Mirza PhD defense on the Ugi reaction for anti-malarial screening

Khalid Baig Mirza

Advisor : Prof. Jean-Claude Bradley

Drexel University

Prof Frank Ji (Chair) , Prof. Susan A. Varnum, Prof. Sally Solomon, Prof Peter Wade, Prof. Louis Scerbo, Prof. Jun Xi

Introduction “Open Notebook Science” is a term used to indicate that the

primary record of a research in its entirety is made public over the internet as soon as it is conducted

Achieves and advocates complete transparency of research being performed

Serve as an open invitation to collaborate with like minded researchers / scientists who are capable and willing to participate in the project

The planned experimental procedure, the log, raw data, and a discussion of the analysis of the data, the assumptions and the conclusions drawn from the specific experiment and the project in general made public in real time.

The UsefulChem project – A collaborative effort to find a treatment/cure to malaria

Tools used – Wikis, blogs, YouTube, FriendFeed, Google Docs, JSpecView, ChemSpider, Mendeley, Online repositories, etc

Why Open Note Book Science?

Justification for the Open Notebook Science approach Oxidation of benzylic alcohols using sodium hydride

NaH is commonly used as a base, eg. enolate chemistry1

References

1 Hudrlik,P., F.; Takacs, J., M.; J. Org. Chem; 43; 20; 3861-3865 (1978)

2 McConaghy Jr. J., S.; Bloomfield, J., J,; J. Org. Chem; 33; 9; 3425-3428 (1968)

O O-Na

+

NaH

To a lesser extent it has also been used as a reducing agent2

Sodium hydride as an oxidizing agent

In a JACS publication last year, Wang, et. al. reported that benzylic type

secondary alcohols under go oxidation when treated with NaH (2eq) in THF at

room temperature. Several examples were given3.

Well picked-up in the blogosphere community, facilitated by social networking

websites like Friendfeed and others.References

3 Wang, X., Zhang, B., Wang, D., Z.; J Amer. Chem Soc; DOI: 10.1021/ja904224y; (2009)

Wang’s ketones...

Experiments repeated by the scientific community from the blogosphere

Totally synthetica :

UsefulChemb:

Experimental details were comprehensively

documented with photos and videos

CH3

OH

CH3

O

2eq NaH

THFx

CH3

OH

Cl

CH3

O

Cl

2eq NaH

THF

15%

a) Paul Docherty- totallysynthetic.com/blog/?p=1903; (12/05/2010)

b) Khalid Mirza, Marshall Moritz, Jean-Claude Bradley; UsefulChem.wikispaces.com/Exp243. Reaction monitored before quenching; (12/05/2010)

CH3

OH

CH3

O

2eq NaH

THF, 19h

1-phenylethanol acetophenone

Green: 1-phenylethanolBlue: acetophenoneRed: product HNMR Overlay

PPM

Green: acetophenoneBlue:1-phenylethanolRed: product

CH3

OH

CH3

O

2eq NaH

THF, 19h

1-phenylethanol acetophenone

IR Overlay

• After 19h no conversion to acetophenone was observed, paper mentioned 75% GC yield

• However, totallysynthetic.com reported at 15% conversion to the ketone. Wang reported 85% isolated yield

• A comment made by a reader on another blog

Carbon based curiositiesa mentions a JOC

paper from 1968 which address the issue.

ahttp://www.coronene.com/blog/?p=842

Results

Lewis E., G.; J. Org. Chem; 30; 7; 2433–2436 (1965)

Scientific communication through social networks- Blog

Wang’s paper withdrawn from JACS

Social networking can/has played an important role in the development of science

Social networking + collaborative effort => unforeseen innovations (Chemistry, Medicine, Engineering, .. data curation as well.

Consequences

The UsefulChem Project

Malaria – Some Facts

Reported cases: 190-311m [ 2008, CDC]Deaths: 708,000 - 1,003,000 [2008, WHO]Half the worlds population (3.3 billion)

susceptible to malaria.[CDC]Second leading cause of deaths from infectious

diseases in Africa after HIV/AIDS [CDC]Approximately 1500 cases of malaria are

reported in the US annually [CDC]Drug resistance of Plasmodium falciparum to

chloroquin, sufladoxine-pyrimethamine are well known. [WHO]

New anti-malarial agents needed.

Collaboration with Find-a-Drug

Obtained a small library of diketopiperazines designed to inhibit enoyl-reducatase, an essential enzyme for the fatty-acid metabolism of Plasmodium falciparium.

Target diketopiperazine

Diketopiperazine library

Reterosynthesis of the target diketopiperazine

•All starting materials except 3, 4-dihydrophenyl acetaldehyde (6) were commercially available.

•Therefore a synthetic procedure had to be designed.

Proposed mechanism for the dehydration of adrenaline to DOPAL

Ugi reaction

OH

OH

O

O

CH3

NH2

N+

-C

O N CH3O

O

NH

S

CH3

NH

OO

CH3CH3

CH3

OH

OHmethanol

OO

CH3CH3

CH3

NH

O OH

S

CH3

The reaction did not produce the desired Ugi product

Monitored the reaction with similar simpler reactants such as phenylacetaldehyde and boc-glycine

Mechanistic understanding of the reaction using NMR monitoring

Reactivity of the alpha protons on phenylacetaldehyde was one of the main reasons for the failure of reaction.

Monitoring the Ugi reaction

A Mechanistic Insight

Imine kinetics

aldehyde amine solvent

Imine formation rate constant 1/(M*min)

veratraldehyde 5-methylfurfurylamine CDCl3 0.01

veratraldehyde 5-methylfurfurylamine CD3OD 0.106

piperonal 5-methylfurfurylamine CDCl3 0.07 a

piperonal 5-methylfurfurylamine CD3OD 0.1552 a

piperonal t-butyl amine CD3OD 0.008 b

3,4,-dihydroxybenzaldehyde 5-methylfurfurylamine CD3OD 0.1043 c

a- Alicia Holsey, b- James Giammarco, c- Sean Gardner

Ugi reactionaldehyde amine carboxylic acid isocyanide Ugi product % Yield

H3CO

H3CO

CHO

O

NH2

CHO

CHO

CHO

O

NH

O

O

CH3

CH3

CH3N

O

O NH

CH3 CH3

CH3

ONH

O

O

CH3

CH3

CH3

N

O

ONH

CH3

CH3

CH3

OCH3H3CO

CH3CH3

CH3

NH O

N

(H2C)6CH3

O

CH3

CH3CH3

CH3

NH O

N

O

CH3

CH3

CH3 CH3

N+ C

-

O

O

CH3

CH3

CH3

NHO

OH

O

O

CH3

CH3

CH3

NHO

OH

CH3

CH3 CH3

N+ C

-

CH3

CH3 CH3

N+ C

-

CH3

CH3 CH3

N+ C

-

NH2

OH

O

CH3

CHO

O

O

CH3

CH3

CH3

NHO

OH

N+

C-

CH3(CH2)5

NH2

OH

O

CH3

O

NH2

CH3

N+

C- ONH

O

O

CH3

CH3

CH3

N

OCH3

ONH

OCH3

OCH3

O

NH2

H3CO

H3CO

CHO

O

O

CH3

CH3

CH3

NHO

OH 37

44

18

31

50

58.5

8

9

10

11

12

13

O

NH2

O

NH

O

O

CH3

CH3

CH3

N

O

O

NH

Fufuryl Cleavage

Attempt at 2,5-diketopiperazine synthesis

The UDC strategy- Ugi deboc cyclization4

R1

CHO

R4

NC-

Ugi reaction (4CR)

methanol, rt

Ugi product

10% TFA / CDCl 3 or CD3OD

10% TFA / CDCl 3 or CD3OD

xO

NH2

CH3 OCH3

N

O

NHO

R1

R4

NHBoc

R3

diketopiperazineCO2H

NHBoc

R3

NH

O

NHO

R1

R4

NH2

R3

OCH3

N O

O NH

R1

R3

4Hulme, C.; Morrissette, M.; Volz, F., A.; Burns, C.; Tett. Lett. 39; 10; 1113-1116; (1998)

CHO

H3CO

H3CO

CHO

O

O

CHO

ONH2

ONH2CH3 BocHN CO2H

NC-

CH3

NC-CH3

CH3

R1

CHO R2

NH2 R3

CO2H R4

NC-

OCH3

N

O

NHO

R1

R4

NHBoc

R3

Furfuryl Cleavage

The unusual elimination was seen in Ugi products containing the N-furfuryl group.

Reactivity of furan derivatives acidic media

Butin, A.,V; Stroganova, T.; Lodina, I., V.; Krapivin, G., D.; Tet. Lett; 42; 10; 2031-2033; (2001)

Reissert modification-indole synthesis

Furan ring opening with lewis acid

Piancatelli, G.; Scettri, A.; David, G.; D’Auria, M.; Tetrahedron; 34; 18; 2775-; (1978)

1, 6-Hoffman elimination of quaternary amines

O

N+(CH3)3

R

OCH2 CH2

R= H, CH3, C6H5

O

x

OO

furan analog of 2,2-[paracyclophane]

polymerization inhibitors20

21

22

23

Appropriately substituted quaternary ammonium hydroxides undergo loss of trimethyl amine across a furan or thiophene to unusual cyclic conjugated trienes, there dimmers and polymers

OCH2 CH2 N

+(CH3)3

HOH-

OCH2 CH2

N(CH3)3+ OH2+

21 2420

Winberg, H., E.; Fawcett, F., S.; Mochel, W.,E.; Thebald, W., C.; J. Amer. Chem .Soc ; 82; 6; 1428-1435; (1959)

Furfuryl cleavage in Ugi products

Observed 5-methyl-2-furfuryl group on nitrogen of a tertiary amide, in the Ugi products.

Undergo a 1-6- elimination to yield a secondary amide when treated with trifluoroacetic acid in CDCl3 or CD3OD

N

R2

R1

OO

CH3

k = 0.9x10-3 min-1

k = 0.9x10-3 min-1

First Order Kinetics for 1,6-methyl furfuryl elimination

y = -0.0009x - 1.5593

-4.5

-4

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 500 1000 1500 2000 2500 3000 3500

time (min)

ln[c

on

c]

(M)

H3CO

H3CO

N

O

NHBoc

NH

O

O

CH3

TFA-

H3CO

H3CO

NH

O

NH3+

NH

O10% TFA /CDCl3

11 13

k = 1.2x10-3 min-1

k = 0.9x10-3 min-1

Proposed mechanism for furfuryl cleavage

TFA-

10% TFA /CDCl 3

R1

N

O

NHBoc

NH

O

O

CH3

R2

R1

N+

O

NH3+

NH

O

O

CH2

R2

H

H

HR

1

NH

O

NHBoc

NH

O

R2

H

CH2

O

CH2

polymer

+10% TFA /CDCl 3

10% TFA /CDCl3

No methyl group- The cleavage is a thousand times slower than the methyl furfuryl analogs cleavage under the same conditions

k= 4x10-6 min-1

A different mechanism needed...

Non-methylated furfuryl analog cleavage

Mechanism for the furfuryl cleavagefor the non methylated analog

50% TFA /CDCl 3

R1

N

O

NHBoc

NH

O

O

R2

R1

N+

O

NH3+TFA

-

NH

O

O

R2

H

H50% TFA /CDCl 3

R1

NH

O

NH3+TFA

-

NH

O

R2

H

CH2

O+

Polymer

+

Optimization of the Ugi reaction

Automated optimization of a Ugi reaction*

The reaction is usually carried-out at room temperature while the product sometimes precipitates out from the reaction mixture.

When it precipitates out, the product is then just filtered and washed with the solvent.

The product precipitation although not consistent is very desirable when performing a scale-up, essentially eliminating an expensive chromatographic purification process.

In order to obtain high yield and purity of the product optimization studies were performed on a Ugi reaction, where the product was filtered directly and washed with the solvent.

Performed in collaboration with Prof. Kevin Owens, Drexel University

Automation performed using the 48-slot Mettler-Toledo MiniBlock quipped with filtration tubes.

Automation

Objective: To optimize the conditions to obtain highest yield

Concentration (0.4, 0.2, 0.07M)Solvent (methanol, ethanol, acetonitrile

and THF)Excess of some reagents (1.2eq)

Parameters under consideration

Reactions performed in little tube with filters at the tipRobot added the four components and solventPrecipitated product then washed, weighed and NMR

analyzed to confirm

Mettler-Toledo MiniBlock System

Results of the Optimization study

Statistical analysis: Prof. Kevin Owens

Results of the Optimization study

Methanol and ethanol best solvents at 0.2M reagent concentration

Yields decreased from 0.2M to 0.07M in MeOH, EtOH & AcCN. Similar yields at 0.2M and 0.4M concentrations in methanol

In EtOH and MeOH higher yields resulted with imine or isonitrile excess. In AcCN amine, aldehyde or isonitrile excess gave better result. In THF imine excess resulted in higher yields. Over yields significantly lower for THF than other solvents

Significant interactions between the solvent choice, reagent concentration and identity of the reagent in excess found

Best yield was 66% at 0.4M in methanol with imine excess (1.2eq).

This was significantly higher than the initial reaction under equimolar conditions (49%)

Library Synthesis

Library synthesisMost reactions were performed for the products which were

predicted to possess some biological activity.

Docking and modeling studies were conducted by collaborators, Dr. Rajarshi Guha and Prof. Andrew Lang

Results of the ranked virtual library of Ugi products were provided in the form of smiles

Based on the ranking Ugi products were synthesized

The compounds were then sent for testing against facipain-2, an enzyme used by the malarial parasite to break down hemoglobin. This was done in collaboration with the Rosenthal group at UCSF

Ugi product precipitation

Very few Ugi reactions yielded solid products

Empirical modeling studies were performed by Prof. Andy Lang (ORU) to predict the likelihood of Ugi product precipitation.

Reactions were performed and the results were incorporated to further improve the model.

Ugi reactions performed – 511Ugi reactions that precipitated a product – 104 (20%)Precipitate confirmed to be Ugi products – 65 (13%)

Precipitate not a Ugi product – 8

Ugi reaction precipitation trendsCorrelation in terms of isocyanide

Isocyanide

Number of

reaction

number of product

precipitations

Percentprecipitat

ion

1,1,3,3-tetramethylbutyl

isocyanide 32 0 0

1-pentyl isocyanide 38 0 0

2-chloro-6-methyl phenyl isocyanide 4 0 0

2-morpholinoethyl isocyanide 13 0 0

benzyl isocyanide 43 3 7

cyclohexyl isocyanide 77 10 13

n-butyl isocyanide 120 5 4

t-butyl isocyanide 215 23 11

tosylmethyl isocyanide 107 0 0Carboxylic acid – boc-glycine 35%

Tosylmethyl isocyanide (TOSMIC)..

TOSMIC was used in 107 Ugi reactions due to its odor-free nature

No solid Ugi products were obtained.

Poor solubility of TOSMIC in most organic solvents

Reactions had to be performed at lower concentrations

Reactivity towards aldimine in basic conditions

Anti-malarial activity results

Anti-malarial activity

Ugi products show inhibition of falcipain-2, cystein protease inhibitor used by the malarial parasite, Plasmodium falciparum to degrade erythrocytic proteins especially hemoglobin.

Show inhibition of the Plasmodium falciparum.

UsefulChem top inhibitorIC50 8.4uM

Solubility

Polyaromatic components in Ugi reactions

1-pyrenebutyric acid

methanol; 1 week

R

NCH3

NH

O

CH3

CH3

CH3

O

O

NCH3 CH3

O

171F/6-4 171K/11-4

RCHO =

O

OH

CH3 NH2

CH3CH3

CH3N

+C

-

RCHO

CH3

O

173B/14

Solubility issues

Solubility of phenanthrene-9-carboxaldehyde

SolventAve. (M)

11,1,2-

trichlorotrifluoroethane 0.02

2 2-propanol 0.07

3 DMF 1.25

4 DMSO 0.77

5 THF 1.29

6 acetonitrile 0.153

7 benzene 0.66

8 chloroform 0.04

9 cyclohexane 0.07

10 cyclopentane 0.03

11 dichloromethane 0

12 diethyl ether 0.1

13 ethanol 0.1

14 ethyl acetate 0.44

15 hexane 0.07

16 methanol 0.104

17 toluene 0.14

Solubility of 1-pyrenebutyric acid

SolventAve. (M)

1 2-propanol 0.06

2 DMF 1.88

3 DMSO 2.067

4 THF 0.55

5 acetonitrile 0

6 benzene 0

7 carbon tetrachloride 0

8 chloroform 0.03

9 cyclohexane 0

10 cyclopentane 0

11 dichloromethane 0.07

12 diethyl ether 0.02

13 ethanol 0.06

14 hexane 0

15 methanol 0.014

16 toluene 0

Solubility..

Poor solubility of reactants resulted in decreased reagent concentration

Reduced likelihood of product precipitation

Need for a solubility model, to predict selective product precipitation

Need to measure solubility of several reactants and Ugi products in different organic solvents at room temperature

ONSChallenge a collaborative effort with Prof. Andew Lang (ORU) and Dr. Rajarshi Guha

Methods for solubility assessment

The shake flask method to saturate a solution

UV-VIS, HPLC, GC

Turbidimetry and nephlometry

Differential Scanning Calorimetry

Methods we used

Speed-Vac method

Hemiacetal of 4-nitrobenzaldehyde in methanol (3:1 aldehyde: hemiacetal ratio)

Solute- Solvent reaction

2-chloro-5-nitrobenzaldehyde in methanol (2:3 aldehyde: hemiacetal ratio)

Solute- Solvent reaction

NMR method and Semi Automated Measurement of Solubility (SAMS)

Web-service created by Prof. Bradley and Prof. Lang using Google spreadsheet.

Based on HNMR of a very small volume of the supernatant in deuterated solvent

JCAMP dx file of the spectrum is uploaded to the server

Density and molar mass of the solute and solvent is incorporated

Predicted densities from ChemSpider.com used for solids

A range in the solute to integrate with the number of corresponding H

A range in the solvent to integrate with the number of corresponding H

SAMS..

Spreadsheet calculates the molar ratio of solute to solvent

Molarity calculation is then done by assuming that the volumes of the two components are additive.

Problem.. Inconsistent integration of aromatic and non aromatic protons

observed Increasing the relaxation delay d1 from 0.3s to 50s resulted in

uniform peak integrations per proton over the entire spectrum

Conclusions

An open, collaborative and more transparent approach to scientific research has been established

A mechanistic understanding in to the ‘furfuryl cleavage’ has been achieved

Conditions necessary for the Ugi reaction has been optimized Library of Ugi reactions have been performed

Synthesis of potential anti-malarial agents accomplished

Solubility of the reagents and Ugi products determined to facilitate the successful construction of a model to predict the selective precipitation of the Ugi products from a reaction

Future work

Convenient web services for solubility measurement and prediction

Virtual library generatorPredictive ToxicologyVirtual dockingIntegration of Multiple Web Services to

expedite the process of drug discover.

Books published..

Acknowledgements

A very patient advisor – Prof. Jean-Claude Bradley

My Parents and families support

Committee members- Prof Frank Ji, Prof. Susan Jansen Varnum, Prof. Sally Solomon, Prof Peter Wade, Prof. Louis Scerbo and Prof. Jun Xi, Prof. Robert Hutchins (Late)

Collaborators – Prof. Kevin Owens, Dr. Rajarshi Guha, Prof. Andrew Lang, ONSChallenge Judges & Tom Osborne.

Prof. Lynn Penn, Prof. Anthony Wambsgans

Ed Doherty, Virginia Nesmith, Tina Lewinski, Ed Thorne, Tim Wade

Co-workers - Alicia Holsey, James Giammarco, Sean Gardner, Emily Messner, Shannon Oseback, Tim Bohinski, Cedric Tschakounte, Marshall Moritz.

Renata Szyszka, Neil Mukherjee, Sudipto Das, David Berke-Schlessel , Kerry Drake, Jonathan Soffer, Addy Kojtari, April Holcomb, Bill Erb, Jim Reiben, Molly O’connor, Tyson Reeves, Christopher Castillo, Nick Paparoidamis, Hung Le, Tom Measy, Andrew Haggarman, Joseph Depasquale, Natalie Dixon, Mukesh Kumar, Ismael Nieto, Marcela Garcia, KimChi Nguyen, Kyle Hess, William Hunt, Arben Kojtari, Michelle Livings, Chi Nguyen, Siobhan Toal and others I missed..

Questions ?

Thank You