Design, synthesis and biological evaluation of novel pyrrolo[2,3-d]pyrimidine as tumor-targeting agents with selectivity for tumor uptake by high affinity folate receptors over the reduced folate carrier

Article abstract of DOI:10.1016/j.bmc.2020.115544

Tumor-targeted 6-substituted pyrrolo[2,3-d]pyrimidine benzoyl compounds based on 2 were isosterically modified at the 4-carbon bridge by replacing the vicinal (C11) carbon by heteroatoms N (4), O (5) or S (6), or with an N-substituted formyl (7), trifluor

Full text of DOI:10.1016/j.bmc.2020.115544

Bioorganic&MedicinalChemistry28(2020)115544
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Design, synthesis and biological evaluation of novel pyrrolo[2,3-d]
pyrimidine as tumor-targeting agents with selectivity for tumor uptake by
high affinity folate receptors over the reduced folate carrier
Lalit K. Golania, Farhana Islama, Carrie O'Connorb, Aamod S. Dekhneb, Zhanjun Houb,d
,
⁎
⁎
Larry H. Matherlyb,c,d,1, , Aleem Gangjeea,1,
a Division of Medicinal Chemistry, Graduate School of Pharmaceutical Sciences, Duquesne University, 600 Forbes Avenue, Pittsburgh, PA 15282, United States
b Department of Oncology, Wayne State University School of Medicine, Detroit, MI 48201, United States
c Department of Pharmacology, Wayne State University School of Medicine, Detroit, MI 48201, United States
d Molecular Therapeutics Program, Barbara Ann Karmanos Cancer Institute, 421 East Canfield, Detroit, MI 48201, United States
A R T I C L E I N F O
A B S T R A C T
Keywords:
Tumor-targeted 6-substituted pyrrolo[2,3-d]pyrimidine benzoyl compounds based on 2 were isosterically
modified at the 4-carbon bridge by replacing the vicinal (C11) carbon by heteroatoms N (4), O (5) or S (6), or
with an N-substituted formyl (7), trifluoroacetyl (8) or acetyl (9). Replacement with sulfur (6) afforded the most
potent KB tumor cell inhibitor, ~6-fold better than the parent 2. In addition, 6 retained tumor transport se-
lectivity via folate receptor (FR) α and -β over the ubiquitous reduced folate carrier (RFC). FRα-mediated cell
inhibition for 6 was generally equivalent to 2, while the FRβ-mediated activity was improved by 16-fold over 2.
N (4) and O (5) substitutions afforded similar tumor cell inhibitions as 2, with selectivity for FRα and -β over
RFC. The N-substituted analogs 7–9 also preserved transport selectivity for FRα and -β over RFC. For FRα-
expressing CHO cells, potencies were in the order of 8 > 7 > 9. Whereas 8 and 9 showed similar results with
FRβ-expressing CHO cells, 7 was ~16-fold more active than 2. By nucleoside rescue experiments, all the
compounds inhibited de novo purine biosynthesis, likely at the step catalyzed by glycinamide ribonucleotide
formyltransferase. Thus, heteroatom replacements of the CH2 in the bridge of 2 afford analogs with increased
tumor cell inhibition that could provide advantages over 2, as well as tumor transport selectivity over clinically
used antifolates including methotrexate and pemetrexed.
Pyrrolo[2,3-d]pyrimidine
Heteroatom bridge
Folate receptor
Selective uptake
Reduced folate carrier
1. Introduction
poorly across cell membranes. Reflecting this, specific transport systems
are needed to achieve sufficient intracellular concentrations of folates
Reduced folates (e.g. 5-methyltetrahydrofolate) play critical roles as
cofactors in the biosynthesis of purines and pyrimidines.1,2 Mammalian
cells lack the enzymatic machinery to synthesize folates de novo. As a
result, it is necessary to obtain folates from dietary sources. Folate co-
factors are bivalent anions and hydrophilic compounds that diffuse
to support cellular anabolism.3
Three specialized folate transport systems exist in mammalian cells
that include the reduced folate carrier (RFC),4 folate receptors (FRs) α
and β,5 and the proton-coupled folate transporter (PCFT).6–8 RFC and
PCFT are facilitative folate transporters,4,7 whereas FRs are glycosyl
Abbreviations: (AICA), 5-Aminoimidazole-4-carboxamide; (AICARFTase), 5-aminoimidazole-4-carboxamide ribonucleotide formyltransferase; (CHO), Chinese
hamster ovary; (FBS), fetal bovine serum; (FDA), Food and Drug Administration; (FF), folate free; (FR), folate receptor; (10-CHOTHF), N10-formyl tetrahydrofolate;
(GAR), glycinamide ribonucleotide; (GARFTase), glycinamide ribonucleotide formyltransferase; (GPI), glycosyl-phosphatidylinositol; (IC50), 50 percent inhibitory
concentration; (IUPAC), Chemical names follow International Union of Pure and Applied Chemistry; (LCV), leucovorin; (MHz), megahertz; (MTX), methotrexate;
(MEM), minimal essential media; (NMR), nuclear magnetic resonance spectra; (NSCLC), non-small cell lung cancer; (N-CHO), N- formyl; (N-COCF3), N-tri-
fluoroacetyl; (N-COCH3), N-acetyl; (ppm), parts per million; (PMX), pemetrexed; (PCFT), proton-coupled folate transporter; (RTX), Raltitrexed; (RFC), reduced folate
carrier; (RMSD), root-mean-square deviation; (RPMI), Roswell Park Memorial Institute; (TLC), thin layer chromatography; (UV), ultraviolet
⁎ Corresponding authors at: Division of Medicinal Chemistry, Graduate School of Pharmaceutical Sciences, Duquesne University, 600 Forbes Avenue, Pittsburgh, PA
15282, United States (A.Gangjee); Barbara Ann Karmanos Cancer Institute, 421 East Canfield, Detroit, MI 48201, United States (L.H. Matherly).
E-mail addresses: lalit2510@gmail.com (L.K. Golani), matherly@karmanos.org (L.H. Matherly), gangjee@duq.edu (A. Gangjee).
1 To whom correspondence should be addressed.
https://doi.org/10.1016/j.bmc.2020.115544
Received 16 December 2019; Received in revised form 29 April 2020; Accepted 30 April 2020
Availableonline06May2020
0968-0896/©2020ElsevierLtd.Allrightsreserved.

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
phosphatidylinositol-linked proteins that mediate uptake of folates into
cells by receptor-mediated endocytosis.5 RFC is the major route for
membrane transport of circulating folates into mammalian tissues. RFC
is expressed in both normal and cancer cells, whereas FRα, FRβ, and
PCFT exhibit narrower patterns of tissue expression and serve more
specialized physiological roles.4,5,7,8 RFC functions optimally at neutral
pH (~7.4) associated with most normal tissues.4
in adenocarcinomas of the ovary, lung, uterus, breast, cervix, kidney,
and colon, as well as testicular choriocarcinoma, ependymal brain tu-
mors, and nonfunctioning pituitary adenocarcinoma.5,17–19,21,24 For
certain leukemias, such as chronic myelogenous leukemia and acute
myelogenous leukemia, FRβ expression is elevated.25 FRβ is also ex-
pressed in tumor-associated macrophages20, and activated synovial
macrophages involved in the pathogenesis of rheumatoid arthritis26
and other inflammatory conditions including psoriasis and Crohn’s
disease.20,27 Finally, FRβ has been reported in gastric acid cancer and
multiple myeloma, along with other human neoplastic tissues.24,28,29
FR-targeted therapeutics based on FR-targeted antibodies, folate-
conjugates, and antifolates are being pursued.18 A FRα-targeting anti-
body-drug conjugate mirvetuximab soravtansine (IMGN853; Im-
munoGen)30 received “Fast Track” designation by the US Food and
Drug Administration (FDA) and entered Phase III clinical trials. FRα-
targeted strategies have used folate conjugates to deliver toxins, lipo-
somes, and cytotoxic agents to FRα-expressing tumors.15 Clinically
tested therapies include EC1456, a folic acid-tubulysin conjugate.17,31
ONX-0801 (BGC 945), a thymidylate synthase (TS) inhibitor,16 com-
pleted a Phase I clinical trial with plans for a randomized biomarker
pre-specified Phase II clinical trial.32
PCFT is expressed at apical brush border membranes of the duo-
denum and proximal jejunum where it functions at low pH and is the
major intestinal transporter for the absorption of dietary folates.9,10
Although PCFT can be detected in other tissues (e.g., liver, kidney),8 the
requirement for an acidic pH (pH < 7, optimum at pH 5–5.5) for
optimal activity7,10 precludes significant levels of PCFT transport in
most normal tissues.6
For over 60 years, antifolates have served important therapeutic
roles as anticancer, antimicrobial and immunomodulatory agents.11–13
Methotrexate (MTX), a dihydrofolate reductase inhibitor, is a well-
known classical antifolate and continues to be used to treat acute
lymphoblastic leukemia and rheumatoid arthritis.13 Other clinically
used antifolates include pralatrexate, a DHFR inhibitor used for per-
ipheral T-cell lymphoma,12 and the thymidylate synthase inhibitors
raltitrexed (for colorectal cancer) and pemetrexed (PMX) (for non-small
cell lung cancer (NSCLC) and malignant pleural mesothelioma)
(Fig. 1).14 PMX also inhibits secondary targets in one-carbon metabo-
lism, including glycinamide ribonucleotide (GAR) and 5-aminoimida-
zole-4-carboxamide (AICA) formyltransferases (GARFTase and AI-
CARFTase, respectively) in de novo purine biosynthesis, and DHFR.14 As
all of these antifolates are excellent substrates for cellular uptake by the
ubiquitously expressed RFC, they are comparatively nonselective for
tumors versus normal tissues and show dose-limiting toxicities when
used as cancer chemotherapeutic agents, resulting in tumor resistance
and in some cases treatment failures.4,8,15
We discovered 6-substituted 2-amino-4-oxo pyrrolo[2,3-d]pyr-
imidines as inhibitors of tumor cell proliferation with cellular uptake by
FRs and PCFT and inhibition of de novo purine biosynthesis at the first
folate-dependent step catalyzed by GARFTase.33–41 Among the most
active analogs of the previously published series were compounds 1 and
3 (Fig. 2), both three atom bridged compounds, with selective transport
into certain tumor cells by FRs and PCFT (Table 1). Upon internaliza-
tion, compounds 1 and 3 inhibited GARFTase in de novo purine nu-
cleotide biosynthesis, resulting in depletion of purine nucleotides.33,42
However, for compounds 1 and 3, selectivity over RFC was less than
absolute.35,42 The non-selective activity of these 3-atom bridged ana-
logs may reflect their shorter bridge lengths, which are similar to those
for 2-atom bridged classic antifolate drugs such as MTX and PMX which
are good RFC substrates and are poor (and non-selective) substrates for
FRs.35,42
There has been an increased recognition of human FRs as targets for
the specific delivery of new classes of antifolates or folate conjugates to
tumors or sites of inflammation.16–18 In humans, three genes encoding
functional FRs termed FRα, FRβ, and FRγ have been recognized.5,19,20
FRα and FRβ are anchored at the plasma membrane via a glycosyl-
phosphatidylinositol (GPI)- anchor, whereas FRγ is secreted due to the
lack of a signal sequence for GPI anchor attachment.5 FRα is detected
on apical surfaces of several epithelial cells (e.g., kidney) where it is
inaccessible to parenterally administered folates and related com-
pounds.5 FRβ is detected in spleen, bone marrow, and thymus5.
Moreover, in many types of cancer and inflammatory diseases, re-
flecting high expression levels of FRα or FRβ, FR-targeted therapies are
expected to be effective.5,17,18,21,22 FR-targeted therapeutics are pre-
dicted to show modest toxicity in normal tissues because of the re-
stricted expression of both FRα and FRβ, and for FRα, limited access to
the circulation.5,17–19,23 FRα has been reported to be highly expressed
Compound 2 is an analog of 1 but it has a 4-carbon bridge and
showed better selectivity for FRs and PCFT than 1 or 3.35,42 The length
of the side chains in the carbon-bridged analogs was previously iden-
tified as an important structural determinant of potency and transport
selectivity.35,40,42 For analogs with bridge lengths from 2 to 8 carbons,
the most potent tumor inhibitors had 3 carbons (e.g., compound 1)
(Fig. 2). However, this was at the expense of transport selectivity.33,42
For the 4-carbon bridge analogs like 2, there was an improvement in FR
selectivity over RFC compared to 1 without loss of anti-tumor activity;
however, this was accompanied by decreased PCFT activity.33,42
We reported35 that the replacement of the benzylic carbon in 1 (3-
carbon bridge) (Fig. 2) with a heteroatom (N, O or S) provides a series
Fig. 1. Structures of classical antifolate drugs, including methotrexate (MTX), pemetrexed (PMX), pralatrexate (PDX), and raltitrexed (RTX).
2

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
Fig. 2. Design of compounds with optimized chain lengths.
of compounds that were more potent toward FRα-expressing KB human
tumor cells, with substantial selectivity for FRs and PCFT over RFC.35
Analysis of the X-ray crystal structures of the bridge with a N-atom
analog 3 (Fig. 2) with human FRα and GARFTase showed that the
bound conformations required flexibility to attach to both FRα (PDB
5IZQ) and GARFTase (PDB 5J9F).35 The X-ray crystal structure of 3
with GARFTase showed that the bridge heteroatoms of the bound in-
hibitor formed intermolecular interactions with GARFTase, mediated
by an ordered water molecule (PDB 5J9F).35 Analogs of 3 with N-
formyl (N-CHO), N-trifluoroacetyl (N-COCF3), or N-acetyl (N-COCH3)
substitutions were designed to mimic the GARFTase substrate (N10-
formyl tetrahydrofolate; 10-CHO-THF) (Fig. 3) and to afford additional
hydrogen bonding within the GARFTase active site.35 Compound 3, in
addition to tumor cell inhibition, was also efficacious in vivo in mice
bearing IGROV1 ovarian tumor xenografts.35
transport selectivity, and the heteroatom bridged analog 3 that in-
creases potency towards tumor cells. These hybrid analogs were ex-
pected to afford improved antitumor potencies and transporter se-
lectivities, compared to PMX. This is anticipated to overcome toxicity
and potential resistance to PMX. The current work involves the synth-
esis of pyrrolo[2,3–d]pyrimidine analogs with the benzylic CH2 of the
4-carbon bridge of 2 replaced with NH (4), O (5), or S (6) (Fig. 2). We
measured the bond lengths and bond angles in compounds 4–6 and
compared them with literature values.43 As flexibility is crucial for the
binding to both FRα (PDB 5IZQ) and GARFTase (PDB 5J9F),35 it is
notable that the increased bond length of 6 (2.93 Å) compared to 2
(2.55 Å) (Fig. 4) provides additional flexibility in the chain region to
afford both FRα (PDB 5IZQ) selectivity over RFC and GARFTase po-
tency. The bond angle of CeSeC in 6 (109.1°) is decreased compared to
CeCeC (112.8°)44 in 2 (Fig. 4), which is responsible for a different
conformation and orientation of the α and γ-carboxylates of the L-glu-
tamate of 6 compared to 2. Compounds 4 and 5 have comparable bond
lengths and bond angles to 2 and are predicted to mimic compound 2
Thus, it was of interest to synthesize 4-atom bridged compounds
with heteroatoms (N, O, S, and N-CHO, N-COCF3, N-COCH3) (Figs. 2
and 3) as hybrids of the 4-carbon bridged analog 2 that improves
Table 1
IC50s (in nM) for 6-substituted pyrrrolo[2,3–d]pyrimidine benzoyl antifolates with heteroatom replacements, and classical antifolates in RFC-, PCFT- and FR-
expressing cell lines. Proliferation assays were performed for CHO sublines engineered to express human RFC (PC43-10), FRα (RT16), FRβ (D4) or PCFT (R2/PCFT4),
and transporter-null (R2) CHO cells, and KB human tumor cells (express RFC, FRα, and PCFT). For the experiments measuring FR-mediated effects, assays were
performed in the presence or absence of 200 nM folic acid (results are shown only for KB cells). Results are presented as IC50 values, corresponding to the
concentrations that inhibit growth by 50% relative to cells incubated without the drug. The data are mean values from 5 to 16 experiments ( standard errors in
parentheses). Some of the data for 3, MTX, and PMX have been previously published.35,51 Results are also summarized for KB cells for the protective effects of
adenosine (60 µM), thymidine (10 µM), glycine (130 μM) or 5-aminoimidazole-4-carboxamide (320 µM). For compounds 2, 4, 5, 6, 7, 8, and 9, folic acid and
nucleoside/AICA protection results are shown in Fig. 7. Methods are summarized in the Experimental Section. Undefined abbreviations: Ade, adenosine; AICA, 5-
aminoimidazole-4-carboxamide; FA, folic acid; gly, glycine; ND, not determined; and Thd, thymidine.
Compounds
R2
RFC
FRα
FRβ
PCFT
RFC/FRα/PCFT
PC43-10
RT16
D4
R2/PCFT4
KB
KB (+FA)
KB + Ade/Thd/AICA/gly
1
2
> 1000
> 1000
> 1000
697(1 5 8)
> 1000
> 1000
> 1000
> 1000
> 1000
216(8.7)
894(93)
649(38)
> 1000
510(90)
670(1 5 0)
> 1000
> 1000
> 1000
> 1000
> 1000
12(1.1)
4.1(1.6)
6.3(1.6)
5.6(1.2)
10(2)
23.0(3.3)
213(28)
1.7(0.4)
1.9(0.7)
> 1000
> 1000
666(46)
> 1000
> 1000
757(41)
> 1000
> 1000
> 1000
20(2.4)
Ade/AICA
Ade/AICA
Ade/AICA
Ade/AICA
Ade/AICA
Ade/AICA
Ade/AICA
Ade/AICA
Ade/AICA
Thd/Ade
3
3.04 (0.71)
1.76(0.43)
4.6(1.3)
0.62(0.20)
1.51(0.43)
5.6(1.4)
87.4(9.9)
235(96)
0.32(0.05)
0.95(0.35)
2.50(0.64)
0.30(0.04)
13.20 (3.32)
5.35 (0.49)
40.11 (9.34)
6.00(0.60)
68(12)
4
5
> 1000
6
2.56(0.73)
35.63 (4.78)
5.78(1.45)
57.24 (6.9)
114(31)
0.60(0.12)
0.61 (0.11)
3.87(0.54)
12.79 (1.90)
106(11)
593(2 0 4)
893(1 6 1)
487(1 7 1)
> 1000
7
8
9
MTX
PMX
121(17)
138(13)
42(9)
60 (8)
13.2(2.4)
327(1 0 3)
Thd/Ade
3

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
Fig. 3. Structure of N10-formyl tetrahydrofolate and proposed mimics.
(Fig. 4). The purpose of this study is to determine the effects of het-
eroatom replacements of the vicinal (C11) carbon to the side chain
phenyl ring in compound 2 by N, O, or S, and N-CHO, N-COCF3, or N-
COCH3 moieties (7–9) on transport specificities by FRα, FRβ, and PCFT,
over RFC, and on the inhibition of de novo purine nucleotide bio-
synthesis on cell proliferation and antitumor activity.
nitrogen and the backbone amide eNHe of Leu899 and the 2-NH2 and
the backbone carbonyls of Glu948 and Leu899. Other interactions in-
clude N3 and the backbone carbonyl of Ala947, and the 4-oxo and the
backbone eNH- of Ala947 and Val950. The N11-H of compound 4 is
hydrogen-bonded via a water-mediated hydrogen-bond with the sub-
strate GAR (Fig. 5A). Similar interactions are predicted to occur in-
volving the heteroatom bridge of 5 (oxygen) (figure not shown). The
flexible four-atom bridges of 4 and 6 orient the side chain phenyl ring
into a hydrophobic cleft comprised of Phe895 (not labeled), Met896,
Leu899, and Ile898. The interactions involving the flexible glutamate
side chains are similar to those reported for the glutamate side chain of
3.35 The α-carboxylic acid of 4 forms salt bridges with Arg871 and
Arg897, and hydrogen bonds with the backbone of Ile898. The γ-car-
boxylic acid forms a salt bridge with Lys844 and Arg897 (Fig. 5A).
Compounds 2 and 4 had docked scores of −14.02 and −14.29 kcal/
mol, respectively, in GARFTase, similar to that for 3 (-14.36 kcal/mol).
Docking studies of 6 suggested an altered orientation of the phenyl ring
and side-chain L-glutamate compared to 4. This is attributed to the
larger size of the sulfur atom, resulting in a different set of interactions
between the active site with the
2. Molecular modeling
To determine the ligand-protein interactions, docking studies were
carried out with compounds 4–9 using X-ray crystal structures of
human GARFTase (PDB 5J9F), FRα (PDB 5IZQ) and FRβ (PDB
4KN2).23,35,37,45 Docking scores are provided in supporting information
(Table 2S).
2.1. Docking studies of compounds 4–9 with human GARFTase (PDB:
5J9F) (Figs. 5A and 5B):
Docking results of compounds 2, 4, and 6 with human GARFTase
(PDB 5J9F) are displayed in Figs. 5A and 5B. The docked poses of 2, 4
and 6 indicate that the pyrrolo[2,3–d]pyrimidine scaffold is buried in
the active site and occupies the same location as the bound ligand 335
(Figs. 5A and 5B; bound ligand 3 not shown for clarity). The bicyclic
pyrrolo[2,3–d]pyrimidine scaffold of 4 and 6 is stabilized by the hy-
drogen bonding interactions, including hydrogen bonds between the N1
carboxylic acid forms a salt bridgeLw-giltuhtaAmrga8te97s,idwehcilheatihneoγf-c6a.rbTohxeylαic-
acid of 6 forms a salt bridge with Arg871 (Fig. 5B). The resulting
docking score of 6 is −14.53 kcal/mol, somewhat better than 2–4,
indicating an increased inhibition of GARFTase. The N11-CHO, COCF3,
and COCH3 moieties in 7, 8, and 9, respectively, are oriented similarly
Fig. 4. Distance and bond angle variations predicted by the nature of the bridge at the benzylic position (X). Distances and angles for X = NH, O, and S were
measured using energy-minimized conformations of compounds with Maestro Schrödinger 2019-1.44
4

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
Fig. 5A. Molecular modeling studies with human
GARFTase (PDB 5J9F).33 Superimposition of the
docked pose of 2 (green), 4 (tan) and GAR (cyan).
Docking scores of
2 and 4 are −14.02 and
−14.29 kcal/mol, respectively. Hydrogen bond in-
teractions, pi-pi interactions, and salt bridges are
depicted as black, cyan and magenta dashed lines,
respectively.
to the N11-H of 4. Compounds 7–9 bind deeper into the formyl transfer
region that is occupied by the formyl group of the natural GARFTase
substrate, N10-formyl tetrahydrofolate (10-CHOTHF) (figure not
shown). The docked scores of 5 and 7–9 (-14.38, −14.79, −14.87, and
PMX, respectively (Figs. 6A, 6B, 6C; 3 and PMX are not shown for
clarity). Figs. 6A and 6B display the docked poses of compounds 2, 4
and 6 in the X-ray crystal structures of human FRα. Molecular modeling
revealed that the pyrrolo[2,3–d]pyrimidine scaffold of compounds 4
and 6 are stacked between the side chains of Tyr85 and Trp171, similar
to the bicyclic scaffold of bound ligand 3 (Figs. 6A and 6B; 3 not shown
for clarity).35 The 2-NH2 and 7-NH of compounds 4 and 6 form hy-
drogen bonds with Asp81. Additional hydrogen bonds are formed be-
tween N3 and Ser174, and between the 4-oxo group and side chains of
Arg103 and Arg106. The α-carboxylic acids of 4 and 6 form hydrogen
bonds with the backbone NH of Gly137 and the pyrrole NH of Trp140,
while the γ-carboxylic acid forms a salt bridge with Lys136. In com-
parison, with the 3-atom linker of 3 (not shown for clarity), the four
atom linkers of 4 and 6 position the phenyl ring in the hydrophobic
−14.12 kcal/mol, respectively), are equivalent to that for
4
(-14.29 cal/mol), which suggest that these compounds should be potent
GARFTase inhibitors.
2.2. Docking studies of compound 4–9 with FRα (PDB 5IZQ) and FRβ (PDB
4KN2) (Figs. 6A-6C):
The docked poses of 2, 4 and 6 with the crystal structures of 3 in
FRα (PDB 5IZQ)35,45 and PMX in FRβ (PDB 4KN2)23 indicated that all
three compounds are accommodated within a similar space as 3 and
Fig. 5B. Molecular modeling studies with human
GARFTase (PDB 5J9F).33 Superimposition of the
docked pose of 2 (green), 6 (pink) and GAR (cyan).
Docking scores of
2 and 6 are −14.02 and
−14.53 kcal/mol, respectively. Hydrogen bond in-
teractions, pi-pi interactions, and salt bridges are
depicted as black, cyan, and magenta dashed lines,
respectively.
5

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
Fig. 6A. Molecular modeling studies using the
human FRα (PDB 5IZQ)35,46 crystal structures. Su-
perimposition of the docked pose of 2 (green) with
the docked pose of 4 (tan) in FRα. Docking scores of
2 and 4 are −11.72 and −12.02 kcal/mol, respec-
tively. Hydrogen bonding interactions, pi-pi inter-
actions, and salt bridges are depicted as black cyan
and magenta dashed lines, respectively.
Fig. 6B. Molecular modeling studies using the
human FRα (PDB 5IZQ)35,46 crystal structures. Su-
perimposition of the docked pose of 2 (green) with
the docked pose of 6 (pink) in FRα. Docking scores
of 2 and 6 are −11.72 and −12.07 kcal/mol, re-
spectively. Hydrogen bondinteractions, pi-pi inter-
actions, and salt bridges are depicted as black, cyan
and magenta dashed lines, respectively.
pocket (Tyr60, Trp102, His135, and Trp138) to make hydrophobic in-
teractions (Figs. 6A and 6B). The results revealed that the expanded
chain length in the 4-atom chain compound (compared to 3) is vital for
positioning the phenyl ring to the hydrophobic pocket for increased
FRα selectivity.
similar binding space as the corresponding L-glutamate of the native
ligand.35 The α-carboxylate forms a salt bridge with Arg152 and hy-
drogen bonds with the pyrrole NH of Trp118, while the γ-carboxylic
acid of 2 and 6 forms hydrogen bonds with the pyrrole NH of Trp156
and a water-mediated hydrogen bond with the backbone of Trp154.
The flexible four atom bridge with the sulfur atom of 2 and 6 helps to
orient the side chain phenyl ring into a hydrophobic cleft of amino acids
Phe78, Trp118, and Trp154. The extended chain length (4-atom chain)
provides FRβ selectivity, mediated by these hydrophobic interactions.
These interactions are less apparent in the 3-atom chain compound 3
(not shown for clarity).
Figs. 6C shows the docked pose of compounds 2 and 6 in the X-ray
crystal structure of human FRβ (PDB 4KN2).23 In this pose, the or-
ientation of the scaffold permits the 2-NH2 and 7-NH moieties of both 2
and 6 to form hydrogen bonds with Asp97, and the 4-oxo moiety to
form hydrogen bonds with the side-chain nitrogen hydrogens of
Arg119, His151 and Tyr101. Hydrogen bonding interaction with
Ser190 and N3 of 2 and 6 were also observed. The pyrrolo[2,3–d]
pyrimidine scaffold is stacked amid the hydrophobic aromatic side
chains of Tyr101 and Trp187. The L-glutamate moiety occupies a
6

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
Fig. 6C. Molecular modeling studies using the
human FRβ (PDB 4KN2)23 crystal structures. Su-
perimposition of the docked pose of 2 (green) with
the docked pose of 6 (pink) in FRβ. Docking scores
of 2 and 6 are −13.92 and −14.84 kcal/mol, re-
spectively. Hydrogen bonding interactions, pi-pi
interactions, and salt bridges are depicted as black,
cyan and magenta dashed lines, respectively.
3. Chemistry
4. Biological evaluation and discussion
The reaction of β-propiolactone 10 (Scheme 1) and ethyl p-amino-
As a part of the study of structure-activity relationships for tumor-
targeted antitumor compounds with specificities for the individual
transporters, we measured the impact of isosteric heteroatom replace-
ments, including N, O, and S (compounds 4, 5, and 6, respectively), of
CH2 at position-11 (compound 2) (Fig. 2). Additional N-substituted
analogs related to 4 with formyl (7), trifluoroacetyl (8), and acetyl (9)
substitutions (Fig. 3), as a mimic of the natural substrate, 10-CHOTHF
cofactor for GARFTase, were also tested.
benzoate 11 afforded 12.46 The carboxylic acid 12 was then treated
with trifluoroacetic anhydride to give the protected amine 13.35 Com-
pound 13 was converted to the acid chloride and immediately reacted
with diazomethane, followed by treatment with silver acetate to pro-
vide the homologated carboxylic acid 14. Carboxylic acid 14 was then
converted to the acid chloride and immediately reacted with diazo-
methane, followed by 48% HBr in water to afford the α-bromoketone
15. Condensation of 2,6-diamino-3H-pyrimidin-4-one with 15 in DMF
at room temperature for 3 days provided the 6-substituted-2-amino-4-
oxo-pyrrolo[2,3–d]pyrimidine 16. Hydrolysis of 16 resulted in the
corresponding free acid 17. Subsequent coupling with L-glutamate di-
methyl ester using 2-chloro-4,6-dimethoxy-1,3,5-triazine as the acti-
vating agent provided the diester 18. Saponification of the diester
yielded the target 4. Compound 4 was then reacted with formic acid,
trifluoroacetic anhydride or acetic anhydride to afford target com-
pounds 7, 8 and 9, respectively.
To screen for antiproliferative activities, we used a panel of isogenic
Chinese hamster ovary (CHO) sublines engineered from RFC-, FR- and
PCFT-null MTXRIIOuaR2-4 CHO cells48 (R2) cells to individually ex-
press FRα (RT16 cells), FRβ (D4), RFC (PC43-10), or PCFT (R2/
PCFT4).33,42,49,50 The CHO cells were cultured with a range of drug
concentrations for up to 96 h and cell viabilities were measured to
calculate IC50 values, corresponding to the concentrations that inhibit
growth by 50%.42 Results are summarized in Table 1 and are compared
to the inhibition of FRα-expressing KB human tumor cells.
The carboxylic acids 22 and 23 (Scheme 2) were obtained via nu-
cleophilic substitution of the alkyl bromide 19 with 20 or 21, respec-
tively, followed by deprotection of the tert-butyl ester with tri-
fluoroacetic acid.47 Compounds 22 and 23 were then converted to the
acid chloride and immediately reacted with diazomethane, followed by
48% HBr in water to provide the corresponding α-bromomethylketones
28 and 29, respectively.35 Condensation of 2,6-diamino-3H-pyrimidin-
4-one with 28 and 29 in DMF at room temperature for 3 days afforded
the 2-amino-4-oxo-6-substituted-pyrrolo[2,3–d]pyrimidines 30 and 31,
respectively. Hydrolysis of 30 and 31 afforded the corresponding free
acids 32 and 33, respectively. Subsequent coupling with
The 6-substituted pyrrolo[2,3–d]pyrimidine analogs with het-
eroatom bridge replacements, 4 (N), 5 (O), and 6 (S), all showed potent
growth inhibition toward FRα-expressing RT16 cells and toward FRβ-
expressing D4 cells with IC50 values generally less than those previously
reported for 2 (Table 1).42 Compounds 4 (N) and 6 (S) displayed 3- and
2-fold better activities respectively than the lead analog 2 in FRα-ex-
pressing RT16 cells. The impact of N-substitutions on compound 4 in-
cluding 7 (N-CHO), 8 (N-COCF3) and 9 (N-COCH3) were disparate.
Compound 8 showed comparable potency to 2 with RT16 cells, whereas
7 and 9 were much less inhibitory. The N-substituted compounds 7 (N-
CHO), 8 (N-COCF3), and 9 (N-COCH3) showed activity in the rank order
of 8 > 7 > 9 for FRα (RT16). Compounds 4, 5 and 6, were highly
active (6-, 2-, and 16-fold, respectively) toward D4 cells (expresses FRβ)
compared to compound 2. For the N-substituted compounds with D4
ester using 2-chloro-4,6-dimethoxy-1,3,5-triazine as theL-galcutitvaamtiantge
agent provided the diesters 34 and 35. Saponification of the diesters
yielded the target compounds 5 and 6.
7

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
Scheme 1. . Reagents and conditions; (a)
acetone, 4 h, reflux, 52%; (b) (CF3CO)2O,
CH2Cl2, 2 h, rt, 88%; (c) oxalyl chloride,
CH2Cl2, reflux,
1 h; (d) diazomethane,
(CH3CH2)2O, rt, 1 h; (e) AgCOOCH3, water,
1–4 dioxane, 1.5 h, reflux, 38%; (f) oxalyl
chloride, CH2Cl2, reflux,
1 h; (g) diazo-
methane, (CH3CH2)2O, rt, 1 h; (h) 48% HBr
in water, reflux, 1 h; (i) 2,6-diaminopyr-
imidin-4(3H)-one, DMF, rt, 3 d, 34%; (j) 1 N
NaOH, rt, 12 h, 88%;(k) N-methylmorpho-
line 2-chloro-4, 6-methoxy-1,3,5-triarzine, L-
glutamate dimethyl ester, N-methyl mor-
pholine, DMF, 12 h, 43%; (l) 1 N NaOH, rt,
4 h, 88%; (m) HCOOH, (CH3CO)2O, reflux,
2 h, 42%; (n) (CF3CO)2O, 12 h, rt, 68%; (o)
(CH3CO)2O, 12 h, rt, 74%.
cells, activity was comparable to (9) or exceeded (~16-fold for 7; ~3-
fold for 8) that for 2.
Adenosine completely reversed the drug effects of all the com-
pounds, whereas thymidine and glycine were ineffective alone and they
did not augment the effects of adenosine in combination (not shown).
Thus, de novo purine biosynthesis rather than thymidylate synthase or
mitochondrial C1 metabolism is the targeted pathway for compounds
4–9. AICA (320 µM) has been used to identify GARFTase inhibitors as
this is metabolized to AICA ribonucleotide (ZMP) to circumvent the
GARFTase step.35,42 As AICA was completely protective for compounds
4–9, we conclude that GARFTase is the primary cellular target for this
series.
Relative inhibition by the heteroatom compounds toward KB human
tumor cells generally paralleled that in FRα-expressing RT16 cells.
Compounds
4 (N) and 6 (S) exhibited sub-nanomolar potencies
(IC50 = 0.95 and 0.30 nM, respectively) in KB tumor cells which were
~2- and ~6-fold, respectively, greater than for the lead compound 2.
From the results with the engineered CHO sublines, compounds 4–9
were all more selective toward FRs compared to PCFT and RFC. Anti-
proliferative activities toward PCFT- (R2/PCFT4) and RFC- (PC43-10)
expressing cells ranged from modest to undetectable and were similar
to those for compound 2. These results contrast with those for the
classic antifolates (MTX, PMX) which inhibit cells regardless of the
expressed transporter (Table 1).
5. Summary
In this report, we identified novel heteroatom substituted pyrrolo
[2,3–d]pyrimidine compounds with N, O, or S replacements for CH2 at
position-11. Our results document remarkably increased in vitro anti-
proliferative activity resulting from heteroatom bridge substitutions
toward CHO cell lines engineered to express human FRα or FRβ, and
FRα-expressing human tumor cells, over the corresponding CH2 analog
2. All these compounds have substantial selectivities toward FRα and
FRβ compared to RFC. PCFT-targeted activity for this series was
modest. All compounds inhibited de novo purine biosynthesis, most
We previously reported that compound 2 was an inhibitor of de novo
purine biosynthesis at the reaction catalyzed by GARFTase.42 We
evaluated inhibition of de novo purine biosynthesis by compounds 4–9
by comparing the extent of growth of KB cells in the presence of ade-
nosine (60 µM) to that with thymidine (10 µM).35,42 We also evaluated
the protective effects of glycine (130 μM) to assess the potential for
inhibition of C1 metabolism in mitochondria.51 Results for the nu-
cleoside/glycine protection experiments are shown in Fig. 7.
8

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
Scheme 2. . Reagent and conditions; (a) Cs2CO3, tetrabutyl ammonium chloride, DMF, 1 h, 100 °C, 47–65%; (b) CF3COOH, CH2Cl2, 84–86%; (c) oxalyl chloride,
CH2Cl2, reflux, 1 h, 79%; (d) diazomethane, (CH3CH2)2O, rt, 1 h; (e) 48% HBr in water, 80 °C, 2 h, 86–94% (crude yield); (f) 2,6-diaminopyrimidin-4(3H)-one, DMF,
rt, 3 d, 31–34%; (g) 1 N NaOH, rt, 12 h, 66–88%; (h) N-methylmorpholine 2-chloro-4, 6-methoxy-1,3,5-triarzine, L-glutamate diethyl ester, DMF, 12 h, 43–45%; (i)
1 N NaOH, rt, 4 h, 85–88%.
likely at GARFTase. Compounds 6 and 7 displayed ~16-fold increased
activity over the CH2 analog (2) and selectivity for FRβ compared to
FRα, along with 6-fold increased activity against KB tumor cells for 6
over 2. In addition to tumor targeting, these results suggest potential
applications for targeting FRβ expressing hematopoietic cancers and
activated macrophages.5,20 As such, dual therapeutic targeting FRα-
expressing tumors (e.g., epithelial ovarian cancer), along with FRβ-
expressing tumor-associated macrophages, would be highly impactful.
Collectively, these compounds provide an exciting platform for con-
tinued discovery of new and highly selective FR-targeted therapeutics
for cancer, as well as other diseases.
NMR spectra. Chemicals and solvents were purchased from Aldrich
Chemical Co. or Fisher Scientific Co. and were used as received.
Elemental analysis (C, H, N, F, and S) was performed by Atlantic
Microlab, Inc. (Norcross, GA). Element compositions were within 0.4%
of the calculated values and confirmed > 95% purity for all the com-
pounds submitted for biological evaluation.
6.1. Synthesis
3-((4-(ethoxycarbonyl)phenyl)amino)propanoic acid (12).
A
mixture of ethyl p-aminobenzoate (11) (1.65 g, 10 mmol) and β-pro-
piolactone (10) (0.72 g, 10 mmol) in 15 mL of acetone was refluxed for
4 h. After evaporation of the solvent under reduced pressure, MeOH
(10 mL) was added, followed by silica gel (5 g). The resulting plug was
loaded on to a silica gel column and eluted with ethyl acetate: hexane
1:10. Fractions with Rf = 0.34 (hexane: EtOAc 1:1) were pooled and
evaporated to afford 12 as yellow semisolid (1.23 g, yield; 52%). TLC
Rf = 0.34 (1:1 hexane:EtOAc); 1H NMR (CDCl3) δ 1.26–1.29 (t, 3H,
OCH2CH3, J = 7 Hz), 2.51–2. 2.53 (t, 2H, CH2CH2NH, J = 6.8 Hz),
3.30–3.32 (m, 2H, CH2CH2NH), 4.18–4.24 (q, 2H, OCH2CH3, J = 7 Hz),
6.59–6.61 (d, 2H, Ar-CH, J = 7.2 Hz), 7.68–7.70 (d, 2H, Ar-CH,
J = 7.2 Hz).
6. Experimental section
A rotary evaporator was used to carry out evaporation in vacuo.
Final compounds and intermediates were dried in a CHEM-DRY drying
apparatus over P2O5 at 80 °C. A MEL-TEMP II melting point with a
FLUKE 51 K/J electronic thermometer apparatus was used and un-
corrected to record melting points. A Bruker WH-400 [400-megahertz
(MHz)] spectrometer or a Bruker WH-500 (500 MHz) spectrometer was
used to record proton nuclear magnetic resonance spectra (1H NMR).
Tetramethylsilane was used as an internal standard to express the
chemical shift in ppm (parts per million): s, singlet; d, doublet; t, triplet;
q, quartet; quin, quintet m, multiplet; and bs, broad singlet. Chemical
names follow International Union of Pure and Applied Chemistry
(IUPAC) nomenclature. Whatman Sil G/UV254 silica gel plates with a
fluorescent indicator were used for performing thin-layer chromato-
graphy (TLC), and the spots were visualized under 254 and 365 nm
illumination. All analytical samples were homogeneous on TLC in three
different solvent systems. Solvents used for TLC were measured in vo-
lume. Columns of silica gel (230–400 mesh) (Fisher, Somerville, NJ)
were used for chromatography. In spite of 24–48 h of drying in vacuo,
fractional moles of water found in the analytical samples of antifolates
3-(N-(4-(ethoxycarbonyl)phenyl)-2,2,2-trifluoroacetamido)
propanoic acid (13). A solution of compound 12 (1.18 g, 5 mmol) in
anhydrous CH2Cl2 (20 mL) was treated with triethylamine (1 mL) and
trifluoroacetic anhydride (5 mL) under the anhydrous condition for 2 h.
After removal of the excess of solvent under reduced pressure at
40–45 °C, the yellowish oily residue was dissolved in dichloromethane
(20 mL) followed by silica gel (3 g). The resulting plug was loaded on to
a silica gel column and eluted with 10% ethyl acetate in hexane.
Fractions with and Rf = 0.59 (1:1 hexane: EtOAc) were pooled and
evaporated to afford 13 (1.47 g, yield; 88%) as yellow semisolid. TLC
Rf = 0.59 (1:1 hexane:EtOAc); 1H NMR (CDCl3) δ 1.40–1.44 (t, 3H,
OCH2CH3, J = 7.2 Hz), 2.68–2.72 (t, 2H, CH2CH2NCOCF3, J = 7.2 Hz),
could not be prevented and were confirmed by their presence in the 1
H
9

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
Fig. 7. Identification of the intracellular target by protection by nucleosides, glycine and AICA. KB cells were incubated with drugs in folate-, nucleoside-, and
glycine-free RPMI 1640 medium with 10% dialyzed fetal bovine serum (FBS), antibiotics, L-glutamine, and 2 nM leucovorin (LCV) with a range of drug con-
centrations in the presence of folic acid (200 nM), adenosine (60 μM), thymidine (10 μM), glycine (130 μM) or AICA (320 μM). Cell proliferation was assayed with
Cell Titer BlueTM (Promega) using a fluorescence plate reader42. Data are mean values for at least triplicate experiments. Error bars represent standard errors.
Experimental details are summarized in the Experimental Section.
4.07–4.10 (t, 2H, CH2CH2NCOCF3, J = 7.2 Hz), 4.39–4.44 (q, 2H,
OCH2CH3, J = 7.2 Hz), 7.33–7.35 (d, 2H, Ar-CH, J = 8.4 Hz),
8.13–8.15 (d, 2H, CH2, Ar-CH, J = 8.4 Hz). Anal. calcd (C14H14F3NO5):
C, 50.46; H, 4.23; N, 4.20; F, 17.10. Found: C, 50.50; H, 4.22; N, 4.19; F,
16.87.
3.82–3.86 (t, 2H, CH2CH2CH2NCOCF3, J = 7.2 Hz), 4.40–4.45 (q, 2H,
OCH2CH3, J = 7.2 Hz), 7.31–7.34 (d, 2H, Ar-CH, J = 8.4 Hz),
8.14–8.16 (d, 2H, CH2, Ar-CH, J = 8.4 Hz). Anal. calcd (C15H16NF3O5):
C, 51.88; H, 4.64; N, 4.03; F, 16.41. Found: C, 51.82; H, 4.83; N, 3.99; F,
16.26.
4-(N-(4-(ethoxycarbonyl)phenyl)-2,2,2-trifluoroacetamido)bu-
tanoic acid (14). Compound 13 (1.33 g, 4 mmol) was dissolved in
20 mL anhydrous dichloromethane and oxalyl chloride (2 mL,
23.30 mmol) was added. The resulting solution was refluxed for 1 h and
then cooled to room temperature. After evaporating the solvent under
reduced pressure, the residue was dissolved in 20 mL of diethyl ether.
The resulting solution was added dropwise to ice-cooled diazomethane
(generated in situ from 10 g of Diazald® by using Aldrich mini Diazald®
Apparatus) in an ice bath over 10 min. The resulting mixture was al-
lowed to stand for 30 min and then stirred for an additional 1 h, excess
diazomethane was decomposed with l mL of acetic acid, and the solu-
tion evaporated to dryness to afford diazo compound. To a stirred
suspension of silver acetate (1.2 g) in 30 mL of water was added a
solution of the diazo compound in 30 mL of 1,4-dioxane. The reaction
mixture was heated to reflux for 1.5 h, and then 1.2 g of sodium car-
bonate was added. The solid obtained was filtered and extracted with
chloroform (3 × 25 mL). The combined chloroform extract was dried
(Na2SO4), filtered, and the solvent was evaporated under reduced
pressure. The residue was recrystallized from ethyl acetate to afford 14
(527 mg, yield over three steps 38%) as yellow semi-solid. TLC
Rf = 0.64 (1:1 hexane:EtOAc); 1H NMR (CDCl3) δ 1.41–1.44 (t, 3H,
OCH2CH3, J = 7.2 Hz), 1.89–1.96 (quin, 2H, CH2CH2CH2NCOCF3,
J = 7.2 Hz), 2.42–2.46 (t, 2H, CH2CH2CH2NCOCF3, J = 7.2 Hz),
4-((3-(2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3–d]pyrimidin-
6-yl)propyl)amino) benzoic acid (17). To 14 (2.4 g, 10 mmol) in a
250 mL flask was added oxalyl chloride (5.14 mL, 60 mmol) and an-
hydrous CH2Cl2 (20 mL). The resulting solution was refluxed for 1 h
and then cooled to room temperature. After the solvent was evaporated
under reduced pressure, the residue was dissolved in 20 mL of diethyl
ether. The resulting solution was added dropwise to ice-cooled diazo-
methane (generated in situ from 15 g of Diazald® by using Aldrich Mini
Diazald® apparatus) in an ice bath over 10 min. The resulting mixture
was allowed to stand for 30 min and then stirred for an additional 1 h.
To this solution was added 48% HBr in water (20 mL). The resulting
mixture was refluxed for 1.5 h. After the mixture was cooled to room
temperature, the organic layer was separated, and the aqueous layer
was extracted with Et2O (2 × 200 mL). The combined organic layer and
Et2O extract was washed with two portions of 10% Na2CO3 solution and
dried over Na2SO4. Evaporation of the solvent afforded 15 in 94% yield.
To a suspension of 2,6-diaminopyrimidin-4-one (1.26 g, 10 mmol) in
anhydrous DMF (25 mL) was added 15 (9.4 mmol). The resulting
mixture was stirred under N2 at room temperature for 3 days. After
evaporation of the solvent under reduced pressure, MeOH (20 mL) was
added followed by silica gel (5 g). The resulting plug was loaded on to a
silica gel column and eluted with CHCl3 followed by 3% MeOH in
CHCl3 and then 5% MeOH in CHCl3. Fractions with and Rf = 0.58
10

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
(TLC) (CHCl3:CH3OH, 5:1) were pooled and evaporated to afford 16
(1.03 g, yield; 25%). Compound 16 (0.7 mmol) was dissolved in MeOH
(10 mL) added 1 N NaOH (10 mL) and the mixture was stirred under N2
at room temperature for 10 h. TLC showed the disappearance of the
starting material and one major spot at the origin (CHCl3: CH3OH, 5:1).
The reaction mixture was dissolved in water (10 mL), the resulting
solution was cooled in an ice bath, and the pH was adjusted to 3–4 with
the dropwise addition of 1 N HCl. The resulting suspension was frozen
in the dry ice-acetone bath, thawed to 4–5 °C in the refrigerator, and
filtered. The residue was washed with a small amount of cold water and
dried in vacuum using P2O5 to afford the target compound 17 (196 mg,
yield 86%) as yellow solid. TLC Rf = 0.19 (5:1 CHCl3:MeOH); mp
154 °C, 1H NMR (DMSO‑d6) δ 1.81–1.88 (quin, 2H, CH2CH2CH2NH,
J = 7.2 Hz), 2.56–2.59 (t, CH2CH2CH2NH, CH2, J = 7.2 Hz), 3.05–3.09
(m, 2H, CH2CH2CH2NH), 5.91 (s, 1H, C5-CH), 6.18 (bs, 2H, 2-NH2,
exch.), 6.54–6.56 (bd, 3H, Ar-CH and NH, J = 8.4 Hz, one proton
exch), 7.64–7.66 (d, 2H, CH2, Ar-CH, J = 8.4 Hz), 10.34 (s, 1H, 3-NH,
exch.), 10.90 (s, 1H, 7-NH, exch.). Anal. calcd (C16H17N5O3·0.4 HCl): C,
56.20; H, 5.13; N, 20.48. Found: C, 56.39; H, 5.27; N, 20.41.
CH2CH2CH2NCHO, J = 7.2 Hz), 3.84–3.88 (t, 2H, CH2CH2CH2NCHO,
J = 7.5 Hz) 4.38–4.43 (m, 2H, α-CH), 5.85 (s, 1H, C5-CH), 5.97 (s, 2H,
2-NH2, exch.), 7.45–7.78 (d, 1.72H, Ar-CH, J = 8.4 Hz, rotamer of
formamide), 7.51–7.53 (d, 0.28H, Ar-CH, J = 8.4 Hz, rotamer of for-
mamide), 7.91–7.95 (d, 2H, Ar-CH, J = 8.4 Hz), 8.64–8.66 (d, 1H, Ar-
CONH, J = 7.6 Hz, exch.), 10.23 (bs, H, 3-NH, exch.), 10.93 (s, H, 7-
NH, exch.). Anal. calcd (C22H24N6O7·1.5H2O2.5 HCl): C, 45.25; H, 5.25;
N, 13.47. Found: C, 45.56; H, 4.89; N, 13.40.
(4-(N-(3-(2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyr-
imidin-6-yl)propyl)-2,2,2-trifluoroacetamido)benzoyl)-L-glutamic
acid (8). 5 mL trifluoroacetic anhydride was added to a solution of 4
(110 mg, 0.25 mmol) in 10 mL dichloromethane, and the reaction
mixture was stirred at 25 °C for 3 h. The solvent was removed under
reduced pressure and the residue was suspended in water. 1 N NaOH
was added dropwise to the suspension and pH was adjusted to 4. The
suspension was stored at 0 °C for 2 h. The yellow solid was collected by
filtration and dried over P2O5 to give 83 mg (yield 76%) of 8. TLC
Rf = 0.21 (5:1 CHCl3:MeOH); mp 158 °C; 1H NMR (DMSO‑d6) δ
1.87–2.11 (m, 4H, CH2CH2CH2NCOCF3 and β -CH2), 2.33–2.37 (t, 2H,
γ-CH2, J = 7.5 Hz), 2.62–2.66 (t, 2H, CH2CH2CH2NCOCF3, J = 7.2 Hz),
3.97–3.99 (t, 2H, CH2CH2CH2NCOCF3, J = 7.2 Hz), 4.38–4.44 (m, 1H,
α-CH), 5.92 (s, 1H, C5-CH), 5.97 (bs, 2H, 2-NH2, exch.), 7.40–7.42 (d,
2H, Ar-CH, J = 8 Hz), 7.94–7.96 (d, 2H, Ar-CH, J = 8 Hz), 8.76–8.78
(d, 1H, Ar-CONH, J = 8 Hz, exch.), 10.39 (bs, H, 3-NH, exch.), 10.99 (s,
H, 7-NH, exch.). Anal. calcd (C23H13N6O7F3·0.5H2O): C, 49.20; H, 4.31;
N, 14.97; F, 10.15. Found: C, 49.16; H, 4.52; N, 15.05; F, 9.82.
(4-(N-(3-(2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyr-
(4-((3-(2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3–d]pyr-
imidin-6-yl)propyl)amino) benzoyl)-L-glutamic acid (4). To
a
250 mL round bottom flask, was added a mixture of compound 17
(100 mg, 0.2 mmol), N-methylmorpholine (0.4 mmol), 2-chloro-4,6-
dimethoxy-1,3,5-triazine (0.4 mmol) and anhydrous DMF (7 mL). The
resulting mixture was stirred at room temperature under the anhydrous
condition for 1.5 h. N-mehtylmorpholine (0.64 mmol) and L-glutamate
diethyl hydrochloride (0.3 mmol) were added in the reaction mixture.
The resulting mixture was then stirred at room temperature under the
anhydrous condition for 12 h. After evaporation of the solvent under
reduced pressure, MeOH (20 mL) was added followed by silica gel (1 g).
The resulting plug was loaded on to a silica gel column and eluted with
CHCl3 followed by 3% MeOH in CHCl3 and then with 5% MeOH in
CHCl3. Fractions with Rf = 0.45 (CHCl3:CH3OH, 5:1) were pooled and
evaporated to afford 18 (96 mg, yield 61%) as solid. Compound 18
(0.15 mmol) was dissolved in MeOH (10 mL) added 1 N NaOH (10 mL)
and the mixture was stirred under N2 at room temperature for 10 h. TLC
showed the disappearance of the starting material and one major spot at
the origin (CHCl3: CH3OH, 5:1). The reaction mixture was dissolved in
water (10 mL), the resulting solution was cooled in an ice bath, and the
pH was adjusted to 3–4 with dropwise addition of 1 N HCl and acetic
acid. The resulting suspension was frozen in the dry ice-acetone bath,
thawed to 4–5 °C in the refrigerator, and filtered. The residue was
washed with a small amount of cold water and dried in vacuum using
P2O5 to afford the target compound 4 (58 mg, yield 84%) as a yellow
powder. TLC Rf = 0.16 (5:1 CHCl3:MeOH); mp 163 °C; 1H NMR
(DMSO‑d6) δ 1.92–1.95 (quin, 2H, CH2CH2CH2NH, J = 7.2 Hz),
1.97–2.09 (m, 2H, β -CH2), 2.32–2.35 (t, 2H, γ-CH2, J = 7.5 Hz),
2.58–2.61 (t, 2H, CH2CH2CH2NH, J = 7.2 Hz), 3.31–3.34 (m, 2H,
CH2CH2CH2NH), 4.34–4.37 (m, 1H, α-CH), 5.91–5.91 (d, 1H, Ar-CH,
imidin-6-yl)propyl)acetamido)benzoyl)-L-glutamic
acid
(9).
Compound 4 (110 mg, 0.25 mmol) was added in 5 mL acetic anhydride,
and the reaction mixture was stirred under the anhydrous condition at
25 °C for 3 h. The excess of acetic anhydride was removed under re-
duced pressure. The residue was suspended in cold water and basified
using 1 N NaOH. The suspension was then filtered and acidified to pH 4
with 0.5 N HCI and stored at 0 °C for 2 h. The white solid was collected
by filtration and dried over P2O5 to give 57 mg (yield 47%) of 9. TLC
Rf = 0.19 (5:1 CHCl3:MeOH); mp 166 °C; 1H NMR (DMSO‑d6) δ
1.93–2.14 (m, 4H, CH2CH2CH2NCOCH3,
β
-CH2), 2.05 (s, 3H,
CH2CH2CH2NCOCH3), 2.35–2.39 (t, 2H, γ-CH2, J = 7.5 Hz), 2.64–2.67
(t, 2H, CH2CH2CH2NCOCH3,
J = 8 Hz), 3.89–3.92 (t, 2H,
CH2CH2CH2NCOCH3, J = 8 Hz), 4.38–4.44 (m, 1H, α-CH), 5.91 (s, 1H,
C5-CH), 5.99 (s, 2H, 2-NH2, exch.), 7.34–7.36 (d, 2H, Ar-CH, J = 8 Hz),
7.93–7.95 (d, 2H, Ar-CH, J = 8 Hz), 8.68–8.70 (d, 1H, Ar-CONH,
J = 8 Hz, exch.), 10.16 (bs, H, 3-NH, exch.), 10.83 (s, H, 7-NH, exch.),
12.45 (bs, 2H, 2COOH, exch.). Anal. calcd (C23H26N6O7·0.5
H2O·1.6HCl): C, 49.27; H, 5.21; N, 14.67. Found: C, 48.99; H, 4.99; N,
15.37.
4-(4-(ethoxycarbonyl)phenoxy)butanoic acid (22). To 100 mL
rbf was added a mixture of compound 20 (1.66 g, 10 mmol), cesium
carbonate (3.26 g, 10 mmol), TBAI (3.70 g, 10 mmol) and anhydrous
DMF (20 mL). Compound 19 (2.23 g, 10 mmol) was added dropwise to
the mixture. The reaction mixture was then stirred at room temperature
for 3 h. Ethyl acetate was added into the reaction mixture. The com-
bined mixture was washed with two portions of water. After evapora-
tion of the solvent under reduced pressure MeOH (20 mL) was added
followed by silica gel (1 g). The resulting plug was loaded on to a silica
gel column and eluted with 1:10 (ethyl acetate: hexane). Fractions with
and Rf = 0.64 (hexane:ethyl acetate 1:1) were pooled and evaporated
to afford tert-butyl esters (1.47 g, yield; 47%). Trifluoroacetic acid was
then added into the tert-butyl esters and mixture was stirred at room
temperature for 30 min. Excess of trifluoroacetic acid was evaporated
and MeOH (20 mL) was added followed by silica gel (1 g). The resulting
plug was loaded on to a silica gel column and eluted with 1:10 (ethyl
acetate: hexane). Fractions with and Rf = 0.45 (TLC) (Hexane: ethy-
lacetate, 1:1) were pooled and evaporated to afford 22 (1 g, yield 83%)
as white solid. TLC Rf = 0.45 (1:1 hexane:EtOAc); mp 112 °C; 1H NMR
(CDCl3) δ 2.13–2.20 (quin, 2H, CH2CH2CH2O, J = 7.2 Hz), 2.60–2.64
J
= 2 Hz), 5.97 (bs, 2H, 2-NH2, exch.), 6.22–6.24 (t, 1H,
CH2CH2CH2NH,
J = 5.5 Hz, exch.), 6.55–6.56 (d, 2H, Ar-CH,
J = 9 Hz), 7.65–7.67 (d, 2H, Ar-CH, J = 9 Hz), 8.09–8.10 (d, 1H, NH,
J = 7.5 Hz, exch.), 10.14 (s, 1H, 3-NH, exch.), 10.85 (s, 1H, 7-NH,
exch.). Anal. calcd (C21H24N6O6·0.5H2O0.5CH3COOH): C, 53.33; H,
5.49; N, 16.96. Found: C, 53.27; H, 5.59; N, 16.61.
(4-(N-(3-(2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyr-
imidin-6- yl)propyl)formamido) benzoyl)-L-glutamic acid (7). To a
solution of 4 (110 mg, 0.25 mmol) in 97% formic acid (5 mL) was
added acetic anhydride (1 mL), and the reaction mixture was stirred at
25 °C for 3 h. The solvent was removed under reduced pressure and the
residue dissolved in 1 N NaOH at 0 °C. The filtrate was acidified to pH 4
with 0.5 N HCI and stored at 0 °C for 2 h. The yellow solid was collected
by filtration and dried over P2O5 to give 40 mg (yield 35%) of 7. TLC
Rf = 0.19 (5:1 CHCl3:MeOH); mp 153 °C; 1H NMR (DMSO‑d6) δ
1.92–1.95 (quin, 2H, CH2CH2CH2NCHO, J = 7.2 Hz), 1.97–2.11 (m,
2H, β -CH2), 2.35–2.38 (t, 2H, γ-CH2, J = 7.5 Hz), 2.44–2.47 (t, 2H,
11

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
(t, 2H, CH2CH2CH2O, J = 7.2 Hz), 3.90 (s, 3H, OCH3), 4.08–4.11 (t,
2H, CH2CH2CH2O, J = 7.2 Hz), 6.91–6.93 (d, 2H, Ar-CH, J = 8.8 Hz),
7.99–8.01 (d, 2H, Ar-CH, J = 8.8 Hz).
(0.48 mmol) and anhydrous DMF (10 mL). The resulting mixture was
stirred at room temperature under the anhydrous condition for 1.5 h. N-
mehtylmorpholine (0.48 mmol) and L-glutamate di-tert-butyl hydro-
chloride (0.36 mmol) were added in reaction mixture. The resulting
mixture was then stirred at room temperature under the anhydrous
condition for 12 h. After evaporation of the solvent under reduced
pressure, MeOH (20 mL) was added followed by silica gel (1 g). The
resulting plug was loaded on to a silica gel column and eluted with
CHCl3 followed by 1% MeOH in CHCl3. Fractions with Rf = 0.52
(CHCl3:CH3OH, 5:1) were pooled and evaporated to afford 34 (56 mg,
yield 45%) as solid. Compound 34 (56 mg, 0.11 mmol) was dissolved in
MeOH (10 mL) added 1 N NaOH (10 mL) and the mixture was stirred
under N2 at room temperature for 10 h. TLC showed the disappearance
of the starting material and one major spot at the origin (CHCl3: MeOH
5:1). The reaction mixture was dissolved in water (10 mL), the resulting
solution was cooled in an ice bath, and the pH was adjusted to 3–4 with
the dropwise addition of 1 N HCl. The resulting suspension was frozen
in the dry ice-acetone bath, thawed to 4–5 °C in the refrigerator, and
filtered. The residue was washed with a small amount of cold water and
dried in vacuum using P2O5 to afford the target compound (5) (34 mg,
yield 68%) as white powder. TLC Rf = 0.12 (5:1 CHCl3:MeOH); mp
173 °C; 1H NMR (DMSO‑d6) δ 1.91–2.09 (m, 4H, CH2CH2CH2O and β
-CH2), 2.33–2.37 (t, 2H, γ-CH2, J = 7.6 Hz), 2.45–2.68 (t, 2H,
4-((4-(ethoxycarbonyl)phenyl)thio)butanoic
acid
(23).
Carboxylic acid 23 was synthesized from compound 21 using the syn-
thetic procedure as for carboxylic acid 22 in 74% yield as white solid
over two-step. TLC Rf = 0.52 (1:1 hexane:EtOAc); mp 76 °C; 1H NMR
(CDCl3) δ 1.39–1.43 (t, 2H, OCH2CH3, J = 7.2 Hz), 2.03–2.06 (quin,
2H, CH2CH2CH2S, J = 7.2 Hz), 2.53–2.60 (t, 2H, CH2CH2CH2O,
J = 7.2 Hz), 3.07–3.10 (t, 2H, CH2CH2CH2S, J = 7.2 Hz), 4.36–4.41 (q,
2H, OCH2CH3, J = 7.2 Hz), 7.33–7.35 (d, 2H, Ar-CH, J = 8.4 Hz),
7.95–7.97 (d, 2H, CH2, Ar-CH, J = 8.4 Hz).
4-(3-(2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyrimidin-
6-yl)propoxy)benzoic acid (32). To 22 (2.52 g, 10 mmol) in a 250 mL
flask was added oxalyl chloride (5.14 g, 60 mmol) and anhydrous
CH2Cl2 (20 mL). The resulting solution was refluxed for 1 h and then
cooled to room temperature. After the solvent was evaporated under
reduced pressure, the residue was dissolved in 20 mL of Et2O. The re-
sulting solution was added dropwise to an ice-cooled diazomethane
(generated in situ from 15 g of Diazald® by using Aldrich Mini Diazald®
apparatus) in an ice bath over 10 min. The resulting mixture was al-
lowed to stand for 30 min and then stirred for an additional 1 h. To this
solution was added 48% HBr (20 mL). The resulting mixture was re-
fluxed for 1.5 h. After the mixture was cooled to room temperature, the
organic layer was separated, and the aqueous layer was extracted with
Et2O (2 × 200 mL). The combined organic layer and Et2O extract was
washed with two portions of 10% Na2CO3 solution and dried over
Na2SO4. Evaporation of the solvent afforded 28 in 94% crude yield. To
a suspension of 2,6-diaminopyrimidin-4-one (1.26 g, 10 mmol) in an-
hydrous DMF (25 mL) was added 28 (9.4 mmol). The resulting mixture
was stirred under N2 at room temperature for 3 days. After evaporation
of the solvent under reduced pressure, MeOH (20 mL) was added fol-
lowed by silica gel (5 g). The resulting plug was loaded on to a silica gel
column and eluted with CHCl3 followed by 3% MeOH in CHCl3 and
then 5% MeOH in CHCl3. Fractions with and Rf = 0.48 (TLC) (CHCl3:
CH3OH, 5:1) were pooled and evaporated to afford 30 (1 g, yield 31%).
Compound 30 (1 g, 2.92 mmol) was dissolved in MeOH (10 mL) added
1 N NaOH (10 mL) and the mixture was stirred under N2 at room
temperature for 10 h. TLC showed the disappearance of the starting
material and one major spot at the origin (CHCl3: MeOH 5:1). The re-
action mixture was dissolved in water (10 mL), the resulting solution
was cooled in an ice bath, and the pH was adjusted to 3–4 with the
dropwise addition of 1 N HCl. The resulting suspension was frozen in
the dry ice-acetone bath, thawed to 4–5 °C in the refrigerator, and fil-
tered. The residue was washed with a small amount of cold water and
dried in vacuum using P2O5 to afford the compound 32 (638 mg, yield
66%) as white solid. TLC Rf = 0.15 (5:1 CHCl3:MeOH); mp 155 °C; 1H
NMR (DMSO‑d6) δ 2.02–2.07 (quin, 2H, CH2CH2CH2O, J = 7.2 Hz),
2.63–2.69 (t, 2H, CH2CH2CH2O, J = 7.2 Hz), 4.03–4.06 (t, 2H,
CH2CH2CH2O, J = 7.2 Hz), 5.95 (s, 1H, C5-CH), 6.60 (bs, 2H, 2-NH2,
exch.), 6.99–7.02 (d, 2H, Ar-CH, J = 8.8 Hz), 7.86–7.89 (d, 2H, Ar-CH,
J = 8.8 Hz), 10.67 (s, 1H, 3-NH, exch.), 11.18 (s, 1H, 7-NH, exch.).
4-((3-(2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyrimidin-
6-yl)propyl)thio)benzoic acid (33). Carboxylic acid 33 was synthe-
sized from compound 23 using the synthetic procedure as for carboxylic
acid 32 from compound 22 in 22% yield as white solid over five steps.
TLC Rf = 0.22 (1:1 hexane:EtOAc); mp 143 °C; 1H NMR (DMSO‑d6) δ
1.89–1.96 (quin, 2H, CH2CH2CH2S, J = 7.2 Hz), 2.64–2.67 (t, 2H,
CH2CH2CH2O,
J = 7.6 Hz), 4.03–4.06 (t, 2H, CH2CH2CH2O,
J = 7.6 Hz), 4.35–4.40 (m, 1H, α-CH), 5.90 (s, 1H, C5-CH), 5.99 (bs,
2H, 2-NH2, exch.), 6.99–7.02 (d, 2H, CH2, Ar-CH, J = 8.8 Hz),
7.84–7.87 (d, 2H, Ar-CH, J = 8.8 Hz), 8.44–8.46 (d, 1H, Ar-CONH,
J = 7.5 Hz, exch.), 10.16 (s, 1H, 3-NH, exch.), 10.89 (s, 1H, 7-NH,
exch.). Anal. calcd (C21H23N5O7 · 2 H2O): C, 51.11; H, 5.51; N, 14.19:
Found: C, 50.94; H, 5.45; N, 14.00.
(4-((3-(2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyr-
imidin-6-yl)propyl)thio)benzoyl)-L-glutamic acid (6). Target com-
pound 6 was synthesized from compound 33 using the synthetic pro-
cedure as for target compound 5 from compound 32 in 38% yield as
buff color solid over two steps. TLC Rf = 0.14 (5:1 CHCl3:MeOH); mp
171 °C; 1H NMR (DMSO‑d6) δ 1.95–2.13 (m, 4H, CH2CH2CH2S and β
-CH2), 2.33–2.37 (t, 2H, γ-CH2, J = 7.2 Hz), 2.62–2.66 (t, 2H,
CH2CH2CH2S,
J = 7.2 Hz), 3.02–3.06 (t, 2H, CH2CH2CH2S,
J = 7.2 Hz), 4.38–4.39 (m, 1H, α-CH), 5.89 (s, 1H, C5-CH), 5.99 (bs,
2H, 2-NH2, exch.), 7.35–7.37 (d, 2H, Ar-CH, J = 8.4 Hz), 7.80–7.82 (d,
2H, Ar-CH, J = 8.4 Hz), 8.57–8.59 (d, 1H, Ar-CONH, J = 7.5 Hz,
exch.), 10.16 (s, 1H, 3-NH, exch.), 10.88 (s, 1H, 7-NH, exch.). Anal.
calcd (C21H23N5O6S·0.5 H2O): C, 52.27; H, 5.01; N, 14.51; S, 6.65.
Found: C, 52.08; H, 4.86; N, 14.51; S, 6.47.
6.2. Molecular modeling and computational studies
All the compounds were docked on to the X-ray crystal structures of
human FRα (PDB 5IZQ, 3.60 Å)35,37,45, FRβ (PDB 4KN2, 2.6 Å)23, and
GARFTase (PDB 5J9F, 2.1 Å)35,37 to analyze the potential binding
modes, binding energies, and favorable or unfavorable interactions. The
energy-minimized crystallized ligand 335 and PMX were redocked with
an RMSD (root-mean-square deviation) of 0.89 and 0.91 for the best-
scored pose, thus validating the docking process. The crystal structure
PDBs were obtained from the protein database. All docking procedures
were performed using various modules of Schrödinger Maestro suite
(Schrödinger, LLC, New York, NY, 2019).44 The polypeptide structures
of FRα, FRβ and GARFTase were optimized and prepared for docking
using the Maestro Protein Preparation Wizard to assess bond order and
missing hydrogens, followed by energy minimization using the OPLS3e
force field. Gaps in the protein structures were not corrected as they
were far from the active site. The Maestro induced-fit Grid Generation
module was then used to define a 15 × 15 × 15 Å grid from the center
of all the ligands. Ligands used in the computational docking study
were built using the Maestro 2D Build module. The Maestro LigPrep
module was then used to generate conformers of each compound
CH2CH2CH2S,
J = 7.2 Hz), 3.03–3.07 (t, 2H, CH2CH2CH2S,
J = 7.2 Hz), 5.95 (s, 1H, C5-CH), 6.02 (bs, 2H, 2-NH2, exch.),
7.34–7.36 (d, 2H, Ar-CH, J = 8.8 Hz), 7.82–7.84 (d, 2H, Ar-CH,
J = 8.8 Hz), 10.47 (s, 1H, 3-NH, exch.), 11.06 (s, 1H, 7-NH, exch.).
(4-(3-(2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyrimidin-
6-yl)propoxy)benzoyl)-L-glutamic acid (5). To a 250 mL rbf, was
added a mixture of compound 32 (80 mg, 0.24 mmol), N-methylmor-
pholine
(0.48
mmol),
2-chloro-4,6-dimethoxy-1,3,5-triazine
12

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
subjected to energy minimization using the OPLS3e force field protocol.
The resulting compounds were docked into the prepared FRα, FRβ, and
GARFTase structures using the Maestro Induced Fit Docking. Induced
Fit Docking was performed with standard precision with flexible ligand
sampling. A total of 20 initial poses were generated for each compound.
Based on the pose score, the top 4 poses were selected and subjected to
energy minimization using the OPLS3e force field. Finally, the top 2
poses per compound were generated and ranked according to Glide
score, which is an approximation of binding energy defined by receptor-
ligand complex energies. The top pose was analyzed and presented in
the Biological Evaluation and Discussion. Docking scores are listed in
Table 2S (supporting information).
Institutes of Health R01 CA53535 (LHM and ZH), R01 CA125153 (AG),
R01 CA152316 (LHM and AG), and R01 CA166711 (AG and LHM), the
Eunice and Milton Ring Endowed Chair for Cancer Research (LHM),
and the Duquesne University Adrian Van Kaam Chair in Scholarly
Excellence (AG). AD was supported by T32 CA009531 (LHM) and F30
CA228221 (AD).
Appendix A. Supplementary material
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmc.2020.115544.
References
6.3. Cell lines and assays of antitumor drug activities
1. Stokstad ELR. Historial perspective on key advances in the biochemistry and phy-
siology of folates. In: Picciano MF, Stokstad ELR, Gregory JF, eds. Contemporary Issues
in Clinical Nutrition. New York: Wiley-Liss; 1990:1–21. Folic Acid Metabolism in
Heath and Disease.
The engineered CHO sublines including RFC-, PCFT-, and FRα-null
MTXRIIOuaR2-4 (R2) and RFC- (PC43-10), PCFT- (R2/PCFT4), FRα-
(RT16), and FRβ- (D4) expressing CHO sublines were previously de-
scribed.33,34,42,48–50 The CHO cells were grown in α-minimal essential
medium (MEM) supplemented with 10% bovine calf serum (Invitrogen,
Carlsbad, CA), L-glutamine (2 mM), penicillin (1000 U/mL), and
streptomycin (1000 μg/mL) at 37 °C with 5% CO2. R2 transfected cells
(PC43-10, RT16, D4, R2/PCFT4) were cultured in complete α-MEM
media plus G418 (1 mg/mL). Prior to the cell proliferation assays (see
below), RT16 and D4 cells were cultured in complete folate-free RPMI
1640 (without added folate), plus 10% dialyzed fetal bovine serum
(FBS) (Sigma-Aldrich) and penicillin/streptomycin for 3 days. Human
KB carcinoma cells were purchased from ATCC (Manassas, VA) con-
tinuously maintained in complete folate-free (FF) RPMI with 10% fetal
bovine serum and 1% penicillin-streptomycin and L-glutamine.
For growth inhibition studies, cells (CHO, KB) were plated in 96
well dishes (~2000 cells/well, total volume of 200 μL) and treated with
a range of drug concentrations (0–1000 nM) in complete folate-free
RPMI 1640 medium with 10% dialyzed FBS, supplemented with 2 nM
(KB, RT16, D4 CHO cells) or 25 nM (PC43-10, R2/PCFT4) leucovorin,
as described.33–35,38–42,50 To confirm FR-mediated drug uptake, 200 nM
folic acid was added to parallel incubations for KB, RT16 and D4 cells.
After 96 h, viable cells were assayed with Cell-Titer BlueTM reagent
(Promega, Madison, WI), and fluorescence was measured with a
fluorescence plate reader. Fluorescence measurements were used for
calculations of IC50 values, corresponding to the drug concentrations at
which cells showed 50% loss of proliferation.
2. Shuvalov O, Petukhov A, Daks A, Fedorova O, Vasileva E, Barlev NA. One-carbon
metabolism and nucleotide biosynthesis as attractive targets for anticancer therapy.
Oncotarget. 2017;8:23955–23977.
3. Hou Z, Matherly LH. Biology of the major facilitative folate transporters SLC19A1
and SLC46A1. Curr Top Membr. 2014;73:175–204.
4. Matherly LH, Hou Z, Deng Y. Human reduced folate carrier: translation of basic
biology to cancer etiology and therapy. Cancer Metastasis Rev. 2007;26:111–128.
5. Elnakat H, Ratnam M. Distribution, functionality and gene regulation of folate re-
ceptor isoforms: implications in targeted therapy. Adv Drug Deliv Rev.
2004;56:1067–1084.
6. Matherly LH, Hou Z, Gangjee A. The promise and challenges of exploiting the proton-
coupled folate transporter for selective therapeutic targeting of cancer. Cancer
Chemother Pharmacol. 2018;81:1–15.
7. Zhao R, Goldman ID. The molecular identity and characterization of a Proton-cou-
pled Folate Transporter–PCFT; biological ramifications and impact on the activity of
pemetrexed. Cancer Metastasis Rev. 2007;26:129–139.
8. Desmoulin SK, Hou Z, Gangjee A, Matherly LH. The human proton-coupled folate
transporter: Biology and therapeutic applications to cancer. Cancer Biol Ther.
2012;13:1355–1373.
9. Qiu A, Jansen M, Sakaris A, et al. Identification of an intestinal folate transporter and
the molecular basis for hereditary folate malabsorption. Cell. 2006;127:917–928.
10. Inoue K, Nakai Y, Ueda S, et al. Functional characterization of PCFT/HCP1 as the
molecular entity of the carrier-mediated intestinal folate transport system in the rat
model. Am J Physiol-Gastrointestinal LiverPhysiol. 2008;294:G660–G668.
11. Chabner BA, Allegra CJ. Cancer chemotherapy and biotherapy principles and practice:
Antifolates. 5 ed. Philadelphia; Baltimore; New York: Wolters Kluwer: Lippincott
Williams & Wilkins; 2011:109–138.
12. Visentin M, Zhao R, Goldman ID. The antifolates. Hematol Oncol Clin North Am.
2012;26:629–648 ix.
13. Wessels JA, Huizinga TW, Guchelaar HJ. Recent insights in the pharmacological
actions of methotrexate in the treatment of rheumatoid arthritis. Rheumatology
(Oxford). 2008;47:249–255.
14. Chattopadhyay S, Moran RG, Goldman ID. Pemetrexed: biochemical and cellular
pharmacology, mechanisms, and clinical applications. Mol Cancer Ther.
2007;6:404–417.
To confirm the targeted pathway or enzyme, in vitro growth in-
hibition of KB tumor cells was measured with complete folate- and
glycine-free RPMI 1640 supplemented with 10% dialyzed fetal bovine
serum in the presence of thymidine (10 μM), adenosine (60 μM) and/or
glycine (130 μM).33–35,38–42,50 For de novo purine biosynthesis in-
hibitors, additional protection experiments used AICA (320 μM) to
distinguish inhibitory effects at GARFTase from those at AI-
CARFTase.33–35,38–42,50
15. Gonen N, Assaraf YG. Antifolates in cancer therapy: structure, activity and me-
chanisms of drug resistance. Drug Resist Updat. 2012;15:183–210.
16. Gibbs DD, Theti DS, Wood N, et al. BGC 945, a novel tumor-selective thymidylate
synthase inhibitor targeted to alpha-folate receptor-overexpressing tumors. Cancer
Res. 2005;65:11721–11728.
17. Vergote IB, Marth C, Coleman RL. Role of the folate receptor in ovarian cancer
treatment: evidence, mechanism, and clinical implications. Cancer Metastasis Rev.
2015;34:41–52.
18. Xia W, Low PS. Folate-targeted therapies for cancer. J Med Chem.
2010;53:6811–6824.
19. Parker N, Turk MJ, Westrick E, Lewis JD, Low PS, Leamon CP. Folate receptor ex-
pression in carcinomas and normal tissues determined by a quantitative radioligand
binding assay. Anal Biochem. 2005;338:284–293.
7. Statistical analysis
Descriptive statistical tests (e.g., t-tests) were conducted using
GraphPad 6.0 software (La Jolla, CA).
20. Puig-Kroger A, Sierra-Filardi E, Dominguez-Soto A, et al. Folate receptor beta is
expressed by tumor-associated macrophages and constitutes a marker for M2 anti-
inflammatory/regulatory macrophages. Cancer Res. 2009;69:9395–9403.
21. Shen J, Hu Y, Putt KS, et al. Assessment of folate receptor alpha and beta expression
in selection of lung and pancreatic cancer patients for receptor targeted therapies.
Oncotarget. 2018;9:4485–4495.
Declaration of Competing Interest
22. Weitman SD, Weinberg AG, Coney LR, Zurawski VR, Jennings DS, Kamen BA.
Cellular Localization of the Folate Receptor: Potential Role in Drug Toxicity and
Folate Homeostasis. Cancer Res. 1992;52:6708–6711.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
23. Wibowo AS, Singh M, Reeder KM, et al. Structures of human folate receptors reveal
biological trafficking states and diversity in folate and antifolate recognition. Proc
Natl Acad Sci. 2013;110:15180–15188.
24. Shen J, Putt KS, Visscher DW, et al. Assessment of folate receptor-beta expression in
human neoplastic tissues. Oncotarget. 2015;6:14700–14709.
Acknowledgments
25. Ross JF, Wang H, Behm FG, et al. Folate receptor type beta is a neutrophilic lineage
This work was supported in part by grants from the National
13

L.K. Golani, et al.
Bioorganic&MedicinalChemistry28(2020)115544
marker and is differentially expressed in myeloid leukemia. Cancer.
purine biosynthesis with selectivity for cellular uptake by high affinity folate re-
ceptors and the proton-coupled folate transporter over the reduced folate carrier. J
Med Chem. 2012;55:1758–1770.
1999;85:348–357.
26. Hu Y, Wang B, Shen J, et al. Depletion of activated macrophages with a folate re-
ceptor-beta-specific antibody improves symptoms in mouse models of rheumatoid
arthritis. Arthritis Res Therapy. 2019;21:143.
39. Wang L, Cherian C, Desmoulin SK, et al. Synthesis and antitumor activity of a novel
series of 6-substituted pyrrolo[2,3-d]pyrimidine thienoyl antifolate inhibitors of
purine biosynthesis with selectivity for high affinity folate receptors and the proton-
coupled folate transporter over the reduced folate carrier for cellular entry. J Med
Chem. 2010;53:1306–1318.
27. Nakashima-Matsushita N, Homma T, Yu S, et al. Selective expression of folate re-
ceptor beta and its possible role in methotrexate transport in synovial macrophages
from patients with rheumatoid arthritis. Arthritis Rheum. 1999;42:1609–1616.
28. Wu D, Zhang P, Ma J, et al. Serum biomarker panels for the diagnosis of gastric
cancer. Cancer Med. 2019;8:1576–1583.
40. Wang L, Wallace A, Raghavan S, et al. 6-Substituted pyrrolo[2,3-d]pyrimidine
thienoyl regioisomers as targeted antifolates for folate receptor alpha and the proton-
coupled folate transporter in human tumors. J Med Chem. 2015;58:6938–6959.
41. Wang Y, Cherian C, Orr S, et al. Tumor-targeting with novel non-benzoyl 6-sub-
stituted straight chain pyrrolo[2,3-d]pyrimidine antifolates via cellular uptake by
folate receptor alpha and inhibition of de novo purine nucleotide biosynthesis. J Med
Chem. 2013;56:8684–8695.
29. Zhou Y, Unno K, Hyjek E, et al. Expression of functional folate receptors in multiple
myeloma. Leukemia lymphoma. 2018;59:2982–2989.
30. Kurkjian C, LoRusso P, Sankhala KK, et al. A phase I, first-in-human studyto evaluate
the safety, pharmacokinetics (PK), and pharmacodynamics (PD) ofIMGN853 in pa-
tients (Pts) with epithelial ovarian cancer (EOC) and other FOLR1-positive solid
tumors. J Clin Oncol. 2013;31(15 Suppl.):2573.
42. Deng YJ, Wang YQ, Cherian C, et al. Synthesis and discovery of high affinity folate
receptor-specific glycinamide ribonucleotide formyltransferase inhibitors with anti-
tumor activity. J Med Chem. 2008;51:5052–5063.
31. Reddy JA, Dorton R, Bloomfield A, et al. Pre-clinical evaluation of EC1456, a folate-
tubulysin anti-cancer therapeutic. Sci Rep. 2018;8:8943.
32. Banerji U, Garces AHI, Michalarea V, et al. An investigator-initiated phase I study of
ONX-0801, a first-in-class alpha folate receptor targeted, small molecule thymidylate
synthase inhibitor in solid tumors. J Clin Oncol. 2017;35(15 suppl):2503.
33. Desmoulin SK, Wang Y, Wu J, et al. Targeting the proton-coupled folate transporter
for selective delivery of 6-substituted pyrrolo[2,3-d]pyrimidine antifolate inhibitors
of de novo purine biosynthesis in the chemotherapy of solid tumors. Mol Pharmacol.
2010;78:577–587.
43. Meanwell NA. Synopsis of some recent tactical application of bioisosteres in drug
design. J Med Chem. 2011;54:2529–2591.
44. Maestro, Schrödinger Release 2016-4; Schrödinger, LLC: New York, NY; 2016.
45. Chen C, Ke JY, Zhou XE, et al. Structural basis for molecular recognition of folic acid
by folate receptors. Nature. 2013;500(7463):486–489.
46. Smart BP, Pan YH, Weeks AK, Bollinger JG, Bahnson BJ, Gelb MH. Inhibition of the
complete set of mammalian secreted phospholipases A(2) by indole analogues: a
structure-guided study. Bioorgan Med Chem. 2004;12:1737–1749.
34. Golani LK, George C, Zhao S, et al. Structure-activity profiles of novel 6-substituted
pyrrolo[2,3-d]pyrimidine Thienoyl antifolates with modified amino acids for cellular
uptake by folate receptors alpha and beta and the proton-coupled folate transporter.
J Med Chem. 2014;57:8152–8166.
47. Flintoff WF, Davidson SV, Siminovitch L. Isolation and partial characterization of
three methotrexate-resistant phenotypes from Chinese hamster ovary cells. Somatic
Cell Genet. 1976;2:245–261.
35. Golani LK, Wallace-Povirk A, Deis SM, et al. Tumor targeting with novel 6-sub-
stituted pyrrolo [2,3-d] pyrimidine antifolates with heteroatom bridge substitutions
via cellular uptake by folate receptor alpha and the proton-coupled folate transporter
and inhibition of de novo purine nucleotide biosynthesis. J Med Chem.
2016;59:7856–7876.
48. Wong SC, Proefke SA, Bhushan A, Matherly LH. Isolation of human cdnas that restore
methotrexate sensitivity and reduced folate carrier activity in methotrexate trans-
port-defective chinese-hamster ovary cells. J Biol Chem. 1995;270:17468–17475.
49. Deng YJ, Zhou XL, Desmoulin SK, et al. Synthesis and biological activity of a novel
series of 6-substituted thieno[2,3-d]pyrimidine antifolate inhibitors of purine bio-
synthesis with selectivity for high affinity folate receptors over the reduced folate
carrier and proton-coupled folate transporter for cellular entry. J Med Chem.
2009;52:2940–2951.
36. Kugel Desmoulin S, Wang L, Hales E, et al. Therapeutic targeting of a novel 6-sub-
stituted pyrrolo [2,3-d]pyrimidine thienoyl antifolate to human solid tumors based
on selective uptake by the proton-coupled folate transporter. Mol Pharmacol.
2011;80:1096–1107.
50. Dekhne AS, Shah K, Ducker GS, et al. Novel pyrrolo[3,2-d]pyrimidine compounds
target mitochondrial and cytosolic one-carbon metabolism with broad-spectrum
antitumor efficacy. Mol Cancer Ther. 2019;18:1787–1799.
37. Ravindra M, Wilson MR, Tong N, et al. Fluorine -substituted pyrrolo[2,3-d]pyr-
imidine analogues with tumor targeting via cellular uptake by folate receptor alpha
and the proton-coupled folate transporter and inhibition of de novo purine nucleo-
tide biosynthesis. J Med Chem. 2018;61:4228–4248.
51. Flintoff WF, Nagainis CR. Transport of methotrexate in Chinese hamster ovary cells: a
mutant defective in methotrexate uptake and cell binding. Arch Biochem Biophys.
1983;223(2):433–440.
38. Wang L, Cherian C, Desmoulin SK, et al. Synthesis and biological activity of 6-sub-
stituted pyrrolo[2,3-d]pyrimidine thienoyl regioisomers as inhibitors of de novo
14