Development of a Non-Hydroxamate Dual Matrix Metalloproteinase (MMP)-7/-13 Inhibitor

Matrix metalloproteinase 7 (MMP-7) is a member of the MMP superfamily and is able to degrade extracellular matrix proteins such as casein, gelatin, fibronectin and proteoglycan. MMP-7 is a validated target for the development of small molecule drugs against cancer. MMP-13 is within the enzyme class the most efficient contributor to type II collagen degeneration and is a validated target in arthritis and cancer. We have developed the dual MMP-7/-13 inhibitor ZHAWOC6941 with IC50-values of 2.2 μM (MMP-7) and 1.2 μM (MMP-13) that is selective over a broad range of MMP isoforms. It spares MMP-1, -2, -3, -8, -9, -12 and -14, making it a valuable modulator for targeted polypharmacology approaches.


Introduction
Matrix metalloproteinases (MMPs) are a family of calcium-and zinc-dependent endopeptidases able to metabolize components of the extracellular matrix (ECM) [1]. In healthy organisms, the activity of MMPs is strongly regulated by the tissue inhibitors of metalloproteinases (TIMPs) [2]. An imbalance in this network can lead to a series of serious diseases including, but not exclusively, different forms of cancer or arthritis [3][4][5][6][7][8]. The enzyme class has been in the focus of the pharmaceutical industry for decades since their first description by Gross and Lapiere [9,10]. Many different MMP inhibitors are known to date, demonstrating a range of potency and selectivity [11][12][13][14][15][16][17][18][19][20][21]. Early inhibitors of the target family incorporated a hydroxamic acid moiety as a strong metal chelating group interacting with the catalytic zinc which is conserved in all MMP isoforms [22]. This led to potent inhibitors, which however did not display satisfying selectivity profiles [23]. In clinical trials those inhibitors failed due to painful side effects such as the joint-stiffening musculoskeletal syndrome (MSS) [24,25]. Further development led to more sophisticated inhibitors with superior selectivity profiles compared to hydroxamate derivatives. For some MMPs very selective inhibitors are available nowadays. For example MMP-13, the key player in collagen degradation and a valid target for arthritis and cancer [7,26], can be inhibited selectively and with high affinity with a variety of ligands [27][28][29].
Matrix metalloproteinase 7 (MMP-7) is a MMP-family member that differs from most of the other isoforms because it lacks the haemopexin-like domain, found in all MMPs except for MMP-7, MMP-23 and MMP-26 [4]. It is capable of activating the pro-forms of MMP-2 and MMP-9 [30]. Overall and Kleifeld have reviewed the role of MMPs as drug targets and anti-targets in relation to cancer therapy [31]. Several investigations indicate that MMP-7 is a validated drug target related to cancer. It is associated with prostate cancer [32,33], tumor proliferation [34], invasion of ovarian cancer [30], gastric cancer [35] as well as colorectal cancer [36]. In mouse models the administration of inhibitors addressing MMP-7 reduced the number of intestinal polyps [37]. Those findings suggest that inhibitors of MMP-7 that spare other MMP isoforms could be promising compounds for the treatment of the diseases.
Compounds 1 (batimastat) and 2 (PDB code: TQJ) ( Figure 1) are potent MMP inhibitors of the hydroxamate class that inhibit MMP-7 with IC 50 -values of 6 nM (1) [38] and 79 nM (2) [39] but also inhibit other members of the target class with the same potency leading to a lack in selectivity [38][39][40]. The non-hydroxamate based compounds 3 (PDB code: TQI) and 4 (PDB code: RSS) are weaker inhibitors with IC 50 -values of 10 µM (3) [39] and 850 nM (4) [41]. Compound 3 shows 10-fold selectivity over MMP-1 and MMP-14 [39]. inhibitors of MMP-7 that spare other MMP isoforms could be promising compounds for the treatment of the diseases. Compounds 1 (batimastat) and 2 (PDB code: TQJ) ( Figure 1) are potent MMP inhibitors of the hydroxamate class that inhibit MMP-7 with IC50-values of 6 nM (1) [38] and 79 nM (2) [39] but also inhibit other members of the target class with the same potency leading to a lack in selectivity [38][39][40]. The non-hydroxamate based compounds 3 (PDB code: TQI) and 4 (PDB code: RSS) are weaker inhibitors with IC50-values of 10 μM (3) [39] and 850 nM (4) [41]. Compound 3 shows 10-fold selectivity over MMP-1 and MMP-14 [39]. As displayed in Figure 2, all co-crystallized MMP-7 inhibitors found in the Protein Data Bank populate the active site, but only 2 (PDB 2Y6D [39], green) penetrates the S1' channel, which is responsible for selective binding among different subtypes in this enzyme class [42].  As displayed in Figure 2, all co-crystallized MMP-7 inhibitors found in the Protein Data Bank populate the active site, but only 2 (PDB 2Y6D [39], green) penetrates the S 1 ' channel, which is responsible for selective binding among different subtypes in this enzyme class [42]. inhibitors of MMP-7 that spare other MMP isoforms could be promising compounds for the treatment of the diseases. Compounds 1 (batimastat) and 2 (PDB code: TQJ) ( Figure 1) are potent MMP inhibitors of the hydroxamate class that inhibit MMP-7 with IC50-values of 6 nM (1) [38] and 79 nM (2) [39] but also inhibit other members of the target class with the same potency leading to a lack in selectivity [38][39][40]. The non-hydroxamate based compounds 3 (PDB code: TQI) and 4 (PDB code: RSS) are weaker inhibitors with IC50-values of 10 μM (3) [39] and 850 nM (4) [41]. Compound 3 shows 10-fold selectivity over MMP-1 and MMP-14 [39]. As displayed in Figure 2, all co-crystallized MMP-7 inhibitors found in the Protein Data Bank populate the active site, but only 2 (PDB 2Y6D [39], green) penetrates the S1' channel, which is responsible for selective binding among different subtypes in this enzyme class [42].  Work conducted previously in our lab led to the potent (IC 50 = 6 nM) and selective MMP-13 inhibitor 5 ( Figure 3) that was selective over all tested subtypes of the target class [43]. Work conducted previously in our lab led to the potent (IC50 = 6 nM) and selective MMP-13 inhibitor 5 ( Figure 3) that was selective over all tested subtypes of the target class [43]. Herein, we present an approach that enabled us to modify the characteristics of the MMP-13 inhibitor 5 towards a dual MMP-7/-13 inhibitor, while conserving the selectivity profile over other MMP isoforms.

Strategy of Inhibitor Development
Compound 5 was originally designed as an MMP-13 inhibitor and displayed very high potency and selectivity against the target enzyme with an IC50-value of 6 nM.
The selectivity profile of 5 (Table 1) demonstrated high selectivity against all examined MMPs. Nevertheless, initial inhibitory activity against MMP-7 was detected at an inhibitor concentration of 10 μM (Table 1). Consecutive dose dependent measurements revealed an IC50-value of 15.7 μM against MMP-7 resulting in a >2600 fold selectivity for MMP-13 over MMP-7. This finding motivated us to investigate structural modifications of 5 with the intention of improving the affinity against MMP-7 while maintaining the appealing selectivity profile over MMP family members being anti-targets in cancer therapy such as MMP (-3, -8, -9, -12 and -14) [31]. The initial scaffold was modified at the positions indicated by R1 and R2 in Figure 4 to probe the effects of structural changes on the enzymatic activity of MMP-7 and the selectivity profile over other MMP isoforms. At position R1 we examined the initial para fluorinated benzyl residue along with a non-fluorinated benzyl and a methyl substituent with the aim of probing the influence of electronic properties and size of the moiety tolerated for optimal interactions. As the S1' pocket of MMP-7 is smaller than in MMP-13, a smaller residue in this area is hypothesized to rather fit to the limited space in MMP-7. At the R2 position, we varied the length of the aliphatic linker between the phenolic oxygen and the carboxylic acid head group from 1-9 CH2 entities for the identification of the optimal chain length for the inhibition of MMP-7.  Herein, we present an approach that enabled us to modify the characteristics of the MMP-13 inhibitor 5 towards a dual MMP-7/-13 inhibitor, while conserving the selectivity profile over other MMP isoforms.

Strategy of Inhibitor Development
Compound 5 was originally designed as an MMP-13 inhibitor and displayed very high potency and selectivity against the target enzyme with an IC 50 -value of 6 nM.
The selectivity profile of 5 (Table 1) demonstrated high selectivity against all examined MMPs. Nevertheless, initial inhibitory activity against MMP-7 was detected at an inhibitor concentration of 10 µM (Table 1). Consecutive dose dependent measurements revealed an IC 50 -value of 15.7 µM against MMP-7 resulting in a >2600 fold selectivity for MMP-13 over MMP-7. This finding motivated us to investigate structural modifications of 5 with the intention of improving the affinity against MMP-7 while maintaining the appealing selectivity profile over MMP family members being anti-targets in cancer therapy such as MMP (-3, -8, -9, -12 and -14) [31]. The initial scaffold was modified at the positions indicated by R 1 and R 2 in Figure 4 to probe the effects of structural changes on the enzymatic activity of MMP-7 and the selectivity profile over other MMP isoforms. At position R 1 we examined the initial para fluorinated benzyl residue along with a non-fluorinated benzyl and a methyl substituent with the aim of probing the influence of electronic properties and size of the moiety tolerated for optimal interactions. As the S 1 ' pocket of MMP-7 is smaller than in MMP-13, a smaller residue in this area is hypothesized to rather fit to the limited space in MMP-7. At the R 2 position, we varied the length of the aliphatic linker between the phenolic oxygen and the carboxylic acid head group from 1-9 CH 2 entities for the identification of the optimal chain length for the inhibition of MMP-7. Work conducted previously in our lab led to the potent (IC50 = 6 nM) and selective MMP-13 inhibitor 5 ( Figure 3) that was selective over all tested subtypes of the target class [43]. Herein, we present an approach that enabled us to modify the characteristics of the MMP-13 inhibitor 5 towards a dual MMP-7/-13 inhibitor, while conserving the selectivity profile over other MMP isoforms.

Strategy of Inhibitor Development
Compound 5 was originally designed as an MMP-13 inhibitor and displayed very high potency and selectivity against the target enzyme with an IC50-value of 6 nM.
The selectivity profile of 5 (Table 1) demonstrated high selectivity against all examined MMPs. Nevertheless, initial inhibitory activity against MMP-7 was detected at an inhibitor concentration of 10 μM (Table 1). Consecutive dose dependent measurements revealed an IC50-value of 15.7 μM against MMP-7 resulting in a >2600 fold selectivity for MMP-13 over MMP-7. This finding motivated us to investigate structural modifications of 5 with the intention of improving the affinity against MMP-7 while maintaining the appealing selectivity profile over MMP family members being anti-targets in cancer therapy such as MMP (-3, -8, -9, -12 and -14) [31]. The initial scaffold was modified at the positions indicated by R1 and R2 in Figure 4 to probe the effects of structural changes on the enzymatic activity of MMP-7 and the selectivity profile over other MMP isoforms. At position R1 we examined the initial para fluorinated benzyl residue along with a non-fluorinated benzyl and a methyl substituent with the aim of probing the influence of electronic properties and size of the moiety tolerated for optimal interactions. As the S1' pocket of MMP-7 is smaller than in MMP-13, a smaller residue in this area is hypothesized to rather fit to the limited space in MMP-7. At the R2 position, we varied the length of the aliphatic linker between the phenolic oxygen and the carboxylic acid head group from 1-9 CH2 entities for the identification of the optimal chain length for the inhibition of MMP-7.

Chemistry
The synthetic routes to the intermediates 9 and 11 are illustrated in Scheme 1. The right hand side fragment could be synthesized from the benzylated bromoalkylalcohol. In case this intermediate could not be purchased, it was synthesized through protection of the according bromoalkylalcohol with benzylbromide. A nucleophilic substitution reaction between the benzylated bromoalkylalcohols and the commercially available methyl 2-(4-hydroxyphenyl)acetate led to the intermediates 8a-i and by a consecutive saponification with potassium hydroxide to the building blocks 9a-i. The left hand side fragments 11a-c were synthesized by alkylation of the commercially available 4-aminophthalimide 10 with the corresponding alkyl halide.
As displayed in Scheme 2, the aniline derivatives 11a-c and the carboxylic acids 9a-i were coupled to form the intermediates 12a-p via the formation of the acid chloride. By debenzylation of the protected alcohol moieties with TMSI for 13a and hydrogen for 13b-p, and consecutive oxidation employing TEMPO, the final compounds 5 and 14a-o could be obtained as the free carboxylic acids in moderate to good yields. Complete analytical data of the synthesized compounds and IC 50 -curves are shown in the Supplementary Materials.

Biological Evaluation
The synthesized compounds were examined in in vitro assays to determine their inhibitory potential against MMP-7 and MMP-13. Table 2 depicts the structure activity relationship of the novel inhibitors.
Compound 14i could be identified as the most potent MMP-7 inhibitor within the series displaying an IC50-value of 2.2 μM ( Figure S1a). In combination with a remaining affinity towards MMP-13 (IC50value of 1.2 μM, Figure S1b), this qualifies 14i as a dual MMP-7/MMP-13 inhibitor in the low micromolar range.

Biological Evaluation
The synthesized compounds were examined in in vitro assays to determine their inhibitory potential against MMP-7 and MMP-13. Table 2 depicts the structure activity relationship of the novel inhibitors.
Compound 14i could be identified as the most potent MMP-7 inhibitor within the series displaying an IC 50 -value of 2.2 µM ( Figure S1a). In combination with a remaining affinity towards MMP-13 (IC 50 -value of 1.2 µM, Figure S1b), this qualifies 14i as a dual MMP-7/MMP-13 inhibitor in the low micromolar range.
Consecutively, we tested inhibitor 14i in single dose assays against a set of MMP isoforms in order to determine its selectivity profile. Table 3 shows the remaining enzymatic activity at an inhibitor concentration of 10 µM.
The remaining enzymatic activities in the range of >70-100% at an inhibitor concentration of 10 µM indicate IC 50 -values against those enzymes higher than 10 µM.   Table 3. Selectivity profile of 14i against a variety of MMPs 1 .

Discussion
The initial compound 5 was modified with the aim of improving its inhibitory potential against MMP-7. A first modification led to compound 14a, here the para-fluorobenzyl residue has been replaced by a methyl group resulting in a total loss of inhibition. This indicated that an aromatic moiety is important for protein ligand interactions such as π-stacking to aromatic amino acids in this area of the receptor. Therefore we decided to conserve a benzyl group in this area of the molecule. We crafted two series of compounds one with the original fluorinated benzyl residue and the second with a non-fluorinated benzyl moiety. In both series the length of the aliphatic linker between the phenolic oxygen and the carboxylic acid head group was varied. The series containing a fluorine atom was equipped with different linkers ranging from 2 to 7 CH 2 entities in length. Here, a tendency could be observed that longer linkers corresponded to better inhibition of MMP-7 with an optimum at six carbons in 14n, which demonstrated remaining enzymatic activity of 47.5% at 10 µM inhibitor concentration. A similar trend could be observed for the ensemble lacking the fluorine atom. A chain length between one and four carbons was not tolerated and resulted in very low to no inhibition. With longer linkers, better inhibition could be achieved with an optimum at 8 CH 2 entities. The most potent inhibitor of the series 14i displayed an IC 50 -value of 2.2 µM against MMP-7 and 1.2 µM against MMP-13. Compared to MMP-7 inhibitors in literature, our compound is with a low single digit micromolar IC 50 amongst the most potent non-hydroxamate compounds [39,41,44], and to the best of our knowledge unique in displaying selectivity over a wide range of other MMP isoforms [39,41].
We performed molecular docking experiments to examine potential binding modes of 14i with MMP-7 and MMP-13. The water molecules present in the used co-crystal structures (PDB: 2Y6D [39] for MMP-7 and 2OW9 [45] for MMP-13) were set to inactive with respect to the force field prior to the docking experiment to enable the ligand to populate all space available in the active site. The water molecule which is expected to populate the remaining coordination site at the zinc(II) ion ( Figure 5a) is not present in the co-crystal structure (2Y6D) used for the docking experiments as this is replaced by the zinc-chelating ligand within the complex.
For MMP-7 two low energy docking poses were found with significantly diverse binding modes. As visible in Figure 5a (the non zinc-binding mode), the ligand interacts with four amino acids of the receptor by the establishment of hydrogen bonds to Ala182, Ala184, Gly244 and Asp245. In this pose the ligand does not coordinate to the catalytic zinc(II) ion. The aliphatic linker between the scaffold and the carboxylic acid populates the hydrophobic S 1 ' channel and the phthalimide blocks the groove at the active site. The benzyl residue can be engaged in π-interactions with Tyr172 and Phe185. In the second pose, depicted in Figure 5c (the zinc-binding mode), the benzyl moiety populates the S 1 ' pocket and interacts with Tyr241 via a face to edge aromatic interaction.
The para-substituted phenyl ring and the aliphatic linker chain populate the cleft where the substrate is recognized and block the active site. The carboxylic acid head group ligates the catalytic zinc(II) ion, which is known to be a strong interaction leading to enhanced affinity between the ligand and the target enzyme. Both of the proposed binding modes qualify as reasonable explanations for the appealing affinity towards MMP-7 without the need of hydroxamic acids as zinc-binding fragments and can be used for further structure-based optimization of the inhibitor. Docking of 14i into MMP-13 ( Figure 5e) revealed a binding mode where the phthalimide oxygens and the amide oxygen form hydrogen bonds to the backbone amino acids Thr224, Thr226 and Met232 and the unsubstituted benzyl ring populates the S 1 '* selectivity loop, buried deep in the S 1 ' channel. The carboxylate head group chelates the catalytic zinc(II) ion and interacts with Ala167 and Glu202 by the formation of hydrogen bonds.

General Information
All reagents and solvents were purchased from Sigma Aldrich (Buchs, Switzerland), TCI (Zwijndrecht, Belgium) or Fluorochem (Hadfield, UK) and used as received. Solvents were stored over 4 Å molecular sieves. NMR spectra were recorded at 25 °C on an AVANCE III HD 500 One Bay

General Information
All reagents and solvents were purchased from Sigma Aldrich (Buchs, Switzerland), TCI (Zwijndrecht, Belgium) or Fluorochem (Hadfield, UK) and used as received. Solvents were stored over 4 Å molecular sieves. NMR spectra were recorded at 25 • C on an AVANCE III HD 500 One Bay spectrometer (Bruker, Fällanden, Switzerland) with a magnetic field of 11.75 T. For 1 H-NMR spectra a frequency of 500 MHz resulted. Chemical shifts are reported in ppm from tetramethylsilane as internal standard. Data are reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, quint. = quintet, br. = broad, m = multiplet), coupling constants (Hz), integration. For 13 C-NMR spectra a frequency of 125 MHz resulted. Chemical shifts are reported in ppm from tetramethylsilane as internal standard. The multiplicities of the signals were determined by DEPT measurements. Low-resolution mass spectrometry was performed on a MSQ Plus device (Thermo Scientific, Reinach, Switzerland). High-resolution mass spectrometry was performed on an 6530 Q-TOF (Agilent Technologies, Basel, Switzerland). NMR spectra, HRMS spectra and IC 50 curves can be found in the Supplementary Materials.

Chemistry
17 mmol) was added to a solution of 4-aminophthalimide 10 (1.00 g, 6.17 mmol) in dimethylformamide (30 mL) and the mixture was stirred at ambient temperature for 2 h. Iodomethane (0.88 g, 6.17 mmol) was added and it was stirred for another 18 h at the same temperature. Water (50 mL) and ethyl acetate (50 mL) was added and the resulting phases were separated. The organic phase was washed with brine, dried over sodium sulfate and concentrated in vacuum. Purification by chromatography on silica gel (Gradient: 0-100% ethyl acetate in cyclohexane) afforded the title compound 11a as a yellow solid (1.00 g, 92% yield): In analogy to ZHAWOC3444 the following derivatives were synthesized, employing the alkyl bromide instead of the alkyl iodide: In analogy to ZHAWOC7100 the following derivatives were synthesized:     13

In Silico Studies
Molecular modeling experiments were performed using the Molecular Operating Environment MOE 2015.10 from Chemical Computing Group (Montreal, QC, Canada). Co-crystal structures of MMP-7 are available from the Protein Data Bank. For the actual work PDB code: 2Y6D was selected for the computational studies. In MOE the pocket was prepared for the dockings via the Protonate 3D method applying the default values for temperature 300 K, pH 7 and salt 0.1. The ligands to be docked to the protein were imported from SD files to receive a MOE compatible molecular database. As the SD files did not contain 3D coordinates, they were generated directly using MOE rebuild3D with an RMSD gradient of 0.1. For docking experiments the Amber10:EHT force field [46,47] was used. The triangle matcher placement was applied with a rigid receptor. The docked poses were subsequently analyzed with respect to their scores and interactions with the target enzyme.

Conclusions
A highly potent and selective MMP-13 inhibitor was modified to obtain a dual MMP-7/-13 inhibitor with selectivity over a variety of MMP isoforms. We were able to modify the original molecule with a focus on gaining potency against MMP-7 while decreasing its potency against MMP-13. The IC 50 -value against MMP-7 could be improved from 15.7 µM to 2.2 µM by removing the fluorine atom from the benzylic ring and by elongating the aliphatic linker between the phenolic oxygen and the carboxylic acid head group from 4 to 8 CH 2 entities. Further improvements with respect to the inhibitor's potency are imaginable by rigidification of the flexible aliphatic linker to yield beneficial entropy terms for the ligand-enzyme complex, for example by the incorporation of non-saturated fragments such as alkenes or alkynes. The improvements towards MMP-7 inhibition decreased the potency on MMP-13 drastically from 6 nM to 1.2 µM resulting in a dual inhibitor in the low micromolar range equally potent against MMP-7 and MMP-13. To our knowledge, this inhibitor is the first of its kind that simultaneously inhibits the two validated drug targets MMP-7 and MMP-13 selectively. This is of utmost interest in polypharmacology, because here two or more targets are addressed at the same time to tackle one disease [48,49].
Supplementary Materials: Supplementary materials related to this article, including complete analytical data of the synthesized compounds and IC 50 -curves, are available online.