth

ating activity

bVME-inhib

proteasoma

system de

own protea

structured p

nce and ab

E gels sh

UbVME inhib

ained SDShod.

y was com

bited protea

l substrate,

eveloped by

somal subs

protein idea

bsence of U

howing dis

bition of the

S PAGE ge

mpletely abo

asomes we

, a GFP-titin

y the Marti

strate, with

al for degra

UbVME we

sappearanc

e proteasom

el of rabbit

olished upo

ere next as

n-cyclinPY f

n group10,1

an engine

adation by t

ere determin

ce of the

me resulted

26S protea

on treatme

ssessed fo

fusion prote

1. This sub

eered degro

he 26S. De

ned by Ima

highly po

d in about h

asome. Pu

nt with UbV

r their abili

ein ubiquitin

bstrate con

on fused to

egradation r

ageJ analys

olyubiquitin

half as slow

109

rified

VME

ity to

nated

tains

titin,

rates

sis of

nated

FtrpqS

Figure A 6.4reated proroteasomeuantitative

SDS PAGE

4: Activity ooteasome . C) SDS ImageJ angels.

of UbVME-tand B) PAGE ge

nalysis. D)

treated proDeubiquit

l of GFP-UQuantitativ

oteasome. Atinase actUbn substve results

A) 20S actitivity of rate degraof UbVME

ivity of UbVUbVME-tre

adation use inhibition

110

VME-eated ed in from

111

degradation compared to uninhibited proteasomes, despite activity of the 20S

core particle being unchanged in the presence of UbVME (Fig 6.4). This would

suggest that deubiquitination by UCH37 and USP14 are coupled to degradation,

a theory still under investigation.

Due to this result, we were curious about the effect of simultaneous

deubiquitinase and 20S inhibition, achieved by the addition of UbVME and the

20S inhibitor MG132. MG132 inhibits the β5 subunit of the 20S core particle,

thereby slowing its proteolytic activities by inhibiting one of the proteolytic

subunits12,13. We were curious as to the contribution of UbVME in slowing

proteasomal degradation compared to MG132, a well-characterized inhibitor.

Incubation of the 26S with inhibitors was done on ice for 2 hours in the presence

of either MG132 or a DMSO control, or UbVME or a buffer control. First, we

tested 20S and deubiquitinase activity by SucLLVY-AMC and UbAMC hydrolysis

(Fig 6.5). As expected, MG132 inhibits 20S activity but not deubiquitinating

activity, and UbVME inhibits deubiquitinating activity but not 20S activity. 20S

activity was slightly enhanced in the presence of UbVME, a phenomenon

previously observed by the Goldberg group (using ubiquitin-aldehyde), but

explained as an increase in AAA ATPase activity upstream14. Interestingly,

MG132 seems to slightly enhance deubiquitinating activity, but is within error,

therefore was not further investigated. These results indicate appropriate levels

of inhibition of the 20S and deubiquitinases by their respective inhibitors, so

FoUsre

Figure A 6.5f UbVME-

UbVME- orubstrate desults of Ub

5: Activity oor MG132

r MG132-tregradation bVME or M

f UbVME- a2-treated preated pro

used in G132 inhib

and MG132proteasomeoteasomes. quantitativeition from S

2-treated pres and B)

C) SDS e ImageJ SDS PAGE

roteasomeDeubiquitinPAGE geanalysis. D

E gels.

. A) 20S acnase activiel of GFPD) Quantit

112

ctivity ty of -Ubn tative

113

these samples were used to test degradation of the GFP-titin-cyclinPY-Ubn

substrate (Fig 6.5).

The rates of degradation of the GFP-Ubn substrate were measured by

running time points on SDS PAGE gels to show disappearance of the heavily

polyubiquitinated band, which were subsequently quantified by ImageJ analysis.

MG132 alone appears to only decrease degradation by about 25% compared to

the uninhibited sample, which is understandable because it only inhibits one of

the catalytic subunits of the 20S, rather than all three. UbVME alone shows a

decrease of 50% in degradation, similar to the results shown above in Figure 6.5.

However, the combination of MG132 and UbAMC slows degradation by >90%

compared to the uninhibited sample. As this amount is even greater than the

additive 75% of the two inhibitors alone, this result indicates that significant

inhibition of degradation is occurring in a coupled deubiquitination-degradation

mechanism.

6.4 Further Directions

Investigation into the coupling of UCH37/USP14 deubiquitination and

degradation by the 26S proteasome could prove vital to our understanding of all

protein degradation, but especially how these deubiquitinases may be the first

step in regulation of this cellular machine. The experiments addressed here

indicate that deubiquitination may be coupled to degradation, however, it is

necessary to separate out the effects of deubiquitination by Rpn11 before any

conclusions can be made. This could be achieved by incubation of proteasomes

114

with 1,10-phenanthroline, a known inhibitor of Rpn1112,15-17. It is necessary to

determine if deubiquitination by the cysteine protease deubiquitinases has a

separate function from that of Rpn11 and which level of deubiquitination

contributes most to slowing the rate of proteasomal degradation.

Additionally, the assays described here rely on SDS PAGE gel analysis of

the disappearance of a band of highly polyubiquitinated GFP substrates,

however, disappearance does not isolate deubiquitination from degradation.

These experiments are currently being pursued by a labmate, Michael Sheedlo,

using a T7 probe and polyubiquitinated Sic1, another proteasome substrate, to

determine the contributions of deubiquitination vs degradation. More specific

answers to these questions are needed before we can definitively say that

deubiquitination by UCH37 and USP14 are indeed coupled to and a contributing

factor during proteasomal degradation.

115

6.5 References

1 Lam, Y. A., DeMartino, G. N., Pickart, C. M. & Cohen, R. E. Specificity of the ubiquitin isopeptidase in the PA700 regulatory complex of 26 S proteasomes. The Journal of biological chemistry 272, 28438-28446 (1997).



3 Yao, T. et al. Proteasome recruitment and activation of the Uch37


4 Besche, H. C. & Goldberg, A. L. Affinity purification of mammalian 26S

proteasomes using an ubiquitin-like domain. Methods in molecular biology 832, 423-432, doi:10.1007/978-1-61779-474-2_29 (2012).

5 Besche, H. C., Haas, W., Gygi, S. P. & Goldberg, A. L. Isolation of

mammalian 26S proteasomes and p97/VCP complexes using the ubiquitin-like domain from HHR23B reveals novel proteasome-associated proteins. Biochemistry 48, 2538-2549, doi:10.1021/bi802198q (2009).

6 Hiyama, H. et al. Interaction of hHR23 with S5a. The ubiquitin-like domain

of hHR23 mediates interaction with S5a subunit of 26 S proteasome. The Journal of biological chemistry 274, 28019-28025 (1999).

7 Elsasser, S. et al. Proteasome subunit Rpn1 binds ubiquitin-like protein

domains. Nature cell biology 4, 725-730, doi:10.1038/ncb845 (2002). 8 Koulich, E., Li, X. & DeMartino, G. N. Relative structural and functional


9 Lee, B. H. et al. Enhancement of proteasome activity by a small-molecule

inhibitor of USP14. Nature 467, 179-184, doi:10.1038/nature09299 (2010). 10 Beckwith, R., Estrin, E., Worden, E. J. & Martin, A. Reconstitution of the

26S proteasome reveals functional asymmetries in its AAA+ unfoldase. Nature structural & molecular biology 20, 1164-1172, doi:10.1038/nsmb.2659 (2013).

116

11 Matyskiela, M. E., Lander, G. C. & Martin, A. Conformational switching of the 26S proteasome enables substrate degradation. Nature structural & molecular biology 20, 781-788, doi:10.1038/nsmb.2616 (2013).

12 Bush, K. T., Goldberg, A. L. & Nigam, S. K. Proteasome inhibition leads to

a heat-shock response, induction of endoplasmic reticulum chaperones, and thermotolerance. J Biol Chem 272, 9086-9092 (1997).

13 Stein, M. L. et al. Systematic comparison of peptidic proteasome inhibitors

highlights the alpha-ketoamide electrophile as an auspicious reversible lead motif. Angewandte Chemie 53, 1679-1683, doi:10.1002/anie.201308984 (2014).

14 Peth, A., Kukushkin, N., Bosse, M. & Goldberg, A. L. Ubiquitinated

proteins activate the proteasomal ATPases by binding to Usp14 or Uch37 homologs. The Journal of biological chemistry 288, 7781-7790, doi:10.1074/jbc.M112.441907 (2013).

15 Cooper, E. M. et al. K63-specific deubiquitination by two JAMM/MPN+

complexes: BRISC-associated Brcc36 and proteasomal Poh1. The EMBO journal 28, 621-631, doi:10.1038/emboj.2009.27 (2009).


degradation by the 26S proteasome. Science 298, 611-615, doi:10.1126/science.1075898 (2002).

17 Guterman, A. & Glickman, M. H. Complementary roles for Rpn11 and

Ubp6 in deubiquitination and proteolysis by the proteasome. J Biol Chem 279, 1729-1738, doi:10.1074/jbc.M307050200 (2004).

VITA

117

VITA

Marie Morrow was born in Winston-Salem, North Carolina to Jamie and

Christine Morrow. She grew up in Jacksonville, Florida and went to Stanton

College Preparatory School where she first came to love chemistry in Missy

Ray’s AP/IB Chemistry class. She went on to study chemistry at the University of

Florida where she did undergraduate research in the lab of Carrie Haskell-

Luevano, developing peptide inhibitors for melanocortin receptors. After

graduating with her Bachelor’s degree, she went to Purdue University to pursue

her Ph.D in chemistry. In Chitta Das’s group, she has studied the structure and

biophysical/biochemical properties of the proteasomal deubiquitinase UCH37.

After graduating, she will start a post-doctoral research position in the lab of

Cynthia Wolberger, Johns Hopkins University School of Medicine, studying the

structure and function of ubiquitination/deubiquitination machinery.

PUBLICATIONS

Stabilization of an Unusual Salt Bridge in Ubiquitin by the ExtraC‑Terminal Domain of the Proteasome-Associated DeubiquitinaseUCH37 as a Mechanism of Its Exo SpecificityMarie E. Morrow,† Myung-Il Kim,† Judith A. Ronau,† Michael J. Sheedlo,† Rhiannon R. White,‡

Joseph Chaney,† Lake N. Paul,§ Markus A. Lill,∥ Katerina Artavanis-Tsakonas,‡ and Chittaranjan Das*,†

†Department of Chemistry, Purdue University, 560 Oval Drive, West Lafayette, Indiana 47907, United States‡Division of Cell and Molecular Biology, Imperial College London, Sir Alexander Fleming Building, Imperial College Road, LondonSW7 2AZ, U.K.§Bindley Biosciences Center, Purdue University, West Lafayette, Indiana 47907, United States∥Department of Medicinal Chemistry and Molecular Pharmacology, Purdue University, 575 Stadium Mall Drive, West Lafayette,Indiana 47907, United States

*S Supporting Information

ABSTRACT: Ubiquitination is countered by a group ofenzymes collectively called deubiquitinases (DUBs); ∼100 ofthem can be found in the human genome. One of the mostinteresting aspects of these enzymes is the ability of somemembers to selectively recognize specific linkage typesbetween ubiquitin in polyubiquitin chains and their endoand exo specificity. The structural basis of exo-specificdeubiquitination catalyzed by a DUB is poorly understood.UCH37, a cysteine DUB conserved from fungi to humans, is aproteasome-associated factor that regulates the proteasome bysequentially cleaving polyubiquitin chains from their distal ends, i.e., by exo-specific deubiquitination. In addition to the catalyticdomain, the DUB features a functionally uncharacterized UCH37-like domain (ULD), presumed to keep the enzyme in aninhibited state in its proteasome-free form. Herein we report the crystal structure of two constructs of UCH37 from Trichinellaspiralis in complex with a ubiquitin-based suicide inhibitor, ubiquitin vinyl methyl ester (UbVME). These structures show thatthe ULD makes direct contact with ubiquitin stabilizing a highly unusual intramolecular salt bridge between Lys48 and Glu51 ofubiquitin, an interaction that would be favored only with the distal ubiquitin but not with the internal ones in a Lys48-linkedpolyubiquitin chain. An inspection of 39 DUB−ubiquitin structures in the Protein Data Bank reveals the uniqueness of the saltbridge in ubiquitin bound to UCH37, an interaction that disappears when the ULD is deleted, as revealed in the structure of thecatalytic domain alone bound to UbVME. The structural data are consistent with previously reported mutational data on themammalian enzyme, which, together with the fact that the ULD residues that bind to ubiquitin are conserved, points to a similarmechanism behind the exo specificity of the human enzyme. To the best of our knowledge, these data provide the only structuralexample so far of how the exo specificity of a DUB can be determined by its noncatalytic domain. Thus, our data show that,contrary to its proposed inhibitory role, the ULD actually contributes to substrate recognition and could be a major determinantof the proteasome-associated function of UCH37. Moreover, our structures show that the unproductively oriented catalyticcysteine in the free enzyme is aligned correctly when ubiquitin binds, suggesting a mechanism for ubiquitin selectivity.

The ubiquitin proteasome system (UPS), present in alleukaryotes, is responsible for the majority of controlled

degradation and recycling of proteins within the cell.1−5

Polyubiquitinated, and to some extent monoubiquitinated,proteins are recognized and degraded by the 26S proteasome, a2.5 MDa self-compartmentalizing proteolytic complex.6−13 It iscomposed of two major units: the 20S core particle (CP)consisting of 28 subunits and the 19S regulatory particle (RP)containing 19 subunits in yeast. The proteolytic active sites arehoused within the luminal chamber of the barrel-shaped CP,capped on both ends by the RP, which contains ubiquitinreceptors and enzymes that prepare substrates for degradation.

Entry of substrates into the CP is regulated by the RP, primarilyby opening and closing of the substrate translocation channel.Before the substrate is translocated into the narrow channelleading to the lumen of the CP, it is obligatorily deubiquitinatedwith the help of the RP-resident JAMM metalloproteaseRpn1114−16 and unfolded by Rpt subunits that sit within thebase subcomplex of the RP.7,9,14 However, additional regulation

Received: March 11, 2013Revised: April 25, 2013Published: April 25, 2013

Article

pubs.acs.org/biochemistry

© 2013 American Chemical Society 3564 dx.doi.org/10.1021/bi4003106 | Biochemistry 2013, 52, 3564−3578

118

is performed by proteasome-associated deubiquitinatingenzymes, whose underlying mechanism is still poorly under-stood.7,17

Attachment of ubiquitin to a lysine residue(s) on targetproteins is catalyzed by the sequential action of three enzymaticsystems: E1 (ubiquitin-activating), E2 (ubiquitin-conjugating),and E3 (ubiquitin-ligating) enzymes.18,19 Usually, ubiquitina-tion of a target protein results in the attachment of apolyubiquitin chain in which successive ubiquitin moieties areattached to one of the seven lysines, or the N-terminal aminogroup of the preceding monomer, to generate a homopolymericstructure.18,20 Polyubiquitin chains of a distinct topology arethus generated depending on which amino group of ubiquitin isused for chain extension (lysines 6, 11, 27, 29, 33, 48, and 63 orthe amino group of Met1). A polyubiquitin chain of a specifictopology is meant for a specific type of functional out-come.20−25 For example, a Lys48 (K48)-linked chain usuallyserves as the signal for proteasomal degradation, whereas K63chains signal other types of functions such as endocytosis, DNArepair, and NF-κB signaling.24,26

Ubiquitination works as a reversible post-translationalmodification, like phosphorylation. Deubiquitinating enzymes,or DUBs, can hydrolytically remove ubiquitin from proteinadducts, thereby opposing the action of ubiquitin conjugatingmachinery.27−33 Consequently, DUBs have been found to playimportant regulatory roles in numerous ubiquitin-dependentcellular processes.32−35 In mechanistic terms, these enzymescan be categorized into two main groups: cysteine proteasesand zinc metalloproteases. The zinc metalloproteases consist ofonly one family, the JAB1/MPN/MOV34 metalloenzymes(JAMMs). The cysteine proteases are further broken down intofour families based on the structure of their catalytic domain:ubiquitin carboxyl-terminal hydrolases (UCHs), ubiquitin-specific proteases (USPs), ovarian tumor proteases (OTUs),and Machado-Josephin domain proteases (MJDs).32

UCH37 (also known as UCHL5) is a 37 kDa DUB of theUCH family and is one of the two proteasome-associatedDUBs, the other being USP14 (Ubp6 in yeast), known toregulate protein degradation by the mammalian protea-some.36−40 These associated DUBs, along with Rpn11, aconstitutive member of the RP, conduct deubiquitination at theproteasome. However, the activities of the three enzymes aredistinct. Rpn11 is responsible for en-bloc removal ofpolyubiquitin chains prior to (or concurrent with) unfoldingand translocation of the substrate into the CP, an activity thatappears to be coupled to substrate degradation.15−17 USP14and UCH37 on the other hand are known to have chain-trimming functions.17,37,41 The importance of these associatedDUBs to proteasome function was revealed throughpharmacological inhibition of these enzymes. A small-moleculeinhibitor of USP14 appears to accelerate proteasomaldegradation of certain substrates, whereas UCH37 inhibitioncan stall proteolysis, consistent with distinct functional rolesplayed by the two enzymes.42−44

UCH37 was first identified as the PA700 isopeptidase, thecysteine DUB tightly associated with the RP, also known asPA700.38,45,46 Like other UCH family members, it contains aconserved catalytic triad of a cysteine, a histidine, and anaspartate. UCH37 has a canonical UCH domain that is 45%similar to UCHL1 and 49% similar to UCHL3, its single-domain family members.47−50 It also has an additional C-terminal tail domain responsible for its interaction with theRpn13 subunit of the RP.51−54 Proteasome-bound UCH37 is

thought to behave as an “editor”, relieving poorly ubiquitinatedsubstrates from degradation by sequentially dismantling theirK48-linked polyubiquitin chains from the very distal end,removing one ubiquitin at a time.37,38,45 Such a type of chaindisassembling activity can be termed as an exo cleavage activityin contrast to the endo activity, which leads to dismantling ofchains by cleavage between internal ubiquitins. Although it hasrespectable UbAMC (ubiquitin aminomethylcoumarin) hydrol-ysis activity in its unbound form, UCH37 has been shown torequire association with the proteasome to cleave diubiquitin(and polyubiquitin) chains.37 Additionally, its UbAMChydrolysis activity is enhanced upon binding with Rpn13.37,54

Interestingly, UCH37 also associates in the nucleus with thehuman Ino80 chromatin remodeling complex, where it is heldin an inactive state compared to the free enzyme.55 It thusserves as an example of a DUB whose catalytic activity is bothpositively and negatively regulated by binding to specificprotein partners, making it an attractive target for structuralstudies. Crystal structures have been determined for both thecatalytic domain and full-length human UCH37;56−58 however,the mechanism of its catalytic regulation upon binding toassociated protein factors is not known. Any mechanisticunderstanding of its regulation must require structuralinformation about UCH37 and its catalytic domain bound toubiquitin, which has yet to be reported.TsUCH37 is a recently characterized lower-organism

homologue of UCH37 from Trichinella spiralis (Ts), aninfectious helminth found nearly worldwide. TsUCH37 wasidentified by White et al. by incubation of the whole-cell lysateof Ts larvae with the HA-UbVME probe (HA, thehemagglutinin epitope, fused with the N-terminus of ubiquitinvinyl methyl ester), an epitope-tagged irreversible inhibitor ofcysteine DUBs.59 Its structural and functional homology withhuman UCH37 was then confirmed by sequence analysis, co-immunoprecipitation with proteasomal subunits, and UbAMChydrolysis assays. TsUCH37 is 45% identical to its humanhomologue and was shown to pull down TsADRM1, thecorresponding Rpn13 homologue, by co-immunoprecipita-tion.59 The sequence and functional conservation betweenthe Ts and human enzymes implies a similar chain-editing roleof the former at the proteasome. To understand themechanisms associated with UCH37, we have crystallized twoconstructs of TsUCH37 bound to ubiquitin vinyl methyl ester.The structures illuminate the mode of ubiquitin recognition inthe enzyme by revealing binding interactions with the catalyticdomain, which are conserved among UCH enzymes, andinteractions unique to UCH37, notably ubiquitin binding bythe ULD, providing further explanation of the proteasome-associated exo-specific deubiquitination activity of the DUB.

■ MATERIALS AND METHODSCloning, Expression, and Purification. TsUCH37cat.

TsUCH37cat (residues 1−226) was subcloned from the full-length construct (residues 1−309) in pET28a(+) into pGEX-6P-1 (GE Biosciences) using BamHI and XhoI restriction sites.The protein was expressed in Escherichia coli Rosetta cells(Novagen) grown at 37 °C in LB medium containing 100 μg/Lampicillin to an OD600 of 1.0 and then induced with 0.5 mMisopropyl β-D-thiogalactoside (IPTG) at 18 °C for 16 h.Harvested cells were resuspended in lysis buffer (1× phosphatebuffered saline and 400 mM KCl) and lysed with a Frenchpress. The lysate was then purified on a glutathione S-transferase (GST) column (GE Biosciences) followed by

Biochemistry Article

dx.doi.org/10.1021/bi4003106 | Biochemistry 2013, 52, 3564−35783565

119

cleavage of the GST tag by PreScission Protease (GEBiosciences) per the manufacturer’s instructions. It was furtherpurified by size exclusion chromatography on a Superdex 75column (GE Biosciences). Intein-fused ubiquitin1−75 in pTXB1was expressed in E. coli Rosetta cells and purified on chitinbeads (New England Biosciences). Ubiquitin vinyl methyl ester(UbVME) was synthesized by overnight incubation of Ub1−75-MESNa (MESNa, sodium mercaptoethanesulfonate) withglycine vinyl methyl ester and then purified on a MonoScation exchange column (GE Biosciences). Glycine vinylmethyl ester was synthesized by a modified, previouslypublished procedure.60 TsUCH37cat was reacted withUbVME for 4 h, followed by purification on a MonoQ anionexchange column (GE Biosciences) to separate any unreactedTsUCH37cat. Selenomethionine TsUCH37cat protein (SeMetTsUCH37cat) was grown in M9 minimal medium supple-mented with selenomethionine, reacted with UbVME, andpurified as described above.TsUCH37ΔC46. TsUCH37FL was subcloned previously into

pET28a(+) with an N-terminally fused His tag (Novagen).TsUCH37FL was expressed in E. coli Rosetta cells, grown at 37°C in LB medium containing 10 μg/L kanamycin to an OD600of 0.8, and then induced with 0.5 mM IPTG at 18 °C for 16 h.Harvested cells were resuspended in lysis buffer [50 mM Tris-HCl (pH 7.6), 200 mM NaCl, and 3 mM β-mercaptoethanol]and lysed with a French press. His-tagged TsUCH37FL waspurified by immobilized metal affinity chromatography (IMAC)and eluted with lysis buffer including 500 mM imidazole. Elutedproteins were further purified by size exclusion chromatography(SEC) on a Superdex 75 column (GE Biosciences) in 50 mMHEPES (pH 7.6) and 3 mM dithiothreitol (DTT). SDS−PAGE on the fractions indicated a cleavage of the full-lengthprotein, so the construct described is actually a proteolyticcleavage product of the full-length protein. The crystal structure(described below) lacks density for the last 46 amino acids fromthe C-terminus; therefore, this construct will hereafter bedescribed as TsUCH37ΔC46. Fractions containing the targetprotein were pooled, concentrated, and reacted with UbVME.UbVME was synthesized and reacted with purifiedTsUCH37ΔC46 as was done with TsUCH37cat. To separateunreacted TsUCH37ΔC46, the complex was further purified bySEC on a Superdex 75 column (GE Biosciences).Crystallization and Structure Determination.

TsUCH37cat−UbVME Complex. The TsUCH37cat−UbVMEcomplex was concentrated to 3 mg/mL in 50 mM Tris (pH7.6), 200 mM NaCl, and 1 mM DTT. Crystals were grown in 2days at room temperature by hanging drop vapor diffusion in 3M ammonium sulfate and 0.1 M bicine (pH 9.0) with 2 mM L-glutathione (mixture of oxidized and reduced) additive.Crystals were cryoprotected in 2.5 M sodium malonate andflash-frozen in liquid nitrogen.61 Diffraction data were collectedon a Mar300 CCD detector (Mar USA) at the 23-ID-Bbeamline at Argonne National Laboratory (Argonne, IL). Dataup to 1.7 Å were collected on SeMet TsUCH37cat−UbVMEcrystals at the selenium peak (0.979 Å) for SAD (single-wavelength anomalous dispersion) phasing. Data wereprocessed with HKL2000.62

The initial model was obtained by Se-SAD phasing in thePhenix AutoSol wizard.63 Its sequence was built in using thePhenix AutoBuild wizard, as well as manual model building inCoot.63,64 Structural refinement was conducted in Phenix usingTLS refinement (with the entire asymmetric unit taken as oneTLS group), as well as optimized weighting for stereochemical

restraints.63 The data were run through Phenix Xtriage, whichconfirmed the chosen space group, C2, and did not detectevidence of crystal twinning.63 The completeness of thecrystallographic data for the TsUCH37cat−UbVME complexwas less than ideal (see Table 2); however, this did not hinderthe determination of the structure or the generation of thestructural model presented herein and can be ascribed to poorcompleteness in the highest-resolution shells.

TsUCH37ΔC46−UbVME Complex. The TsUCH37ΔC46−UbVME complex was concentrated to 5 mg/mL in 50 mMHEPES (pH 7.6) and 2 mM DTT. Crystals were grown in 60days at room temperature in 0.2 M ammonium chloride (pH5.8) and 18% PEG3350. Crystals were cryoprotected inethylene glycol and flash-frozen in liquid nitrogen. Diffractiondata were collected on a Mar300 CCD detector (Mar USA) atthe 23-ID-B beamline at Argonne National Laboratory. Data upto 2.0 Å were collected on TsUCH37ΔC46−UbVME crystals at1.033 Å. Data were processed with HKL2000.62

The initial model was obtained by molecular replacementusing the Phenix AutoMR wizard, with a monomer of theTsUCH37cat−UbVME complex as the search model.63 Manualmodel building was conducted in Coot, and structuralrefinement was conducted initially in Refmac using TLSrefinement and then using simulated annealing and individual Bfactor refinement in Phenix.63,64 The data were run throughPhenix Xtriage, which confirmed the chosen space group, R3,and did not detect any evidence of crystal twinning.63

UbAMC Hydrolysis Assay. TsUCH37cat was diluted inreaction buffer [50 mM Tris (pH 7.6), 0.5 mM EDTA, 0.1%bovine serum albumin, and 5 mM DTT] to a final reactionconcentration of 7 nM and preincubated at 30 °C for 5 minprior to the addition of the UbAMC substrate (BostonBiochem). UbAMC cleavage was measured on a Tecan(Mannedorf, Switzerland) fluorescence plate reader with 380nm excitation and 465 nm emission wavelengths at 30 °C. Datawere fit to Michaelis−Menten kinetics in SigmaPlot (SystatSoftware, San Jose, CA).

Analytical Ultracentrifugation. Sedimentation velocityexperiments were conducted with the Beckman-Coulter XLAanalytical ultracentrifuge. The sample was extensively dialyzedagainst 50 mM Tris-HCl, 200 mM NaCl, and 1 mM DTT (pH7.4). The TsUCH37cat and TsUCH37cat−UbVME complexconcentration ranged from 10 to 32 μM. The samples werecentrifuged at 50000 rpm using a two-sector 1.2 cm path-lengthcarbon-filled Epon centerpiece. The experiments were con-ducted on an An-50 Ti rotor at 20 °C. The density and relativeviscosity of the buffers were calculated using SEDNTERPversion 1.09 (http://www.rasmb.bbri.org/rasmb/windows/sednterp-philo): 1.0079 g/mL and 0.01036 P, respectively.The partial specific volume (vbar) of the protein was alsocalculated from the protein sequence using SEDNTERP(0.7340 mL/g for TsUCH37cat and 0.7317 mL/g for theTsUCH37cat−UbVME complex). The samples were monitoredat 280 nm with a 4 min delay and 150 scans. The c(s)distributions were analyzed using SEDFIT version 13.0b.65

Molecular Dynamics Simulations. A model of full-lengthTsUCH37 was generated by the SwissModel homologymodeling server using the structure of the full-length humanprotein as a template.66 Missing ULD residues produced by themodel were appended to the TsUCH37ΔC46−UbVMEstructure in Coot, and a single round of refinement wasconducted in Phenix, to produce a final model hereafter termed“the system”.63,64 The system was solvated in a box of TIP3P



120

water with the minimal distance between any solute atom andthe boundary of the box set to 10 Å. The system wasneutralized with 15 Na+ ions, which were automaticallypositioned by the tleap program. Molecular dynamics (MD)simulations were performed using Amber 10 with Amber forcefield ff03.67 Periodic boundary conditions were applied, and thefull electrostatic energy was calculated using the particle meshEwald (PME) method.68 The simulation consisted of threesequential steps: energy minimization for 5000 steps (2500steps of steepest-descent followed by 2500 steps of conjugategradient minimization), equilibration for 100 ps of solvent withthe protein restraint with a force constant of 5 kcal mol−1 Å−1,and a final MD simulation for 2 ns. All simulations wereconducted at 300 K with a constant volume. A time step of 2 fswas used, and the SHAKE algorithm was applied to constrainthe bonds involving hydrogen atoms.69

■ RESULTS

TsUCH37, like its mammalian counterpart, contains a catalyticUCH domain, and an additional polypeptide chain following itcalled the C-terminal tail comprising the conserved UCH37-like domain (ULD) followed by a putative KEKE motif (Figure1b).37,51,70,71 The ULD in human UCH37 is thought to have aninhibitory role, presumably by folding onto the catalytic domainthereby occluding ubiquitin binding.37 However, how ubiquitinbinds to UCH37 has not been structurally characterized. Togain insight into how ubiquitin is recognized by TsUCH37, weaimed to crystallize both the catalytic domain of TsUCH37

bound to ubiquitin vinyl methyl ester (UbVME) (theTsUCH37cat−UbVME complex) and the UbVME complex ofthe full-length protein. UbVME is a suicide substrate of cysteineDUBs, which react with the former via nucleophilic attack ofthe catalytic cysteine at the vinyl group of the VME moiety,resulting in an irreversible modification whereby a covalentbond is formed between the catalytic cysteine and the VMEportion of the inhibitor (Figure 1a).36,48,60 This covalent adductis thought to mimic the acyl-enzyme intermediate formedduring deubiquitination reactions catalyzed by the DUB (II andIV in Figure 1a). If diubiquitin is used as the substrate, thedistal ubiquitin moiety is the acyl component of the acyl-enzyme intermediate, with the proximal ubiquitin acting as theleaving group during isopeptide bond hydrolysis (in diubiqui-tin, a lysine residue of one ubiquitin, called the proximalubiquitin, is linked via an isopeptide bond to the C-terminalcarboxylate group of another ubiquitin, called the distalubiquitin) (III in Figure 1a).The TsUCH37cat−UbVME complex crystallized in the C2

space group with two molecules of the complex in theasymmetric unit. Our attempts to crystallize the full-lengthversion, however, were met with limited success, the full-lengthprotein being susceptible to proteolysis as indicated by at leasttwo closely migrating bands in an SDS−PAGE gel (data notshown). While attempting to purify the full-length construct,we managed to retrieve a truncated version of the proteinlacking 46 amino acids from the C-terminal end of the protein(see Materials and Methods). This truncated protein was

Figure 1. (a) Schematic structures representing inhibition of UCH37 by UbVME (I and II). Definition of proximal and distal ubiquitin in adiubiquitin substrate (III). Schematic structure of the acyl-enzyme intermediate formed during deubiquitination catalyzed by a cysteine DUB (IV).The UbVME adduct (II) mimics the acyl-enzyme intermediate (IV), as shown in yellow. (b) Domain diagrams of TsUCH37 constructs compared toother UCH family members with UCH domains boxed in gray and additional domains boxed and labeled as shown. (c) Kinetic assay of UbAMChydrolysis by TsUCH37cat. (d) Analytical ultracentrifugation profiles of TsUCH37cat (left) and the TsUCH37cat−UbVME complex (right),indicating that both are monomeric in solution.



121

purified by Ni affinity chromatography and reacted withUbVME, and the complex was purified using ion-exchangechromatography. This complex, hereafter termed theTsUCH37ΔC46−UbVME complex (TsUCH37 missing the last46 residues), crystallized in the R3 space group with onecomplex in the asymmetric unit.The catalytic activity of TsUCH37cat was measured with a

UbAMC hydrolysis assay (Figure 1c), which yieldedMichaelis−Menten parameters as shown in Table 1. Compared

to the catalytic domain of human UCH37, TsUCH37cat has anapproximately 20-fold lower KM, indicating a higher affinity forthis substrate compared to that of the human protein, but a100-fold lower kcat, a substantially lower turnover number.Consequently, TsUCH37cat is nearly 5-fold less efficient thanthe UCH domain of human UCH37.Crystals of the TsUCH37cat−UbVME complex diffracted to

1.7 Å. The structure was determined by single-wavelengthanomalous dispersion (SAD) using anomalous scattering fromselenium (TsUCH37cat was labeled with selenium). Manualmodel building using Coot, followed by multiple rounds ofrefinement using Phenix, produced a final model with an Rfactor of 17.4% and an Rfree of 21% (see Table 2 forcrystallographic and refinement parameters).63,64 The finalrefined model corresponding to the asymmetric unit consists oftwo copies of the TsUCH37cat−UbVME complex, composed ofTsUCH37cat, residues 1−226, covalently connected via athioether bond linking the catalytic cysteine with the VMEgroup of UbVME (residues 1−75 of ubiquitin attached toGlyVME as the 76th residue, which is modeled as methyl 4-amino butanoate). The refined model was of high stereo-chemical quality, with <0.2% of residues in the disallowedregion of the Ramachandran plot and scoring in the upper 98%according to Molprobity evaluation.72 The structure of theTsUCH37ΔC46−UbVME complex (2.0 Å resolution) wasdetermined by molecular replacement using theTsUCH37cat−UbVME structure as the search model (Table2). The final refined model with good stereochemical quality(<0.2% of residues in the disallowed region of theRamachandran plot and Molprobity score of 63%) has aminoacids 5−263 of the protein and one UbVME linked via athioether bond to the catalytic cysteine. The structures of theUCH domain in the two constructs are very similar, except fortwo loop regions (see below), with Cα root-mean-squaredeviations (rmsds) of 0.32 Å between the two (the loop regionswere excluded from the calculation of the rmsd). Whendiscussing the structure of the UCH domain alone or itsinteraction with UbVME, we will therefore use the structure ofthe TsUCH37cat−UbVME complex because its resolution ishigher while specifically mentioning any structural feature thatis different in the UCH domain of TsUCH37ΔC46−UbVMEcomplex.Initial analysis of the structure revealed that the two copies in

the asymmetric unit of TsUCH37cat−UbVME crystals are

linked by a disulfide bond between Cys71 of the twoTsUCH37cat chains (Figure 2a). It is possible that the disulfidebond forms because the protein exists as a dimer in solution,bringing the cysteines into proximity of each other, or is a resultof crystallographic packing. To determine if this disulfide is acrystallographic artifact or a biologically relevant association, wedetermined the oligomerization state of complexed anduncomplexed (apo) TsUCH37cat by sedimentation velocityanalytical ultracentrifugation (AUC). We found that both thecomplex and the apo protein exist as monomers in solutionwith sedimentation coefficients (S20,w) of 3.3 and 2.8,respectively (Figure 1d), indicating that this disulfide is likelya result of crystal packing. TsUCH37 is expected to bepredominantly localized to the cytosol, a reducing environment,and therefore should not rely on disulfide-mediated dimeriza-tion for catalytic activity. Moreover, the observation that the

Table 1. Kinetic Parameters for TsUCH37cat

enzyme KM (nM) kcat (s−1) kcat/KM (×105 M−1 s−1)

TsUCH37cat 1085 0.37 3.4UCH37N240a 21493 34 16UCHL3a 77.1 19 2414UCHL1a 47.0 0.03 7.4

aKinetic parameters previously determined, from ref 75.

Table 2. Crystallographic and Refinement Statisticsa

SeMet TsUCH37cat−UbVME

TsUCH37ΔC46−UbVME

Data Collectionspace group C121 R3cell dimensions

a, b, c (Å) 171.2, 55.8, 73.9 147.4, 147.4, 40.5α, β, γ (deg) 90, 113.4, 90 90, 90, 120

wavelength (Å) 0.979 1.033resolution (Å) 50.00−1.70 (1.73−1.70) 50−2.0 (2.03−2.00)Rsym or Rmerge

b (%) 8.7 (50.0) 8.5 (83.8)I/σI 15.9 (3.0) 4.9 (4.1)completeness (%) 88.5 (42.0) 100.0 (100.0)redundancy 6.8 (3.5) 5.8 (5.7)

Refinementresolution (Å) 27.9−1.7 38.6−2.0no. of uniquereflections

62326/3126 22270/2002

Rworkc/Rfree

d 17.4/21.1 19.3/24.0no. of atoms

proteine 4650 2428ligand 24 8water 437 100

average B factor(Å2)

protein 36.2 43.4ligand 36.5 32.4water 44.0 44.3

rmsdbond lengths(Å)

0.013 0.009

bond angles(deg)

1.48 1.07

Ramachandran plot(%)

favored 98.1 97.7allowed 1.6 1.0outliers 0.4 1.3

aNumbers in parentheses refer to data in the highest-resolution shell.bRmerge = ∑|Ih − ⟨Ih⟩|/∑Ih, where Ih is the observed intensity and ⟨Ih⟩is the average intensity. cRwork = ∑||Fobs| − k|Fcal||/∑|Fobs|.

dRfree is thesame as Robs for a selected subset (5 and 9%) of the reflections thatwas not included in prior refinement calculations. eOrdered residues:Pro3−Gly141 and Lys153−Asp224 in chain C and Pro3−Gly141 andGln152−Gln225 in chain A of the SeMet TsUCH37cat−UbVMEstructure and Gly4−Lys57, Thr72−Gly141, and Glu157−Ala263 ofthe TsUCH37ΔC46−UbVME structure.



122

TsUCH37ΔC46−UbVME complex is a monomer in theasymmetric unit and that the segment of residues 57−71,which is used as a part of the dimer interface in the crystals ofthe TsUCH37cat−UbVME complex, is disordered in theTsUCH37ΔC46−UbVME structure (Figures 2 and 3) supports

the notion that the dimer observed in the TsUCH37cat−UbVME structure is a crystallographic dimer and may not existin solution. The two copies of the complex in the dimerobserved in the crystals of the TsUCH37cat−UbVME complexhave very similar structures with an rmsd of 0.39 Å between Cαatoms. We will therefore focus on one of them in discussionspresented below.Overall Structure of the UCH Domain of TsUCH37.

The overall structure of the TsUCH37 catalytic domain issimilar to that of other structurally characterized UCHenzymes.49,50,73 It has the classical αβα fold, in which a centralsix-stranded β-sheet is surrounded by six α-helices, five on oneside (α1−α5) and one on the other (α6) (Figure 3). The

overall architecture of TsUCH37cat can be seen as bilobal, withone of the lobes comprising helices α1−α5 and the othercomprising the β-sheets and helix α6. The active site is locatedat the interface of the two lobes, with Cys85 from helix α2 inone lobe and His161 from β3 in the other forming the catalyticCys-His pair. An adjacent loop provides the third member ofthe triad, Asp176. Most of the secondary structural elementsseen in TsUCH37cat are conserved in UCHL1 and UCHL3,with the only noticeable difference being the conformation of asegment following β2, residues 57−71. This segment is a helixin UCHL1 and UCHL3 and is in somewhat of an extendedlooplike conformation in human UCH37 (hUCH37) but isfairly ordered; in the structure of the various constructs ofhuman UCH37 determined so far, this loop has been found tobe in a similar conformation regardless of crystallographicpacking (Figure S1 of the Supporting Information).56−58,74 Incontrast, this segment appears to be flexible in TsUCH37 and isvisualized only in the TsUCH37cat−UbVME structure, in whichit forms the dimer interface between the two subunits in theasymmetric unit. In the TsUCH37ΔC46−UbVME complex, acrystallographic monomer, this loop is disordered (Figure 3).Although the possibility that its binding can influence the loopdynamics cannot be ruled out, it is unlikely that UbVME hasanything to do with the dynamic behavior of the loop because itdoes not bind to it. We therefore propose that the loop isintrinsically flexible in TsUCH37 but can become orderedunder certain circumstances, such as under the constraints ofcrystallographic packing.It is possible that the corresponding loop segment in

hUCH37 is somewhat dynamic as well, but it appears to besignificantly more flexible in TsUCH37. The significance of thisdifference in dynamics between the two proteins is not clear atthe moment. Intriguingly, the loop’s dynamic behavior appearsto have an effect on the conformation of a tryptophan residue(Trp55) adjacent to the active site (Figure S2 of the SupportingInformation). This tryptophan is conserved among Schizo-saccharomyces pombe (Sp), Ts, and human UCH37 (Figure S3

Figure 2. Crystal structures of TsUCH37 constructs bound to UbVME. (a) Dimeric structure of TsUCH37cat bound to UbVME (orange) incrystals. Monomers are colored teal (chain A) and gray (chain C). The inset shows the disulfide bridge that links the two subunits via Cys71. Theelectron density is rendered from the 2Fo − Fc map contoured at 1σ. (b) Monomer of the TsUCH37cat−UbVME structure. (c) Structure of theTsUCH37ΔC46−UbVME complex, with TsUCH37ΔC46 colored olive and UbVME orange.

Figure 3. Secondary structures of TsUCH37 constructs.TsUCH37ΔC46−UbVME and TsUCH37cat−UbVME complexes aresuperposed with α-helices and 310-helices colored pale yellow, β-sheetsblue, loops green, and UbVME orange. Arrows indicate where theTsUCH37ΔC46−UbVME structure lacks density, compared to theTsUCH37cat−UbVME structure, from residue 57 to 71.



123

of the Supporting Information). In the TsUCH37ΔC46−UbVME complex, Trp55 makes contact with the OMe groupof the suicide inhibitor, which in the actual substrate (aubiquitinated protein or the diubiquitin motif of a polyubiquitinchain) would be replaced by the hydrocarbon portion of theisopeptide-linked lysine side chain (Figure 1a). The sameresidue in the TsUCH37cat−UbVME complex shows a differentorientation with respect to the OMe group and appears to haveadopted a more open position for interaction with theisopeptide unit (Figure S2 of the Supporting Information).Therefore, Trp55 not only may provide important contactswith the isopeptide link to hold it in place near the active sitebut also may confer a certain plasticity to the active site ofUCH37, which may be useful for an induced-fit type ofengagement with the substrate.As stated before, in the TsUCH37ΔC46−UbVME structure,

we are able to visualize 40 additional amino acids after theUCH domain, the first 41 amino acids (residues 223−263) ofthe ULD in TsUCH37. The polypeptide chain, after emergingfrom the C-terminus of the UCH domain, adopts a helicalstructure of six turns (α7), takes a U-turn, and then continuesas a helix (α8). α7 and α8 are arranged as a helix−turn−helixmotif with a number of interhelical contacts, and this motifadopts a similar orientation with respect to the UCH domain asobserved in hUCH37 (Figures 3 and 4b).57 The only differencein this motif between TsUCH37 and hUCH37 is that it issomewhat shorter in the former. The ULD in TsUCH37appears to have a proteolytically susceptible region afterAla263, perhaps immediately following it, producing the C-terminal truncation we are observing here. When we model themissing part of the ULD, using the structure of hUCH37 as atemplate (see Materials and Methods), it is apparent that α8could have continued on after the cleavage site (Figure 4b)almost as a long helix all the way up to residue 285, except foran interruption at Arg268 where four successive residues,

including the arginine, adopt nonhelical dihedral anglesproducing a kink (a kink featuring equivalent residues is alsoseen in the template structure). As expected from the hUCH37structure, the model shows that after the interruption, the helixwould terminate at or near amino acid 285 (Figure 4), wherethe polypeptide chain reverses its direction as a turn segmentthat appears to cap the C-terminus of the helix. The putativeKEKE motif was not modeled because it is absent in thetemplate structure. Interestingly, the structure of theTsUCH37ΔC46−UbVME complex reveals side chains from α8making contact with ubiquitin, specifically with its Lys48residue, an interaction that may explain the distal end specificitydisplayed by UCH37 (discussed in more detail below). Also,the side chains from the modeled part of the ULD, missing inour structure, appear to present themselves for additionalcontacts with ubiquitin. Indeed, the two most conservedresidues in the ULD, Glu265 and Asn272, are facing ubiquitinand lie within contact distances (Figure 4c). Thus, it is possiblethat they may actually bind to ubiquitin. Alternatively, incontrast to what is predicted by the model, these residues maybe used for making contact with Rpn13, explaining why theyare conserved.

Active-Site Geometry. The catalytic triad in this cysteineprotease assumes a canonical arrangement in the ubiquitin-bound complex. The distance between the catalytic cysteineand histidine is 3.9 Å (Nδ−Sγ distance) in both structures, andthat between the histidine and aspartate is 2.8 Å (Nε−Oδ) inthe TsUCH37cat−UbVME complex and 2.9 Å in theTsUCH37ΔC46−UbVME complex. The distance between theCεH group of the catalytic histidine and the side chain carbonyloxygen of the oxyanion stabilizing glutamine (Gln79) is 3.3 Åin the TsUCH37cat−UbVME complex and 3.1 Å in theTsUCH37ΔC46−UbVME complex, suggesting a significantCH···O interaction between them, an interaction seen inother cysteine proteases as well.75 We were unable to crystallize

Figure 4. ULD−ubiquitin interactions. (a) Sequence alignment of the ULD of UCH37 highlighting conserved residues in UCH37 homologues.Glu265 and Asn272 (according to Ts numbering) are absolutely conserved (highlighted in red). (b) Superposition of the TsUCH37ΔC46−UbVMEcomplex (ULD colored olive and UbVME orange), human UCH37 (ULD colored purple, PDB entry 3IHR), and TsUCH37 with the entire ULDmodeled (cyan) based on the structure of the ULD in human UCH37. The model was generated using SwissModel and MD simulation (please seeMaterials and Methods). This model is taken from a snapshot collected at 1.3 ns during a 2 ns MD simulation run. (c) Structure of the TsUCH37−ubiquitin complex with the entire ULD modeled as shown in panel b, showing that the conserved residues of the ULD could make additionalcontacts with ubiquitin. The regions marked i and ii are expanded in the panels below. The UCH domain is colored gray.



124

the apo form of either TsUCH37cat or TsUCH37ΔC46. In itsplace, we use the structure of apo human UCH37 to gaininsight into structural changes in the active-site region that mayoccur upon ubiquitin binding.57,58 Comparison with thestructures of human UCH37 reveals that the catalytic cysteinehas changed its orientation, going from the apo form to theubiquitin-bound form, adopting a more productive orientationin the latter, an orientation in which the catalytic cysteine’s sidechain faces the catalytic cleft (Figure 5g). This analysis suggeststhat UCH37 exists in an unproductive form in the absence of

ubiquitin, with the catalytic thiol facing the interior of theprotein rather than the open space in the catalytic cleft,58 but isinduced to adopt a more productive form upon its binding.Thus, UCH37 may offer yet another example of a UCH DUBthat undergoes substrate-induced reorganization to a moreproductive form.48,76

Crossover Loop Flexibility. A common structural featurepresent in all UCH enzymes is the crossover loop, which inTsUCH37 spans residues 141−157 (connecting α5 with β3). Itstraddles the active-site cleft as a flexible loop and is known to

Figure 5. Recognition of ubiquitin by TsUCH37. (a) Surface rendering of TsUCH37cat (cyan) with ubiquitin binding regions highlighted. The distalsite is colored yellow and the active-site cysteine red, and resolved portions of the crossover loop are colored pink. (b) Surface rendering ofTsUCH37ΔC46 (green) with ubiquitin binding regions highlighted as in panel a, except with additional C-terminal tail ubiquitin binding residuescolored blue. (c) Interactions near the active-site cleft with the C-terminal hexapeptide tail of ubiquitin. UbVME residues are colored orange,TsUCH37 residues teal, and human UCH37 residues purple. (d) Interactions of Arg72 of ubiquitin with surrounding residues of TsUCH37cat.Density from the 2Fo − Fc map is contoured at 1σ (blue mesh). (e) UCH37 distal-site binding residues, with TsUCH37 colored teal and humanUCH37 purple. (f) Ile44 patch interacting residues, with UbVME colored orange, TsUCH37 teal, and human UCH37 purple. Waters involved inbinding are also shown, enveloped with density from the 2Fo − Fc map contoured at 1σ. Sequence alignment of this region in TsUCH37 comparedto human UCH37 is shown as an inset. (g) Active site of TsUCH37 (teal), showing the catalytic residues, compared to human UCH37 (purple),with UbVME colored orange.



125

provide steric constraint, limiting the size of the leaving groupat the C-terminus of ubiquitin.74,77 Accordingly, UCH enzymes,such as UCHL1 and UCHL3, can cleave only small leavinggroups from the C-terminus of ubiquitin, not large proteins oranother ubiquitin.47 However, UCH37 is known to cleavediubiquitin (and polyubiquitin chains), but only when it isassociated with the RP, being activated upon binding to itsprotein cofactor, Rpn13.37 All previously determined structuresof UCH enzymes bound to ubiquitin have shown a resolvedcrossover loop, which makes contact with at least one residuefrom the C-terminal tail of ubiquitin. In the apo form ofUCHL3, the closest homologue of UCH37, the loop isdisordered but becomes ordered when ubiquitin is bound.48,50

The ubiquitin-bound structures of PfUCHL3 and the yeastubiquitin hydrolase Yuh1 show an ordered crossover loopmaking contacts with side chains on the C-terminal tail ofubiquitin.73,78 In contrast, the structures of the TsUCH37−UbVME constructs present the only examples so far of a UCHDUB in which the crossover loop is still disordered even afterubiquitin is bound, indicating that the loop is flexible and doesnot contribute to ubiquitin binding. A small network of van derWaals interactions and hydrogen bonds seem to stabilize part ofthe crossover loop (residues 152−157) in a short helicalconformation in the structure of the TsUCH37cat−UbVMEcomplex, but the same segment in the TsUCH37ΔC46−UbVMEstructure is disordered and hence not visible, supportingdynamic sampling of conformations by this loop. Theobservation that the crossover loop is flexible despite thebound ubiquitin may be related to its activation by itsproteasome cofactor Rpn13.37 By not engaging with ubiquitin,the loop is available to freely interact with the cofactor, whichmay stabilize it in a conformation that leaves the active sitemaximally open to accommodate the isopeptide bond betweentwo ubiquitins or between ubiquitin and an acceptor protein.Interactions with Ubiquitin. The interaction of UbVME

with the TsUCH37cat UCH domain buries a total of 2355 Å2 ofsolvent accessible surface area, a value comparable to theamount buried in other UCH domain ubiquitin complexes (theburied accessible surface area in the TsUCH37ΔC46−UbVMEcomplex is 2479 Å2).48,76 The interaction is predominantlylocalized at two areas on TsUCH37, the active-site cleft and thedistal site (Figure 5a,b) The active-site cleft engages the C-terminal hexapeptide segment, Leu71ArgLeuArgGly-Gly76VME,of UbVME with numerous intermolecular contacts that includevan der Waals, hydrogen bonding, electrostatic, and water-mediated interactions (Figure 5c). This segment sits in theactive-site cleft with an extended conformation to maximizeinteractions with both backbone and side chain atoms of nearbyresidues of the enzyme. As seen in other UCH structures, thenarrowest part of the active-site cleft surrounds the terminalGly-Gly motif, with the last Gly (GlyVME in this case) beingplaced immediately adjacent to the Sγ atom of the catalyticcysteine, precisely located for nucleophilic attack on the scissilepeptide bond (Figure 5a,b). It is interesting to note that Arg72of UbVME is engaged in at least three major interactions(Figure 5d), suggesting that it contributes significantly tostabilizing the enzyme−substrate complex. The interactionswith Arg72 imply that TsUCH37 will find NEDD8 (neuralprecursor cell expressed, developmentally downregulated 8, astructurally similar ubiquitin-like protein modifier with asequence that is 60% identical with that of ubiquitin) as apoorer substrate because this arginine is replaced with alaninein NEDD8. Indeed, TsUCH37 does not cleave NEDD8-AMC

(see Figure S4 of the Supporting Information). Many of theactive-site interactions observed in the ubiquitin-boundstructures of UCHL1, UCHL3, PfUCHL3, and Yuh1 areconserved in both TsUCH37 structures. Additionally, thoseresidues surrounding the C-terminal hexapeptide tail ofubiquitin are strongly conserved between the Ts and humanprotein (Figure 5c).The interactions at the active-site cleft appear to be necessary

for precise cleavage at the terminal glycine residue of ubiquitin,while the distal site provides additional interaction to stabilizethe enzyme−substrate complex (Figure 5e,f). The distal siteengages the N-terminal β-hairpin of ubiquitin, which docks byutilizing interactions primarily involving the two-residue β-turnsegment, Leu8 and Thr9 of ubiquitin. These interactions aremostly hydrophobic in nature, involving van der Waals contactof Leu8 and Thr9 with Val35, Leu36, Ile206, Phe216, andLeu218, residues that constitute the surface-exposed hydro-phobic crevice that is the distal site. Leu36, Ile206, Phe216, andLeu218 are conserved among Sp, Ts, Pf, and human UCH37(Figure S3 of the Supporting Information), suggesting theimportance of distal-site binding in enzyme−substrate recog-nition.Ile44 of ubiquitin, a residue widely used in recognition by

ubiquitin-binding proteins, including DUBs, is seen making vander Waals contacts with Val34 on a greasy loop in TsUCH37,residues 34−36 (residues Val35 and Leu36 extend into thedistal-site pocket) (Figure 5f). A similar motif is used in otherUCH enzymes to bind to Ile44 of ubiquitin. Val34 ofTsUCH37 also makes contacts with His68 and Val70, which,together with Ile44 and Leu8 from the N-terminal β-hairpinturn, form the so-called Ile44 patch on ubiquitin. Thus, thebinding potential of the Ile44 patch on ubiquitin appears to befully satisfied in structures of the two complexes presented here,with each residue in the patch making at least one contact withthe enzyme. The structural data presented here are supportedby previously reported mutational analysis of the PA700isopeptidase. Replacing Ile44 and Leu8 from the Ile44 patchwith alanine in the distal ubiquitin of a diubiquitin substrateresults in significantly impaired catalysis with no detectablehydrolysis product.45 Val34 and Val35 are replaced withtryptophan and serine, respectively, going from Ts to humanUCH37 (Figure 5f) (Val34 provides additional contacts withVal70 of UbVME). These residues also show variability amongother UCH family members. Subtle differences in the Ile44patch-binding residues could be one of the contributing factorsin the difference in KM between human and Ts UCH37,especially as most of the residues at the active site are conservedbetween the two.There appear to be no striking conformational changes

between the ubiquitin-bound form of TsUCH37 and apohUCH37 except for the aforementioned reorientation of thecatalytic cysteine. However, we cannot rule out the possibilitythat significant conformational changes might have occurred asa result of ubiquitin binding in the Ts enzyme because we couldnot crystallize its apo form.

Ubiquitin Binding by the ULD. As mentioned earlier, theULD of hUCH37 was thought to have an inhibitory role,presumably by folding onto the catalytic domain andobstructing substrate binding.37 In contrast, the structure ofthe TsUCH37ΔC46−UbVME complex provides crystallographicevidence that the ULD can actually contribute to ubiquitinbinding and therefore can play a productive role in catalysis.Arg261 and Tyr262 on α8 of the ULD approach ubiquitin to



126

engage in van der Waals contact with of three of its side chains,Lys48 (with Arg261) and Gln49 and Arg72 (both with Tyr262)(Figure 6). Most notably, Arg261 is oriented in such a way toengage in close van der Waals contact with the hydrocarbonportion of the Lys48 side chain, forcing it to adopt an unusualconformation that allows an intramolecular salt bridgeinteraction with Glu51. This interaction is not observed inany of the 39 other ubiquitin-bound DUB structures currentlyfound in the PDB, catalogued in Table 3; the Lys48−Glu51distance is greater than 5.8 Å in all. Figure 6b shows theorientation of the same lysine in the TsUCH37cat−UbVMEcomplex. Clearly, the orientation is different in this structure,and the intramolecular salt bridge in ubiquitin is absent,suggesting that Arg261 of the ULD plays a role in inducing theunusual orientation of Lys48 of ubiquitin. Arg261 is conservedamong Sp, Ts, and human UCH37 (Figure 7) but is replacedwith leucine in PfUCH37 (also known as PfUCH54). Tyr262 isconserved in human and Ts forms but is substituted withtryptophan in Sp and PfUCH37. Inspection of the structurereveals that the van der Waals contact with Lys48 is still feasiblewith leucine in place of arginine and tryptophan canconservatively replace tyrosine as well. Thus, it is likely that

ULD binding with Lys48 and subsequent formation of theintramolecular salt bridge we are observing here are conservedfeatures of UCH37 in general.UCH37, as a part of PA700, is known to selectively cleave

polyubiquitin chains from the very distal end, sequentiallyremoving one ubiquitin at a time.38 The structural basis of thisexo cleavage specificity is not yet known. The uniqueorientation of Lys48 stabilized by Arg261 leading to theintramolecular salt bridge may explain this selectivity. Wepropose that although a similar type of interaction betweenArg261 and ubiquitin’s Lys48 is possible with an internalubiquitin, the intramolecular salt bridge will be absent in thiscase because the amino group of the lysine is acylated andhence not charged. Thus, it is the lack of an additionalinteraction with an internal ubiquitin that makes binding toLys48 of the terminal ubiquitin more favored, hence the exoselectivity.

■ DISCUSSION

UCH37 is a proteasome-associated UCH DUB known to havepolyubiquitin chain-editing function. It preferentially cleavesthe chain from its very distal tip.38 Such a function might rescue

Figure 6. ULD of TsUCH37 binding to ubiquitin. (a) TsUCH37ΔC46 (olive) ULD residues interacting with UbVME (orange). The inset showsinteractions of Arg261 and Tyr262 with UbVME, as well as the intramolecular salt bridge formed between Lys48 and Glu51 of ubiquitin. Densityrendered from the 2Fo − Fc map contoured to 0.7σ. (b−e) Comparison of the Lys48−Glu51 distances in ubiquitin observed in other DUB−ubiquitin structures. (b) Lys48 and Glu51 form a 3.7 Å salt bridge in the TsUCH37ΔC46−UbVME structure (olive), but not in the TsUCH37cat−UbVME structure (9.9 Å). (c) The same distance in all other UCH−ubiquitin structures is ≥9 Å: UCHL3−UbVME (yellow, PDB entry 1XD3),UCHL1−UbVME (red, PDB entry 3KW5), Yuh1−Ubal (pink, PDB entry 1CMX), PfUCHL3−UbVME (purple, PDB entry 2WDT), andTsUCH37cat−UbVME (teal). (d) The same distance is 8.7 Å in the Otu1−ubiquitin structure (dark red, PDB entry 3BY4) and 6.0 Å in the DUBA−Ubal structure (pale yellow, PDB entry 3TMP). (e) The same distance is 10 Å in the HAUSP/USP7−Ubal structure (blue, PDB entry 1NBF) and10.9 Å in the USP14−Ubal structure (brown, PDB entry 2AYO).



127

certain substrates from being committed to further downstreamaction of the proteasome.38 It is also possible that certainsubstrates carry inappropriate polyubiquitin tags that are notoptimal for their degradation. The chain-editing function mightbe essential for releasing these substrates to clear up ubiquitinreceptors for binding to productive substrates. A regulator ofproteasome function, it is itself regulated by binding to theproteasome: UCH37 is activated upon binding to Rpn13, asubunit of PA700 (the 19S proteasome or RP), the mechanism

of which is not understood. We report here the structure of twoconstructs of UCH37 from the infectious helminth T. spiralis(Ts) bound to the suicide inhibitor UbVME. This workconstitutes the first structural analysis of a ubiquitin pathwayprotein in the organism showing how ubiquitin is recognized bythis UCH family DUB in Ts. The structures reveal strikingconservation of the ubiquitin binding mode among UCHDUBs, from lower eukaryotes to human (Figure S5 of theSupporting Information). It also shows important structuraldifferences between other UCH DUBs, such as UCHL1 andUCHL3, some of which could be used for the specializedfunction of UCH37. While revealing interesting differences, theTs structures provide a number of details that may also holdtrue for the human enzyme, advancing our understanding ofUCH37 in general.The active-site cysteine may undergo ubiquitin-mediated

reorientation to a more productive form (Figure 5g), makingUCH37 yet another example of a UCH DUB that showsregulation of activity by ubiquitin, a feature that may provideselectivity to this group of cysteine proteases. Structures of thetwo constructs reveal invariant parts of the enzyme, likely lessdynamic parts, while also revealing parts that are more dynamicin nature, such as the segment of residues 57−71 and Trp55.Future studies should reveal the role of such dynamic parts incatalysis or regulation thereof.Importantly, the structure of the construct with the

additional 40 amino acids after the UCH domain reveals thatthe ULD could contribute to ubiquitin binding (Figures 4 and6), an unexpected finding because it was thought to beinhibitory in the human enzyme.37 The interaction of Arg261on the ULD appears to engage Lys48 of the distal ubiquitin in away that would be energetically most favored with the veryterminal ubiquitin in a polyubiquitin chain, possibly explainingthe exo specificity displayed by mammalian UCH37. Thesestructural data are consistent with previously reported muta-tional analysis probing substrate specificity of the PA700isopeptidase: mutation of Lys48 to cysteine on the distalubiquitin of a diubiquitin substrate results in severely impairedcatalysis.45 Apart from the broad agreement with theaforementioned experimental work, this observation of theintramolecular Lys48−Glu51 salt bridge in the distal ubiquitin,apparently induced by Arg261, is purely crystallographic at thispoint, although it seems unlikely that lattice forces haveanything to do with it. Even if the opposite is true, the fact thatsuch interactions are physiologically relevant cannot be ignored.The lack of an intramolecular Lys48−Glu51 salt bridge in anyother ubiquitin-bound DUB structures to date (Table 3) makesthis unusual interaction more intriguing, and worth additionalstudy. This observation therefore lays the structural ground-work for future mutational analysis aimed at validating theirexistence in solution and their role in substrate specificity.It is interesting to note that a salt bridge interaction, albeit an

intermolecular one, involving Lys48 of ubiquitin and an acidicside chain of the enzyme is also seen in the structure of USP7bound to ubiquitin aldehyde (the Lys48 side chain of the distalubiquitin is interacting with Asp305 and Glu308 of USP7).79

Such bifurcated salt bridges will perhaps contribute substan-tially to the binding of the enzyme to distal ubiquitin in a K48-linked chain, based on which one may predict that USP7 willalso exhibit exo specificity. This needs to be examined.Preferential cleavage from the very distal tip of a Lys48-linkedpolyubiquitin chain may be a feature common to DUBs thatwork on chains of this topology. Lys48-linked chains are known

Table 3. Lys48−Glu51 Distances for All DUB−UbiquitinComplexes

PDBentry DUB−ubiquitin complex

Lys48−Glu51distancea (Å)

UCH Family1XD3 UCHL3−UbVME 9.1, 11.51CMX YUH1−Ubal 10.82WDT PfUCHL3−UbVME 7.3, 9.63IFW UCHL1 S18Y−UbVME 8.73KVF UCHL1 I93M−UbVME 12.33KW5 UCHL1−UbVME 13.1

USP Family3TMP DUBA−Ubal 8.32Y5B USP21−linear diUbal 7.8, 10.8, 6.61NBF HAUSP−Ubal 10.0, 10.72AYO USP14−Ubal 10.93MHS SAGA complex (UBP8)−Ubal 9.22HD5 USP2, Ub 9.02G45 IsoT, Ub 8.6, 8.92J7Q M48 USP−UbVME 9.9, 10.83V6E USP2, Ub variant 7.03V6C USP2, Ub variant 7.43IHP USP5, Ub covalent 9.6, 6.33MTN USP21, ubiquitin-based USP21-specific

inhibitor8.5

3IT3 USP21, Ub covalent 7.4, 7.5, 7.4, 7.43N3K USP8, covalent Ub-like variant 10.43NHE USP2a, Ub 7.42IBI USP2, Ub covalent 7.6

OTU Family4IUM arterivirus papain-like protease 2, Ub 7.43ZNH CCHF viral, Ub-propargyl 9.54I6L OTUB1, Ub 9.13PT2 viral OTU, Ub 10.74HXD Nairovirus viral OTU, Ub 8.3, 10.13BY4 OTU, Ub 6.03PRM CCHF viral OTU, Ub 9.5, 10.03PRP CCHF viral OTU, Ub 10.3, 10.93C0R OTU, Ub 8.44DHZ h/ceOTUB1-ubiquitin aldehyde-

UBC13∼Ub9.2, 8.8

4DHJ ceOTUB1 ubiquitin aldehydeUBC13∼Ub complex

11.4, 8.5, 6.9, 11.8, 9.8,10.1

4DDI OTUB1/UbcH5b∼Ub/Ub 7.5, 12.6, 12.6, 7.5,12.6, 7.5

4DDG OTUB1/UbcH5b∼Ub/Ub 11.3, 7.1, 8.03PHW OTU domain of CCHF virus, Ub 7.6, 7.2, 9.0, 5.8

MJD Family3O65 Ataxin-3-like, Ub 9.6, 11.4, 12.8, 11.52JRI Ataxin 3, Ub 12.3, 12.9

JAMM Family2ZNV AMSH-LP, Lys63-linked diubiquitin 9.0, 8.3, 11.9, 11.3

aMultiple distance entries refer to those in the other subunits of thecrystallographic asymmetric unit.



128

to adopt a compact structure.80 However, the terminalubiquitin, being less packed than the internal ones (packedfrom both sides), is more likely to fray and be susceptible toDUB cleavage for stereochemical reasons. Certain DUBs mayhave evolved a mechanism for grabbing onto those fraying endsand start disassembling chains from there. There may be otherDUBs that prefer internal ubiquitins, or the terminal ones onthe other extreme end of the chain, such as isopeptidase-T(USP5),81 and there may be some with no preference at all.The structure of AMSH-LP (a Lys63-linked chain-specificDUB) in complex with Lys63-linked diubiquitin shows thatLys63 on the distal ubiquitin is not engaged by the enzyme,suggesting it is unlikely to show any preference between theterminal and internal cleavage sites.82 This is consistent withthe structure of a Lys63-linked chain, which adopts a moreextended conformation in crystals and perhaps in solution aswell.83−85 Future structural studies should reveal more detailsexplaining exo and endo specificity seen in certain DUBs.The structural analysis, combined with MD simulation,

shows the contribution of the ULD in ubiquitin binding. Intheory, certain residues in TsUCH37’s ULD, missing in ourstructure, also appear to be correctly positioned for contacting

ubiquitin. Notably, the modeling study provides a possibleexplanation of why Glu265 and Asn272 are so strictlyconserved in UCH37 from different organisms, with virtuallyno exception. Contributing to ubiquitin binding, as suggestedby our modeling studies, may be one of the functionalconstraints underlying the conservation of the amino acids,although one cannot rule out whether binding to other proteinssuch as Rpn13 may be involved. It should be noted that Bap1, aUCH DUB mutated in several cancers, also features aULD.70,71,86 Like UCH37, Bap1 becomes activated uponbinding to a larger protein complex, demonstrated with theDrosophila orthologue, Calypso, binding to the polycombrepressor DUB complex.87 Interestingly, the putative ubiquitin-binding residues of the ULD of UCH37 are also conserved inBap1 (data not shown), suggesting a role in ubiquitin bindingfor Bap1’s ULD as well (in some Bap1 orthologues, theglutamate corresponding to TsUCH37’s Glu265 is replacedwith an aspartate). However, human Bap1 has a linker ofapproximately 300 amino acids separating the UCH domainand its ULD. It will be interesting to see how the ULDpositions itself to bind ubiquitin, if it does. Of more interest is

Figure 7. Sequence alignment of TsUCH37 and other homologues: human UCH37, S. pombe Uch2, and Plasmodium falciparum UCH54. Secondarystructures for the two TsUCH37 structures are annotated above (e.g., α1, α-helix 1; β1, β-sheet 1; η1, 310-helix 1). α2′ and η2 are not resolved in theTsUCH37ΔC46−UbVME structure, and helices α7 and α8 are not present in the TsUCH37cat−UbVME construct.



129

knowing whether the ULD has independent ability to bind toubiquitin.

■ ASSOCIATED CONTENT*S Supporting InformationSupporting figures and NEDD8 hydrolysis data (Figure S4).This material is available free of charge via the Internet athttp://pubs.acs.org.Accession CodesCoordinates and structure factors have been deposited in theProtein Data Bank as entries 4I6N and 4IG7.

■ AUTHOR INFORMATIONCorresponding Author*Brown Laboratory of Chemistry, 560 Oval Dr., WestLafayette, IN 47907. E-mail: [email protected]. Phone: (765)494-5478. Fax: (765) 494-0239.Author ContributionsM.E.M. and M.-I.K. contributed equally to this work.FundingFinancial support from the National Institutes of Health(1R01RR026273, C.D.) is gratefully acknowledged. TheGeneral Medicine and Cancer Institutes Collaborative AccessTeam (GM/CA CAT) of the Advanced Photon Source atArgonne National Laboratory has been funded in whole or inpart with Federal funds from the National Cancer Institute (Y1-CO-1020) and the National Institute of General MedicalSciences (Y1-GM-1104).NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSWe acknowledge Venugopalan Nagarajan, Ruslan Sanishvili,and Craig Ogata at beamline 23-ID-B of the Advanced PhotonSource for assistance with data collection. Use of the AdvancedPhoton Source was supported by the U.S. Department ofEnergy, Basic Energy Sciences, Office of Science, underContract DE-AC02-06CH11357. We thank Emma DeWaltand Garth Simpson from the Department of Chemistry, PurdueUniversity, for their assistance with SONICC imaging of initialcrystals.

■ ABBREVIATIONSSDS−PAGE, sodium dodecyl sulfate−polyacrylamide gelelectrophoresis; DTT, dithiothreitol; IPTG, isopropyl β-D-1-thiogalactopyranoside; UbVME, ubiquitin vinyl methyl ester;UbAMC, ubiquitin aminomethylcoumarin; UCH37, ubiquitincarboxyl-terminal hydrolase 37; DUB, deubiquitinating enzymeor deubiquitinase; PDB, Protein Data Bank.

■ REFERENCES(1) Ciechanover, A. (2005) Proteolysis: From the lysosome toubiquitin and the proteasome. Nat. Rev. Mol. Cell Biol. 6, 79−87.(2) Ciechanover, A., and Schwartz, A. L. (2002) Ubiquitin-mediateddegradation of cellular proteins in health and disease. Hepatology 35,3−6.(3) Varshavsky, A. (1997) The ubiquitin system. Trends Biochem. Sci.22, 383−387.(4) Wilkinson, K. D. (2000) Ubiquitination and deubiquitination:Targeting of proteins for degradation by the proteasome. Semin. CellDev. Biol. 11, 141−148.(5) Goldberg, A. L. (2003) Protein degradation and protectionagainst misfolded or damaged proteins. Nature 426, 895−899.

(6) Pickart, C. M., and Cohen, R. E. (2004) Proteasomes and theirkin: Proteases in the machine age. Nat. Rev. Mol. Cell Biol. 5, 177−187.(7) Finley, D. (2009) Recognition and processing of ubiquitin-protein conjugates by the proteasome. Annu. Rev. Biochem. 78, 477−513.(8) Baumeister, W., Walz, J., Zuhl, F., and Seemuller, E. (1998) Theproteasome: Paradigm of a self-compartmentalizing protease. Cell 92,367−380.(9) Matyskiela, M. E., and Martin, A. (2012) Design principles of auniversal protein degradation machine. J. Mol. Biol. 425, 199−213.(10) Lander, G. C., Estrin, E., Matyskiela, M. E., Bashore, C.,Nogales, E., and Martin, A. (2012) Complete subunit architecture ofthe proteasome regulatory particle. Nature 482, 186−191.(11) Lasker, K., Forster, F., Bohn, S., Walzthoeni, T., Villa, E.,Unverdorben, P., Beck, F., Aebersold, R., Sali, A., and Baumeister, W.(2012) Molecular architecture of the 26S proteasome holocomplexdetermined by an integrative approach. Proc. Natl. Acad. Sci. U.S.A.109, 1380−1387.(12) Goldberg, A. L. (2007) Functions of the proteasome: Fromprotein degradation and immune surveillance to cancer therapy.Biochem. Soc. Trans. 35, 12−17.(13) Demartino, G. N., and Gillette, T. G. (2007) Proteasomes:Machines for all reasons. Cell 129, 659−662.(14) Guterman, A., and Glickman, M. H. (2004) Deubiquitinatingenzymes are IN/(trinsic to proteasome function). Curr. Protein Pept.Sci. 5, 201−211.(15) Yao, T., and Cohen, R. E. (2002) A cryptic protease couplesdeubiquitination and degradation by the proteasome. Nature 419,403−407.(16) Verma, R., Aravind, L., Oania, R., McDonald, W. H., Yates, J. R.,III, Koonin, E. V., and Deshaies, R. J. (2002) Role of Rpn11metalloprotease in deubiquitination and degradation by the 26Sproteasome. Science 298, 611−615.(17) Lee, M. J., Lee, B. H., Hanna, J., King, R. W., and Finley, D.(2011) Trimming of ubiquitin chains by proteasome-associateddeubiquitinating enzymes. Mol Cell Proteomics. 10, R110.003871.(18) Pickart, C. M. (2001) Mechanisms underlying ubiquitination.Annu. Rev. Biochem. 70, 503−533.(19) Schulman, B. A. (2011) Twists and turns in ubiquitin-likeprotein conjugation cascades. Protein Sci. 20, 1941−1954.(20) Pickart, C. M. (2000) Ubiquitin in chains. Trends Biochem. Sci.25, 544−548.(21) Haglund, K., and Dikic, I. (2005) Ubiquitylation and cellsignaling. EMBO J. 24, 3353−3359.(22) Ikeda, F., and Dikic, I. (2008) Atypical ubiquitin chains: Newmolecular signals. ‘Protein Modifications: Beyond the Usual Suspects’review series. EMBO Rep. 9, 536−542.(23) Fushman, D., and Wilkinson, K. D. (2011) Structure andrecognition of polyubiquitin chains of different lengths and linkage.F1000 Biol. Rep. 3, 26.(24) Komander, D., and Rape, M. (2012) The ubiquitin code. Annu.Rev. Biochem. 81, 203−229.(25) Kulathu, Y., and Komander, D. (2012) Atypical ubiquitylation:The unexplored world of polyubiquitin beyond Lys48 and Lys63linkages. Nat. Rev. Mol. Cell Biol. 13, 508−523.(26) Chen, Z. J., and Sun, L. J. (2009) Nonproteolytic functions ofubiquitin in cell signaling. Mol. Cell 33, 275−286.(27) Komander, D., Clague, M. J., and Urbe, S. (2009) Breaking thechains: Structure and function of the deubiquitinases. Nat. Rev. Mol.Cell Biol. 10, 550−563.(28) Komander, D. (2010) Mechanism, specificity and structure ofthe deubiquitinases. Subcell. Biochem. 54, 69−87.(29) Amerik, A. Y., and Hochstrasser, M. (2004) Mechanism andfunction of deubiquitinating enzymes. Biochim. Biophys. Acta 1695,189−207.(30) Wilkinson, K. D. (1997) Regulation of ubiquitin-dependentprocesses by deubiquitinating enzymes. FASEB J. 11, 1245−1256.(31) Wilkinson, K. D. (2009) DUBs at a glance. J. Cell Sci. 122,2325−2329.



130

(32) Nijman, S. M., Luna-Vargas, M. P., Velds, A., Brummelkamp, T.R., Dirac, A. M., Sixma, T. K., and Bernards, R. (2005) A genomic andfunctional inventory of deubiquitinating enzymes. Cell 123, 773−786.(33) Love, K. R., Catic, A., Schlieker, C., and Ploegh, H. L. (2007)Mechanisms, biology and inhibitors of deubiquitinating enzymes. Nat.Chem. Biol. 3, 697−705.(34) Reyes-Turcu, F. E., Ventii, K. H., and Wilkinson, K. D. (2009)Regulation and cellular roles of ubiquitin-specific deubiquitinatingenzymes. Annu. Rev. Biochem. 78, 363−397.(35) Tsou, W. L., Sheedlo, M. J., Morrow, M. E., Blount, J. R.,McGregor, K. M., Das, C., and Todi, S. V. (2012) Systematic analysisof the physiological importance of deubiquitinating enzymes. PLoSOne 7, e43112.(36) Borodovsky, A., Kessler, B. M., Casagrande, R., Overkleeft, H. S.,Wilkinson, K. D., and Ploegh, H. L. (2001) A novel active site-directedprobe specific for deubiquitylating enzymes reveals proteasomeassociation of USP14. EMBO J. 20, 5187−5196.(37) Yao, T., Song, L., Xu, W., DeMartino, G. N., Florens, L.,Swanson, S. K., Washburn, M. P., Conaway, R. C., Conaway, J. W., andCohen, R. E. (2006) Proteasome recruitment and activation of theUch37 deubiquitinating enzyme by Adrm1. Nat. Cell Biol. 8, 994−1002.(38) Lam, Y. A., Xu, W., DeMartino, G. N., and Cohen, R. E. (1997)Editing of ubiquitin conjugates by an isopeptidase in the 26Sproteasome. Nature 385, 737−740.(39) Hanna, J., Hathaway, N. A., Tone, Y., Crosas, B., Elsasser, S.,Kirkpatrick, D. S., Leggett, D. S., Gygi, S. P., King, R. W., and Finley,D. (2006) Deubiquitinating enzyme Ubp6 functions noncatalyticallyto delay proteasomal degradation. Cell 127, 99−111.(40) Hu, M., Li, P., Song, L., Jeffrey, P. D., Chenova, T. A.,Wilkinson, K. D., Cohen, R. E., and Shi, Y. (2005) Structure andmechanisms of the proteasome-associated deubiquitinating enzymeUSP14. EMBO J. 24, 3747−3756.(41) Koulich, E., Li, X., and DeMartino, G. N. (2008) Relativestructural and functional roles of multiple deubiquitylating proteinsassociated with mammalian 26S proteasome. Mol. Biol. Cell 19, 1072−1082.(42) Lee, B. H., Lee, M. J., Park, S., Oh, D. C., Elsasser, S., Chen, P.C., Gartner, C., Dimova, N., Hanna, J., Gygi, S. P., Wilson, S. M., King,R. W., and Finley, D. (2010) Enhancement of proteasome activity by asmall-molecule inhibitor of USP14. Nature 467, 179−184.(43) D’Arcy, P., Brnjic, S., Olofsson, M. H., Fryknas, M., Lindsten, K.,De Cesare, M., Perego, P., Sadeghi, B., Hassan, M., Larsson, R., andLinder, S. (2011) Inhibition of proteasome deubiquitinating activity asa new cancer therapy. Nat. Med. 17, 1636−1640.(44) D’Arcy, P., and Linder, S. (2012) Proteasome deubiquitinases asnovel targets for cancer therapy. Int. J. Biochem. Cell Biol. 44, 1729−1738.(45) Lam, Y. A., DeMartino, G. N., Pickart, C. M., and Cohen, R. E.(1997) Specificity of the ubiquitin isopeptidase in the PA700regulatory complex of 26 S proteasomes. J. Biol. Chem. 272, 28438−28446.(46) Stone, M., Hartmann-Petersen, R., Seeger, M., Bech-Otschir, D.,Wallace, M., and Gordon, C. (2004) Uch2/Uch37 is the majordeubiquitinating enzyme associated with the 26S proteasome in fissionyeast. J. Mol. Biol. 344, 697−706.(47) Larsen, C. N., Krantz, B. A., and Wilkinson, K. D. (1998)Substrate specificity of deubiquitinating enzymes: Ubiquitin C-terminal hydrolases. Biochemistry 37, 3358−3368.(48) Misaghi, S., Galardy, P. J., Meester, W. J., Ovaa, H., Ploegh, H.L., and Gaudet, R. (2005) Structure of the ubiquitin hydrolase UCH-L3 complexed with a suicide substrate. J. Biol. Chem. 280, 1512−1520.(49) Das, C., Hoang, Q. Q., Kreinbring, C. A., Luchansky, S. J.,Meray, R. K., Ray, S. S., Lansbury, P. T., Ringe, D., and Petsko, G. A.(2006) Structural basis for conformational plasticity of the Parkinson’sdisease-associated ubiquitin hydrolase UCH-L1. Proc. Natl. Acad. Sci.U.S.A. 103, 4675−4680.

(50) Johnston, S. C., Larsen, C. N., Cook, W. J., Wilkinson, K. D.,and Hill, C. P. (1997) Crystal structure of a deubiquitinating enzyme(human UCH-L3) at 1.8 Å resolution. EMBO J. 16, 3787−3796.(51) Hamazaki, J., Iemura, S., Natsume, T., Yashiroda, H., Tanaka, K.,and Murata, S. (2006) A novel proteasome interacting protein recruitsthe deubiquitinating enzyme UCH37 to 26S proteasomes. EMBO J.25, 4524−4536.(52) Husnjak, K., Elsasser, S., Zhang, N., Chen, X., Randles, L., Shi,Y., Hofmann, K., Walters, K. J., Finley, D., and Dikic, I. (2008)Proteasome subunit Rpn13 is a novel ubiquitin receptor. Nature 453,481−488.(53) Schreiner, P., Chen, X., Husnjak, K., Randles, L., Zhang, N.,Elsasser, S., Finley, D., Dikic, I., Walters, K. J., and Groll, M. (2008)Ubiquitin docking at the proteasome through a novel pleckstrin-homology domain interaction. Nature 453, 548−552.(54) Qiu, X. B., Ouyang, S. Y., Li, C. J., Miao, S., Wang, L., andGoldberg, A. L. (2006) hRpn13/ADRM1/GP110 is a novelproteasome subunit that binds the deubiquitinating enzyme,UCH37. EMBO J. 25, 5742−5753.(55) Yao, T., Song, L., Jin, J., Cai, Y., Takahashi, H., Swanson, S. K.,Washburn, M. P., Florens, L., Conaway, R. C., Cohen, R. E., andConaway, J. W. (2008) Distinct modes of regulation of the Uch37deubiquitinating enzyme in the proteasome and in the Ino80chromatin-remodeling complex. Mol. Cell 31, 909−917.(56) Nishio, K., Kim, S. W., Kawai, K., Mizushima, T., Yamane, T.,Hamazaki, J., Murata, S., Tanaka, K., and Morimoto, Y. (2009) Crystalstructure of the de-ubiquitinating enzyme UCH37 (human UCH-L5)catalytic domain. Biochem. Biophys. Res. Commun. 390, 855−860.(57) Burgie, S. E., Bingman, C. A., Soni, A. B., and Phillips, G. N.(2012) Structural characterization of human Uch37. Proteins 80, 649−654.(58) Maiti, T. K., Permaul, M., Boudreaux, D. A., Mahanic, C.,Mauney, S., and Das, C. (2011) Crystal structure of the catalyticdomain of UCHL5, a proteasome-associated human deubiquitinatingenzyme, reveals an unproductive form of the enzyme. FEBS J. 278,4917−4926.(59) White, R. R., Miyata, S., Papa, E., Spooner, E., Gounaris, K.,Selkirk, M. E., and Artavanis-Tsakonas, K. (2011) Characterisation ofthe Trichinella spiralis deubiquitinating enzyme, TsUCH37, anevolutionarily conserved proteasome interaction partner. PLoSNeglected Trop. Dis. 5, e1340.(60) Borodovsky, A., Ovaa, H., Kolli, N., Gan-Erdene, T., Wilkinson,K. D., Ploegh, H. L., and Kessler, B. M. (2002) Chemistry-basedfunctional proteomics reveals novel members of the deubiquitinatingenzyme family. Chem. Biol. 9, 1149−1159.(61) Holyoak, T., Fenn, T. D., Wilson, M. A., Moulin, A. G., Ringe,D., and Petsko, G. A. (2003) Malonate: A versatile cryoprotectant andstabilizing solution for salt-grown macromolecular crystals. ActaCrystallogr. D59, 2356−2358.(62) Otwinowski, Z., and Minor, W. (1997) Processing of X-rayDiffraction Data Collected in Oscillation Mode.Methods Enzymol. 276,307−326.(63) Adams, P. D., Afonine, P. V., Bunkoczi, G., Chen, V. B., Davis, I.W., Echols, N., Headd, J. J., Hung, L.-W., Kapral, G. J., and Grosse-Kunstleve, R. W. (2010) PHENIX: A comprehensive Python-basedsystem for macromolecular structure solution. Acta Crystallogr. D66,213−221.(64) Emsley, P., and Cowtan, K. (2004) Coot: Model-building toolsfor molecular graphics. Acta Crystallogr. D60, 2126−2132.(65) Schuck, P. (2000) Size-distribution analysis of macromoleculesby sedimentation velocity ultracentrifugation and lamm equationmodeling. Biophys. J. 78, 1606−1619.(66) Arnold, K., Bordoli, L., Kopp, J., and Schwede, T. (2006) TheSWISS-MODEL workspace: A web-based environment for proteinstructure homology modelling. Bioinformatics 22, 195−201.(67) Case, D., Darden, T., Cheatham, T., III, Simmerling, C., Wang,J., Duke, R., Luo, R., Walker, R., Zhang, W., and Merz, K. (2012)AMBER 12, University of California, San Francisco.



131

(68) Darden, T., York, D., and Pedersen, L. (1993) Particle meshEwald: An N·log (N) method for Ewald sums in large systems. J.Chem. Phys. 98, 10089.(69) Ryckaert, J.-P., Ciccotti, G., and Berendsen, H. J. (1977)Numerical integration of the cartesian equations of motion of a systemwith constraints: Molecular dynamics of n-alkanes. J. Comput. Phys. 23,327−341.(70) Misaghi, S., Ottosen, S., Izrael-Tomasevic, A., Arnott, D.,Lamkanfi, M., Lee, J., Liu, J., O’Rourke, K., Dixit, V. M., and Wilson, A.C. (2009) Association of C-terminal ubiquitin hydrolase BRCA1-associated protein 1 with cell cycle regulator host cell factor 1. Mol.Cell. Biol. 29, 2181−2192.(71) Sanchez-Pulido, L., Kong, L., and Ponting, C. P. (2012) Acommon ancestry for BAP1 and Uch37 regulators. Bioinformatics 28,1953−1956.(72) Chen, V., Arendall, W., Headd, J., Keedy, D., Immormino, R.,Kapral, G., Murray, L., Richardson, J., and Richardson, D. (2010)MolProbity: All-atom structure validation for macromolecularcrystallography. Acta Crystallogr. D66, 12−21.(73) Johnston, S. C., Riddle, S. M., Cohen, R. E., and Hill, C. P.(1999) Structural basis for the specificity of ubiquitin C-terminalhydrolases. EMBO J. 18, 3877−3887.(74) Zhou, Z. R., Zhang, Y. H., Liu, S., Song, A. X., and Hu, H. Y.(2012) Length of the active-site crossover loop defines the substratespecificity of ubiquitin C-terminal hydrolases for ubiquitin chains.Biochem. J. 441, 143−149.(75) Boudreaux, D. A., Chaney, J., Maiti, T. K., and Das, C. (2012)Contribution of active site glutamine to rate enhancement in ubiquitinC-terminal hydrolases. FEBS J. 279, 1106−1118.(76) Boudreaux, D. A., Maiti, T. K., Davies, C. W., and Das, C.(2010) Ubiquitin vinyl methyl ester binding orients the misalignedactive site of the ubiquitin hydrolase UCHL1 into productiveconformation. Proc. Natl. Acad. Sci. U.S.A. 107, 9117−9122.(77) Popp, M. W., Artavanis-Tsakonas, K., and Ploegh, H. L. (2009)Substrate filtering by the active site crossover loop in UCHL3 revealedby sortagging and gain-of-function mutations. J. Biol. Chem. 284,3593−3602.(78) Artavanis-Tsakonas, K., Weihofen, W. A., Antos, J. M., Coleman,B. I., Comeaux, C. A., Duraisingh, M. T., Gaudet, R., and Ploegh, H. L.(2010) Characterization and structural studies of the Plasmodiumfalciparum ubiquitin and Nedd8 hydrolase UCHL3. J. Biol. Chem. 285,6857−6866.(79) Hu, M., Li, P., Li, M., Li, W., Yao, T., Wu, J. W., Gu, W., Cohen,R. E., and Shi, Y. (2002) Crystal structure of a UBP-familydeubiquitinating enzyme in isolation and in complex with ubiquitinaldehyde. Cell 111, 1041−1054.(80) Eddins, M. J., Varadan, R., Fushman, D., Pickart, C. M., andWolberger, C. (2007) Crystal structure and solution NMR studies ofLys48-linked tetraubiquitin at neutral pH. J. Mol. Biol. 367, 204−211.(81) Reyes-Turcu, F. E., Shanks, J. R., Komander, D., and Wilkinson,K. D. (2008) Recognition of polyubiquitin isoforms by the multipleubiquitin binding modules of isopeptidase T. J. Biol. Chem. 283,19581−19592.(82) Sato, Y., Yoshikawa, A., Yamagata, A., Mimura, H., Yamashita,M., Ookata, K., Nureki, O., Iwai, K., Komada, M., and Fukai, S. (2008)Structural basis for specific cleavage of Lys 63-linked polyubiquitinchains. Nature 455, 358−362.(83) Datta, A. B., Hura, G. L., and Wolberger, C. (2009) Thestructure and conformation of Lys63-linked tetraubiquitin. J. Mol. Biol.392, 1117−1124.(84) Komander, D., Reyes-Turcu, F., Licchesi, J. D., Odenwaelder, P.,Wilkinson, K. D., and Barford, D. (2009) Molecular discrimination ofstructurally equivalent Lys 63-linked and linear polyubiquitin chains.EMBO Rep. 10, 466−473.(85) Tenno, T., Fujiwara, K., Tochio, H., Iwai, K., Morita, E. H.,Hayashi, H., Murata, S., Hiroaki, H., Sato, M., Tanaka, K., andShirakawa, M. (2004) Structural basis for distinct roles of Lys63- andLys48-linked polyubiquitin chains. Genes Cells 9, 865−875.

(86) Ventii, K. H., Devi, N. S., Friedrich, K. L., Chernova, T. A.,Tighiouart, M., Van Meir, E. G., and Wilkinson, K. D. (2008) BRCA1-associated protein-1 is a tumor suppressor that requires deubiquitinat-ing activity and nuclear localization. Cancer Res. 68, 6953−6962.(87) Scheuermann, J. C., de Ayala Alonso, A. G., Oktaba, K., Ly-Hartig, N., McGinty, R. K., Fraterman, S., Wilm, M., Muir, T. W., andMuller, J. (2010) Histone H2A deubiquitinase activity of thePolycomb repressive complex PR-DUB. Nature 465, 243−247.



132

research papers

J. Synchrotron Rad. (2013). 20, 531–540 doi:10.1107/S0909049513007942 531

Journal of

SynchrotronRadiation

ISSN 0909-0495

Received 12 October 2012

Accepted 22 March 2013

# 2013 International Union of Crystallography

Printed in Singapore – all rights reserved

Integrated nonlinear optical imaging microscopefor on-axis crystal detection and centering at asynchrotron beamline

Jeremy T. Madden,a Scott J. Toth,a Christopher M. Dettmar,a Justin A. Newman,a

Robert A. Oglesbee,a Hartmut G. Hedderich,a R. Michael Everly,a Michael Becker,b

Judith A. Ronau,a Susan K. Buchanan,c Vadim Cherezov,d Marie E. Morrow,a

Shenglan Xu,b Dale Ferguson,b Oleg Makarov,b Chittaranjan Das,a

Robert Fischettib and Garth J. Simpsona*

aDepartment of Chemistry, Purdue University, 560 Oval Drive, West Lafayette, IN 47906, USA,bGM/CA@APS, Advanced Photon Source, Argonne National Laboratory, 9700 South Cass Avenue,

Argonne, IL 60439, USA, cNIDDK, National Institutes of Health, Building 50, Room 4503, 50 South

Drive, Bethesda, MD 20814, USA, and dDepartment of Molecular Biology, The Scripps Research

Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037, USA. E-mail: [email protected]

Nonlinear optical (NLO) instrumentation has been integrated with synchrotron

X-ray diffraction (XRD) for combined single-platform analysis, initially

targeting applications for automated crystal centering. Second-harmonic-

generation microscopy and two-photon-excited ultraviolet fluorescence micro-

scopy were evaluated for crystal detection and assessed by X-ray raster

scanning. Two optical designs were constructed and characterized; one

positioned downstream of the sample and one integrated into the upstream

optical path of the diffractometer. Both instruments enabled protein crystal

identification with integration times between 80 and 150 ms per pixel,

representing a �103–104-fold reduction in the per-pixel exposure time relative

to X-ray raster scanning. Quantitative centering and analysis of phenylalanine

hydroxylase from Chromobacterium violaceum cPAH, Trichinella spiralis

deubiquitinating enzyme TsUCH37, human �-opioid receptor complex kOR-

T4L produced in lipidic cubic phase (LCP), intimin prepared in LCP, and �-cellulose samples were performed by collecting multiple NLO images. The

crystalline samples were characterized by single-crystal diffraction patterns,

while �-cellulose was characterized by fiber diffraction. Good agreement was

observed between the sample positions identified by NLO and XRD raster

measurements for all samples studied.

Keywords: XRD; NLO; SHG; SONICC; centering; protein; TPE-UVF; microscopy;LCP; two-photon.

1. Introduction

The high photon flux and energy tunability of synchrotron

radiation sources have made them indispensable tools for

X-ray analysis, with applications spanning protein structure

determination through materials science and nanotechnology

(Rasmussen et al., 2011; Moukhametzianov et al., 2008; Bates

et al., 2006; Berger et al., 2010; Dauter, 2006; Ihee et al., 2010;

le Maire et al., 2011; Parker et al., 2006; Riekel et al., 2005).

The increasing drive toward tighter focusing has enabled

structure determination on ever-smaller crystals and sub-

domains within materials, but presents growing challenges for

reliable crystal centering. These challenges are particularly

relevant for protein crystal diffraction, in which the drive

toward fully automated X-ray diffraction analysis at

synchrotron sources has introduced bottlenecks in sample

positioning (Andrey et al., 2004; Moukhametzianov et al., 2008;

Pothineni et al., 2006; Aishima et al., 2010; Cherezov et al.,

2009; Stepanov et al., 2011a). Diffraction-quality protein

crystals are typically obtained through crystallization screen-

ings, followed by optimization, and then are placed into cryo-

loops, which are flash-cooled in liquid nitrogen to reduce

X-ray damage and aid in sample handling (Dobrianov et al.,

1999; Karain et al., 2002). High-throughput methods for

automated crystal positioning are frustrated by complications

of reliable centering of smaller and smaller protein crystals

within more complex and turbid matrices. The current most

reliable methods for crystal centering involve rastering the

133

sample using a focused X-ray beam (Accardo et al., 2010;

Hilgart et al., 2011; Cherezov et al., 2009; Stepanov et al., 2011a;

Aishima et al., 2010; Song et al., 2007). From the resulting

X-ray diffraction images recorded as a function of sample

position in the beam, protein crystals are centered based on

the locations of strongest Bragg-like diffraction. X-ray fluor-

escence raster is also relatively fast, but it requires a conve-

nient X-ray fluorescent element to be present in the crystal

(Stepanov et al., 2011a).

While generally successful, X-ray raster scanning suffers

from several limitations. First, the method is relatively slow,

often utilizing >2 s per pixel (raster cell), corresponding to

analysis times from several minutes up to an hour depending

on the number of cells in the raster grid and on the exposure

time (Aishima et al., 2010). Rastering is commonly performed

first with a coarse grid, and then a finer grid, to minimize the

number of cells, and to increase speed. The total pixel number

is in turn dependent on the size of the X-ray beam, the speed

of the detector and analysis, as well as the scanned size of the

cryo-loop and the crystal itself (Cherezov et al., 2009; Song

et al., 2007). Recent advances in diffraction image read times

using single-photon-counting arrays (pixel array detectors)

(Broennimann et al., 2006), allowing integration times as low

as 2 ms per image (Aishima et al., 2010), can significantly

reduce the time frame for raster scanning measurements.

However, the time required for raster scanning will still ulti-

mately be limited by the collective times required to obtain

sufficient signal to noise (S/N) in a given pixel, to translate the

sample through the X-ray source, and to reconstruct the

crystal positions based on automated analysis of the compiled

diffraction images. Diffraction is a relatively inefficient process

with far more X-ray photons absorbed or inelastically scat-

tered than detected for diffraction analysis, contributing to

sample damage, even under the cryogenic conditions typically

utilized. With small crystals or beams, incident X-ray inten-

sities must be increased accordingly to achieve diffracted

intensities equivalent to those for large crystals, thereby

increasing absorbed X-ray dose and exacerbating damage.

Alternative methods for automated loop centering based on

optical imaging include bright-field image analysis and ultra-

violet fluorescence (UVF) microscopy, which takes advantage

of intrinsic fluorescent properties of protein crystals (Jain &

Stojanoff, 2007; Vernede et al., 2006; Pohl et al., 2004; Andrey

et al., 2004; Pothineni et al., 2006). However, algorithms for

protein crystal centering (e.g. based on crystal edge-finding

algorithms) are error-prone for microcrystals and turbid

matrices, such as lipidic cubic phase (LCP). Methods opti-

mized for analysis within the mother liquor often prove

unreliable for a loop-mounted crystal, in part because algo-

rithms often cannot easily distinguish between the loop,

features in the cryo-cooled mother-liquor and the crystal.

Furthermore, both bright-field and UVF imaging are chal-

lenging to reliably implement in turbid matrices, where optical

scattering frustrates reliable crystal imaging. UVF also has

a potential disadvantage of inducing UV photodamage to

samples from long exposures, or in highly labile proteins, but

the exposure times required for imaging are typically short

enough to minimize such effects (Vernede et al., 2006; Chen et

al., 2009; Nanao & Ravelli, 2006).

More recently, nonlinear optical imaging (NLO) methods

such as second-harmonic generation (SHG) and two-photon-

excited UV fluorescence (TPE-UVF) have emerged as viable

alternatives for high-contrast crystal visualization (Kissick et

al., 2010; Madden et al., 2011). SHG, or the frequency doubling

of light, is symmetry forbidden in disordered media (e.g.

amorphous protein aggregates or proteins in solution) but is

allowed for certain classes of crystals (Haupert & Simpson,

2011). Fortuitously, the chirality intrinsic to proteins typically

results in the adoption of SHG-active crystal classes. Recent

quantum chemical calculations suggest an SHG coverage of

approximately 84% of protein crystals in the Protein Crystal

Database using an optimized instrument (Haupert et al.,

2012). TPE-UVF provides a complimentary method to SHG

for protein crystal detection, with contrast dependent on the

presence of aromatic side-chains (primarily tryptophan),

independent of crystallinity. Crystals that are weakly active to

SHG imaging but contain fluorescent amino acid residues can

be detected (Madden et al., 2011). Furthermore, TPE-UVF

can aid in distinguishing SHG-active small-molecule and salt

crystals from protein crystals.

The high selectivity for crystals and negligible background

from disordered protein aggregates typically produces high-

contrast SHG images, which are highly compatible with

automated image analysis algorithms designed for protein

crystal detection and centering (Haupert & Simpson, 2011).

SHG measurements have recently enabled crystal detection

for diffraction centering using off-line instrumentation

(Kissick et al., 2013), in which protein crystals were first

imaged under cryogenic conditions with an SHG microscope,

and then manually compared with diffraction images obtained

by X-ray raster scanning with good agreement. A major

benefit of NLO instruments is the reduction in time required

to determine crystal locations with high contrast, as

measurements for an entire loop can be obtained in as little as

a few seconds, compared with tens of minutes routinely

required for X-ray raster imaging. The spatial resolution of

NLO instruments is also high (�1–2 mm), whereas X-ray

diffraction (XRD) rastering with this type of resolution would

take substantially longer to scan an area equivalent to that of

the entire NLO image (>72 h at 1 s per pixel for a 512 � 512

pixel image). Furthermore, reducing the reliance on X-ray

raster imaging would minimize X-ray-induced sample damage

(Hilgart et al., 2011; Ravelli & Garman, 2006).

By integrating SHG and TPE-UVF imaging directly into a

synchrotron X-ray diffraction beamline, the robotic controls,

automated positioning capabilities, cryogenics and other

beamline utilities of high-throughput synchrotron facilities can

be leveraged. However, the spatial constraints of a typical

synchrotron X-ray experimental hutch represent a nontrivial

hurdle for development of compatible NLO instrumentation.

Typical research NLO instruments occupy a large footprint

(an optical table approximately 120 cm � 300 cm), far greater

than the space available on a typical beamline. In this

work, two complementary prototypes for an on-line compa-

research papers

532 Jeremy T. Madden et al. � Integrated nonlinear optical imaging microscope J. Synchrotron Rad. (2013). 20, 531–540

134

tible instrument combining synchrotron

XRD and NLO imaging are described.

Assessment of these systems was

performed by direct comparisons

between NLO images and those

obtained by X-ray diffraction rastering.

2. Experimental methods

Two separate instruments were

designed and constructed for inte-

grating XRD and NLO imaging, each

with its own advantages and limitations.

The upstream version introduced the

incident light coaxial and parallel with

the direction of the X-ray beam path,

while the downstream system was

coaxial and anti-parallel. The upstream

version was designed to fully integrate

with the existing optical path, while the

downstream version was optimized for

high flexibility and compatibility with

diverse beamline configurations. Both

systems were rated as Class I laser

systems on-site, with enclosed beam

paths, shutters and interlocks to ensure

no exposed collimated optical radiation.

The integrated NLO microscopes were

installed at beamlines 23-ID-B and 23-

ID-D at the Advanced Photon Source

(APS) at Argonne National Laboratory

in Argonne, IL, USA. A basic schematic

of the instruments and beam paths as they were installed on

the synchrotron beamline can be seen in Fig. 1. Detailed

descriptions and photographs are provided.

2.1. Integrated nonlinear optical microscope designs

The upstream illumination NLO system was designed to sit

above the existing instrumentation at GM/CA beamline 23-

ID-B at the APS, and couple directly into the existing optical

path. A Fianium FemtoPower 1060 ultrafast fiber laser was

utilized, producing �160 fs pulses centered around 1060 nm,

with a 50 MHz repetition rate, maximum power of 1.5 W,

allowing for a maximum power of �140 mW at the sample,

with 80% of the overall loss arising from the objective. The

Fianium source was composed of an oscillator coupled via a

1.5 m fiber to a dispersion compensator and free-space coupler

unit, with dimensions of approximately 15 cm� 13 cm� 8 cm.

A heated doubling crystal (Newlight Photonics Inc.,

SHG1663-IM, HTS 85141000) was permanently assembled in

the beam path, with the fundamental beam focused into the

crystal with a plano-convex lens ( f = 35 mm) and collimated

with another plano-convex lens ( f = 100 mm) after the

doubling crystal. The efficiency of SHG from the doubling

crystal was controlled by either introducing or removing a

1064 nm zero-order half-wave plate using a flip mount (New

Focus, 8892-K). The scanning assembly consisted of a

galvanometer mirror (Cambridge Technology, 6210H) and

resonant scanning mirror (Cambridge Technology, 1-003-

3002509), controlling the beam position on the horizontal

slow-scan and vertical fast-scan axes, respectively. The beam

was directed into a telocentric lens pair consisting of two

plano-convex lenses ( f = 75 mm and f = 250 mm) leading to an

additional 3.3� beam expansion after the scan head. The

incident light then reflected off a dichroic mirror stack

(Semrock, PBP01-529/23-25x36 and Chroma, 900dcsp)

designed to reflect 1060 nm and s-polarized 530 nm incident

light. The p-polarized component of the returning 530 nm light

was transmitted by this same dichroic for epi-detected SHG

(i.e. SHG detected in the backward direction through the same

objective as the incident light). High-reflectivity dichroic

mirrors for both 1060 nm and 530 nm light (Semrock, FF550-

Di01-25x36) delivered both wavelengths to the back aperture

of the 10� objective (Optem, 28-21-10), which was modified

with a �1.2 mm hole bored through the center to allow

X-ray access. In epi, the p-polarized SHG returning through

the dichroic mirror was passed through a bandpass filter

set (Chroma, HQ530/30m and CVI, 03FCG567/KG3) and into

a compact photomultiplier tube (PMT) module (Hamamatsu,

H10722-10). SHG and TPE-UVF were collected in the

transmission direction by a plano-convex lens ( f = 25.4 mm)

research papers

J. Synchrotron Rad. (2013). 20, 531–540 Jeremy T. Madden et al. � Integrated nonlinear optical imaging microscope 533

Figure 1(a) Schematic of the downstream NLO microscope; (b) schematic of the upstream NLOmicroscope; (c) close-up view of the downstream NLO microscope, with the solid arrowrepresenting incident laser propagation (red, 1060 nm) and dashed arrows representing thefrequency-doubled signal (green, SHG at 530 nm); (d) close-up view of the upstream NLOmicroscope, with solid arrows representing incident laser propagation (red, 1060 nm; green, 530 nm)and dashed arrows representing the measured signal (green, SHG at 530 nm; blue, TPE-UVF).

135

affixed to a right-angle prism using optical epoxy (Norland

Optical Adhesive 63). Another plano-convex lens ( f =

25.4 mm) coupled the detected light into a near-UV-compa-

tible liquid light guide (Oriel Instruments, 77554) collimated

with a plano-convex lens ( f = 25.4 mm) into the detection

assembly. Both the SHG and TPE-UVF were then reflected

off a primary dichroic beam splitter (Semrock, FF555-Di03-

25x36), then separated at a second dichroic beam splitter

(Chroma, z1064rdc-sp) for selective detection of SHG

(through Chroma, HQ530/30m and CVI, 03FCG567/KG3

filters) and TPE-UVF (through Semrock, SP01-532RS-25 and

FF01-440/SP-25 filters). Both the SHG and TPE-UVF were

focused onto the faces of the PMT modules (Hamamatsu,

H10722-10) by a plano-convex lens ( f = 60 mm) positioned

between the primary and secondary dichroic beam splitters.

Backlight illumination was achieved using an LED (ThorLabs,

MCWHL2) passing through the primary dichroic beam

splitter and into the liquid light guide. The illumination light

was then focused through the trans-SHG/TPE-UVF collection

optics and onto the sample.

The downstream NLO system was also designed with the

optical axis of the objective co-axial with the axis of X-ray

propagation [Figs. 1(a) and 1(b)], using a similar laser source.

The size constraints associated with this beamline, specifically

the restrictions imposed by the support structure of the

beamline and the area and instruments surrounding the

sample, limited the available footprint of the NLO system to

39 cm � 19 cm. The scanning assembly was composed of dual

galvanometers (Cambridge Technologies, 6210HSM40B),

mounted in a two-dimensional galvo 30 mm cage cube

(Thorlabs, GCM002), with each scanning mirror rotating

along either the x or y axis. With the scan head inducing a 90�

turn into the beam path, the incident light was directed

through a telocentric lens pair, mounted in a 30 mm cage cube,

and composed of an aspheric lens ( f = 10 mm) and a plano-

convex lens ( f = 50 mm), leading to a 5� beam expansion. The

incident light was then focused onto the sample by a long-

working-distance IR 10� objective (Mitutoyo, NT46-403)

generating SHG at 530 nm. Up to 650 mW of 1064 nm light

could be delivered to the sample with this system with the use

of the IR objective (compared with 140 mWwith the upstream

system). The SHG was detected in the epi-direction, collected

through the incident objective and reflected through a filter set

and onto a compact PMT module (Hamamatsu, H10722-10)

by a dichroic mirror (Omega Optical, 580DCLP) centered

around 532 nm and mounted in a rotatable kinematically

controlled cage cube platform. The SHG signal was detected

through a filter set composed of a KG3 (Thorlabs, FGS900)

and 530 nm filter (Chroma, z532/10x). Bright-field images

were also collected in the epi-direction using a module

composed of an aspheric lens ( f = 20 mm) and a CMOS

camera (Thorlabs, DCC1645C), manually inserted when

bright-field images were desired. Including the laser source,

the total footprint of the microscope was 25 cm � 15 cm �15 cm. The microscope was translated to the sample, at a

height of 1.4 m, to perform SHG detection and centering

measurements. The foundation of the microscope was a

high-precision long-travel translation stage (Newport,

M-IMS300V), and its electronics box (Newport, ESP 300,

three-axis motion controller), capable of translating the laser

pulse-compressor/output coupler, the microscope and the

support structure to and from the sample between X-ray

measurements, corresponding to approximately 20 cm of

travel, with an absolute accuracy of 2 mm.

The electronics package was designed and constructed in

collaboration with the Jonathan Amy Facility for Chemical

Instrumentation at Purdue University (JAFCI). The electro-

nics package integrated the electronics associated with the

microscope, including the power supplies, control boards and

data acquisition card (National Instruments), into a compact

housing for easy mounting and transport, with a footprint of

46 cm� 61 cm � 31 cm. Data were acquired as photon counts

using a gated multi-scalar card (Becker & Hickl, PMS-400a),

controlled using a custom-designed Labview program, which

was also written in collaboration with JAFCI. Data recon-

struction and imaging were completed through ImageJ (NIH,

2011).

2.2. X-ray raster scan scheme

XRD analysis and NLO images were acquired on all

samples studied on 23-ID-B. Diffraction of kOR-T4L was

acquired with a 5 mm-diameter X-ray beam, 5� 5 mm cell size,

12.0 keV X-ray beam, with 1 s exposure times, a photon flux

of 2.7 � 1010 photons s�1 (full unattenuated beam) and a

detector distance of 300 mm. Diffraction of TsUCH37 was

acquired with a 10 mm-diameter X-ray beam, a 10 � 10 mmcell, a photon flux of 1.3 � 1010 photons s�1 (10-fold

attenuation) and detector distance of 300 mm. Diffraction of

�-cellulose was acquired with a 10 mm-diameter X-ray beam,

a 10 � 10 mm X-ray beam with a photon flux of 2.7 �109 photons s�1 (50-fold attenuation) and detector distance of

300 mm. The resulting NLO images and XRD raster

measurements were compared using ImageJ and JBluIce

(Hilgart et al., 2011), which employs DISTL (Zhang et al.,

2006), to assess the degree of correlation of the sample posi-

tion within the loop. The boundaries of the raster grids and

raster cell sizes were defined using the software GUI JBluIce

(Stepanov et al., 2011b). Bragg candidates, which estimate the

number of well-ordered reflections, were generated for each

X-ray diffraction image; they are shown color-coded in the

figures as unsmoothed XRD raster images. The X-ray beam

size was adjusted using a mini-beam collimator (Fischetti et al.,

2009).

3. Sample materials

Phenylalanine hydroxylase from Chromobacterium violaceum

(cPAH) was purified as a glutathione s-transferase (GST)

fusion protein. The GST tag was cleaved with PreScission

protease (GE Biosciences). For crystallization, cPAH was

concentrated to 10 mg ml�1 in a solution of 5 mMHEPES, pH

7.4. Crystals of cPAH were obtained at ambient temperature

utilizing hanging-drop vapor diffusion from solution 43 of

research papers


136

Hampton Research’s PEG/Ion 2 screen [0.1 M Na-HEPES,

pH 7.0, 0.01 M magnesium chloride hexahydrate, 0.005 M

nickel (II) chloride hexahydrate and 15% w/v PEG 3350] with

8.3 mM hexammine cobalt (III) chloride and 8.3 mM guani-

dine hydrochloride as additives. Crystals were briefly soaked

in 25% ethylene glycol and then flash-cooled in liquid

nitrogen.

Crystals of human �-opioid receptor in complex with an

antagonist JDTic were obtained as described by Wu et al.

(2012). Briefly, the human �-opioid receptor sequence was

modified by fusing T4 lysozyme (T4L) into intracellular loop 3

(Gly261–Arg263), performing N/C-terminal truncations

(�Glu2Ala42, �Arg359Val380) and introducing a single

point mutation Ile1353.29Leu. The resulting construct kOR-

T4L was expressed in baculovirus infected sf9 insect cells.

Receptor was extracted from isolated membranes using

dodecylmaltoside/cholesterol hemisuccinate detergent

mixture, purified by metal-affinity chromatography, and

concentrated to 40 mg ml�1. Lipidic cubic phase crystal-

lization was performed as previously described (Caffrey &

Cherezov, 2009; Cherezov et al., 2004), by mixing protein

solution with 10% cholesterol in monoolein at 2/3 protein

solution/lipid ratio, and dispensing 50 nL protein laden LCP

boluses overlaid with 800 nL precipitant solutions in a 96-well

glass sandwich plate (Marienfeld) (Cherezov & Caffrey, 2003)

using a NT8-LCP crystallization robot (Formulatrix). Crystals

were obtained in 100 mM sodium citrate pH 5.8–6.4, 28–32%

(v/v) PEG 400, 350–450 mM potassium nitrate, and were

harvested directly from LCP matrix using MiTeGen micro-

mounts and flash-cooled in liquid nitrogen.

The catalytic domain of Trichinella spiralis deubiquitinating

enzyme UCH37 was expressed in E. coli as a GST-fused

construct, purified on a glutathione-agarose column,

complexed with ubiquitin vinyl methyl ester (UBVME), and

subsequently purified by ion-exchange chromatography.

Crystals of this complex, hereafter referred to simply as

TsUCH37-UbVME complex, were grown by hanging-drop

vapor diffusion in 3 M ammonium sulfate, 0.1 M bicine pH 9.0,

and 2 mM l-glutathione (mixture of reduced and oxidized)

over two days at room temperature.

The �-cellulose was prepared from pulpwood that under-

went both the Kraft process and subsequent mercerization

(Sixta et al., 2004; Takai & Colvin, 1978).

A construct encoding the membrane domain of E. coli

O157:H7 intimin was expressed, purified and crystallized as

described previously (Fairman et al., 2012). Briefly, Int208-449

was expressed in the outer membranes of E. coli BL21(DE3)

cells, extracted with the detergent Elugent (Calbiochem), and

purified by Ni-NTA affinity and anion-exchange chroma-

tography using buffers containing dodecyl maltoside

(Anatrace). Size-exclusion chromatography was used as a final

purification step and served to exchange the detergent to

lauryl dimethyl amine oxide (LDAO, Anatrace) using a buffer

containing 50 mM Tris-HCl, pH 7.5, 200 mM NaCl, 0.01%

NaN3 and 0.05% LDAO. The protein was concentrated to

20 mg ml�1, heptanetriol was added at 3% w/v, and the solu-

tion was mixed with monoolein at a 2/3 protein-to-lipid ratio.

A Mosquito LCP robot (TTP Labtech) was used to dispense

100 nL protein–lipid droplets, overlaid with 750 nL well

solutions. Intimin crystals grew from 100 mM sodium citrate,

pH 4.5–5.5, 50–100 mM NaCl, 100–150 mM MgCl2 and 30–

34% PEG 400. Crystals were mounted directly from the LCP

mixture and flash-cooled in liquid nitrogen.

4. Results and discussion

Data were acquired with both downstream and upstream

versions of the NLO instrument, and schematic representa-

tions along with photographs of the beam paths are shown

in Fig. 1.

Fig. 2 (acquired via the upstream system) shows a large

TsUCH37-UbVME crystal. Both the presence and position of

the crystal can be independently confirmed with bright-field

imaging (a), NLOmicroscopy and XRDmeasurements. Signal

intensities of the corresponding epi-SHG (b), transmission-

SHG (c) and TPE-UVF (d) were measured and processed in

ImageJ. Although the crystal is visible using conventional

optical imaging approaches, NLO microscopy produced

substantial improvements in contrast compared with bright-

field imaging. An X-ray diffraction raster was acquired (e) and

a representative diffraction image is shown ( f).

Intimin protein crystals in LCP were examined using the

upstream NLO system. In Fig. 3 the bright-field image is

shown in (a), with the corresponding trans-SHG image (b),

and X-ray raster acquired with a 5 � 5 mm beam, confirming

the presence of a protein crystal (c), with the spot having

greatest protein-like diffraction circled and the resulting

diffraction pattern provided (d). All protein crystals identified

by SHG and XRD were accurate for absolute position within

the resolution of the 5 mm X-ray beam.

In Fig. 4 (acquired via the upstream system) a bright-field

image of a kOR-T4L crystal within frozen lipidic cubic phase

is shown (a). As often arises with lipidic mesophase crystal-

lizations, the looped droplets exhibited high optical scattering

upon freezing that frustrated conventional bright-field

imaging approaches for crystal positioning. Transmission SHG

(b) and TPE-UVF (c) images were acquired, exhibiting loca-

lized areas (�2–5 mm) of signal within the loop, suggesting the

presence of a crystal. Crystals were confirmed via a 5 mm-

diameter X-ray beam and 5 � 5 mm cell X-ray raster scan (d),

in which several pixels exhibit weak, but detectable, diffrac-

tion with Bragg analysis consistent with the presence of a

protein crystal. Diffraction patterns for the brightest spot are

shown in Fig. 4(e). However, signal is observed in the trans-

SHG and TPE-UVF images that does not correspond to areas

of protein-like diffraction in the X-ray raster image. This

signal discrepancy is tentatively attributed to protein crystals

that are too small to produce Bragg peaks by XRD, or to the

presence of other ordered materials arising in a false positive.

False negatives for particular focal planes were also observed,

in which analysis of the diffraction patterns obtained from the

raster image indicates the presence of protein-like diffraction

located in areas that did not exhibit substantial SHG or TPE-

UVF due to the finite depth of field (�25 mm). However,

research papers


137

acquisition of multiple focal planes through samples has been

observed to recover crystal locations more quantitatively (not

shown).

In SHG measurements the possibility of false positives

exists from other SHG-active structures. Most notably, some

salts commonly used in crystallization screening can adopt

non-centrosymmetric SHG-active lattices and produce bright

SHG. Alternatively, noncrystalline structures exhibiting

molecular ordering over distances significantly greater than

the wavelength of light can also potentially produce false

positives for SHG. An example of a false positive, from a

noncentrosymmetric vanadate salt crystal, is shown in Fig. S1

of the supplementary information1 in which a cryo-loop

containing a crystal grown in LCP was examined with the

upstream NLO instrument, and yielded substantial signal in

the epi- and transmission-SHG directions. X-ray raster scans

suggested the presence of salt-like diffraction, in addition to

ice diffraction, as there was ice present on the sample loop.

Key signatures for an SHG-active salt were found to be bright

epi-SHG and little to no detectable TPE-UVF. These salt

crystal signatures can be exploited to reduce the likelihood of

false positives. False positives can arise

using TPE-UVF if there is protein

aggregate located within the loop

because TPE-UVF probes the presence

of aromatic residues and is not crystal

specific. Salt crystals and protein

aggregates are common occurrences

with protein crystal growth, generating

false positives for SHG and TPE-UVF

measurements, respectively. Fortu-

nately, most simple salts adopt SHG-

inactive centrosymmetric structures.

Complementary use of these two tech-

niques can significantly reduce the

likelihood of false positives and false

negatives.

Combined NLO imaging and XRD

was also applied to studies of �-cellu-lose, which exhibits fiber-like diffrac-

tion. NLO measurements performed

on loop-mounted cellulose generated

moderate S/N for multiple fibers within

the sample loop (Fig. 5, acquired via

the upstream system). Although fiber

diffraction was evident from the cellu-

lose samples, the DISTL algorithm used

in raster scanning, which searches for

discrete Bragg reflections or spots and

not fiber diffraction, does not indicate

these areas, but rather seems to show

that no measurable sample is present.

Manual inspection of the individual

diffraction patterns was performed to discern the presence of

fiber diffraction.

cPAH crystals ranging in size from 50 mm to 200 mm in

length were imaged with both the downstream instrument

with epi-only detection and X-ray raster scanning [Fig. S2

(supplementary information)]. The locations of intense

protein-like Bragg diffraction typically agreed well with those

of brightest epi-SHG for both large and small cPAH crystals

(e.g. Fig. S2). However, departures between the two were also

observed. Several explanations for the differences were

considered. First, the presence of multiple crystalline domains

within the crystal (e.g. from twinning) may cause the diffrac-

tion spot total to deviate from indicating optimal protein

ordering. Second, inhomogeneous optical scattering of the

incident or detected light can potentially impact the contrast

through effects unrelated to the crystal SHG activity.

However, bright-field images do not suggest substantial

differences in optical transmissivity across the crystal that

might have influenced contrast. Finally, NLO measurements

probe a much narrower depth of field than X-ray diffraction,

which is penetrating. If a particular crystal was not positioned

within the depth of field of the beam-scanning NLO micro-

scope, the SHG efficiency will be substantially reduced or

entirely absent within the detection limits of the instrument.

Despite the quantitative discrepancies, the presence of SHG

research papers


Figure 2(a) Bright-field image of a T. spiralis UCH37 1-226/UbVME complex crystal (�100 mm thick) andthe corresponding (b) epi-SHG, (c) trans-SHG, (d) TPE-UVF and (e) X-ray raster scan within the300 � 300 mm box. ( f ) X-ray diffraction of a representative 10 mm-diameter area from (e). X-rayenergy: 12 keV; exposure time: 1 s; photon flux: 2.7 � 109 photons s�1 (10-fold attenuation);detector distance: 300 mm; maximum theoretical resolution: 2.25 A. The large difference in the epi-and trans-SHG signals is expected for thick samples owing to the difference in the forward andbackward coherence length. The intensities of the two directions will approach equality as thesample thickness approaches the backwards coherence length (�100 nm). Scale bars are 100 mm.(Three darkened spots, apparent in this figure, arose from separate X-ray ‘burn tests’ to assess X-raydamage, the results of which will be published in a future study.)

1 Supplementary data for this paper are available from the IUCr electronicarchives (Reference: WA5051). Services for accessing these data are describedat the back of the journal.

138

signals above the background correlated with the areas of

the crystal generating a detectable protein-like diffraction,

providing preliminary confirmation of the ability of the

downstream instrument to rapidly generate information for

crystal position as a complement to X-ray raster scanning.

The polyimide loops (MiTeGen) were found to undergo

noticeable deformation with less than 100 mW incident power

using the downstream system, whereas the nylon loops were

more robust, and were not damaged at these powers. No

noticeable damage could be induced in either loop types using

the upstream system during either SHG or TPE-UVF

measurements (120 mW and 90 mW, respectively). Several

mechanisms were considered for the observed laser-induced

damage to the polyimide loops when measured with the

downstream system. Previous studies suggest that damage

from multi-photon absorption and plasma formation was

found to be an important, if not dominant, mechanism for

damage in biological NLO imaging (Sacconi et al., 2006).

However, those measurements were performed under condi-

tions of tight focusing [high numerical aperture (NA)] and

on live cells/tissues. However, alternative mechanisms may

dominate in the present low-NA studies of purified protein

crystals maintained under cryogenic conditions. Local heating

was also considered as a possible damage mechanism, arising

from either one- or two-photon absorption of the incident

beam. The marked difference in damage susceptibilities

between the upstream and downstream systems is consistent

with this mechanism, differing notably in the use of a resonant

research papers


Figure 4(a) Bright-field image of a membrane protein (human �-opioid receptorcomplex) crystal in lipidic cubic phase and the corresponding (b) trans-SHG and (c) TPE-UVF, with (d) an X-ray raster summary overlayshowing corrected Bragg-like reflection counts. (e) X-ray diffraction ofthe 5 mm-diameter area corresponding to the red circles in each image.X-ray energy: 12.0 keV; exposure time: 1 s; photon flux: 2.7 �1010 photons s�1 (unattenuated beam); sample-to-detector distance:300 mm; maximum theoretical resolution: 2.25 A. Scale bars are 20 mm.Cross-hairs were added to (b) and (c) to assist in orienting the fields ofview with respect to the bright-field and diffraction raster images.

Figure 3(a) Bright-field for an intimin protein crystal generated in LCP withcorresponding (b) trans-SHG and (c) X-ray raster summary overlayshowing corrected Bragg-like reflection counts. (d) X-ray diffraction ofthe 5 mm-diameter area corresponding to the red circles in each image,with X-ray energy 12.0 keV, exposure time 1 s, photon flux 2.7 �1010 photons s�1 (unattenuated beam), sample-to-detector distance of300 mm, resulting in a maximum theoretical resolution of 2.25 A. Scalebars are 50 mm. Cross-hairs were added to (a) and (b) to assist in orientingthe field of view with respect to the diffraction raster images.

139

8 kHz scan mirror for the upstream system and a galvan-

ometer-driven mirror operating at 200 Hz on the downstream

system. Rapid beam-scanning using a resonant scanner

combined with long-wavelength (>1 mm) incident light was

shown previously to have no detectable effect on crystal

diffraction quality using a variety of protein crystals, including

myoglobin crystals containing heme groups exhibiting strong

visible light absorption (Kissick et al., 2013). Myoglobin was

specifically chosen, as the color center was anticipated to be

highly susceptible to light-induced perturbation (Banerjee et

al., 1969). However, no statistically significant structural

changes to the lattice were observed in laser-exposed versus

unexposed regions of single crystals (Kissick et al., 2013).

The susceptibility for damage using the polyimide loops

increased notably for TPE-UVF, as the optical transparency

was substantially reduced at 530 nm. Whereas loop absorption

is negligible at 1 mm for SHG, roughly 30% of the incident

530 nm light for TPE-UVF is absorbed by the standard

yellow-tinted polyimide loop material (MiTeGen, http://www.

mitegen.com/). By positioning the loop to avoid the outer

turning points of the fast-scan mirror or blocking the beam at

those locations, no noticeable damage could be induced in the

polyimide loops during TPE-UVF imaging.

Both of the NLO imaging systems presented in this paper

have strengths and limitations, and either could be utilized as

a method for locating and centering protein crystals on a

synchrotron beamline. With a small footprint and the ability to

insert and remove the instrument, there is potential for a

single design of the downstream instrument to be utilized on a

variety of different beamlines. However, the time required for

translating the entire microscope to and from the sample

increases the total time for collecting SHG images and XRD

of the protein. Indeed, the microscope positioning required

substantially more time (�2 min) than the sample imaging

(�40 s). Furthermore, the absolute accuracy of the translation

stage (in this case,�2 mm) can ultimately dictate the precision

in crystal positioning. In addition, the downstream instrument

did not have transmission-SHG detection capabilities. For

protein crystals, detection in transmission provides substantial

improvements in detection limits for weakly SHG-active

proteins, as thickness greater than the crystals’ coherence

lengths can decrease the overall SHG intensity in the epi

direction (Boyd, 2009; Kestur et al., 2012). The absence of

transmission detection could potentially be remedied by

introducing additional optics or integrating into existing

optical paths.

The direct integration of the upstream system eliminated

the need for a translation stage for inserting the microscope,

as was used with the downstream system. This significantly

reduced the time between imaging and XRD, which allowed

for a marked improvement on throughput of data collection.

The upstream system did still require the transmission detec-

tion optics to translate in and out for XRD collection in

transmission, but epi-detected SHG can be performed

concurrently with X-ray diffraction, with only a factor of three

reduction in signal intensity with the mini-beam collimator

in place. The positioning of the collection optics does not,

however, require precise realignment allowing for a significant

improvement on the translation time, as compared with the

downstream instrument, where the entire microscope requires

translation with high precision. The upstream system had

some design trade-offs to accommodate the existing optical

path, which in part accounted for the lower infrared (IR)

throughput and available power in the upstream system. The

biggest losses came from the incident objective in which 80%

of the IR power was lost from reflections because it was not

designed for IR incident light. Choosing optics with a more

broadband anti-reflective coating (ARC) will significantly

improve the power throughput. Testing performed in-house,

with an IR-ARC objective, resulted in a doubling of the

IR transmittance, corresponding to an anticipated four-fold

improvement in signal at the sample (unpublished). The

multiple imaging modes (SHG and TPE-UVF), as well as both

epi and transmission detection, improves the ability of the

upstream system to detect protein crystals that could other-

wise be missed on the downstream system.

research papers


Figure 5(a) Bright-field image of �-cellulose fibers and the corresponding (b) epi-SHG and (c) trans-SHG images, all 300� 300 mm. (d) X-ray diffraction ofa 10 mm-diameter area within the red circle of each image. X-ray energy:12.0 keV; exposure time: 1 s; photon flux: 2.7 � 1010 photons s�1

(unattenuated beam); sample-to-detector distance: 300 mm; maximumtheoretical resolution: 2.25 A. Scale bars are 100 mm. Cross-hairs wereadded to (b) and (c) to assist in orienting the fields of view with respect tothe bright-field image.

140

Based on these combined results, integrating a NLO

microscope with a synchrotron XRD instrument complements

stand-alone X-ray raster scanning for crystal centering in three

key respects. First, it is expected to minimize radiation-

induced sample damage compared with X-ray raster techni-

ques for X-ray labile crystals or small crystals difficult to

quickly detect at low X-ray flux (Kissick et al., 2013). Second,

NLO microscopy significantly increases the spatial resolution

and reduces the total acquisition time for the determination of

crystal location. For a large sample area (150 � 150 mm)

scanned with a small beam size (5 � 5 mm), X-ray raster

images for the protein crystals typically required approxi-

mately 30 min to acquire with a 1 s X-ray exposure time. For

NLO measurements on identical samples, the acquisition time

for the collection of each image was typically <10 s. The

downstream NLO system allows 512 � 512 pixel images with

40 s acquisitions, and the upstream system allows 150 � 150

pixel images with 1 s acquisitions, which is roughly a >104-fold

reduction in the per-pixel acquisition time compared with the

X-ray raster acquisition time per cell (�3 s per pixel, corre-

sponding to a 1 s exposure, with 2 s of dead-time between

pixel acquisitions). The theoretical resolution of the objective

was 1.6 mm with 2 mm measured spatial resolution. The

downstream NLO system required a total time of 2.5 min for

translation of the microscope from its resting position to the

sample and then back to the resting position following NLO

measurements, resulting in a total acquisition time for each

sample of the order of 3 min, which is still significantly faster

and of higher resolution compared with X-ray raster scan

measurements performed on the same sample. In the

upstream system, no dead-time was required for epi-detection

(in fact, SHG imaging can be performed while acquiring

diffraction measurements), and only a few seconds of trans-

lation time were required to raise and lower the collection

optics in transmission. Third, for weakly diffracting systems

where rapid automated diffraction scoring is challenging,

NLO measurements may significantly increase the ability to

locate protein crystals.

5. Conclusion

Two different designs of integrated NLO instruments were

constructed and characterized targeting applications for

automated sample positioning. The systems were evaluated

using protein crystals (TsUCH37-UbVME, kOR-T4L, cPAH,

Intimin) and fibers (�-cellulose). Both NLO and XRD

exhibited good agreement for crystal positioning, consistent

with previous off-line measurements specifically targeting

protein crystals (Kissick et al., 2013). The integrated NLO and

synchrotron XRD instrument was found to enable precise

centering of �-cellulose samples for fiber diffraction without

requiring the development of an application-specific analysis

algorithm. The NLO instrument produced images with <10 s

image acquisition times, compared with 3–60 min for X-ray

rastering performed at much lower spatial resolution. By

nature of the higher resolution of NLO image acquisition, the

per-pixel raw data acquisition time was approximately five

orders of magnitude faster than X-ray raster scanning. Once

fully developed, NLO imaging may serve to identify regions of

interest for targeted X-ray scanning, or ultimately serve as the

sole or primary method for precise automated crystal posi-

tioning, such that all of the X-rays striking the crystal are

dedicated to structure elucidation.

Despite these successes, a relatively small variety of crystals

were used to characterize the instruments in this initial study.

Further studies on a greater diversity of protein crystals will

help define the scope of use for NLO methods in automated

centering. Additionally, the present study focused exclusively

on the hardware for visualization, and not on subsequent

algorithms for image analysis and automated crystal posi-

tioning. Higher contrast afforded by NLO imaging has the

potential to significantly improve the reliability of such algo-

rithms if the combined techniques of SHG and TPE-UVF

provide sufficient protein crystal coverage for general-purpose

use.

These studies provided a foundation for future efforts

combining NLO measurements with synchrotron X-ray

diffraction. The data presented here support the use of the

NLO microscopy for automated or manual crystal centering

prior to or in lieu of raster scanning. Potential scope of use

where all optical crystal positioning would be preferred

includes the analysis of smaller crystals (<5 mm), where the

low crystal volume may present challenges for rapid crystal

positioning by X-ray raster scanning. SHG also enables posi-

tioning of fibrous material exhibiting fiber diffraction, such as

cellulose, collagen, chitin etc. Further potential applications

include defect studies, X-ray damage studies and studies of

active pharmaceutical ingredients.

The authors acknowledge Huixian Wu, Victoria J. Hall,

Emma L. DeWalt, Valerie Pye, Martin Caffrey, Nicholas

Noinaj and James W. Fairman for aiding in sample prepara-

tion. Instrumentation development was supported in part by

the Center for Direct Catalytic Conversion of Biomass to

Biofuels (C3Bio), an Energy Frontier Research Center funded

by the US Department of Energy, Office of Science, Office of

Basic Energy Sciences, Award No. DE-SC0000997, and by the

NIH-NIGMS through the R01GM-103401. GM/CA@APS has

been funded in whole or in part with federal funds from the

National Cancer Institute (Y1-CO-1020) and the National

Institute of General Medical Sciences (Y1-GM-1104). Use

of the Advanced Photon Source was supported by the US

Department of Energy, Basic Energy Sciences, Office of

Science, under contract No. DE-AC02-06CH11357. Support is

also acknowledged from the NIH Common Fund in Structural

Biology, grant P50 GM073197. SKB is supported by the

Intramural Research Program of the NIH, National Institute

of Diabetes and Digestive and Kidney Diseases.

References

Accardo, A., Gentile, F., Mecarini, F., De Angelis, F., Burghammer,M., Di Fabrizio, E. & Riekel, C. (2010). Langmuir, 26, 15057–15064.

research papers


141

Aishima, J., Owen, R. L., Axford, D., Shepherd, E., Winter, G., Levik,K., Gibbons, P., Ashton, A. & Evans, G. (2010). Acta Cryst. D66,1032–1035.

Andrey, P., Lavault, B., Cipriani, F. & Maurin, Y. (2004). J. Appl.Cryst. 37, 265–269.

Banerjee, R., Alpert, Y., Leterrier, F. & Williams, R. J. (1969).Biochemistry, 8, 2862–2867.

Bates, S., Zografi, G., Engers, D., Morris, K., Crowley, K. & Newman,A. (2006). Pharm. Res. 23, 2333–2349.

Berger, M. A., Decker, J. H. & Mathews, I. I. (2010). J. Appl. Cryst.43, 1513–1518.

Boyd, R. (2009). J. Biomed. Opt. 14, 029902.Broennimann, Ch., Eikenberry, E. F., Henrich, B., Horisberger, R.,Huelsen, G., Pohl, E., Schmitt, B., Schulze-Briese, C., Suzuki, M.,Tomizaki, T., Toyokawa, H. & Wagner, A. (2006). J. SynchrotronRad. 13, 120–130.

Caffrey, M. & Cherezov, V. (2009). Nat. Protoc. 4, 706–731.Chen, J., Callis, P. R. & King, J. (2009). Biochemistry, 48, 3708–3716.Cherezov, V. & Caffrey, M. (2003). J. Appl. Cryst. 36, 1372–1377.Cherezov, V., Hanson, M. A., Griffith, M. T., Hilgart, M. C., Sanishvili,R., Nagarajan, V., Stepanov, S., Fischetti, R. F., Kuhn, P. & Stevens,R. C. (2009). J. R. Soc. Interface, 6, S587–S597.

Cherezov, V., Peddi, A., Muthusubramaniam, L., Zheng, Y. F. &Caffrey, M. (2004). Acta Cryst. D60, 1795–1807.

Dauter, Z. (2006). Acta Cryst. D62, 1–11.Dobrianov, I., Caylor, C., Lemay, S., Finkelstein, K. & Thorne, R.(1999). J. Cryst. Growth, 196, 511–523.

Fairman, J. W., Dautin, N., Wojtowicz, D., Liu, W., Noinaj, N.,Barnard, T. J., Udho, E., Przytycka, T. M., Cherezov, V. &Buchanan, S. K. (2012). Structure, 20, 1233–1243.

Fischetti, R. F., Xu, S., Yoder, D. W., Becker, M., Nagarajan, V.,Sanishvili, R., Hilgart, M. C., Stepanov, S., Makarov, O. & Smith,J. L. (2009). J. Synchrotron Rad. 16, 217–225.

Haupert, L. M., DeWalt, E. L. & Simpson, G. J. (2012). Acta Cryst.D68, 1513–1521.

Haupert, L. M. & Simpson, G. J. (2011). Methods, 55, 379–386.Hilgart, M. C., Sanishvili, R., Ogata, C. M., Becker, M., Venugopalan,N., Stepanov, S., Makarov, O., Smith, J. L. & Fischetti, R. F. (2011).J. Synchrotron Rad. 18, 717–722.

Ihee, H., Wulff, M., Kim, J. & Adachi, S. (2010). Int. Rev. Phys. Chem.29, 453–520.

Jain, A. & Stojanoff, V. (2007). J. Synchrotron Rad. 14, 355–360.Karain, W. I., Bourenkov, G. P., Blume, H. & Bartunik, H. D. (2002).Acta Cryst. D58, 1519–1522.

Kestur, U. S., Wanapun, D., Toth, S. J., Wegiel, L. A., Simpson, G. J. &Taylor, L. S. (2012). J. Pharm. Sci. 101, 4201–4213.

Kissick, D. J., Dettmar, C. M., Becker, M., Mulichak, A. M., Cherezov,V., Ginell, S. L., Battaile, K. P., Keefe, L. J., Fischetti, R. F. &Simpson, G. J. (2013). Acta Cryst. D69, 843–851.

Kissick, D. J., Gualtieri, E. J., Simpson, G. J. & Cherezov, V. (2010).Anal. Chem. 82, 491–497.

Madden, J. T., DeWalt, E. L. & Simpson, G. J. (2011). Acta Cryst.D67,839–846.

Maire, A. le, Gelin, M., Pochet, S., Hoh, F., Pirocchi, M., Guichou,J.-F., Ferrer, J.-L. & Labesse, G. (2011). Acta Cryst. D67, 747–755.

Moukhametzianov, R., Burghammer, M., Edwards, P. C., Petit-demange, S., Popov, D., Fransen, M., McMullan, G., Schertler,G. F. X. & Riekel, C. (2008). Acta Cryst. D64, 158–166.

Nanao, M. H. & Ravelli, R. B. (2006). Structure, 14, 791–800.NIH (2011). ImageJ, http://rsbweb.nih.gov/ij/.Parker, S., Kenney, C., Gnani, D., Thompson, A., Mandelli, E.,Meddeler, G., Hasi, J., Morse, J. & Westbrook, E. (2006). IEEETrans. Nucl. Sci. 53, 1676–1688.

Pohl, E., Ristau, U., Gehrmann, T., Jahn, D., Robrahn, B., Malthan,D., Dobler, H. & Hermes, C. (2004). J. Synchrotron Rad. 11, 372–377.

Pothineni, S. B., Strutz, T. & Lamzin, V. S. (2006). Acta Cryst. D62,1358–1368.

Rasmussen, S. G., DeVree, B. T., Zou, Y., Kruse, A. C., Chung, K. Y.,Kobilka, T. S., Thian, F. S., Chae, P. S., Pardon, E., Calinski, D.,Mathiesen, J. M., Shah, S. T., Lyons, J. A., Caffrey, M., Gellman,S. H., Steyaert, J., Skiniotis, G., Weis, W. I., Sunahara, R. K. &Kobilka, B. K. (2011). Nature (London), 477, 549–555.

Ravelli, R. B. & Garman, E. F. (2006). Curr. Opin. Struct. Biol. 16,624–629.

Riekel, C., Burghammer, M. & Schertler, G. (2005). Curr. Opin.Struct. Biol. 15, 556–562.

Sacconi, L., Dombeck, D. A. & Webb, W. W. (2006). Proc. Natl Acad.Sci. USA, 103, 3124–3129.

Sixta, H., Harms, H., Dapia, S., Parajo, J., Puls, J., Saake, B., Fink, H. &Roder, T. (2004). Cellulose, 11, 73–83.

Song, J., Mathew, D., Jacob, S. A., Corbett, L., Moorhead, P. & Soltis,S. M. (2007). J. Synchrotron Rad. 14, 191–195.

Stepanov, S., Hilgart, M., Yoder, D. W., Makarov, O., Becker, M.,Sanishvili, R., Ogata, C. M., Venugopalan, N., Aragao, D., Caffrey,M., Smith, J. L. & Fischetti, R. F. (2011a). J. Appl. Cryst. 44, 772–778.

Stepanov, S., Makarov, O., Hilgart, M., Pothineni, S. B., Urakhchin,A., Devarapalli, S., Yoder, D., Becker, M., Ogata, C., Sanishvili, R.,Venugopalan, N., Smith, J. L. & Fischetti, R. F. (2011b). Acta Cryst.D67, 176–188.

Takai, M. & Colvin, J. R. (1978). J. Polym. Sci. 16, 1335–1342.Vernede, X., Lavault, B., Ohana, J., Nurizzo, D., Joly, J., Jacquamet, L.,Felisaz, F., Cipriani, F. & Bourgeois, D. (2006). Acta Cryst. D62,253–261.

Wu, H., Wacker, D., Mileni, M., Katritch, V., Han, G. W., Vardy, E.,Liu, W., Thompson, A. A., Huang, X. P., Carroll, F. I., Mascarella,S. W., Westkaemper, R. B., Mosier, P. D., Roth, B. L., Cherezov, V.& Stevens, R. C. (2012). Nature (London), 485, 327–332.

Zhang, Z., Sauter, N. K., van den Bedem, H., Snell, G. & Deacon,A. M. (2006). J. Appl. Cryst. 39, 112–119.

research papers


142

Structural and biophysical analysis of the proteasomal ...

Documents