Selective Water-Soluble Gelatinase Inhibitor Prodrugs

Abstract

SB-3CT (1), a selective and potent thiirane-based gelatinase inhibitor, is effective in animal models of cancer metastasis and stroke; however, it is limited by poor aqueous solubility and extensive metabolism. We addressed these issues by blocking the primary site of metabolism and capitalizing on a prodrug strategy to achieve >5000-fold increased solubility. The amide prodrugs were quantitatively hydrolyzed in human blood to a potent gelatinase inhibitor, ND-322 (3). The arginyl amide prodrug (ND-478, 5d) was metabolically stable in mouse, rat, and human liver microsomes. Both 5d and 3 were non-mutagenic in the Ames II mutagenicity assay. The prodrug 5d showed moderate clearance of 0.0582 L/min/kg, remained mostly in the extracellular fluid compartment (Vd = 0.0978 L/kg), and had a terminal half-life of >4 h. The prodrug 5d had superior pharmacokinetic properties than 3, making the thiirane class of selective gelatinase inhibitors suitable for intravenous administration in treatment of acute gelatinase-dependent diseases.

INTRODUCTION

Matrix metalloproteinases 2 and 9 (MMP-2 and MMP-9) are zinc-dependent endopeptidases that are also known as the gelatinases. The unregulated activities of these enzymes have been implicated in a host of human diseases, including tumor invasion and metastasis, cardiovascular and neurological diseases, and inflammation to name a few.^1-4 Gelatinases were shown to be essential for leukocyte penetration into the brain parenchyma in a mouse model of experimental autoimmune encephalomyelitis.⁵ All these diseases involve restructuring of the extracellular matrix for manifestation of the disease,^{4, 6-8} hence selective gelatinase inhibitors are highly sought.^{9, 10}

We have described the first highly selective mechanism-based gelatinase inhibitor, (4-phenoxyphenylsulfonyl)methylthiirane (SB-3CT, 1).¹¹ This inhibitor undergoes a chemical transformation within the active site of gelatinases, a process that is likely at the root of the selectivity that the compound exhibits in its potent inhibition.¹² Compound 1 is efficacious in animal models of prostate cancer metastasis to the bone,¹³ breast cancer metastasis to the lungs¹⁴ and T-cell lymphoma metastasis to the liver.¹⁵ Compound 1 also blocks the MMP-9-dependent degradation of the extracellular matrix protein laminin, and thus rescues neurons from apoptotic cell death in a mouse model of transient focal cerebral ischemia.¹⁶ In addition, 1 prevents laminin degradation and neuronal death in a rat model of subarachnoid hemorrhage,¹⁷ a type of hemorrhagic stroke. Compound 1 significantly reduces flow-induced vascular remodeling, an adaptive process that allows blood vessels to normalize hemodynamic stress in response to increased blood flow in atherosclerosis, aneurysms, and brain arteriovenous maltransformations, in mice.¹⁸ In a rat model of spinal cord injury, treatment with 1 results in decreases in MMP-9-activity, extravasation, and apoptotic cell death.¹⁹ This novel inhibitor has found use in many other biological systems, which we will not outline here in the interest of brevity.

Despite documented excellent biological activity, administration of 1 to animals is problematic due to its poor water solubility of 2.3 μg/mL.²⁰ Because of poor solubility, the compound is typically dosed as a suspension intraperitoneally. Furthermore, it is rapidly metabolized by hydroxylation at the para position of the terminal phenyl ring (2) to a more potent gelatinase inhibitor than the parent 1, and by oxidation at the α-position to the sulfonyl that leads to the formation of the inactive sulfinic acid.²¹ We addressed oxidation at the α-position to the sulfonyl by addition of a methyl substituent at that site; however, this modification led to a 10-fold decrease in inhibitory potency.²² On the other hand, derivatization at the para-position of the terminal phenyl ring with methanesulfonate did not only improve metabolic stability, but also enhanced potency.²³

Our interest in developing therapeutics for the treatment of acute neurological conditions such as stroke, aneurysm, and traumatic brain injury, for which intravenous infusion is the preferred route of administration, led us to explore increasing the water solubility of 1. For parenteral administration, a solubility of 10 mg/mL is usually needed to adequately formulate a drug.²⁴ This presented a delivery challenge in the case of 1, which was administered intraperitoneally as a suspension in studies of animal models of disease. A prodrug strategy is useful to solve formulation problems associated with poor aqueous solubility. A prodrug (drug + promoiety) is a chemical entity with little or no pharmacological activity, which undergoes transformation (biological or otherwise) to a therapeutically active drug.²⁵ The majority of prodrugs for parenteral administration have modifications where a polar ionizable promoiety is introduced to increase water solubility.²⁵ Amino-acid promoieties have been employed to improve the aqueous solubility of drugs containing an alcohol or amine,^{26, 27} and are rapidly and quantitatively converted to the active drug by the action of esterases and/or peptidases in blood. We surmise that functionalization at the para-position of the terminal phenyl ring with water-solubilizing agents to create prodrugs might address the solubility problem of 1, as well as block the primary site of metabolism. We used amino acids (and a peptide) as promoieties to impart water solubility to the gelatinase inhibitor. The p-hydroxy derivative 2 is a good candidate for ester conjugation, as it was previously shown to be a more potent gelatinase inhibitor than 1.²¹ We also explored the use of the p-amino derivative 3, which was found to have potent gelatinase inhibitory activity, to synthesize amino-acid amide conjugates. The resulting ester and amide prodrug series (Table 1) are expected to undergo hydrolysis in plasma and blood to release the active gelatinase inhibitors 2 or 3. The syntheses and evaluation of the properties of these prodrug series, compounds 4 and 5, are the subjects of this report.

Table 1.

Structures of Ester and Amide Prodrugs

X
Gly	4a	5a
l-Lys	4b	5b
l-Glu	4c	5c
l-Arg	4d	ND-478 (5d)
l-Arg- l-Arg	4e	5e

RESULTS AND DISCUSSION

Syntheses of Prodrugs

The p-aminophenoxybenzene scaffold in 3 was assembled by the methodology developed by Ikejiri et al.,²⁸ which is the coupling of 1-fluoro-4-nitrobenzene (6) and 4-(allylthio)phenol (7) under basic condition, followed by reduction of the nitro group to the amine over elemental zinc in acetic acid (Scheme 1). The resulting amine 8 was treated with di-tert-butyl dicarbonate to give the Boc-protected compound 9. The Boc group was chosen as the amine-protecting group, since sulfonylmethylthiirane is relatively stable under the acidic condition that is used for the removal of Boc at the end of the synthesis (see transformation of 10 to 3). The rest of the transformations, including oxidation to sulfone, epoxide formation using m-chloroperoxybenzoic acid (m-CPBA), and thiirane conversion using thiourea, were performed by the methodology developed by us earlier.^{29, 30} Removal of the Boc group in compound 10 was carried out in the presence of 4 N HCl in 1,4-dioxane at room temperature for 48 h to yield 3 as the HCl salt in 82% yield.

Scheme 1.

Synthesis of 3^a

^a Reagents and condition: (a) (i) Cs₂CO₃, room temperature, 2 h; (ii) Zn, AcOH, 0 °C to room temperature, 2 h, 79%. (b) Boc₂O, Et₃N, MeOH, 60 °C, 2 h, 82%. (c) m-CPBA, CH₂Cl₂, room temperature, 72 h, 81%. (d) thiourea, MeOH/CH₂Cl₂, room temperature, 18 h, 72%. (e) 4 N HCl in 1,4-dioxane, ethyl acetate/CH₂Cl₂, 0 °C to room temperature, 48 h, 82%.

The syntheses of amino-acid conjugates of phenol 2 and of aniline 3 as prodrugs of gelatinase inhibitors are outlined in Scheme 2. Compound 2, prepared according to the literature,²³ was acylated with Boc-protected N-hydroxysuccinimide (NHS) esters of amino acids (Gly, L-Lys, L-Glu, L-Arg) or peptide (L-Arg-L-Arg) in the presence of 4-(dimethylamino)pyridine (DMAP) to give the esters 11. The Boc group in compounds 11, in turn, was readily removed by treatment with 4 N HCl in 1,4-dioxane to result in the ester prodrugs 4. The acid treatment for removal of the Boc group in the presence of thiirane worked well, as described in Scheme 1 for the synthesis of 3.

Scheme 2.

Syntheses of the Ester and Amide Prodrugs^a

^a Reagents and condition: (a) DMAP, iPr₂EtN, CH₂Cl₂, room temperature, 3 h, 72 - 81%. (b) 4 N HCl in 1,4-dioxane, ethyl acetate/CH₂Cl₂, 0 °C to room temperature, 48 h, 71-79%. (c) THF, −20 °C to room temperature, 1 h, 61-79%.

When 3 was subjected to the same acylation conditions (using NHS esters of amino acid), the desired amide bond was not formed. Instead, it gave only the recovery of the two starting materials. The Boc-protected amino acid was activated to a mixed anhydride by treatment of isobutyl chloroformate in the presence of N-methylmorpholine, and was allowed to react with 3 to give the desired amide linkage. The removal of the Boc group in compounds 12 was carried out by acid treatment resulting in the amide prodrugs 5.

Aqueous Solubility and Ex vivo Hydrolyses of Prodrugs

The aqueous solubilities of the prodrugs were assessed by UV spectroscopy (see Experimental Section). The solubility of compounds 4 and 5 exceeded 10,000 μg/mL, which indicates a significant enhancement in aqueous solubility over that of the parent 1 (2.3 μg/mL) of over 5000-fold. Next, we explored the stability of compounds 4 and 5 in aqueous solution. The ester prodrugs 4 were unstable under aqueous conditions and within 2 h the aqueous solutions turned turbid. The precipitate was analyzed and was confirmed as compound 2. The instability of the ester prodrugs in aqueous solution might be due to promotion of a water molecule by the aminoacyl moiety itself for hydrolysis of the ester group or by intermolecular aminolysis of the ester moiety by the free amino group(s) in the promoiety. Hence, the ester derivatives 4 would appear to be inherently unstable under these conditions. However, the corresponding amide derivatives 5 were stable in aqueous solution for over one month, as assessed by HPLC.

We subsequently explored the properties of the prodrugs in the presence of human plasma and human whole blood. The ester prodrugs 4 hydrolyzed completely in human plasma within 2 min. In contrast, the amide derivatives 5 displayed increased stability in human plasma. The half-lives of the amide prodrugs in human plasma and whole blood are summarized in Table 2. The amide prodrugs 5a-d quantitatively released the active constituent 3 in human plasma within 30 min. Whereas the half-life for hydrolysis of the dipeptide variant 5e was shorter in the series, the hydrolytic enzymes in plasma first gave rise to the mono-arginine variant 5d, which experienced further hydrolysis to 3. Hydrolyses of the amide conjugates were in general more rapid in whole blood than in plasma, significantly so in the case of compounds 5b and 5e. The time-dependent release of 3 from the arginyl amide prodrug 5d is depicted in Figure 1.

Table 2.

Ex vivo Hydrolyses of Amide Prodrugs of 3

Amide Prodrugs	Half-life at 37 oC (min)		Aqueous Solubility (mg/mL)
	human plasma	human blood
5a	32.4 ± 2.7	27.1 ± 2.5	>10a
5b	25.1 ± 5.2	15.3 ± 1.9	>10a
5c	32.8 ± 3.6	27.9 ± 3.8	>10a
5d	29.5 ± 0.7	25.4 ± 0.1	>10a
5e	11.8 ± 0.5	5.9 ± 0.6	>10a

^aDue to limited sample supply, no further effort was made to measure the maximum solubility.

Figure 1.

Ex vivo hydrolysis of 5d in human blood generates 3.

Gelatinase Inhibition

The kinetics of gelatinase inhibition of the amide prodrugs is given in Table 3. The amide prodrugs showed slow-binding behavior in inhibition of gelatinases, but the activities were consistently worse for the prodrugs, compared to the active constituent 3, which is released from them. Several general observations can be made of these compounds. First, the rate constants for the onset of inhibition (k_on) are rapid and the rate constants for the reversal of inhibition (k_off) are slow. From the ratio of k_off/k_on the dissociation constants (K_i) are evaluated. Compound 3, the species that is released from the prodrugs, is the most effective of the compounds listed in Table 3, inhibiting the enzymes with K_i values in the nanomolar range.

Table 3.

Kinetic Parameters for Inhibition of Gelatinases

Amide Prodrugs	MMP-2			MMP-9a
	10−3kon (M−1s−1)	104koff (s−1)	Ki (μM)	10−3kon (M−1s−1)	104koff (s−1)	Ki (μM)
5a	1.9 ± 0.6	8.2 ± 0.9	0.44 ± 0.05	0.21 ± 0.04	6.5 ± 0.6	3.1 ± 0.3
5b	3.4 ± 0.1	2.1 ± 0.9	0.062 ± 0.025	0.27 ± 0.04	4.2 ± 0.7	1.6 ± 0.3
5c	2.5 ± 0.1	1.8 ± 1.6	0.069 ± 0.064	0.28 ± 0.02	12.0 ± 0.7	4.3 ± 0.4
5d	1.5 ± 0.1	7.6 ± 2.9	0.52 ± 0.20	0.031 ± 0.006	6.0 ± 0.3	19.6 ± 1.0
5e	1.5 ± 0.1	6.1 ± 2.1	0.42 ± 0.15	0.044 ± 0.004	8.3 ± 2.3	18.7 ± 5.6
3	8.9 ± 0.2	2.1 ± 1.3	0.024 ± 0.015	0.61 ± 0.01	5.3 ± 0.7	0.87 ± 0.11

^aCatalytic domain

Metabolic Stability in Mouse, Rat, and Human Liver Microsomes

The in vitro metabolism of 5d and 3 was further evaluated in mouse, rat, and human liver microsomal incubations. The prodrug 5d was metabolically stable, as indicated by half lives of 60 min or longer in mouse, rat, and human liver microsomes (Table 4). The prodrug 5d was more metabolically stable than 3. Compound 3 was stable in human liver microsomes, with t_½ >60 min; the rates of metabolism in mouse and rat liver microsomes were faster, with t_½ values of 27.3 and 23.4 min, respectively. Compound 3 was significantly more metabolically stable than 1, which had a t_½ of 12.0 min in rat liver microsomes.²⁰

Table 4.

Half-life and Apparent Intrinsic Clearance in Liver Microsomes

	5d		3
Species	Half-life (min)	CLint,app ( mL/min/kg)	Half-life (min)	CLint,app ( mL/min/kg)
Mouse	>60	<47	27.3 ± 2.1	102.8 ± 7.9
Rat	59.3 ± 4.5	23.7 ± 1.8	23.4 ± 0.7	60.0 ± 1.8
Human	>60	<10	>60	<10

Intrinsic clearance (CL_int), a measure of how readily the liver metabolizes a compound, was calculated from the in vitro half-lives. The prodrug 5d had intrinsic clearances of <47, 23.7, and <10 mL/min/kg in mouse, rat, and human, respectively, which were <50% of hepatic blood flow (86, 66, and 21 mL/min/kg in mouse, rat, and human, respectively),³¹ indicating that the compound would not be readily metabolized by the liver. Intrinsic clearances for 3 were 103, 60, and <10 mL/min/kg in mouse, rat, and human, respectively. Intrinsic clearances of 3 in the mouse and rat, in contrast to the case in human, were >70% hepatic blood flow, indicating that 3 was a higher-clearance compound in these species. In contrast, the prototype inhibitor 1 was readily metabolized and had an intrinsic clearance of 117 mL/min/kg in the rat, significantly higher than the hepatic blood flow of 66 mL/min/kg.²⁰ These data collectively indicate that 5d has superior pharmacokinetic properties compared to 3 and 1.

Metabolism Pathway

The metabolism of 5d was similar across species. The major metabolic pathway of 5d was hydrolysis to the active constituent 3 (Figure 2). Minor oxidation at the alpha position to the sulfonyl group to lead to the corresponding sulfinic acid M1 and N-acetylation of 3 to give M3 (Figure 2) were observed. M1 is not observed on the reconstructed ion chromatogram as it is a minor metabolite and it readily experiences hydrolysis to give M2, which was detected. The prodrug 5d was hydrolyzed to 3 by amidases present in microsomes and further oxidation gave its sulfinic acid metabolite M2. The identities of the sulfinic acids M1 and M2 were assigned based on MS/MS spectra (Figure 3). We had previously observed a similar fragmentation pattern for the sulfinic acid of 1.²¹ The identity of M3 was confirmed by comparison of HPLC retention time and MS/MS spectrum (Figure 3) with those of an authentic synthetic standard, prepared as reported earlier.²⁸

Figure 2.

Reconstructed ion chromatogram following 60-min incubation of 5d with rat S9 liver microsomes.

Figure 3.

Product ion mass spectra of (A) metabolite M1 (MH⁺, m/z 406) (B) M2 (MH⁺, m/z 250) and (C) M3 (MH⁺, m/z 364).

To gain further insight into the metabolism of 5d after removal of the promoiety, we investigated the metabolism of 3. Incubation of 3 with liver microsomes showed oxidation at the α-position to the sulfonyl to give the sulfinic acid derivative M2. To detect other minor metabolites of 3, the ion chromatogram at m/z 338 (MH + 16)⁺ was reconstructed (Figure 4, inset). The chromatogram displayed two peaks, labeled as M4 and M5, with distinct fragmentation patterns. Product ion mass spectrometric analysis of M5 showed cleavage of the sulfonyl-methylenethiirane linkage giving an ion at m/z 248, followed by a subsequent loss of sulfur to give a fragment at m/z 216, and further loss of two oxygens to generate the fragment at m/z 184. The product ion mass spectrum of 3 produced identical fragments at m/z 248, 216 and 184, indicating that the terminal phenyl ring and the middle phenyl ring were intact, and that oxidation of the thiirane sulfur was the most likely transformation to account for the increase in mass of 16 a.m.u. in M5. Loss of 16 a.m.u. from MH⁺ to give a fragment at m/z 322, confirmed the assignment of M5 as the sulfoxide of 3. We had seen this fragmentation previously for the sulfoxide metabolite of 1.²¹ The MS/MS spectrum of M4, on the other hand, indicated cleavage of the sulfonyl-methylenethiirane linkage to generate a fragment at m/z 264, 16 a.m.u. higher than that of 3. Loss of SO₂ gave an ion at m/z 200. These data suggested the potential oxidation of the amino group or oxidation at the terminal phenyl or the middle phenyl rings. While hydroxylation of the middle ring is possible, we had not observed such oxidative pathway for 1 and its derivatives previously.^{21, 22} Since the fragmentation pattern could not discriminate between the variant 13 and the N-hydroxylamine 14, we prepared the authentic synthetic standards for each of compounds 13 and 14 to confirm the identity of the minor metabolite M4.

Figure 4.

Product ion mass spectra of (A) metabolite M4 and (B) M5 (Inset: Reconstructed ion chromatogram of m/z 338 (MH + 16)⁺, following 30-min incubation of 3 with rat liver microsomes).

The synthesis of derivative 13 started by base mediated coupling of the activated fluorobenzene 15 and phenolic 16 to yield 17, followed by reduction of the nitro group to the amine and subsequent N-Boc protection (Scheme 3). Basic hydrolysis of 18 at reflux selectively deprotected the thiophenol, which was then allowed to react with epichlorohydrin to generate the epoxide 19. The succeeding steps involved oxidation to the sulfone 20, conversion to the thiirane 21, and final MOM and N-Boc removal by acid reflux to give 13. A typical method to prepare N-hydroxylamine is partial reduction of the nitro compound using zinc dust in the presence of ammonium chloride.³² The transformation of the nitro derivative 22, prepared as published,²⁸ to N-hydroxylamine 14 proceeded without decomposition of the thiirane moiety.

Scheme 3.

Syntheses of 13 and the N-Hydroxylamine 14 Derivative.^a

^a Reagents and condition: (a) Cs₂CO₃, DMF, room temperature, 24 h, 88%. (b) Fe, NH₄Cl, MeOH/H₂O, reflux, 2 h, 83%. (c) Boc₂O, Et₃N, MeOH, room temperature, 24 h, 76%. (d) KOH, MeOH, reflux, 4 h; then epichlorohydrin, room temperature, 10 min, 68%. (e) m-CPBA, CH₂Cl₂, 0 °C to room temperature, 10 min, 89%. (f) thiourea, MeOH/CH₂Cl₂, room temperature, 24 h, 83%. (g) HCl, MeOH, reflux, 1 h, 95%. (h) Zn, NH₄Cl, CH₂Cl₂/H₂O, room temperature, 0.5 h, 69%.

HPLC analysis revealed that M4 had an identical retention time as the synthetic 13 (Figure 5). Identification of M4 as derivative 13 was further confirmed by MS/MS analyses of M4 and the synthetic standards (Figure 6). Compound 13 had the identical product ion mass spectrum as metabolite M4. The MS/MS spectrum of the synthetic N-hydroxylamine 14 showed similar fragments (at m/z 264 and 200) as the metabolite. However, the ions at m/z 321 due to loss of OH from the molecular ion, and at m/z 109, corresponding to the phenylhydroxylamine fragment, established that M4 was not the N-hydroxylamine derivative 14. These data conclusively identified metabolite M4 as compound 13.

Figure 5.

Identification of M4 by comparison of HPLC retention time to synthetic standards. (A) Reconstructed ion chromatogram of m/z 338 following 30-min incubation of 3 with rat liver microsomes and (B) HPLC chromatogram of a mixture of synthetic standards 13 and 14 using UV absorbance at 245 nm.

Figure 6.

Product ion mass spectra of (A) metabolite M4, (B) synthetic standard 13, and (C) synthetic standard 14.

The metabolism pathway for 5d is summarized in Figure 7. Prodrug 5d is hydrolyzed to the active gelatinase inhibitor 3 as the major biotransformation. A minor metabolism pathway of 5d and 3 is oxidation at the α-methylene to give rise to the sulfinic acids M1 and M2. In another minor pathway, the gelatinase inhibitor 3 is oxidized at the terminal phenyl ring to the 3-hydroxy analog 13. N-Oxidation of the aromatic amine of 3 to N-hydroxylamine 14 does not take place.

Figure 7.

Metabolism pathway of 5d. Active gelatinase inhibitors are boxed; the major pathway is represented by a thick arrow, minor pathways are depicted by thin arrows, minute pathways are represented by broken thin arrows. The crossed red arrow indicates the absence of 14.

Exploration of Potential Toxicity

Aromatic amines could be toxic due to oxidation by cytochromes P450 to reactive N-hydroxylamine derivatives, which can be conjugated with DNA and proteins or oxidized to nitrosoarenes, which are reactive entities in their own right.³³ These undesired reactivities could result in toxicity and mutagenicity. While 3 contains an aromatic amine, we know definitively from the foregoing experiments that it is not oxidized to the corresponding N-hydroxylamine derivative 14 in vitro. To confirm that the N-hydroxylamine derivative (compound 14) is not generated in vivo, we dosed mice with 5d or 3 and screened plasma and urine samples for the presence of compound 14. The HPLC and mass spectrometry experiments, along with direct comparison to the authentic synthetic sample, documented the absence of 14 in vivo.

Notwithstanding the total absence of evidence for the existence of the N-hydroxylamine derivative 14, we also investigated the potential mutagenicity of both 5d and 3 by the Ames II mutagenicity test.³⁴ This assay evaluates the mutagenic potential of a compound by measuring its ability to induce reverse mutations at selected loci of Salmonella typhimurium mixed and tester strains in the presence and absence of rat S9 liver activation (see Experimental Section). The mixed strain contains an equimolar mixture of six strains, which are individually designed to revert by only one specific base-pair substitution. Thus, when mixed, all base-pair mutations are represented in one culture. The tester strain is used to detect frame-shift mutations. Both 5d and 3 gave negative (non-mutagenic) responses in the Ames II mutagenecity assay with and without S9 metabolic activation up to concentrations of 1 mg/mL (equivalent to 1.8 and 2.8 mM of 5d and 3, respectively). These results substantiate the non-mutagenic and non-toxic properties of the prodrug 5d, as well as the active gelatinase inhibitor 3.

MMP Selectivity

The arginyl amide prodrug 5d and 3 were evaluated for selectivity towards inhibition of MMPs. Although 5d showed slow-binding behavior against MMP-14_cat, its high K_i of 28.9 ± 2.3 μM essentially renders it inactive against the enzyme (Table 5). Furthermore, 5d was an extremely poor linear competitive inhibitor of MMP-1_cat, MMP-3_cat and MMP-7. As with gelatinases, 3 also showed good slow-binding inhibition of MMP-14_cat with K_i of 0.21 ± 0.02 μM. MMP-14 is the cell surface activator of pro-MMP-2,³⁵ cleaving the N-terminal prodomain of pro-MMP-2 and generating the activated intermediate, which matures into the fully active form of MMP-2.³⁶ However, 3 extremely poorly inhibited MMP-1_cat, MMP-3_cat and MMP-7.

Table 5.

Kinetic Parameters for Inhibition of MMPs

	10−3kon (M−1 s−1)	104koff (s−1)	Ki (μM)
Prodrug 5d
MMP-14a	0.042 ± 0.001	12.0 ± 0.1	28.9 ± 2.3
MMP-1a	-	-	43% inhibition at 1 mM
MMP-3a	-	-	40.5 ± 1.2
MMP-7	-	-	160 ± 14
Compound 3
MMP-14a	6.0 ± 0.5	12.7 ± 0.8	0.21 ± 0.02
MMP-1a	-	-	10% inhibition at 250 μM
MMP-3a	-	-	23.4 ± 1.6
MMP-7	-	-	16% inhibition at 250 μM
Compound 13 (M4)
MMP-2	1.7 ± 0.1	1.4 ± 0.7	0.078 ± 0.043
MMP-9a	1.1 ± 0.1	5.3 ± 1.0	0.49 ± 0.11
MMP-14a	2.7 ± 0.3	7.0 ± 1.6	0.26 ± 0.07
MMP-1a	-	-	11% inhibition at 300 μM
MMP-3a	-	-	54% inhibition at 300 μM
MMP-7	-	-	116 ± 13
M3
MMP-2	1.2 ± 0.3	1.3 ± 0.3	0.11 ± 0.04b
MMP-9	-	-	0.13 ± 0.01b
MMP-14	-	-	0.68 ± 0.05b
MMP-1	-	-	5.4 ± 0.4b
MMP-3	-	-	12.2 ± 0.9b
MMP-7	-	-	39 ± 3b

^aCatalytic domains

^bReported earlier, and given here for the sake of side-by-side comparison.²⁸

As observed for the sulfinic acid and sulfoxide derivatives of 1,²¹ the minor sulfinic acid metabolites of 5d, M1 and M2, as well as the sulfoxide M5, are also expected to be poor inhibitors of gelatinases, as the critical thiirane warhead in these compounds is missing. The N-acetyl metabolite (M3) is, however, a potent inhibitor of gelatinases.²⁸ Compound 13 (M4) was also tested for inhibitory activities toward MMPs. It displayed a similar slow binding inhibition profile as 3, with inhibition constants (K_i) of 0.078 ± 0.043, 0.49 ± 0.11, and 0.26 ± 0.07 μM for MMP-2, MMP-9_cat, and MMP-14_cat, respectively, while sparing MMP-1_cat, MMP-3_cat and MMP-7 (Table 5). The concept of the prodrug strategy is well demonstrated by 5d, which is hydrolyzed in the bloodstream to the active 3, which, in turn, gets further metabolized to two additional minor metabolites, M3 and M4, both of which are potent gelatinase inhibitors in their own right. Note that while M4 was detected as a minor metabolite in vitro, it was not observed in vivo.

Pharmacokinetics

The pharmacokinetics of 5d and 3 were separately evaluated in mice, following single bolus intravenous dose administration at 12.5 mg/kg. Concentrations of 5d, 3, and the p-N-acetyl derivative (M3) are summarized in Table 6. We did not determine the blood levels of the other metabolites, because their concentrations were quite low, including M4 which was not observed in plasma. As predicted from the ex vivo experiments, administration of 5d to mice resulted in hydrolysis to 3, which further metabolized to the p-N-acetyl derivative (M3). Pharmacokinetic parameters are calculated in Table 6.

Table 6.

Pharmacokinetic Parameters of 5d and 3 after a Single Bolus Intravenous Dose to Mice

Time (min)	IV dose of 5d			IV dose of 3
	5d (μM)	3 (μM)	M3 (μM)	3 (μM)	M3 (μM)
0	198	25.7		36.4
2-3	35.1± 9.65	12.3± 3.79	0.0354± 0.0184	18.4± 3.56	0.197± 0.0616
4-5	7.32± 5.76	10.4± 2.31	0.231± 0.0197	13.4± 3.29	0.462± 0.313
7-8	1.28± 0.0544	4.50± 0.361	0.176± 0.0913	5.33± 0.863	0.192± 0.0121
11-12	1.58± 0.423	2.45± 0.664	0.149± 0.0243	3.44± 0.775	0.150± 0.0272
22	0.138± 0.0451	0.787± 0.373	0.0627± 0.0169	-	-
30	-	-	-	1.11± 1.11	0.0958±0.0830
42	0.0872±0.00313	0.207±0.0388	0.0358± 0.0178	-	-
62-66	0.0385±0.00582	0.0908±0.0403	0.0121±0.00181	0.138± 0.0609	0.0273±0.00628
120-123	0.0171±0.00652	0.0248±0.00652	NQa	0.0190±0.00350	0.00277±0.00480
240	0.00830±0.00820	0.0145± 0.00573	NQq	NQa	NQa
AUC0-last (μM · min)	360	137	4.22	221	7.71
AUC0-∞ (μM · min)	363	138	NCb	221	NCb
CL (L/min/kg)	0.0582	-	-	0.0875	-
Vd (L/kg)	0.0978	-	-	0.532	-
t½ α (min)	0.88	2.9	12	3.0	5.1
t½ β (min)	260	69	NCb	21	NCb

^aNQ = not quantifiable;

^bNC = not calculated, since the low levels observed did not allow calculation of terminal half life and AUC_last-∞.

Following intravenous administration of 5d, plasma levels of 5d were initially 198 μM and declined rapidly, with a t_{½ α} (distribution half life) of less than 1 min and a longer elimination half life (t_{½ β}) of >4 h. Systemic exposure (AUC) of 5d was 363 μM·min. The calculated clearance (CL) of 0.0582 L/min/kg was lower than hepatic blood flow of 0.086 L/min/kg,³¹ indicating moderate clearance of the prodrug from systemic circulation. The volume of distribution (Vd) of 0.0978 L/kg (equivalent to 2.30 mL) was significantly lower than total body water of 14.5 mL,³¹ suggesting that 5d was poorly distributed to tissues and remained primarily in the extracellular fluid compartment. The low volume of distribution for the prodrug 5d was attributed to the fact that this compound was designed to be readily hydrolyzed to 3 by amidases present in blood. Thus, 5d is not anticipated to be distributed to tissues and is expected to remain in the extracellular fluid compartment, as it was found experimentally. This is consistent with prodrugs that are readily hydrolyzed in blood to the active drugs,^{37, 38} and have volumes of distribution significantly lower than total body water. For example, fospropofol is a phosphate ester water-soluble prodrug which is readily hydrolyzed to propofol in blood. Fospropofol shows a Vd of 0.25 L/kg in rats and 0.11 L/kg in humans, while propofol has a Vd of 2.3 L/kg in rats and 12.4 L/kg in humans.^{37, 38} Systemic exposure of 3 was 138 μM·min, 2.6-fold lower than that of 5d. The terminal half life of 3 was 69 min or 3-fold shorter than that of 5d. The p-N-acetyl derivative (M3), generated from N-acetylation of 3, had a distribution half life of 12 min and a systemic exposure of 4.22 μM·min, which was significantly lower (33-fold) than systemic exposure of 3. This is consistent with the observation of N-acetylation as a minor metabolism pathway. Mice that received 5d had quantifiable levels of both the prodrug and the released active constituent 3 up to 4 h post administration of 5d.

In contrast, intravenous administration of 3 resulted in initial plasma levels of 36.4 μM and systemic exposure of 221 μM·min. The terminal half life of 21 min was considerably shorter than that of 3 generated from administration of 5d. Clearance of 0.0875 L/min/kg was >70% hepatic blood flow of 0.086 L/min/kg³¹, indicating high clearance of 3. The volume of distribution was 0.532 L/kg (equivalent to 12.5 mL) and approximated total body water of 14.5 mL, indicating that 3 was distributed to tissues. Moreover, the resulting p-N-acetyl metabolite (M3) had an initial half life of 5.1 min, 2.4-fold shorter than that generated after administration of 5d. Systemic exposure of 7.71 μM·min, 29-fold lower than that of 3, further indicated that N-acetylation was a minor metabolism pathway and/or that the p-N-acetyl metabolite was cleared from systemic circulation.

Comparison of the blood levels of 3 after a single bolus intravenous administration of 5d and 3 to mice are shown graphically in Figure 8. More favorable pharmacokinetic properties were observed following intravenous administration of 5d than 3. In particular, the terminal half life of 3 generated from 5d was more than 3-fold longer than that after administration of 3 itself. The in vivo pharmacokinetic data were consistent with the in vitro results, and indicated that 5d had superior pharmacokinetic properties than 3. The >4 hr elimination half life for 5d indicates that this prodrug can be administered as an intravenous infusion, which is the preferred route of administration for the treatment of acute gelatinase dependent diseases, such as stroke, aneurysm, and traumatic brain injury.

Figure 8.

Plasma concentration-time curves of 3 after a single bolus intravenous dose of 5d and 3 to mice.

Plasma levels of 3 remained above the K_i for MMP-2 for 2 h and were higher than the K_i for MMP-9 for 20 min. However, we had previously demonstrated that the mechanism of inhibition of the thiirane class of gelatinase inhibitors is a rate-limiting gelatinase-mediated generation of a thiolate within the enzyme active site.¹² Once the thiolate is generated, it is significantly tighter binding than the parental thiirane, as documented for the prototype 1, which shows a 26-fold more potent inhibition of MMP-2 compared to the thiirane (K_i of 530 pM for the thiolate versus 14 nM for the thiirane).¹² The ability of gelatinases to perform this chemistry within the active site is the reason for the selectivity that this class of inhibitors enjoys in inhibition of gelatinases. Similarly, once 3 is generated from 5d in vivo, we anticipate that on binding to the active site of gelatinases, the same chemistry leads to the formation of the tight-binding thiolate that ties up the enzyme in manifestation of the biological activity of this class of compounds in vivo.

CONCLUSION

Gelatinases are important targets of inhibition in many diseases, in particular acute neurological conditions such as stroke, aneurysm, and traumatic brain injury. For treatment of these diseases, intravenous infusion is the preferred route of administration, which requires that a drug possess a water solubility of 10 mg/mL. We addressed the poor water solubility and extensive metabolism of the prototype thiirane inhibitor 1 by blocking the primary site of metabolism and capitalizing on a prodrug strategy. The prodrugs increased aqueous solubility by more than 5000-fold. The amide prodrugs were quantitatively hydrolyzed in human blood to a potent and selective gelatinase inhibitor. The arginyl amide prodrug 5d was metabolically stable in mouse, rat, and human liver microsomes. The major metabolism pathway of 5d was hydrolysis to the corresponding p-amino derivative 3; minor pathways were N-acetylation, oxidation at the terminal phenyl ring, and oxidation at the α-methylene. Both N-acetylation and hydroxylation at the terminal phenyl ring generated active gelatinase inhibitors. N-Hydroxylation of the aromatic amine was not observed in vitro or in vivo. Both 5d and 3 were non-mutagenic in the Ames II mutagenicity assay. Following a single intravenous bolus administration of 5d to mice, systemic exposures of 5d, 3, and the p-N-acetyl derivative were 363, 138, and 4.22 μM·min, respectively. The prodrug 5d showed moderate clearance of 0.0582 L/min/kg, remained mostly in the extracellular fluid compartment (Vd = 0.0978 L/kg), and had a moderate terminal half life of >4 h. The prodrug 5d had superior pharmacokinetic properties than 3. The high water solubility, chemical stability, quantitative enzymatic hydrolysis of the amide prodrugs, metabolic stability, lack of toxicity, and favorable pharmacokinetic properties make the thiirane class of selective gelatinase inhibitors suitable for intravenous administration in treatment of acute gelatinase-dependent diseases

EXPERIMENTAL SECTION

Chemistry

Organic reagents and solvents were purchased from either Sigma-Aldrich Chemical Company (St. Louis, MO, USA) or Acros Organics (Geel, Belgium), unless otherwise stated and were used without further purification. All reactions were performed under an atmosphere of nitrogen, unless noted otherwise. ¹H and ¹³C NMR spectra were recorded on a Varian INOVA-500 (Varian Inc., Palo Alto, CA, USA), or Varian DirectDrive 600 spectrometer (Varian Inc., Palo Alto, CA, USA). Thin-layer chromatography was performed with Whatman reagents 0.25 mm silica gel 60-F plates. Flash chromatography was carried out with silica gel 60, 230-400 mesh (0.040-0.063 mm particle size) purchased from EM Science (Gibbstown, NJ, USA). High-resolution mass spectra were obtained at the Department of Chemistry and Biochemistry, University of Notre Dame by FAB ionization, using a JEOL AX505HA mass spectrometer (Tokyo, Japan) or ESI ionization, using Bruker micrOTOF/Q2 mass spectrometer (Bruker Daltonik, Bremen, Germany). Purity of the prepared compounds was generally >95%, confirmed by HPLC. Detailed condition is given in HPLC section.

t-Butyl [4-(4-allylthiophenoxy)phenyl]carbamate (9)

Compound 8 (10.3 g, 40.0 mmol), which was prepared by a literature method,²⁸ was dissolved in a mixture of MeOH and triethylamine (7:1, 80 mL), and di-t-butyl dicarbonate (17.5 g, 80 mmol) was added. The resulting solution was stirred for 2 h at 60 °C and was concentrated under reduced pressure. The crude material was purified by column chromatography on silica gel to give the title compound (11.7 g, 82%). ¹H NMR (600 MHz, CDCl₃) δ 1.53 (s, 9H), 3.48 (d, J = 6.7 Hz, 2H), 4.96 - 5.09 (m, 2H), 5.86 (m, J = 16.7, 10.3 Hz, 1H), 6.47 (br. s, 1H), 6.89 (d, J = 8.8 Hz, 2H), 6.96 (d, J = 8.8 Hz, 2H), 7.24 - 7.39 (m, 4H); ¹³C NMR (151 MHz, CDCl₃) δ 28.5, 38.9, 80.7, 117.7, 118.6, 120.3, 120.5, 128.9, 133.3, 134.0, 134.5, 152.2, 153.1, 157.4; HRMS-FAB (m/z): [M]⁺, calcd for C₂₀H₂₃NO₃S, 357.1399; found, 357.1402.

t-Butyl [4-(4-((thiiran-2-yl)methylsulfonyl)phenoxy)phenyl]carbamate (10)

To a solution of compound 9 (10.0 g, 28.0 mmol) in CH₂Cl₂ (100 mL) was added a solution of m-CPBA (31.2 g, 140 mmol, 77%) in an ice water bath. After 72 h, the suspension was filtered and the filtrate was diluted with EtOAc and washed with 10% aqueous sodium thiosulfate, followed by washes with saturated sodium bicarbonate and brine. The organic layer was dried over MgSO₄ and was concentrated. The product was purified by silica gel chromatography to yield the oxirane (7.4 g, 81%) with recovery of some of the allylsulfone derivative (2.1 g, 19%). ¹H NMR (600 MHz, CDCl₃) δ 1.50 (s, 9H), 2.45 (dd, J = 5.0, 1.2 Hz, 1H), 2.79 (dd, J = 4.4, 2.6 Hz, 1H), 3.24 - 3.34 (m, 3H), 6.85 (s, 1H), 6.99 (d, J = 8.8 Hz, 2H), 7.02 (d, 2H), 7.42 (d, J = 8.2 Hz, 2H), 7.84 (d, J = 8.8 Hz, 2H); ¹³C NMR (151 MHz, CDCl₃) δ 28.4, 46.0, 46.0, 59.7, 117.2, 120.5, 121.3, 130.6, 132.2, 136.0, 149.8, 153.0, 163.4; HRMS-FAB (m/z): [M]⁺, calcd for C₂₀H₂₃NO₆S, 405.1246; found, 405.1230.

Thiourea (2.4 g, 31.5 mmol) was added to a solution of the oxirane (6.0 g, 14.8 mmol), obtained above, in a 1:1 mixture of methanol and CH₂Cl₂ (50 mL). The reaction mixture was stirred at room temperature for 18 h. The solvent was removed under reduced pressure. The residue was partitioned between CH₂Cl₂ and water, the organic layer was washed with brine and water, dried (MgSO₄) and the suspension was filtered. Evaporation of solvent gave the crude product, which was purified by column chromatography on silica gel to give thiirane 10 (4.4 g, 72%). ¹H NMR (600 MHz, CDCl₃) δ 1.53 (s, 9H), 2.16 (dd, J = 5.1, 1.6 Hz, 1H), 2.54 (dd, J = 6.2, 1.2 Hz, 1H), 3.06 (dq, J = 7.6, 5.7 Hz, 1H), 3.17 (dd, J = 14.1, 7.9 Hz, 1H), 3.53 (dd, J = 14.1, 5.6 Hz, 1H), 6.61 (br. s, 1H), 7.06 (d, J = 8.8 Hz, 2H), 7.03 (d, J = 8.8 Hz, 2H), 7.44 (d, J = 8.5 Hz, 2H), 7.84 (d, J = 8.8 Hz, 2H); ¹³C NMR (151 MHz, CDCl₃) δ 24.5, 26.3, 28.5, 62.8, 81.0, 117.4, 120.6, 121.4, 130.9, 131.8, 136.0, 150.0, 153.0, 163.6; HRMS-FAB (m/z): [M]⁺, calcd for C₂₀H₂₃NO₅S₂, 421.1018; found, 421.1009.

4-[4-((Thiiran-2-yl)methylsulfonyl)phenoxy]benazamine · HCl salt (3)

The thiirane (10, 4.0 g, 9.5 mmol) was dissolved in a 1:1 mixture of CH₂Cl₂ and ethyl acetate (40 mL) and HCl (10 mL, 4 N in dioxane) was added. The reaction mixture was stirred for 48 h and concentrated under reduced pressure. The resulting crude compound was triturated with diethyl ether and the product was obtained by filtration (2.8 g, 82%). ¹H NMR (500 MHz, D₂O) δ 2.03 (d, J = 4.8 Hz, 1H), 2.42 (d, J = 6.4 Hz, 1H), 2.94 (m, J = 6.4, 6.4 Hz, 1H), 3.42 (dd, J = 14.4, 6.4 Hz, 1H), 3.61 (dd, J = 14.4, 6.4 Hz, 1H), 7.11 (d, J = 8.0 Hz, 2H), 7.16 (d, J = 8.0 Hz, 2H), 7.35 (d, J = 8.0 Hz, 2H), 7.80 (d, J = 8.0 Hz, 2H); ¹³C NMR (126 MHz, D₂O) δ 23.5, 26.0, 61.4, 118.6, 121.9, 125.2, 126.8, 130.8, 131.1, 155.2, 162.5; HRMS-FAB (m/z): [M+H]⁺, calcd for C₁₅H₁₆NO₃S₂, 322.0572; found, 322.0569.

Syntheses of ester prodrugs (4)

Synthesis of the lysine prodrug (4b) is given here as a representative example.

To a solution of 2 (0.20 g, 0.62 mmol) in CH₂Cl₂ (5 mL), which was prepared by a literature method,²³ was added N_α,N_ε-di-Boc-l-lysine hydroxysuccinimide ester (0.41 g, 0.93 mmol) at room temperature. DMAP (76 mg, 0.62 mmol) and diisopropylethyl amine (108 uL, 0.62 mmol) were added and the reaction mixture was stirred for 3 h at room temperature. After concentrating the solution under reduced pressure, the crude product was purified by column chromatography on silica gel to give the desired product (11b, 0.30 g, 74%). The ester (0.30 g, 0.46 mmol) was dissolved in CH₂Cl₂ and ethyl acetate (1:1, 4 mL) and HCl (2 mL, 4 N in dioxane) was added. The reaction mixture was stirred for 48 h and was concentrated under reduced pressure. The resulting crude compound was triturated with diethyl ether and the product (4b) was obtained by filtration (0.18 g, 75%).

Syntheses of amide prodrugs (5)

Synthesis of the lysine prodrug (5b) is shown as a representative example.

i Butylchloroformate (83 μL, 0.53 mmol) was added to a mixture of N_α,N_ε-di-Boc-l-lysine dicyclohexylamine salt (0.28 g, 0.64 mmol) and N methylmorpholine (140 μL, 1.3 mmol) in THF (4 mL) at −15 °C. After stirring for 0.5 h at the same temperature, the suspension of 3 (0.19 g, 0.53 mmol) and N-methylmorpholine (58 μL, 0.53 mmol) in THF (2 mL) was added to the reaction mixture. Stirring was continued for 1 h, while temperature was gradually increased to room temperature. The reaction mixture was diluted with CH₂Cl₂/water and layers were separated. The aqueous layer was washed with CH₂Cl₂ and the combined organic layer was dried (MgSO₄) and evaporated to dryness. The crude material was purified by column chromatography on silica gel to afford the desired product (12b, 0.21 g, 64%). The amide (0.20 g, 0.30 mmol) was dissolved in CH₂Cl₂ and ethyl acetate (1:1, 3 mL) and HCl (2 mL, 4 N in dioxane) was added. The reaction mixture was stirred for 48 h and concentrated under reduced pressure. The resulting crude compound was triturated with diethyl ether and the product (5b) was obtained by filtration (0.11 g, 71%).

Spectral data of Gly ester prodrug compound 11a. ¹H NMR (500 MHz, CDCl₃) δ 1.48 (s, 9H), 2.17 (dd, J = 5.0, 1.8 Hz, 1H), 2.55 (dd, J = 6.2, 1.8 Hz, 1H), 3.04 - 3.10 (m, 1H), 3.20 (dd, J = 14.4, 7.8 Hz, 1H), 3.52 (dd, J = 14.2, 5.6 Hz, 1H), 4.19 (d, J = 5.8 Hz, 2H), 7.11 (d, J = 9.0 Hz, 2H), 7.12 (d, J = 8.8 Hz, 2H), 7.18 (d, J = 9.0 Hz, 2H), 7.88 (d, J = 9.0 Hz, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 24.4, 26.2, 28.5, 42.7, 62.7, 80.5, 117.9, 121.5, 123.3, 131.0, 132.4, 147.3, 152.6, 162.8, 169.3; HRMS-ESI (m/z): [M+H]⁺, calcd for C₂₂H₂₆N₂O₆S₂, 480.1145; found, 480.1145. Compound 4a. ¹H NMR (500 MHz, D₂O) δ 2.17 (dd, J = 5.2, 1.2 Hz, 1H), 2.56 (d, J = 6.2 Hz, 1H), 3.09 (quin, J = 6.2 Hz, 1H), 3.57 (dd, J = 14.6, 7.2 Hz, 1H), 3.74 (dd, J = 14.6, 6.4 Hz, 1H), 4.25 (s, 2H), 7.21 - 7.35 (m, 6H), 7.93 (d, J = 8.8 Hz, 2H); HRMS-ESI (m/z): [M+H]⁺, calcd for C₁₇H₁₈NO₅S₂, 380.0621; found, 380.0597.

Spectral data of Lys ester prodrug compound 11b. ¹H NMR (500 MHz, CDCl₃) δ 1.45 (s, 9H), 1.47 (s, 9H), 1.52 (m, 2H), 1.56 (m, 4H), 1.85 (m, 1H), 1.99 (m, 1H), 2.17 (dd, J = 5.2, 1.8 Hz, 1H), 2.55 (dd, J = 6.2, 1.8 Hz, 1H), 3.07 (m, 1H), 3.17 (m, 2H), 3.19 (dd, J = 14.2, 7.8 Hz, 1H), 3.53 (dd, J = 14.2, 5.6 Hz, 1H), 4.49 (m, 1H), 4.61 (m, 1H), 5.23 (d, J = 7.4 Hz, 1H), 7.10 (d, J = 9.0 Hz, 1H), 7.11 (d, J = 8.8 Hz, 2H), 7.17 (d, J = 9.0 Hz, 1H); ¹³C NMR (126 MHz, CDCl₃) δ 22.7, 24.4, 26.3, 28.5, 28.6, 29.9, 32.1, 40.1, 53.8, 62.8, 79.5, 80.3, 81.3, 117.9, 121.6, 123.4, 131.0, 132.4, 147.6, 152.6, 162.9, 171.8; HRMS-ESI (m/z): [M+H]⁺, calcd for C₃₁H₄₃N₂O₉S₂, 651.2404; found, 651.2414. Compound 4b. ¹H NMR (500 MHz, 10% CDCl₃ in CD₃OD) δ 1.62 - 1.84 (m, 4H), 2.05 - 2.26 (m, 2H), 2.14 (dd, J = 5.0, 1.4 Hz, 1H), 2.52 (dd, J = 6.3, 1.1 Hz, 1H), 3.01 (t, J = 7.6 Hz, 2H), 3.06 (m, 1H), 3.46 (dd, J = 14.6, 7.0 Hz, 1H), 3.54 (t, J = 6.6 Hz, 1H), 4.40 (t, J = 6.5 Hz, 1H), 7.19 (d, J = 9.0 Hz, 2H), 7.23 (d, J = 9.0 Hz, 2H), 7.32 (d, J = 9.2 Hz, 2H), 7.94 (d, J = 9.0 Hz, 2H); HRMS-ESI (m/z): [M+H]⁺, calcd for C₂₁H₂₇N₂O₅S₂, 451.1356; found, 541.1298.

Spectral data of Glu ester prodrug compound 11c. ¹H NMR (500 MHz, CDCl₃) δ 1.45 (s, 18H), 2.09 (m, 1H), 2.15 (dd, J = 5.2, 1.8 Hz, 1H), 2.29 (m, 1H), 2.43 (q, J = 7.4 Hz, 2H), 2.52 (dd, J = 6.2, 1.8 Hz, 1H), 3.05 (m, 1H), 3.18 (dd, J = 14.4, 7.8 Hz, 1H), 3.51 (dd, J = 14.4, 5.8 Hz, 1H), 4.52 (m, 1H), 5.26 (d, J = 8.2 Hz, 1H), 7.08 (d, J = 9.0 Hz, 2H), 7.09 (d, J = 9.0 Hz, 2H), 7.17 (d, J = 9.2 Hz, 2H), 7.86 (d, J = 9.0 Hz, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 24.4, 26.2, 27.3, 28.2, 28.4, 31.7, 53.5, 62.7, 80.4, 81.1, 117.9, 121.4, 123.3, 130.9, 132.3, 147.5, 152.5, 155.6, 162.8, 171.3, 172.1; HRMS-ESI (m/z): [M+H]⁺, calcd for C₂₉H₃₈NO₉S₂, 608.1982; found, 608.1981. Compound 4c. ¹H NMR (500 MHz, 10% CDCl₃ in CD₃OD) δ 2.14 (dd, J = 5.1, 1.3 Hz, 1H), 2.37 (m, 2H), 2.52 (dd, J = 6.2, 1.0 Hz, 1H), 2.68 (td, J = 7.1, 1.6 Hz, 2H), 3.05 (quin, J = 6.2 Hz, 1H), 3.44 (dd, J = 14.4, 7.0 Hz, 1H), 3.54 (dd, J = 14.4, 6.4 Hz, 1H), 4.43 (t, J = 6.8 Hz, 1H), 7.19 (d, J = 8.8 Hz, 2H), 7.22 (d, J = 9.2 Hz, 2H), 7.32 (d, J = 9.0 Hz, 2H), 7.93 (d, J = 9.0 Hz, 2H); HRMS-ESI (m/z): [M+H]⁺, calcd for C₂₀H₂₂NO₇S₂, 452.0832: found, 452.0769.

Spectral data of Arg ester prodrug compound 11d. ¹H NMR (500 MHz, CDCl₃) δ 1.46 (s, 9H), 1.50 (s, 9H), 1.77 (m, 1H), 1.94 (m, 1H), 2.13 (dd, J = 5.0, 1.6 Hz, 1H), 2.51 (dd, J = 6.2, 1.6 Hz, 1H), 3.03 (m, 1H), 3.17 (dd, J = 14.2, 7.8 Hz, 1H), 3.49 (dd, J = 14.2, 5.6 Hz, 1H), 3.84 (m, 1H), 3.96 (m, 1H), 4.50 (m, 1H), 5.83 (d, J = 8.2 Hz, 1H), 7.07 (d, J = 9.0 Hz, 2H), 7.08 (d, J = 8.8 Hz, 2H), 7.14 (d, J = 9.0 Hz, 1H), 7.85 (d, J = 9.0 Hz, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 24.3, 26.2, 28.1, 28.3, 28.5, 44.1, 53.5, 60.5, 62.7, 84.1, 117.8, 121.4, 123.4, 130.9, 132.3, 147.7, 152.4, 155.0, 155.6, 160.6, 162.8, 163.6, 171.2; HRMS-ESI (m/z): [M+H]⁺, calcd for C₃₆H₅₁N₄O₁₁S₂, 779.2990; found, 779.3001. Compound 4d. ¹H NMR (500 MHz, 10% CDCl₃ in CD₃OD) δ 1.78 - 2.26 (m, 4H), 2.14 (dd, J = 5.1, 1.3 Hz, 1H), 2.52 (dd, J = 6.2, 0.8 Hz, 1H), 3.04 (quin, J = 6.2 Hz, 1H), 3.27 (m, 2H), 3.43 (dd, J = 14.4, 7.0 Hz, 1H), 3.52 (dd, J = 14.8, 6.6 Hz, 1H), 4.39 (t, J = 6.5 Hz, 1H), 7.17 (d, J = 9.0 Hz, 1H), 7.21 (d, J = 9.0 Hz, 1H), 7.30 (d, J = 9.0 Hz, 2H), 7.92 (d, J = 8.8 Hz, 2H); HRMS-ESI (m/z): [M+H]⁺, calcd for C₂₁H₂₇N₄O₅S₂, 479.1417; found, 479.1423.

Spectral data of Arg Arg ester prodrug compound 11e. ¹H NMR (500 MHz, CDCl₃) δ 1.39 - 1.53 (m, 45H), 1.54 - 2.11 (m, 8H), 2.16 (dd, J = 5.2, 1.4 Hz, 1H), 2.54 (d, J = 5.2 Hz, 1H), 3.06 (m, 1H), 3.19 (dd, J = 14.2, 7.8 Hz, 1H), 3.47 (dd, J = 12.8, 7.0 Hz, 2H), 3.52 (dd, J = 14.4, 5.6 Hz, 1H), 3.74 - 3.95 (m, 2H), 4.32 (dd, J = 14.8, 6.4 Hz, 1H), 4.77 (m, 1H), 5.92 (d, J = 7.8 Hz, 1H), 7.07 - 7.12 (m, 4H), 7.16 (d, J = 9.0 Hz, 2H), 7.33 (d, J = 7.6 Hz, 1H), 7.87 (d, J = 8.8 Hz, 2H), 8.38 (t, J = 4.9 Hz, 1H), 9.30 (br. s, 2H), 11.50 (s, 1H); ¹³C NMR (126 MHz, CDCl₃) δ 24.4, 24.9, 25.6, 26.2, 28.2, 28.2, 28.5, 28.6, 29.3, 40.4, 44.1, 52.5, 54.3, 62.8, 79.5, 80.2, 83.4, 83.5, 84.2, 117.9, 121.5, 121.5, 123.4, 130.8, 131.0, 132.3, 147.5, 152.5, 153.4, 155.0, 156.4, 160.8, 160.9, 162.9, 163.7, 170.6, 172.7; HRMS-ESI (m/z): [M+H]⁺, calcd for C₅₂H₇₉N₈O₁₆S₂, 1135.5050; found, 1135.5039. Compound 4e. ¹H NMR (500 MHz, 10% CDCl₃ in CD₃OD) δ 1.71 - 2.23 (m, 8H), 2.16 (dd, J = 5.0, 1.2 Hz, 1H), 2.54 (dd, J = 6.2, 0.8 Hz, 1H), 3.03 - 3.10 (m, 1H), 3.23 - 3.32 (m, 4H), 3.47 (dd, J = 14.4, 7.0 Hz, 1H), 3.54 (dd, J = 14.2, 6.4 Hz, 1H), 4.16 (t, J = 6.4 Hz, 1H), 4.72 (dd, J = 9.1, 4.5 Hz, 1H), 7.15 - 7.23 (m, 4H), 7.25 (d, J = 9.0 Hz, 2H), 7.94 (d, J = 8.8 Hz, 2H); HRMS-ESI (m/z): [M+H]⁺, calcd for C₂₇H₃₉N₈O₆S₂, 635.2428; found, 635.2449.

Spectral data of Gly amide prodrug compound 12a. ¹H NMR (500 MHz, CDCl₃) δ 1.37 (s, 9H), 2.05 (dd, J = 5.2, 1.8 Hz, 1H), 2.44 (dd, J = 6.2, 1.8 Hz, 1H), 2.94 (m, 1H), 3.14 (dd, J = 14.4, 7.6 Hz, 1H), 3.41 (dd, J = 14.4, 6.0 Hz, 1H), 3.80 (d, J = 5.2 Hz, 2H), 3.82 (s, 1H), 6.96 (d, J = 15.2 Hz, 2H), 6.98 (d, J = 15.4 Hz, 2H), 7.52 (d, J = 8.8 Hz, 2H), 7.75 (d, J = 9.0 Hz, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 24.0, 25.9, 28.2, 44.2, 62.5, 80.4, 117.4, 121.0, 121.8, 130.7, 131.4, 150.7, 163.3, 168.3, 172.9; HRMS-ESI (m/z): [M+Na]⁺, calcd for C₂₂H₂₆N₂O₆S₂Na, 501.1124; found, 501.1112. Compound 5a. ¹H NMR (500 MHz, 10% CDCl₃ in CD₃OD) δ 2.14 (dd, J = 5.2, 1.6 Hz, 1H), 2.51 (dd, J = 6.3, 1.5 Hz, 1H), 3.04 (m, J = 7.0, 5.2 Hz, 1H), 3.42 (dd, J = 14.4, 7.0 Hz, 1H), 3.53 (dd, J = 14.4, 6.4 Hz, 1H), 4.14 (s, 2H), 7.17 (d, J = 9.0, 2.0, 1.8 Hz, 1H), 7.20 (d, J = 9.2 Hz, 2H), 7.29 (d, J = 9.2 Hz, 2H), 7.92 (d, J = 9.2 Hz, 1H); HRMS-ESI (m/z): [M+H]⁺, calcd for C₁₇H₁₉N₂O₄S₂, 379.0781; found, 379.0805.

Spectral data of Lys amide prodrug compound 12b. ¹H NMR (500 MHz, CDCl₃) δ 1.41 (s, 9H), 1.43 (s, 9H), 2.02 (s, 3H), 2.13 (dd, J = 5.0, 1.6 Hz, 1H), 2.50 (dd, J = 6.2, 1.6 Hz, 1H), 3.02 (m, 1H), 3.09 (m, 2H), 3.16 (dd, J = 14.4, 7.8 Hz, 1H), 3.50 (dd, J = 14.4, 5.8 Hz, 1H), 4.28 (m, 1H), 4.75 (m, 1H), 5.53 (d, J = 7.4 Hz, 1H), 6.93 (d, J = 8.4 Hz, 2H), 7.00 (d, J = 8.6 Hz, 2H), 7.55 (d, J = 8.8 Hz, 2H), 7.82 (d, J = 9.0 Hz, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 24.4, 26.2, 28.5, 28.6, 29.8, 31.9, 39.8, 55.3, 62.7, 79.3, 80.5, 117.4, 121.1, 121.7, 130.8, 131.8, 135.6, 150.5, 156.4, 156.6, 163.3, 171.2; HRMS-ESI (m/z): [M+Na]⁺, calcd for C₃₁H₄₃N₃O₈S₂Na, 672.2384; found, 672.2393. Compound 5b. ¹H NMR (500 MHz, 10% CDCl₃ in CD₃OD) δ 1.59 - 1.86 (m, 4H), 2.06 - 2.25 (m, 2H), 2.14 (dd, J = 5.2, 1.4 Hz, 1H), 2.52 (dd, J = 6.4, 1.4 Hz, 1H), 3.00 (t, J = 7.6 Hz, 2H), 3.05 (m, 1H), 3.45 (dd, J = 14.4, 6.8 Hz, 1H), 3.53 (dd, J = 14.6, 6.6 Hz, 1H), 4.39 (t, J = 6.5 Hz, 1H), 7.18 (d, J = 8.8 Hz, 2H), 7.22 (d, J = 9.2 Hz, 2H), 7.32 (d, J = 9.0 Hz, 2H), 7.93 (d, J = 9.0 Hz, 2H); HRMS-ESI (m/z): [M+H]⁺, calcd for C₂₁H₂₈N₃O₄S₂, 450.1516; found, 450.1485.

Spectral data of Glu amide prodrug compound 12c. ¹H NMR (500 MHz, CDCl₃) δ 1.47 (s, 9H), 1.49 (s, 9H), 2.00 (m, 1H), 2.17 (dd, J = 5.2, 1.8 Hz, 1H), 2.41 (m, 1H), 2.56 (m, 2H), 3.07 (m, 1H), 3.18 (dd, J = 14.3, 7.9 Hz, 1H), 3.54 (dd, J = 14.1, 5.5 Hz, 1H), 4.27 (m, 1H), 7.07 (d, J = 8.6 Hz, 2H), 7.08 (d, J = 9.0 Hz, 2H), 7.61 (d, J = 9.0 Hz, 1H), 7.86 (d, J = 9.0 Hz, 1H); ¹³C NMR (126 MHz, CDCl₃) δ 24.5, 26.3, 28.2, 28.5, 37.1, 51.5, 62.8, 82.3, 117.6, 121.3, 121.9, 130.9, 131.9, 135.2, 151.0, 163.3, 169.3, 171.7; HRMS-ESI (m/z): [M+Na]⁺, calcd for C₂₉H₃₈N₂O₈S₂Na, 629.1962: found, 629.1969. Compound 5c. ¹H NMR (500 MHz, 10% CDCl₃ in CD₃OD) δ 2.32 (m, 1H), 2.42 (m, 1H), 2.51 (dd, J = 6.2, 1.4 Hz, 1H), 2.68 (dt, J = 7.2, 6.8, 2.0 Hz, 2H), 3.04 (m, 1H), 3.44 (dd, J = 14.4, 7.0 Hz, 1H), 3.53 (dd, J = 14.4, 6.4 Hz, 1H), 4.43 (t, J = 6.8 Hz, 1H), 4.87 (s, 10H), 7.19 (d, J = 9.2 Hz, 2H), 7.22 (d, J = 9.2 Hz, 2H), 7.32 (d, J = 9.2 Hz, 2H); HRMS-ESI (m/z): [M+H]⁺, calcd for C₂₀H₂₃N₂O₆S₂, 451.0992; found, 451.1004.

Spectral data of Arg amide prodrug compound 12d. ¹H NMR (500 MHz, CDCl₃) δ 1.38 (s, 9H), 1.48 (s, 9H), 1.53 (s, 9H), 1.59 - 1.96 (m, 4H), 2.16 (dd, J = 5.2, 1.8 Hz, 1H), 2.54 (dd, J = 6.2, 1.6 Hz, 1H), 3.06 (m, 1H), 3.17 (dd, J = 14.2, 8.0 Hz, 1H), 3.53 (dd, J = 14.2, 5.6 Hz, 1H), 3.72 (m, 1H), 4.12 (s, 1H), 4.55 (m, 1H), 5.99 (d, J = 8.2 Hz, 1H), 7.05 (d, J = 9.0 Hz, 2H), 7.07 (d, J = 8.8 Hz, 2H), 7.53 (d, J = 8.6 Hz, 2H), 7.85 (d, J = 9.0 Hz, 2H), 9.13 (br. s, 1H); ¹³C NMR (126 MHz, CDCl₃) δ 24.5, 24.7, 26.3, 28.2, 28.6, 29.3, 44.1, 54.0, 62.8, 84.6, 117.6, 121.2, 123.4, 130.9, 132.0, 135.1, 151.3, 155.0, 161.3, 163.2, 163.4, 171.1; HRMS-ESI (m/z): [M+H]⁺, calcd for C₃₆H₅₂N₅O₁₀S₂, 778.3150; found, 778.3154. Compound 5d. ¹H NMR (500 MHz, 10% CDCl₃ in CD₃OD) δ 1.71 - 2.11 (m, 4H), 2.16 (d, J = 5.2 Hz, 1H), 2.54 (dd, J = 6.2, 1.2 Hz, 1H), 3.06 (quin, J = 6.2 Hz, 1H), 3.29 (m, 1H), 3.45 (m, 2H), 3.54 (dd, J = 14.6, 6.4 Hz, 1H), 4.14 (m, 1H), 7.15 (d, J = 9.0 Hz, 2H), 7.15 (d, J = 9.0 Hz, 2H), 7.77 (d, J = 9.0 Hz, 2H), 7.92 (d, J = 8.8 Hz, 2H); HRMS-ESI (m/z): [M+H]⁺, calcd for C₂₁H₂₈N₅O₄S₂, 478.1577; found, 478.1580.

Spectral data of Arg Arg amide prodrug compound 12e. ¹H NMR (500 MHz, CDCl₃) δ 1.40 - 1.54 (5 × s, 45H), 1.46 - 1.90 (m, 8H), 2.16 (dd, J = 5.2, 1.8 Hz, 1H), 2.53 (dd, J = 6.2, 1.4 Hz, 1H), 3.05 (m, 1H), 3.16 (dd, J = 14.2, 7.8 Hz, 1H), 3.39 (m, 1H), 3.53 (dd, J = 14.2, 5.6 Hz, 1H), 3.55 (m, 2H), 3.82 (t, J = 6.8 Hz, 2H), 4.28 (dd, J = 13.8, 6.4 Hz, 1H), 4.59 (q, J = 7.4 Hz, 1H), 6.10 (d, J = 7.4 Hz, 1H), 7.03 (d, J = 9.0 Hz, 2H), 7.05 (d, J = 8.8 Hz, 2H), 7.22 (d, J = 6.4 Hz, 1H), 7.61 (d, J = 8.4 Hz, 2H), 7.84 (d, J = 9.0 Hz, 2H), 8.41 (t, J = 5.6 Hz, 1H), 8.84 (s, 1H), 9.26 (br. s, 1H), 9.35 (br. s, 1H), 11.47 (s, 1H); ¹³C NMR (126 MHz, CDCl₃) δ 22.8, 24.5, 25.4, 26.1, 26.2, 28.2, 28.2, 28.4, 28.4, 28.6, 29.2, 31.8, 34.8, 40.1, 44.1, 53.8, 55.0, 62.8, 79.6, 80.6, 83.4, 84.3, 117.5, 121.2, 122.5, 130.9, 131.8, 135.4, 150.9, 153.4, 154.9, 156.6, 160.8, 163.4, 163.4, 163.5, 169.6; HRMS-ESI (m/z): [M+H]⁺, calcd for C₅₂H₈₀N₉O₁₅S₂, 1134.5210; found, 1134.5215. Compound 5e. ¹H NMR (500 MHz, 10% CDCl₃ in CD₃OD) δ 1.68 - 2.05 (m, 8H), 2.13 (dd, J = 5.1, 1.3 Hz, 1H), 2.51 (dd, J = 6.2, 1.0 Hz, 1H), 3.17 (m, 1H), 3.22 - 3.28 (m, 4H), 3.42 (dd, J = 14.6, 7.0 Hz, 1H), 3.50 (dd, J = 14.0, 7.0 Hz, 1H), 4.10 (t, J = 6.3 Hz, 1H), 4.52 (dd, J = 8.7, 5.1 Hz, 1H), 7.10 (t, J = 8.9 Hz, 4H), 7.66 (d, J = 9.0 Hz, 2H), 7.89 (d, J = 9.0 Hz, 2H); HRMS-ESI (m/z): [M+H]⁺, calcd for C₂₇H₄₀N₉O₅S₂, 634.2588; found, 634.2585.

4-Fluoro-2-(methoxymethoxy)-1-nitrobenzene (15)

5 Fluoro 2 nitrophenol (10 g, 63 mmol, 99%) was dissolved in anhydrous DMF (100 mL) in an oven-dried flask. The solution was stirred under an atmosphere of nitrogen and cooled in an ice-water bath. Sodium hydride (3.0 g, 75 mmol) was added with stirring, after which chloromethyl methyl ether (5.3 g, 66 mmol) was added dropwise. The resulting mixture was aged at room temperature for 1 h. The reaction was quenched with MeOH, diluted with ether and washed with water and brine. The ether layer was dried over anhydrous Na₂SO₄ and concentrated in vacuo. Purification of the product by silica gel chromatography (hexanes/EtOAc = 1/7) gave the title compound in 88% yield. ¹H NMR (500 MHz, CDCl₃) δ 3.53 (s, 3H), 5.30 (d, J = 2.0 Hz, 2H), 6.78 (dd, J = 7.3, 9.1 Hz, 1H), 7.06 (dt, J = 2.3, 10.4 Hz, 1H), 7.91 (dd, J = 6.0, 9.0 Hz, 1 H); ¹³C NMR (126 MHz, CDCl₃) δ 57.0, 95.6, 105.0 (d, J = 26.3 Hz), 108.8 (d, J = 23.0 Hz), 127.8 (d, J = 10.7 Hz), 152.8 (d, J = 11.5 Hz), 165.5 (d, J = 255.9 Hz); HRMS-FAB (m/z): [M+Na]⁺, calcd for C₈H₈FNNaO₄, 224.0330, found 224.0328.

S-(4-(3-(Methoxymethoxy)-4-nitrophenoxy)phenyl) dimethylcarbamothioate (17)

To a round bottom flask was added dimethylcarbamothioate 16 (2.1 g, 11 mmol), compound 15 (2.5 g, 12 mmol), DMF (50 mL) and Cs₂CO₃ (7.1 g, 22 mmol). The resulting mixture was stirred for 24 h, followed by filtration through a layer of silica gel. The solvent was evaporated and the product was purified by silica gel chromatography (hexanes/EtOAc = 2/1) to give 17 as a solid in 88% (3.5 g). ¹H NMR (500 MHz, CDCl₃) δ 3.05 (br. s., 3H), 3.12 (br. s., 3H), 3.53 (s, 3H), 5.28 (s, 2H), 6.61 (ddd, J = 0.8, 2.5, 9.1 Hz, 1H), 6.99 (d, J = 2.4 Hz, 1H), 7.02 - 7.12 (m, 2H), 7.50 - 7.56 (m, 2H), 7.88 - 7.93 (m, 1 H); ¹³C NMR (126 MHz, CDCl₃) δ 37.1, 37.2, 57.1, 95.5, 107.1, 110.4, 120.5, 125.1, 127.9, 135.7, 137.9, 153.0, 156.1, 162.0, 166.9; HRMS-FAB (m/z): [M+H]⁺, calcd for C₁₇H₁₉N₂O₆S, 379.0958, found 379.0945.

t-Butyl (4-(4-((dimethylcarbamoyl)thio)phenoxy)-2-(methoxymethoxy)phenyl)carbamate (18)

A solution of 17 (2.5 g, 6.6 mmol) in MeOH/H₂O (30 mL/15 mL) was treated with elemental iron (1.9 g, 34 mmol) and NH₄Cl (0.35 g, 6.5 mmol) and the resulting mixture was heated at reflux for 2 h. The crude reaction mixture was filtered through Celite and the filtrate was concentrated in vacuo. The residue was dissolved in EtOAc and was washed with water. The organic layer was separated and the aqueous layer was washed with EtOAc. The combined organic portions were dried over anhydrous Na₂SO₄. The solvent was removed under reduced pressure to provide the amine as a solid (1.9 g, 5.5 mmol, 83%). This product was dissolved in MeOH (30 mL), was treated with TEA (0.6 g, 5.9 mmol) and Boc₂O (1.9 g, 8.4 mmol, 97%) and was stirred at room temperature for 24 h. The solvent was removed under reduced pressure and the product was purified by silica gel chromatography (hexanes/EtOAc = 3/1) to give the title compound (1.9 g, 76%). ¹H NMR (500 MHz , CDCl₃) δ 1.54 (s, 9H), 3.02 (br. s., 3H), 3.08 (br. s., 3H), 3.48 (s, 3H), 5.18 (s, 2H), 6.71 (dd, J = 2.6, 8.8 Hz, 1H), 6.90 (d, J = 2.6 Hz, 1H), 6.92 - 6.97 (m, 2H), 7.00 (br. s., 1H), 7.36 - 7.42 (m, 2H), 8.05 (br. s., 1 H); ¹³C NMR (126 MHz, CDCl₃) δ 28.4, 36.9, 56.5, 80.5, 95.3, 107.2, 113.6, 117.7, 119.3, 121.6, 125.3, 137.4, 146.5, 150.6, 152.9, 159.4, 167.4; HRMS-FAB (m/z): [M+H]⁺, calcd for C₂₂H₂₉N₂O₆S, 449.1741, found 449.1735.

t-Butyl (2-(methoxymethoxy)-4-(4-((oxiran-2-ylmethyl)thio)phenoxy)phenyl)carbamate (19)

To a solution of compound 18 (0.65 g, 1.5 mmol) in anhydrous MeOH (10 mL) was added potassium hydroxide (0.42 g, 7.5 mmol). The mixture was refluxed for 4 h and then cooled to room temperature. Epichlorohydrin (0.21 g, 2.3 mmol) was added dropwise and the mixture was stirred at room temperature for 15 min, after which the solvent was removed under reduced pressure. The concentrate was diluted with water and extracted with EtOAc. The combined organic extracts were dried over anhydrous Na₂SO₄, the suspension was filtered, followed by concentration of the filtrate in vacuo. Purification by silica gel chromatography (hexanes/EtOAc = 4/1) gave 19 in 68% yield (0.43 g) ¹H NMR (500 MHz , CDCl₃) δ 1.53 (s, 9H), 2.47 (dd, J = 2.6, 5.0 Hz, 1H), 2.77 (t, J = 4.4 Hz, 1H), 2.85 (dd, J = 6.1, 13.9 Hz, 1H), 3.05 - 3.12 (m, 1H), 3.12 - 3.18 (m, 1H), 5.18 (s, 2H), 3.48 (s, 3H), 6.66 (dd, J = 2.6, 8.8 Hz, 1H), 6.85 - 6.92 (m, 3H), 6.98 (br. s., 1H), 7.37 - 7.44 (m, 2 H), 8.03 (br. s., 1H); ¹³C NMR (126 MHz, CDCl₃) δ 28.5, 38.2, 47.6, 51.3, 56.5, 80.6, 95.4, 106.8, 113.1, 118.4, 119.4, 125.1, 128.0, 131.8, 133.8, 146.6, 151.2, 152.9, 157.9; HRMS-FAB (m/z): [M+H]⁺, calcd for C₂₂H₂₈NO₄S, 434.1632, found 434.1637.

t-Butyl (2-(methoxymethoxy)-4-(4-((oxiran-2-ylmethyl)sulfonyl)phenoxy)phenyl)carbamate (20)

Compound 19 (0.44 g, 1.0 mmol) was dissolved in CH₂Cl₂ (10 ml) and cooled in an ice-water bath. It was treated with m-CPBA (0.48 g, 2.1 mmol, 77%) and the resulting white suspension was stirred at room temperature for 10 min. The mixture was filtered and the filtrate was washed with saturated Na₂S₂O₃, followed by saturated NaHCO₃. The organic layer was separated and the aqueous layer was washed with CH₂Cl₂. The combined organic solution was dried over anhydrous Na₂SO₄, the suspension was filtered and the filtrate was concentrated in vacuo. Purification of the product by silica gel chromatography (hexanes/EtOAc = 2/1) gave the title compound in 89% yield (0.42 g). ¹H NMR (500 MHz, CDCl₃) δ 1.54 (s, 9H), 2.48 (dd, J = 2.2, 4.8 Hz, 1H), 2.82 (t, J = 4.5 Hz, 1H), 3.19 - 3.28 (m, 1H), 3.28 - 3.38 (m, 2H), 3.49 (s, 3H), 5.20 (s, 2H), 6.73 (dd, J = 2.6, 8.8 Hz, 1H), 6.91 (d, J = 2.6 Hz, 1H), 7.03 (s, 1H), 7.04 - 7.09 (m, 2H), 7.83 - 7.89 (m, 2H), 8.13 (br. s., 1 H); ¹³C NMR (126 MHz, CDCl₃) δ 28.5, 46.0, 46.1, 56.6, 59.8, 80.9, 95.4, 107.5, 114.2, 117.2, 119.5, 126.3, 130.6, 132.3, 146.7, 149.4, 152.9, 163.5; HRMS-FAB (m/z): [M+H]⁺, calcd for C₂₂H₂₈NO₈S, 466.1530, found 466.1543.

t-Butyl (2-(methoxymethoxy)-4-(4-((thiiran-2-ylmethyl)sulfonyl)phenoxy)phenyl)carbamate (21)

To a solution of 20 (0.40 g, 0.86 mmol) in CH₂Cl₂ (5 mL) was added a mixture of thiourea (0.10 g, 1.3 mmol, 99%) in methanol (5 mL). The resulting mixture was stirred for 24 h at room temperature, after which the solvent was removed under reduced pressure. The residue was partitioned between CH₂Cl₂ and water. The organic layer was dried over anhydrous Na₂SO₄ and was filtered. Evaporation of solvent and purification by silica gel chromatography (hexanes/EtOAc = 4/1) gave 0.34 g of 21 (83%). ¹H NMR (500 MHz, CDCl₃) δ 1.53 (s, 9H), 2.14 (dd, J = 1.8, 5.0 Hz, 1H), 2.52 (dd, J = 1.5, 5.9 Hz, 1H), 3.04 (dq, J = 5.6, 7.8 Hz, 1H), 3.15 (dd, J = 8.0, 14.2 Hz, 1H), 3.47 (s, 3H), 3.51 (dd, J = 5.6, 14.4 Hz, 1H), 5.19 (s, 2H), 6.72 (dd, J = 2.6, 8.8 Hz, 1H), 6.91 (d, J = 2.6 Hz, 1H), 7.01 - 7.08 (m, 3H), 7.80 - 7.85 (m, 2H), 8.08 - 8.16 (m, 1 H); ¹³C NMR (126 MHz, CDCl₃) δ 24.4, 26.2, 28.5, 56.6, 62.7, 80.8, 95.3, 107.3, 114.1, 117.3, 119.4, 126.2, 130.8, 131.6, 146.6, 149.3, 152.8, 163.5; HRMS-FAB (m/z): [M+H]⁺, calcd for C₂₂H₂₈NO₇S₂, 482.1302, found 482.1301.

2-Amino-5-(4-((thiiran-2-ylmethyl)sulfonyl)phenoxy)phenol (13)

To a solution of compound 21 (0.10 g, 0.21 mmol) in methanol (5 mL) was added a few drops of conc. HCl. After the mixture was stirred at reflux for 1 h, the solvent was evaporated in vacuo. The crude product was taken up into water and EtOAc. Layers were separated and the aqueous layer was washed with EtOAc. The combined organic layers were dried over anhydrous Na₂SO₄, the suspension was filtered and the filtrate was concentrated in vacuo. Purification by silica gel chromatography (hexanes/EtOAc = 1/1) gave 13 in 95% yield (66 mg). ¹H NMR (500MHz, CDCl₃) δ 2.14 (dd, J = 1.3, 4.9 Hz, 1H), 2.18 (d, J = 1.0 Hz, 1H), 2.49 - 2.56 (m, 1H), 3.00 - 3.08 (m, 1H), 3.17 (dd, J = 7.8, 14.4 Hz, 1H), 3.51 (dd, J = 5.6, 14.2 Hz, 2H), 6.46 - 6.54 (m, 2H), 6.73 - 6.80 (m, 1H), 6.97 - 7.07 (m, 2H), 7.80 (m, 2 H); ¹³C NMR (126MHz, CDCl₃) δ 24.4, 26.3, 62.8, 108.5, 113.1, 117.2, 117.9, 130.7, 131.1, 131.9, 145.9, 147.3, 164.1; HRMS-FAB (m/z): [M+H]⁺, calcd for C₁₅H₁₆NO₄S₂, 338.0515, found 338.0507.

N-(4-(4-((Thiiran-2-ylmethyl)sulfonyl)phenoxy)phenyl)hydroxylamine (14)

Compound 22 (75 mg, 0.21 mmol), prepared as described previously,²⁸ in CH₂Cl₂/H₂O (0.2 mL/0.1mL) was treated with Zn (50 mg, 0.76 mmol) and NH₄Cl (25 mg, 0.47 mmol). The resulting mixture was stirred for 30 min at room temperature. It was then filtered over Celite and the filtrate was concentrated in vacuo. Purification by silica gel chromatography (CH₂Cl₂) gave 14 in 69% yield (50 mg). ¹H NMR (500 MHz, CDCl₃) δ 2.16 (dd, J = 1.8, 5.2 Hz, 1H), 2.53 (dd, J = 1.0, 6.2 Hz, 1H), 3.01 - 3.09 (m, 1H), 3.17 (dd, J = 7.9, 14.3 Hz, 1H), 3.53 (dd, J = 5.8, 14.2 Hz,1H), 6.98 - 7.10 (m, 6H), 7.81 - 7.88 (m, 2 H); ¹³C NMR (126 MHz, CDCl₃) δ 24.5, 26.3, 62.8, 116.5, 117.3, 121.6, 130.9, 131.5, 147.4, 149.4, 163.8; HRMS-FAB (m/z): [M+H]⁺, calcd for C₁₅H₁₆NO₄S₂, 338.0515, found 338.0501.

Solubility Determination and Solution Stability

The water solubilities of the prodrugs were measured at room temperature by using a 10 mg/mL solution of each compound in water. The molar extinction coefficient was calculated at 245 nm. The aqueous solutions were centrifuged and the absorbance at 245 nm of the supernatant was recorded. The concentration was calculated using Beer’s Law. Assays were performed in triplicate. The ester and amide prodrugs have solubilities of over 10 mg/mL. Over a period of one month, appropriately diluted aliquots were analyzed by HPLC for the presence of the prodrug and the active gelatinase inhibitors (compound 2 from the esters or compound 3 from the amides).

Ex vivo Hydrolyses of the Amide Prodrugs

Fresh blood was collected in heparinized syringes from human volunteers. Blood collection from human volunteers was approved by the University of Notre Dame Institutional Review Board. The blood was centrifuged to obtain plasma. The stability of the prodrugs in plasma and whole blood (1 mL) was determined by spiking plasma or blood with 10 μL of 10 mM of prodrug and incubating at 37 °C. Aliquots were drawn periodically and processed for analysis. For the plasma samples, the reaction was immediately terminated by the addition of two volumes of acetonitrile. For the blood samples, aliquots were first centrifuged at 4 °C, and the resultant plasma was treated with two volumes of acetonitrile. The precipitated protein was centrifuged (10 min, 10 500 g) and the supernatant was analyzed by reversed phase HPLC with UV detection. First-order rate constants for the disappearance of the prodrugs were calculated, from which the half-lives were determined.

Enzyme Inhibition Studies

Human recombinant active MMP-2 and MMP-7 and catalytic domains of MMP-3 and MMP-14 were purchased from EMD Biosciences (La Jolla, CA, USA). The catalytic domains of human recombinant MMP-1 and MMP-9 were from Biomol International (Plymouth Meeting, PA, USA). Fluorogenic substrate MOCAcPLGLA₂pr(Dnp)AR-NH₂ was purchased from Peptides International (Louisville, KY, USA); MOCAcRPKPVE(Nva)WRK(Dnp)-NH₂ and (Dnp)P(Cha)GC(Me)HAK(NMa)NH₂ were obtained from R&D Systems (Minneapolis, MN, USA). The K_m values used for the reaction of MMP-2, MMP-9 and MMP-14 with the fluorogenic substrate MOCAcPLGLA₂pr(Dnp)AR-NH₂ were 18.9 ± 1.0, 5.0 ± 0.1, 5.6 ± 0.4 μM, respectively. Inhibitor stock solutions (10 mM) were prepared in water (except for compound 13, which was prepared in DMSO). The methodology for enzyme inhibition and assays was the same as reported previously.²⁸ Substrate hydrolysis was measured with a Cary Eclipse fluorescence spectrophotometer (Varian, Walnut Creek, CA, USA).

Microsomal Incubations

Mouse, rat, and human liver microsomes and rat S9 were purchased from BD Biosciences (Woburn, MA, USA). Incubations consisted of liver microsomes (1.0 mg), NADPH (0.5 mM, final concentration), and 100 μM compound (final concentration) in potassium phosphate buffer (100 mM, pH 7.4) at 37 °C in a total volume of 1 mL. Aliquots were drawn at different time points and the reactions were terminated by addition of one volume of acetonitrile containing 25 μM (S)-2-((S)-1-((4-phenoxyphenyl)sulfonyl)ethyl)thiirane, prepared as published,²² as internal standard. The precipitated protein was centrifuged (10 min, 10 500 g) and the supernatant was analyzed by reversed-phase HPLC with UV detection.

Apparent Intrinsic Clearance (CL_int,app) Determination in Microsomes

Apparent intrinsic clearance was calculated by the following equation:³⁹

$${\mathrm{CL}}_{\mathrm{int},\mathrm{app}}=(0.693∕\text{in vitro}{t}_{1∕2})\times (\text{incubation volume}∕\mathrm{mg}\text{of microsomal protein})\times (45\mathrm{mg}\text{microsomal protein}∕\text{gram of liver})\times ({20}^a\mathrm{g}\text{of liver}∕\mathrm{kg}\text{body weight})$$

a: 20, 45 and 90 g of liver/kg of body weight were used for human, rat and mouse, respectively.

High-Performance Liquid Chromatography

The HPLC system consisted of a Perkin-Elmer series 200 quaternary pump (Perkin-Elmer Corp., Norwalk, CT, USA), a Perkin-Elmer UV detector series 200, and a Perkin-Elmer autosampler series 200 with a thermostated tray. Samples were analyzed on a YMC 5 μm basic column (4.6 mm i.d. × 150 mm, YMC Inc., Wilmington, NC, USA) connected to a YMC 5 μm basic guard cartridge (4.0 mm i.d. × 20 mm). The mobile phase consisted of elution at 1 mL/min with 75% A/25% B for 3 min, followed by a 17-min linear gradient to 10% A/90% B, then 10% A/90% B for 5 min (A = 100 mM NH₄OAc pH 4, B = acetonitrile). Effluent was monitored by UV detection at 245 nm.

Liquid Chromatography-Mass Spectrometry

The LC system consisted of a Dionex Ultimate 3000 RSLC (Dionex, Sunnyvale, CA, USA) equipped with Dionex Ultimate 3000 autosampler and Dionex Ultimate 3000 diode array detector. All mass spectrometric experiments were performed with a Bruker MicroTOF-Q II (Billerica, MA, USA) monitored with Hystar 3.2 software. Mass spectrometry acquisition was performed using electrospray ionization with the following parameters: end plate offset voltage = −0.5 kV, capillary voltage = 4.5 kV, nitrogen as both nebulizer and dry gas (8.0 L/min flow rate, 200 °C dry temperature, 4.0 bar). Samples were analyzed on Acclaim RSLC 120 C18 column (2.2 μm, 2.1 mm i.d. × 100 mm, Dionex, Sunnyvale, CA, USA). The mobile phase consisted of elution at 0.5 mL/min with 85% A/15% B for 3 min, followed by a 9-min linear gradient to 10% A/90% B, then 10% A/90% B for 2 min (A = 0.1% formic acid in water, B = 0.1% formic acid in acetonitrile).

AMES II™ Mutagenecity Assay

The Ames II™ mutagenecity assay was carried out by BioReliance (Rockville, MD, USA). Briefly, the strains included the Ames II™ mixed strains (TA7001 to TA7006) and tester strain TA98, as described.³⁴ The strains were mixed with the test article (5d or 3) and appropriate metabolic activation mixture. The mixture was incubated at 37 °C for 90 min. The mixture was then mixed with indicator medium and the contents were plated in 384-well plates. After 48 h incubation, the plates were scored for positive wells. The test article was tested at a maximum of six dose levels (3.3, 10.0, 33.3, 100.0, 333.3, 1000 μg/mL) along with appropriate negative and positive controls with mixed Ames II™ strains and tester strain TA98, with and without S9 activation. All dose levels of test article, negative controls and positive controls were tested in triplicate. 2-Aminoanthracene (5.0 μg/mL) and 4-nitroquinoline N-oxide (1.0 μg/mL) were used as positive controls with S9 activation. 2-Nitrofluorene (2.0 μg/mL) was used as a positive control without S9 activation.

Animals

Male C57Bl/6J mice (The Jackson Laboratory, Bar Harbor, ME, USA), 6-8 weeks old, 22-25 g body weight, specific pathogen free, were provided with Laboratory 5001 Rodent Diet (PMI, Richmond, IN, USA) and water ad libitum. Animals were maintained in polycarbonate shoebox cages with hardwood bedding in a room under a 12:12 h light:dark cycle and at 72 ± 2 °F. Animal studies were approved by the University of Notre Dame Institutional Animal Care and Use Committee.

Animal Dosing and Sample Collection

Compounds 5d and 3, as HCl salts, were dissolved in saline to a final concentration of 2.5 mg/mL. The solution was sterilized by passage through an Acrodisc^® syringe filter (Pall Life Sciences, 0.2 μm, 13 mm diameter, PTFE membrane). Mice (n = 3 per time point per compound) received a single 120 μL dose of 5d or 3 intravenously (equivalent to 12.5 mg/kg) by tail vein injection. Mice used for later time points (1 h, 2 h, and 4 h) were housed in Nalgene rat metabolic cages equipped with flooring grates and feeders for mice (Tecniplast, West Chester, PA, USA), for easy collection of urine. Terminal blood samples were collected at various time points by cardiac puncture following CO₂ asphixation, using heparin as anticoagulant. Blood was stored on ice and centrifuged to obtain plasma. Urine samples were collected, pooled by time point (1 h, 2 h, and 4 h) and stored on ice.

Sample Analysis

A 200-μL aliquot of plasma was mixed with 400 μL of 1 μM (S)-2-((S)-1-((4-phenoxyphenyl)sulfonyl)ethyl)thiirane as internal standard in acetonitrile. The sample was centrifuged at 10 000 g for 20 min. The supernatant was concentrated, reconstituted to 120 μL with water and analyzed by reversed phase LC with MS detection at m/z 478.1577 (for 5d), m/z 322.0566 (for 3), and m/z 364.0672 (for the p-N-acetyl derivative, M3). Standard curves of 5d, 3, and p-N-acetyl derivative in blank mouse plasma were prepared by fortification of 200 μL of plasma with each compound at concentrations of 0, 0.005, 0.01, 0.025, 0.05, 0.1, 0.25, 0.5, 1, 2.5, 5, 10, 25, and 50 μM. Quantification of 5d, 3, and the p-N-acetyl derivative (M3) was performed using peak area ratios relative to the internal standard and linear regression parameters calculated from the calibration curve standards prepared in blank mouse plasma. The assay was linear from 0.005 to 50 μM, with coefficients of determination, R², of 0.998 for 5d, 0.999 for 3, and 0.999 for the p-N-acetyl derivative.

Urine samples were centrifuged at 10 000 g for 10 min, and a 50-μL aliquot of each supernatant was analyzed by LC/MS. In addition to analyzing for 5d, 3 and p-N-acetyl M3, the plasma and urine samples were screened specifically for N-hydroxylamine 14 with MS detection at m/z 338.0515.

Pharmacokinetic Parameters

Pharmacokinetic parameters were determined by noncompartmental analyses of the concentration-time data for 5 and 3. Area under the concentration-time curve (AUC) was determined by the linear trapezoidal rule. The concentration at time = 0 (C₀) was estimated from the first sampling times by back-extrapolation using log-linear regression analysis. AUC_0-∞ was calculated as AUC_0-t + (C_last/k), where AUC_0-t is the area under the mean plasma concentration-time curve up to the last quantifiable sample and C_last is the concentration at the last quantifiable sampling time. Half-lives (t_½) were estimated from the terminal linear portion of the concentration time data by linear regression, where the slope of the line was the rate constant k and t_½ = ln 2/k. Volume of distribution (Vd) was calculated as dose divided by initial concentration (C₀). Clearance (CL) was calculated from dose divided by AUC_0-∞. Concentrations of 3 and the p-N-acetyl derivative (M3) after intravenous administration of 3 at 12.5 mg/kg were normalized to 8.13 mg/kg (equivalent to 12.5 mg/kg intravenous administration of 5d).

ABBREVIATIONS USED

a.m.u.

atomic mass unit

AUC

Area under the concentration-time curve

Boc

tert-butoxycarbonyl

CL

clearance

DMAP

4-(dimethylamino)pyridine

DMSO

dimethyl sulfoxide

Dnp

2,4-dinitrophenyl

HPLC

high-performance liquid chromatography

M1(−5)

metabolite 1(−5)

m-CBPA

meta-chloroperbenzoic acid

MMP

matrix metalloproteinase

MOCAc

(7-methoxycoumarin-4-yl)acetyl

MOM

methoxymethyl

MS

mass spectrometry

NADPH

nicotinamide adenine dinucleotide phosphate reduced

NHS

N-hydroxysuccinimide

Nva

l-norvalyl

Vd

volume of distribution