Targeting the Hemopexin-like Domain of Latent Matrix Metalloproteinase-9 (proMMP-9) with a Small Molecule Inhibitor Prevents the Formation of Focal Adhesion Junctions

Article abstract of DOI:10.1021/acschembio.7b00758

A lack of target specificity has greatly hindered the success of inhibitor development against matrix metalloproteinases (MMPs) for the treatment of various cancers. The MMP catalytic domains are highly conserved, whereas the hemopexin-like domains of MMPs are unique to each family member. The hemopexin-like domain of MMP-9 enhances cancer cell migration through self-interaction and heterointeractions with cell surface proteins including CD44 and α4β1 integrin. These interactions activate EGFR-MAP kinase dependent signaling that leads to cell migration. In this work, we generated a library of compounds, based on hit molecule N-[4-(difluoromethoxy)phenyl]-2-[(4-oxo-6-propyl-1H-pyrimidin-2-yl)sulfanyl]-acetamide, that target the hemopexin-like domain of MMP-9. We identify N-(4-fluorophenyl)-4-(4-oxo-3,4,5,6,7,8-hexahydroquinazolin-2-ylthio)butanamide, 3c, as a potent lead (K_d = 320 nM) that is specific for binding to the proMMP-9 hemopexin-like domain. We demonstrate that 3c disruption of MMP-9 homodimerization prevents association of proMMP-9 with both α4β1 integrin and CD44 and results in the dissociation of EGFR. This disruption results in decreased phosphorylation of Src and its downstream target proteins focal adhesion kinase (FAK) and paxillin (PAX), which are implicated in promoting tumor cell growth, migration, and invasion. Using a chicken chorioallantoic membrane in vivo assay, we demonstrate that 500 nM 3c blocks cancer cell invasion of the basement membrane and reduces angiogenesis. In conclusion, we present a mechanism of action for 3c whereby targeting the hemopexin domain results in decreased cancer cell migration through simultaneous disruption of α4β1 integrin and EGFR signaling pathways, thereby preventing signaling bypass. Targeting through the hemopexin-like domain is a powerful approach to antimetastatic drug development.

Full text of DOI:10.1021/acschembio.7b00758

Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES
Article
Targeting the Hemopexin-like Domain of Latent Matrix
Metalloproteinase-9 (proMMP-9) with a Small Molecule
Inhibitor Prevents the Formation of Focal Adhesion Junctions
Vincent M. Alford, Anushree Kamath, Xiaodong Ren, kunal kumar, Qianwen Gan, Monaf
Awwa, Michael Tong, Markus A Seeliger, Jian Cao, Iwao Ojima, and Nicole S Sampson
ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.7b00758 • Publication Date (Web): 25 Sep 2017
Downloaded from http://pubs.acs.org on September 26, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Chemical Biology is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 44
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
Targeting the Hemopexin-like Domain of Latent Matrix Metalloproteinase-9
(proMMP-9) with a Small Molecule Inhibitor Prevents the Formation of Focal
Adhesion Junctions
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Short title: PEX-9 Inhibitors Prevent Formation of Focal Adhesion Junctions
Vincent M. Alford,1,3 Anushree Kamath,2 Xiaodong Ren,2 Kunal Kumar,2 Qianwen Gan,2 Monaf
Awwa,2 Michael Tong,1 Markus A. Seeliger,1,4 Jian Cao,3 Iwao Ojima,2,4 Nicole S. Sampson 2,4*
1Department of Molecular and Cellular Pharmacology, Stony Brook University, Stony Brook,
NY, USA
2Department of Chemistry, Stony Brook University, Stony Brook, NY, USA
3Department of Medicine, Stony Brook University, Stony Brook, NY, USA
4Institute of Chemical Biology & Drug Discovery, Stony Brook, NY, USA
*Corresponding author. Tel. & Fax: (631) 632ꢀ7952 | (631) 632ꢀ5731 Email:
nicole.sampson@stonybrook.edu
1
ACS Paragon Plus Environment

ACS Chemical Biology
Page 2 of 44
1
2
3
4
5
6
7
8
ABSTRACT: A lack of target specificity has greatly hindered the success of inhibitor
development against matrix metalloproteinases (MMPs) for the treatment of various cancers.
The MMP catalytic domains are highly conserved, whereas the hemopexinꢀlike domain of
MMPs are unique to each family member. The hemopexinꢀlike domain of MMPꢀ9 enhances
cancer cell migration through selfꢀinteraction and heteroꢀinteractions with cell surface proteins
including CD44 and α4β1 integrin. These interactions activate EGFRꢀMAP kinase dependent
signaling that leads to cell migration. In this work, we generated a library of compounds, based
on hit molecule, Nꢀ[4ꢀ(difluoromethoxy)phenyl]ꢀ2ꢀ[(4ꢀoxoꢀ6ꢀpropylꢀ1Hꢀpyrimidinꢀ2ꢀyl)sulfanyl]ꢀ
acetamide, that target the hemopexinꢀlike domain of MMPꢀ9. We identify Nꢀ(4ꢀfluorophenyl)ꢀ4ꢀ
(4ꢀoxoꢀ3,4,5,6,7,8ꢀhexahydroquinazolinꢀ2ꢀylthio)butanamide, 3c, as a potent lead (Kd = 320
nM) that is specific for binding to proMMPꢀ9 hemopexinꢀlike domain. We demonstrate that 3c
disruption of MMPꢀ9 homodimerization prevents association of proMMPꢀ9 with both α4β1
integrin and CD44 and results in dissociation of EGFR. This disruption results in decreased
phosphorylation of Src and its downstream target proteins focal adhesion kinase (FAK) and
paxillin (PAX), which are implicated in promoting tumor cell growth, migration, and invasion.
Using a chicken chorioallantoic membrane in vivo assay, we demonstrate that 500 nM 3c
blocks cancer cell invasion of the basement membrane and reduces angiogenesis. In
conclusion, we present a mechanism of action for 3c whereby targeting the hemopexin domain
results in decreased cancer cell migration through simultaneous disruption of α4β1 integrin
and EGFR signaling pathways, thereby preventing signaling bypass. Targeting through the
hemopexinꢀlike domain is a powerful approach to antiꢀmetastatic drug development.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Keywords: MMPꢀ9; hemopexin domain; PEXꢀ9 inhibitors; focal adhesions; metastasis.
2
ACS Paragon Plus Environment

Page 3 of 44
ACS Chemical Biology
1
2
3
INTRODUCTION
4
5
6
7
8
9
Metastasis accounts for 90% of all human cancer related deaths yet we lack adequate
drugs to target this biological process often associated with aggressive cancers.1 Accumulating
evidence suggests an important role for matrix metalloproteinases (MMPs) in promoting
cancer progression whereby they modify their surrounding environment to promote the growth
and spread of tumor cells.2ꢀ5 Although MMPs represent the most prominent family of
proteinases associated with tumorigenesis, drugs designed to inhibit their proteolytic activities
largely failed in clinical trials due to issues with selectivity for individual MMPs.6 The highly
conserved catalytic domain within this family of zymogens required a paradigm shift to the
development of novel MMP inhibitors (MMPIs) targeting less conserved, nonꢀcatalytic
functional domain(s) of the proteases to increase target specificity and selectivity.7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The biological functions of MMPs extend beyond just their proteolytic function, and
include induction of complex signaling cascades.8ꢀ10 Homodimerization of secreted proMMPꢀ9
hemopexin domains is sufficient to induce cancer cell migration independent of proteolytic
activity.11 Using an shRNA approach the mechanism behind this phenotype was dissected and
found to be dependent on an interaction between MMPꢀ9 and CD44 at the cell surface. This
interaction required an intact MMPꢀ9 hemopexin domain (PEXꢀ9). Interaction between CD44
and PEXꢀ9 results in increased epidermal growth factor receptor (EGFR) phosphorylation and
subsequent activation of mitogenꢀactivated protein kinase (MAPK)/extracellular signal–
regulated kinases (ERK) signaling, thereby enhancing the migratory capacity of proMMPꢀ9
expressing cancer cells.12 Signaling was lost upon deletion or swapping of PEXꢀ9 with the PEX
domain of its closest homolog MMPꢀ2.
The Xꢀray crystal structure of the PEX domain of MMPꢀ9 has been solved. PEXꢀ9 forms
a propellerꢀlike structure composed of four blades, which converge and are linked between
blades I and IV through disulfide bonds.13 Each blade of the PEX domain is composed of a
3
ACS Paragon Plus Environment

ACS Chemical Biology
Page 4 of 44
1
2
3
4
5
single αꢀhelix and four antiꢀparallel βꢀstrands. They form a scaffold for substrate recognition
and docking14
6
7
8
9
Published reports have shown that PEXꢀ9 can interact with different integrin subunits to
promote enhanced cancer cell migration, invasion, and survival in various cancer cell types.15ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
20
Of particular interest to our study was a recent finding that peptide sequence
FPGVPLDTHDVFQYREKAYFC within the central cavity of PEX9/blade IV is a binding site for
α4β1 integrin. Treatment with synthetic peptide was found to prevent Bꢀcell chronic
lymphocytic leukemia transendothelial cell migration and survival.21, 22 Mutation of either
aspartic acid (Asp) residue (Aspꢀ660 and Aspꢀ663) in this peptide sequence abolished
inhibition supporting the importance of Asp residues as key recognition sites for integrins.23, 24
Furthermore, the LDT motif in this peptide sequence is a highly conserved motif found in a
variety of wellꢀknown α4β1 integrin ligands.23, 25, 26
We previously utilized an in silico DOCKing approach for finding small molecules that
bind to MMPꢀ9 and identified a substituted pyrimidone, 1a, which docked to the nonꢀcatalytic
PEX domain of MMPꢀ9.27 Compound 1a inhibited proMMPꢀ9ꢀmediated cancer cell migration
and proliferation in vitro. Importantly, this compound did not inhibit MMPꢀ9 enzymatic activity.
Pyrimidone 1a specifically binds to the hemopexin domain of MMPꢀ9 and does not bind to
closely related MMPs. In addition to inhibiting proMMP9ꢀmediated cell migration, compound 1a
effectively prevented cancer metastasis in an in vivo mouse xenograft model.27
In this current study, we generated an in silico library of analogs of compound 1a to find
a more potent compound for potential clinical application. After screening for optimal DOCKing
scores in silico, the top 14 compounds were synthesized and tested in a 2ꢀD collagen dot
migration assay. A high nanomolar inhibitor that maintained MMP specificity was identified
(compound 3c). Our data demonstrate that PEXꢀ9 inhibitor 3c selectively inhibits cell migration
induced by proMMPꢀ9 leading to reduced cell invasion predominately through disruption of
4
ACS Paragon Plus Environment

Page 5 of 44
ACS Chemical Biology
1
2
3
4
α4β1 integrin focal adhesion complexes. Thus PEXꢀ9 inhibitors rectify the abnormal migratory
phenotype associated with high MMPꢀ9 expression.
5
6
7
8
9
RESULTS
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Development of a second generation hit compound. We designed analogs of compound
1a based on possible derivative syntheses that utilized an amide and a thioether coupling.
Thus, three moieties were varied: the amine (R1), the linker ([CH2]n), and the
sulfhydrylpyrimidone (R2) (Scheme). To rationally prioritize the synthesis of compound 1a
analogs, the University of California San Francisco (UCSF) DOCK (6.7) algorithm was used to
rank various structures for predicted binding affinity. The proposed inhibitors were DOCKed to
the PEXꢀ9 domain pocket (PDB Code 1ITV) (Fig. 1A and 1B). This pocket is at the center of
the four blades of the hemopexin domain.
Fig. 1. PEX-9 inhibitor DOCKing to PEX-9 and binding site analysis. (A – B) PEXꢀ9 inhibitor derivatives were
DOCKed within the central cavity of the hemopexin domain of PEXꢀ9 MMPꢀ9 and hydrogen bonding was
5
ACS Paragon Plus Environment

ACS Chemical Biology
Page 6 of 44
1
2
3
visualized with dashed red lines. (C) A cluster of top hit compounds (2c, 3b, 3c, 3a, 4a, 4d) were found to share a
similar binding orientation. (D) The binding orientation of compound 1a (green) relative to compound 3c (red).
4
5
6
7
8
9
Amongst a library of 66 analogs, 14 (compounds 1a-d, 2a, 2c, 3a-d, 4a-e) were chosen
for synthesis based on superior DOCK algorithm scores (Scheme, Table S1). One motif that
demonstrated consistent improvement in predicted binding affinity was modification of the 6ꢀ
propylpyrimidone group to bicyclic pyrimidone systems.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chemical synthesis of PEXꢀ9 parental compound 1a and analogs. First, the early lead
compound 1a was reꢀsynthesized to confirm the chemical structure based on full
characterization, using the synthetic route shown in the Scheme. The synthesis commenced
with the reaction of aniline 5a with bromoacetic acid in the presence of EDC—HCl to afford
bromoacetanilide 6a as white solid in 67% yield. Subsequently, 6a was reacted with thiouracil
o
7 in aqueous NaOH and heated to 70 C for 48 hours to give compound 1a in 63% yield as a
white solid after recrystallization. In addition, newly designed compounds 1b and 1c were
synthesized using the same synthetic process (Scheme).
Closely related compound 1d bearing a benzimidazole group was synthesized using a
different route. In this synthesis, thiouracil 7 was converted to the corresponding thioacetic acid
8, followed by the amide coupling with 2ꢀaminobenzimidazole to give compound 1d in modest
yield. The attempted preparation of 2ꢀbromoacetylaminobenzimidazole, following the synthetic
route for compound 1a did not yield pure compound. Thus, we judged that the synthetic
strategy which coupled R1 last would enable a better modular synthesis of analogs.
With this modular synthesis protocol in hand, we undertook synthesis of the analogs
based on the original R1 moiety and two bicyclic sulfhydrylpyrimidones, as well as the linkers
with different lengths and a variety of aromatic R2 substituents. Synthesis of compounds 2a
and 2c followed the same strategy as that for compound 1d by converting 7 to the
corresponding thiobutanoic acid 9, followed by amide formation by coupling with 4ꢀsubstituted
6
ACS Paragon Plus Environment

Page 7 of 44
ACS Chemical Biology
1
2
3
Scheme. Synthesis of Inhibitors
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Reagents and conditions: (i) 2ꢀbromoacetic acid, EDC—HCl, DCM, room temp. 24 h: 67% for 6a, 47% for 6c, 28% for 6b;
(ii) NaOH, H2O, THF, 70 oC, 2 h: 63% for 1a, 35% for 1b, and 69% for 1c. (iii) Ethyl 2ꢀbromoacetate, K2CO3, MeOH, H2O,
rt, 15 min, 67%; (iv) LiOH, MeOH, H2O, rt, 2 days, 28% for 8, 66% for 9; (v) (a) EDC.HCl, DMAP, room temp. 24 h: 40%
for 1d, 41% for 2a, 39% for 2c, 52% for 3a, 46% for 3b, 70% for 3c, 22% for 3d, 36% for 4b, 46% for 4c, 42% for 4a,
26% for 4d; (b) CDI, room temp., 24 h, 45% for 4e. (vi) methyl 4ꢀbromobutanoate, K2CO3, MeOH, H2O, reflux: 47% for 9
(overnight), 73% for 12 (overnight), 81% for 14 (1 h); (vii) thiourea, MeONa, MeOH, reflux, overnight, 93%; (viii) NaOH,
EtOH, RT, overnight, 99%; %; (ix) LiOH, DMSO, H2O, room temp., 2h, 86%.
anilines 5a and 5c to give 2a and 2c in modest yields.
In a similar manner, compounds 3a-d were synthesized in moderate yields through the
coupling of anilines and 2ꢀaminobenzimidazole with thiobutanoic acid 12, prepared from 2ꢀ
sulfhydrylꢀ5,6,7,8ꢀtetrahydroquinazolinꢀ4(3H)ꢀone (11) as shown. Compound 11 was readily
prepared by reacting keto ester 10 with thiourea and a base. Compounds 4a-e were
synthesized in moderate yields using the same strategy except for starting from 2ꢀ
sulfhydryltetrahydroquinazolinone (13) and employed anilines and 2ꢀaminoimidazole for the
amide formation.
7
ACS Paragon Plus Environment

ACS Chemical Biology
Page 8 of 44
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 2. Structure activity relationship analysis of all synthesized PEX-9 inhibitor derivatives. (A)
Representative 20x magnification DAPI images of COSꢀ1 cells engineered to stably express either vector control
or proMMPꢀ9 cDNA vectors in a twoꢀdimensional migration assay after treatment with 50 ꢁM of 1f, 1a, or 3c.
Scale bar = 250 ꢁm. (B) Quantitative analysis of migrated COSꢀ1 vector control and proMMPꢀ9 expressing cells
after treatment with 50 ꢁM of inhibitors. (C) Tryptophan fluorescence emission of proMMPꢀ9 or proMMPꢀ9/PEX2
8
ACS Paragon Plus Environment

Page 9 of 44
ACS Chemical Biology
1
2
3
4
(ex 280 nm, em 330 nm) was monitored and inhibitor 3c titrated into the protein solution. A maximal λMax shift of 9
nm was observed after treatment with 25 ꢁM inhibitor. Solid line is a curve fit for Kd for compound 3c was
determined. (D) Kd for each inhibitor was plotted against percent inhibition of proMMP9ꢀmediated cell migration
for each inhibitor.
5
6
7
8
9
Structure activity relationship analysis of PEXꢀ9 derivative compounds. COSꢀ1 cells
were chosen for this set of experiments due to their low migratory capacity and complete lack
of MMPꢀ9 expression thereby minimizing potential artifacts when interpreting the data.
Overexpression of proMMPꢀ9 results in enhanced cell migration in a twoꢀdimensional dot
migration assay.27 This assay was initially used to test the efficacy of analogs at a single
concentration, 50 ꢂM. This concentration was used as a maximal cutoff because
concentrations above 50 ꢂM compound 1a did not increase inhibitor effectiveness in COSꢀ1
cells.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cells were mixed with type I collagen solution with or without compound, dotted onto a
96ꢀwell plate, and allowed to migrate for 18 hours. They were then fixed, stained and scored
for extent of migration (Fig. 2A and 2B). 2ꢀ(4ꢀOxoꢀ6ꢀpropylꢀ1,4ꢀdihydropyrimidinꢀ2ꢀylthio)ꢀ
N(methyl)ꢀphenylacetamide (1f) was used as a negative control.27 Compound 3c was the most
effective and was selected for further study. The migration assay was repeated in HT1080
fibrosarcoma cells, which endogenously express MMPꢀ9, and the same inhibition potency was
observed for both compound 1a and 3c relative to control groups (Supplemental Fig. 1A and
1B).
Compound affinity for PEXꢀ9 binding. A blue shift in proMMPꢀ9 tryptophan fluorescence
was monitored to determine the binding affinity of all derivative compounds as previously
described.27 The Kd for compound 3c binding to proMMPꢀ9 is 0.32 ꢂM (Fig. 2C), four times
tighter than parent compound 1a. Moreover, a clear correlation between binding affinity and
inhibition potency in the migration assay was observed (Fig 2D). Thus, the biophysical SAR is
maintained in the cellular assay and increased inhibition of cell migration is MMP9ꢀdependent
and due to optimization of chemical structure for binding the PEXꢀ9 domain.
9
ACS Paragon Plus Environment

ACS Chemical Biology
Page 10 of 44
1
2
3
4
5
6
7
8
PEXꢀ9 binding specificity. A chimera of proMMPꢀ9 and MMPꢀ2 in which the MMPꢀ9 PEX
domain was replaced by that of MMPꢀ2, (proMMPꢀ9/MMPꢀ2PEX)11 was tested for analog
binding. The tryptophans in MMPꢀ9 PEX domain are conserved in the MMPꢀ2 PEX domain.
Upon titration of compound 3c with the chimera, no shift in fluorescence was observed
suggesting that 3c does not bind to MMP2ꢀPEX (Fig. 2C).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As an additional test of binding selectivity, saturated transfer difference (STD) NMR was
performed with purified recombinant proMMPꢀ2 and proMMPꢀ9 protein. ProMMPꢀ2 and
proMMPꢀ9 are highly similar in structure and function, belonging to the same subfamily of
MMPs known as the type IV collagenases and/or gelatinases. Thus proMMPꢀ2 constitutes the
best control for testing the selectivity of the PEXꢀ9 inhibitor.28 Selective saturation of the 1.12
ppm resonance (proMMPꢀ2 and proMMPꢀ9) was used for STD NMR spectra. STD peaks were
observed in the presence of proMMPꢀ9, but not in the presence of proMMPꢀ2, indicating that
derivative compound 3c is specific for the PEXꢀ9 target (Supplemental Fig. 2A - 2F). Thus,
binding specificity is maintained upon increasing binding affinity.
PEXꢀ9 inhibitor target specificity. To test whether compound 3c is specific for inhibiting
only proMMP9ꢀmediated migration, a twoꢀdimensional dot migration assay was performed
using COSꢀ1 cells engineered to stably express vector control pQCXIP, proMMPꢀ2, proMMPꢀ
9, or proMMPꢀ14 cDNA vectors (Fig. 3A). An immunoblot was performed to confirm protein
expression and proper localization of each MMP (MMPꢀ2/ꢀ9 are secreted while MMPꢀ14 is
membrane bound) (Fig. 3B). As expected, expression of these three MMPs resulted in
enhanced cellular migration in COSꢀ1 cells.11 Treatment with either compound 1a or 3c only
inhibited proMMP9ꢀmediated migration: compound 3c was significantly more effective at
inhibiting migration (Fig. 3C).
10
ACS Paragon Plus Environment

Page 11 of 44
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 3. Determination of PEX-9 inhibitor specificity in the presence of homologous and non-homologous
MMPs. (A) COSꢀ1 cells transiently transfected with either vector control, proMMPꢀ2, proMMPꢀ9, or MMPꢀ14
cDNAs were treated with 50 ꢁM compound 3c and then used in a twoꢀdimensional dot migration assay.
Representative 10x DAPI magnification images were taken and are shown. Scale bar = 250 ꢁm. (B) MMP
expression and proper localization was confirmed by immunoblot analysis of cell lysates and protein precipitated
serumꢀfree medium. pQCXIP is the vector only control. (C) Quantification of inhibition proMMPꢀ9 mediated cell
migration by compounds 1a and 3c. Compound 1f was used as a negative control. (D) Assay of MMPꢀ9 catalytic
activity using a fluorogenic peptide degradation assay and activated MMPꢀ9. proMMPꢀ9 protein was activated
with 1 mM APMA overnight. 100 ng of activated MMPꢀ9 was incubated with 10 ꢀM fluorogenic peptide (excitationꢀ
320 nm; emissionꢀ 405 nm) in the presence of buffer, 50 ꢂM compound 1a, 3c, 1f, or 250 nM broadꢀspectrum
catalytic inhibitor Marimastat. Readings were performed every 10 minutes over the course of 2 hours. (E) Cell
viability was assayed with MTT. COSꢀ1 cells engineered to stably express either vector control or proMMPꢀ9
cDNA were treated for 24 hours with DMSO control, 50 ꢂM compound 3c or 100 nM staurosporine (STS) as a
positive control for acute cell toxicity.
11
ACS Paragon Plus Environment

ACS Chemical Biology
Page 12 of 44
1
2
3
4
Effect of PEXꢀ9 inhibitor on MMPꢀ9 catalytic activity. Recombinant proMMPꢀ9 protein
o
was purified and incubated overnight in a 37 C water bath in TNC buffer + 1 mM APMA to
5
6
7
8
9
artificially activate the protein.29, 30 The next day, the purified protein was incubated with 10 ꢂM
fluorogenic peptide (McaꢀProꢀLeuꢀGlyꢀLeuꢀDpaꢀAlaꢀArgꢀNH2), and catalytic activity was
monitored over 2 hours in the presence of compounds 1a, 3c, 1f, and marimastat, a broadꢀ
spectrum MMP catalytic inhibitor. No inhibition of the catalytic activity was observed with 1a or
3c, corroborating that PEXꢀ9 inhibitors do not modulate the catalytic activity of the protease. In
the presence of marimastat (Kd = 80 nM) no substrate cleavage was observed (Fig. 3D).
Therefore, inhibition of proMMP9ꢀmediated cell migration by compound 1a or 3c is not due to
loss of proteolytic activity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cell viability analysis. Compound 3c was evaluated for acute cytotoxicity. COSꢀ1 cells
stably expressing either vector control or proMMPꢀ9 cDNA were treated with 50 ꢂM 3c for 24
hours. As negative and positive controls, cells were treated with DMSO alone or 100 nM
staurosporine (STS), an apoptosis inducer, respectively.31 Significant cell death was only
observed after treatment with STS. A nonꢀstatistically significant 8% decrease in cell growth
occurred after treatment with 50 ꢂM of compound 3c (Fig. 3E). These experiments were
repeated in endogenous MMPꢀ9 expressing HT1080 fibrosarcoma cells, which were analyzed
for chronic toxicity (cells were reꢀtreated every three days for a total of nine days). No cell
death was observed at 100 ꢂM 3c (data not shown).
PEXꢀ9 inhibitors decrease MMPꢀ9 dimer formation and association with EGFR. Human
fibroscarcoma HT1080 cells were coꢀtransfected with MMPꢀ9 cDNAs tagged with either a Myc
or HA tag as previously reported.12 The next day, cells were switched to serum free media and
treated with 50 ꢂM of either compound 1a or 3c overnight. Using Myc antibody for
immunoprecipitation and HA antibody for immunoblotting, the formation of dimeric MMPꢀ9 was
monitored. Compound 3c significantly blocked MMPꢀ9 dimer formation as compared to control
12
ACS Paragon Plus Environment

Page 13 of 44
ACS Chemical Biology
1
2
3
4
5
6
7
8
compound 1f. In addition, inhibition of PEXꢀ9 with 1a or 3c reduced endogenous CD44 and
EGFR coꢀimmunoprecipitation (Supplemental Fig. 3A). Moreover, these PEXꢀ9 inhibitors
reduced phosphorylation of EGFRTyr1068 and its downstream targets AKTSer473 and
Erk1/2Thr202/Tyr204 (Supplemental Fig. 3B).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
PEXꢀ9 inhibitors disrupt MMPꢀ9 interaction with α4β1 integrin. HT1080 cells transiently
coꢀtransfected with tagged MMPꢀ9 cDNAs were prepared as described above.12 Inhibition of
PEXꢀ9 with 50 ꢂM of either compound 1a or 3c reduced endogenous α4β1 integrin coꢀ
immunoprecipitation (Fig. 4A). Inhibition of PEXꢀ9 with 50 ꢂM of either compound 1a or 3c
reduced SrcTyr418 phosphorylation in addition to phosphorylation of Src downstream targets
FAKTyr 576/577 and PAXTyr118 (Fig. 4B.)
13
ACS Paragon Plus Environment

ACS Chemical Biology
Page 14 of 44
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 4. Treatment with PEX-9 inhibitor results in decreased interaction between MMP-9 and α4β1 integrin
and attenuated downstream signaling. (A) CoꢀIP with HT1080 cells transiently overexpressing MMPꢀ9/MYC
cDNA was performed in the presence of 50 ꢁM compound 3c. MMPꢀ9/MYC was immunoprecipitated using an
antiꢀMyc antibody and served as an input control while α4 and β1 integrin interactions were probed via
immunoblot. A decrease in MMPꢀ9 interacting with both integrin subunits was observed as well as a (B) decrease
in SrcTyr418 and downstream target proteins FAKTyr 576/577/PAXTyr118 phosphorylation. (C) 60x confocal images were
captured and representative images of MMPꢀ9 (green) and each integrin subunit (red) were taken after treatment
of HT1080 cells with 50 ꢁM of either compound 1a or 3c compound 1f. Scale bar = 100 ꢁm. (D) 60x confocal
images were captured and representative images of pꢀFAKTyr 576/577 (red) and pꢀPAXTyr118 (green) were taken after
treatment of HT1080 cells with 50 ꢁM of either compound 1a or 3c compound 1f. Scale bar = 100 ꢁm.
Evaluation of MMPꢀ9 cell surface localization upon treatment with PEXꢀ9 inhibitors.
HT1080 cells were fixed in 4% paraformaldehyde then subsequently incubated with antiꢀMMPꢀ
9 in addition to antiꢀα4 and/or β1 integrin antibodies to monitor relative levels of localization of
14
ACS Paragon Plus Environment

Page 15 of 44
ACS Chemical Biology
1
2
3
4
5
6
7
8
these proteins at the cell surface. The cells were counterstained with DAPI nuclear dye.
Treatment with either compound 1a or 3c resulted in loss of MMPꢀ9 from the cell surface (Fig.
4C). However, treatment with either solvent control DMSO or negative control compound 1f did
not disrupt the localization of MMPꢀ9 on the cell surface.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Evaluation of focal adhesion complexes in the presence of PEXꢀ9 inhibitors. HT1080
cells were fixed after overnight treatment with 50 ꢂM PEXꢀ9 inhibitors, probed for either FAKTyr
576/577 or PAXTyr118 and counterstained with DAPI. A decrease in the formation of FAK and PAX
adhesion junctions, normally depicted as “large punctae”, was observed in cells after treatment
with PEXꢀ9 inhibitor compound 1a or 3c (Fig. 4D).
Validation of MMPꢀ9’s role in signaling. MMPꢀ9 expression was silenced using an
shRNA approach as previously described.12 HT1080 cell lysates were collected and analyzed
by immunoblot. Attenuation of MMPꢀ9 expression resulted in a decrease in the
576/577
phosphorylation of EGFRTyr1068
,
FAKTyr
,
PAXTyr118
,
SrcTyr418
,
AKTSer473
,
and
Erk1/2Thr202/Tyr204 (Supplemental Fig. 4A). These findings are corroborated by Kinexus
antibody microarray screening of phosphorylationꢀdependent signaling pathways in MMPꢀ9ꢀ
transfected COSꢀ1 cells12 in which Src, FAK, and PAX phosphorylation increase in proMMPꢀ9
overexpressing cells (Supplemental Figure 4B).
Treatment with PEXꢀ9 inhibitor prevents association of EGFR with CD44 and β1
integrin. HT1080 cells were treated with 50 ꢂM 3c, 1a, 1f or DMSO alone. Cell lysates were
collected and analyzed by a coꢀIP assay in which endogenous EGFR was captured with an
antiꢀEGFR antibody and served as an input control. CD44 and β1 integrin were individually
probed on the immunoblot. Decreased interaction between EGFRꢀCD44 and EGFRꢀβ1 integrin
was observed after treatment with 3c (Supplemental Fig. 5A).
Src activation is a PEXꢀ9 dependent process. NonꢀMMPꢀ9 expressing MCFꢀ7 breast
cancer cells were transiently transfected with either vector control pcDNA3.1, proMMPꢀ9, or
15
ACS Paragon Plus Environment

ACS Chemical Biology
Page 16 of 44
1
2
3
4
5
6
7
8
chimeric proMMPꢀ9/MMPꢀ2PEX cDNA constructs. Phosphorylation of SrcTyr418 was detected
using an immunoblotting assay. Increased SrcTyr418 phosphorylation was observed in proMMPꢀ
9 overexpressing cells. PEX domain swapping with MMPꢀ2 has no effect on SrcTyr418
phosphorylation similar to vector control pcDNA3.1 expressing cells (Supplemental Fig. 5B).
Effect of PEXꢀ9 inhibitors on cell adhesion. A 96ꢀwell plate was coated with a thin layer
of either collagen or fibronectin (5 ꢁg/ml in PBS) substrate before use. HT1080 cells were then
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
o
seeded onto wells and incubated at 37 C for 30 min in the presence of various inhibitors.
Wells were washed and cells were fixed in 4% paraformaldehyde and stained with DAPI.
Images of each 96ꢀwell were microscopically captured and counted for the number of cells still
adherent to the coated surface using automated computer software. Compound 1a or 3c
inhibited cellular adhesion whereas treatment with Marimastat, a broadꢀspectrum MMP
catalytic inhibitor, had no observable effect on cell attachment. Solvent DMSO and compound
1f also had no affect on cell adhesion (Supplemental Fig. 6A).
Effect of broadꢀspectrum MMP catalytic inhibitor marimastat on focal adhesion
complexes. HT1080 cells treated with 80 nM of broadꢀspectrum MMP inhibitor marimastat
were analyzed by immunofluorescence microscopy for both pꢀFAKTyr 576/577 and pꢀPAXTyr118
.
Treatment with marimastat did not prevent the formation of focal adhesion contact sites in
HT1080 cells (Supplemental Fig. 6B).
Treatment with PEXꢀ9 inhibitor prevents both angiogenesis and invasion in a
chorioallantoic membrane (CAM) assay. A CAM assay was carried out to evaluate angiogenic
and invasive potential of HT1080 cells after treatment with 3c. HT1080 cells were adsorbed
onto a gelatin sponge and implanted onto the surface of the chicken embryo CAM followed by
treatment with 100 ꢂM stock solution compound 1f, compound 1a or 3c (final drug
concentration is estimated to be 0.5 ꢂM). After a 4ꢀday incubation, neovascularization was
imaged and quantified. Treatment with either compound 1a or 3c significantly reduced the
16
ACS Paragon Plus Environment

Page 17 of 44
ACS Chemical Biology
1
2
3
angiogenic potential of HT1080 cells as compared to solvent DMSO and negative compound
4
5
1f control treated cells (Fig. 5A - 5B).
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 5. Compound 3c prevents angiogenesis and invasion through the basement membrane in a CAM
assay. (A) HT1080 cells were adsorbed onto a gelatin sponge, implanted atop the CAM, and were then
o
subsequently treated with DMSO control, 0.5 ꢂM compound 1f, 1a, or 3c. After 4 days of incubation at 37 C,
representative images of the angiogenic bed surrounding the gelatin sponge were taken. (B) Analysis and
quantification of the number of blood vessels in the area surrounding the implanted sponge. (C ꢀ D) The CAM was
inoculated with HT1080 GFP+ cells mixed with a type I collagen solution and treated with DMSO control, 0.5 ꢂM
o
compound 1f, 1a, or 3c. After 7 days of incubation at 37 C, CAM membranes were isolated, sectioned, H&E
stained, and imaged under 20x magnification.
HT1080 cells stably expressing GFP were mixed with a type I collagen solution (3
mg/mL) and seeded atop the CAM to ensure the cells localized to a specific area. Then they
were treated with a 100 ꢂM stock solution of compound 1f, compound 1a or 3c (final drug
17
ACS Paragon Plus Environment

ACS Chemical Biology
Page 18 of 44
1
2
3
4
5
6
7
8
concentration is estimated to be 0.5 ꢂM). The invasion of cancer cells through the epithelium
and basement membrane of the upper CAM into connecting tissue was examined by
hematoxylin and eosin staining (Fig. 5C) in addition to fluorescence microscopy for GFP+ cells
(Fig. 5D). Treatment of HT1080 cells with either compound 1a or 3c reduced invasion of CAM
tissue compared to treatment with either DMSO or compound 1f control.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DISCUSSION
MMPs are crucial for the degradation of ECM and blood vessel linings. Degradation
allows tumor cells to escape their primary location and enter the circulatory system where they
can travel to seed secondary tumors elsewhere in the body.32, 33 Therefore, targeting MMPs
with inhibitors is of interest for cancer drug development.34 Original inhibition attempts focused
on synthetic collagenꢀmimicking hydroxamate inhibitors that chelate the zinc ion in the active
site upon binding.35 These drugs largely failed in clinical trials due to lack of selectivity for
individual MMPs. Due to the specificity issues inherent in generating inhibitors of MMP
proteolytic activity, we developed inhibitors which target MMPꢀ9’s hemopexin domain. MMP
hemopexin domains share a common tertiary structure, yet the surface residues of MMP
hemopexin domains are distinct and one hemopexin domain cannot functionally substitute for
another.
Identification of a more potent PEXꢀ9 inhibitor. To identify compounds that bind more
tightly to the hemopexin domain of MMPꢀ9 than our original hit,27 structural diversity was
introduced at three positions and computational DOCKing utilized to prioritize synthesis and
testing of a small subset. The similarities between their predicted and measured binding
affinities indicate that the DOCKed structures may be used to infer important binding
modalities. Compounds that displayed the greatest inhibition of proMMPꢀ9 mediated migration
had longer linkers with three methylenes between the amide and the sulfide moieties and at
18
ACS Paragon Plus Environment

Page 19 of 44
ACS Chemical Biology
1
2
3
4
5
6
7
8
least 1 fluorine in the R1 group. A cluster of derivative compounds (2c, 3a-c, 4a, 4d), shared a
similar binding geometry within the hemopexin domain of MMPꢀ9 (Fig. 1C). In all the tightest
binding PEXꢀ9 inhibitor derivatives, the aryl fluorine atoms were found deep within the binding
pocket in contrast to the original hit 1a, in which the difluoromethoxyphenol moiety remained
solvent exposed. The STDꢀNMR data support the binding pose in which the aniline is deeply
buried in the pocket (Supplemental Fig 2A-F). In addition, fusion of a second ring to the
pyrimidone moiety to form a quinazolinone significantly improved inhibitor efficiency and
affinity. The length of the aliphatic chain between the two ring moieties is key for optimizing
binding interactions. Comparison of DOCKed compounds 1a and 3c highlights differences in
binding orientation that may contribute to the higher affinity of 3c (Fig. 1D).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Binding affinity for PEXꢀ9 correlates with inhibition of proMMP9ꢀmediated migration
(Fig. 2D) across our 16ꢀcompound series that spans 1.5 logs of binding affinity. These
structureꢀactivity results support that the mechanism of inhibition of proMMP9ꢀdependent cell
migration is through targeting PEXꢀ9 on the cell surface.
proMMPꢀ9 homodimerization is a critical step for interaction with CD44 at the cell
surface which results in decreased phosphorylation of EGFRTyr1068 and its downstream targets
AKTSer473 and Erk 1/2Thr202/Tyr204 phosphorylation.12 Compound 3c demonstrated enhanced
efficacy of blocking proMMPꢀ9 homodimerization and activation of downstream protein targets
(Fig 4, Supplemental Fig. 3). Moreover, our most potent compound, 3c, retained specificity
for a single MMP as demonstrated by the lack of inhibition in proMMPꢀ2 or proMMPꢀ14
mediated cell migration (Fig. 3C). Our structure activity correlation (Fig. 2D) and the specificity
of inhibition for proMMPꢀ9 mediated cell migration corroborate that compound 3c targets PEXꢀ
9 in vitro and in vivo and not a downstream signaling target common to other mediators of cell
migration.
19
ACS Paragon Plus Environment

ACS Chemical Biology
Page 20 of 44
1
2
3
4
5
6
7
8
Recently GarciaꢀPardo and coworkers reported that the central cavity of PEXꢀ9 interacts
with α4β1 integrin as a mechanism for crossꢀtalk between MMPs and integrin focal adhesion
formation.17, 18, 21 In turn, Src kinase interacts with the cytoplasmic tail of α4β1 integrin
subunits, which is critical for integrin function to promote and regulate focal adhesion
turnover.36ꢀ42 Interestingly, PEX9ꢀα4β1 integrin downstream signaling increases activation of
Fyn, a Src family kinase, and drives chronic lymphocytic leukemia B cell survival.17 Therefore,
we examined the consequences of binding 3c to PEXꢀ9 on α4β1 integrinꢀmediated Src
signaling.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Using a coꢀimmunoprecipitation approach, we found that PEXꢀ9 inhibitors interfere with
binding between proMMPꢀ9 and α4β1 integrin and can block crossꢀtalk signaling. This
disruption is accompanied by decreases in phosphorylation levels of SrcTyr418 and its
downstream substrates FAKTyr 576/577 and PAXTyr118 (Fig. 4B). Immunofluorescence imaging in
HT1080 cells further supported the idea that focal adhesion junctions (“large punctae”) were
lost upon treatment with PEXꢀ9 inhibitor (Fig. 4C-D). We also demonstrated that
overexpression of proMMPꢀ9 in MCFꢀ7 breast cancer cells enhanced Src activation.
Overexpression of proMMPꢀ9/MMP2ꢀPEX in MCFꢀ7 breast cancer cells or silencing of
endogenous MMPꢀ9 expression in HT1080 attenuated Src activation. Taken together, the data
suggest that the proMMPꢀ9 hemopexin domain is required for enhanced cellular adhesion and
migration.
Potential mechanism by which Src is activated. Our data are in agreement with the
literature showing that α4β1 integrin clustering is sufficient for activation of Src kinase.37, 43
However, in addition to integrin clustering, EGFR can directly activate Src kinase and EGFR
interacts with β1 integrin.44ꢀ46
It is proposed that proMMPꢀ9 forms a large charged surface for complex interactions
with multiple surface proteins involved in outsideꢀin signaling.47 The hemopexin domain of
20
ACS Paragon Plus Environment

Page 21 of 44
ACS Chemical Biology
1
2
3
4
5
6
7
8
proMMPꢀ9 may act as a scaffold to enhance clustering of α4β1 integrinꢀEGFRꢀCD44 at the cell
surface and result in aberrant Src activation.48 GarciaꢀPardo and coworkers demonstrated that
a peptide mimicking blade 1 of PEXꢀ9 inhibits interaction with integrin, and that a second
peptide mimicking blade 4 of PEXꢀ9 inhibits interaction with CD44.21, 22 We reasoned that our
inhibitor which binds directly to PEXꢀ9 could simultaneously disrupt both interactions. Indeed,
our coꢀimmunoprecipitation experiments demonstrate that addition of 3c selectively disrupts
proMMPꢀ9 interaction with both α4β1 integrin (Fig. 4) and CD44 (Supplemental Fig. S3).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 6. Overview of the mechanism by which MMP-9 regulates formation of focal adhesion junctions. (A)
During exocytosis and upon being secreted into the extracellular space, proMMPꢀ9 forms a large complex with
itself that can then act as a scaffold for promoting outsideꢀin signaling. In our model, PEXꢀ9 scaffolding promotes
the association of β1 integrinꢀEGFRꢀCD44. Interaction with CD44 results in enhanced EGFR activation in addition
to increased phosphorylation of its downstream targets AKT and Erk 1 + 2 (through the MAPK/Erk pathway).
21
ACS Paragon Plus Environment

ACS Chemical Biology
Page 22 of 44
1
2
3
4
5
6
7
8
While complexed to β1 integrin, EGFR then goes on to transactivate Src kinase after it is recruited to the α4
integrin subunit during the generation of a new focal adhesion contact site. After activation, Src can then directly
phosphorylate FAK currently associated to the β1 integrin subunit resulting in its maximal catalytic activity. This
active SrcꢀFAK complex can then bind and activate PAX resulting in a mature FAKꢀPAX complex necessary for
the formation of focal adhesion junctions. Formation of a complex between FAK and PAX results in final
translocation to ECMꢀintegrin junctions at the cell surface where they regulate cytoskeletal interactions resulting in
enhanced cellular adhesion, migration, and invasion of cancer cells. (B) Treatment with PEXꢀ9 inhibitor (depicted
as red triangle) prevents MMPꢀ9 scaffolding thereby preventing downstream signaling driven by EGFR activation.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Importantly, our inhibitor is specific for MMPꢀ9. Targeting the hemopexin domain to
investigate the role of MMPꢀ9 in promoting cellular adhesion has identified a connection to
integrin and Src in a complex adhesion network as depicted in Figure 6. There are currently
no effective cancer therapeutics designed to target adhesion receptors because cancer cells
can change the mechanism by which they migrate depending upon their environmental
constraints.49 Our in vivo CAM experiments demonstrate that we can inhibit two key
determinants of tumor cell metastatic potential, angiogenesis and invasion (Fig. 5), through
PEX domain targeting. Our small molecule inhibitor paves the way for designing new
therapeutic strategies for the treatment of metastatic disease and opens the door for the future
development of new technologies and tools for understanding the role of individual MMPs in
cancer progression. Development of PEX inhibitors for other MMPs will provide understanding
of their nonꢀcatalytic functions and a strategy for reducing undesirable offꢀtarget effects
previously observed in broadꢀspectrum MMP catalytic inhibitor clinical trials.
EXPERIMENTAL METHODS
Materials. Collagen type I (acetic acidꢀextracted native type I collagen from rat tail
tendon) was obtained from BD Bioscience Discovery Labware (Franklin Lakes, NJ).
Fibronectin solution was purchased from SigmaꢀAldrich (St. Louis, MO). Hoechst nuclear stain
was obtained from Invitrogen (Grand Island, NY). Rabbit anti–pꢀAKTSer473, antiꢀAKT (total),
antiꢀβ1 integrin, antiꢀpꢀErk1/2Thr202/Tyr204, antiꢀpꢀFAKTyr 576/577, antiꢀFAK (total), antiꢀpꢀPAXTyr118
,
antiꢀPAX (total) and antiꢀEGFR (total) antibodies were all purchased from Cell Signaling
22
ACS Paragon Plus Environment

Page 23 of 44
ACS Chemical Biology
1
2
3
4
5
6
7
Technology (Danvers, MA). Mouse antiꢀErk1/2 (total), antiꢀβꢀactin and anti–α/βꢀtubulin
antibodies were also purchased from Cell Signaling Technology (Danvers, MA). Rabbit antiꢀpꢀ
SrcTyr418 and antiꢀSrc (total) in addition to mouse antiꢀβ1 integrin were obtained from Abcam
8
9
(Cambridge, MA). Mouse antiꢀα4 integrin and antiꢀCD44 in addition to rabbit antiꢀα4 integrin
were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit antiꢀMMPꢀ9 was
purchased from Millipore (Darmstadt, Germany). Rabbit pꢀEGFRTyr1068 was obtained from
Enzo Life Sciences (Ann Arbor, MI). Rat antiꢀCD44 was obtained through BD Pharmingen
(San Diego, CA). Mouse antiꢀHA was purchased from InvitrogenꢀThermoFisher Scientific
(Waltham, MA). Rabbit antiꢀMYC and antiꢀHA in addition to horseradish peroxidase conjugated
antiꢀrabbit, antiꢀmouse, and antiꢀrat antibodies were obtained from Rockland
Immunochemicals (Gilbertsville, PA). Mouse antiꢀMYC was purchased from Roche Life
Sciences (Branford, CT). Alexa Fluor 488 and 568 antiꢀrabbit/ꢀmouse antibodies were
purchased from Molecular Probes, Life Technologies (Grand Island, NY). proMMPꢀ9 used in
kinetic experiments was purchased from Biolegends (San Diego, CA). McaꢀPLGLꢀDpaꢀARꢀNH2
Fluorogenic MMP Peptide Substrate was purchased from R&D Systems (Minneapolis, MN).
Chemicals were purchased from SigmaꢀAldrich, Fisher Scientific, and VWR International and
used as received or purified before use by standard methods. Tetrahydrofuran was freshly
distilled from sodium and benzophenone. Dichloromethane was also distilled immediately prior
to use under nitrogen from calcium hydride. 2ꢀ(4ꢀOxoꢀ6ꢀpropylꢀ1,4ꢀdihydropyrimidinꢀ2ꢀylthio)ꢀ
N(methyl)ꢀphenylacetamide (1f) was purchased from ChemBridge Corp. (San Diego, CA).
General Methods. 1H and 13C NMR spectra were measured on a Bruker 300, 400, 500,
or 700 MHz spectrometer. Melting points were measured on a ThomasꢀHoover capillary
melting point apparatus and are uncorrected. TLC was performed on Sorbent Technologies
aluminumꢀbacked Silica G TLC plates (Sorbent Technologies, 200ꢂm, 20 cm × 20 cm) and
column chromatography was carried out on silica gel 60 (Merck, 230 −400 mesh ASTM). Highꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
23
ACS Paragon Plus Environment

ACS Chemical Biology
Page 24 of 44
1
2
3
4
5
6
7
resolution mass spectrometry analysis was carried out on an Agilent LC−UV−TOF mass
spectrometer at the Mass Spectrometry facility of the Institute of Chemical Biology and Drug
Discovery, Stony Brook University.
8
9
Chemical Synthesis of PEX-9 Inhibitor Compounds and Characterization Data:
N-(4-Difluoromethoxyphenyl)-2-(4-oxo-6-propyl-1,4-dihydropyrimidin-2-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ylthio)acetamide (1a). To a solution of bromoacetic acid (170 mg, 1.25 mmol) and EDC—HCl
(239 mg, 1.25 mmol) in dichloromethane (6 mL) was added aniline (5a, 200 mg, 1.25 mmol)
and stirred at room temperature overnight. After the completion of the reaction, the reaction
mixture was diluted with dichloromethane and washed twice with 1 N hydrochloric acid, twice
with saturated solution of sodium bicarbonate and twice with brine. The organic layers were
collected, dried over magnesium sulfate, filtered and concentrated to afford an offꢀwhite crude
product. The crude product was recrystallized from hexanes/dichloromethane to give
bromoacetanilide 6a as white solid (227 mg, 65 % yield): 1H NMR (300 MHz, CDCl3) δ 4.27 (s,
2 H), 6.30ꢀ6.79 (s, 1 H), 7.21 (d, 2 H, J = 9Hz), 7.63 (d, 2 H, J = 9Hz).
To a solution of sodium hydroxide (7.2 mg, 0.18 mmol) in water (2 mL) was added 2ꢀ
hydrosulfanylꢀ4ꢀoxoꢀ6ꢀpropylꢀ1,4ꢀdihydropyrimidine (7) (30 mg, 0.18 mmol) and the mixture
was stirred till all of 1 were dissolved. To the mixture was added a solution of 6a (50 mg, 0.18
mmol) in tetrahydrofuran (2 mL) and heated at 60 °C for 2 h and then at 70 °C overnight. After
the completion of reaction white precipitate was observed which was collected on a filter and
o
recrystallized from ethanol to afford 1a as white solid (42 mg, 63 % yield): mp. 180 C
(decomp.); 1H NMR (700 MHz, DMSOꢀd6) δ 0.73 (t, J = 7.0 Hz, 3H), 1.51 (m, 2H), 2.32 (t, J =
7.0 Hz, 2H), 4.05 (s, 2H), 5.97 (s, 1H), 6.97–7.29 (m, 3H), 7.62 (d, J = 8.4 Hz, 2H), 10.44 (s,
1H); 13C NMR (175 MHz, DMSOꢀd6) δ 13.4, 20.6, 35.1, 38.2, 116.5.0 (t, J = 256.0 Hz), 119.5,
120.4, 136.4, 146.3, 165.9; 19F NMR (376 MHz, DMSOꢀd6) δ ꢀ81.6; HRMS (ESI+) calcd for
C16H18F2N3O3S [M + H]+ 370.1031, found 370.1039 (∆= ꢀ2.02 ppm).
24
ACS Paragon Plus Environment

Page 25 of 44
ACS Chemical Biology
1
2
3
In the same manner, 1b and 1c were synthesized.
4
5
2-(4-Oxo-6-propyl-1,4-dihydropyrimidin-2-ylthio)-N-phenylacetamide (1b). White
6
7
8
9
o
1
1
solid; 35% yield; mp. 170 C (decomposed); H NMR (700 MHz, DMSOꢀd6) δ H NMR (700
MHz, DMSOꢀd6) δ 0.83 (t, J = 7.0Hz, 3H), 1.46 ꢀ 1.56 (m, 2H), 2.31 (t, J = 7.0Hz, 2H), 4.04 (s,
2H), 5.90 (s, 1H), 7.04 (t, J = 7.7Hz, 1H), 7.30 (t, J = 7.7Hz, 2H), 7.56 (t, J = 7.7Hz, 2H), 10.33
(s, 1H), 12.67 (s, 1H). 13C NMR (175 MHz, DMSOꢀd6) δ 13.4, 20.5, 35.1, 38.5, 119.0, 123.3,
128.7, 139.0, 165.9; HRMS (ESI+) calcd for C15H18N3O2S [M + H]+ 304.1114, found 304.1113
(∆= 0.29 ppm).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N-(4-Fluorophenyl)-2-(4-oxo-6-propyl-1,4-dihydropyrimidin-2-ylthio)acetamide
(1c). White solid; 69 % yield; mp. 191ꢀ194 oC; 1H NMR (500 MHz, DMSOꢀd6) δ 0.73 (t, J = 7.0
Hz, 3H), 1.45 ꢀ 1.56 (m, 2H), 2.31 (t, J = 7.0Hz, 2H), 4.02 (s, 2H), 6.94 (s, 1H), 7.10 ꢀ 7.19 (m,
13
2H), 7.55 ꢀ 7.62 (m, 2H), 10.30 (s, 1H), 12.54 (s,1H). C NMR (500 MHz, DMSOꢀd6) δ 13.4,
20.5, 35.0, 38.5, 115.3 (d, J = 22.1Hz), 120.7 (d, J = 7.7Hz), 135.5 (d, J = 2.5Hz), 158.0 (d, J =
238.4Hz), 165.8; 19F NMR (376 MHz, DMSOꢀd6) δ ꢀ119.3; HRMS (ESI+) calcd for
C15H17FN3O2S [M + H]+ 322.1020, found 322.1020 (∆= ꢀ0.03 ppm).
N-(1H-benzo[d]imidazol-2-yl)-2-(4-oxo-6-propyl-1,4-dihydropyrimidin-2-
ylthio)acetamide (1d). To a solution of thiouracil 7 (1.70 g, 10.0 mmol) and potassium
carbonate (2.07 g, 15.0 mmol) in MeOH (10 mL) and water (20 mL) was added ethyl 2ꢀ
bromoacetate (1.42 g, 8.5 mmol). The mixture was stirred at room temperature for 15 min and
50 mL water was added. Then, the reaction mixture was extracted with ethyl acetate (5 × 80
mL). The organic layers were combined and washed with water (100 mL) and brine (100 mL),
dried over MgSO4, and the solvent was removed by rotary evaporator. The crude product was
purified by flash column chromatography on silica gel using MeOH/DCM as eluent to give ethyl
2ꢀ(4ꢀoxoꢀ6ꢀpropylꢀ1,4ꢀdihydropyrimidinꢀ2ꢀylthio)acetate as white solid (1.7 g, 67% yield): mp.
117ꢀ118 oC; 1H NMR (300 MHz, CDCl3) δ 0.93 (t, J = 7.2 Hz, 3H), 1.27 (t, J = 7.2 Hz, 3H), 1.59
25
ACS Paragon Plus Environment

ACS Chemical Biology
Page 26 of 44
1
2
3
4
5
6
7
8
– 1.65 (m, 2H), 2.43 (t, J = 7.2 Hz, 2H), 3.93 (s, 2H), 4.20 (q, J = 7.2 Hz, 2H), 6.05 (s, 1H),
13.17 (s, 1H). 13C NMR (175 MHz, CDCl3) δ 13.6, 14.1, 20.8, 32.8, 39.5, 61.9, 108.2, 158.9,
165.5, 168.2, 169.2; HRMS (ESI+) calcd for C11H17N2O3S [M + H]+ 257.0954, found 257.0953
(∆= 0.48 ppm).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To a solution of ethyl 2ꢀ((4ꢀoxoꢀ6ꢀpropylꢀ1,4ꢀdihydropyrimidinꢀ2ꢀyl)thio)acetate (1.20 g,
4.68 mmol) in MeOH and water was added lithium hydroxide (0.2 g, 9.36 mmol). The mixture
was stirred at room temperature for 2 days. The reaction mixture was adjusted to pH 1 by 1N
hydrochloric acid, and extracted with ethyl acetate (5 × 80 mL). The organic layers were
combined and dried over MgSO4, and the solvent was removed by rotary evaporator to afford
o
thioacetic acid 8 (0.30 g, 28%) as white solid: mp. 154ꢀ155 C; 1H NMR (700 MHz, DMSOꢀd6)
δ 0.86 (t, J = 7.0 Hz, 3H), 1.56 (m, 2H), 2.30 (t, J = 7.0 Hz, 2H), 3.47 (s, 2H), 5.79 (s, 1H). 13C
NMR (175 MHz, DMSOꢀd6) δ 13.6, 20.7, 36.1, 38.2, 107.4, 163.7, 164.3, 166.3, 171.2; HRMS
(ESI+) calcd for C9H13N2O3S [M + H]+ 229.0641, found 229.0644 (∆= ꢀ1.11 ppm).
To a mixture of 2ꢀ(6ꢀoxoꢀ4ꢀpropylꢀ1,6ꢀdihydropyrimidinꢀ2ꢀylthio)acetic acid (8) (228 mg, 1.0
mmol), 1Hꢀbenzo[d]imidazolꢀ2ꢀamine (5d, 146 mg, 1.1 mmol) and DMAP (134 mg, 1.1 mmol)
in DMF (3 mL) was added EDC.HCl (211 mg, 1.1 mmol), and the mixture was stirred at RT for
24 h. 50 mL water was added, and the resulting precipitate was filtered to give a solid, which
was recrystallized from 1,4ꢀdioxane to give compound 1d (136 mg, 40 %) as offꢀwhite solid:
o
1
mp. 217 C (decomposed); H NMR (700 MHz, DMSOꢀd6) δ 0.63 (t, J = 7.0 Hz, 3H), 1.50 –
1.40 (m, 2H), 2.27 (t, J = 7.0 Hz, 2H), 4.14 (s, 2H), 5.93 (s, 1H), 7.05ꢀ7.09 (m, 2H), 7.40ꢀ7.43
13
(m, 2H), 12.10 (s, 3H); C NMR (175 MHz, DMSOꢀd6) δ 13.3, 20.5, 34.5, 38.4, 106.5, 111.5,
114.1, 120.3, 121.0, 136.3, 146.7, 161.6, 162.3, 163.9, 167.7; HRMS (ESI+) calcd for
C16H18N5O2S [M + H]+ 344.1176, found 344.118 (∆= ꢀ1.1 ppm).
N-(4-Difluoromethoxyphenyl)-4-(4-oxo-6-propyl-1,4-dihydropyrimidin-2-
ylthio)butanamide (2a). To a solution of thiouracil 7 (0.85 g, 5 mmol) and potassium
26
ACS Paragon Plus Environment

Page 27 of 44
ACS Chemical Biology
1
2
3
4
5
6
7
8
carbonate (1.03g, 7.5 mmol) in methanol (5 mL) and water (10 mL) was added methyl 4ꢀ
bromobutanoate (1.36g, 7.5 mmol). The solution was heated to reflux overnight. The reaction
mixture was cooled to room temperature, 50 mL water was added, and the mixture was
extracted with ethyl acetate (3 × 50 mL). The organic layers were combined and washed with
water (50 mL) and brine (50 mL), dried over MgSO4, the solvent was removed by rotary
evaporator. The crude product was purified by flash column chromatography on silica gel using
MeOH/DCM as eluent to give methyl 4ꢀ(4ꢀoxoꢀ6ꢀpropylꢀ1,4ꢀdihydropyrimidinꢀ2ꢀylthio)butanoate
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
o
as white solid (0.64 g, 47% yield): mp. 129ꢀ130 C; 1H NMR (500 MHz, CDCl3) δ 0.95 (t, J =
7.0 Hz, 3H), 1.64ꢀ1.71(m, 2H), 2.06 (m, 2H), 2.46 (m, 4H), 3.24 (t, J = 7.0 Hz, 2H) , 3.68 (s,
3H) , 6.04 (s, 1H), 12.99 (s, 1H); 13C NMR (125 MHz, CDCl3) δ 13.8, 21.0, 24.7, 29.9, 32.7,
39.7, 51.8, 108.0, 160.2, 165.6, 169.5, 173.4; HRMS (ESI+) calcd for C12H19N2O3S [M + H]+
271.1111, found 271.111(∆= 0.51ppm).
To a solution of methyl 4ꢀ(4ꢀoxoꢀ6ꢀpropylꢀ1,4ꢀdihydropyrimidinꢀ2ꢀylthio)butanoate (500
mg, 1.84 mmol) in MeOH and water was added lithium hydroxide (88.1 mg, 3.68 mmol), and
the mixture was stirred at room temperature for 2 days. The mixture was adjusted to pH 1 by
1N hydrochloric acid to form white precipitate. The precipitate was collected on a filter to afford
4ꢀ(4ꢀoxoꢀ6ꢀpropylꢀ1,4ꢀdihydropyrimidinꢀ2ꢀylthio)butanoic acid 9 as white solid (308.0 mg, 66%
o
1
yield): mp. 139ꢀ141 C; H NMR (300 MHz, CDCl3) δ 0.95 (t, J = 7.3 Hz, 3H), 1.68 (m, 2H),
2.01ꢀ2.14(m, 2H), 2.49 (m, 4H), 3.28 (t, J = 6.5 Hz, 2H), 6.04 (s, 1H); 13C NMR (175 MHz,
CDCl3) δ 13.8, 21.1, 24.5, 29.9, 32.8, 39.7, 107.7, 160.3, 165.9, 170.2, 178.5; HRMS (ESI+)
calcd for C11H17N2O3S [M + H]+ 257.0954, found 257.0958(∆= ꢀ1.33ppm).
To a mixture of 9 (190 mg, 0.74 mmol), 4ꢀ(difluoromethoxy)aniline (5a, 130.5 mg, 0.82
mmol) and 4ꢀ(dimethylamino)pyridine (DMAP) (99.4 mg, 0.81 mmol) in dimethylformamide
(DMF) (3 mL) was added N’ꢀethylcarbodiimide hydrochloride (EDC—HCl (156 mg, 0.81 mmol),
the mixture was stirred at room temperature for 1 h. Then, 100 mL water was added to the
27
ACS Paragon Plus Environment

ACS Chemical Biology
Page 28 of 44
1
2
3
4
5
6
7
reaction mixture to form offꢀwhite precipitate, which was collected on a filter to give an offꢀwhite
solid. The solid was dissolved in ethyl acetate and dried over MgSO4, filtered, and the solvent
was removed by rotary evaporator. The crude product was purified by recrystallization from
8
9
o
1
ethyl acetate/hexane to afford 2a as white solid (120.4 mg, 41% yield): mp. 174ꢀ177 C; H
NMR (500 MHz, CDCl3) δ 0.92 (t, J = 7.5 Hz, 3H), 1.64 (m, 2H), 2.06 – 2.22 (m, 2H), 2.42 (t, J
= 7.5 Hz, 2H), 2.49 (t, J = 6.9 Hz, 2H), 3.31 (t, J = 5.7 Hz, 2H), 6.01 (s, 1H), 6.45 (t, J = 74.0
Hz, 1H), 7.07 (d, J = 8.6 Hz, 2H), 7.55 (d, J = 8.6 Hz, 2H), 7.92 (s, 1H), 12.69 (s, 1H). 13C NMR
(125 MHz, CDCl3) δ 13.8, 21.1, 24.8, 30.2, 35.5, 39.7, 107.8, 116.1(t, J = 259 Hz), 120.6,
121.3, 135.6, 147.3, 160.2, 165.5, 169.9, 170.5; 19F NMR (376 MHz, CDCl3) δ ꢀ80.7; HRMS
(ESI+) calcd for C18H22F2N3O3S [M + H]+ 398.1344, found 398.1345 (∆ = ꢀ0.03 ppm).
In the same manner, compound 2c was synthesized.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N-(4-Fluorophenyl)-4-(4-oxo-6-propyl-1,4-dihydropyrimidin-2-ylthio)butanamide
(2c). White solid; 39% yield. mp. 180ꢀ182 oC; 1H NMR (300 MHz, CDCl3) δ 0.83ꢀ0.96 (t, J = 7.3
Hz, 3H), 1.57ꢀ1.69 (m, 2H), 2.05ꢀ2.21 (m, 2H), 2.33ꢀ2.55 (m, 4H), 3.31 (t, J = 6.3 Hz, 2H), 6.02
(s, 1H), 7.00 (t, J = 8.7 Hz, 2H), 7.49 (m, 2H), 7.64 (s, 1H), 12.45 (s, 1H). 13C NMR (175 MHz,
DMSOꢀd6) δ 13.4, 20.5, 24.8, 29.2, 34.9, 38.5, 115.2 (d, J = 22.0Hz), 120.7 (d, J = 7.7Hz),
19
135.7 (d, J = 2.3Hz), 157.1, 158.5, 170.3; F NMR (376 MHz, CDCl3) δ ꢀ119.7; HRMS (ESI+)
calcd for C17H21FN3O2S [M + H]+ 350.1333, found 350.134 (∆ = ꢀ2.12ppm).
4-(4-Oxo-3,4,5,6,7,8-hexahydroquinazolin-2-ylthio)-N-phenylbutanamide (3b). Ethyl 2ꢀ
oxocyclohexaneꢀ1ꢀcarboxylate 10 (1.70 g, 10.0 mmol) and thiourea (1.52 g, 20.0 mmol) were
added to a sodium methoxide solution (0.58 g sodium in 20 mL methanol), and the mixture
was refluxed overnight. The solvent was removed and the residue was dissolved in hot water
(50 mL). Glacial acetic acid (6 mL) was added dropwise to make the solution acidic until a
white solid was formed. The precipitate was collected by a filter and washed successively with
saturated sodium bicarbonate (3 × 10 mL) and water (3 × 10 mL), and recrystallized from ethyl
28
ACS Paragon Plus Environment

Page 29 of 44
ACS Chemical Biology
1
2
3
4
5
6
7
8
acetate (30 mL) to afford 2ꢀsulfhydrylꢀ5,6,7,8ꢀtetrahydroquinazolinꢀ4(3H)ꢀone (11) (1.70 g,
93%) as white solid: mp. > 230 oC; 1H NMR (700 MHz, DMSOꢀd6) δ 1.53ꢀ1.76 (m, 4H), 2.16 (t,
J = 6.0 Hz, 2H), 2.36 (t, J = 6.0 Hz, 2H), 12.09 (s, 1H), 12.25 (s, 1 H); 13C NMR (175 MHz,
DMSOꢀd6) δ 20.5, 20.7, 20.9, 25.5, 111.6, 149.5, 161.3, 173.9; HRMS (ESI+) calcd for
C8H11N2OS [M + H]+ 183.0587, found 183.0587 (∆= ꢀ0.44 ppm).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To a solution of 11 (1.59 g, 8.72 mmol) in anhydrous methanol (15 mL) was added methyl
4ꢀbromobutanoate (1.72 g, 9.55 mmol) and potassium carbonate (1.31 g, 9.48 mmol). The
reaction mixture was heated to reflux for overnight. After cooled to room temperature, the
reaction mixture was concentrated in vacuo to remove the solvent. To the residue water (50
mL) was added, and extracted with dichloromethane (4 × 40 mL). The combined organic
phases were washed with brine (5 mL), dried over MgSO4, and concentrated in vacuo to give a
crude product, which was purified by recrystallization from ethyl acetate to afford methyl 4ꢀ(4ꢀ
oxoꢀ3,4,5,6,7,8ꢀhexahydroquinazolinꢀ2ꢀylthio)butanoate (1.80 g, 73%) as white solid: mp. 149
– 150 °C; 1H NMR (700 MHz, CDCl3) δ 1.69 ꢀ 1.79 (m, 4H), 2.01 ꢀ 2.06 (m, 2H), 2.47 (m, 4H),
13
2.59 (t, J = 7.0 Hz, 2H), 3.22 (t, J = 7.0 Hz, 2H), 3.67 (s, 3H); C NMR (175 MHz, CDCl3) δ
21.7, 21.8, 22.2, 24.7, 29.8, 31.6, 32.7, 51.8, 117.6, 156.4, 162.0, 165.0, 173.5; HRMS (ESI+)
calcd for C13H19N2O3S [M + H]+ 283.1111, found 283.1117 (∆ = ꢀ2.09 ppm).
To a mixture of 2 N sodium hydroxide (1.51 g in 19 mL water) and ethanol (22 mL) was
added methyl 4ꢀ(4ꢀoxoꢀ3,4,5,6,7,8ꢀhexahydroquinazolinꢀ2ꢀylthio)butanoate (1.34 g, 4.75 mmol)
at room temperature. The mixture was stirred for overnight at room temperature. The reaction
mixture was adjusted to pH 5 with 1N hydrochloric acid, and extracted with dichloromethane (5
× 40 mL). The combined organic phases were dried over MgSO4, and concentrated in vacuo to
give a residue, which was recrystallized from dichloromethane to afford 4ꢀ(4ꢀoxoꢀ3,4,5,6,7,8ꢀ
hexahydroquinazolinꢀ2ꢀylthio)butanoic acid 12 (1.26 g, 99% yield) as white solid: mp. 176 –
29
ACS Paragon Plus Environment