Supplementary Information for
Epigenetic inactivation of oncogenic brachyury (TBXT) by H3K27 histone demethylase controls chordoma cell survival
Lucia Cottone, Edward S Hookway, Adam Cribbs, Graham Wells, Patrick Lombard, Lorena Ligammari, Anthony Tumber, Roberto Tirabosco, Fernanda Amary, Tamas Szommer, Catrine Johannson, Paul E. Brennan, Nischalan Pillay, Udo Oppermann, Adrienne M. Flanagan
Corresponding authors
Adrienne M Flanagan [email protected] and Udo Oppermann [email protected]
This PDF file includes:
Figs. S1 to S6 Tables S1 to S5 Captions for Datasets S1 to S2 Supplementary text References for SI reference citations
Other supplementary materials for this manuscript include the following:
Datasets S1 to S2
1
Supplementary Figures
Fig. S1. Ligand structure of human KDM6C/UTY in complex with KDOBA67. Crystal structure of UTY/KDM6C in complex with the inhibitor KDOBA67a shows identical ligand binding as KDM6B/GSK-J1. (a) Representation of the ligand binding pocket in KDM6C/UTY (pdb 5A1L, rendered in pink) in complex with KDOBA67a (cyan). Amino acid side-chains lining the active site are depicted as sticks. The position of the catalytic metal ion and water molecules are shown as orange/brown spheres. The 2FoFc electron density map is contoured at 1.0 sigma. (b) KDOBA67 occupies part of the cofactor site in KDM6 enzymes, demonstrated through an overlay of UTY (pdb 5A1L, pink) with KDM6B (grey) in complex with the 2'oxoglutarate cofactor (yellow) (PDB 2XUE). (c) KDOBA67 shows equivalent occupation and residue contacts in the active site of KDM6 enzymes, demonstrated through an overlay of KDM6C/KDOBA67 and KDM6B/GSK-J1 (magenta) (PDB 4ASK) complexes.
2
A DMSO GSK-J4 GSK-J5 KDOBA67 B MUG 72h
100
** ***
s
l l
e 75 *
c
f
G
o
e
U g
a 50 t
M ***
n
e
c
r e
P 25 * *** 0
DMSO GSK-J4 GSK-J5 KDOBA67
t n
u UM-Chor 72h o 100
C ***
s
r
l
l o
e 75
c
h
f
o
C
e
- g
a 50
t
M n
e
c
U
r e
P 25
*** 0 DMSO GSK-J4 GSK-J5 KDOBA67
SubG1 S G2 G1 Propidium Iodide
C DMSO GSK-J4 GSK-J5 KDOBA67 D MUG 72h
20
s l
l **** e
c 15 ***
+
I
P
G
+ V
U 10
n
i
x
M
e n
n 5
A
e
% d i 0
d DMSO GSKJ4 GSKJ5 KDOBA67
o
I
m u
i UmChor 72h
d i
p 20 ****
o
s
l
l
r
e
c r
P 15
+
I
o
P
h +
V 10
C
n
i
- x
e ***
n M
n 5
A
U % 0 DMSO GSKJ4 GSKJ5 KDOBA67
Annexin V - APC E F
UCH1 48h MUG 48h UM-Chor 48h UCH1 48h MUG 48h UMChor 48h 100 100 100 ** *** ** * 15 10 6
** * * ****
s s s
s
l l l s
l **
s
l l l
l **
l l
l
l e e e
e 8
e *
e c c
75 75 75 c
c
c
c
+ + +
f
f I I I
f 10 4
o
o
P P P
o * 6
+ + +
e
e e
V V V
g
g
g
n n n
a
a
i i 50 i
50 a 50
t
t
t
x x
x 4
n
n
n
e e e
e 5 2
e
e
n n n
c
c
c
n n n
r
r
r
e A A
A 2
e
e P
P 25
25 P 25
% % % 0 0 0 ** DMSO GSKJ4 GSKJ5 KDOBA67 DMSO GSKJ4 GSKJ5 KDOBA67 DMSO GSKJ4 GSKJ5 KDOBA67 0 0 0 DMSO GSK-J4 GSK-J5 KDOBA67 DMSO GSK-J4 GSK-J5 KDOBA67 DMSO GSK-J4 GSK-J5 KDOBA67
Fig. S2. H3K27 lysine demethylase inhibitors induce cell cycle arrest and apoptosis in chordoma cell lines. Cell cycle changes (A-B) and cell death analysis (C-D) in response to the compounds after 72 hours of treatment in UM-Chor1 and MUG-Chor1. By 72 hours, there is evidence of apoptosis as determined by an increase in sub- G1 and in AnnexinV-PI positive population in response to GSK-J4 and KDOBA67. Cell cycle changes (E) and cell death analysis (F) in response to the compounds after 48 hours of treatment in UCH1, UM-Chor and MUG- Chor. By 48 hours, there is evidence of G1 arrest as determined by an increase in G1 population and reduction in 3
the S population together with a slight increase AnnexinV-PI positive population in response to GSK-J4 and KDOBA67. Representative histogram/dot plots and quantification of 3 independent experiments, with 3 replicates per condition. *p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001.
.
4
A UCH1 B UM-Chor l l l l l l tr O tr O tr tr O O tr tr O O C K C K C C K K C C K K
KDM6A KDM6B KDM6A KDM6B
C Ctrl KDM6A KO KDM6B KO KDM6A/B KO
UCH1
D UM-Chor E Ctrl KDM6A KO KDM6B KO KDM6A/B KO **** 250 n.s.
) n.s m
n 200 .
0
9
5
/ 0
6 150
5
(
e c
n 100
e
c
e r
o 50
u
l F 0 Ctrl KDM6A KO KDM6B KO KDM6A/B KO UM-Chor
F MUG G Ctrl KDM6A KO KDM6B KO J
400 * TBXT - MUG-Chor
H )
**** D 2.0 P
m ***
A
n
0 G
9 300
r 5
e 1.5
/
v
0
o
6
5
n
( o
200 i 1.0
e
s
c
s
n
e
r
e
p c
x 0.5
e r
100 e
o
A
u
l
N F
R 0.0
0 m DMSO KDOBA67 Ctrl KDM6A KO KDM6B KO MUG-Chor H I K
TBXT - UM-Chor TBXT - UM-Chor
H
H D
D 6 * 1.5
P P
A ***
A
G
G
r
r
e
e v
v 4 1.0
o
o
n
n
o
o
i
i
s
s
s
s
e
e r
r 2 0.5
p
p
x
x
e
e
A
A
N N
R 0.0
R 0 m m Control KDM6A/B KO DMSO KDOBA67
Fig. S3. CRISPR/Cas9 mediated KDM6A and KDM6B knockout. (A-B) Cleavage assay for UCH1 and UM- Chor for cells transduced with empy-Lenti-Cas9-vector, guideRNA-directed at KDM6A or KDM6B, show efficient gene editing at the predicted site. (C) Bright field images of UCH1 with CRISPR editing of KDM6A/B or both. (D-E) CRISPR/Cas9 knock-out of KDM6A and KMD6B induces cell growth arrest in UM-Chor chordoma cell line. Cell viability and representative images of chordoma cells transduced with empy-Lenti-Cas9-vector, guideRNA-directed towards KDM6A, KDM6B or co-transduced with both. (F-G) In MUG-Chor cell line knockout of KDM6A is sufficient to reduce cell viability as MUG-Chor contains a biallelic deletion of KDM6B (H) as visualised by genome sequencing generated from the ChIP input. (I) Expression of TBXT is reduced in UM-Chor upon double KO of KDM6A/B, as assessed by qPCR. Quantification of 2 independent experiments, with 2
5
replicates per condition. (J-K) Expression of TBXT is reduced in chordoma cell lines as assessed by qPCR. *p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001.
6
Fig. S4. Methylation profiles of TBXT promoter in a range of solid tumours. Data from the Cancer Tumor Genome Atlas (TCGA) showed that the promoter of TBXT is unmethylated is a variety of malignancies, known not to express TBXT. Low levels of methylation are prominent in cervical squamous cell carcinomas, diffuse B cell lymphomas and adenocarcinomas of the gastrointestinal tract, tumours which are rarely reported to express TBXT protein
7
A B MUG-Chor C MUG-Chor (replicate)
D UCH1 E UCH7
Fig. S5. Transcriptomic analysis of chordoma cell lines in response to KDOBA67. (A) PCA plot of all RNA sequencing data for UCH1, UCH7, MUG-Chor, UM-Chor. Two independent datasets were generated for MUG- Chor, and the very close clustering shows high reproducibility of the data. (B-E) Volcano plot summarising differential expression in chordoma cells following treatment with KDOBA67.
8
A B C D MUG 0.6 0.20 UM-Chor
INHBE H
Mt1X H
H
H
D
D
D
D P
15 60 P 0.15
P
P A
**** 0.4 A
A
A
G
G
G
G
r
r
e
e r
*** r TBXT
v
v e
e 0.10
o
v o 10 v 40 ATF4
o 0.2
o n
n TBXT
o
n
n
o i
DDIT3 i
o
o
s
s i
i 0.05 ATF4
s
s
s s
e INHBE
e
s
s r
r DDIT3
e
e
p
r r
5 20 p
x
x p
p 0.0003 INHBE
e
e
x
x
e
e e
e 0.0003
n
n
A
A e
0.0002 e
N N
G 0.0002
G R
0 R 0 m DMSO KDOBA67 m DMSO KDOBA67 0.0000 0.0000 0 3 6 24 48 0 3 6 24 48 Time (hours) Time (hours)
E DDIT3 F ATF4 DDIT3 G ATF4 DDIT3
H
H
H
H
H
D
D
D D
D 8 8 40 3 10
P
P
P P
P **** *
A
A
A
A
A
O
O
O O O * *
G ** *
G G
**** *** G
G
S
S S
* * S
S
r
r r
r 8
r
M
M
M
M
e
e M
6 e e 30
e 6
D
v
v
D
D
v
D
v v
D 2
o
o
o
o
o
o
o
o
o
o
t
t t
t 6
n
t
n
n
n
n
o
d
o
d
o
d
o
d
i
i
o
d i
4 i e
e 20
i e
4 e
s
s
e
s
s
s
s
s
s
s
s
i
s
i
s
s
s
i
i
l
l s
i 4
l
l
e
e
l
e
e
a
a
r
e
r a
a 1
r
r
a
r
p
p
m
p
p
m
m
m
p r
2 x
m
x
r
x
r x
2 r 10
x
r
e
o
e
o e
e 2
o
o
e
o
n
n
n
n
A
A
n
A
A
A
N
N
N
N N
R 0
R 0 R R 0 0
R 0 m
m 7 7 m O B 7 7 m 7 7 O IB 6 6 7 7 m 7 7 I 6 6 O IB 6 S O IB 6 6 O IB 6 S R A A S 6 R A A S S 6 M S R A A M IS B B R A A R A A D I B B M IS B B D M IS B B M IS B B O O D O O D D D D O O D D O O O O D D K K D D D D K K K K K K K K + + + + + IB B B B B R I I I I S R R R R I IS IS IS IS
INHBE INHBE TBXT INHBE TBXT
H
H
H
H
H
D
D
D D
2.0 15 1.5 D 3 1.5
P
P
P
P
P
A
A
A
A
A
O
O O O n.s.
** ** O *** ***
G
G G
G n.s.
G
S S
** S S
S ****
r
r
r
r
r
M
M
M
M
e e
1.5 e
M
e
e
v
D
v
D
v
D v
D ** * v
1.0 D 1.0
o o 10 o
o 2
o
o
o
o
o
o
t
t
t
t
t
n
n
n
n
n
o
d
o
d
o
d
o
d
i
i
o
d i
1.0 i
e
e
e
i
e
s
s
s
e
s
s
s
s
s
s
s
i
s
i
s
i
s
s
i
l
l
l
s
i
l
e
e
e
l
e
a
a
a
e
r
r
r a
r 0.5 0.5
5 a
r 1
p
p
p
p
m
m
m
m
p
r r
0.5 r
x
x
m
x
r
x
r
x
e
e
e
o
o
o
e
o
e
o
n
n
n
n
A
A
A
n
A
A
N
N
N
N
N
R R 0.0 0.0 R 0.0
R 0
R 0
m
m m O B 7 7 m 7 7 O B 7 7 O B 7 7 S I 6 6 O IB 6 6 S I 6 6 m 7 7 S I 6 6 R A A S R R A A O IB 6 6 R A A M S B B M A A M S B B S R M S B B D I IS B B D I M A A D I O O D O O O O IS B B O O D D D D D D D O O D D K K K K K K D D K K + + + K K + IB B IB + IB R I R B R S R S I S I IS I R I IS
Fig. S6. The ATF4-response is activated in patient-derived primary chordoma cell cultures and two additional chordoma cell lines but does not contribute to TBXT suppression in response to KDOBA67. (A- B) Upregulation of ATF4-target genes in patient-derived primary chordoma cultures as assessed by qPCR (4 primary samples, at least 3 replicates per condition). (C-D) ATF4, DDIT3, INHBE and TBXT regulation over a time course in response to KDOBA67 in MUG-Chor and UM-Chor as assessed by qPCR. Results from 2 independent experiments, with 3 replicates per condition. (E-G) Co-treatment of KDOBA67 and ISRIB reduces the induction of ATF4 and target genes but does not prevent the downregulation of TBXT induced by KDOBA67, as assessed by qPCR in UCH1 (E) MUG-Chor (F) and UM-Chor (G). Results from 3 independent experiments, with 3 replicates per condition. *p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001.
9
Supplementary Tables
Table S1. Table of IC50 values against selected KDM members as determined by Alphascreen (values in M). ______
GSK-J1 KDOBA67 ______
KDM3 KDM4 KDM5 KDM5 KDM5 KDM6 JmjD1 JmjD2C JARID1A JARID1B JARID1C JmjD3 GSK-J1 >10 >10 2.5 0.4 1.3 0.16 GSK-J4 >100 >100 >100 >100 >100 >100 KDOBA67 >100 >100 24.0 28.2 62 0.24
10
Table S2: X-ray data collection and refinement statistics for human KDM6C/UTY in complex with KDOBA67 (pdb 5A1L).
Data collection Space group P212121 Cell dimensions a, b, c (Å) 90.71, 110.5 ,119.46 () 90, 90, 90 Resolution (Å) 29.31 - 2 (2.0 - 2.11) Rmerge 0.019 (0.89) Mean I / I 2.3 Completeness (%) 99.4 (99.4) Redundancy 6.2 (5.8)
Refinement Resolution (Å) 2.0 No. reflections 76980 Rwork / Rfree 0.188/0.225 No. atoms Protein 7044 Ligand/ion 104 Water 680 B-factors Protein 36.70 Ligand/ion 51.00 Water 42.70 R.m.s deviations Bond lengths (Å) 0.019 Bond angles () 1.33
11
Table S3. List of antibodies used in this study
Protein Kda Brand Cat. Species Clonality Dilution Number ChIP HA tag Abcam Ab9110 Rabbit H3 Total Abcam Ab 1791 H3K27me3 Millipore 07-449 Rabbit IgG control (anti-mouse Dako Z0259 Rabbit Polyclonal IgG (Z0259) H3K9me2 Millipore 17-648 H3K4me3 Millipore 07-473 Rabbit Polyclonal H3K4me3 Abcam ab8580 Western Blot B-actin 42 Sigma A5441 Mouse Monoclonal 1:5000 Brachyury (A-4) 47 Santa Cruz Sc-374321 Mouse Monoclonal 1:1000 Biotechnology Immunofluorescence Brachyury (A-4) Santa Cruz Sc-374321 Mouse Monoclonal 1:250 Biotechnology Anti-mouse-Alexa488 Thermo Fisher Rabbit 1:500 Scientific
12
Table S4. List of primers used in this study.
Gene Application Fw Primer Rev primer Brachyury qPCR CCCGTCTCCTTCAGCAAAGTC TGGATTCGAGGCTCATACTTATGC GAPDH qPCR GCACCGTCAAGGCTGAGAAC TGGTGAAGACGCCAGTGGA ATF4 qPCR ATGACCGAAATGAGCTTCCTG GCTGGAGAACCCATGAGGT DDIT3 qPCR ATGAACGGCTCAAGCAGGAA GCAGATTCACCATTCGGTCAA INHBE qPCR CTCTGCCCTCTAGTGGCTTGA CGCCTCGGTTGTCCAGTAA MT1X qPCR ACGCTTTTCATCTGTCCCGC GGGTCCATTTCGAGGCAAGG KDM6A qPCR AGCGCAAAGGAGCCGTGGAAAA GTCGTTCACCATTAGGACCTGC KDM6B qPCR GACCCTCGAAATCCCATCACAG GTGCGAACTTCCACGGTGTGTT Brachyury ChIP-PCR GAGAAGGGAGAGAAGGAGCC TTTCCCGCGGACATGGTAT promoter 1Kb Brachyury ChIP-PCR AAGCATGTTTTCGGTGGGAC TAACAATGCGCTTCCTTGGG promoter 2Kb Brachyury exon 3 ChIP-PCR AACGCCATGTACTCCTTCCT CCGTTGAGCTTGTTGGTGAG CRISPR guide CRISPR CAGCATTATCTGCATACCAG KDM6A CRISPR guide CRISPR GACAAAAGTACTGTTATCGG KDM6A
13
Table S5. STR (Short Tandem Repeat) analysis results for UCH1, MUG-Chor, UM-Chor and UCH7 used in the study (latest results obtained in March 2018). UCL profiles Database profiles Cell line Marker Allele 1 Allele 2 Allele 1 Allele 2 UCH1 AMEL X Y X Y CSF1PO 10 11 10 11 D13S317 11 13 11 13 D16S539 12 13 12 13 D18S51 15 15 15 15 D21S11 28 29 28 29 D3S1358 15 15 15 15 D5S818 11 12 11 12 D7S820 9 12 9 12 D8S1179 10 15 10 15 FGA 20 21 20 21 Penta D 11 11 11 11 Penta E 7 10 7 10 TH01 7 7 7 7 TPOX 8 11 8 11 vWA 17 17 17 17 AMEL X X X X MUG-Chor CSF1PO 11 11 11 11 D13S317 11 11 11 11 D16S539 11 14 11 14 D18S51 17 23 17 23 D21S11 29 33.2 29 33.2 D3S1358 14 17 14 17 D5S818 11 12 11 12 D7S820 8 11 8 11 D8S1179 11 12 11 12 FGA 21 26 21 26 Penta D 13 13 13 13 Penta E 5 12 5 12 TH01 9.3 9.3 9.3 9.3 TPOX 8 8 8 8 vWA 15 15 15 15 UM-Chor AMEL X Y X Y CSF1PO 11 12 11 12 D13S317 12 12 12 12 D16S539 12 12 12 12 D18S51 14 14 14 14 D21S11 27 31 27 31 D3S1358 18 18 18 18 D5S818 9 13 9 13 D7S820 11 11 11 11 D8S1179 12 13 12 13 FGA 23 23 19 23 Penta D 8 9 8 9 Penta E 7 7 7 7 TH01 7 9.3 7 9.3 TPOX 8 9 8 9 vWA 15 15 15 15 UCH-7 AMEL 17 CSF1PO 7 D13S317 30 33.2 D16S539 16 D18S51 12 15 D21S11 12 13 D3S1358 10 D5S818 10 11 14
D7S820 10 12 D8S1179 12 FGA 12 13 Penta D X Y Penta E 16 17 TH01 14 16 TPOX 8 11
15
Captions for Supplementary Datasets
Dataset S1. List of compounds used in the compound drug screen including screening concentration, supplier, references and proliferation raw data (related to Fig. 1A). Attached as separate file (Excel).
Dataset S2. List of differentially expressed genes in response to treatment with KDOBA67 for 48 hours. Attached as separate file (Excel).
16
Supplementary Text
SI Materials and Methods
Cell lines - Five human chordoma cell lines, U‐CH1, U‐CH2, U‐CH7, MUG‐Chor and UM‐Chor were cultured as described previously (1). All chordoma cell lines derive from sacral tumours except UM- Chor which was generated from a clival chordoma (www.chordomafoundation.org). U2OS (ATCC® HTB96™, ATCC, VA, USA), an osteosarcoma cell line was cultured according to ATCC guidelines. Cells lines were quality controlled by short‐tandem‐repeat (STR) analysis (DNA Diagnostic Centre, London, UK) (last report in Table S5) and were regularly tested to ensure that they were mycoplasma- free. In order to perform ChIP for H3.3, cells lines (UCH1, UM-Chor, MUG-Chor and U2OS) were transduced with a lentiviral vector containing full lenght human H3F3A tagged with Hemagglutinin (HA). Transduced cells were FACS-sorted based on a fluorescent marker and processed for ChIP analysis.
Chordoma primary tumour samples - Fresh sacral chordoma samples were obtained with consent from patients being treated at the Royal National Orthopaedic Hospital (RNOH), London, United Kingdom. Tumour diagnoses were made according to the WHO classification (2). Ethical approval was obtained from the Cambridgeshire 2 Research Ethics Service (reference 09/H0308/165). The samples were collected through the UCL Biobank for Health and Disease which is covered by the Human Tissue Authority licence 12055: project EC17.1. Fresh chordoma samples were obtained within 30 minutes from being removed from the patient, washed in PBS supplemented with antibiotics (Primocin™, 100 µg/ml, Invivogen, Tolouse, France), minced and incubated with Collagenase II (100 mg / ml, 17101-015, Thermo Fisher Scientific) for one hour. The digested tissue samples was then passed through a 70 μM cell strainer, centrifuged and then resuspended in chordoma medium (as described in reference (1)). When confluence was reached, cells were detached, passaged once and then plated at equal density for KDOBA67 treatment. The assay was repeated with 4 independent donors, each with at 3 three replicates.
Screening of small molecule inhibitors - Chordoma cells were seeded at a density of 5000 cells per well in a 96 well plate and allowed to adhere overnight. Cells were treated with either a compound or vehicle control for 3-6 days as indicated in Dataset S1. Cell viability was measured using Presto Blue 17
Cell Viability Reagent (Thermo Fisher Scientific, Loughborough, UK) according to the manufacturer’s instructions. Viability was normalised to vehicle controls within the same plate. Each assay was repeated three times, using 3 replicates and the mean decrease in viability calculated. For IC50 determination, cells were plated as above and treated with halving the dilution of the compounds with the highest concentration being 50 M. Viability was measured by Presto Blue Cell Viability Reagent (Thermo
Fisher Scientific) as above. IC50 was calculated by fitting 7-point dose response data using Prism version 5. Quoted values represent the mean of 3 independent experiments, each with 3 replicates.
Flow cytometry - Experiments were performed on an LSR Fortessa™ (Becton Dickinson, USA) running FACSDiva Software version 6 with 104 events recorded for each sample. Cell cycle analysis – Propidium Iodide (PI) staining: Sub-confluent cultures of cells were treated with compounds at the IC50 concentration for between 47 and 72 hours. Cells were harvested, washed once in phosphate buffer saline (PBS, Thermo Fisher Scientific) and counted. 2 x105 cells were fixed in 70 % ethanol in PBS on ice for 30 minutes. Fixed cells were centrifuged at 3000 rpm for 5 minutes and washed with PBS. To ensure that only DNA was stained the pellet was treated with RibonucleaseA (100 g/ml in PBS, Thermo Fisher Scientific) and subsequently stained with PI solution in PBS (50 µg/ml, Sigma- Aldrich) at room temperature in the dark for 30 minutes prior to analysis. Apoptosis Detection: Apoptosis was determined by detecting phosphatidylserine by APC-conjugated Annexin V using the APC Annexin V Apoptosis Detection Kit with PI (Biolegend, CA, USA). Briefly, following compound treatment as above, cells were harvested, washed once in PBS and 2 x105 cells resuspended in 250 µl of binding buffer containing 5 µL Annexin V-APC and 10 µl PI solution. Cells were incubated in the dark for 15 minutes before being analysed. Each assay was repeated 3 times, each with 3 replicates.
CRISPR/Cas9 knockout - Target guide sequences validated against KDM6A or KDM6B were designed as previously published (3, 4). Guides were cloned in a CRISPR/Cas9 lentiviral backbone containing a Puromycin and a Blasticidin resistance cassette, respectively, as described previously in Shalem et al (5). LentiCRISPR v2-Blast (Addgene plasmid # 83480) was a gift from Mohan Babu. LentiCRISPR v2 (Addgene plasmid # 52961) was a gift from Feng Zhang (3). The empty vectors were used as a non- targeting control. Lentiviral particles were produced as described previously (6). Chordoma cells were trasnduced with an MOI of 1 and subjected to antibiotic selection (Puromycin, sc-108071, Insight
18
Biotechnology LTD, at 2 µg / ml for UCH1 and UM-Chor, 4 µg / ml for MUG-Chor and Blasticidine, A1113903, Thermo Fisher Scientific at 10 µg / ml for UCH1 and UM-Chor, 20 µg / ml for MUG-Chor) using constructs in a sequential manner. Each knockout experiment was repeated twice for each three chordoma cell lines, with at least four replicates per condition per assay. Genome targeting efficiency was assessed in a cleavage assay using the T7 Endonuclease I according to manufacturer’s instructions (M0302, New England Biolabs, USA). Cell proliferation was assessed 72 hours after the transduction of the second CRISPR construct by Presto Blue Viability Assay (Thermo Fisher Scientific) according to manufacturer’s instruction. Each assay was repeated twice using at least five replicates. Viability was normalized to vehicle controls within the same plate.
Chromatin Immunoprecipitation - Cells were fixed in 1% formaldehyde for 10 minutes prior to quenching with excess glycine. Cells were resuspended on ice in a cell lysis buffer (10 mM Tris-HCL, 10 mM Nacl, 0.5 % NP-40, all reagents were from Sigma-Aldrich) for 15 minutes, centrifuged, and resuspended in nuclear lysis buffer (50 mM Tris-HCl, 10 mM EDTA, 1 % SDS, all from Sigma-Aldrich) for 10 minutes. Samples were diluted with 2 volumes of dilution buffer (16.7 mM Tris-HCL, 167 mM Nacl, 1.2 mM EDTA, 0.01 % SDS, all from Sigma-Aldrich) and sonicated on a Bioruptor Pico (Diagenode, Belgium) for 10 cycles of 30 seconds on/30 seconds off. Sonication efficiency was checked by running a sample of decrosslinked material on a 1 % agarose gel. Samples were incubated overnight rotating at 4ºC using antibody pre-incubated with protein A/G magnetic beads. Beads were washed sequentially with buffer 1 (50 mM Tris, 500 mM NaCL, 1 mM EDTA, 1 % Triton X-100, 0.1 % sodium deoxycholate, 0.1 % SDS, all from Sigma-Aldrich) three times, buffer 2 (20 mM Tris, 1 mM EDTA, 250 mM LiCl, 0.5% NP-40, 0.5% sodium deoxycholate, all from Sigma-Aldrich) three times and twice with TE buffer + 50 mM NaCl. DNA was eluted in buffer containing 50 mM Tris, 10 mM EDTA and 1% SDS before treatment with proteinase K and RNAse-A (both Thermo Fisher Scientific). DNA was eluted and decrosslinked from beads prior to purification using home made Solid Phase Reversible Immobilization (SPRI) magnetic beads (prepared as previously described (7)).
ChIPseq library preparation - ChIP was performed as above with the exception that a fixed ratio of Drosophila S2 cells (20% of chordoma cells) was spiked in prior to fixation to allow for normalization using the method of Orlando et. al (8). Libraries were prepared using the NEBNext Ultra 2 DNA Library Preparation Kit (New England Biolabs, MA, USA) and sequenced on a Nextseq500 (Illumina, CA, USA). 19
RNAseq library preparation - Cells were lysed in Trizol and total RNA extracted using the Direct-zol kit (Zymo Research, CA, USA) including an on-column DNA digest. Poly(A) RNA was selected using the NEBNext Poly(A) mRNA Magnetic Isolation Module (New England Biolabs) and a first strand library prepared using NEBNext Ultra Directional RNA Library Prep Kit (New England Biolabs) and sequenced on a Nextseq 500 (Illumina).
Next generation sequencing data processing - Computational pipelines were used by calling scripts from the CGAT toolkit to analyze the next generation sequencing data (https://github.com/cgat- developers) (9). Briefly, FASTQ files were assessed for quality using FASTQC, aligned to GRCh37 in the case of RNAseq and a concatenated genome consisting of GRCh37 using the default options of HISAT2 (10). Differential gene expression analysis was performed using DESeq2 using cell type and treatment as factors in the model. Genes were considered to be differentially regulated based on log2 fold change and p-value < 0.05. Analysis of ChIPseq data was performed as in Tumber et al (11).
DNA Methylation Analysis - DNA was extracted from either formalin-fixed paraffin-embedded (FFPE) or fresh frozern tumours according to the manufacturer’s instructions. Between 0.5ug and 1ug was for the Infinium Methylation EPIC array (Illumina) or 450K Infinium Human Methylation array (Illumina). The DNA was run on the Infinium Methylation EPIC array (Illumina) or 450K Infinium Human Methylation array (Illumina). Data were assembled using BeadStudio (Illumina) and processed using Minfi1 using the functional normalisation protocol. Beta values were downloaded from the Cancer Genome Atlas for cancers other than chordoma.
Quantification and Statistical Analysis - Statistical parameters including the exact value of n, precision measures (mean ± SD) and statistical significance are reported in the Figures and Figure Legends. Data were judged to be statistically significant when p < 0.05 by two-tailed Student’s t test. In figures, asterisks denote statistical significance as calculated by Student’s t test (*p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001). Statistical analysis was performed in GraphPad PRISM 5.0.
Data availability - Sequencing experiments have been deposited under GEO accession number GSE120214. Methylation array data have been deposited with GEO accession number GSE119462. 20
Coordinates and structure factors for the KDM6C:KDOBA67 ligand complex are deposited with RCSB under PDB ID 5A1L.
Synthesis of compound: KDOBA67
Reagents and conditions: (a) LiAlH4, Et2O, THF, 61%; THF, 34%. 3-{[2-(pyridin-2-yl)-6-(1,2,4,5-tetrahydro-3H-benzo[d]azepin-3-yl)pyrimidin-4-yl]amino}propan-1-ol (KDOBA67).
Method: GSK-J4 was synthesised as previously reported (12). To a mixture of LiAlH4 (28 mg, 0.72 mmol) in 3 mL of Et2O at 0 °C was added dropwise a solution of GSK-J4 (100 mg, 0.24 mmol) in 2 mL of Et2O and 2 mL of THF. After 1 h at 0 °C the mixture was filtered and the filtrate was washed with
CH2Cl2 (2x 20mL). The combined organic extracts were dried over Na2SO4, filtered, and evaporated in vacuo. The residue was purified by flash chromatography on silica gel (CH2Cl2/MeOH 30:1) to afford the title compound (59 mg, 61%). + MS (ESI): m/z calcd. for (C22H25N5O+H) 375.5, found 376.2. 1 H-NMR (300 MHz, CDCl3) δ 8.74 (m, 1H), 8.43 (m, 1H), 7.83 – 7.72 (m, 1H), 7.32 (m, 1H), 7.16 – 7.09 (m, 4H), 5.54 (s, 1H), 5.17 (s, 1H), 3.96 – 3.84 (m, 4H), 3.71 – 3.58 (m, 4H), 3.03 – 2.92 (m, 4H), 1.82 – 1.66 (m, 2H).
IC50 Determination of GSK-J1 analogs - The potency of the GSK-J1 bioisostere KDOBA67 was determined against JMJD3, JARID1A, JARID1B, JARID1C, JMJD2C and JMJD1A (KDM families 3- 6) using the ALPHAscreen (Amplified Luminescence Proximity Homogenous Assay) as previously described (13). This assay detects the product methyl mark using a methyl mark selective antibody coupled to a protein A acceptor bead and a streptavidin donor bead that captures a biotinylated peptide. Recombinant Jmj enzymes were produced in Ecoli or a baculovirus system (13, 14). Enzyme reactions
21
were stopped by addition of ethylenediaminetetraacetic acid (EDTA) and the presence of the biotinylated product methyl mark was detected using ALPHAscreen platform (Perkin Elmer). Inhibition assays were performed in a 384-well plate format using white ProxiPlates (Perkin Elmer). Hydroxy ethyl piperazine ethane sulfonic acid (HEPES) buffer was from Life Technologies. Anti Histone-H3-K9Me1 and anti Histone-H3-K9Me2 antibodies were from Abcam, anti Histone-H3-K27Me2 antibody was from Millipore, and anti Histone-H3-K4Me2 antibody was from Cell Signaling Technology. Alphascreen General IgG detection kit was from Perkin Elmer. L-ascorbic acid (L-AA), ferrous ammonium sulphate (FAS), α-Ketoglutaric acid (α-KG) and bovine serum albumin (A7030) were from Sigma Aldrich. All in vitro assays were performed in assay buffer (50 mM HEPES pH7.5 containing 0.1% BSA and 0.01% Tween-20). Serial titrations of compounds (0.1 µl) were transferred into wells of a 384-well ProxiPlate using an Echo Acoustic dispenser (Labcyte) and Jmj enzyme (4.9 µl at 2X final assay concentration) was dispensed into each well and allowed to pre-incubate with compound for 15 minutes. Enzyme reactions were initiated by addition of 5 µl of a 2x substrate mix consisting of FAS, L-AA, α-KG and biotinylated peptide and the enzyme reaction progressed for the indicated times. The enzyme reaction was stopped by addition of 5 µl of assay buffer containing EDTA (7.5 mM final assay concentration) and NaCl (200 mM final assay concentration). Antibody to the product methyl mark was pre-incubated in the presence of Protein A acceptor beads (0.08 mg/ml) and Streptavidin donor beads (0.08 mg/ml) for 1 hour and the presence of biotin-H3-product in the assay was detected by addition of 5 µl of the pre-incubated alphascreen beads (final concentration of 0.02 mg/ml for acceptor and donor beads). After 1 hour incubation at room temperature the assay plates were read in a BMG Pherastar FS plate reader. Data were normalized to the no-enzyme control and the IC50 were determined from the nonlinear regression curve fit using GraphPad Prism 5.
Structure determination of KDM6C (UTY) in complex with KDOBA67 - A construct of human KDM6C/UTY (encompassing aa residues Leu878–Ser1347) was expressed in Sf9 cells with a tobacco etch virus (TEV) protease-cleavable C-terminal His10-tag, essentially as described (4). Generation of recombinant baculo virus, insect cell culture, and infections were performed according to the manufacturer’s instructions (Invitrogen). The cells were suspended in a buffer containing 50 mM HEPES pH 7.5, 500 mM NaCl, 10 mM imidazole, 5% glycerol, 0.5 mM TCEP and a protease inhibitor mix (Calbiochem), and purified using nickel affinity chromatography with a stepwise gradient of imidazole after sonication. The eluted protein was then incubated with TEV protease at 4 °C overnight, followed by size exclusion chromatography (Superdex 200). The TEV protease and uncleaved protein were 22
removed using nickel affinity chromatography, and the mass of the cleaved KDM6/UTY protein was verified by electrospray mass ionization time-of-flight mass spectrometry (Agilent LC/MSD).
Protein crystallisation, X-ray Data collection and structure determination - To crystallise KDM6C/UTY with KDOBA67a, the protein (11 mg/mL) was preincubated with 1 mM of the inhibitor and 2 mM MnCl2. The protein-compound mixture was then transferred to crystallization plates using the sitting drop vapour diffusion method at 4 °C. Protein crystals were obtained in a drop consisting of 100 nL protein-compound and 100 nL of a precipitant consisting of 25% (w/v) PEG 3350, and 0.1 M bis-tris pH 6.5. Crystals were cryoprotected with mother liquor supplemented with 25% ethylene glycol before they were flash-frozen in liquid nitrogen. A data set with a diffraction limit of 2.0 Å was collected on beamline I04-1 at the Diamond Light source. The data were processed with XDS, scaled and merged using Aimless. The structure was determined by molecular replacement using UTY PDB ID 3ZLI (15) as a search model. The protein-ligand model was further improved by subsequent cycles of model building in coot and refinement in REFMAC (16). The refinement was terminated when there were no significant changes in the Rwork and Rfree values and inspection of the difference density maps suggested that no further corrections or additions were justified. The quality of the structure was assessed with the MolProbity server and deposited in Protein Data Bank with the PDB code 5A1L. Data collection and refinement statistics are given in Table S2.
23
Supplementary Acknowledgements We would like to thank the UCL Experimental Cancer Centre which provided support from the Genomics and Genome Engineering, Microscopy and Imaging, Pathology Services and Flow Cytometry Core facilities. AMF is a NIHR senior investigator and was also supported by the National Institute for Health Research, UCLH Biomedical Research Centre. The Structural Genomics Consortium is a registered charity (number 1097737) that receives funds from Abbvie, Bayer Healthcare, Boehringer Ingelheim, the Canadian Institutes for Health Research, the Canadian Foundation for Innovation, Eli Lilly and Company, Genome Canada, GlaxoSmithKline, the Ontario Ministry of Economic Development and Innovation, Janssen, the Novartis Research Foundation, Pfizer, Takeda, and the Wellcome Trust.
24
Supplementary References
1. Scheipl S, et al. (2016) EGFR inhibitors identified as a potential treatment for chordoma in a focused compound screen. J Pathol 239(3):320–334. 2. Flanagan AM, Yamaguchi T (2013) Chordoma. WHO Classification of Tumours of Soft Tissue and BoneTitle, pp 328–329. 3. Sanjana NE, Shalem O, Zhang F (2014) Improved vectors and genome-wide libraries for CRISPR screening. Nat Methods 11(8):783–784. 4. Ohguchi H, et al. (2017) KDM6B modulates MAPK pathway mediating multiple myeloma cell growth and survival. Leukemia 31(12):2661–2669. 5. Shalem O, et al. (2014) Genome-scale CRISPR-Cas9 knockout screening in human cells. Science (80- ). doi:10.1126/science.1247005. 6. Lampada A, et al. (2017) MTORC1-independent autophagy regulates receptor tyrosine kinase phosphorylation in colorectal cancer cells via an mTORC2-mediated mechanism. Cell Death Differ 24(6):1045–1062. 7. Rohland N, Reich D (2011) Cost-effective , high-throughput DNA sequencing. Genome Res:939–946. 8. Orlando DA, et al. (2014) Quantitative ChIP-Seq normalization reveals global modulation of the epigenome. Cell Rep 9(3):1163–1170. 9. Sims D, et al. (2014) Genome analysis CGAT : computational genomics analysis toolkit. 30(9):1–2. 10. Pertea M, Kim D, Pertea GM, Leek JT, Salzberg SL (2016) Transcript-level expression analysis of RNA-seq experiments with HISAT, StringTie and Ballgown. Nat Protoc 11(9):1650–1667. 11. Tumber A, et al. (2017) Potent and Selective KDM5 Inhibitor Stops Cellular Demethylation of H3K4me3 at Transcription Start Sites and Proliferation of MM1S Myeloma Cells. Cell Chem Biol 138(0):981–992. 12. Kruidenier L, et al. (2012) A selective jumonji H3K27 demethylase inhibitor modulates the proinflammatory macrophage response. Nature 488(7411):404–408. 13. Kawamura A, et al. (2010) Development of homogeneous luminescence assays for histone demethylase catalysis and binding. Anal Biochem 404(1):86–93. 14. Hatch SB, et al. (2017) Assessing histone demethylase inhibitors in cells: Lessons learned. Epigenetics and Chromatin 10(1):1–17. 15. Walport LJ, et al. (2014) Human UTY(KDM6C) is a Male-Specific Nϵ-Methyl Lysyl- Demethylase. J Biol Chem. doi:10.1074/jbc.M114.555052. 16. Lo WW, Pinnaduwage D, Gokgoz N, Wunder JS, Andrulis IL (2014) Aberrant hedgehog signaling and clinical outcome in osteosarcoma. Sarcoma 2014:1–10.
25