Discovery of 2-((1H-benzo[d]imidazol-1-yl)methyl)-4H-pyrido[1,2-a] pyrimidin-4-ones as novel PKM2 activators

Article abstract of DOI:10.1016/j.bmcl.2013.03.090

The M2 isoform of pyruvate kinase is an emerging target for antitumor therapy. In this letter, we describe the discovery of 2-((1H-benzo[d]imidazol-1- yl)methyl)-4H-pyrido[1,2-a]pyrimidin-4-ones as potent and selective PKM2 activators which were found to have a novel binding mode. The original lead identified from high throughput screening was optimized into an efficient series via computer-aided structure-based drug design. Both a representative compound from this series and an activator described in the literature were used as molecular tools to probe the biological effects of PKM2 activation on cancer cells. Our results suggested that PKM2 activation alone is not sufficient to alter cancer cell metabolism.

Full text of DOI:10.1016/j.bmcl.2013.03.090

Accepted Manuscript
Discovery of 2-((1H-benzo[d]imidazol-1-yl)methyl)-4H-pyrido[1,2-a]pyrimi‐
din-4-ones as Novel PKM2 Activators
Chuangxing Guo, Angelica Linton, Mehran Jalaie, Susan Kephart, Martha
Ornelas, Mason Pairish, Samantha Greasley, Paul Richardson, Karen Maegley,
Michael Hickey, John Li, Xin Wu, Christy Ji, Zhi Xie
PII:
S0960-894X(13)00418-6
DOI:
http://dx.doi.org/10.1016/j.bmcl.2013.03.090
BMCL 20326
Reference:
To appear in:
Bioorganic & Medicinal Chemistry Letters
Received Date:
Revised Date:
Accepted Date:
4 February 2013
20 March 2013
22 March 2013
Please cite this article as: Guo, C., Linton, A., Jalaie, M., Kephart, S., Ornelas, M., Pairish, M., Greasley, S.,
Richardson, P., Maegley, K., Hickey, M., Li, J., Wu, X., Ji, C., Xie, Z., Discovery of 2-((1H-benzo[d]imidazol-1-
yl)methyl)-4H-pyrido[1,2-a]pyrimidin-4-ones as Novel PKM2 Activators, Bioorganic & Medicinal Chemistry
Letters (2013), doi: http://dx.doi.org/10.1016/j.bmcl.2013.03.090
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Graphic Abstract
Discovery of 2-((1H-benzo[d]imidazol-1-yl)methyl)-4H-pyrido[1,2-a]pyrimidin-4-ones as
Novel PKM2 Activators
Chuangxing Guo,* Angelica Linton,* Mehran Jalaie, Susan Kephart, Martha Ornelas, Mason Pairish, Samantha
Greasley, Paul Richardson, Karen Maegley, Michael Hickey, John Li, Xin Wu, Christy Ji, Zhi Xie
Pfizer Global Research and Development, 10770 Science Center Drive, San Diego, CA 92121
O
N
N
N
N
SBDD
O
N
N
N
From HTS
N
8
1
3.5uM
0.159uM
PKM2 AC50
Max. Act. 148%
MW 380.44
Max. Act. 192%
MW 290.3
CLogP: 3.92
LipE=1.54
CLogP: 2.08
LipE=4.7
2-((1H-benzo[d]imidazol-1-yl)methyl)-4H-pyrido[1,2-a]pyrimidin-4-ones were identified as novel PKM2 activators with a novel
binding mode. The original lead was optimized into an efficient series via computer-aided structure-based drug design. Biological
studies on the representative PKM2 activators suggest that PKM2 activation along is not sufficient to alter cancer cell metabolism.

Discovery of 2-((1H-benzo[d]imidazol-1-yl)methyl)-4H-pyrido[1,2-a]pyrimidin-4-ones as
Novel PKM2 Activators
Chuangxing Guo,* Angelica Linton,* Mehran Jalaie, Susan Kephart, Martha Ornelas, Mason
Pairish, Samantha Greasley, Paul Richardson, Karen Maegley, Michael Hickey, John Li, Xin
Wu, Christy Ji, Zhi Xie
Oncology Medicinal Chemistry, Pfizer Worldwide Research & Development, San Diego, CA
92121, USA
Abstract
The M2 isoform of pyruvate kinase is an emerging target for antitumor therapy. In this
paper, we describe the discovery of 2-((1H-benzo[d]imidazol-1-yl)methyl)-4H-pyrido[1,2-a]pyrimidin-4-
ones as potent and selective PKM2 activators which were found to have a novel binding mode. The
original lead identified from high throughput screening was optimized into an efficient series via
computer-aided structure-based drug design. Both a representative compound from this series and an
activator described in the literature were used as molecular tools to probe the biological effects of PKM2
activation on cancer cells. Our results suggested that PKM2 activation along is not sufficient to alter
cancer cell metabolism.
In recent years there has been increased interest in the mechanisms of fundamental metabolic
alterations during malignant transformation. The reprogramming of cellular energy to support abnormal
growth and proliferation of cancer cells is considered an emergent hallmark of cancer.1
Understanding the biological differences between normal cells and cancer cells is crucial for
formulating strategies to treat cancer. More than 75 years ago the observation was made that tumor cells
have an elevated rate of glucose uptake but reduced rates of oxidative phosphorylation,2 now known as
the Warburg effect. Recently, the involvement of the M2 splice isoform of pyruvate kinase in the shift in
cellular metabolism to aerobic glycolysis has been proposed.3 Tumor cells exclusively express the
embryonic M2 isoform of pyruvate kinase (PKM2),4 which can switch from a less active dimeric form to
a highly active tetrameric form. PKM2 is proposed to shuttle between these two forms to regulate the
metabolism of glucose. The low activity dimeric form supports cell growth by increasing glycolytic
intermediates necessary for biosynthetic processes, but when energy levels fall, the enzyme can switch to
the high activity tetrameric form and facilitate oxidative phosphorylation.3,4 While it has been reported
that PKM2 in tumors exists predominantly as the less active dimer, the exact mechanism of its
involvement in tumorigenesis and the Warburg effect is still not well understood. Consequently, there

have been significant research efforts to elucidate pyruvate kinase regulation and to assess PKM2 as a
target for cancer therapy.5
Christofk et al.3 showed that switching pyruvate kinase expression to the M1 isoform results in a
reversal of the Warburg effect, as measured by reduced lactate production and increased oxygen
consumption. These observations correlated to a reduced ability to form tumors in nude mouse
xenografts.5d It has been hypothesized that pharmacological intervention to restore the activity of tumor
PKM2 to levels seen with PKM1 in normal cells may force cancer cells to adopt a metabolic state
characteristic of normal cells, thus suggesting PKM2 activation as a novel strategy to inhibit tumor
growth. Several small molecule activators of PKM2 have been reported in literature.6
Here we report the discovery of a series of pyridopyrimidinone analogs as potent activators of
PKM2 with good HLM stability,7 permeability, solubility and selectivity. Lead compounds resulting
from our discovery program were used as tool molecules to probe the effects of PKM2 activation on
cancer cell metabolism and proliferation. The resulting structure-activity relationships, effects on oxygen
consumption and lactate production, and cancer cell proliferation are discussed.
PKM2 catalyzes the conversion of phosphoenolpyruvate (PEP) to pyruvate and generates ATP in
the process. A high-throughput screening (HTS) assay that monitored production of ATP by
luminescence was developed, and Compound 1 (Figure 1) was identified as one of the most active hits,
showing 155% activation at 10 M. The active compounds in the HTS assay were further characterized
using a lactate dehydrogenase (LDH)-coupled enzyme assay. In this assay, Compound 1 gave an AC50 of
3.5 M and a maximal activation of 148% compared to FBP.8 Compound 1 was subsequently modeled 9
into the activator binding site of the published PKM2 co-crystal structure (Figure 1).10 According to
previous reports, the activator binding site is located at the dimer-dimer interface.5e In this model of
Compound 1, it appeared that the benzyl group presented a sub-optimal interaction with the PKM2
protein and in order to accommodate the benzyl group, two water molecules which are part of a H-bond
network to the carbonyl main chain of ASN350 need to be displaced. However, these water molecules are
part of an intricate network of H-bonds, and displacement by the benzyl moiety would leave an
unsatisfied H-bond network (generating a desolvation penalty), ultimately leading to suboptimal binding
affinity of Compound 1.

N
N
O
N
MW 380.44
CLogP: 3.92
N
1
3.5uM
LDH AC50
Max. Act. 148%
LE=0.26
LipE=1.54
Figure 1 Compound 1 modeled into the reported activator binding site. Compound 1 (yellow) is docked into the
activator binding site of published PKM2 co-crystal structure PDB IDs = 3GQY.
Several simple analogs were designed and synthesized to test the effect of replacing the sub-
optimal benzyl group with a smaller substituent. The analog lacking the phenyl group lost biochemical
activity (2, Figure 2, max. activation: 73% @ 10 uM), but further removal of the C-2 methyl substituent
resulted in a 5.4 fold improvement in potency (3, Figure 2, AC50 = 755 nM). The simple unsubstituted
core was found to be 10-fold more potent in the PKM2 activation assay (4, AC50 = 340 nM, Figure 2),
leading to a significant increase in efficiency (LipE = 5.14 vs 1.54 )11 compared to the original HTS hit 1.
Figure 2. Initial optimization of HTS Pyridopyrimidinone lead 1.
An X-ray co-crystal structure of Compound 412 with PKM2 protein was solved, revealing the
binding conformation (Figure 3). When compared to several published co-crystal structures including
13
3GR4 / 3H6O / 3GQY, (Figure 3),
the 2-((1H-benzo[d]imidazol-1-yl)methyl)-4H-pyrido[1,2-
a]pyrimidin-4-one core of compound 4 adopts an unique binding mode. The benzimidazole fragment

reaches a region of the activator pocket that is not occupied by an activator in any of the published crystal
structures. Furthermore, one of the imidazole nitrogens forms a hydrogen-bond to LYS311 which is
normally part of a salt bridge to ASP354. Such an interaction has not been observed in other published
co-crystal structures. In general, the ligand is bound to the activator site in two separate conformations
with equal occupancy, with the core rotated by 47 degrees in relation to other activators.14 For simplicity
figure 3 shows only the ligand bound to the A/B dimer activator site and not the one bound to the C/D
dimer. The pyrimidone ring sits between the two Phe26 residues from monomers A (pink) and B (cyan)
forming  interactions among the aromatic rings. The carbonyl interacts with a bridging water
molecule, which in turn interacts with the carbonyl of the backbone of TYR390 (Figure 4). The
observation of a new binding mode presented new opportunities for the design of more potent PKM2
activators.
Figure 3. Comparison of X-Ray co-crystal structure of compound 4 (PDB: 4JPG, magenta) and published activator
structures 3GR4 (gray) / 3H6O (orange) / 3GQY (green). PKM2 monomer A (pink), PKM2 monomer B (cyan).
The preparation of these pyridopyrimidinone analogs is outlined in Scheme 1. Nucleophilic
displacement of various commercially available 6-(chloromethyl)pyrimidines 6 with the anion of 1H-
benzimidazole 515 provides compounds of the general structure 7 in good yields.

Scheme 1: Synthesis of pyridopyrimidinones analogs.
In order to further improve the PKM2 activation potency, the bound conformation of 4 was
compared with other published conformations.13,9 The benzimidazole ring of 4 appears to make favorable
interactions with the neighboring residues of PKM2, therefore, we decided to keep this portion of the
ligand constant in our optimization (figure 4). Opportunities to optimize hydrophobic interactions to the
enzyme around the pyridopyrimidinone ring were identified and modeled. On the basis of docking scores,
various analogs (compounds 8-13, Table 1) were identified and synthesized following the route shown in
scheme 1. In general, small substituents such as methyl, chloro or methoxy groups improved potency.
small alkyl substitutions at C6 (R1) or C9 (R4) of the pyridopyrimidinone ring were especially effective.
Figure 4. Interactions of benzimidazole ring. Crystal structure of compound 4 (magenta) showing the 
interaction between Phenylalanines 26A (pink) and 26B (cyano), imidazole NH interaction to Lys B-311 and
carbonyl interaction to Tyr390 via a water network. Only one binding mode shown.
Due to the presence of interactions between the activators and phenylalanines 26A and 26B (Figure
4), we attempted to improve binding efficiency by altering the electron density of the pyridopyrimidinone
portion of the activator core. Analogs with higher electron density (12, methoxy analog) or lower
electron density (compounds 14-19, thiazolo or thiodiazolo analogs, Table 2) were synthesized. The
thiazolopyrimidinone ring appeared to have an electron density slightly more favorable for the
interaction with the phenyl groups (e.g. compound 18, AC50 = 266 nM, compared to compound 4,
AC50 = 340 nM). Consistent with earlier observations, small substitutions off the thiazolo portion of the
thiazolo pyrimidinone ring seem to increase potency (e.g. 14, AC50 = 86 nM) whereas larger substitutions
decrease potency (e.g. 15, AC50 = 1.03 ). Modeling compounds 12, 15 and 19 into compound 4
crystal structure reveals a potential clash with residue GLN393 thus explaining their lost of potency.

Tables 1. SAR and selected Pyridopyrimidinone analogs
a AC50 values were determined in the lactate dehydrogenase (LDH) enzyme assay (ref.8). b PKM2 asymptote max
represents the % activation at infinite compound concentration. c HLM corresponds to the intrinsic metabolic
d
clearance (Clint) in microsomes (ref.7). RRCK represents the passive permeability from A to B direction of a
transwell assay. e kinetic solubility @ pH = 7.4. f ClogP calculated with ChemBioDraw by Cambridge Soft.
Table 2. SAR of Thiazolo and Thiodiazolo Pyridopyrimidinone analogs

a AC50 values were determined in the lactate dehydrogenase (LDH) enzyme assay (ref.8). b PKM2 asymptote max
represents the % activation at infinite compound concentration. c HLM corresponds to the intrinsic metabolic
d
clearance (Clint) in microsomes (ref.7). RRCK represents the passive permeability from A to B direction of a
transwell assay. e kinetic solubility @ pH = 7.4. f ClogP calculated with ChemBioDraw by Cambridge Soft.
Through two rounds of optimization, both the potency and ligand efficiency were significantly improved
from the original HTS hit (e.g. compound 1, AC50 = 3.47 uM, LE16=0.26, LipE=1.54 vs compound 8
AC50 0.159 uM, LE = 0.43, LipE = 4.7). Since many of the pyridopyrimidinone analogs are ligand
efficient PKM2 activators with low molecular weight and lipophilicity, they generally demonstrated good
ADME property profiles and favorable metabolic stability,17 high permeability, and excellent aqueous
solubility. Although the catalytic site of PKM2 is distinct from that of typical protein kinases,18 we
evaluated off-target liabilities by testing compound 8 across a selected panel of 35 kinases19 at Life
technologies, against a broad panel of pharmacological assays at CEREP,20 as well as in the Ames

mutagenicity21 assay. No positive findings were observed in any of these assays, suggesting compound 8
has no immediate evident liabilities.
Figure 5. Profile comparison of pyridopyrimidinones lead 8 and a sulfonamide analog (20) previously reported to be
a PKM2 activator (ref. 6c).
Compound 20, a lead compound from another known series,6c and compound 8 were used as
small molecule tools to probe the biological effects of PKM2 activation. Both compounds activated
PKM2 in Huh7 cells [8, Cell PK EC50 (Huh7) = 70 nM; 20, Cell PK EC50 (Huh7) = 22 nM, Figure 5].22 As
shown in Figure 6a, both compounds induced formation of PKM2 tetramers (240 kd vs. 60 kd monomer)
in Huh7 cells. However, neither compound promoted any significant change in oxygen consumption
(OCAR change) or lactate production (ECAR change) in Snu-182 (Figure 6b).23 Furthermore, neither
compound effectively suppressed cancer cell growth in Snu182, Snu449, Huh1 or JHH-2 cell lines, even
at 10uM.24

Figure 6. The Effect of PKM2 Activators 8 and 20 on PKM2. (a) Effect on tetramer (240kd band) formation in
Huh7. (b) Effect in bioenergetics of Snu-182 cell line.
In summary, 2-((1H-benzo[d]imidazol-1-yl)methyl)-4H-pyrido[1,2-a]pyrimidin-4-ones are potent and
selective PKM2 activators with a novel binding mode to the PKM2 protein. Both a representative
compound from this series and an activator described in the literature6c were used as molecular tools to
probe the biological effects of PKM2 activation on cancer cells. The two compounds activate PKM2 both
enzymatically and cellularly, and induce tetramer formation, but they do not affect cancer cell glycolysis
(OCAR/ECAR), and are not effective in suppressing cancer cell growth in vitro. While it is a formal
possibility other chemotypes could induce a different tetrameric conformation, one that is functionally
equivalent to the preferential expression of PKM1, our studies suggest that PKM2 activation alone is not
sufficient to alter cancer cell metabolism and reverse the Warburg effect.
Supplementary material
Supplementary material associated with this article can be found in the online version, at
http://dx.doi.org/XXXXXXXX. Material provided includes an illustration of the 4/PKM2 co-crystal
structure showing both ligand binding conformations, as well as synthetic procedures and supporting
spectral data for all numbered compounds and novel reactants.
References and notes:
1 Hanahan, D.; Weinberg, R. A. Cell 2011, 144, 646.

2 Warburg, O. Science 1956, 123, 309.
3 Christofk, H. R.; Vander Heiden, M. G.; Harris, M. H.; Ramanathan, A.; Gerszten, R. E.; Wei, R.; Fleming, M. D.;
Schreiber, S. L.; Cantley, L. C. Nature 2008, 452, 230.
4 (a) ref 3; (b) Mazurek, S.; Boschek, C. B.; Hugo, F.; Eigenbrodt, E. Semin Cancer Biol. 2005, 15, 300. (c)
Dombrauckas, J. D.; Santarsiero, B. D.; Mesecar, A. D. Biochemistry 2005, 44, 9417. (d) Mazurek, S. Int. J.
Biochem. Cell Biol. 2011, 43, 969.
5 (a) Levine, A. J.; Puzio-Kuter, A. M. Science 2010, 330, 1340. (b) Vander Heiden, M. G.; Cantley, L. C.;
Thompson, C. B. Science 2009, 324, 1029. (c) Luo, W.; Hu, H.; Chang, R.; Zhong, J.; Knabel, M.; O'Meally, R.;
Cole, R. N.; Pandey, A.; Semenza, G. L. Cell 2011, 145, 732. (d) Gao, X.; Wang, H.; Yang, J. J.; Liu, X.; Liu, Z. R.
Mol. Cell 2012, 45, 598. (e) Anastasiou, D.; Yu, Y.; Israelsen, W. J.; Jiang, J.-K.; Boxer, M. B.; Hong, B. S.;
Tempel, W.; Dimov, S.; Shen, M.; Jha, A.; Yang, H.; Mattaini, K. R.; Metallo, C. M.; Fiske, B. P.; Courtney, K. D.;
Malstrom, S.; Khan, T. M.; Kung, C.; Skoumbourdis, A. P.; Veith, H.; Southall, N.; Walsh, M. J.; Brimacombe, K.
R.; Leister, W.; Lunt, S. Y.; Johnson, Z. R.; Yen, K. E.; Kunii, K.; Davidson, S. M.; Christofk, H. R.; Austin, C. P.;
Inglese, J.; Harris, M. H.; Asara, J. M.; Stephanopoulos, G.; Salituro, F. G.; Jin, S.; Dang, L.; Auld, D. S.; Park, H.-
W.; Cantley, L. C.; Thomas, C. J.; Vander Heiden, M. G. Nature Chem. Biol. 2012, 8, 839. (f) Cortés-Cros, M.;
Hemmerlin, C.; Ferretti, S.; Zhang, J.; Gounarides, J. S.; Yin, H.; Muller, A.; Haberkorn, A.; Chene, P.; Sellers, W.
R.; Hofmann, F. Proc. Natl. Acad. Sci. U.S.A. 2013, 110, 489. (g) Baas, T. SciBX, 2013, 6(2);
doi:10.1038/scibx.2013.28. Published online Jan. 17, 2013.
6 (a) Boxer, M. B.; Jiang, J. K.; Vander Heiden, M. G.; Shen, M.; Skoumbourdis, A. P.; Southall, N.; Veith, H.;
Leister, W.; Austin, C. P.; Park, H. W.; Inglese, J.; Cantley, L. C.; Auld, D. S.; Thomas, C. J. J. Med. Chem. 2010,
53, 1048. (b) Jiang, J.-K.; Boxer, M. B.; Vander Heiden, M. G.; Shen, M.; Skoumbourdis, A. P.; Southall, N.;
Veith, H.; Leister, W.; Austin, C. P.; Park, H. W.; Inglese, J.; Cantley, L. C.; Auld, D. S.; Thomas, C. J. Bioorg.
Med. Chem. Lett. 2010, 20, 3387. (c) Walsh, M. J.; Brimacombe, K. R.; Veith, H.; Bougie, J. M.; Daniel, T.;
Leister, W.; Cantley, L. C.; Israelsen, W. J.; Vander Heiden, M. G.; Shen, M.; Auld, D. S.; Thomas, C. J.; Boxer, M.
B. Bioorg. Med. Chem. Lett. 2011, 21, 6322.
7 HLM corresponds to the intrinsic metabolic clearance (Clint) in microsomes, in uL/min/mg of microsomal
protein. Data interpretation as following: Clint <15 uL/min/mg (low clearance); Clint 15-40 uL/min/mg (moderate
clearance); Clint >40 (high clearance).
8 The Pyruvate Kinase-catalyzed production of ATP that accompanies phosphoryl transfer from PEP substrate to
ADP is coupled to the oxidation of -NADH through the action of lactate dehydrogenase (LDH). NADH

conversion to NAD+ is monitored by the decrease in absorbance at 340 nm (e = 6.22 mM-1cm-1) using a Molecular
Devices Spectramax Plus plate reader. Typical reaction solutions contain 2nM PK enzyme (human PKM2 (cleaved
N-term his, 1-531), (~2X Km concentration), 2 mM PEP (~Km of PKM2 minus FBP), 0.35 mM NADH, and 4
units/mL LDH in 50 mM Tris buffer, pH 7.5 containing 100mM KCl2 and 5 mM MgCl2. Compound or DMSO is
added to appropriate wells and allowed to incubate for 10 minutes at 32 oC. The reaction is initiated with 0.5mM
ADP. Percent inhibition/activation is determined in duplicate at an appropriate inhibitor concentration. For
IC50/AC50 determinations, compounds are typically run at 10 concentration points (including no inhibitor control, no
enzyme control, and activator control) using a 1:3 serial dilution starting from 100 µM (for original HTS “hits” and
exploratory compounds) or 1:2 serial dilution from 10 µM (for more potent compounds from targeted libraries).
9 Docking methods and examples see following references. (a) Guo, C.; Hou, X.; Dong, L.; Dagostino, E.; Greasley,
S.; Ferre, R.; Marakovits, J.; Johnson, M.C.; Matthews, D.; Mroczkowski, B.; Parge, H.; VanArsdale, T.; Popoff, I.;
Piraino, J.; Margosiak, S.; Thomson, J.; Los, G.; Murray, B.W. Bioorg. Med. Chem. Lett. 2009, 19, 5613. (b) Dong,
L.; Marakovits, J.; Hou, X.; Guo, C.; Greasley, S.; Dagostino, E.; Ferre, R.; Johnson, M.C.; Kraynov, E.; Thomson,
J.; Pathak, V.; Murray, B.W. Bioorg. Med. Chem. Lett. 2010, 20, 2210.
10 Published PKM2 co-crystal structures are found in the Protein Data Bank. The protein shown corresponds to
PDB IDs = 3GQY.
11 a) Ryckmans, T.; Edwards, M.E.; Horne, V.A.; Correia, A.M.; Owen, D.R.; Thompson, L.R., Tran, I.; Tutt, M.F.;
Young, T. Bioorg. Med. Chem. Lett. 2009, 19, 4406. b) Guo, C.; Linton, A.; Kephart, S.; Ornelas, M.; Pairish, M.;
Gonzalez, J.; Greasley, S.; Nagata, A.; Burke, B.J.; Edwards, M.; Hosea, N.; Kang, P.; Hu, W.; Engebretsen, J.;
Briere, D.; Shi, M.; Gukasyan, H.; Richardson, P.; Dack, K.; Underwood, T.; Johnson, P.; Morell, A.; Felstead, R.;
Kuruma, H.; Matsimoto, H.; Zoubeidi, A.; Gleave, M.; Los, G., Fanjul, A.N. J. Med. Chem. 2011, 54, 7693. c) Guo,
C.; Kephart, S.; Ornelas, M.; Gonzales, J.; Linton, A.; Pairish, M.; Nagata, A.; Greasley, S.; Elleraas, J.; Hosea, N.;
Engebretsen, J.; Fanjul, A.N. Bioorg. Med. Chem. Lett. 2012, 22, 1230. d) Guo, C.; Pairish, M.; Linton, A.;
Kephart, S.; Ornelas, M.; Nagata, A.; Burke, B.; Dong, L.; Engbretsen, J.; Fanjul, A.N. Bioorg. Med. Chem. Lett.
2012, 22, 2572. e) Guo, C.; McAlpine, I.; Zhang, J.; Knighton, D.; Kephart, S.; Johnson, M.C.; Li, H., Bouzida, D.;
Yang, A.; Dong, L.; Marakovits, J., Tikhe, J., Richardson, P., Guo, L.; Kania, R.; Edwards, M.; Kraynov, E.;
Christensen, J.; Piraino, J.; Lee, J.; Dagostino, E.; Del-Carmen, C., Deng, Y.; Smeal,T.; Murray, B.W. J. Med.
Chem. 2012, 55, 4728.
12 The PDB code of the structure is 4JPG. The ligand was solved with a 2.33 Å resolution. The ligand is bound in
the A/B dimer activator site, bound in 2 conformations with equal occupancy. The ligand was also found in the C/D
dimer. Compared to apo and other activator-bound structures, there were no significant changes in the conformation
of residues in the activator site.

13 Published PKM2 co-crystal structures are found in the Protein Data Bank. Structures shown corresponds to PDB
IDs = 3GR4, 3H6O and 3GQY.
14 For density map of the two ligand binding conformations see Figure S1 in the Supplementary Material.
15 For experimental procedures, including synthesis of non-commercially available reactants, see Supplementary
Material.
16 Kuntz, I. D.; Chen, K.; Sharp, K. A.; Kollman, P. A. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 9997.
17 The 2-((1H-benzo[d]imidazol-1-yl)methyl)-4H-pyrido[1,2-a]pyrimidin-4-ones core was found to be a substrate of
aldehyde oxydase (AO), which may present challenges in clinical development. The AO assay was described in
following reference. Linton, A.; Kang, P.; Ornelas, M.; Kephart, S.; Hu, Q.; Pairish, M.; Jiang, Y.; Guo, C. J. Med.
Chem. 2011, 54, 7705.
18 PKM2 is one of the Pyruvate Kinases. Pyruvate Kinases do not phosphorylate other proteins, rather a pyruvate
kinase catalyzes the transfer of a phosphate group from phosphoenolpyruvate (PEP) to ADP, generating ATP and
pyruvate. Its substrate is PEP, not ATP (protein kinase substrate). Therefore, the catalytic site of Pyruvate Kinases
(including PKM2) is distinct from that of protein kinases. For example, the PKM2 catalytic site does not contain a
hinge-binding motif, a highly conserved area for classic protein kinases.
19 Selected kinase panel: ABL1, AKT1 (PKB alpha), AURKA (Aurora A), BTK, CAMK2A (CaMKII alpha),
CDK2/cyclin A, CHEK1 (CHK1), CHEK2 (CHK2), CSNK1A1 (CK1 alpha 1), CSNK2A2 (CK2 alpha 2), EGFR
(ErbB1), EPHA2, FGFR1, GSK3B (GSK3 beta), INSR, JAK3, KDR (VEGFR2), LCK, MAP4K4 (HGK), MAPK1
(ERK2), p38, MAPKAPK2, MARK1 (MARK), MET (cMet), MST4 ,MYLK2 (skMLCK), NEK2, NTRK1
(TRKA), PAK4, PDK1 Direct, PIM2, PRKACA, (PKA), PRKCB2 (PKC beta II), ROCK1, SGK (SGK1), SRC,
STK3 (MST2), TAOK2 (TAO1), TEK (Tie2), ZAP70, mTOR, PIK3C2A (PI3K-C2 alpha).
20 CEREP panel: Alpha 1 (non-selective) (antagonist radioligand), Beta 2 (h) (agonist radioligand), D1 (h)
(antagonist radioligand), Dopamine transporter (h) (antagonist radioligand), Muscarinic M1 (h) (antagonist
radioligand), mu opiod (MOP) (h) (agonist radioligand), Ca2+ channel (L, diltiazem site) (benzothiazepines)
(antagonist radioligand), Na+ channel (site 2) (antagonist radioligand), norepinephrine transporter (h) (antagonist
radioligand), 5-HT transporter (h) (antagonist radioligand), Histamine H1 (h) (antagonist radioligand), 5-HT2B (h)
(agonist radioligand), PDE3B (h), Cannabiniod CB1, GABA-BZD (central) rat cerebral cortex.

21 Ames assay conditions were reported previously. Palmer, C.; Pairish, M.; Kephart, S.; Bouzida, D.; Cui, J.; Deal,
J.; Dong, L.; Gu, D.; Linton, A.; McAlpine, I.; Yamazaki, S.; Smith, E.; John-Baptiste, A.; Bagrodia, S.; Kania, R.;
Guo, C. Bioorg. Med. Chem. Lett. 2012, 22, 7605.
22 In this assay, cells from the human HCC cell line, Huh7 were treated with the test compounds, and then lysed
cells were added to PEP and ADP to generate pyruvate and ATP. The generated pyruvate was oxidized by pyruvate
oxidase to produce fluorescence. Since the increase in fluorescence intensity is proportional to the increase in
pyruvate concentration, the pyruvate kinase activity from cells was accurately measured.
23 Compound Effect on Bioenergetic Profile: To determine compounds’ effect on bioenergetic profile, the oxygen
consumption rate (OCR) and extracellular acidification rate (ECAR) of Snu182 cells treated with compounds was
meassured. Snu182 cells were seeded at density of 12k/well with RPMI growth media (Life Technology, #21870)
supplemented with 10%FBS and 2 mM Glutamine in 96-well polyethylene terephthalat (PET) microplates (Seahorse
Bioscience, #101107-001). After overnight culture, 10 μM compounds were added to the media. After 18 hours
culture at 37°C and 5% CO2, the media was changed to XF DMEM media (Seahorse, #100965-000) that was
supplemented 5 mM Glucose and 10 μM compounds, OCR and ECAR readouts were assayed on the Seahorse XF96
Analyzer.
24 Anti-proliferation assay. Cells were seeded in 96-well plates with Optimem-GlutaMax media (Invitrogen). After
overnight culture, 10 μM compound 8 and 20 were added to the cells. Cell viability was measured by Cyto Tox-Glo
kit (Promega) after 5-day compound treatment.