MCT1-mediated transport of a toxic molecule is an effective
strategy for targeting glycolytic tumors
Kivanc Birsoy,1,2 Tim Wang,1,2,3 Richard Possemato,1,2 Omer H. Yilmaz,1,2 Catherine E.
Koch,1,2 Walter W. Chen,1,2 Amanda W. Hutchins,1,2 Yetis Gultekin,1,2 Tim R. Peterson,1,2
Jan E. Carette,1,6 Thijn R. Brummelkamp,1,7 Clary B. Clish,3 and David M. Sabatini1,2,3,4,5
1Whitehead Institute for Biomedical Research, Nine Cambridge Center, Cambridge, MA 02142, USA.
2Department of Biology, Massachusetts Institute of Technology (MIT), Cambridge, MA 02139, USA.
3Broad Institute, Seven Cambridge Center, Cambridge, MA 02142, USA.
4David H. Koch Institute for Integrative Cancer Research at MIT, 77 Massachusetts Avenue, Cambridge, MA 02139, USA.
5Howard Hughes Medical Institute, MIT, Cambridge, MA 02139, USA.
6Current address: Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA 94305, USA.
7Current address: Department of Biochemistry, Netherlands Cancer Institute, Plesmanlaan 121 1066 CX, Amsterdam, The Netherlands.
To whom correspondence should be addressed. Email: [email protected]
Nature Genetics: doi:10.1038/ng.2471
cleavedCASPASE3
RPS6
Clone BClone A
WT KBM7
0.66% 3.06%
2.94%96.34% 65.16% 18.9%
97.26% 96.17%
1.93% 1.99%
1.65% 1.25%
95.42% 94.60%
3.01% 3.66%
1.26% 1.09%
7AA
D
Annexin V-PE
PBS 50uM 3-BrPa
Wild
Typ
eC
lone
AC
lone
B
a b
0 200 400 600 800 1000 1200
-log
(p_v
alue
)
Genes ranked by chromosomal position
0.1
1
10
100
1000
SLC16A1
BSG
69 insertions
26 insertions
Supplementary Figure 1 - Loss of SLC16A1 and BSG prevents 3-BrPA induced cell death.
a. SLC16A1 and BSG genes contain the highest degree of insertional enrichment in 3-BrPA selected cells compared to the unselected control cells (p=4.7E-121 and p=5E-29, respectively). Y axis represents the inverse logarithm of p values, calculated by Fisher Exact Test. The X-axis represents the insertion sites ordered by their genomic position. The diameter of the bubbles denotes the number of insertions for each gene. b. Wild type and MCT1 null KBM7 cells were treated for 3 hours with 3-BrPA (50uM) and FACS analysis have been performed using 7AAD and Annexin V staining.
Nature Genetics: doi:10.1038/ng.2471
0 10 20 30 40 500
50
100
150
OC
R(%
)
Clone A50uM 3-BrPa
Wild Typ
e
Clone A
0
200
400
600
800
AUC
Ano
va E
CAR
(mpH
)Wild
Type
Clone A
0
2000
4000
6000
8000
AUC
Ano
va O
CR
(pM
oles)
OXYGEN CONSUMPTION RATE
p=0.21ns
EXTRACELLULAR ACIDIFICATION RATER
elativ
e Lac
tate P
rodu
ction
Wild Typ
e
Clone A
0.0
0.5
1.0
1.5
a
b
LACTATE PRODUCTION
Supplementary Figure 2 - Metabolic characterization of Wild Type and MCT1 null KBM7 cells
a. Lactate production of wild type and MCT1 null KBM7 cells were measured using a colorimetric assay and normalized by cell number. ECAR and OCR reading were measured using Seahorse Extracellular Flux Analyzer. AUC (area under curve) converts OCR rate data to accumulation of total oxygen consumed upon three separate readings.b. Extracellular Flux Analysis of wild type and MCT1 null KBM7 cells upon 3-BrPA (50 uM) addition. Changes in OCR, a proxy for oxygen consumption, were monitored upon the addition of 50 uM 3-BrPA. Results are displayed as a percentage of the OCR reading immediately before 3-BrPA addition. Error bars are SEM.c. Relative cellular ATP levels during a time course following 3-BrPA treatment. 50uM 3-BrPA was added onto wild type and MCT1 null KBM7 cells. Relative ATP levels were measured using a luciferase based assay for 2 hours. Error bars are SEM.
p=0.14ns
p=0.39ns
c
*
0 25 50 75 100 1250
50
100
150
Time (min)
% A
TP le
vels
WT KBM7Clone A
WT KBM7
Time (min)
Nature Genetics: doi:10.1038/ng.2471
KBM7
MDA-MB-468
GAPDH
Raptor
shRNA: GAPDH_1GAPDH_2
GAPDH_1GAPDH_2
GFPGFP
KBM7
PBS 3-BrPA0
50
100
150
Rela
tive C
ell N
umbe
r
shGFPshGAPDH_1shGAPDH_2
MDA-MB-468
PBS 3-BrPA0
50
100
150
Rela
tive C
ell N
umbe
r
shGFPshGAPDH_1shGAPDH_2
GAPDH
Raptor
shRNA:
Supplementary Figure 3 - Effect of GAPDH silencing on 3-BrPA sensitivity Silencing GAPDH expression in cancer cell lines using RNAi causes a decrease in cell proliferation. Cells were infected with corresponding shRNAs and grown in RPMI media in presence and absence of 3-BrPA. 50uM and 10uM 3-BrPA were used for MDA-MB-468 and KBM7, respectively. Note the synergistic effect of GAPDH inhibition and 3-BrPA treatment on cell number.
Nature Genetics: doi:10.1038/ng.2471
pH.7.
0pH
.5.5
3-BrPa (
10mM)
Pyruva
te (10
mM)
Lacta
te (10
mM)0
100
200
300
0.1 1 10 100 10000
50
100
150 Pyruvate
L-Lactate
D-Lactate
3-BrPa
a b
Supplementary Figure 4 - Pyruvate and Lactate can compete with 3-BrPA uptake in KBM7 cells.
a. pH dependence of MCT1 mediated 3-BrPA transport. Wild Type KBM7 cells were incubated with 100 uM of radiactively labeled 3-BrPA for 20 minutes in HBSS in presence of indicated pH conditions and various monocarboxylates (n=3). b. Dose dependent inhibition of labeled 3-BrPA transport by D-Lactate, L-Lactate and Pyruvate. Wild Type KBM7 cells were incubated with 100 uM radiactively labeled 3-BrPA for 20 minutes in presence of different concentrations of indicated monocarboxylates.
% re
sidu
al 1
4C-3
-BrP
A si
gnal
% re
sidu
al 1
4C-3
-BrP
A si
gnal
Nature Genetics: doi:10.1038/ng.2471
1.0 1.5 2.0 2.5 3.00
5
10
15 Slc16a3
1.0 1.5 2.0 2.5 3.04
5
6
7
8 Slc16a7
1.0 1.5 2.0 2.5 3.03.43.63.84.04.24.44.6 Slc16a8
1.0 1.5 2.0 2.5 3.03.5
4.0
4.5
5.0 Slc5a8
1.0 1.5 2.0 2.5 3.03.03.23.43.63.84.04.2 Slc5a12
r=0.11p=0.97
r=-0.47p=0.07
r=-0.2p=0.45
r=-0.08p=0.75
r=-0.01p=0.94
MCTs
SMCTs
Supplementary Figure 5 - Correlation of other MCT transporter expression levels with 3-BrPA sensitivity
The concentration of 3-BrPA at which 50% cell growth inhibition occurred after 3 days of administration (IC50) was determined for a panel of cancer cell lines. These values were correlated with transcriptome-wide mRNA expres-sion data from the Cancer Cell Line Encyclopedia (CCLE) and the resulting Pearson correlation coefficients were plotted. Relative expression levels for known lactate transporters (MCTs and SMCTs) were correlated with the corresponding IC50 values for 3-BrPA of each cell line.
Rel
ativ
e e
xpre
ssio
n (lo
g2)
Log[IC50] Rel
ativ
e e
xpre
ssio
n (lo
g2)
Rel
ativ
e e
xpre
ssio
n (lo
g2)
Rel
ativ
e e
xpre
ssio
n (lo
g2)
Rel
ativ
e e
xpre
ssio
n (lo
g2)
Log[IC50] Log[IC50]
Log[IC50] Log[IC50]
Nature Genetics: doi:10.1038/ng.2471
CNS
Cell Li
ne
Haemato
poeti
c
Cell Li
neKidn
ey
Cell Li
neLiv
er
Cell Li
ne
Pancre
as
Cell Li
ne
Prostat
e
Cell Li
ne Skin
Cell Li
neTes
tis
Cell Li
ne Lung
Cell Li
neBrea
st
Cell Li
neColo
n
Cell Li
ne
Thyroi
d
Cell Li
ne
Stomac
h
Cell Li
ne
Saliva
ry
Cell Li
ne
6
8
10
12
Tumor Cells and Normal Counterparts
Rel
ativ
e Ex
pres
sion
(Log
2)
a
GFP MCT1
MDA-MB-231 tumor xenograftsan
ti-M
CT1
c
d
Supplementary Figure 6 - MCT1 expression in cancers
a. IHC staining of MDA-MB-231 tumors expressing GFP and MCT1 cDNAs using an anti-MCT1 antibody.b. Time course weight measurements of tumors formed by MDA-MB-231 cells expressing the MCT1 or GFP cDNA during 3 weeks of treatment with vehicle or 3-BrPA (8 mg/kg). Error bars are ± SD (n = 5).c. Relative expression of MCT1 in cancer cell lines and their normal tissue of origins. Data was collected from CCLE and GEO browser and quantile normalized. d. Top 40 genes correlating with low (glycolytic) and high (OXPHOS) OCR/ECAR values.
TIMP2 PIM1 FN3KRP CATSPERBC12orf75 ORAI2 HS6ST3 ALDH3B2CLIP4 TNFAIP3 PPP2R2C KIAA1324RNF145 SCHIP1 ST6GALNAC2 GPR81UBE2E3 TBC1D1 TMEM141 MUM1L1BCAP29 PRKD3 SLC25A21 UCK1TSPAN3 SLC16A1 ZNF350 MYRIPNAB1 PDLIM1 LOC150622 ANKRD30ABET1 PDE8A CRAT UMODL1TMEM55A C20orf30 BRF1 MUCL1MDFIC CTSC GLS2 WDR85KLHL5 ROCK2 CNTNAP2 RAB3DGSTP1 SGCB ZNF485 PATZ1LTBP4 IFNAR2 NRK ZNF480C6orf145 ANXA11 CLCA2 ZMYND19CLN5 MTMR2 AR ZNF90PGM1 MFGE8 BDH1 RETPRKAA1 CDK6 RNF208 LOC643072IFT57 TYW3 LOC254559 AKT1BRI3 LDHB CEBPA ALDH4A1
Low
OC
R/E
CAR
Hig
h O
CR
/EC
AR
cDNA:
b
GFPMCT1GFPMCT1
cDNA Treatment3-BrPA
PBS
weeks0 1 3
0
50
100
150
200
Tum
or si
ze (
mm
3)
Start of thetreatment
Nature Genetics: doi:10.1038/ng.2471