A fragment-like approach to PYCR1 inhibition

Article abstract of DOI:10.1016/j.bmcl.2019.07.047

Pyrroline-5-carboxylate reductase 1 (PYCR1) is the final enzyme involved in the biosynthesis of proline and has been found to be upregulated in various forms of cancer. Due to the role of proline in maintaining the redox balance of cells and preventing apoptosis, PYCR1 is emerging as an attractive oncology target. Previous PYCR1 knockout studies led to a reduction in tumor growth. Accordingly, a small molecule inhibitor of PYCR1 could lead to new treatments for cancer, and a focused screening effort identified pargyline as a fragment-like hit. We report the design and synthesis of the first tool compounds as PYCR1 inhibitors, derived from pargyline, which were assayed to assess their ability to attenuate the production of proline. Structural activity studies have revealed the key determinants of activity, with the most potent compound (4) showing improved activity in vitro in enzyme (IC₅₀ = 8.8 μM) and pathway relevant effects in cell-based assays.

Full text of DOI:10.1016/j.bmcl.2019.07.047

Journal Pre-Proof
A Fragment-Like Approach to PYCR1 Inhibition
Kirsty Milne, Jianhui Sun, Esther A. Zaal, Jenna Mowat, Patrick H.N. Celie,
Alexander Fish, Celia R. Berkers, Giuseppe Forlani, Fabricio Loayza-Puch,
Craig Jamieson, Reuven Agami
PII:
S0960-894X(19)30503-7
https://doi.org/10.1016/j.bmcl.2019.07.047
BMCL 26588
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
19 June 2019
23 July 2019
24 July 2019
Please cite this article as: Milne, K., Sun, J., Zaal, E.A., Mowat, J., Celie, P.H.N., Fish, A., Berkers, C.R., Forlani,
G., Loayza-Puch, F., Jamieson, C., Agami, R., A Fragment-Like Approach to PYCR1 Inhibition, Bioorganic &
Medicinal Chemistry Letters (2019), doi: https://doi.org/10.1016/j.bmcl.2019.07.047
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will
undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing
this version to give early visibility of the article. 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.
© 2019 Published by Elsevier Ltd.

JOURNAL PRE-PROOF
A Fragment-Like Approach to PYCR1 Inhibition
Kirsty Milnea, Jianhui Sunb, Esther A. Zaalc, Jenna Mowata, Patrick H. N. Celieb,d, Alexander Fishb,d, Celia R. Berkersc,e,
Giuseppe Forlanif, Fabricio Loayza-Puchb*, Craig Jamiesona*, and Reuven Agamib,g*
aDepartment of Pure and Applied Chemistry, Thomas Graham Building, University of Strathclyde, Glasgow, G1 1XL;
bH5 Division of Oncogenomics, Oncode Institute, The Netherlands Cancer Institute,121 Plesmanlaan,1066 CX,
Amsterdam, The Netherlands; cBiomolecular Mass Spectrometry and Proteomics, Bijvoet Center for Biomolecular
Research, Utrecht University, Padualaan 8, 3584 CH, Utrecht, The Netherlands; dNKI Protein Facility, Division of
Biochemistry, The Netherlands Cancer Institute,121 Plesmanlaan,1066 CX, Amsterdam, The Netherlands; eDepartment
of Biochemistry and Cell Biology, Faculty of Veterinary Medicine, Utrecht University, Yalelaan 2, 3584 CM, Utrecht,
The Netherlands; fDepartment of Life Science and Biotechnology, University of Ferrara, 44121 Ferrara, Italy;
1gDepartment of Molecular Genetics, Erasmus MC, Rotterdam University, The Netherlands
KEYWORDS: inhibitor, proline modulation, oncology, tool compound, fragment-based drug discovery
GRAPHICAL ABSTRACT
ABSTRACT: Pyrroline-5-carboxylate reductase 1 (PYCR1) is the final enzyme involved in the biosynthesis of proline
and has been found to be upregulated in various forms of cancer. Due to the role of proline in maintaining the redox
balance of cells and preventing apoptosis, PYCR1 is emerging as an attractive oncology target. Previous PYCR1
knockout studies led to a reduction in tumor growth. Accordingly, a small molecule inhibitor of PYCR1 could lead to
new treatments for cancer, and a focused screening effort identified pargyline as a fragment-like hit. We report the design
and synthesis of the first tool compounds as PYCR1 inhibitors, derived from pargyline, which were assayed to assess
their ability to attenuate the production of proline. Structural activity studies have revealed the key determinants of
activity, with the most potent compound (4) showing improved activity in vitro in enzyme (IC50 = 8.8 µM) and pathway
relevant effects in cell-based assays.
Pyrroline-5-carboxylate reductase 1 (PYCR1) is the final enzyme involved in the biosynthesis of proline from both
glutamic acid and ornithine, as outlined in Scheme 1.1 Glutamic acid is firstly phosphorylated by glutamate 5-kinase
(G5K), before being dephosphorylated by gamma-glutamyl phosphate reductase (γ-GPR) to produce glutamate-γ-
semialdehyde, which exists in an equilibrium with pyrroline-5-carboxylate (P5C). Ornithine is also transformed to the
same intermediate through the action of ornithine amino transferase (OAT). P5C is then finally reduced to proline by
PYCR1 using the reduced form of nicotinamide adenine dinucleotide phosphate (NADPH) as a cofactor, although in
vitro nicotinamide adenine dinucleotide (NADH) can also serve as a co-factor.
Proline is essential for protein synthesis and plays a role in the secondary structure of proteins.2 This amino acid and its
derivatives are also the main residues found in collagen, the most abundant protein found within the body.3 However, it
also plays a role in maintaining the redox balance of cells through a process known as the proline cycle, outlined in
Figure 1.3–5
*corresponding authors: C. Jamieson (craig.jamieson@strath.ac.uk); R. Agami; (r.agami@nki.nl); F. Loayza-Puch
f.loayza-puch@dkfz-heidelberg.de;
Current address (F. Loayza-Puch): Translational Control and Metabolism, Deutsches rebsforschungszentrum
Im Neuenheimer, Feld 280, 69120 Heidelberg

JOURNAL PRE-PROOF
2-
O
O
O
OPO3
ATP
ADP
O
NADPH
O
H3N
H3N
G5K
NADP+
O
O
O
NADP+
-GPR
NADPH
glutamic acid
O
H2
N
O
H
N
H3N
O
PYCR1
O
O
H3N
P5C
proline
O
OAT
O
glutamate--semialdehyde
H3N
O
ornithine
Scheme 1: The proline biosynthesis pathway (adapted from reference 1).
P5C is regenerated in the mitochondria by the oxidation of proline by proline dehydrogenase (PRODH), generating
adenosine triphosphate (ATP), in the process. Outside of the mitochondria, P5C can be reconverted to proline. This
produces a molecule of nicotinamide adenine dinucleotide phosphate (NADP+) which is available for use by the pentose-
phosphate pathway. The pentose-phosphate pathway will eventually produce ribose-5-phosphate (R-5-P), which can be
used to synthesize nucleotides or undergo further transformations to eventually reach fructose-6-phosphate (F-6-P), that
is able to produce ATP through glycolysis. These three process have an essential role in the survival and proliferation of
cells.6–8 The pentose-phosphate pathway also reduces NADP+ back to NADPH which then supports the disulfide
reduction system via thioredoxin reductase and glutathione reductase, minimising production of reactive oxygen species
(ROS), again contributing to cell survival.
Figure 1: The proline cycle and pentose-phosphate pathway (adapted from reference 2).
In healthy cells, these processes are highly regulated and essential for maintaining normal function.5 However, in certain
cancers; such as breast, prostate and some lung and skin cancers, PYCR1 is found to be upregulated.9–15 This leads to
higher levels of proline and exacerbated effects of the proline cycle, with the cells effectively using this method to
increase cell survival.5,15 If a method of reducing or inhibiting PYCR1 could be discovered, it could provide a new means
of treating cancer.

JOURNAL PRE-PROOF
We and others have reported studies in both breast cancer12 and human prostate14 cell lines showing that PYCR1
knockout causes phenotypic changes in the cell. In the prostate cancer studies, this resulted in an increase in cell cycle
arrest and apoptosis in vitro, while in the breast cancer studies a reduction in tumor size in vivo was observed. These
experiments have shown that modulating PYCR1 directly affects the survival rate of some cancers, validating PYCR1 as
an emerging oncology target.
While omission of a gene is a useful tool in its own right, the process is complex and careful selection of vectors and
delivery vehicles are required to minimize inflammatory and off-target effects, which can make the process lengthy and
expensive.16–18 Furthermore, gene therapy is not yet an officially approved treatment for any disease and until more is
known about the human genome, this is not likely to be approved in the near future.19 A more tractable approach would
utilize a small molecule tool compound, potentially leading to a new approved therapy for cancer treatment.
In order to identify a chemical starting point, a commercially available library of pharmaceutically active compounds
(LOPAC®1280, Sigma Aldrich)20 was screened against PYCR1. Based on this screening campaign, pargyline was
identified as a fragment-like hit (1, Figure 2). Pargyline had a modest IC50 of 198 µM, however, displayed an
encouraging ligand efficiency (LE) of 0.42.21
N
IC50 = 198 M
LE = 0.42
1
Figure 2: Structure and activity of pargyline.
These considerations, coupled with its low molecular weight, made it an attractive fragment-like hit and a number of
analogues were prepared in order to assess the structure activity relationship (SAR). Starting from the appropriate amine,
the target analogues were synthesized via alkylation or reductive amination with the corresponding bromide or aldehyde,
respectively, as outlined in Scheme 2.
A.
K2CO3, 0-10 mol% KI
HN
N
+
Br
MeCN, R.T. - 40 o
C
B.
i) AcOH, dry DCM, R.T.
ii) STAB, 0 oC - R.T.
HN
N
+
O
Scheme 2: A: Alkylation conditions; bromide (0.25 mmol), amine (0.25 mmol), acetonitrile (0.8 M), with potassium
iodide added for less activated bromides. B: reductive amination conditions; aldehyde (3 mmol), amine (1 mmol), acetic
acid (1 mmol), dry dichloromethane (0.05 M) then sodium triacetoxyborohydride (3 mmol).
The modular structure of pargyline makes it amenable to targeted elaboration of three principal regions: the benzyl, N-
methyl and propargyl groups. As no crystallographic information of the binding site pargyline occupies within PYCR1
was available, a stepwise approach was adopted in order to assess the impact of changing these substituents on enzyme
inhibition using the in vitro compound screening assay (see Supplementary Material).
Initially, changes to the benzyl group were examined. Due to their relative abundance in pharmaceutically active
compounds, the presence of halogens in various positions around the phenyl ring of the benzyl group was assessed first
(Figure 3A).22 Pleasingly almost all of the compounds were found to be more active than pargyline at PYCR1, with the
exception of the 4-fluoro, 2-chloro and 2-bromo derivatives (2, 8 and 9, respectively) which showed reduced activity. It
was also found that the influence on activity was greatest when the halogen was in the 4-position (3-5) and weakest in the
2-position (7 and 9), with the 3-position being between the two potencies (6 and 7). The size of the halogen in the 4-
position was also found to have an effect with a general increase in potency observed moving down the group (2-5) , with
the optimum being the 4-bromo system, which was very similar in potency to the 4-iodo moiety, with both of these
surpassing the 4-chloro derivative.
In order to follow up on these observations, a number of different functional groups on the benzyl ring were studied, as
shown in Figure 3B. A range of electron donating and withdrawing groups were analyzed in various positions around the
ring. Again, the 4-position was favored with a similar pattern of activity being observed with the nitro group (14-16) as
with the halogens. Unfortunately, none of the compounds matched the potency of compound 4, with no clear preference
for electron donating (10-12) or withdrawing groups (13-18) noted from this study.

JOURNAL PRE-PROOF
With more potent analogues observed with increasing size of halogen, it was reasoned that larger groups on the benzyl
ring could also result in an increase in potency. Due to the similarities in volume of iodine and phenyl moieties23,24 some
biphenyls (19 and 20) were prepared by Suzuki coupling of compound 4 (See Supplementary Material). Unfortunately,
these analogues were less active than both pargyline and compound 4, suggesting that there are other properties beyond
the size of substituent contributing to the increase in potency. This could be linked to a halogen bonding effect, where the
strength of an interaction increases with the size of the halogen atom, as observed with analogues 2-4.22

JOURNAL PRE-PROOF
A
N
N
N
N
I
F
Cl
Br
2
3
4
5
IC50 = 311 M
LE = 0.37
IC50 = 13 M
LE = 0.51
IC50 = 8.8 M
LE = 0.53
IC50 = 9.7 M
LE = 0.53
Cl
Br
N
N
N
N
Cl
Br
6
7
8
9
IC50 = 52 M
LE = 0.45
IC50 = 276 M
LE = 0.37
IC50 = 139 M
LE = 0.41
IC50 = 200 M
LE = 0.39
B
N
N
N
N
O
F3C
10
11
12
13
IC50 = 40 M
LE = 0.46
IC50 = 38 M
LE = 0.43
IC50 = 142 M
LE = 0.35
IC50 = 65 M
LE = 0.35
O2N
N
N
N
N
O2N
NO2
16
O
14
15
17
IC50 = 43 M
LE = 0.40
IC50 = 394 M
LE = 0.31
IC50 = >500 M
LE = n.d.
IC50 = 184 M
LE = 0.34
N
N
N
NC
18
19a
20a
F
IC50 = >500 M
IC50 = 315 M
LE = 0.27
IC50 = 211 M
LE = 0.26
LE = n.d.
C
Cl
F3C
Cl
Cl
Cl
Cl
N
N
N
N
Cl
21
22
23
24
Cl
IC50 = 43 M
LE = 0.35
IC50 = 13.6 M
LE = 0.48
IC50 = >500 M
IC50 = >500 M
LE = n.d.
LE = n.d.
Cl
O
Br
N
N
N
N
Cl
O
F
25
26
27
28
IC50 = 9.2 M
LE = 0.49
IC50 = 55 M
LE = 0.42
IC50 = >500 M
LE = n.d.
IC50 = >500 M
LE = n.d.
Br
F
Cl
Br
N
N
N
N
Cl
Br
Br
Br
29
30
31
32
IC50 = 66 M
LE = 0.41
IC50 = 42 M
LE = 0.43
IC50 = 34 M
LE = 0.44
IC50 = 228 M
LE = 0.36
O
N
N
N
Br
Br
33
34
35
IC50 = 38 M
IC50 = 21 M
IC50 = 21 M
LE = 0.43
LE = 0.43
LE = 0.28
D
Br
N
N
N
N
Br
O
N
S
Br
36
37
38
39
IC50 = >500 M
IC50 = 316 M
LE = 0.34
IC50 = >500 M
IC50 = >500 M
LE = n.d.
LE = n.d.
LE = n.d.
Figure 3: A: SAR of halogen substitutions (blue). B: SAR of non-halogen substitutions (blue). C: SAR of disubstitutions
(blue). D: SAR of non-benzyl substitutions (blue). a prepared by Suzuki-Miyaura cross-coupling.

JOURNAL PRE-PROOF
Compound 4 was found to have the highest potency, also improving the LE, showing that the presence of a bromine atom
on the 4-position of the benzyl group may be beneficial to binding.
With the previous exemplars only representing a single substitution on the benzyl group, it was decided to incorporate a
variety of disubstituted analogues as shown in Figure 3C. Initially, this was probed using dichlorobenzyl moieties due to
their synthetic availability. In comparison to compound 3, having substitution of the 3- and 4-positions resulted in a
compound with a similar potency to the monosubstituted 4-position, while the 3,5- and 2,6- derivatives (compounds 23
and 24, respectively) were inactive.
Exchanging the 3-chloro of compound 22 for a trifluoromethyl group (21) resulted in a less active compound, again
suggesting the importance of halogens. However, the 2,4-dichloro species 25 had a much greater potency than compound
3 and a similar potency to compound 4. This was surprising as a monosubstituted compound in the 2-position was found
to be the least active regioisomer in the initial halogen screen (7 and 9). This increase in potency could be due to an extra
interaction at the, as yet unknown, binding site of the enzyme which complements the interaction at the 4-position. Other
disubstituted modifications (26 – 34) and naphthyl (25) resulted in lower levels of activity.
The final modifications to the benzyl group involved exchanging it for a completely different functional handle to assess
its necessity for activity, as outlined in Figure 3D. Homologation (36) and exchange of the 6-membered benzyl group for
5-membered heterocycles (38 and 39) resulted in compounds that were inactive, indicating that the benzyl group is
essential for activity. Incorporation of a branched methyl in the benzyl position (37) reduced the activity of the
compound, suggesting that there may be a steric clash at the binding site.
With the 4-bromobenzyl moiety identified as the optimal group for PYCR1 inhibition, and this motif was carried on
throughout the rest of the SAR exploration.
The second group assessed was the N-methyl moiety as outlined in Figure 4. There was markedly less tolerance in this
group than with the benzyl group with only four analogues showing any measurable activity against the enzyme. Carbon
chains longer than an ethyl group (42 and 43) were found to be inactive as were branched moieties (44 and 45). Indeed,
of the alkyl substituents only ethyl (41) was active, albeit less so than the methyl substituent of compound 4. Having no
substituent on the nitrogen core (40) also resulted in a less potent compound. Larger groups such as 4-bromo benzyl (47)
and isoxazolyl (46) were moderately active. However, the higher molecular weight of these compounds, resulted in a
lower LE. Larger substituents such as 48 were found not to be tolerated. Tethering the N-group to the benzyl group was
also not tolerated with the isoindoline and tetrahydroisoquinoline (49 and 50, respectively) both being inactive. This
could be due to the molecules being constrained in the wrong conformation for binding to the target. The final changes
assessed were to the nature of the nitrogen core, with the introduction of an amide (51) or sulfonamide (52) in the place
of a basic tertiary amine. Both of these were inactive suggesting that a basic amine is necessary for activity.
The last modifications examined were alterations to the propargyl moiety, as outlined in Figure 5. As with the N-methyl
group, there was limited tolerance in changes to this group. Fully reducing (55) or completely removing the propargyl
group (53) resulted in inactive derivatives, while reduction to the propenyl analogue (54) drastically reduced activity.
Introduction of a benzyl (57) and cyclopropyl group (58) also resulted in inactive compounds and suggests the alkyne
component of the molecule is required. Homologation (56) and incorporation of a branched methyl to the propargyl unit
(61) resulted in a considerable reduction in potency. In the case of the branched analogue, this could be due to a steric
restriction in the binding site, however further cements the need for a propargyl amine for optimum potency. Finally, the
terminal alkyne was exchanged for internal alkynes 59 and 60. Pleasingly, these compounds retained some degree of
potency which decreased when the size of the capping group increased. However, they remained less potent than
compound 4, which again could be attributed to a steric constraint. It should be noted that further structural data would be
required to corroborate this hypothesis.

JOURNAL PRE-PROOF
N
N
N
N
H
Br
Br
Br
Br
40
41
42
43
IC50 = 99 M
IC50 = 58 M
IC50 = >500 M
IC50 = >500 M
LE = 0.46
LE = 0.41
LE = n.d.
LE = n.d.
N
N
N
N
O
Br
Br
Br
Br
N
44
45
46
47
Br
IC50 = >500 M
IC50 = >500 M
IC50 = 100 M
LE = 0.29
IC50 = 92 M
LE = 0.28
LE = n.d.
LE = n.d.
N
O
N
N
N
Br
Br
Br
Br
O
O
tBu
48
49
50
51
IC50 = >500 M
IC50 = >500 M
LE = n.d.
IC50 = >500 M
LE = n.d.
IC50 = >500 M
LE = n.d.
LE = n.d.
N
O
S O
Br
52
IC50 = >500 M
LE = n.d.
Figure 4: Structure and activity of alterations to the N-methyl group (red).
NH
N
N
N
Br
Br
Br
Br
53a
54
55
56
IC50 = 161 M
IC50 = >500 M
IC50 = 337 M
IC50 = >500 M
LE = 0.40
LE = n.d.
LE = 0.34
LE = n.d.
N
N
N
N
Br
Br
Br
Br
57
58
59
60
IC50 = >500 M
LE = n.d.
IC50 = >500 M
IC50 = 16 M
LE = 0.47
IC50 = 71 M
LE = 0.30
LE = n.d.
N
Br
61
IC50 = 394 M
LE = 0.33
Figure 5: Structure and activity of changes to the propargyl group (purple). a Commercially available feedstock
Compound 4 was found to be the most active of all the analogues synthesized and accordingly was advanced for further
biological evaluation as the lead compound.
The first study considered was levels of whole cell proline. As the inhibition of PYCR1 should prevent the formation of
proline, compound 4 should lower the concentration of intracellular proline. Thus, a human breast cancer cell line (SUM-
159-PT) was incubated with both 100 µM pargyline and compound 4 at concentrations of 1, 5, 20 and 100 µM,
respectively. The cells were lysed and the quantity of proline (Pro) and essential amino acids were acquired by LCMS.
As outlined in Figure 6a, pargyline is a modest inhibitor of PYCR1 with approximately 50% reduction in proline levels
at a concentration of 100 µM, in comparison to the control, with no effect on the quantities of the other amino acids
measured (Figure S1, Supplementary Material), with no effects on celluar toxicity apparent at this concentration.
Pleasingly, a similar effect is shown with compound 4, with approximately 50 % reduction in proline shown at 1 µM and
no effect on the other amino acid levels. Furthermore, a concentration dependent response was observed with the proline
levels decreasing with increasing concentration of compound 4 until 100 µM, where it then increases. The increase in
proline levels could be attributed to the limited solubility of compound 4 in the buffer at higher concentrations. In order
to show if the lower levels of proline are a result of PYCR1 inhibition, we performed a 13C-glutamine tracer experiment.
By culturing cells with [U-13C]-glutamine we followed the incorporation of 13C from glutamine, via glutamate, in proline
(M+5), Figure 6b. After 24 hours incubation with compound 4, less proline M+5 is present compared to the controls.
These data further show that lower levels of proline can be attributed to significant inhibition of synthesis from
glutamine.

JOURNAL PRE-PROOF
All of the above data indicates that both pargyline and compound 4 not only inhibit PYCR1, but are also inducing a
measurable, pathway-relevant, biological response. To the best of our knowledge this represents the first time such a
phenotypic response been observed using small molecules as PYCR1 inhibitors.
In an additional study, two breast cancer cell lines, MDA-MB-231 and SUM-159-PT, were incubated with and without
the presence of compound 4 and the number of cells monitored as a measure of the percentage confluence as shown in
Figure 7. Both cell lines have relatively high PYCR1 levels, and depend on PYCR1 activity for proliferation and tumor
formation.12
The control experiment, without compound 4, was run both with and without exogenous proline. Here, there was no
difference between the two experiments with both types of cell showing that exogenous proline had no effect on the
growth of the cells. However, when compound 4 was dosed at 10 µM with no exogenous proline, the number of cells
was significantly reduced in both cases, with a 40% reduction in the MDA-MD-231 cells and a 30% reduction in the
SUM159PT cell line; the effect of compound 4 was negated by supplementing proline to the medium. This demonstrated
unequivocally that compound 4 is having a phenotypic effect on the cell cycle of both cell types. As normal growth is
observed in the presence of exogenous proline it may be reasoned that compound 4 is diminishing the levels of proline
within the cell and causing the retardation in growth. As compound 4 is derived from pargyline it does retain residual
monoamine-B activity (IC50 = 0.3 μM), however, these experiments indicated that the phenotyptic effects observed are
pathway relevant driven through inhibition of PYCR1. Having stated this, the effects of the compound as an inhibitor at
the related PYCR2 and PYCR3 enzymes cannot be excluded at present.
Figure 6. (a) Results of the LC-MS based amino acid assay, showing the quantity of proline present in lysed human
breast cancer cells (SUM-159-PT) after incubation with pargyline and compound 4. (b) Results of the glutamine flux
study showing lower levels of M+5 proline after incubation with compound 4 (5 μM).

JOURNAL PRE-PROOF
Figure 7: Results of the cell proliferation assay for both MDA-MB-321 and SUM159PT breast cancer cell lines.
Controls were performed both with and without exogenous proline showing normal growth in both cell lines. Compound
4 (10 µM) showed a 40% and 30% reduction in the MDA-MB-231 and SUM159PT cells, respectively. Cell growth was
completely rescued in the presence of exogenous proline.
In conclusion, over 60 potential small molecule inhibitors of PYCR1 were synthesized in order to probe the SAR around
PYCR1 inhibition. Of all the synthesized analogues, compound 4 (termed Proline Production Inhibitor-1, ProPI-1) was
identified as the most potent in the PYCR1 inhibition assay with an IC50 of 8.8 µM, approximately 20 times more potent
than pargyline, with the most efficient binding as evidenced by a LE of 0.53. As a result of this PPI-1 was taken forward
as a lead into a series of pathway-relevant biological tests and, was found to significantly reduce the levels of proline
within a breast cancer cell line and reduce cell proliferation in two different breast cancer cell lines – the first time this
has been achieved using a small molecule as a PYCR1 inhibitor. Further work is ongoing to obtain an X-ray crystal
structure of ProPI-1 and PYCR1, as well as further target validation through demonstration of efficacy in vivo.
ACKNOWLEDGEMENTS
GlaxoSmithKline/Engineering and Physical Sciences Research Council UK (KM); The Dutch Cancer Society KWF
2017-11037 (RA). The National Mass Spectrometry Facility at Swansea University for provision of HRMS data.
SUPPLEMENTARY MATERIAL
Experimental procedures and characterization data of all pargyline analogues and protocol for PYCR1 testing.
REFERENCES AND NOTES
1 .
2 .
Christensen, E. M.; Patel, S. M.; Korasick, D. A.; Campbell, A. C.; Krause, K. L.; Becker, D. F.; Tanner, J. J. J.
Biol. Chem. 2017, 292, 7233.
Deber, C. M.; Brodsky, B.; Rath, A. In Encyclopedia of Life Sciences; John Wiley & Sons, Ltd: Chichester, UK,
2010.
3.
4.
5.
6.
Phang, J. M.; Donald, S. P.; Pandhare, J.; Liu, Y. Amino Acids 2008, 35, 681.
Liang, X.; Zhang, L.; Natarajan, S. K.; Becker, D. F. Antioxid. Redox Signal. 2013, 19, 998.
Tanner, J. J.; Fendt, S. M.; Becker, D. F. Biochemistry 2018, 57, 3433.
Song, S.; Jacobson, K. N.; McDermott, K. M.; Reddy, S. P.; Cress, A. E.; Tang, H.; Dudek, S. M.; Black, S. M.;
Garcia, J. G. N.; Makino, A.; Yuan, J. X.-J. Am. J. Physiol. - Cell Physiol. 2015, 310, ajpcell.00092.2015.
Lane, A. N.; Fan, T. W.-M. Nucleic Acids Research. Oxford University Press 2015, 2466.
Kim, J. H.; Nam, B.; Choi, Y. J.; Kim, S. Y.; Lee, J. E.; Sung, K. J.; Kim, W. S.; Choi, C. M.; Chang, E. J.;
Koh, J. S.; Song, J. S.; Yoon, S.; Lee, J. C.; Rho, J. K.; Son, J. Cancer Res. 2018, 78, 4482.
7.
8

JOURNAL PRE-PROOF
9.
Craze, M. L.; Cheung, H.; Jewa, N.; Coimbra, N. D. M.; Soria, D.; El-Ansari, R.; Aleskandarany, M. A.; Wai
Cheng, K.; Diez-Rodriguez, M.; Nolan, C. C.; Ellis, I. O.; Rakha, E. A.; Green, A. R. Br. J. Cancer 2018, 118,
258.
10.
11.
12.
13.
Cai, F.; Cai, F.; Miao, Y.; Miao, Y.; Liu, C.; Liu, C.; Wu, T.; Wu, T.; Shen, S.; Shen, S.; Su, X.; Su, X.; Shi, Y.;
Shi, Y. Oncol. Lett. 2018, 15, 731.
Ding, J.; Kuo, M.-L.; Su, L.; Xue, L.; Luh, F.; Zhang, H.; Wang, J.; Lin, T. G.; Zhang, K.; Chu, P.; Zheng, S.;
Liu, X.; Yen, Y. Carcinogenesis 2017, 38, 519.
Loayza-Puch, F.; Rooijers, K.; Buil, L. C. M.; Zijlstra, J.; Oude Vrielink, J. F.; Lopes, R.; Ugalde, A. P.; Van
Breugel, P.; Hofland, I.; Wesseling, J.; Van Tellingen, O.; Bex, A.; Agami, R. Nature 2016, 530, 490.
De Ingeniis, J.; Ratnikov, B.; Richardson, A. D.; Scott, D. A.; Aza-Blanc, P.; De, S. K.; Kazanov, M.;
Pellecchia, M.; Ronai, Z.; Osterman, A. L.; Smith, J. W. PLoS One 2012, 7, e45190.
Zeng, T.; Zhu, L.; Liao, M.; Zhuo, W.; Yang, S.; Wu, W.; Wang, D. Med. Oncol. 2017, 34, 27.
Cai, F.; Miao, Y.; Liu, C.; Wu, T.; Shen, S.; Su, X.; Shi, Y. Oncol. Lett. 2017, 15, 731.
Selkirk, S. M. Postgraduate Medical Journal. 2004, 560.
Kang, X.; He, W.; Huang, Y.; Yu, Q.; Chen, Y. J. Assist. Reprod. Genet. 2016, 33, 581.
Liang, P.; Xu, Y.; Zhang, X.; Ding, C.; Huang, R.; Zhang, Z.; Lv, J.; Xie, X.; Chen, Y.; Li, Y.; Sun, Y.; Bai, Y.;
Songyang, Z.; Ma, W.; Zhou, C.; Huang, J. Protein Cell 2015, 6, 363.
14.
15.
16.
17.
18.
19.
20.
21.
Petropoulos, S.; Edsgärd, D.; Reinius, B.; Deng, Q.; Panula, S. P.; Codeluppi, S.; Plaza Reyes, A.; Linnarsson,
S.; Sandberg, R.; Lanner, F. Cell 2016, 165, 1012.
LOPAC®1280
Sigma-Aldrich
https://www.sigmaaldrich.com/catalog/product/sigma/lo3300?lang=en&region=GB (accessed Oct 5, 2018).
Chen, H.; Engkvist, O.; Kogej, T. In The Practice of Medicinal Chemistry: Fourth Edition; Academic Press,
2015, 379.
22.
23.
24.
Ford, M. C.; Ho, P. S. J. Med. Chem. 2016, 59, 1655.
Batsanov, S. S. Inorg. Mater. 2001, 37, 871.
Stepan, A. F.; Subramanyam, C.; Efremov, I. V.; Dutra, J. K.; O’Sullivan, T. J.; Dirico, K. J.; McDonald, W. S.;
Won, A.; Dorff, P. H.; Nolan, C. E.; Becker, S. L.; Pustilnik, L. R.; Riddell, D. R.; Kauffman, G. W.; Kormos,
B. L.; Zhang, L.; Lu, Y.; Capetta, S. H.; Green, M. E.; Karki, K.; Sibley, E.; Atchison, K. P.; Hallgren, A. J.;
Oborski, C. E.; Robshaw, A. E.; Sneed, B.; O’Donnell, C. J. J. Med. Chem. 2012, 55, 3414.
Highlights



higher levels of protein are found in certain types of cancer
this is attributed to overexpression of a key protein involved in proline biosynthesis
focused screening has for the first time identified a small fragment-like inhibitor (ProPI-1) of
this protein target

JOURNAL PRE-PROOF


mechanistic studies show ProPI-1 can attenuate proline biosynthesis in relevant cell types
phenotypic studies reveal that ProPI-1 can control cell proliferation in two different breast
cancer cell types