Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in Mutant KRAS-Driven Cancers

Article abstract of DOI:10.1016/j.chembiol.2018.03.005

Activating KRAS mutations are major oncogenic drivers in multiple tumor types. Synthetic lethal screens have previously been used to identify targets critical for the survival of KRAS mutant cells, but their application to drug discovery has proven challenging, possibly due in part to a failure of monolayer cultures to model tumor biology. Here, we report the results of a high-throughput synthetic lethal screen for small molecules that selectively inhibit the growth of KRAS mutant cell lines in soft agar. Chemoproteomic profiling identifies the target of the most KRAS-selective chemical series as dihydroorotate dehydrogenase (DHODH). DHODH inhibition is shown to perturb multiple metabolic pathways. In vivo preclinical studies demonstrate strong antitumor activity upon DHODH inhibition in a pancreatic tumor xenograft model. Koundinya et al. show that inhibitors of the pyrimidine biosynthetic enzyme dihydroorotate dehydrogenase selectively inhibit the growth of KRAS mutant cell lines. Differential sensitivity of the mutant lines correlates with differential effects of the inhibitors on primary energy metabolism and glutamine levels, and the inhibitors synergize with some clinically used anticancer agents.

Full text of DOI:10.1016/j.chembiol.2018.03.005

Article
Dependence on the Pyrimidine Biosynthetic Enzyme
DHODH Is a Synthetic Lethal Vulnerability in Mutant
KRAS-Driven Cancers
Graphical Abstract
Authors
Malvika Koundinya, Judith Sudhalter,
Albane Courjaud, ...,
Ivan Cornella-Taracido,
Rosalia Arrebola, Aaron Morris
Correspondence
sslicht@gmail.com (S.L.),
aaronjmorris@gmail.com (A.M.)
In Brief
Koundinya et al. show that inhibitors of
the pyrimidine biosynthetic enzyme
dihydroorotate dehydrogenase
selectively inhibit the growth of KRAS
mutant cell lines. Differential sensitivity of
the mutant lines correlates with
differential effects of the inhibitors on
primary energy metabolism and
glutamine levels, and the inhibitors
synergize with some clinically used
anticancer agents.
Highlights
d
d
d
DHODH inhibitors selectively inhibit the growth of KRAS
mutant cell lines
The inhibitors affect energy metabolism and glutamine levels
in sensitive lines
DHODH inhibitors synergize with some clinically used
anticancer agents
Koundinya et al., 2018, Cell Chemical Biology 25, 1–13
June 21, 2018 ª 2018 Elsevier Ltd.
https://doi.org/10.1016/j.chembiol.2018.03.005

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Cell Chemical Biology
Article
Dependence on the Pyrimidine Biosynthetic Enzyme
DHODH Is a Synthetic Lethal Vulnerability in Mutant
KRAS-Driven Cancers
Malvika Koundinya,1 Judith Sudhalter,1 Albane Courjaud,2 Bruno Lionne,2 Gaetan Touyer,2 Luc Bonnet,2
Isabelle Menguy,2 Isabelle Schreiber,2 Christelle Perrault,2 Stephanie Vougier,2 Brigitte Benhamou,2 Bailin Zhang,3
Timothy He,3 Qiang Gao,3 Patricia Gee,3 Daniel Simard,3,4 M. Paola Castaldi,4 Ronald Tomlinson,4 Stephan Reiling,5
Matthieu Barrague,6 Richard Newcombe,1 Hui Cao,7 Yanjun Wang,8 Fangxian Sun,8 Joshua Murtie,8 Mark Munson,6
Eric Yang,7 David Harper,1 Monsif Bouaboula,1 Jack Pollard,7 Claudine Grepin,2 Carlos Garcia-Echeverria,3 Hong Cheng,3
4
2
1,
Francisco Adrian,1 Christopher Winter,1 Stuart Licht,3,9, Ivan Cornella-Taracido, Rosalia Arrebola, and Aaron Morris
1Cancer Biology, Oncology Division, Sanofi, Cambridge, MA 02138, USA
2LGCR-LIT, Sanofi, Vitry-Sur-Seine 94400, France
*
*
3Biochemistry, Bioanalytics, and Chemical Biology, Oncology Division, Sanofi, Cambridge, MA 02138, USA
4Chemistry, Oncology Division, Sanofi, Cambridge, MA 02138, USA
5LGCR-SDI, Oncology Division, Sanofi, Cambridge, MA 02138, USA
6LGCR, Oncology Division, Sanofi, Cambridge, MA 02138, USA
7TEM-BioInformatics, Oncology Division, Sanofi, Cambridge, MA 02138, USA
8In Vivo Pharmacology, Oncology Division, Sanofi, Cambridge, MA 02138, USA
9Lead Contact
*Correspondence: sslicht@gmail.com (S.L.), aaronjmorris@gmail.com (A.M.)
https://doi.org/10.1016/j.chembiol.2018.03.005
SUMMARY
(Hunter et al., 2014; Ostrem et al., 2013; Sun et al., 2012),
KRAS inhibitors have not yet entered the clinic. Synthetic
Activating KRAS mutations are major oncogenic lethality screens have frequently been used to find alternative
points of intervention in mutant KRAS-driven cancers, but the re-
sults of these screens have not yet been translated to clin-
drivers in multiple tumor types. Synthetic lethal
screens have previously been used to identify targets
critical for the survival of KRAS mutant cells, but their
application to drug discovery has proven chal-
lenging, possibly due in part to a failure of monolayer
cultures to model tumor biology. Here, we report the
results of a high-throughput synthetic lethal screen
for small molecules that selectively inhibit the
ical benefit.
Previous synthetic lethal screening efforts (Cox et al., 2014)
have mainly used RNAi as a means of identifying potential tar-
gets (Barbie et al., 2009; Kim et al., 2016; Luo et al., 2009; Scholl
et al., 2009), although a few screens (Shaw et al., 2011; Steckel
et al., 2012) have been performed with small chemical libraries.
In all cases, these screens have relied on adherent monolayer
(2D) cell cultures, and in most cases have used only a matched
pair of mutant/wild-type (WT) cell lines. To improve the probabil-
growth of KRAS mutant cell lines in soft agar. Che-
moproteomic profiling identifies the target of the
most KRAS-selective chemical series as dihydrooro- ity of finding translatable targets, it might be beneficial to use
more physiologically relevant cellular models and to screen
larger panels of cell lines (Cox et al., 2014).
tate dehydrogenase (DHODH). DHODH inhibition is
shown to perturb multiple metabolic pathways.
In an effort to discover translatable mutant KRAS synthetic le-
In vivo preclinical studies demonstrate strong anti-
thal targets, we have performed a 3D clonogenic synthetic lethal
screen with a diverse small-molecule library of 280,000 com-
tumor xenograft model.
tumor activity upon DHODH inhibition in a pancreatic
pounds across six mutant KRAS-dependent cell lines and four
KRAS-independent cell lines. We report here the results from
this screen, and the identification and characterization of dihy-
droorotate dehydrogenase (DHODH) as the target of one of
the most KRAS mutant-selective compounds. As expected,
INTRODUCTION
Despite the advance of pathway-targeted therapeutics, KRAS DHODH inhibition prevents de novo pyrimidine biosynthesis,
mutant cancers still remain a largely unmet medical need. The which appears to be a particularly important pathway in the
prevalence of activating KRAS mutations is particularly high growth/survival of KRAS mutant cells. DHODH inhibition also
(greater than 90%) in pancreatic ductal adenocarcinoma substantially decreases cellular levels of glutamine and gluta-
(Bardeesy and DePinho, 2002), which has a notoriously poor mate, suggesting a connection to the previously established
prognosis. KRAS mutations are also prevalent in lung and colon glutamine dependence of KRAS mutant cells (Son et al., 2013;
cancers, and the presence of these mutations negatively im- Weinberg et al., 2010). We also show that brequinar, a DHODH
pacts therapeutic efficacy (Linardou et al., 2008). While direct inhibitor that has previously failed to demonstrate efficacy in clin-
targeting of the KRAS protein has shown renewed promise ical trials as an anticancer agent (Moore et al., 1993), exhibits
Cell Chemical Biology 25, 1–13, June 21, 2018 ª 2018 Elsevier Ltd.
1

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Figure 1. Determination of KRAS Dependence for a Panel of Cell Lines Cultured in 3D
(A) RNAi knockdown of KRAS impairs growth and viability of HCT116 (mutant KRAS) cells in 3D culture, but SK-OV-3 (KRAS WT) cells are not affected. Non-
targeted (NT) shRNA was used as a negative control. Dox, doxycycline; RFU, relative fluorescence units.
(B) KP-4 cells (KRAS mutant) in 3D culture (circles) are more sensitive to a MEK inhibitor (pimasertib) than the same cells in 2D culture (squares).
(C) Panel of validated mutant KRAS-dependent and -independent cell lines.
Error bars represent the standard error of the mean. See also Figures S1 and S2. EGFR, epidermal growth factor receptor.
strong in vivo antitumor activity in a KRAS mutant pancreatic tu- KRAS-independent cell lines were defined as those in which
mor xenograft model. The links observed in vitro between KRAS KRAS knockdown had no significant effect (Figure 1A).
status and a requirement for metabolic flux through the de novo In testing growth conditions, we discovered that culturing cells
pyrimidine biosynthetic pathway suggest new strategies for the in soft agar (3D culture) had a dramatic effect on their sensitivity
clinical application of potent DHODH inhibitors against KRAS to inhibition of Ras-MEK-ERK signaling. Cells cultured in 3D
mutant cancers.
exhibited a half maximal inhibitory concentration (IC50) for the
MEK inhibitor pimasertib that was approximately two orders of
magnitude less than the same cells cultured on plastic (2D cul-
ture) (Figure 1B). Since 3D culture is likely to be a more faithful
model for tumors in vivo, these conditions were chosen for
RESULTS
KRAS Dependency of Cell Lines Used for Screening
The panel of cell lines for screening was selected based on the screening.
dependence of growth and viability on mutant KRAS status.
An additional important consideration in selecting cell lines
Because of the important contribution of oncogenic KRAS was finding KRAS WT cell lines with 3D growth capacity and ki-
signaling to robust anchorage-independent growth (Campbell netics comparable to KRAS mutant cell lines. Based on the
et al., 2007; Guerrero et al., 2000), we assessed the KRAS criteria of KRAS dependence and growth rate, we selected
dependence of cell lines based on the impact of KRAS knock- a panel that included six KRAS-dependent and four KRAS-
down in a soft agar clonogenic growth assay. KRAS mutant independent cell lines (Figure 1C). All the KRAS-independent
cell lines showing a reduction in growth upon KRAS knockdown cell lines chosen had WT KRAS genes. Notably, one cell
greater than 50% were designated as KRAS-dependent, while line (H1299) presented an activating mutation of NRAS, which
2
Cell Chemical Biology 25, 1–13, June 21, 2018

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
provided a unique control for the distinct oncogenic impact of that the thiazolimines inhibit a pathway required for survival of
mutant KRAS relative to other Ras isoforms.
KRAS mutant cell lines.
A Screen for Small Molecules with Selective Cytotoxicity
Chemoproteomic Discovery of Potential Targets of the
Thiazolimine Series
toward KRAS Mutant Cell Lines
The chosen cell lines were used in a high-throughput screen of Based on the promising results from the synthetic lethal screen
a 280,000-compound chemical library (Figure S1). Following an and the cell line profiling, the thiazolimine series was chosen for
initial primary screen for effects on viability in five KRAS mutant compound deorphaning via a chemoproteomics approach. For
lines, a KRAS WT line was added to the panel, and actives this chemoproteomics approach, it was judged likely that, for
were confirmed and tested for KRAS mutant selectivity. The cells that exhibited thiazolimine sensitivity in both 2D and 3D cul-
list of active compounds was filtered to remove frequent hitters ture, the specific molecular target of the thiazolimine series would
and other compounds known to be problematic in high- be the same in both culture systems, although the details of the
throughput screens. Potency of cell growth inhibition (IC50
)
phenotypic consequences of target inhibition might differ. In or-
against the five KRAS mutant lines/one KRAS WT line was der to build a solid-supported affinity reagent for use in protein
measured for confirmed actives of interest, and compounds target pull-downs, the lead thiazolimine was functionalized with
were chosen for further analysis based on their selectivity for a short amine-terminated linker at various positions. Compound
mutant lines.
14, in which a phenol group was O-alkylated with the linker, re-
To find compounds with high selectivity for KRAS mutant tained activity in DLD-1 and KP-4 cells, thus validating the phenol
lines, potencies and efficacies were plotted (Figure S2). The moiety as a suitable attachment point for the solid support.
molecules exhibiting both high potency and high efficacy Based on these results, an affinity reagent (compound 18), in
were selected. The list was further curated by removing promis- which 14 was attached to Sepharose beads, was prepared.
cuous and/or chemically intractable compounds, leaving a total
A SILAC-based (Mann, 2006) strategy was used to identify
of 277 compounds. Those were tested against an expanded proteins that bind specifically to the thiazolimine-derivatized
panel of six KRAS mutant and four KRAS WT cell lines. Based solid support (Figure 3A). To provide labeled proteome samples
on their differential activities against mutant and WT lines in for affinity pull-downs, KP-4 cells were grown in either 2D or 3D
this broader panel, the compounds were divided into tiers cell culture using media containing amino acids with either a light
(Figure S2B).
or heavy stable isotope label. In the first of three related experi-
The compounds exhibiting high potency and KRAS-selectivity ments, SILAC pull-downs were performed at two concentrations
came from a number of chemical series. As expected, a MEK in- of the active competitor compound 2 (13 or 203 its half maximal
hibitor, PD-0325901, was one of the most selective inhibitors of effective concentration [EC50]; Table S1), comparing the pull-
KRAS mutant lines. It showed activity against all six KRAS down efficiency in the presence of the competitor to the effi-
mutant lines, but was only active against one WT line. However, ciency in the presence of DMSO as a control (Figure 3B).
the most potent and KRAS mutant-selective compounds be- Dihydroorotate dehydrogenase (DHODH; Uniprot ID PYRD)
longed to another series, based on a thiazolimine scaffold. Active was among the proteins most highly differentially competed
members of this series were chosen for characterization in a sec- from the affinity matrix. DHODH was identified from multiple
ond screen against a large cancer cell line panel.
peptides in the SILAC experiments. The SILAC ratios observed
Because of the difficulty of screening a larger cell line panel in were qualitatively dependent on the concentration of competitor
the 3D format (the automated work stream used for large-scale compound 2. These observations are consistent with a specific
screening was customized for 2D cultures and many WT lines interaction between DHODH and the affinity reagent-derivatized
failed to grow robustly in 3D culture), only the thiazolimine series support.
was tested against this panel of lines (374 diverse cancer cell
In a second experiment, a four-plex tandem mass tag pull-
lines characterized with respect to genetic mutations, copy down experiment on 2D-cultured KP-4 cells was used to quan-
number variations, and mRNA expression levels) in 2D culture. tify the dose-response trend observed in the SILAC experiment.
An analysis of the results (Figure 2A) showed that the thiazoli- The decrease in DHODH affinity pull-down correlated well with
mines were among the most KRAS mutant-selective agents the concentration of free competitor compound 2 (Figure 3C).
tested; the KRAS mutant lines in the panel were, on average, That observation further supports the hypothesis that DHODH
more sensitive to the thiazolimines, and this selectivity was binds specifically to the derivatized solid support.
comparable to that observed for the MEK inhibitor pimasertib.
A third experiment was designed to determine whether
In a complementary analysis, the known mutations in the cell DHODH would also exhibit affinity for a biologically inactive ho-
line panel were examined to determine which were most highly molog of the active thiazolimine compound 2. Both the active
enriched among cell lines that responded to the thiazolimine compound 2 and its close analog 3 (Table S1), which does not
compound 1 (Table S1). This analysis showed that KRAS was inhibit KP-4 cell growth, were used as competitors in a SILAC
the gene in which mutations were most strongly associated experiment. The results of this experiment (Figure 3D) show
with sensitivity to the compound (Figure 2B). In addition, many that DHODH was selectively competed off the affinity matrix by
lung and pancreatic cell lines in the panel were sensitive to 2 relative to 3. This observation suggests that within the thiazo-
compound 1, while cell lines from tissues of origin where limine series, bioactivity against KP-4 cells is associated with
KRAS mutation is less prevalent tended to be less sensitive to binding affinity for DHODH. It is consistent with the hypothesis
it (Figure 2C), consistent with a substantial effect of the KRAS that the thiazolimine series exerts its effect on cell growth by in-
mutation on differential sensitivity. These data thus suggest hibiting DHODH.
Cell Chemical Biology 25, 1–13, June 21, 2018
3

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Figure 2. Phenotypic Profiling of Chemical Hit Series across a Large Cancer Cell Line Panel Identifies One Hit Series as Highly Selective for
KRAS Mutant Cell Lines
(A) Compounds from the thiazolimine series are among the most selective for KRAS mutant cell lines. Statistical significance is plotted as a function of selectivity,
as expressed as the ratio of average potency against KRAS WT/KRAS mutant lines.
(B) KRAS mutation status is the best predictor of cell sensitivity to the thiazolimine series among the mutations represented. Statistical significance is plotted
against mutant/WT differential sensitivity to thiazolimine compound 1.
(C) Compound 1 (red symbols) is highly potent against many pancreatic and lung cancer cell lines. Pimasertib (blue symbols) is plotted for comparison.
Mechanistic Validation of the DHODH Target Hypothesis
ences between 2D and 3D cultures becomes relevant; however,
To determine whether binding of the thiazolimines inhibits by choosing cells that exhibit a growth/survival phenotype in 2D
DHODH’s enzymatic activity, we investigated the inhibitory po- culture, we can correlate DHODH inhibition with cellular pheno-
tency of a number of thiazolimines against purified recombinant type in a 2D system where the phenotype is similar to what is
DHODH. The tested compounds were highly potent on average observed in 3D cultures. A correlation analysis reveals a
(a median IC50 of 11 nM), with the more potent members of the Spearman’s coefficient of 0.76 between these two potencies
series being comparably potent to brequinar, a previously char- (p = 1 3 10À3), consistent with the hypothesis that DHODH inhi-
acterized DHODH inhibitor with an IC50 of 5–10 nM (Chen et al., bition is a key driver of cellular potency. However, there is a sub-
1992; McLean et al., 2001). These results indicate that the thia- stantial (ꢀ100-fold) shift toward lower potency when comparing
zolimines bind to DHODH in a mode that prevents enzymatic the biochemical with the cellular assay (e.g., compounds with
catalysis.
We next investigated whether the compounds’ DHODH inhib- of ꢀ1 mM in the cellular assay).
ꢀ10 nM IC50 values in the biochemical assay exhibit IC50 values
itory potencies were correlated with their potencies for cell As an additional and more direct test of whether the thiazoli-
growth inhibition (Figure 4A). Here, the issue of biological differ- mines also potently inhibit DHODH in a cellular context, a
4
Cell Chemical Biology 25, 1–13, June 21, 2018

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Figure 3. Chemoproteomic Deorphaning of the Thiazolimine Series Identifies DHODH (PYRD) as a Candidate Target for the Series
(A) Outline of SILAC experiments in which light/heavy isotopically labeled media encodes the presence of active/inactive compound (or vehicle).
(B) SILAC experiments in which active compound (compound 2) is used as a competitor at two concentrations (13 EC50 or 203 EC50). KP-4 cells were cultured in
2D (blue graphs, left-hand side) or 3D (red graphs, right-hand side).
(C) Quantitative dose-response relationship for a soluble competitor compound by iTRAQ.
(D) SILAC experiment to determine ratio of proteins competed by active (compound 2)/inactive (compound 3) free compound. Lysates from 2D cultures (blue
graph, left-hand side) or 3D cultures (red graph, right-hand side) were used.
See also Table S1.
mass spectrometry-based pharmacodynamic (PD) assay for di- that for all DHODH inhibitors examined, potencies in the three
hydroorotate (DHO) was used to measure the accumulation of assays are mutually correlated with each other (Spearman
this DHODH substrate in cells. The results (Figure 4B) show coefficients of 0.7, 0.8, and 0.96 for biochemical/PD, biochem-
that the thiazolimines’ potencies for cellular inhibition of DHODH ical/phenotypic, and PD/phenotypic correlations, respectively).
activity are highly correlated with their potencies for inhibition of Thus, on-target effects of DHODH enzyme inhibitors appear to
cell growth (Spearman’s coefficient of 0.96, p = 3 3 10À3). These contribute to growth inhibition of KRAS mutant cells.
data thus support the hypothesis that thiazolimines inhibit cell
The hypothesis that the thiazolimines act via DHODH inhibition
growth in KRAS mutant cells by inhibiting DHODH enzymatic ac- suggests that supplementation with exogenous pyrimidines may
tivity. Biochemical and cellular potencies are also reasonably rescue cell growth inhibition. Treatment of cells with exogenous
well-correlated for DHODH inhibitors from other chemotypes uridine, which can complement deficits in de novo pyrimidine
(e.g., brequinar and analogs) (Figure S3A).
biosynthesis (Connolly and Duley, 1999), was used to test this
Dose-response relationships were measured in all three as- prediction. As predicted, addition of uridine rescued cell growth
says (biochemical, cell PD, and cell viability) for representative in thiazolimine-treated cells (Figure 4D). A similar rescue effect
molecules in each series. The data (Figures 4C and S3B) indicate was observed for cells treated with brequinar (Figure S4).
Cell Chemical Biology 25, 1–13, June 21, 2018
5

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Figure 4. Validation of DHODH as a Biologi-
cally Relevant Target for the Thiazolimine
Series in a KRAS Mutant Cell Line
(A) Correlation of IC50 in the biochemical assay with
EC50 in the cell growth inhibition assay (values
expressed as the logarithm).
(B) Correlation of EC50 in the pharmacodynamic
(PD) assay with EC50 in the cell growth inhibition
assay.
(C) Correlation of biochemical IC50, PD EC50, and
cell growth inhibition EC50
.
(D) Dose-dependent rescue of compound
2-induced cell growth inhibition by uridine. Blue
symbols: thiazolimine dose-response in the
absence of uridine. Red symbols: 100 mM uridine.
Green symbols, 500 mM uridine.
Error bars represent the standard error of the
mean. See also Figures S3 and S4.
fumarate, and aspartate, and decreases
in the levels of glutamate, succinate, and
aKG. Brequinar also induces a substantial
dose-dependent decrease in the steady-
state level of glutamine (Figure 5B); a
Notably, the concentration of uridine required for complete comparable but more potent decrease is observed using
rescue (>100 mM) is greatly in excess of the normal physiological compound 2.
serum concentration (ꢀ5 mM; Connolly and Duley, 1999).
To further investigate how DHODH inhibition might differen-
tially affect KRAS WT and mutant cells, we carried out metabolo-
mics experiments to compare the effects of brequinar on KRAS
WT and mutant DLD-1 cells, as well as to compare metabolite
Mitochondrial Metabolic Effects and Other Cellular
Consequences of DHODH Inhibition
In addition to the role of DHODH in de novo pyrimidine biosyn- levels between the WT and mutant cells in the absence of drug
thesis, its role in mitochondrial electron transport may also (Figures S6 and S7). Consistent with results in other cell lines,
contribute to these inhibitors’ effects.
the mutant DLD-1 cells exhibited a greater response to brequinar
To investigate whether DHODH inhibitors affect the mitochon- and compound 2 than otherwise isogenic cells containing the WT
drial membrane potential, we measured mitochondrial potential allele (Figure S6). In the KRAS mutant DLD-1 cells at the 4-hr time
in KP-4 cells as a function of DHODH inhibitor dose (Figure 5A). point after drug treatment (i.e., before effects on cell viability are
The results show that DHODH inhibitors from two distinct chem- observed), the effects include an increase in the steady-state
ical series potently and dose-dependently induce mitochondrial concentrations of fructose 1,6-bisphosphate; this increase is
depolarization. The MEK inhibitor pimasertib, which also inhibits not observed in the WT cell line (Figure S7). In addition, the flux
KP-4 cell growth, does not affect mitochondrial membrane of glutamate (as judged by the change in isotopic incorporation
potential. In addition to the inactive compound 3, teriflunomide, from 13C-glutamine) increases in response to brequinar treat-
which is a much less potent DHODH inhibitor than brequinar ment, both for KRAS mutant and KRAS WT cells. In KRAS mutant
(Davis et al., 1996; Erra et al., 2011; McLean et al., 2001) or the cells, brequinar treatment is also associated with a large increase
best thiazolimine compounds, also has a minimal effect on mito- in AMP and ADP concentrations and a concomitant decrease
chondrial membrane potential (data not shown). These results in the concentration of ATP, indicative of energy stress. In
suggest that mitochondrial depolarization may contribute to contrast, brequinar treatment of KRAS WT cells leads to at
the effects of potent DHODH inhibitors in cells.
most a modest increase in AMP and ADP, with an increase in
Inhibition of DHODH might also affect steady-state metabolite ATP observed as well, suggesting that the mutant cells are differ-
levels, since glutamine is a starting metabolite for de novo pyrim- entially sensitive to energy stress induced by DHODH inhibition.
idine biosynthesis and DHODH is also linked to the citric acid cy- DHODH inhibition does not affect the ratio of reduced/oxidized
cle via its role in mitochondrial electron transport. For these glutathione at the 4-hr time point, but, consistent with previous
detailed metabolomics experiments and subsequent in vivo studies (Yun et al., 2015), the reduced/oxidized glutathione ratio
experiments, we chose to use brequinar rather than a thiazoli- is markedly smaller in KRAS mutant than KRAS WT cells.
mine compound due to the extensive previous preclinical and
In addition to perturbing metabolite levels, DHODH inhibition
clinical characterization of brequinar as a DHODH inhibitor. also induces cell-cycle arrest. Flow cytometric analysis of
Mass spectrometric analysis indicates that treatment with cellular DNA content after treatment of KP-4 cells with com-
brequinar perturbs the levels of a number of metabolites (Figures pound 2 (Figures 5C and S8) indicates that the cells become
5B and S5). In KP-4 cells at the 24-hr time point after drug arrested in S phase. Inhibition of pyrimidine metabolism and/or
treatment, the effects include increases in the levels of malate, other metabolic perturbations may mediate this cell-cycle arrest.
6
Cell Chemical Biology 25, 1–13, June 21, 2018

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Figure 5. Effects of DHODH Inhibition on Cellular Metabolism and the Cell Cycle
(A) Potent DHODH inhibitors induce mitochondrial depolarization in KP-4 cells. Magenta symbols: compound 2; green symbols, brequinar; orange symbols,
pimasertib; black symbols, inactive thiazolimine (compound 3).
(B) DHODH inhibitors (inactive thiazolimine 3, active thiazolimine 2, and brequinar) decrease cellular levels of glutamine.
(C) Treatment of KP-4 cells with thiazolimine compound 2 for 24–48 hr induces arrest in S phase. DNA content was quantified using propidium iodide. Pink
histogram: cells treated with compound 2. Blue histogram: DMSO-treated cells.
Error bars represent the standard error of the mean. See also Figures S5–S8.
In Vivo Antitumor Activity of a DHODH Inhibitor
3 mg/kg and 10 mg/kg doses that supported substantial tumor
Although DHODH inhibition has well-defined effects on cancer growth inhibition, body weight loss was minimal (at 3 mg/kg) or
cells in vitro, the potential availability of metabolites from other transient (at 10 mg/kg) over the course of the study, defining a
tissues in vivo might limit the applicability of the in vitro findings. therapeutic window for dosing in this system.
To address this issue, we carried out in vivo studies of the effects
of brequinar on xenografts of KP-4 cells. Measurements of Synergistic Combinations of a DHODH Inhibitor with
metabolite concentration in the tumors (Figure 6A) establish Other Drugs
that the DHODH substrate dihydroorotate (DHO) in the tumor Because DHODH inhibitors have already been clinically tested
builds up following accumulation of brequinar, consistent with without demonstrating efficacy (Moore et al., 1993; Peters
in vivo target engagement. Similarly, measurement of the pyrim- et al., 1992), we investigated whether brequinar (the DHODH
idine biosynthesis product uridine as an additional PD marker inhibitor for which the most clinical data is available) might
(Figure 6B) indicates that it becomes depleted. In addition, as exhibit greater efficacy in vitro in combination with another
observed in vitro, glutamine was depleted in the tumor after bre- drug. A set of 23 potential enhancer compounds with known
quinar treatment (Figure 6C).
mechanisms of action was chosen based on previous
In addition to demonstrating target engagement, the xenograft screening results against KRAS mutant cells and on an
studies also indicate that brequinar suppresses tumor growth assessment of whether the compounds’ mechanisms of action
in vivo in a dose-dependent fashion (Figure 6D). Consistent were likely to synergize with DHODH inhibition. These com-
with previous reports (Moore et al., 1993; Peters et al., 1990, pounds were screened against KP-4 cells in a 12 3 12 dose
1992), brequinar also exhibits dose-dependent toxicity as matrix with brequinar (in 2D culture) to test for synergistic
measured by body weight loss in treated mice (Figure 6E); at response.
the 30 mg/kg dose, this body weight loss was large enough to The combination screen identified a number of drugs that syn-
require termination of that arm of the study. However, at the ergize with brequinar in vitro (Figures 7A and 7B) using at least
Cell Chemical Biology 25, 1–13, June 21, 2018
7

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Figure 6. Brequinar Exhibits an Antitumor Effect In Vivo Against KP-4 Xenograft Tumors
(A) Accumulation of DHODH substrate DHO in tumors is correlated with the concentration of brequinar in the tumor.
(B) Accumulation of DHO is correlated with depletion of uridine in vivo. Uridine concentration is presented in a.u. corresponding to ion counts from mass
spectrometry.
(C) Depletion of tumor glutamine is associated with brequinar treatment and occurs with a time course similar to uridine depletion.
(D) Brequinar suppresses the growth of KP-4 xenografts. Po, by mouth; qd, every day.
(E) Brequinar exhibits a therapeutic window in the KP-4 xenograft model.
Error bars represent the standard error of the mean.
one standard criterion (Loewe [Chou and Talalay, 1984] and/or second-generation GLS1 inhibitor exhibits strong synergy with
HSA [Berenbaum, 1989] drug synergy scores; none of the brequinar (data not shown).
drugs listed scored as synergistic using the Bliss synergy score
[Bliss, 1939]). Notably, two of these (floxuridine, gemcitabine) act DISCUSSION
directly on enzymes involved in pyrimidine nucleotide biosyn-
thesis, and a third (everolimus) is expected to affect pyrimidine The current study integrates results from both a new screen in 3D
biosynthesis indirectly (Ben-Sahra et al., 2013; Robitaille culture and more conventional screens in 2D culture. Screening
et al., 2013).
in 3D culture has the advantage of sensitizing KRAS mutant cells
Because glutamate is a product of de novo pyrimidine biosyn- to Raf-MEK-ERK pathway inhibition (Figure 1D), so that the phe-
thesis, inhibiting pyrimidine biosynthesis may affect the steady- notypes observed may be more directly translatable to in vivo
state glutamate concentration. Experimentally, brequinar as a systems. Empirically, primary screening in 3D culture enabled
single agent induces a decrease in steady-state glutamate con- the discovery of KRAS synthetic lethal mechanisms that had
centration that, surprisingly, is comparable to the effect of an in- not previously been reported. However, screening in large
hibitor of GLS1, a major contributor to glutamate flux in many numbers of different cell lines proved to be more technically trac-
cancer cells (Figure 7C). The combination of a GLS1 inhibitor, table for 2D cell cultures. While results from both kinds of
compound 4 (Table S1) (Gross et al., 2014), with brequinar screens were broadly consistent with each other, responses of
was observed to decrease glutamate in KP-4 cells to lower 2D-cultured lines to synthetic lethal perturbation varied widely
levels than observed for equivalent concentrations of either (see, e.g., Figure 2A). These results suggest that 3D culture sys-
agent individually (Figure 7C); however, it is not yet known tems are useful for studying KRAS mutant phenotypes, and that
whether this effect is due to additivity or synergism between results from small numbers of cell lines cultured in 2D may or may
the two agents. No synergistic effect on cell growth was not be predictive of results from 3D cultures and/or in vivo
observed for 4; however, preliminary results indicate that a biology.
8
Cell Chemical Biology 25, 1–13, June 21, 2018

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Figure 7. Screening for Agents That Synergize with a DHODH Inhibitor
(A) Plot of statistical significance (adjusted false discovery rate, FDR) versus HSA synergy score.
(B) Compounds exhibiting significant synergy as judged by two different synergy scores (Loewe and HSA). mTOR, mammalian target of rapamycin.
(C) The combination of brequinar sodium and a GLS1 inhibitor (compound 4) decreases steady-state levels of glutamate in KP-4 cells to a greater extent than
either compound alone. Glutamate is quantified as intensity of MS signal (ion counts). Error bars represent the standard error of the mean.
The primary finding of the current work is that KRAS mutant (Patel et al., 2016). Deprivation of pyrimidine nucleotides is thus
cells exhibit a synthetic lethal sensitivity to inhibitors of DHODH. likely to be a major contributor to the differential effect of DHODH
However, it is clear that inhibition of DHODH has pleiotropic ef- inhibitors on cell viability in KRAS mutant cell lines.
fects in cancer cells. Multiple downstream consequences of
DHODH inhibition may thus contribute to the selective effect of (Figure 5A) may also contribute to the selective effects of
DHODH inhibitors. DHODH inhibitors on KRAS mutant cells. If so, the effect may
Mitochondrial depolarization induced by DHODH inhibition
Results in cells and in vivo (Figures 5, 6A, and 6B) indicate that, arise more from pro-apoptotic effects of mitochondrial depolar-
as expected, DHODH inhibition leads to decreased flux through ization (Lemasters et al., 1998) than from impaired oxidative
the de novo pyrimidine biosynthetic pathway. In addition, cell phosphorylation. Oncogenic KRAS mutations have been associ-
growth inhibition by DHODH inhibitors is rescued by uridine (Fig- ated with decreased, rather than increased, oxidative phosphor-
ure 4D), suggesting that inhibition of de novo pyrimidine biosyn- ylation: expression of mutant KRAS leads to an increased
thesis contributes to cytotoxicity. The latter observation further dependence on anabolic glucose metabolism (Weinberg et al.,
suggests that de novo pyrimidine biosynthesis is required for 2010; Ying et al., 2012), while ablation of mutant KRAS leads
growth and survival in KRAS mutant cells.
to an increased dependence on oxidative phosphorylation (Viale
Previous studies have also linked Ras-Raf-MEK-ERK signaling et al., 2014). However, to the extent that KRAS mutant cells
with increased de novo pyrimidine biosynthesis. Decreased exhibit enhanced flux through de novo pyrimidine biosynthesis,
expression of an inducible mutant KRAS downregulates pyrimi- they may also exhibit a greater dependence on electron trans-
dine biosynthetic genes at the transcriptional level (Ying et al., port through DHODH to maintain consistently hyperpolarized
2012). In addition, MAPK signaling abolishes feedback inhibition mitochondrial membrane potentials and a pro-survival pheno-
of carbamoyl phosphate synthetase, which catalyzes the rate- type. A precedent for an association between hyperpolarized
limiting step in de novo pyrimidine biosynthesis (Graves et al., mitochondrial potentials and a pro-survival phenotype can be
2000). Notably, glutamine deprivation in KRAS-driven cancer found in the example of CD8+ memory T cells, for which surplus
cell lines has recently been shown to interfere with deoxynucleo- capacity for oxidative phosphorylation enhances cellular survival
tide biosynthesis, leading to S-phase arrest and replicative stress (van der Windt et al., 2012).
Cell Chemical Biology 25, 1–13, June 21, 2018
9

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Measurements of metabolite levels in the current study are NADH, these metabolites must partition between the two pro-
also consistent with the idea that KRAS mutant cells are differen- survival pathways. DHODH activity may thus be required to
tially dependent on mitochondrial electron transport via DHODH, enable high flux through both of these pathways simultaneously.
which may contribute to differential effects of DHODH inhibition.
KRAS mutant cells, unlike WT cells, also exhibit a large bre-
In KRAS mutant cells, brequinar treatment is associated with en- quinar-mediated increase in fructose-1,6-bisphosphate levels
ergy stress (Figure S7). However, brequinar does not induce (Figure S7). However, there does not appear to be a clear mech-
similar energy stress in WT cells. Taken together, these data sup- anism for DHODH inhibition to mediate this effect directly. One
port the hypothesis that KRAS mutant cells are differentially possibility is that developing arrest of the cells in S phase leads
dependent on flux through DHODH for survival, due in part to to upregulation of glycolytic flux (Kaplon et al., 2015) and an in-
its role in primary energy metabolism. One possibility is that in crease in fructose-1,6-bisphosphate (which is downstream of
KRAS mutant cells, increased electron transport via DHODH the rate-limiting enzymes in this pathway). In that case, the upre-
partially compensates for decreased electron transport via aero- gulation of glycolytic flux would appear to be mostly through the
bic glucose metabolism.
anaerobic (lactic acid) branch of the pathway; brequinar treat-
In addition to a role for DHODH inhibition in primary energy ment is associated with a decrease in the steady-state level of
metabolism, the observation of decreased steady-state levels citrate without a concomitant decrease of isotopic exchange
of glutamine and glutamate and increased isotopic exchange into this metabolite (Figure S7), suggesting a decreased flux of
into glutamate (a proxy for metabolic flux) upon treatment with metabolites into the citric acid cycle.
a DHODH inhibitor (Figures 5C, 6C, 7G, S5, and S7) suggests
The relative functional importance of DHODH’s pyrimidine
a link between DHODH activity and glutamine metabolism, biosynthetic, energy-producing, and redox-modulating func-
which was previously demonstrated to be crucial for the growth tions is likely to depend on cellular context. DHODH has previ-
and survival of KRAS mutant cancer cells (Son et al., 2013; Wein- ously been characterized as a synthetic lethal target in several
berg et al., 2010). These observations, as well as the increased different oncogenic cell contexts, including the BRAF V600E mu-
steady-state level of aspartate observed upon DHODH inhibi- tation (White et al., 2011), PTEN deficiency (Mathur et al., 2017),
tion, are consistent with increased conversion of glutamine and and triple-negative breast cancer cell lines (Brown et al., 2017). In
oxaloacetate to aKG and aspartate via GOT1. Therefore, the the case of the BRAF V600E mutation, oncogenic mutations of
decreased steady-state levels of glutamine appear not to be both BRAF and KRAS are gain-of-function with respect to the
caused by a lack of available glutamine as a metabolic substrate, Ras-Raf-MEK-ERK pathway, which may account for the shared
but rather by increased consumption of glutamine.
phenotype with respect to DHODH inhibition. However, KRAS
Since the decrease in glutamine levels appears to be associ- signaling affects multiple signaling pathways in addition to Raf-
ated with increased glutaminolysis rather than a decreased abil- MEK-ERK, so, in general, interventions that affect KRAS mutant
ity to accumulate glutamine, it might be hypothesized that this cells may or may not affect BRAF mutant cells. Recently, de novo
decrease in glutamine levels does not contribute to impaired pyrimidine metabolism through carbamoyl phosphate generated
growth and survival in KRAS mutant cells. However, the obser- in the urea cycle has been shown to be important for growth/sur-
vation that uridine rescues both the effect of brequinar on gluta- vival in the KRAS/LKB1 double-mutant context (Kim et al., 2017),
mate flux and its cellular phenotypic effect suggests that the suggesting that DHODH inhibition might be useful specifically in
metabolic effects of DHODH do contribute to its broader biolog- this cellular context as well.
ical effects. In addition, the results are consistent with synergism
The in vitro and in vivo results presented here suggest that
between GLS1 inhibition and DHODH inhibition both at the level DHODH inhibition may be clinically useful in the treatment of
of regulating glutamine levels and (at least for some GLS1 inhib- KRAS mutant tumors. However, in previous clinical trials, brequi-
itors) at the level of growth and survival. One hypothesis that ac- nar failed to demonstrate efficacy in multiple tumor types,
counts for the current observations is that DHODH activity helps including pancreatic and colorectal, a substantial fraction of
to mediate the balance between two distinct glutaminolysis- which almost certainly harbored KRAS mutations (Moore et al.,
dependent redox metabolic pathways that have been impli- 1993; Peters et al., 1992). Understanding the basis for the
cated in growth/survival in KRAS mutant cells: generation of discrepancy between in vitro/in vivo models and clinical results
mitochondrial reactive oxygen species (ROS) (Weinberg et al., might enable the design of new and more effective clinical treat-
2010) and flux through GOT1 enabling generation of NADPH ment strategies.
and maintenance of reduced glutathione (Son et al., 2013; Yun
et al., 2015).
One possible explanation for the lack of observed clinical
effectiveness for brequinar is that target engagement in the
Electron transport via DHODH activity will allow the generation clinical studies was insufficient to induce tumor cell death. In vivo
of mitochondrial ROS without the consumption of a-ketogluta- studies suggest that prolonged exposure to brequinar is
rate or NADH. Glutamine-dependent flux through the TCA cycle required to achieve the metabolite depletion effects required
and oxidative phosphorylation can also enable generation of for maximum efficacy (Dexter et al., 1985; Schwartsmann
mitochondrial ROS (Weinberg et al., 2010). The latter process et al., 1988). However, the clinical trials of brequinar were de-
requires partitioning of aKG derived from glutamine into the signed following a traditional chemotherapeutic regimen, in
TCA cycle to generate reducing equivalents in the form of which the maximum tolerated dose was delivered, followed by
NADH, leading ultimately to reduction of molecular oxygen by an intervening recovery period. Pharmacokinetic data from these
Complex III to generate ROS (Weinberg et al., 2010). However, clinical trials (Arteaga et al., 1989) reveals that between doses,
since production of NADPH via GOT1/MDH1/ME1 (Son et al., the plasma drug concentrations dropped below the efficacy
2013; Yun et al., 2015) also requires consumption of aKG and threshold seen in our preclinical studies, and investigators in
10 Cell Chemical Biology 25, 1–13, June 21, 2018

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
B Mass Spectrometry Analysis and Data Interpretation
B DHODH Biochemical Assay
the clinical studies speculated that maintaining effective drug
concentrations over a longer period of time might be required
to achieve clinical efficacy (Braakhuis et al., 1990).
B Cell Cycle Analysis
B Mitochondrial Membrane Potential Assay
B LC-MS/MS Mass Spectrometric Metabolite Analysis of
Cells and Tumor Tissues
Another potential approach to achieving clinical benefit from
DHODH inhibition would be to employ combination therapies.
The results reported here suggest that promising combination
candidates include floxuridine (an inhibitor of pyrimidine biosyn-
thesis via thymidylate synthase) and gemcitabine (an inhibitor of
ribonucleotide reductase), both of which also target nucleotide
metabolism. Gemcitabine is a particularly intriguing potential
combination partner because it is already in use as a first-line
therapy for pancreatic cancer. Results on GLS1 inhibitors sug-
gest that such agents might also be useful combination agents
for DHODH inhibitors, although highly potent GLS1 inhibitors
may be required for a synergistic effect to be observable at the
level of growth inhibition.
B CE-TOFMS Mass Spectrometric Metabolite Analysis
of Cells and Tumor Tissues
B Synergistic Combinations in KRAS Mutant Cell
Context
B General Synthetic Methods
B Synthetic Protocols
d QUANTIFICATION AND STATISTICAL ANALYSIS
SUPPLEMENTAL INFORMATION
Supplemental Information includes seven figures and one table can be found
with this article online at https://doi.org/10.1016/j.chembiol.2018.03.005.
The current study provides evidence for a previously unknown
role for DHODH as a growth/survival factor in KRAS mutant tu-
mors. While many aspects of the link between DHODH inhibition
and survival of KRAS mutant cells remain to be established,
de novo pyrimidine biosynthesis appears to be a crucial pathway
for KRAS mutant cells. The connection between KRAS mutant
status and a requirement for DHODH enzymatic activity may
enable new therapeutic approaches for KRAS mutant cancers.
ACKNOWLEDGMENTS
We thank Dr. Laura Shelton (Human Metabolome Technologies) for helpful dis-
cussions. We thank Dr. Dietmar Hoffmann and the Sanofi Molecular Cloning
group for generation of inducible shRNA vectors for KRAS knockdown. We
thank Dr. Joern Hopke for microscopic image acquisition and analysis.
AUTHOR CONTRIBUTIONS
SIGNIFICANCE
M.K., J.S., A.C., B.L., G.T., L.B., I.M., I.S, C.P., S.V., B.B., B.Z., T.H., Q.G., P.G,
D.S., M.P.C., R.T., M. Barrague, R.N., Y.W., F.S., J.M., E.Y., and D.H. designed
and carried out experiments. S.R., H.C., M.M., E.Y., and J.P. analyzed data. M.
Barrague, M. Bouaboula, J.P., C.G., C. G-E., H.C., F.A., C.W., S.L., I. C-T. R.A.,
and A.M. designed and interpreted experiments. M.K., J.S., B.Z., Q.G., D.S.,
R.N., S.L., and A.M. wrote the manuscript, with contributions from all other
authors.
A synthetic lethal screen for small molecules that selectively
inhibit the growth of KRAS mutant cell lines identifies inhib-
itors of dihydroorotate dehydrogenase (DHODH) as poten-
tial therapies for KRAS mutant cancers. An investigation of
the metabolic effects of these inhibitors on KRAS mutant
cell lines points to perturbation of primary energy meta-
bolism and glutamine metabolism as possible contributors
to the observed KRAS mutant selectivity. The results sug-
gest that alternative clinical trial designs for potent clin-
ical-stage DHODH inhibitors such as brequinar in KRAS
mutant cancers may be worth evaluating.
DECLARATION OF INTERESTS
The authors are current or former employees and/or shareholders of Sanofi.
Received: November 23, 2016
Revised: December 29, 2017
Accepted: March 8, 2018
Published: April 5, 2018
STAR+METHODS
REFERENCES
Detailed methods are provided in the online version of this paper
and include the following:
Arteaga, C.L., Brown, T.D., Kuhn, J.G., Shen, H.S., O’Rourke, T.J., Beougher,
K., Brentzel, H.J., Von Hoff, D.D., and Weiss, G.R. (1989). Phase I clinical and
pharmacokinetic trial of brequinar sodium (DuP 785; NSC 368390). Cancer
Res. 49, 4648–4653.
d KEY RESOURCES TABLE
d CONTACT FOR REAGENT AND RESOURCE SHARING
d EXPERIMENTAL MODEL AND SUBJECT DETAILS
B In Vivo Animal Studies
Bantscheff, M., Eberhard, D., Abraham, Y., Bastuck, S., Boesche, M., Hobson,
S., Mathieson, T., Perrin, J., Raida, M., Rau, C., et al. (2007). Quantitative
chemical proteomics reveals mechanisms of action of clinical ABL kinase in-
hibitors. Nat. Biotechnol. 25, 1035–1044.
B Cell Lines
d METHOD DETAILS
Barbie, D.A., Tamayo, P., Boehm, J.S., Kim, S.Y., Moody, S.E., Dunn, I.F.,
Schinzel, A.C., Sandy, P., Meylan, E., Scholl, C., et al. (2009). Systematic
RNA interference reveals that oncogenic KRAS-driven cancers require
TBK1. Nature 462, 108–112.
B Reagents
B Compound Screening in 3D Cultures
B KRAS Knockdown: shRNA Expression and Gene
Down-Regulation
Bardeesy, N., and DePinho, R.A. (2002). Pancreatic cancer biology and ge-
B Cell Line Sensitivity Profiling
netics. Nat. Rev. Cancer 2, 897–909.
B Biochemical Pull-Down Experiment with Thiazolimine-
Conjugated Beads
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.
B SDS/PAGE and In-Gel Digestion
B iTRAQ Labeling of the Digested Pull-Down Samples
Berenbaum, M.C. (1989). What is synergy? Pharmacol. Rev. 41, 93–141.
Cell Chemical Biology 25, 1–13, June 21, 2018 11

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Bliss, C.I. (1939). The toxicity of poisons applied jointly. Ann. Appl. Biol. 26,
pools and DNA synthesis in KRAS/LKB1 mutant lung cancer cells. Nature
585–615.
546, 168–172.
Lemasters, J.J., Nieminen, A.L., Qian, T., Trost, L.C., Elmore, S.P., Nishimura,
Y., Crowe, R.A., Cascio, W.E., Bradham, C.A., Brenner, D.A., et al. (1998). The
mitochondrial permeability transition in cell death: a common mechanism in
necrosis, apoptosis and autophagy. Biochim. Biophys. Acta 1366, 177–196.
Braakhuis, B.J., van Dongen, G.A., Peters, G.J., van Walsum, M., and Snow,
G.B. (1990). Antitumor activity of brequinar sodium (Dup-785) against human
head and neck squamous cell carcinoma xenografts. Cancer Lett. 49,
133–137.
Linardou, H., Dahabreh, I.J., Kanaloupiti, D., Siannis, F., Bafaloukos, D.,
Kosmidis, P., Papadimitriou, C.A., and Murray, S. (2008). Assessment of so-
Brown, K.K., Spinelli, J.B., Asara, J.M., and Toker, A. (2017). Adaptive reprog-
ramming of de novo pyrimidine biosynthesis is a metabolic vulnerability in tri-
ple-negative breast cancer. Cancer Discov. 7, 391–399.
matic k-RAS mutations as
a mechanism associated with resistance to
EGFR-targeted agents: a systematic review and meta-analysis of studies in
advanced non-small-cell lung cancer and metastatic colorectal cancer.
Lancet Oncol. 9, 962–972.
Campbell, P.M., Groehler, A.L., Lee, K.M., Ouellette, M.M., Khazak, V., and
Der, C.J. (2007). K-Ras promotes growth transformation and invasion of
immortalized human pancreatic cells by Raf and phosphatidylinositol 3-kinase
signaling. Cancer Res. 67, 2098–2106.
Luo, J., Emanuele, M.J., Li, D., Creighton, C.J., Schlabach, M.R., Westbrook,
T.F., Wong, K.K., and Elledge, S.J. (2009). A genome-wide RNAi screen iden-
tifies multiple synthetic lethal interactions with the Ras oncogene. Cell 137,
835–848.
Chen, S.F., Perrella, F.W., Behrens, D.L., and Papp, L.M. (1992). Inhibition of
dihydroorotate dehydrogenase activity by brequinar sodium. Cancer Res.
52, 3521–3527.
Mann, M. (2006). Functional and quantitative proteomics using SILAC. Nat.
Chou, T.C., and Talalay, P. (1984). Quantitative analysis of dose-effect rela-
tionships: the combined effects of multiple drugs or enzyme inhibitors. Adv.
Enzyme Regul. 22, 27–55.
Rev. Mol. Cell Biol. 7, 952–958.
Mathur, D., Stratikopoulos, E., Ozturk, S., Steinbach, N., Pegno, S.,
Schoenfeld, S., Yong, R., Murty, V.V., Asara, J.M., Cantley, L.C., and
Parsons, R. (2017). PTEN regulates glutamine flux to pyrimidine synthesis
and sensitivity to dihydroorotate dehydrogenase inhibition. Cancer Discov.
7, 380–390.
Connolly, G.P., and Duley, J.A. (1999). Uridine and its nucleotides: biological
actions, therapeutic potentials. Trends Pharmacol. Sci. 20, 218–225.
Cox, A.D., Fesik, S.W., Kimmelman, A.C., Luo, J., and Der, C.J. (2014).
Drugging the undruggable RAS: mission possible? Nat. Rev. Drug Discov.
13, 828–851.
McLean, J.E., Neidhardt, E.A., Grossman, T.H., and Hedstrom, L. (2001).
Multiple inhibitor analysis of the brequinar and leflunomide binding sites on hu-
man dihydroorotate dehydrogenase. Biochemistry 40, 2194–2200.
Davis, J.P., Cain, G.A., Pitts, W.J., Magolda, R.L., and Copeland, R.A. (1996).
The immunosuppressive metabolite of leflunomide is a potent inhibitor of hu-
man dihydroorotate dehydrogenase. Biochemistry 35, 1270–1273.
Moore, M., Maroun, J., Robert, F., Natale, R., Neidhart, J., Dallaire, B., Sisk, R.,
and Gyves, J. (1993). Multicenter phase II study of brequinar sodium in patients
with advanced gastrointestinal cancer. Invest. New Drugs 11, 61–65.
Dexter, D.L., Hesson, D.P., Ardecky, R.J., Rao, G.V., Tippett, D.L., Dusak,
B.A., Paull, K.D., Plowman, J., DeLarco, B.M., Narayanan, V.L., et al. (1985).
Activity of a novel 4-quinolinecarboxylic acid, NSC 368390 [6-fluoro-2-(2’-flu-
oro-1,1’-biphenyl-4-yl)-3-methyl-4-quinolinecarb oxylic acid sodium salt],
against experimental tumors. Cancer Res. 45, 5563–5568.
Ostrem, J.M., Peters, U., Sos, M.L., Wells, J.A., and Shokat, K.M. (2013).
K-Ras(G12C) inhibitors allosterically control GTP affinity and effector interac-
tions. Nature 503, 548–551.
Patel, D., Menon, D., Bernfeld, E., Mroz, V., Kalan, S., Loayza, D., and Foster,
D.A. (2016). Aspartate rescues S-phase arrest caused by suppression of
glutamine utilization in KRas-driven cancer cells. J. Biol. Chem. 291,
9322–9329.
Erra, M., Moreno, I., Sanahuja, J., Andres, M., Reinoso, R.F., Lozoya, E.,
Pizcueta, P., Godessart, N., and Castro-Palomino, J.C. (2011). Biaryl ana-
logues of teriflunomide as potent DHODH inhibitors. Bioorg. Med. Chem.
Lett. 21, 7268–7272.
Peters, G.J., Kraal, I., and Pinedo, H.M. (1992). In vitro and in vivo studies on
the combination of brequinar sodium (DUP-785; NSC 368390) with 5-fluoro-
uracil; effects of uridine. Br. J. Cancer 65, 229–233.
Graves, L.M., Guy, H.I., Kozlowski, P., Huang, M., Lazarowski, E., Pope, R.M.,
Collins, M.A., Dahlstrand, E.N., Earp, H.S., 3rd, and Evans, D.R. (2000).
Regulation of carbamoyl phosphate synthetase by MAP kinase. Nature 403,
328–332.
Peters, G.J., Schwartsmann, G., Nadal, J.C., Laurensse, E.J., van Groeningen,
C.J., van der Vijgh, W.J., and Pinedo, H.M. (1990). In vivo inhibition of the py-
rimidine de novo enzyme dihydroorotic acid dehydrogenase by brequinar so-
dium (DUP-785; NSC 368390) in mice and patients. Cancer Res. 50,
4644–4649.
Gross, M.I., Demo, S.D., Dennison, J.B., Chen, L., Chernov-Rogan, T., Goyal,
B., Janes, J.R., Laidig, G.J., Lewis, E.R., Li, J., et al. (2014). Antitumor activity
of the glutaminase inhibtor CB-839 in triple-negative breast cancer. Mol.
Cancer Ther. 13, 890–901.
Robitaille, A.M., Christen, S., Shimobayashi, M., Cornu, M., Fava, L.L., Moes,
S., Prescianotto-Baschong, C., Sauer, U., Jenoe, P., and Hall, M.N. (2013).
Quantitative phosphoproteomics reveal mTORC1 activates de novo pyrimi-
dine synthesis. Science 339, 1320–1323.
Guerrero, S., Casanova, I., Farre, L., Mazo, A., Capella, G., and Mangues, R.
(2000). K-ras codon 12 mutation induces higher level of resistance to
apoptosis and predisposition to anchorage-independent growth than codon
13 mutation or proto-oncogene overexpression. Cancer Res. 60, 6750–6756.
Scholl, C., Frohling, S., Dunn, I.F., Schinzel, A.C., Barbie, D.A., Kim, S.Y.,
Silver, S.J., Tamayo, P., Wadlow, R.C., Ramaswamy, S., et al. (2009).
Synthetic lethal interaction between oncogenic KRAS dependency and
STK33 suppression in human cancer cells. Cell 137, 821–834.
Hesson, D.P. (1987). 2-Phenyl-4-quinolinecarboxylic acids and pharmaceu-
tical compositions thereof. US patent US4680299A, filed July 14, 1987.
Hunter, J.C., Gurbani, D., Ficarro, S.B., Carrasco, M.A., Lim, S.M., Choi, H.G.,
Xie, T., Marto, J.A., Chen, Z., Gray, N.S., et al. (2014). In situ selectivity profiling
and crystal structure of SML-8-73-1, an active site inhibitor of oncogenic
K-Ras G12C. Proc. Natl. Acad. Sci. USA 111, 8895–8900.
Schwartsmann, G., Peters, G.J., Laurensse, E., de Waal, F.C., Loonen, A.H.,
Leyva, A., and Pinedo, H.M. (1988). DUP 785 (NSC 368390): schedule-depen-
dency of growth-inhibitory and antipyrimidine effects. Biochem. Pharmacol.
37, 3257–3266.
Kaplon, J., van Dam, L., and Peeper, D. (2015). Two-way communication be-
tween the metabolic and cell cycle machineries: the molecular basis. Cell
Cycle 14, 2022–2032.
Shaw, A.T., Winslow, M.M., Magendantz, M., Ouyang, C., Dowdle, J.,
Subramanian, A., Lewis, T.A., Maglathin, R.L., Tolliday, N., and Jacks, T.
(2011). Selective killing of K-ras mutant cancer cells by small molecule in-
ducers of oxidative stress. Proc. Natl. Acad. Sci. USA 108, 8773–8778.
Kim, J., McMillan, E., Kim, H.S., Venkateswaran, N., Makkar, G., Rodriguez-
Canales, J., Villalobos, P., Neggers, J.E., Mendiratta, S., Wei, S., et al.
(2016). XPO1-dependent nuclear export is
a druggable vulnerability in
Son, J., Lyssiotis, C.A., Ying, H., Wang, X., Hua, S., Ligorio, M., Perera, R.M.,
Ferrone, C.R., Mullarky, E., Shyh-Chang, N., et al. (2013). Glutamine supports
pancreatic cancer growth through a KRAS-regulated metabolic pathway.
Nature 496, 101–105.
KRAS-mutant lung cancer. Nature 538, 114–117.
Kim, J., Hu, Z., Cai, L., Li, K., Choi, E., Faubert, B., Bezwada, D., Rodriguez-
Canales, J., Villalobos, P., Lin, Y.F., et al. (2017). CPS1 maintains pyrimidine
12 Cell Chemical Biology 25, 1–13, June 21, 2018

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Steckel, M., Molina-Arcas, M., Weigelt, B., Marani, M., Warne, P.H.,
Kuznetsov, H., Kelly, G., Saunders, B., Howell, M., Downward, J., et al.
(2012). Determination of synthetic lethal interactions in KRAS oncogene-
dependent cancer cells reveals novel therapeutic targeting strategies. Cell
Res. 22, 1227–1245.
Weinberg, F., Hamanaka, R., Wheaton, W.W., Weinberg, S., Joseph, J.,
Lopez, M., Kalyanaraman, B., Mutlu, G.M., Budinger, G.R., and Chandel,
N.S. (2010). Mitochondrial metabolism and ROS generation are essential for
Kras-mediated tumorigenicity. Proc. Natl. Acad. Sci. USA 107, 8788–8793.
White, R.M., Cech, J., Ratanasirintrawoot, S., Lin, C.Y., Rahl, P.B., Burke, C.J.,
Langdon, E., Tomlinson, M.L., Mosher, J., Kaufman, C., et al. (2011). DHODH
modulates transcriptional elongation in the neural crest and melanoma. Nature
471, 518–522.
Sun, Q., Burke, J.P., Phan, J., Burns, M.C., Olejniczak, E.T., Waterson, A.G.,
Lee, T., Rossanese, O.W., and Fesik, S.W. (2012). Discovery of small mole-
cules that bind to K-Ras and inhibit Sos-mediated activation. Angew. Chem.
Int. Ed. 51, 6140–6143.
Ying, H., Kimmelman, A.C., Lyssiotis, C.A., Hua, S., Chu, G.C., Fletcher-
Sananikone, E., Locasale, J.W., Son, J., Zhang, H., Coloff, J.L., et al. (2012).
Oncogenic Kras maintains pancreatic tumors through regulation of anabolic
glucose metabolism. Cell 149, 656–670.
van der Windt, G.J.W., Everts, B., Chang, C.-H., Curtis, J.D., Freitas, T.C.,
Amiel, E., Pearce, E.J., and Pearce, E.L. (2012). Mitochondrial respiratory ca-
pacity is a critical regulator of CD8+ T cell memory development. Immunity
36, 68–78.
Viale, A., Pettazzoni, P., Lyssiotis, C.A., Ying, H., Sanchez, N., Marchesini, M.,
Carugo, A., Green, T., Seth, S., Giuliani, V., et al. (2014). Oncogene ablation-
resistant pancreatic cancer cells depend on mitochondrial function. Nature
514, 628–632.
Yun, J., Mullarky, E., Lu, C., Bosch, K.N., Kavalier, A., Rivera, K., Roper, J.,
Chio, , II, Giannopoulou, E.G., Rago, C., et al. (2015). Vitamin C selectively kills
KRAS and BRAF mutant colorectal cancer cells by targeting GAPDH. Science
350, 1391–1396.
Cell Chemical Biology 25, 1–13, June 21, 2018 13

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
STAR+METHODS
KEY RESOURCES TABLE
REAGENT or RESOURCE
Antibodies
SOURCE
IDENTIFIER
Anti-Ras Antibody, clone RAS10
Monoclonal Anti-b-Actin
Chemicals, Peptides, and Recombinant Proteins
Recombinant DHODH
PD0325901 MEK Inhibitor
Pimasertib (AS-703026) MEK Inhibitor
Brequinar
Millipore Sigma
Millipore Sigma
Cat#05-516; RRID: AB_11211664
Cat#A5316; RRID: AB_476743
Creative Biomart
Selleck Chemicals
Selleck Chemicals
Sanofi internal collection
Sanofi internal collection
Selleck Chemicals
Selleck Chemicals
Selleck Chemicals
Sigma
Cat#DHODH-2486
Cat#S1036
Cat#S1475
N/A
Teriflunomide
N/A
Floxuridine
Cat#S1299
Cat#S1149
Cat#S1120
Cat#62996-74-1
Cat#50302
Cat#H21486
Cat#M7512
Gemcitabine
Everolimus (RAD001)
Staurosporine
SeaPrep Agarose
Lonza
Hoeschst 35480
Invitrogen
MitoTracker Red
Thermofisher
Critical Commercial Assays
Trans-Lentiviral shRNA Packaging Kit
KRAS Taqman Gene Expression assay
b Actin Taqman Gene Expression assay
TaqMan Gene Expression Cells-to-CT Kit
iTRAQꢀ Reagents
Open Biosystems
Thermofisher
Thermofisher
Thermofisher
SCIEX
Cat#TLP5912
Cat#Hs00270666_m1
Cat#Hs01102345_m1
Cat#AM1729
iTRAQꢀ Reagents
Cat#G7570
Cell Titer Glo
Promega
Experimental Models: Cell Lines
KP-4
JCRB
JCRB0182
DLD-1
ATCC
ATCC CCL-221
ATCC CRL-1469
ATCC CRL-1918
ATCC CCL-247
ATCC CRL-1420
ATCC CRL-5803
CVCL_D249
PANC-1
ATCC
CFPAC-1
ATCC
HCT116
ATCC
MIA PaCa-2
ATCC
H1299
ATCC
UISO-BCA-1
Southern Research Institute
ATCC
ACHN
ATCC CRL-1611
ATCC HTB-77
HD 105-002
SK-OV-3
ATCC
DLD1 KRAS (+/-)
Horizon Discovery
Horizon Discovery
DLD1 KRAS (G13D/-)
Experimental Models: Organisms/Strains
Female SCID mice
HD 105-011
Charles River Laboratories
This paper, Sanofi
N/A
N/A
Oligonucleotides
shRNA KRAS B sequence : CACCGTT
GGAGCTGGTGGtGTgGGCTTCAA
GAGAGCCTACGCCACCAGCTCCAAC
shRNA KRAS H sequence : CACCGGT
CCTGCTGACAAATCAgGATTCAAGAG
ATCTTGATTTGTCAGCAGGACC
This paper, Sanofi
N/A
(Continued on next page)
e1 Cell Chemical Biology 25, 1–13.e1–e11, June 21, 2018

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Continued
REAGENT or RESOURCE
Recombinant DNA
SOURCE
IDENTIFIER
N/A
pLKO_IRES-Puro_DEST
Sanofi internal collection
Corning
Other
Ultra low-attachment 384 well plates
Cat#3827
CONTACT FOR REAGENT AND RESOURCE SHARING
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Stuart
Licht (sslicht@gmail.com).
EXPERIMENTAL MODEL AND SUBJECT DETAILS
In Vivo Animal Studies
All in vivo experiments were performed according to institutional guidelines as approved by the Sanofi Institutional Animal Care and
Use Committee. 5x106 KP-4 cells were suspended in 50% matrigel and 50%DPBS and implanted subcutaneously into the right flank
of 7ꢀ8 week old female Severe Combined Immuno Deficiency (SCID) mice (Charles River Laboratories). The tumors were allowed to
establish until average tumor volume reached about 200-250 mm3. Subsequently mice were randomized into control and treated
groups (n=10/group), Tumor growth was monitored over time using caliper measurements. Subcutaneous tumor volume was calcu-
lated using the formula: length 3 width2/2. Brequinar was formulated in 20% HPbCD in 10mM phosphate buffer. Animals were orally
administered brequinar at 1, 3, 10, 30 mg/kg once daily for 14 days. For the pharmacodynamic studies (Figures 6A–6C), three mice
were used for each time point. In the efficacy study (Figure 6D), ten mice were used in each group.
Cell Lines
The following cell lines were used as a panel for initial testing in soft agar : PANC-1 (pancreatic), KP-4 (pancreatic), CFPAC-1 (pancre-
atic), MiaPaCa-2 (pancreatic), HCT-116 (colon), DLD-1 (colon), SK-OV-3 (ovarian), ACHN (kidney), H1299 (lung), UISO-BCA-1
(breast). Chemoproteomic experiments, cell cycle analysis, mitochondrial potential experiments, metabolite profiling, and drug
synergy experiments were carried out using KP-4 cells. For capillary electrophoresis-based metabolomics experiments, DLD-1 cells
bearing either a single mutant or single wild-type allele were used (DLD1 KRAS (+/-) HD-105-002 and DLD1 KRAS (G13D/-) HD-105-
011, Horizon Discovery). All cells were cultured at 37ꢁC in humidified incubators at 5% CO2.
METHOD DETAILS
Reagents
Biochemical reagents were obtained from Sigma-Aldrich unless otherwise specified. Brequinar was synthesized as previously
described (Hesson, 1987). MEK inhibitor PD0325901 was purchased from Selleck Chemicals.
Compound Screening in 3D Cultures
SeaPrep agarose (Lonza #50302) was used for cell culture. A 6% (w/v) agarose solution in PBS (Gibco#10010) was prepared by mix-
ing the components in an autoclavable bottle via magnetic stirring. The suspension was autoclaved, then stirred again until homo-
geneous. The final concentration diluted in media for cell culture was 1%; volumes of the 6% stock solution (1.05 g/mL) were
measured gravimetrically.
For assay purposes, a mixture of cells and agarose (final concentration of agarose 1% in complete media, 40,000 cells/ml) was
prepared from cell and agarose stocks pre-warmed to 37ꢁC, and kept at 37ꢁC until it was dispensed via an automated liquid handling
system (Thermo Scientific Multidrop Combi) into 384-well ultra low-attachment plates (Corning, #3827); low-attachment plates were
used to ensure that any cells that settle will remain rounded rather than attaching to the plate, such plates eliminate the need for the
traditional bottom agar layer. The plates were incubated at 4ꢁC for 15-30 minutes to allow gelling to occur. After gelling of the agarose,
the cultures were incubated at 5% CO2 and 37ꢁC in a humidified incubator. Clonogenicity of cell lines was determined using this cul-
ture system for cell lines of interest.
Compound plates for screening were prepared in advance using acoustic dispensing (Labcyte Echo). On day 4 after plating of
cells, compounds to be screened (in DMSO stock) were diluted in media and added to the culture plates using an automated liquid
dispenser (BioMek FX). On day 7 after plating of cells,Alamar Blue dye (10 ml) was added to culture plates using an automated liquid
dispenser (BioMek FX), and the plates were incubated at 37ꢁC (incubation periods were optimized for each cell line to maximize the
Cell Chemical Biology 25, 1–13.e1–e11, June 21, 2018 e2

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
fluorescence signal and avoid reagent depletion). Fluorescence (Ex/Em 530 nm/590 nm) was measured (Envision 2104 multi-label
reader) to determine cell viability.
KRAS Knockdown: shRNA Expression and Gene Down-Regulation
shRNAs against KRAS were cloned in a Tet-inducible pLKO vector (pLKO_IRES_Puro_Dest). The shRNA sequences used were the
following: shRNA KRAS B, CACCGTTGGAGCTGGTGGtGTgGGCTTCAAGAGAGCCTACGCCACCAGCTCCAAC ; shRNA KRAS H,
CACCGGTCCTGCTGACAAATCAgGATTCAAGAGATCTTGATTTGTCAGCAGGACC.
The shRNA-expressing lentiviral vectors were packaged in 293T cells using the Trans-Lentiviral packaging kit (OpenBiosystems).
Culture medium containing virus was collected and filtered. Cells were transduced with virus and subjected to puromycin selection.
After selection, the surviving cells were grown in 3D culture as previously described, and doxycycline was added to the culture to
induce shRNA expression. Cells were re-fed with doxycycline every 2 or 3 days.
Downregulation of KRAS at the mRNA level was determined using a KRAS Taqman Gene Expression assay - Hs00270666_m1
(Thermofisher) and Cells-To-Ct kit (Ambion/ABI, #4399002). RPL37A - Hs01102345_m1 (ThermoFisher) was used as a housekeeping
gene for normalization. Cells were plated in a 96-well format. Seventy-two hours after shRNA induction, cells were washed with 1X
PBS. cDNA preparation from cell lysates and qPCR were carried out according to the manufacturer’s instructions.
To determine down-regulation of KRAS protein, cells were plated in 6-well plates. Seventy-two hours after shRNA induction, the
cells were washed with 1X PBS and lysed in 1X RIPA lysis buffer (Thermo 89901) containing 1X Halt Protease and phosphatase in-
hibitor (Thermo 78440). Samples were run on NuPage Novex 4-12% Bis-Tris polyacrylamide protein gels, and transferred to mem-
branes using the iBlot system (Invitrogen). Membranes were incubated with an anti-Ras antibody (Millipore Sigma #: 05-516) and
probed using secondary antibodies (LiCOR). b-actin (Millipore sigma #A5316) was used a loading control.
Cell Line Sensitivity Profiling
A set of 640 mechanistically diverse compounds were obtained from commercial sources or an internal library. A total of 374 cancer
cell lines of diverse origin were obtained from commercial vendors. Cell lines were chosen to sample diverse gene expression pro-
files. The cell lines represented 27 tissue types, with some lineages, such as liver, lung, pancreas, and stomach, being more highly
represented. The cell lines were cultured in vendor-recommended media using standard culturing techniques, either in a biological
safety cabinet or using an automated cell culture system (CompacT, Sartorius, Wilmington DE.). For dose-response studies, all com-
pounds were dissolved in DMSO and diluted in an 8-point dilution series, and an aliquot (10 mL) of each dilution was placed in a Lab-
cyte-certified LDV 384-well polystyrene plate. The starting stock concentrations used were between 0.1 mM to 10 mM, depending on
compound activity, solubility, and availability.
To screen the compounds for bioactivity against the cell line panel, a Labcyte Echo (Sunnyvale CA) liquid dispensing system was
used to transfer compound solution (25 nL) to 384-well assay plates (Greiner Bio-One, Monroe, NC). In addition to library compounds,
each assay plate also contained DMSO vehicle controls and a staurosporine (Sigma, St Louis MO) positive control. All liquid
dispensing and assay steps subsequent to the initial compound transfer were performed using a custom robotics platform (GNF,
San Diego CA). Cell suspension (25 mL) was added to each well, using a seeding density between 250 and 1500 cells per well, de-
pending on the cell line. Assay plates were incubated at 37ꢁC for 72-168 hours. Cell numbers were determined by adding
Hoechst34580 DNA staining dye (5 mL; Invitrogen) to a final concentration of 2 mg/mL. Assay plates were incubated for 3 hours at
37ꢁC, 9% paraformaldehyde (10 mL; EMS Hatfield, PA ) was added, and the plates were incubated at room temperature for 4 h.
Following fixation, cell counts were determined using an Acumen microplate cytometer (TTP Labtech Cambridge MA). Cell number
values were corrected for pattern artifacts, and normalized to the vehicle control and the number of untreated cells initially plated in
order to quantify effects on growth inhibition; growth inhibition was calculated as (median cell number-treated/median cell number
untreated)*100. All dose-response fitting was performed on normalized data using either a four-parameter logistic nonlinear regres-
sion model or a constant model (Genedata AG).
For DLD-1 cell lines, cells were plated at a density of 3000 cells/well and cultured for 2 days before drug and/or uridine treatment;
after 72 hours of treatment, viable cell density was quantified using the CellTiter Glo assay (Promega).
A subset of data containing a matrix of growth inhibition potencies from 372 cell lines (53 of which contain mutant KRAS) and 489
compounds was used for the analysis of the effect of KRAS mutation status on compound bioactivity. A t-test was performed for each
compound, comparing the mean of potencies in KRAS-mutant and wild-type cell lines. The Bonferroni correction for multiple com-
parisons was applied to the final p-values. The result is shown in a volcano plot. In a similar analysis, the effects of mutation status of
multiple oncogenes on the potencies of the lead thiazolimine was also compared.
Biochemical Pull-Down Experiment with Thiazolimine-Conjugated Beads
Protein pull-down using chemically modified beads was performed as described previously (Bantscheff et al., 2007). KP-4 cells sepa-
rately SILAC-labeled with either L-arginine and L-lysine (light) or L-arginine-13C6 and L-lysine-13C6-15N2 (heavy) were prepared in lysis
buffer containing 1% Nonidet P-40, 50 mM Tris/HCl, pH 7.5, 5% glycerol, 1.5 mM MgCl2, 150 mM NaCl, 1 mM EDTA, and protease
inhibitors (Complete tablets, Roche Applied Science, IN). Lysates were vortexed intermittently and chilled on ice for 10 min, followed
by centrifugation at 14,0003g. Protein concentrations of lysates were measured using the Protein BCA Assay (Thermo Pierce, CA),
e3 Cell Chemical Biology 25, 1–13.e1–e11, June 21, 2018

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
and lysates were diluted to a final protein concentration of 5 mg/ml in the lysis buffer. The heavy amino acid-labeled cell lysate was
pre-incubated with the active competitor compound Compound 2 (50 nM or 1 mM, in separate aliquots) for 1 h. The light amino acid-
labeled cell lysate was treated with DMSO as a control. Beads conjugated with the thiazolimine (200 ml) were added to each lysate
and incubated on an end-over-end rotator for 4 h at 4ꢁC. After incubation, the bead/lysate mixture was centrifuged at 10003g and the
supernatant was aspirated. The beads from each lysate were washed with lysis buffer to remove excess soluble small molecule
competitor. Beads from the heavy and light amino acid-labeled lysates were combined and washed (3X) with lysis buffer supple-
mented with 0.1% SDS buffer. After the final wash, beads were collected by centrifugation at 10003g, and the residual buffer
was carefully aspirated to avoid disturbing the beads.
SDS/PAGE and In-Gel Digestion
Proteins bound to thiazolimine-conjugated beads were reduced and alkylated on-bead in 10 mM DTT and 25 mM iodoacetamide
respectively. NuPAGE LDS sample buffer (Invitrogen) was added to bead mixtures, followed by heating to 60ꢁC for 40 min. Eluted
proteins were resolved on a 4n) was added to bead Bis-Tris gel with MES running buffer (Nupage, Invitrogen), and stained with Coo-
massie Blue (Simply Blue, Invitrogen). Gel lanes were excised into twelve pieces, each of which was cut into 1 x 1 mm cubes. The gel
pieces were further destained three times in a solution containing 50% acetonitrile and 50% ammonium bicarbonate (50 mM), then
dehydrated in 100% acetonitrile and completely dried in a vacuum concentrator (SpeedVac, Thermo Scientific) before addition of
sufficient trypsin solution (12.5 ng/L) to swell the gel pieces. The gel pieces were incubated in the trypsin solution at 37ꢁC overnight.
The peptides were extracted three times from the gel in 2:1 (vol/vol) acetonitrile/5% formic acid and dried in a vacuum concentrator to
remove organic solvents. The peptides were re-dissolved in an aqueous solution containing 0.1% formic acid and 5% acetonitrile for
mass spectrometry analysis.
iTRAQ Labeling of the Digested Pull-Down Samples
iTRAQ labeling was performed using the iTRAQ Reagent kit (AB SCIEX, CA) according to the manufacturer’s instructions. The pep-
tides extracted from the SDS-PAGE gel were dried and reconstituted in the labeling buffer (0.5 M tetraethylammonium bromide) pro-
vided with the kit. The peptides from the control lysate (treated with DMSO) were labeled with iTRAQ reagent 114, and the peptides
from lysate treated with 10 nM, 100 nM and 1 mM active competitor Compound 2 were labeled with iTRAQ reagents 115, 116, and
117, respectively. After labeling with iTRAQ reagents, the peptides were incubated at room temperature for 2 h and pooled.
Mass Spectrometry Analysis and Data Interpretation
LC-MS/MS analysis was carried out on a LTQ OrbitrapVelos mass spectrometer (Thermo Scientific, MA) equipped with a
Thermo nanospray source (Thermo, San Jose, CA) and coupled to an Eksigent NanoLC Ultra 2D system (ABsciex). Separation
˚
of peptide mixtures was achieved on a 75 mm I.D. microcapillary trap column packed with 18 cm of Magic C18resin (3 mm, 200 A,
MichromBioresources) with a gradient of acetonitrile (3% -35%) in 0.1% FA for 145 minutes at a flow rate of 300 nL/min.
A voltage of 3 kV was applied to generate electrospray. All the MS and MS/MS spectra were acquired in the data-dependent
mode. The full-scan spectra (from m/z 300-1800) were acquired at resolution 60,000 at m/z 400, with an accumulation target value
of 1 x 106. The top 20 most intense ions at a threshold above 500 counts were selected for fragmentation by CID at a normalized
collision energy of 40%. Dynamic exclusion was applied to reject ions from repeated MS/MS selection for 90 ms using monoisotopic
precursor selection. Singly-charged ions and those with unassigned charge state were also excluded from MS/MS.
The mass spectrometry data were processed using the Proteome Discoverer 1.3 software (Thermo Scientific). The MS/MS peak
lists were extracted and searched by MASCOT against a concatenated forward and reversed version of the Swiss-prot human data-
base. The precursor and fragment mass tolerances were set to 10 ppm and 0.5 Da, respectively. Cysteine carbamidomethylation was
searched as a fixed modification, while N-acetyl protein, N-pyroglutamine, oxidized methionine, phosphorylation of serine, threonine,
and tyrosine and SILAC labels on arginine and lysine were searched as variable modifications. The raw MS files were loaded to PD
1.3, and peptides were quantified with the MSQuant software embedded in the PD. MSQuant was used to calculate the ratio aver-
aged over the peptide elution time, and the assignments used for quantification were visually displayed and validated. To minimize
the false-discovery rate (FDR), all phosphopeptide identifications were filtered by MASCOT score. We accepted peptides based on
the criterion that the number of forward hits in the database was at least 200-fold higher than the number of reversed database hits,
which gives an estimated FDR of less than 1% (p < 0.01). The final filtering criteria were MASCOT score R 20 (p < 0.01), identification
of the peptide in the correct SILAC form (heavy or light), and presence of the correct number of lysine and arginine residues specified
by the mass difference observed in the full scan between the SILAC partners.
DHODH Biochemical Assay
DHODH activity was measured using a coupled assay with 2,6-dichloroindophenol (DCIP) as the terminal electron acceptor. Reac-
tions were carried out in UV-transparent plates (Corning #3675). The reaction mixture comprised HEPES buffer (100 mM, pH 7.5),
sodium chloride ( 150 mM), Triton X-100 (0.01%), dihydroorotate (10 mM), decylubiquinone (10 mM), and DCIP (60 mM). The reaction
was initiated by the addition of recombinant dihydroorotate dehydrogenase (5 nM, Creative Biomart part #DHODH-2486). Reaction
progress was monitored by measuring the absorbance at 600 nm (extinction coefficient of the reduced DCIP product, 18.8 M-1 cm-1;
absorbance measured using an Envision 2104 plate reader equipped with a monochromator).
Cell Chemical Biology 25, 1–13.e1–e11, June 21, 2018 e4

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Cell Cycle Analysis
KP-4 cells were treated with Compound 2 or DMSO (vehicle control) for 24 or 48 h, then collected and fixed with ice-cold ethanol.
Fixed cells were treated with RNAse, stained with propidium iodide, and subjected to flow cytometric analysis in duplicate to quantify
DNA content.
Mitochondrial Membrane Potential Assay
KP-4 cells were treated with DHODH inhibitors at the defined concentrations for 72 h. After treatment, cells were stained with
MitoTracker dye (ThermoFisher M7512; 300nM), washed with 1X PBS, fixed with paraformaldehyde (3.7%), and stained with
Hoechst dye (Life Technologies H1399). Image acquisition was performed using a Perkin Elmer Opera QEHS automated microscope
equipped with an Olympus 10X UPlanSApo objective. Hoechst33342 and Mitotracker Red CMXRos epifluorescence were detected
using 450/50 nm and 600/40 nm bandpass filters, respectively, with excitation from 405 nm and 561 nm solid-state lasers. Mitochon-
drial depolarization was quantified by measuring the decrease in Mitotracker Red CMXRos relative fluorescence intensity (RFI). Flu-
orescene intensity data were analyzed using Perkin Elmer Columbus software: nuclei were segmented based on Hoechst staining,
and through dilation of the nuclei segmentation mask, a ring region was created in which Mitotracker mean relative fluorescence
intensity was calculated.
LC-MS/MS Mass Spectrometric Metabolite Analysis of Cells and Tumor Tissues
LC-MS/MS analysis was carried out using an ABSciex QTrap 5500 mass spectrometer coupled to a 1290 Infinity UHPLC system
(Agilent). Samples were injected using a PAL HTC autosampler (LEAP Technologies). Separation of DHO and metabolites in the py-
rimidine synthesis pathway was achieved on an ACQUITY UPLC CSH Fluoro-Phenyl column (50 mm 3 2.1 mm, 1.7 mm Waters) at a
flow rate of 0.50 ml/min. The elution protocol used was the following: holding mobile phase A (0.1% acetic acid in water) at 95% for
0.5 min; ramping mobile phase B (0.1% acetic acid in acetonitrile) from 5 to 95% over 4 min; holding mobile phase B at 90%
for 0.5 min.
DHO and orotate were detected via their MRM transitions in the ESI negative mode; these were m/z 157 / 113 and m/z 155 /
111, for DHO and orotate, respectively. The transition for uridine was m/z 245 / 113 in the ESI positive mode. Instrument parameters
were the following: spray voltage, 4000 V; source temperature, 450ꢁC; ion source gas, 30 liter/h; and collision energy, 17 eV. Mass
spectrometric data were acquired and processed using the Analyst software (ABSciex).
For analysis of cellular samples, cells were treated with compounds of interest for 24 h. Cell pellets (ꢀ1 million cells) were sus-
pended in 80% methanol/water (500 ml) that had been chilled on dry ice. After brief sonication, the samples were centrifuged and
the supernatants were collected. The pellets were resuspended in the methanol/water mixture used previously, and the extraction
was repeated two additional times. The combined supernatants were dried in a centrifugal evaporator. The residues were reconsti-
tuted in water, and the resulting solutions were injected directly into the LC/MS-MS instrument for analysis of metabolites.
For samples from in vivo studies, tissue blocks were homogenized in 80% methanol/water (1 ml) using Precelly-24 with 2.8 mm
ceramic beads (5500 rpm for 2 x 30 sec). The homogenate was centrifuged at 14000 rpm for 10 min and the supernatants were trans-
ferred. The methanol was removed by evaporation under vacuum, and the residue was reconstituted in water (2 ml) for the LC/MS/
MS analysis.
CE-TOFMS Mass Spectrometric Metabolite Analysis of Cells and Tumor Tissues
DLD1 KRAS (+/-) HD-105-002 and DLD1 KRAS (G13D/-) HD-105-011 were obtained from Horizon Discovery. Cells were plated in
RPMI 1640 with glutamine. Culture medium was aspirated the next day and cells treated with 2 mM Brequinar in absence or presence
of 500 mM uridine in C13 glutamine labeled RPMI for 4 hours. For extraction of intracellular metabolites, culture medium was aspirated
and cells were washed twice with a 5% mannitol solution. The cells were treated with 800 mL of methanol and 550 mL of Milli-Q water
containing internal standards (H3304-1002, Human Metabolome Technologies, Inc., Tsuruoka, Japan). The extract was centrifuged
at 2,300 3g and 4ꢁC for 5 min and then 800 mL of upper aqueous layer was centrifugally filtered through a Millipore 5-kDa cutoff filter
at 9,100 3g and 4ꢁC for 120 min to remove proteins. The filtrate was centrifugally concentrated and re-suspended in 50 mL of Milli-Q
water for CE-MS analysis.
Metabolome measurements were carried out through a facility service at Human Metabolome Technology Inc., Tsuruoka, Japan.
Hierarchical cluster analysis (HCA) and principal component analysis (PCA) were performed using proprietary software: PeakStat
and SampleStat, respectively. Detected metabolites were plotted on metabolic pathway maps using VANTED (Visualization and
Analysis of Networks containing Experimental Data) software.
Synergistic Combinations in KRAS Mutant Cell Context
A 12x12 dose matrix of co-plated enhancer and enhancee compounds was used to evaluate whether combination exhibited synergy.
Cell culture was carried out in 2D format, as described in ‘‘Cell Line Sensitivity Profiling.’’
Compounds were diluted in a 12-point dilution series and dispensed in 40 ml aliquots into each well of a Labcyte-certified 384-well
polypropylene plate. The enhancee compounds, brequinar and a DHODH inhibitor from the thiazolimine series (Compound 2), were
diluted and plated (in 25 nL aliquots) in the same format as the enhancer compounds using a Labcyte Echo (Sunnyvale CA)
dispensing system. Cell culture and cell-based screening for growth inhibition effects (using KP-4 cells at a density of 500 cells/
well, and a 72-hour treatment period) were carried out as described for ‘‘Cell Line Sensitivity Screening.’’
e5 Cell Chemical Biology 25, 1–13.e1–e11, June 21, 2018

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Three types of synergy scores, including Loewe, Bliss and HSA, were generated using software for analysis of in vitro pharma-
cology data (GeneData AG). For each type of synergy score (Loewe, Bliss, and HSA), a robust Z-test was performed to identify po-
tential combination partners for Compound 2. Specifically, a robust Z score was calculated for each compound combination as the
difference between the synergy score from given combination and the median of synergy scores from self-cross experiments,
divided by the median absolute deviation of synergy scores from self-cross experiments. The distributions of synergy scores from
the self-cross experiments were found to be approximately normal for all three synergy score types, enabling the generation of
two-tailed p values for each robust Z score. Since multiple robust Z tests were performed, multiple testing adjustment of p values
was performed, using the Benjamini-Hochberg method. Any compound combination with a synergy score satisfying the criteria
of Z > 0 and adjusted p value <= 0.05 was flagged as ‘‘synergistic.’’
General Synthetic Methods
Reagents and solvents were purchased from commercial sources and were used without further purification. All reactions were run
under nitrogen atmosphere unless otherwise noted. Purification of reaction products was carried out by flash column chromatog-
raphy using Teledyne Isco Combiflash Rf and Redisep normal-phase silica flash columns (230-400 mesh). Analytical thin layer chro-
matography was performed on EMD Millipore DC-Kieselgel 60 F254 plates. Visualization was accomplished with UV light, potassium
permanganate, ninhydrin or Hanessian stain followed by heating. Proton nuclear magnetic resonance spectra (1H NMR) were
recorded on a Varian Inova 400MHz spectrometer and are reported in ppm using solvent as an internal standard (DMSO-d6 at
2.50 ppm, methanol-d4 at 3.31 ppm, acetone-d6 at 2.09 ppm). Data are reported as br = broad peak, s = singlet, d = doublet,
t = triplet, q = quadruplet, p = quintuplet, dd = doublet of doublets, ddd = doublet of doublet of doublets, m = multiplet; integration;
coupling constant(s) in Hz. HPLC-MS analysis was carried out on a Waters Acquity UPLCꢀ with autosampler, PDA detector and SQ
mass spectrometer using Acquity UPLCꢀ BEH C18 1.7um 2.1 x 50mm column. Preparative HPLC was performed on a Gilson Pre-
parative HPLC system using a SunFireTM Prep C18 OBDTM 5uM 19x100mm column.
Synthetic Protocols
Scheme 1. Synthetic Routes
Cell Chemical Biology 25, 1–13.e1–e11, June 21, 2018 e6

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Procedure 1a
1-(4-Methyl-2-phenylthiazol-5-yl)ethanone (5)
To a solution of benzothioamide (5 g, 36.44 mmol) in ethanol (36.44 ml, 36.44 mmol) was added 3-chloro-2,4-pentanedione (4.24 ml,
36.44 mmol) at room temperature. The mixture was stirred at reflux for 4 h. An LC/MS indicated complete reaction. The mixture
was allowed to cool down to room temperature. A precipitate formed. It was concentrated under reduced pressure. The residue
1
was recrystallized from hexanes to provide 1-(4-methyl-2-phenylthiazol-5-yl)ethanone (3.6 g, 16.57 mmol, 45.5 % yield). H NMR
(400 MHz, DMSO-d6) d 8.07 – 7.93 (m, 2H), 7.61 – 7.46 (m, 3H), 2.71 (s, 3H), 2.58 (s, 3H). ESI-MS m/z 218.2 [(M + H)+; calcd. for
C
12H12NOS: 218.06].
Procedure 1b
2-Bromo-1-(4-methyl-2-phenylthiazol-5-yl)ethan-1-one (6)
To a solution of 1-(4-methyl-2-phenylthiazol-5-yl)ethanone (3 g, 13.81 mmol) in dichloromethane (69.03 ml, 13.81 mmol) was added
bromine (924.67 ml, 17.95 mmol). The solution was stirred at 40ꢁC overnight. The mixture was concentrated under reduced pressure.
The residue was triturated in methanol then dried under vacuum to give 2-bromo-1-(4-methyl-2-phenylthiazol-5-yl)ethan-1-one
(3.26 g, 10.99 mmol, 79.6 %). 1H NMR (400 MHz, DMSO-d6) d 7.97 (m, 2H), 7.64 – 7.40 (m, 3H), 4.75 (s, 2H), 2.69 (s, 3H). ESI-MS
m/z 296.1 [(M + H)+; calcd. for C12H11BrNOS: 295.97].
2-Bromo-1-(4-methyl-2-(pyrazin-2-yl)thiazol-5-yl)ethan-1-one
This compound was obtained from a commercial source (Oakwood Chemical, West Columbia, South Carolina).
Procedure 2
1-(4-Hydroxy-2-methylphenyl)-3-methylthiourea (7)
To a solution of 4-amino-3-methylphenol (1 g, 8.12 mmol) in THF (27.07 ml, 8.12 mmol) was added isothiocyanatomethane (610.36 ml,
8.93 mmol). The mixture was stirred at 60ꢁC overnight then concentrated under reduced pressure. The residue was purified by flash
chromatography (using a gradient of 10% ethyl acetate in hexanes to 100% ethyl acetate) to afford 1-(4-hydroxy-2-methylphenyl)-3-
methylthiourea (1.05 g, 5.35 mmol, 65.9 % yield) 1H NMR (400 MHz, DMSO-d6) d 9.30 (s, 1H), 8.83 (s, 1H), 6.98 (d, J = 57.0 Hz, 1H),
e7 Cell Chemical Biology 25, 1–13.e1–e11, June 21, 2018

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
6.84 (d, J = 8.4 Hz, 1H), 6.60 (d, J = 2.6 Hz, 1H), 6.54 (dd, J = 8.4, 2.7 Hz, 1H), 2.81 (d, J = 4.5 Hz, 3H), 2.02 (s, 3H). ESI-MS m/z
197.1 [(M + H)+; calcd. for C9H13N2OS: 197.08].
1-(2-Chloro-4-hydroxyphenyl)-3-methylthiourea (8)
To a solution of 4-amino-3-chlorophenol (0.749 g, 5.22 mmol) in THF (17.39 ml, 5.22 mmol) was added isothiocyanatomethane
(427.79 ml, 6.26 mmol). The solution was stirred at 60ꢁC overnight. The resulting suspension was cooled to room temperature and
filtered. The remaining solid was washed with ethanol to produce 1-(2-chloro-4-hydroxyphenyl)-3-methylthiourea (785 mg,
3.62 mmol, 69.4 % yield) 1H NMR (400 MHz, methanol-d4) d 7.16 (d, J = 8.7 Hz, 1H), 6.88 (d, J = 2.7 Hz, 1H), 6.73 (dd, J = 8.7,
2.7 Hz, 1H), 2.97 (s, 3H). ESI-MS m/z 217.1 [(M + H)+; calcd. for C8H10ClN2OS: 217.02].
Tert-butyl (3-(3-(4-hydroxy-2-methylphenyl)thioureido)propyl)carbamate (9)
To a solution of 4-amino-3-methylphenol (50 mg, 406.00 mmol) in THF (2.0 ml, 24.41 mmol) was added tert-butyl (3-isothiocyanato-
propyl)carbamate (131.73 mg, 609.00 mmol). The mixture was stirred at room temperature for 3 hours. An LC/MS indicated reaction
completion. The mixture was concentrated under reduced pressure. The residue was purified by flash chromatography (using a
gradient of 10% ethyl acetate in hexanes to 100% ethyl acetate) to provide tert-butyl (3-(3-(4-hydroxy-2-methylphenyl)thioureido)
propyl)carbamate (117 mg, 344.67 mmol, 84.9 % yield). 1H NMR (400 MHz, acetone-d6) d 8.44 (br s, 1H), 8.30 (br s, 1H), 6.99
(d, J = 8.4 Hz, 1H), 6.83 (br s, 1H), 6.75 (d, J = 2.7 Hz, 1H), 6.68 (dd, J = 8.4, 2.8 Hz, 1H), 6.09 (br s, 1H), 3.63 (q, J = 6.3 Hz, 2H),
3.06 (q, J = 6.3 Hz, 3H), 2.17 (s, 3H), 1.74 – 1.61 (m, 2H), 1.36 (s, 9H). ESI-MS m/z 340.2 [(M + H)+; calcd. for C16H26N3O3S: 340.17].
Procedure 3
(Z)-3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenol (10)
A suspension of 1-(2-chloro-4-hydroxyphenyl)-3-methylthiourea (35 mg, 161.52 mmol) and 2-bromo-1-(4-methyl-2-(pyrazin-2-yl)
thiazol-5-yl)ethanone (48.16 mg, 161.52 mmol) in isopropanol (2.0 mL) was stirred with microwave heating at 100ꢁC for 1 h. The
mixture was then dissolved in methanol (1.0 mL) and triethylamine (1.0 mL). The solution was concentrated under reduced pressure.
The residue was purified via flash chromatography on silica gel (using a gradient of 30% to 100% of ethyl acetate in hexanes, followed
by 10% methanol in dichloromethane) to give (Z)-3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)
phenol (43 mg, 103.39 mmol, 64.0 % yield). 1H NMR (400 MHz, DMSO-d6) d 9.47 (s, 1H), 9.32 (d, J = 1.5 Hz, 1H), 8.78 (d, J =
2.5 Hz, 1H), 8.73 (dd, J = 2.5, 1.5 Hz, 1H), 6.93 (d, J = 8.6 Hz, 1H), 6.86 (d, J = 2.7 Hz, 1H), 6.71 (dd, J = 8.6, 2.7 Hz, 1H),
6.58 (s, 1H), 3.26 (s, 3H), 2.46 (s, 3H). ESI-MS m/z 416.2 [(M + H)+; calcd. for C18H15ClN5OS2: 416.04].
Cell Chemical Biology 25, 1–13.e1–e11, June 21, 2018 e8

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
(Z)-3-(2-((4-hydroxy-2-methylphenyl)imino)-4’-methyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-3(2H)-yl)propan-1-aminium
2,2,2-trifluoroacetate (11)
2-bromo-1-(4-methyl-2-(pyrazin-2-yl)thiazol-5-yl)ethanone (102.77 mg, 344.67 mmol) and tert-butyl (3-(3-(4-hydroxy-2-methyl-
phenyl)thioureido)propyl)carbamate (117 mg, 344.67 mmol) were dissolved in dry isopropanol (2 mL). The solution was stirred with
microwave heating at 80ꢁC for 8 hours. Triethylamine (1.0 mL) was added. The mixture was concentrated under reduced pressure.
The residue was purified by flash chromatography (using a gradient of 100% dichloromethane to 15% methanol in dichloromethane).
The product was dissolved in dichloromethane (2.0 mL). TFA (1.0 mL) was added. The mixture was stirred at room temperature for
2 h. LC-MS indicated reaction completion. The mixture was concentrated under reduced pressure. The residue was dissolved in
DMSO and purified by preparative HPLC to afford (Z)-4-((3-(3-aminopropyl)-4’-methyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-yli-
dene)amino)-3-methylphenol 2,2,2-trifluoroacetate (88 mg, 112.73 mmol, 32.7 % yield) after lyophilization of the fraction containing
1
the pure product. H NMR (400 MHz, DMSO-d6) d 9.36 (d, J = 1.5 Hz, 1H), 8.83 (d, J = 2.5 Hz, 1H), 8.77 (dd, J = 2.5, 1.5 Hz, 1H),
7.78 (br s, 3H), 7.04 (br s, 1H), 6.77 (d, J = 2.2 Hz, 1H), 6.70 (dd, J = 8.5, 2.3 Hz, 1H), 4.00 – 3.90 (m, 2H), 2.95 – 2.68 (m, 2H),
2.54 (s, 1H), 2.48 (s, 3H), 2.18 (s, 3H), 1.93 (p, J = 7.9 Hz, 2H). ESI-MS m/z 439.2 [(M + H)+; calcd. for C21H23N6OS2: 439.14
(Z)-N-(3-(2-((4-hydroxy-2-methylphenyl)imino)-4’-methyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-3(2H)-yl)propyl)
acetamide (12)
Acetic anhydride (1.90 ml, 20.18 mmol was added dropwise to a solution of (Z)-4-((3-(3-aminopropyl)-4’-methyl-2’-(pyrazin-2-yl)-[4,5’-
bithiazol]-2(3H)-ylidene)amino)-3-methylphenol 2,2,2-trifluoroacetate (15 mg, 19.22 mmol) in pyridine (64.05 ml, 19.22 mmol) ). The
mixture was stirred at room temperature for 30 minutes. An LC/MS indicated reaction completion. The reaction mixture was concen-
trated under reduced pressure. The residue was purified using flash chromatography (with a gradient of 30% to 100% ethyl acetate in
hexanes) to provide (Z)-N-(3-(2-((4-hydroxy-2-methylphenyl)imino)-4’-methyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-3(2H)-yl)propyl)acet-
amide (4.73 mg, 9.84 mmol, 51.2 % yield). 1H NMR (400 MHz, methanol-d4) d 9.37 (d, J = 1.4 Hz, 1H), 8.68 (d, J = 2.5 Hz, 1H),
8.66 (dd, J = 2.5, 1.5 Hz, 1H), 6.83 (d, J = 8.4 Hz, 1H), 6.73 (d, J = 2.7 Hz, 1H), 6.65 (dd, J = 8.4, 2.8 Hz, 1H), 6.43 (s, 1H), 3.95 –
3.81 (m, 2H), 3.14 (t, J = 6.5 Hz, 2H), 2.66 (s, 1H), 2.50 (s, 3H), 2.16 (s, 3H), 1.90 (p, J = 6.8 Hz, 2H), 1.75 (s, 3H). ESI-MS m/z
481.2 [(M + H)+; calcd. for C23H25N6O2S2: 481.15].
(Z)-3-chloro-4-((3,4’-dimethyl-2’-(4-(trifluoromethyl)phenyl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenol (13)
This compound was synthesized as previously described for (Z)-3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-yli-
dene)amino)phenol. 1H NMR (400 MHz, DMSO-d6) d 9.93 (br s, 1H), 8.15 (d, J = 8.1 Hz, 2H), 7.87 (d, J = 8.2 Hz, 2H), 7.13 (d, J = 8.5 Hz,
1H), 6.94 (d, J = 2.7 Hz, 1H), 6.89 (br s, 1H), 6.79 (dd, J = 8.6, 2.7 Hz, 1H), 3.37 (s, 3H), 2.42 (s, 3H). ESI-MS m/z 482.0 [(M + H)+; calcd.
for C21H16ClF3N3OS2: 482.04].
(Z)-3-chloro-4-((3,4’-dimethyl-2’-phenyl-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenol (14)
This compound was synthesized as previously described for (Z)-3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-yli-
dene)amino)phenol. 1H NMR (400 MHz, DMSO-d6) d 7.97 (dd, J = 6.5, 3.1 Hz, 2H), 7.64 – 7.47 (m, 3H), 7.22 (s, 1H), 6.99 (d, J = 2.5 Hz,
1H), 6.84 (dd, J = 8.5, 2.7 Hz, 1H), 3.42 (s, 3H), 2.43 (s, 3H). ESI-MS m/z 414.0 [(M + H)+; calcd. for C20H17ClN3OS2: 414.05].
e9 Cell Chemical Biology 25, 1–13.e1–e11, June 21, 2018

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
Procedure 4
(Z)-tert-butyl (3-(3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenoxy)propyl)
carbamate (15)
To a solution of (Z)-3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenol (50 mg, 120.22 mmol)
in THF (0.4 mL) was added triphenylphosphine (63.06 mg, 240.43 mmol) and tert-butyl (3-hydroxypropyl)carbamate (41.10 ml,
240.43 mmol). (E)-diisopropyl diazene-1,2-dicarboxylate (47.34 ml, 240.43 mmol) was then added dropwise. The reaction mixture
was stirred at room temperature for 3 h. Upon reaction completion, the mixture was purified by preparative HPLC, using a gradient
of 10% acetonitrile in water (containing 0.1% trifluoroacetic acid) to 100% in acetonitrile (containing 0.1% trifluoroacetic acid), to give
(Z)-tert-butyl (3-(3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenoxy)propyl)carbamate (43 mg,
75.03 mmol, 62.4 % yield). 1H NMR. ESI-MS m/z 573.4 [(M + H)+; calcd. for C26H30ClN6O3S2: 573.15].
Procedure 5
(Z)-3-(3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenoxy)propan-1-amine
bis(2,2,2-trifluoroacetate) (16)
To a solution of (Z)-tert-butyl (3-(3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenoxy)propyl)
carbamate (43 mg, 75.03 mmol) in dichloromethane (2 mL) was added trifluoroacetic acid (2 mL). The mixture was stirred at room
temperature for 30 minutes. LC/MS indicated reaction completion. The solvents were removed under reduced pressure to afford
(Z)-3-(3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenoxy)propan-1-amine bis(2,2,2-trifluoroa-
cetate) (49 mg, 69.89 mmol, 93.2 % yield). 1H NMR. ESI-MS m/z 473.3 [(M + H)+; calcd. for C21H22ClN6OS2: 473.10].
Procedure 6
(Z)-N-(3-(3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenoxy)propyl)
acetamide (17)
(Z)-3-chloro-4-((3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenol (5.8 mg, 13.95 mmol), tert-butyl (4-hy-
droxypropyl)carbamate (5.28 mg, 30.13 mmol), and triphenylphosphine (7.32 mg, 27.89 mmol) were dissolved in THF (46.48 ml,
13.95 mmol). (E)-diisopropyl diazene-1,2-dicarboxylate (5.49 ml, 27.89 mmol) was added dropwise at 0ꢁC. The mixture was stirred
at room temperature for 2 h. LC/MS indicated reaction completion. The reaction mixture was concentrated under reduced pressure.
The residue was dissolved in dichloromethane (1.0 mL). TFA was added (1.0 mL) and the mixture was stirred for 30 minutes. LC/MS
indicated full deprotection. The mixture was concentrated under reduced pressure. The residue was dissolved in pyridine (1.0 mL).
Acetic anhydride (7 ml, 74.19 mmol) was added. The reaction mixture was stirred at room temperature for 20 min and concentrated
under reduced pressure. The residue was purified using preparative HPLC, with a gradient of 10% acetonitrile in water (containing
Cell Chemical Biology 25, 1–13.e1–e11, June 21, 2018 e10

Please cite this article in press as: Koundinya et al., Dependence on the Pyrimidine Biosynthetic Enzyme DHODH Is a Synthetic Lethal Vulnerability in
Mutant KRAS-Driven Cancers, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.005
0.1% trifluoroacetic acid) to 100% in acetonitrile (containing 0.1% trifluoroacetic acid), to give (Z)-N-(3-(3-chloro-4-((3,4’-dimethyl-2’-
(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)amino)phenoxy)propyl)acetamide 2,2,2-trifluoroacetate (5.23 mg, 8.31 mmol, 59.6 %
1
yield). H NMR (400 MHz, DMSO-d6) d 9.33 (d, J = 1.5 Hz, 1H), 8.79 (d, J = 2.5 Hz, 1H), 8.74 (dd, J = 2.5, 1.5 Hz, 1H), 7.89 (br t,
J = 5.4 Hz, 1H), 7.17 (br d, J = 8.7 Hz, 1H), 7.12 (d, J = 2.8 Hz, 1H), 6.94 (dd, J = 8.8, 2.8 Hz, 1H), 6.79 (br s, 1H), 3.99 (t, J =
6.3 Hz, 2H), 3.34 (s, 3H), 3.17 (q, J = 6.8 Hz, 2H), 2.46 (s, 3H), 1.88 – 1.73 (m, 5H). ESI-MS m/z 515.3 [(M + H)+; calcd. for
C23H24ClN6O2S2: 515.11].
Procedure 7
Loading of Thiazolimine onto Sepharose FastFlow Beads (18)
Sepharose beads (8 mL of a 1:1 suspension in isopropanol) were separated in two 15 mL conical tubes (Falcon) and centrifuged at
1200 rpm for 2 minutes. The supernatant was removed. DMSO (10 mL) was added to each tube. The beads were resuspended and
the tubes were centrifuged as previously described. This centrifugation and re-suspension process was repeated twice, then once
more with the exception that anhydrous DMSO was used to suspend the beads. The supernatant and excess beads were removed to
adjust the volume to 2 mL of beads in each tube. Anhydrous DMSO (2.0 mL) was added to each tube, followed by a 50 mM DMSO
solution of (Z)-4-(3-aminopropoxy)-2-chloro-N-(3,4’-dimethyl-2’-(pyrazin-2-yl)-[4,5’-bithiazol]-2(3H)-ylidene)aniline 2,2,2-trifluoroa-
cetate (a total of 200.00 ml, 10 mmol). Triethylamine (60 ml, 430.48 mmol) was added to each tube and the beads were resuspended.
The suspension was centrifuged at 1200 rpm for 2 minutes. An aliquot (20 mL) of the supernatant in each tube was diluted 1:1 with
DMSO. The samples were analyzed using LC-MS to quantify the initial amount free amine in the reaction.
The beads were stirred at room temperature via inversion overnight (16 h) then centrifuged. An LC/MS of the aliquot of the super-
natants no longer showed the presence of free amine. Ethanolamine (200 ml, 3.31 mmol) was then added to each tube. The mixture
was stirred at room temperature overnight. The beads were centrifuged. The supernatant was removed, and the beads were resus-
pended in anhydrous DMSO (10 mL) and centrifuged. The supernatant was removed and the beads were washed three times with
anhydrous ethanol (10 mL portions). The recovered beads from both tubes were combined and suspended in anhydrous ethanol in a
1:1 ratio.
QUANTIFICATION AND STATISTICAL ANALYSIS
Statisitical details can be found in the STAR Methods sub-section for each experiment, or in the Results section. Uncertainties are
expressed as the standard error in the mean. For in vitro experiments, n represents the number of independent biological replicates;
for in vivo experiments, n represents the number of animals. For correlations of potencies in biochemical/cell-based experiments,
Spearman’s rank order correlation was used, with use of the t-test for calculation of p values (reported in the Results section).
e11 Cell Chemical Biology 25, 1–13.e1–e11, June 21, 2018