NNC 26-9100. 17.88. 13.05. 4.6. 0.771. 41 .... Carlsbad, CA) was used to synthesize cDNA, and cDNA libraries were generated using Nextera XT DNA.
Cell Reports, Volume 22
Supplemental Information
Small-Molecule Screen Identifies De Novo Nucleotide Synthesis as a Vulnerability of Cells Lacking SIRT3 Karina N. Gonzalez Herrera, Elma Zaganjor, Yoshinori Ishikawa, Jessica B. Spinelli, Haejin Yoon, Jia-Ren Lin, F. Kyle Satterstrom, Alison Ringel, Stacy Mulei, Amanda Souza, Joshua M. Gorham, Craig C. Benson, Jonathan G. Seidman, Peter K. Sorger, Clary B. Clish, and Marcia C. Haigis
Figure S1
20 0
-4
-8 -7 -6 -5 log[Vinblastine Sulfate], M
WT KO
40 20 0
-8
-7
-6
-5
-4
1.8
1.2
0.6
0.0
log[Vincristine Sulfate], M
I
*
1.2
Relative NH4+ production
0.9 0.6 0.3 0.0
KO
+SIRT3
shNS
J
**
1.2
-7 -6 -5 log[Vindesine], M
1.2
shNS shSIRT3
SIRT3
0.6
tubulin shNS
shSIRT3
KO + pBabe
2000
KO + SIRT3 ***
1500
KO +SIRT3
0.6
1000
SIRT3
0.3
tubulin KO
+SIRT3
-4
p = 0.08
K
0.9
0.0
-8
G
1.8
0.0
shSIRT3
-9
**
500 0
D
Relative glutamine uptake
H
*
0
2
60
F Relative glutamine uptake
% of control
80
30
-4
Relative NH4+ production
E
60
ay
D
-9
90
D
-7 -6 -5 log[Decitabine], M
40
WT KO
1
-8
120
ay
0
60
C
D
30
WT KO
0
60
80
ay
90
B
% of control
WT KO
% proliferation
% of control
120
% of control
A
Figure S1. Chemical screen identifies glutamine metabolism as a vulnerability in SIRT3 KO cells; related to Figure 1. (A) Dose response curves of WT and SIRT3 KO MEF growth after treatment with Decitabine, (B) Vinblastine Sulfate, (C) Vindesine, and (D) Vincristine Sulfate for 72 hours. (E) Glutamine uptake and (F) ammonia production in control (shNS) or SIRT3 knockdown (shSIRT3) MCF10A cells. (G) Immunoblots of SIRT3 and tubulin in control (shNS) or SIRT3 knockdown (shSIRT3) MCF10A cells. (H) Glutamine uptake and (I) ammonia production in SIRT3 KO MEFs expressing empty vector or SIRT3. (J) Immunoblots of SIRT3 and tubulin in SIRT3 KO MEFs without and with SIRT3 overexpression. (K) Cell proliferation after SIRT3 overexpression in SIRT3 KO MEFs.
Figure S2
A
one carbon pool by folate amino sugar and nucleotide sugar metabolism
0.001
de novo pyrimidine synthesis
glutamine + bicarbonate + water + ATP
B
p-value
C
pyrimidine salvage/degradation
Pathway name
carbamoyl phosphate
0.99
aspartate
CAD
carbamoyl-aspartate
DHODH
orotate
UMP CTPS
CTP
UTP
uracil
dUDP
cytidine
dCDP
dUTP
cytosine
dCMP
NME6
DCTD
ß-alanine
RNR
CDP
ß-alanine
uridine
UDP
CMP
TMP
DUT TYMS
dUMP
dTMP
thymidine
dCytidine
* 0.5
0.0
1.0
0.5
0.0
1.0
0.5
0.0
2.0
** ***
1.0 0.5
SIRT3
+
+
-
-
azaserine
-
+
-
+
2.0
**
*
1.5 1.0 0.5 0.0
SIRT3
+
+
-
-
SIRT3
+
+
-
-
SIRT3
+
+
-
-
SIRT3
+
+
-
-
azaserine
-
+
-
+
azaserine
-
+
-
+
azaserine
-
+
-
+
azaserine
-
+
-
+
*
1.0
0.5
0.0
2.5
Relative levels
1.5
Relative 3H-thymidine incorporation to DNA
J
WT
KO
2.0
synthesis
**
*
salvage/degradation
K
WT KO
1.5 1.0 0.5 0.0
shNS
*** * *** *
sh #1
sh #2
WT KO WT KO WT KO PPAT
n.s.
actin
U M P U D P dU M P C D P C M cy P tid cy ine to th sin ym e id -a th ine m ym in o ine ac isob id ut yr ic
I
***
1.5
0.0
H
1.5
Relative malate levels
Relative succinate level
Relative orotate level
1.0
G
1.5
0.026
thymine
Relative 3H-adenine incorporation to DNA
F
1.5
Phenylalanine, tyrosine and tryptophan
5,10-methylene-THF DHF
ß-aminoisobutyric acid
E
0.0001
N-carbamoyl-L-aspartate Relative levels
OMP
0.00006
Alanine, Aspartate and glutamate
D
PRPP UMPS
0.000004
Pyrimidine metabolism
dihydroorotate
starch and sucrose pentose and glucuronate metabolism interconversions
p-value
Arginine and proline metabolism
Figure S2. Azaserine inhibits increased de novo pyrimidine synthesis in SIRT3 KO MEFs; related to Figure 2. (A) Diagram generated by MetaboAnalyst 3.0 analysis highlighting metabolic pathways significantly affected by azaserine in SIRT3 KO MEFs. (B) Schematic of pyrimidine metabolism. Shaded in pink is the de novo pyrimidine synthesis pathway, and shaded in blue are intermediates in pyrimidine degradation or the salvage pathway. (C) MetaboAnalyst 3.0 analysis comparing SIRT3 WT and KO MEFs. (D) N-carbamoyl aspartate, (E) orotate, (F) succinate (G) malate levels in WT and SIRT3 KO MEFs treated with 30 µM azaserine (or DMSO as a control) for 6 hours. (H) Incorporation of 3H-adenine, or (I) 3H-thymidine into DNA in WT and SIRT3 KO cells. WT and SIRT3 KO MEFs were serum starved for 16 hours and then labeled and stimulated by insulin for 8 hours. Induction with insulin compared to serum starved values were graphed. (J) Steady-state levels of metabolites in pyrimidine metabolism in WT and SIRT3 KO MEFs. (K) Immunoblot of control (shNS) and PPAT knockdown (sh#1 or sh#2) in WT and SIRT3 KO MEFs.
Figure S3
9 6 3
WT KO
4
***
B
KO
1.5
KO +SIRT3
ATF4 tubulin
2
SIRT3
1.0
0.5
0.0
0
serum starved
serum starved basal
*
basal
KO
+
0
6
KO SI RT 3
**
WT KO
ATF4/tubulin
12
Relative pS6/S6 levels
Relative pS6K/S6K levels
A
WT
KO
rapamycin
WT
KO
pS6 S6
2.0 1.5 1.0 0.5
SIRT3 rapamycin
F Cell Size (μm)
19 18 17 16 15 10 8 6 4 2 0
KO
WT
*
n.s.
+
+
-
-
-
+
-
+
G
WT
16.5
WT KO
16.0
***
basal
15.0 10 5 0
serum starved
KO
aspartate serine glutamine asparagine threonine methionine lysine leucine-isoleucine proline arginine
rapamycin WT + Torin/ WT KO + Torin/ KO
**
0.6
****
***
0.4 0.2 0.0
Day 2
Day 3
***
15.5
Min
0.8
Relative growth
E
*
0.0
α-tubulin
H
***
Average Cell Size (μm)
control
N-carbamoyl-L-aspartate Relative levels
D
C
Day 4
Figure S3. Hyperactive mTORC1 signaling in the absence of SIRT3 contributes to increased de novo purine synthesis, related to Figure 3. (A) Quantification of immunoblots in Figure 3A. (B) Immunoblots and quantification of ATF4 normalized to tubulin in SIRT3 KO and SIRT3 KO MEFs with SIRT3 re-expression. (C) Immunoblots of WT and SIRT3 KO MEFs treated with DMSO or 100 nM rapamycin for 24 hours. (D) N-carbamoyl aspartate levels in WT and SIRT3 KO MEFs treated with DMSO (control) or 100 nM rapamycin. (E) Cell size of WT and SIRT3 KO MEFs under serum starvation or basal conditions (n = 3). Cell size was measured using a Beckman Coulter Z2 cell counter. (F) Cell size of WT and SIRT3 KO MEFs treated with DMSO as a control, or 100 nM rapamycin for 24 hours. (G) Amino acid levels examined by steady state metabolomics in WT and SIRT3 KO MEFs. (H) Relative cell number after Torin treatment normalized to WT or KO cells; related to figure 3E.
Max
basal
Supplemental Table S1 Top 50 hits from small molecule screen to selectively inhibit SIRT3 KO MEF growth # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
Compound name Azaserine Decitabine Vinblastine Sulfate Vindesine Vincristine Sulfate Podofilox Ch 55 Mebendazole Nocodazole Vinorelbine Bitartrate Picropodophyllotoxin Retinoic acid p-hydroxyanilide Fenbendazole IC261 SB 225002 Albendazole cis-(Z)-Flupenthixol dihydrochloride JNK Inhibitor IX Amiodarone·HCl Fluphenazine Dihydrochloride 82640-04-8 Aclarubicin Zucapsaicin Prenylamine Lactate LY 2183240 Suloctidil Nortriptyline Methiothepin Mesylate Salmeterol Duloxetine Hydrochloride Itoraconazole S 14506 Hydrochloride Kenpaullone Capsaicin Spermidine Trihydrochloride Perphenazine AMN082 Plicamycin TGF-b RI Inhibitor III NNC 26-9100 Iodophenpropit Dihydrobromide Fenretinide Triamterene; 396-01-0 Flunarizine dihydrochloride Furoxan Bleomycin AG 494 WIN 55 Dipyridamole 24(S)-hydroxycholesterol
WT IC50 (µM) >29 >25 0.02 0.06 25 4.34 0.22 0.17 5.81 10.75 >17 4.84 4.29 >19 15.92 2.24 19.78 11.62 26.57 0.03 24.94 10.56 0.87 7.48 45.33 15.92 27.01 11.37 >25 56.76 29.95 28.64 8.95 15.07 33.46 1.07 28.61 17.88 18.57 >13 >25 32.55 3.96 17.21 >25 24.72 >25 >31
KO IC50 (µM) 2.91 1.28 0.01 0.01 25 1.02 0.06 0.15 2.53 4.56 2.70 0.87 1.83 4.30 12.18 1.20 10.80 9.81 15.21 0.02 13.21 9.95 0.59 5.70 35.75 10.60 19.69 9.60 >25 25.02 15.87 10.00 7.62 11.91 18.39 0.82 30.15 13.05 10.45 8.56 35.04 22.84 3.58 17.09 >25 19.02 >25 >31
AUC 63.3 60.2 46.6 45.5 42.3 41.5 39.9 38.7 35.1 34.0 30.4 28.9 28.5 26.3 24.5 23.4 23.0 22.7 21.3 20.3 18.8 18.3 18.0 17.5 15.6 15.1 12.6 12.6 11.9 9.8 9.1 8.7 8.5 8.4 8.1 7.6 7.3 5.2 5.0 4.6 4.4 3.9 3.5 3.2 1.8 0.2 -1.3 -5.3 -6.2 -12.2
p-value 0.011 0.006 0.010 0.005 0.003 0.002 0.018 0.004 0.004 0.009 0.006 0.052 0.010 0.020 0.059 0.036 0.097 0.002 0.230 0.021 0.101 0.091 0.025 0.079 0.164 0.530 0.168 0.279 0.241 0.288 0.436 0.238 0.259 0.358 0.472 0.345 0.566 0.797 0.694 0.771 0.789 0.426 0.642 0.812 0.850 0.971 0.964 0.663 0.343 0.108
Table S1. Top 50 hits from small molecule screen to selectively inhibit SIRT3 KO MEF growth. Related to Figure 1.
Supplemental Table S2
Metabolite glutamate diadenosine triphosphate N-acetylneuraminate 5-methyltetrahydrofolate fructose ADP choline uracil thymidine xanthosine 5'-monophosphate 5'- GMP acetylcarnitine (C2) orotate CDP-choline 2'-deoxyuridine guanosine 5'-diphospho-fucose NAD+ leucylglutamate AMP 3-hydroxydecanoate uridine gamma-glutamylthreonine
Fold change relative to WT Metabolic Pathway WT WT + azaserine KO KO + azaserine Glutamate Metabolism 1 0.12 1.66 0.17 Purine Metabolism 1 0.78 1.48 0.80 Aminosugar Metabolism 1 0.82 1.04 0.62 Folate Metabolism 1 0.69 0.62 0.42 Fructose Metabolism 1 1.43 0.92 2.74 Purine Metabolism 1 1.10 1.00 0.53 Phospholipid Metabolism 1 0.99 1.14 1.55 Pyrimidine Metabolism 1 1.13 0.91 1.51 Pyrimidine Metabolism 1 1.72 1.33 6.68 Purine Metabolism 1 0.80 1.48 0.55 Purine Metabolism 1 0.78 1.10 0.62 Fatty Acid Metabolism 1 0.95 0.95 0.73 Pyrimidine Metabolism 1 0.52 0.57 0.28 Phospholipid Metabolism 1 0.90 1.13 0.56 Pyrimidine Metabolism 1 0.00 0.78 2.69 Nucleotide Sugar 1 0.85 0.99 0.79 Nicotinate Metabolism 1 0.94 1.00 0.70 Dipeptide 1 0.81 1.37 1.02 Purine Metabolism 1 0.91 1.21 0.81 Fatty Acid 1 0.94 1.18 0.90 Pyrimidine Metabolism 1 1.11 0.91 1.34 Gamma-glutamyl Amino Acid 1 1.21 1.06 1.47
Table S2. Significantly affected metabolites by azaserine in SIRT3 KO MEFs. Related to Figure 2.
p-value 0.000 0.001 0.002 0.002 0.002 0.004 0.005 0.007 0.009 0.009 0.015 0.018 0.020 0.021 0.022 0.028 0.028 0.029 0.029 0.039 0.041 0.044
Supplemental Experimental Procedures. Cell Culture For 3D cultures, 4000-6000 single cells were seeded on top of a thin layer of Engelbreth-Holm-Swarm (EHS) tumor extract (Corning Matrigel) to which media containing 2% EHS was added. All 3D cultures were maintained for 10-16 days, and media was changed every 2 days after day 4.
SIRT3-overexpressing cells were generated by retroviral infection with empty pBabe vector or pBabe vector containing human WT SIRT3 with a C-terminal FLAG tag (Schwer et al., 2002). Knockdown cell lines were generated by lentiviral infection with shRNAs obtained from The RNAi Consortium (TRC) at the Broad Institute. Clone TRCN0000038892 was used to silence SIRT3 as previously described (Finley et al., 2011).
Reagents The following antibodies were obtained from Cell Signaling (unless noted otherwise) and were used for immunoblotting: SIRT3 (#5490), phospho-CAD (Ser1859; #12662), CAD (#11933), phospho-S6K (#9234), S6K (#9202), phospho-S6 (#5364), S6 (#2317), α-tubulin (Santa Cruz sc-8035), actin (Sigma A5441), PPAT (Proteintech, 15401-1-AP). Cells were treated with Azaserine (Cayman Chemical 14834), insulin (Sigma I0516), Rapamycin (Sigma R0395), or Torin (4247, R&D Systems). Radioactive metabolites were purchased from Perkin Elmer: U14C-aspartic acid (NEC268E050UC), 6-3H-thymidine (NET355250UC), and U-14C-glycine (NEC276E250UC), and 2,8-3H-adenine (NET063001MC). For tracing experiments we used 15N amide-glutamine (Sigma, 490024).
Glutamine and Ammonium Measurements Cells were plated in 6-well plates and allowed to adhere. Cells were washed once with PBS prior to change of media. Glutamine and ammonia concentrations in fresh and culture media were measured after 8 to 24 hours using the BioProfile FLEX analyzer (NOVA Biomedical). Measurements were normalized to cell number.
Metabolite Profiling Metabolites were extracted from cells in 80% ice-cold methanol on dry ice and metabolite profiles were obtained as previously described (Ben-Sahra et al., 2013; Finley et al., 2011). Metabolite levels were normalized by protein content. For glutamine labeling experiments, cells were plated and after 24 hours, cells were washed twice with PBS prior to addition of glutamine-free media supplemented with 15N -glutamine, 10% serum, and penicillin/streptomycin for the specified time. Cells were washed with cold PBS prior to polar metabolite extraction (as already described).
Aspartate and Glycine incorporation into nucleic acid Cells were plated in 6-well plates (in triplicate) and after 24 hours, cells were washed once with PBS and serum starved for 16 hours. Cells were treated with media containing 2 µCi U-14C-glycine, 2 µCi 2,8-3Hadenine, 2 µCi U-14C-aspartic acid, or 1 µCi 6-3H-thymidine, and incubated for the specified time. DNA was extracted from cells using the DNeasy Blood and Tissue kit (Qiagen) based on manufacturer’s instructions. Nucleic acid concentration was determined using a Nanodrop spectrophotometer. Isolated nucleic acid was added to scintillation vials and radioactivity was measured using a liquid scintillation counter. Data was normalized to total nucleic acid concentration.
RNA Sequencing RNA was extracted from immortalized WT and SIRT3 KO MEFs as described. Library preparation and RNA sequencing was performed as previously described (Green et al., 2016). Briefly, 2 µg RNA was incubated with poly-T oligo beads and unbound RNA was washed two times to remove ribosomal RNA and enrich for mRNA transcript pools. The Superscript III First-Strand Synthesis System (Invitrogen, Carlsbad, CA) was used to synthesize cDNA, and cDNA libraries were generated using Nextera XT DNA sample preparation kit. Libraries were sequenced on the Illumina NextSeq and data were processed as previously described (Christodoulou et al., 2014). Data were normalized to the total number of reads per kilobase of exon per million (RPKM), and reads were aligned to the mm10 mouse genome using Spliced Transcripts Alignment to a Reference (STAR) (Dobin et al., 2013).
RNA Sequencing Data Analysis To identify genes differentially expressed genes between WT and SIRT3 KO MEF samples, we utilized the DESeq2 differential gene expression algorithm as previously described (Love et al., 2014). We filtered genes significantly and differentially expressed based on a false discovery rate of less than 0.05 and an absolute log2 fold change greater than 0.05 (absolute fold change of 1.3). We also filtered for genes that were expressed at a minimum of 1 RPKM. The list of differentially expressed genes was used to identify enrichment of transcription factor binding motifs using the previously described MotifADE method (Mootha et al., 2004).
Immunoblotting Cells were washed with cold PBS and lysed on ice for 15 min in RIPA buffer (50 mM Tris, pH7.4, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS) supplemented with 1mM DTT, protease inhibitor cocktail (Roche), and phosphatase inhibitors (Sigma). Lysates were cleared by centrifugation at 16,000 x g for 10 min at 4 °C. Protein was quantified by a BCA protein assay (Thermo Fisher Scientific) and equal amounts of protein were run on a 10-20% Tris-Glycine gel (Bio-Rad) and transferred to nitrocellulose membrane (Bio-Rad) for analysis by immunoblotting. Membranes were blocked in 5% BSA in TBS containing 0.1% Tween 20 for 1 hour at room temperature and incubated with primary antibody at 4°C overnight. Band intensities were quantified using ImageJ.
Immunohistochemistry Tumors were fixed in 10% buffered formalin and submitted to the Dana-Farber/Harvard Cancer Center facility for embedding in paraffin, sectioning, and hematoxylin and eosin staining. Cyclic immunofluorescence was performed as previously described (Lin et al., 2015), using the following antibodies: phospho-S6 (Cell Signaling #5018), phospho-mTOR (EBioscience #50-9718-41) and Ki67 (Cell Signaling #11882).
Clonogenic growth assay
Colony formation assays were performed with KRAS G12V transformed WT and SIRT3 KO MEFs. 100 cells were seeded in 2 mL of media in six well plates. Media was replaced every 2-3 days. Colonies were stained with crystal violet after 8-10 days as previously described (Jeong et al., 2013). For quantification, crystal violet from the colonies was solubilized using 10% acetic acid. The absorbance of the extracted crystal violet solution at a wavelength of 590 nm was measured using a Varian Cary 50 Bio UV-visible spectrophotometer.
Supplemental References: Ben-Sahra, I., Howell, J.J., Asara, J.M., and Manning, B.D. (2013). Stimulation of de novo pyrimidine synthesis by growth signaling through mTOR and S6K1. Science 339, 1323-1328. Christodoulou, D.C., Wakimoto, H., Onoue, K., Eminaga, S., Gorham, J.M., DePalma, S.R., Herman, D.S., Teekakirikul, P., Conner, D.A., McKean, D.M., et al. (2014). 5'RNA-Seq identifies Fhl1 as a genetic modifier in cardiomyopathy. The Journal of clinical investigation 124, 1364-1370. Dobin, A., Davis, C.A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut, P., Chaisson, M., and Gingeras, T.R. (2013). STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15-21. Finley, L.W., Carracedo, A., Lee, J., Souza, A., Egia, A., Zhang, J., Teruya-Feldstein, J., Moreira, P.I., Cardoso, S.M., Clish, C.B., et al. (2011). SIRT3 opposes reprogramming of cancer cell metabolism through HIF1alpha destabilization. Cancer cell 19, 416-428. Green, E.M., Wakimoto, H., Anderson, R.L., Evanchik, M.J., Gorham, J.M., Harrison, B.C., Henze, M., Kawas, R., Oslob, J.D., Rodriguez, H.M., et al. (2016). A small-molecule inhibitor of sarcomere contractility suppresses hypertrophic cardiomyopathy in mice. Science 351, 617-621. Jeong, S.M., Xiao, C., Finley, L.W., Lahusen, T., Souza, A.L., Pierce, K., Li, Y.H., Wang, X., Laurent, G., German, N.J., et al. (2013). SIRT4 has tumor-suppressive activity and regulates the cellular metabolic response to DNA damage by inhibiting mitochondrial glutamine metabolism. Cancer cell 23, 450-463. Lin, J.R., Fallahi-Sichani, M., and Sorger, P.K. (2015). Highly multiplexed imaging of single cells using a high-throughput cyclic immunofluorescence method. Nature communications 6, 8390. Love, M.I., Huber, W., and Anders, S. (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome biology 15, 550. Mootha, V.K., Handschin, C., Arlow, D., Xie, X., St Pierre, J., Sihag, S., Yang, W., Altshuler, D., Puigserver, P., Patterson, N., et al. (2004). Erralpha and Gabpa/b specify PGC-1alpha-dependent oxidative phosphorylation gene expression that is altered in diabetic muscle. Proceedings of the National Academy of Sciences of the United States of America 101, 6570-6575. Schwer, B., North, B.J., Frye, R.A., Ott, M., and Verdin, E. (2002). The human silent information regulator (Sir)2 homologue hSIRT3 is a mitochondrial nicotinamide adenine dinucleotide-dependent deacetylase. The Journal of cell biology 158, 647-657.
