Modeling human hereditary syndromes using … human hereditary syndromes using CRISPR/Cas9 mediated genome editing in Xenopus tropicalis Kris Vleminckx ... • Heteroduplex Mobility

Modeling human hereditary

syndromes using CRISPR/Cas9 mediated genome editing in

Xenopus tropicalis

Kris Vleminckx

Ghent University

Belgium

Nov - 2017

- High number offspring (> 1000 eggs)

- embryogenesis proceeds rapidly

- Vertebrate – tetrapod (aquatic)

- embryos have a large size

Xenopus – main advantages

- External embryonic development

- Xenopus tropicalis is true diploid

- High synteny of genome

- Targeted injections are possible >< X. laevis and zebrafish

= X. laevis and zebrafish

- They don’t smell (>< the mouse)

Nakayima et al., 2005

Hellsten et al., 2010

see http://www.xenbase.org

Xenopus – targeted injections

Injections Cas9 and CRISPR guide RNAs (2–8 cell stage)

Phenotypic Readout

Xenopus – injections CRISPR/Cas9

Design gRNA (CRISPRscan)

Primer ligation + direct T7 RNA synthesis (cloning free)

Inject gRNA + Cas9 protein

Assess efficiency (NGS – MiSeq) Mostly > 70%

Raise or Adjust protocol

Analysis F0 mosaic (e.g. cancer models)

F0 F0

F1 phenotyping

< 2

we

eks

~

10

day

s

~ 6

mo

nth

s

Experimental Pipeline

< 2

mo

nth

s

> 80 genes targeted

Gene KO by Non-Homologous End Joining (NHEJ)

F0 mosaic

F1 offspring

Amaya lab

Methods for genotyping

F0 mosaics • Heteroduplex Mobility Assay (HMA) • TA-cloning and Sanger sequences • NGS + Batch-GE analysis

F1 offspring • High Resolution Melting Analysis (HRMA)

• Sanger Sequencing - Tracking of Indels by DEcomposition (TIDE) analysis

https://tide-calculator.nki.nl/

• Heteroduplex Mobility Assay (HMA) (for F0)

• TA-cloning and Sanger sequences

CRISPR/Cas efficiency analysis by NGS

Decreasing cost

Speed High throughput

Availability

PCR NGS (MiSeq) Data analysis

Slides courtesy of Annekatrien Boel (CMGG)

Decreasing cost

Speed High throughput

Availability

PCR NGS (MiSeq) Data analysis

BATCH-GE

CRISPR/Cas efficiency analysis by NGS

BATCH-GE

Methods for genotyping

F0 mosaics • Heteroduplex Mobility Assay (HMA) • TA-cloning and Sanger sequences • NGS + Batch-GE analysis

F1 offspring • High Resolution Melting Analysis (HRMA)

• Sanger Sequencing - Tracking of Indels by DEcomposition (TIDE) analysis

https://tide-calculator.nki.nl/

• High Resolution Melting Analysis (HRMA) (for F1)

Genotyping work flow for analysis of F1-offspring

HRMA

TIDE

Modeling human (mono)genetic disease

Pseudoxanthoma Elasticum (collab. O. Vanakker)

Abcc6 (full KO)

IPL INL

PRL ONL

RPE

OPL

RCBTB1-mosaic

Left eye

Right eye

IPL

PRL

OPL INL

ONL

GCL

Inherited Retinal Dystrophy (collab. E. De Baere)

Rcbtb1 (mosaic KO) – Unilateral injection

Naert et al., unpublished

Limb defects Ptch1 (mosaic KO)

Visel et al., Nature 2009

Regulation Shh expression (limb enhancer (ZRS) in LMBR1 intron)

Targeting Anterior-Posterior polarity – Shh

ZRS deletion

ZRS point mutations

ZRS = ZPA (Zone of Polarizing Activity) Regulation Sequences

Smith et al., Nat. Genet. 2013

Regulation Shh expression (limb enhancer (ZRS) in LMBR1 intron)

Targeting Anterior-Posterior polarity – Shh

Targeting shh limb enhancer (ZRS) in lmbr1 intron

Targeting Anterior-Posterior polarity – Shh

• APC and Familial Adenomatous Polyposis

Modeling human cancer

HD MT

binding EB1

binding

Arm MCR

15 AA repeat (β-catenin) 20 AA repeat (β-catenin) SAMP repeat (Axin)

Apc

TCTTCAGTACACCATACACGGACTAAAAACAACAGACTTCAAACATCAAA TCTTCAGTACACCATACACGGACT -AAAACAACAGACTTCAAACATCAAA (∆ 1) x2 TCTTCAGTACACCATACA - - - - - - - - - - - - - -ACAGACTTCAAACATCAAA (∆ 13) x2 TCTTCAGTACACCATACACGGACT - - - - - - - - - - - - - - -TCAAACATCAAA (∆14)

Targeting APC MCR in X. tropicalis (TALEN)

Familial Adenomatous Polyposis in Xenopus (targeting APC tumor suppressor gene)

Van Nieuwenhuysen et al., Oncoscience 2015

• Extra-colonic manifestations:

– Congenital Hypertrophy of Retinal Pigment Epithelium (CHRPE) √

– Desmoid tumors √

– Brain tumors (medulloblastoma) √

– Abnormal dentition

– Osteomas

– Epidermal cysts √

Familial Adenomatosis Polyposis in Xenopus (targeting APC tumor suppressor gene)

• Retinoblastoma (Rb1 + Rbl1)

Modeling human cancer

A Xenopus tropicalis retinoblastoma model H&E of Rb1/Rbl1 model

Naert et al., Sci. Rep. 2016

A Xenopus tropicalis retinoblastoma model H&E of Rb1/Rbl1 model

Naert et al., Sci. Rep. 2016

A Xenopus tropicalis retinoblastoma model H&E of Rb1/Rbl1 model

Naert et al., Sci. Rep. 2016

– Medulloblastoma (Wnt, Shh, other) √

– Neurofibromatosis type 1 (NF1) /

– Neurofibromatosis type 2 (NF2) (√)

– Bladder cancer (NF1, PTEN) √

– Leukemia (Notch1, FBXW7, PTEN) √

cancer models under development?

Add. KOs : p53, Rag2

• T-ALL (Notch1* + PTEN)

Modeling human cancer

WT froglet Notch1* + PTEN mosaic

Well you have tumors in your frogs …. So what??

BC-2059 : inhibitor of TBL1 dependent β-catenin signaling

application 1 : Desmoid tumor preclinical compound testing

application 2 : Identification of Driver Mutations of Modifiers

Yes, this is still empty

Dissect tumor Isolate DNA Sequence Gene X

Gene X disrupted In tumor?

YES

Gene X not important for tumorigenesis

NEVER

Gene X REQUIRED for tumorigenesis

Gene X = therapeutic target

Genome modification of APC and

gene x

application 3 : Desmoid tumor therapeutic target identification via gRNA multiplexing

workflow

Midkine (MDK)

0

10

20

30

40

50

60

70

80

90

100

DT1 DT2 DT3 DT4 DT5 DT6 DT7 DT8 DT9

APC

MDK

bi-allelic mutations in MDK detected in desmoid tumors (7); MDK is not essential for in vivo desmoid tumor initation

CRISPR/Cas9 editing efficiencies

Naert et al. unpublished

apc + ptt CRISPR/Cas9

Targeted deep sequencing of apc and ptt loci Dissect tumor

Identification of genes essential for desmoid tumorigenesis

0,00%

10,00%

20,00%

30,00%

40,00%

50,00%

60,00%

Percentage of tumors exhibiting bi-allelic out-of-frame mutations

n= 3 12 6 3 5 9 6 22 9 12

Domain 1 does not drive negative selection and can tolerate in-frame mutations

0

10

20

30

40

50

60

70

80

90

DT1 DT2 DT3 DT4 DT5 DT6 DT7 DT8 DT9 DT10DT11DT120

10

20

30

40

50

60

70

80

90

DT1 DT2 DT3 DT4 DT5 DT6 DT7 DT8 DT9 DT10 DT11 DT12

APC

EZH2

Raw targeted deep sequencing data In frames removed

Per

cen

tage

of

mu

tan

t se

qu

enci

ng

read

s

Perc

en

tage

of

mu

tan

t se

qu

enci

ng

read

s A

fter

rem

ovi

ng

in f

ram

es m

uta

nt

read

s

User CRISPR for Domain Mapping Negative selection to small in frames within functional domain reveals this domain as essential for tumorigenesis

D1 D2 D3 D5 D4

User CRISPR for Domain Mapping Negative selection to small in frames within functional domain reveals this domain as essential for tumorigenesis

Domain 5 targeting results in desmoid tumors that retain one fully wild type allele

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

DT1 DT3 DT5 DT7 DT9 DT11DT13DT15DT17DT19DT21DT23DT25

Raw targeted deep sequencing data

Per

cen

tage

of

mu

tan

t se

qu

enci

ng

read

s

D1 D2 D3 D5 D4

• Extremely simple, cheap and fast (>< mouse)

• Diploidy required (>< zebrafish)

• Targeted injections are great benefit (>< zebrafish)

Why do this in Xenopus tropicalis?

• So far primarily KO experiments

• Can we do KI ??

– HDR with repair template – using oocyte transfer

– Microhomology Mediated Endjoining (PITCh)

generation KO generation KI using host transfer

Generation Xenopus knock-in models

Slide courtesy of Marjolein Carron

Nr2e3 cDNA Mut or VUS 2A RFP

PITCh-MMEJ

Nr2e3 cDNA WT 2A ZsGreen1 STOP

Acknowledgements

Unit Developmental Biology Tom Van Nieuwenhuysen Hong Thi Tran Thomas Naert Suzan Demuynck Rivka Noelanders Dionysia Dimitrakopoulou UZ Gent pathology department Dr. David Creytens Stanford Clinical pathology Prof. Matt van de Rijn Joanna Przybyl

Collaborations: Center Medical Genetics Ghent Paul Coucke Annekatrien Boel Wouter Steyaert Andy Willaert Elfride De Baere Pieter Van Vlierberghe Gnommix Marnik Vuylsteke