Introduction

Matrix metalloproteinases (MMPs) comprise a family of Zn(II)-dependent endopeptidases involved in the cleavage of extracellular proteins [1–3]. MMPs are secreted as zymogens and become activated by a variety of pathways through proteolytic cleavage of the propeptide domain by proteases, other MMPs, and reactive-oxygen species (ROS) [1, 3]. Owing to the ability of MMPs to cleave proteins in the extracellular matrix, overexpression and misregulation of MMPs has been associated with a variety of pathologic disorders, including arthritis, cancer, and cardiovascular disease [2, 4–6]. Of particular note is MMP activation associated with stroke, which is a leading cause of disablement and death [7–10]. In stroke, the inflammatory response induced by ischemia initiates the formation of ROS, which activates MMPs, and can lead to the breakdown of the blood–brain barrier, resulting in cell death and tissue damage [7–11]. The association of ROS with MMP activation after ischemia has generated interest in using MMP inhibitors (MMPi) to treat reperfusion injury associated with stroke and other cardiovascular ailments [12]. It has been shown that MMP inhibition with broad-based inhibitors upon the onset of stroke can reduce ischemic-related brain injury [9, 13, 14]. However, because MMPs have both beneficial and pathogenic roles with respect to normal physiologic function and disease progression, systemic inhibition of MMPs can lead to side effects such as musculoskeletal syndrome [1, 4, 5, 15–18]. Hence, there is an impetus to control both the spatial and temporal activity of these inhibitors [19].

To localize the activity of MMPi, a prodrug approach can be employed to develop latent forms of the inhibitor (i.e., proinhibitor) that are activated under specific, desired conditions. Prochelators activated by biological or chemical stimuli have been explored as a means to modulate metal ion chelation in various disease states [20–25]. By blocking the zinc-binding group (ZBG) of the inhibitor [which coordinates to the catalytic Zn(II) ion] with a stimulus-responsive protecting group, one can greatly attenuate or abolish the inhibitory activity of the MMPi. Under the appropriate biological stimulus, the proinhibitor can be activated, releasing the MMPi, thereby allowing for localized delivery and inhibition. In general, metalloenzyme proinhibitors have not been widely investigated, with only a few systems reported for MMPs [26–31]. To establish the use of ZBG protecting groups for the development of MMP proinhibitors, MMPi containing carbohydrate protecting groups that are cleaved in the presence of β-glucosidase were described [30]. Quotient IC50 values, which represent the ratio of the IC50 values of the MMPi prodrug (proMMPi) to IC50 values of the active inhibitor [32], were excellent at greater than 1,000 with MMP-8 [30]. Similarly, proMMPi based on concepts from prochelators and fluorescent probes sensitive to ROS [33, 34] were developed for potential use in ischemic injury [31]. Boronic ester protecting groups appended via a self-immolative linker to the ZBG of MMPi were reported by our group and showed rapid activation in the presence of H2O2, leading to the release of the free MMPi [31].

We sought to identify additional protecting groups to develop proMMPi that could be triggered in an ischemic setting by ROS. Fluorescent probes incorporating sulfonate ester protecting groups have been shown to be sensitive for detecting several ROS (hydrogen peroxide, superoxide anion) [35–37]. These studies prompted the use of sulfonate esters as selective protecting groups for proMMPi that could be cleaved under conditions of oxidative stress (e.g., ischemia). The work described herein demonstrates that various ZBGs could be protected with sulfonate esters, creating molecules that can be activated with H2O2. Four different ZBGs employing oxygen donor atoms were selected for this study. The ability of H2O2 to liberate the ZBGs was assessed, and suitable candidates were developed into full-length proMMPi that show increased potency against MMP-12 in the presence of H2O2.

Materials and methods

General

Starting materials and solvents were purchased from commercial suppliers (Sigma–Aldrich, Alfa Aesar, Fisher, and others) and used as received. 1H/13C NMR spectra were recorded at ambient temperature with a 400 or 500 MHz Varian Fourier transform NMR instrument or a 500 MHz JEOL instrument, located in the Department of Chemistry and Biochemistry at the University of California, San Diego. Mass spectra were obtained at the Molecular Mass Spectrometry Facility in the Department of Chemistry and Biochemistry at the University of California, San Diego. Elemental analyses was preformed by NuMega Resonance Labs, San Diego.

General procedure for the synthesis of sulfonate ester ZBGs

The ZBG compound was dissolved in pyridine on ice. To this was added the desired sulfonyl chloride. The reaction flask was removed from the ice bath and left stirring overnight under nitrogen while warming to room temperature. The pyridine was removed by rotary evaporation and the resulting oil was redissolved in dichloromethane and washed with 1 M HCl (approximately 30 mL), water, and brine. The organic layer was dried over MgSO4, filtered, and then concentrated via rotary evaporation. The product was purified on a silica gel column and eluted with 1% MeOH in dichloromethane unless otherwise noted.

2-Oxopyridin-1(2H)-yl benzenesulfonate

2-Hydroxypyridine-1-oxide (ZBG1; 1.0 g, 9.1 mmol) was reacted with benzenesulfonyl chloride (1.27 mL, 10.0 mmol) in 75 mL of pyridine to afford 2-oxopyridin-1(2H)-yl benzenesulfonate (PZBG-1a) in 77% yield (1.75 g, 7.0 mmol). 1H NMR (500 MHz, CDCl3) δ = 8.02 (d, J = 8.0 Hz, 2H), 7.75 (t, J = 8.0 Hz, 1H), 7.59 (m, 3H), 7.28 (dt, J 1 = 6.9 Hz, J 2 = 2.3 Hz, 1H), 6.52 (d, J = 9.8 Hz, 1H), 6.15 (t, J = 7.5 Hz, 1H). 13C NMR (100 MHz, CDCl3) δ = 157.0, 139.7, 137.1, 136.0, 133.8, 130.0, 129.5, 123.4, 105.4. Electrospray ionization mass spectrometry (ESI-MS)(+): m/z 252.01 [M + H]+, 273.95 [M + Na]+. Anal. calcd for C11H9NO4S: C, 52.58; H, 3.61; N, 5.57. Found: C, 52.21; H, 3.99; N, 5.44.

2-Oxopyridin-1(2H)-yl 4-methylbenzenesulfonate

ZBG1 (0.5 g, 4.5 mmol) was reacted with p-toluenesulfonyl chloride (2.57 g, 13.5 mmol) in 40 mL of pyridine to afford 2-oxopyridin-1(2H)-yl 4-methylbenzenesulfonate (PZBG-1b) in 89% yield (1.06 g, 4.0 mmol). 1H NMR [500 MHz, dimethyl sulfoxide (DMSO)-d6] δ = 7.82 (d, J = 8.6 Hz, 2H), 7.75 (dd, J1 = 7.5 Hz, J2 = 1.8 Hz, 1H), 7.49 (d, J = 8 Hz, 2H), 7.41 (dt, J1 = 7.5 Hz, J2 = 1.7 Hz, 1H), 6.48 (dd, J1 = 9.2 Hz, J2 = 1.8 Hz, 1H), 6.21 (dt, J1 = 7.5 Hz, J2 = 1.8 Hz, 1H), 2.42 (s, 3H, CH3). 13C NMR (100 MHz, CDCl3) δ = 156.6, 148.9, 141.2, 138.1, 131.1, 130.5, 130.0, 122.9, 106.1, 22.0. ESI-MS(+): m/z 266.10, [M + H]+, 287.99 [M + Na]+. Anal. calcd for C12H11NO4S: C, 54.33; H, 4.18; N, 5.28. Found: C, 54.24; H, 4.35; N, 5.24.

2-Oxopyridin-1(2H)-yl 4-nitrobenzenesulfonate

ZBG1 (0.5 g, 4.5 mmol) was reacted with 4-nitrobenzenesulfonyl chloride (3.0 g, 13.5 mmol) in 40 mL of pyridine to afford 2-oxopyridin-1(2H)-yl 4-nitrobenzenesulfonate (PZBG-1c) in 92% yield (1.23 g, 4.2 mmol). 1H NMR (500 MHz, CDCl3) δ = 8.42 (d, J = 8.6 Hz, 2H), 8.23 (d, J = 9.2 Hz, 2H), 7.65 (dd, J1 = 7.5 Hz, J2 = 1.8 Hz, 1H), 7.34 (dt, J1 = 9.2 Hz, J2 = 1.7 Hz, 1H), 6.52 (d, J = 9.2 Hz, 1H), 6.22 (t, J = 7.7 Hz, 1H). 13C NMR (100 MHz, DMSO) δ = 156.6, 152.3, 141.6, 139.4, 138.3, 131.8, 125.6, 122.8, 106.4. ESI-MS(+): m/z 297.28 [M + H]+, 319.02 [M + Na]+.

2-Oxopyridin-1(2H)-yl 2,4-dinitrobenzenesulfonate

ZBG1 (0.5 g, 4.5 mmol) was reacted with 2,4-dinitrobenzenesulfonyl chloride (1.32 g, 5.0 mmol) in 40 mL of pyridine to afford 2-oxopyridin-1(2H)-yl 2,4-dinitrobenzenesulfonate (PZBG-1d) in 31% yield (0.48 g, 1.4 mmol). 1H NMR (500 MHz, CDCl3) δ = 8.96 (d, J = 2.3 Hz, 1H), 8.42 (dd, J1 = 9.2 Hz, J2 = 2.3 Hz, 1H), 7.68 (dd, J1 = 6.9 Hz, J2 = 1.7 Hz, 1H), 7.50 (dt, J1 = 7.5 Hz, J2 = 2.3 Hz, 1H), 7.06 (d, J = 8.1 Hz, 1H), 6.82 (dd, J1 = 9.2 Hz, J2 = 1.8 Hz, 1H), 6.35 (dt, J1 = 6.9 Hz, J2 = 1.8 Hz, 1H). 13C NMR (100 MHz, CDCl3) δ = 157.0, 155.6, 140.5, 135.1, 129.6, 124.0, 122.9, 116.0, 106.6.

4-(((2-Oxopyridin-1(2H)-yl)oxy)sulfonyl) benzoic acid

ZBG1 (0.21 g, 1.9 mmol) was reacted with 4-(chlorosulfonyl) benzoic acid (0.62 g, 2.8 mmol) in 5 mL of pyridine. The solvent was evaporated, leaving a yellow oil. Addition of 5 mL of dichloromethane followed by the addition of 5 mL of ethyl acetate allowed for precipitation of 4-(((2-oxopyridin-1(2H)-yl)oxy)sulfonyl)benzoic acid (PZBG-1e) in 30% yield (0.17 g, 0.57 mmol). 1H NMR (400 MHz, DMSO) δ = 8.18 (d, J = 8.8 Hz, 2H), 8.08 (d, J = 8.4 Hz), 7.88 (dd, J1 = 7.6 Hz, J2 = 2 Hz), 7.45 (td, J1 = 8.2 Hz, J2 = 2 Hz, 1 H), 6.50 (dd, J1 = 9.2 Hz, J2 = 1.6 Hz, 1 H), 6.26 (td, J1 = 7 Hz, J2 = 1.6 Hz, 1 H). 13C NMR (100 MHz, CDCl3) δ = 166.5, 156.6, 141.5, 138.2, 137.9, 137.3, 131.2, 130.4, 122.8, 106.2. ESI-MS(−): m/z 294.26 [M − H]−. Anal. calcd. for C12H9NO6S: C, 48.81; H, 3.07; N, 4.74. Found: C, 48.91; H, 3.37; N 4.84.

2-Methyl-4-oxo-4H-pyran-3-yl benzenesulfonate

3-Hydroxy-2-methyl-4H-pyran-4-one (ZBG2; 1.0 g, 7.9 mmol) was reacted with benzenesulfonyl chloride (3.0 mL, 23.7 mmol) in 75 mL of pyridine to afford 2-methyl-4-oxo-4H-pyran-3-yl benzenesulfonate (PZBG-2a) in 71% yield (1.50 g, 5.6 mmol). 1H NMR (500 MHz, CDCl3) δ = 8.12 (d, J = 8.6 Hz, 2H), 7.69 (t, J = 7.5 Hz, 1H), 7.65 (d, J = 5.8 Hz, 1H), 7.58 (t, J = 8.0 Hz, 2H), 6.33 (d, J = 5.2 Hz, 1H), 2.46 (s, 3H, CH 3). 13C NMR (100 MHz, CDCl3) δ = 172.1, 163.1, 154.3, 138.4, 136.6, 134.7, 129.2, 129.0, 117.7, 16.3. ESI-MS(+): m/z 267.06 [M + H]+, 289.03 [M + Na]+.

2-Methyl-4-oxo-4H-pyran-3-yl 4-methylbenzenesulfonate

ZBG2 (0.5 g, 4.0 mmol) was reacted with p-toluenesulfonyl chloride (2.27 g, 11.9 mmol) in 40 mL of pyridine to afford 2-methyl-4-oxo-4H-pyran-3-yl 4-methylbenzenesulfonate (PZBG-2b) in 64% yield (0.71 g, 2.5 mmol). 1H NMR (500 MHz, CDCl3) δ = 7.99 (d, J = 8 Hz, 2H), 7.64 (d, J = 5.8 Hz, 1H), 7.37 (d, J = 8 Hz, 2H), 6.34 (d, J = 5.8 Hz, 1H), 2.46 (s, 3H, CH3), 2.45 (s, 3H, CH3). 13C NMR (100 MHz, CDCl3) δ = 127.2, 163.2, 154.2, 145.8, 138.4, 133.6, 129.8, 129.0, 117.7, 22.0, 16.3. ESI-MS(+): m/z 281.01 [M + H]+, 303.03 [M + Na]+.

2-Methyl-4-oxo-4H-pyran-3-yl 4-nitrobenzenesulfonate

ZBG2 (0.5 g, 4.0 mmol) was reacted with 4-nitrobenzenesulfonyl chloride (0.88 g, 4.0 mmol) in 15 mL of pyridine to afford 2-methyl-4-oxo-4H-pyran-3-yl 4-nitrobenzenesulfonate (PZBG-2c) in 58% yield (0.71 g, 2.3 mmol). 1H NMR (500 MHz, CDCl3) δ = 8.42 (d, J = 9.2 Hz, 2H), 8.31 (d, J = 9.2 Hz, 2H), 7.69 (d, J = 5.7 Hz, 1H), 6.34 (d, J = 5.8 Hz, 1H), 2.53 (s, 3H, CH3). 13C NMR (100 MHz, CDCl3) δ = 171.8, 163.5, 154.6, 151.2, 142.3, 138.6, 130.4, 124.3, 117.6, 16.3. ESI-MS(+): m/z 312.06 [M + H]+.

2-Methyl-4-oxo-4H-pyran-3-yl 2,4-dinitrobenzenesulfonate

ZBG2 (0.5 g, 4.0 mmol) was reacted with 2,4-dinitrobenzenesulfonyl chloride (1.58 g, 5.9 mmol) in 40 mL of pyridine to afford 2-methyl-4-oxo-4H-pyran-3-yl 2,4-dinitrobenzenesulfonate (PZBG-2d) in 28% yield (0.39 g, 1.1 mmol). 1H NMR (500 MHz, CDCl3) δ = 8.72 (d, J = 2.3 Hz, 1H), 8.56 (dd, J1 = 9.2 Hz, J2 = 2.3 Hz, 1H), 8.44 (d, J = 8.6 Hz, 1H), 7.71 (d, J = 5.8 Hz, 1H), 6.31 (d, J = 5.8 Hz, 1H), 2.53 (s, 3H, CH3). 13C NMR (100 MHz, CDCl3) δ = 171.6, 163.3, 154.9, 139.3, 136.4, 133.8, 126.8, 120.5, 117.5, 16.1. ESI-MS(+): m/z 357.03 [M + H]+, 378.99 [M + Na]+.

1,2-Dimethyl-4-oxo-1,4-dihydropyridin-3-yl benzenesulfonate

3-Hydroxy-1,2-dimethylpyridin-4(1H)-one (ZBG-3; 0.5 g, 3.6 mmol) was reacted with benzenesulfonyl chloride (0.51 mL, 4.0 mmol) in 40 mL of pyridine to afford 1,2-himethyl-4-oxo-1,4-dihydropyridin-3-yl benzenesulfonate (PZBG-3a) in 43% yield (0.43 g, 1.5 mmol). 1H NMR (500 MHz, CDCl3) δ = 8.19 (d, J = 6.9 Hz, 2H), 7.66 (t, J = 7.5 Hz, 1H), 7.57 (t, J = 7.5 Hz, 2H), 7.23 (d, J = 8 Hz, 1H), 6.34 (d, J = 8.1 Hz, 1H), 3.63 (s, 3H, NCH 3), 2.49 (s, 3H, CH 3). 13C NMR (100 MHz, CDCl3) δ = 171.3, 144.8, 140.7, 139.9, 137.3, 134.3, 129.0, 128.9, 118.3, 41.8, 14.7. ESI-MS(+): m/z 280.09 [M + H]+.

1,2-Dimethyl-4-oxo-1,4-dihydropyridin-3-yl 4-methylbenzenesulfonate

ZBG-3 (0.2 g, 1.4 mmol) was reacted with p-toluenesulfonyl chloride (0.82 g, 4.3 mmol) in 10 mL of pyridine to afford 1,2-dimethyl-4-oxo-1,4-dihydropyridin-3-yl 4-methylbenzenesulfonate (PZBG-3b) in 86% yield (0.35 g, 1.2 mmol). 1H NMR (500 MHz, CDCl3) δ = 7.99 (d, J = 8.4 Hz, 2H), 7.40 (d, J = 7.6 Hz, 1H), 7.33 (d, J = 8 Hz, 2H), 6.40 (d, J = 7.2 Hz, 1H), 3.64 (s, 3H), 2.42 (s, 3H), 2.41 (s, 3H). 13C NMR (100 MHz, CDCl3) δ = 170.71, 145.67, 145.54, 141.48, 134.07, 129.72, 128.96, 126.07, 117.53, 42.25, 22.00, 14.7. ESI-MS(+): m/z 294.05 [M + H]+, 315.97 [M + Na]+.

1,2-Dimethyl-4-oxo-1,4-dihydropyridin-3-yl 4-nitrobenzenesulfonate

ZBG-3 (0.2 g, 1.5 mmol) was reacted with 4-nitrobenzenesulfonyl chloride (0.488 g, 2.2 mmol) in 10 mL of pyridine to afford 1,2-dimethyl-4-oxo-1,4-dihydropyridin-3-yl 4-nitrobenzenesulfonate (PZBG-3c) in 50% yield (0.23 g, 0.7 mmol). 1H NMR (400 MHz, CDCl3) δ = 8.41–8.34 (m, 4H), 7.26 (d, J = 7.6 Hz, 1H), 6.33 (d, J = 7.6 Hz, 1H), 3.67 (s, 3H), 2.53 (s, 3H). 13C NMR (100 MHz, CDCl3) δ = 170.9, 150.9, 144.9, 143.2, 140.9, 140.1, 130.4, 124.0, 118.4, 41.9, 14.6. ESI-MS(+): m/z 325.11 [M + H]+, 346.96 [M + Na]+.

1,2-Dimethyl-4-oxo-1,4-dihydropyridin-3-yl 2,4-dinitrobenzenesulfonate

ZBG-3 (0.10 g, 0.73 mmol) was reacted with 2,4-dinitrobenzenesulfonyl chloride (0.3 g, 1.1 mmol) in 10 mL of pyridine to afford 1,2-dimethyl-4-oxo-1,4-dihydropyridin-3-yl 2,4-dinitrobenzenesulfonate (PZBG-3d) in 23% yield (0.07 g, 0.2 mmol). 1H NMR (400 MHz, CDCl3) δ = 8.83 (d, J = 2.8 Hz 1H), 8.37 (dd, J1 = 9.6 Hz, J2 = 2.8 Hz, 1 H), 7.82 (d, J = 7.6 Hz, 1 H), 7.03 (d, J = 9.2 Hz, 1 H), 6.25 (d, J = 7.6 Hz, 1 H), 3.68 (s, 3H), 2.31 (s, 3H). 13C NMR (100 MHz, CDCl3) δ = 169.35, 154.98, 144.00, 142.61, 141.42, 140.76, 138.54, 129.69, 122.24, 118.15, 116.78, 59.63, 13.29.

7-Oxocyclohepta-1,3,5-trien-1-yl benzenesulfonate

2-Hydroxycyclohepta-2,4,6-trienone (ZBG4; 0.2 g, 1.7 mmol) was reacted with benzenesulfonyl chloride (0.63 mL, 4.9 mmol) in 5 mL of pyridine to afford 7-oxocyclohepta-1,3,5-trien-1-yl benzenesulfonate (PZBG-4a) in 76% yield (0.33 g, 1.3 mmol). 1H NMR (400 MHz, DMSO) δ = 7.95 (d, J = 7.6 Hz, 2H), 7.80 (t, J = 7.6 Hz, 1H), 7.67 (t, J = 8 Hz, 2H), 7.43–7.37 (m, 2H), 7.25 (t, J = 8.4 Hz, 1H), 7.14–7.09 (m, 2H). 13C NMR (100 MHz, CDCl3) δ = 179.24, 154.79, 141.19, 138.21, 136.37, 136.33, 135.57, 131.96, 130.92, 130.27, 128.76. ESI-MS(+): m/z 262.97 [M + H]+, 279.72 [M + NH4]+, 284.99 [M + Na]+.

7-Oxocyclohepta-1,3,5-trien-1-yl 4-methylbenzenesulfonate

ZBG4 (0.2 g, 1.7 mmol) was reacted with p-toluenesulfonyl chloride (0.41 g, 2.0 mmol) in 10 mL of pyridine to afford 7-oxocyclohepta-1,3,5-trien-1-yl 4-methylbenzenesulfonate (PZBG-4b) in 63% yield (0.04 g, 0.1 mmol). 1H NMR (400 MHz, CDCl3) δ = 7.92 (d, J = 8.4 Hz, 2H), 7.46 (d, J = 9.2 Hz, 1H), 7.35 (d, J = 8.4 Hz, 2H), 7.26–7.16 (m, 2H), 7.13–7.06 (m, 1H), 6.98 (t, J = 10 Hz, 1H), 2.45 (s, 3H). 13C (125 MHz, CDCl3) δ = 179.41, 155.15, 145.50, 141.23, 136.32, 134.61, 133.41, 130.81, 130.00, 129.60, 128.59, 21.78. ESI-MS(+): m/z 277.21 [M + H]+, 293.99 [M + NH4]+.

7-Oxocyclohepta-1,3,5-trien-1-yl 4-nitrobenzenesulfonate

ZBG4 (0.2 g, 1.7 mmol) was reacted with 4-nitrobenzenesulfonyl chloride (1.1 g, 4.9 mmol) in 5 mL of pyridine. Addition of 10 mL of water allowed for precipitation of 7-oxocyclohepta-1,3,5-trien-1-yl 4-nitrobenzenesulfonate (PZBG-4c) in 71% yield (0.38 g, 1.2 mmol) without the need for further purification. 1H NMR (400 MHz, DMSO) δ = 8.45 (d, J = 8.8 Hz, 2H), 8.22 (d, J = 8.8 Hz, 2H), 7.52 (d, J = 9.2 Hz, 1H), 7,46 (dd, J1 = 10.2 Hz, J2 = 3.6 Hz, 1 H), 7.31 (dd, J1 = 9.8 Hz, J2 = 2 Hz, 1 H), 7.15 (dd, J1 = 10.6 Hz, J2 = 3.6 Hz, 1 H). 13C (100 MHz, CDCl3) δ = 179.09, 154.86, 151.50, 141.95, 141.27, 138.59, 136.87, 132.06, 131.48, 130.42, 125.42. ESI-MS(+): m/z 308.01 [M + H]+, 324.73 [M + NH4]+.

7-Oxocyclohepta-1,3,5-trien-1-yl 2,4-dinitrobenzenesulfonate

ZBG4 (0.2 g, 1.7 mmol) was reacted with 2,4-dinitrobenzenesulfonyl chloride (0.54 g, 2.0 mmol) in 10 mL of pyridine to afford 7-oxocyclohepta-1,3,5-trien-1-yl 2,4-dinitrobenzenesulfonate (PZBG-4d) in 6% yield (0.04 g, 0.1 mmol). 1H NMR (400 MHz, DMSO) δ = 9.00 (d, J = 2.4 Hz 1H), 8.66 (dd, J1 = 8.8 Hz, J2 = 2.4 Hz, 1 H), 8.38 (d, J = 8.8 Hz 1H), 7.66 (d, J = 9.6 Hz 1H), 7.50 (td, J1 = 8.4 Hz, J2 = 3.6 Hz, J3 = 1.2 Hz, 1 H), 7.35 (td, J1 = 8.4 Hz, J2 = 2.4 Hz, 1 H), 7.20–7.15 (m, 2H). 13C (125 MHz, DMSO) δ = 178.71, 155.30, 151.20, 147.90, 140.96, 138.83, 137.10, 134.58, 133.22, 131.93, 131.55, 127.92, 121.11. ESI-MS(+): m/z 353.15 [M + H]+, 375.11 [M + Na]+.

6-(([1,1′-Biphenyl]-4-ylmethyl)carbamoyl)-2-oxopyridin-1(2H)-yl benzenesulfonate

1,2-HOPO-2 (Scheme 1) was prepared as previously reported [38]. In a 10-mL round-bottom flask was dissolved 0.05 g (0.16 mmol) of 1,2-HOPO-2 in 3 mL of pyridine. To this was added 60 µL (0.5 mmol) of benzenesulfonyl chloride. The reaction mixture was left stirring under nitrogen at room temperature overnight. After 16 h, the solvent was evaporated to leave an orange oil which was dissolved in dichloromethane and washed once with 1 M HCl then brine. The organic layer was dried over MgSO4, filtered, and concentrated. The product was purified on a silica gel column and eluted with 1% MeOH in dichloromethane to afford 6-(([1,1′-biphenyl]-4-ylmethyl)carbamoyl)-2-oxopyridin-1(2H)-yl benzenesulfonate (1a) in 67% yield (0.05 g, 0.1 mmol). 1H NMR (400 MHz, CDCl3) δ = 8.01 (d, J = 7.6 Hz, 2H), 7.76 (t, J = 7.6 Hz, 1 H), 7.61 (m, 6H), 7.47 (m, 4H), 7.38–7.31 (m, 2H), 6.62 (d, J = 8.0 Hz, 1H), 6.57 (d, J = 6.4 Hz, 1H), 4.63 (d, J = 5.2 Hz, 2H). 13C (100 MHz, CDCl3) δ = 159.2, 157.0, 142.7, 141.2, 140.8, 138.9, 136.0, 134.5, 129.9, 129.5, 129.1, 128.9, 127.8, 127.7, 127.3, 125.2, 107.6, 44.5. ESI-MS(+): m/z 461.13 [M + H]+, 483.13 [M + Na]+. Anal. calcd for C25H20N2O5S·0.5 H2O: C, 63.95; H, 4.51; N, 5.99. Found: C, 63.68; H, 5.14; N, 5.99.

4-(((6-(([1,1′-Biphenyl]-4-ylmethyl)carbamoyl)-2-oxopyridin-1(2H)-yl)oxy)sulfonyl) benzoic acid

1,2-HOPO-2 (0.20 g, 0.6 mmol) was dissolved in 5 mL of pyridine. To this was added 4-(chlorosulfonyl) benzoic acid (0.21 g, 1.0 mmol). The reaction was allowed to proceed overnight at room temperature. The solvent was evaporated and to the remaining oil was added 5 mL of dichloromethane followed by 5 mL of ethyl acetate, allowing for precipitation. The solid white product was filtered off and collected to afford 4-(((6-(([1,1′-biphenyl]-4-ylmethyl)carbamoyl)-2-oxopyridin-1(2H)-yl)oxy)sulfonyl) benzoic acid (1b) in 14% yield (0.04 g, 0.08 mmol). 1H NMR (400 MHz, DMSO) δ = 9.36 (t, J = 5.6 Hz, 1H, NH), 8.18 (d, J = 8.8 Hz, 2H), 8.06 (d, J = 8.8 Hz, 2H), 7.66–7.61 (m, 4H), 7.53 (dd, J1 = 7.8 Hz, J2 = 2.8 Hz, 1H), 7.45 (t, J = 7.6 Hz, 2H), 7.38–7.33 (m, 3H), 6.67 (dd, J1 = 9.2 Hz, J2 = 1.6 Hz, 1H), 6.47 (dd, J1 = 6.4 Hz, J2 = 1.2 Hz, 1H), 4.27 (d, J = 5.6 Hz, 2H). 13C NMR (100 MHz, DMSO) δ = 166.55, 159.45, 156.86, 143.21, 140.61, 140.54, 139.68, 138.26, 138.02, 137.61, 130.97, 130.09, 129.62, 128.73, 128.08, 127.33, 127.28, 124.19, 107.11, 43.05. ESI-MS(−): m/z 502.89 [M − H]−. Anal. calcd for C26H20N2O7S·0.25 HCl: C, 60.80; H, 3.97; N, 5.45. Found: C, 60.92; H, 4.30; N, 5.73.

2-(([1,1′-Biphenyl]-4-ylmethyl)carbamoyl)-4-oxo-4H-pyran-3-yl benzenesulfonate

PY-2 (Scheme 1) was prepared as previously reported [38]. In a 50-mL round-bottom flask was dissolved 0.20 g (0.6 mmol) of PY-2 in 15 mL of pyridine. To this was added 240 µL (1.8 mmol) of benzenesulfonyl chloride. The reaction mixture was left stirring under nitrogen at room temperature overnight. After 16 h, the solvent was evaporated to leave a red oil which was dissolved in dichloromethane and washed once with 1 M HCl then brine. The organic layer was dried over MgSO4, filtered, and concentrated. The product was precipitated from MeOH to afford 2-(([1,1′-biphenyl]-4-ylmethyl)carbamoyl)-4-oxo-4H-pyran-3-yl benzenesulfonate (2a) in 12% yield (0.03 g, 0.07 mmol). 1H NMR (400 MHz, CDCl3) δ = 8.10 (d, J = 8.4 Hz, 2H), 7.80 (d, J = 5.6 Hz, 1 H), 7.71 (t, J = 7.6 Hz, 1H), 7.61 (m, 6H), 7.47–7.36 (m, 6H), 6.47 (d, J = 6.0 Hz, 1H), 4.67 (d, J = 6.0 Hz, 2H). 13C NMR (100 MHz, CDCl3) δ = 172.7, 157.3, 154.8, 151.3, 141.2, 140.8, 139.2, 136.0, 135.8, 135.2, 129.3, 129.2, 129.1, 128.9, 127.8, 127.7, 127.3, 118.4, 44.3. ESI-MS(+): m/z 461.98 [M + H]+, 484.02 [M + Na]+.

UV–vis spectroscopy

Absorption spectra of compounds were taken with a PerkinElmer Lambda 25 UV–vis spectrophotometer. To a 1.0 mL solution at 0.05 mM concentration of each compound in N-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid (HEPES) buffer (50 mM, pH 7.5) was added H2O2 (10 µL, 0.09 M in HEPES) and absorption spectra were monitored over time at room temperature. Hydrolytic stability was measured by monitoring each sample in HEPES buffer over a 24-h period.

Calculation of rate constant

The pseudo-first-order rate constant was calculated following a literature procedure [20]. To a 1.0 mL solution of PZBG-1a, PZBG-1b, PZBG-1e, 1a, and 1b in HEPES buffer at 50 µM was added H2O2 to final concentrations of 150, 250, 500, 750, and 900 µM. Spectra were monitored over 15–30 min at room temperature, with at least 50 spectra recorded at every concentration. The change in absorbance at 298 nm was monitored for PZBG-1a and PZBG-1b, whereas the change in absorbance at 288 nm was recorded for PZBG-1e, and at 310 nm for 1a and 1b. The rate constant (k obs) was found from the linear slope of ln[(A − A ZBG)/(A o − A ZBG)] versus time, where A ZBG is the absorbance of a 50 µM sample of the ZBG or full-length inhibitor and A o is the initial absorbance of PZBG-1a, PZBG-1b, PZBG-1e, 1a, and 1b. The rate of conversion was determined from the slope of the line of k obs versus H2O2 concentration.

High-performance liquid chromatography

Analytical High-performance liquid chromatography (HPLC) was performed with a HP series 1050 system equipped with a Vydac® C18 reverse-phase column (218TP, 250 mm × 4.6 mm, 5 µm). Separation was achieved with a flow rate of 1 mL min−1 and the following solvents: solvent A was 5% MeOH and 0.1% formic acid in H2O and solvent B was 0.1% formic acid in MeOH. Starting with 95% solvent A and 5% solvent B, we ran an isocratic gradient for 15 min to a final solvent mixture of 5% solvent A and 95% solvent B, which was held for 5 min before ramping back down to 95% solvent A and 5% solvent B in 2 min and holding for an additional 4 min. ZBG-1a and PZBG-1a were prepared in HEPES buffer (50 mM, pH 7.5) at a concentration of 1 mM and retention times were determined. To evaluate cleavage by H2O2, a 1 mM solution of PZBG-1a in HEPES buffer was reacted with a 20-fold excess of H2O2 before analysis under identical HPLC conditions as before.

Inhibition assays

MMP-12 (catalytic domain, human recombinant) was purchased from Enzo Life Sciences. The assays were carried out in a 96-well plate using a Bio-Tex Flx 800 plate reader. The activity of MMP-12 was evaluated after a 30-min incubation in the presence of H2O2 and proMMPi. The concentration of proMMPi used was selected to be close to the IC50 value of the parent full-length inhibitors, 1,2-HOPO-2 and PY-2 [38]. In each well, 1 µL of proinhibitors 1a, 1b, and 2a and the inhibitors 1,2-HOPO-2 and PY-2 in DMSO (5 µM) were incubated for 30 min at 37 °C with 20 µL of MMP-12 (0.35 U mL−1), 10 µL H2O2 (1 mM in HEPES buffer, pH 7.5), and MMP assay buffer (50 mM HEPES, 10 mM CaCl2, 0.10% Brij-35, pH 7.5) for a total volume of 99 µL. A control sample containing 10 µL H2O2 (1 mM in HEPES buffer, pH 7.5) in MMP assay buffer was also prepared to confirm that H2O2 did not inhibit MMP-12. The reaction was initiated by the addition of 1 µL (400 µM) of the fluorescent substrate Mca-Pro-Leu-Gly-Leu-Dpa-Ala-Arg-NH2 [where Mca is (7-methoxycoumarin-4-yl)acetyl and Dpa is N-3-(2,4-dinitrophenyl)-l-α,β-diaminopropionyl)] after the 30-min incubation period and kinetic activity was measured every minute for 20 min with excitation and emission wavelengths at 335 and 405 nm, respectively. Enzyme activity with the inhibitor was calculated with respect to the control experiment––no inhibitor present. Measurements were performed in duplicate in two independent experiments.

Results and discussion

Assessment of sulfonate esters as suitable protecting groups for ZBGs

Previous studies utilizing fluorescent probes have shown sulfonate esters to be suitable protecting groups of hydroxyl groups that show a turn-on response upon exposure to ROS, including H2O2 and superoxide anion [35–37]. To investigate the use of sulfonate ester protecting groups for the development of ROS-activated proMMPi, a small library of compounds was synthesized. As shown in Fig. 1, sulfonate esters with different substituents were appended to ZBG1-4 to evaluate which protected ZBGs (PZBGs) provided efficient activation in the presence of H2O2. These PZBGs were readily prepared by combining a ZBG with the appropriate sulfonyl chloride in pyridine. In total, 17 PZBGs (PZBG-1a–PZBG-1e, PZBG-2a–PZBG-2d, PZBG-3a–PZBG-3d, and PZBG-4a–PZBG-4d) were prepared and tested for cleavage in the presence of H2O2.

Fig. 1
figure 1

Zinc-binding groups (ZBGs) and their sulfonate ester derivatives (protected ZBGs; PZBGs) examined in this study

Full size image

To evaluate cleavage of the compounds by ROS, a sample of each PZBG in HEPES buffer (50 mM, pH 7.5) was activated with excess H2O2 (0.9 mM, 18 equiv) and the change in absorbance was monitored over time via electronic spectroscopy. Surprisingly, only compounds derived from ZBG-1 showed a change in absorbance (corresponding to the formation of the free ZBG) upon exposure to H2O2. The fact that only the five PZBG-1 derivatives (out of 17 total combinations) were cleaved in the presence of H2O2 strongly suggests that the N–O group is essential for the observed reactivity. Figure 2 shows representative absorption spectra of PZBG-1a in the presence of H2O2. A decrease in absorbance at 298 nm over time is noted, representing the disappearance of PZBG-1a and a gradual increase in absorbance at 312 nm is observed, indicating the emergence of ZBG-1. In addition, analytical HPLC was used to confirm that ZBG-1 was the product after reaction with H2O2 (Fig. 2). Upon the addition of H2O2 to PZBG-1a for 60 min, a peak with a retention time of 5.0 min was observed, which is identical to an authentic sample of ZBG-1. Similarly, treatment of PZBG-1b with H2O2 resulted in nearly identical spectra as found with PZBG-1a, whereas PZBG-1c showed rapid hydrolysis upon the addition of H2O2 (data not shown). It should be noted that the absorption spectra of PZBG-1d were not readily interpreted, owing to the overlapping of absorption profiles of the protecting group and the free ZBG; however, thin-layer chromatography showed the emergence of the free ZBG, demonstrating rapid hydrolytic cleavage of the protecting group (even in the absence of H2O2). These findings taken together prompted the synthesis of PZBG-1e, a more water-soluble alternative to PZBG-1c, with a carboxylic acid attached to the para position of the sulfonate ester. A substantial increase in solubility in buffered solution was noted, and the cleavage behavior was similar to that of the other compounds reported (see below), making PZBG-1e an attractive candidate for development into a full-length proMMPi. As mentioned above, sulfonate ester derivatives of ZBG-2a–ZBG-2d, ZBG-3a–ZBG-3d, and ZBG-4a–ZBG-4d did not show any change in absorbance over a period of 1 h with an 18 M excess of H2O2. These findings suggest that the N–O bond in ZBG-1 is required for facile cleavage of the sulfonate ester group in this series of ligands, although this will require verification by additional studies. It is interesting to note that the compounds tested that did not contain the N–O moiety (those based on PZBG-2, PZBG-3, and PZBG-4) appeared to be stable in aqueous buffer (over at least a 1-h period).

Fig. 2
figure 2

Analysis of PZBGs in the presence of H2O2. a Absorption spectra of PZBG-1a [0.05 mM, 50 mM N-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid (HEPES) buffer, pH 7.5] in the presence of H2O2 (0.9 mM, 18 equiv) monitored every 5 min for 60 min. The dashed line represents the initial spectrum, and an authentic sample of ZBG-1 is shown in red. The arrows indicate changes in spectra over time. b High performance liquid chromatography chromatograms of PZBG-1a, PZBG-1a + H2O2, and ZBG-1. The retention times are 11.5 min for PZBG-1a and 5.0 min for ZBG-1. c–e Absorption spectra of PZBG-2a, PZBG-3a, and PZBG-4a, respectively (0.05 mM, 50 mM HEPES buffer, pH 7.5), in the presence of H2O2 (0.9 mM, 18 equiv) monitored every 5 min for 60 min. The overlapping spectra indicate that no cleavage of the protecting group is occurring in the presence of H2O2

Full size image

One key factor for any prodrug approach is the stability of the protecting group in the absence of the triggering stimuli. To test the stability of the sulfonate esters in buffer, absorption spectra for PZBG-1a, PZBG-1b, and PZBG-1e were collected over 24 h. These stability studies showed approximately 50% cleavage of PZBG-1a and PZBG-1e in 6 h, whereas PZBG-1b was approximately 30% cleaved in 24 h. The rates of conversion of PZBG-1a, PZBG-1b, and PZBG-1e were determined by monitoring the change in absorbance using pseudo-first-order reaction conditions with an excess of H2O2 as previously reported [20]. The calculated rate constants indicate that PZBG-1e had the fastest rate constant at 1.3 M−1 s−1, whereas rate constants of 0.7 and 0.3 M−1 s−1 were determined for PZBG-1a and PZBG-1b, respectively. It should be noted that the rate constants determined do take into account background hydrolysis and reactivity with H2O2; however, all kinetic measurements were taken over a 15–30-min period, which is before a measurable amount of hydrolysis was observed. Experiments with PZBG-1c and PZBG-1d showed the fastest cleavage kinetics upon exposure to H2O2, with complete dissociation achieved in less than 3 min (no rate constants determined). The rates of conversion for the PZBG-1 compounds are consistent with the nature of the respective substituents on the leaving groups. When substituents are varied on an aromatic ring, the change in free energy of activation for a given reaction is proportional to the change in Gibbs free energy, as summarized by the Hammett equation [39]. Of the molecules for which rate constants were obtained, PZBG-1e, which contains an electron-withdrawing group (–CO2H) in the para position, dissociates the fastest. PZBG-1b, on the other hand, with an electron-donating group (–CH3) in the para position, had the slowest rate of the compounds tested. PZBG-1a, with no substituents, had a cleavage rate falling between the others, consistent with the Hammett relationship. The results with PZBG-1c and PZBG-1d are also consistent with this relationship, with the strongly electron withdrawing nitro groups producing the fastest rates, resulting in the inability to acquire precise values.

Development of full-length proMMPi

Having demonstrated the ability of sulfonate esters to act as cleavable protecting groups for ZBG-1, we incorporated this chelator into a full-length proMMPi. The corresponding full-length inhibitor of ZBG-1 with a hydrophobic biphenyl backbone, 1,2-HOPO-2, has been previously prepared and studied [38]. 1,2-HOPO-2 is an effective inhibitor of MMP-3, MMP-8, and MMP-12, with IC50 values under 100 nM [38]. Two full-length MMPi (1,2-HOPO-2 and PY-2) were prepared by previously reported procedures and then protected in pyridine with an excess of the appropriate sulfonyl chloride to generate proMMPi 1a, 1b, and 2a (Scheme 1). The proMMPi 1a and 1b were evaluated for activation by H2O2 via electronic spectroscopy in the same manner as with the PZBG compounds. As shown in Scheme 1, the activation of 1a and 1b to the known MMPi 1,2-HOPO-2 is achieved upon the addition of H2O2. Figure 3 shows the absorption spectra of 1b, for which a decrease at 310 nm over time is observed, indicating the disappearance of the protected MMPi, and a gradual increase in absorbance at 350 nm is observed, demonstrating the emergence of 1,2-HOPO-2. Pseudo-first-order rate constants were determined with an excess of H2O2 as described earlier. Rate constants of 0.3 and 1.1 M−1 s−1 were obtained for 1a and 1b, respectively, which are in good agreement with the rate constants determined for the protected chelators PZBG-1a and PZBG-1e. To evaluate the stability of the sulfonate esters, absorption spectra of 1a and 1b were collected in buffer alone. Compounds 1a and 1b showed approximately 50% hydrolysis after 9 and 3 h, respectively.

Scheme 1
scheme 1

Activation of proinhibitor 1a or 1b with H2O2 results in generation of the inhibitor 1,2-HOPO-2. In contrast, treatment of proinhibitor 2a with H2O2 did not result in production of the inhibitor PY-2

Full size image
Fig. 3
figure 3

Absorption spectra of 1b (0.05 mM, 50 mM HEPES buffer, pH 7.5) in the presence of H2O2 (0.9 mM, 18 equiv excess) monitored every 5 min for 60 min. The dashed line represents the initial spectrum, and an authentic sample of 1,2-HOPO-2 is shown in red. The arrows indicate changes in spectra over time

Full size image

To further confirm that the behavior of the protected chelators (PZBGs) was readily translated to a complete proMMPi, a proinhibitor based on PZBG-2a was synthesized. Compound 2a, which contains a biphenyl backbone like 1a, was prepared. Upon cleavage of the protecting group, 2a should produce PY-2, a known inhibitor of several MMPs [38]. Treatment of 2a with excess H2O2 over the course of 60 min did not result in cleavage of the sulfonate ester (Scheme 1), as evidenced by absorption spectroscopy (data not shown). This negative result is consistent with all of the findings described above, showing that the cleavage behavior of the PZBG is retained in its full-length proMMPi.

MMP inhibition studies

To monitor the ability of the protected compounds to inhibit MMP-12 in the presence of H2O2, a fluorescence-based assay was used [40]. Compounds 1a, 1b, and 2a were tested at concentrations close to the IC50 values of their active parent molecules, 1,2-HOPO-2 and PY-2, against MMP-12. The percent inhibition of proinhibitors 1a, 1b, and 2a at 50 nM was evaluated after 30 min of incubation with and without H2O2 (Fig. 4). Before treatment with H2O2, 1a and 1b were shown to exhibit approximately 20% and 30% inhibition of MMP-12, respectively. This significant level of inhibition is likely due to the hydrolysis of these compounds to the active MMPi during the incubation period, as described earlier for the stability studies. In addition, the inhibition assays were performed at a higher temperature than our kinetic experiments, which may result in an increase in the rate of hydrolysis and therefore lead to higher inhibition than might otherwise be expected. After treatment with 100 µM H2O2, the percent inhibition by proinhibitors 1a and 1b increased to approximately 30% and 50%, respectively, indicative of activation to 1,2-HOPO-2. Proinhibitor 2a was used as a negative control, and as expected, displayed no significant change in inhibitory activity upon exposure to H2O2.

Fig. 4
figure 4

Percent inhibition of MMP-12 with proinhibitors 1a, 1b, and 2a tested at 50 nM in the absence (gray bars) and presence (black bars) of H2O2 after 30 min of activation

Full size image

Conclusions

In summary, we have demonstrated an alternative method to prepare proMMPi that are activated in the presence of ROS. The addition of sulfonate ester protecting groups to the ZBG blocks the chelating ability of the inhibitors, thereby attenuating their activity. The sulfonate ester proMMPi are easy to synthesize, have tunable rates of cleavage (modulated by appropriate substituents on the leaving group), and can be made reasonably water soluble by addition of substituents, such as carboxylic acids, on the aromatic ring of the protecting group. Unfortunately, the limited hydrolytic stability of the sulfonate ester protected ZBGs is a significant drawback. Nearly complete hydrolysis of the sulfonate ester proMMPi was observed after 24 h. In contrast, self-immolative boronic ester derivatized proMMPi were entirely stable under identical conditions [31]. In addition, the sulfonate esters were only cleaved from one (ZBG-1) of four different chelators tested, suggesting that the present strategy may not be readily applied to inhibitors with other ZBGs. Nonetheless, complete proMMPi based on these sulfonate ester protected chelators were activated after treatment with H2O2 at relatively fast rates, resulting in an increase in inhibition of MMP-12. It is expected that combining sulfonate esters with a self-immolative strategy may produce more stable, and more rapidly cleavable proinhibitors [31, 41, 42]. These and related experiments are under way.