Design, synthesis and biological evaluations of a long-acting, hypoxia-activated prodrug of fasudil, a ROCK inhibitor, to reduce its systemic side-effects

Abstract

ROCK, one of the downstream regulators of Rho, controls actomyosin cytoskeleton organization, stress fiber formation, smooth muscle contraction, and cell migration. ROCK plays an important role in the pathologies of cerebral and coronary vasospasm, hypertension, cancer, and arteriosclerosis. Pharmacological-induced systemic inhibition of ROCK affects both the pathological and physiological functions of Rho-kinase, resulting in hypotension, increased heart rate, decreased lymphocyte count, and eventually cardiovascular collapse. To overcome the adverse effects of systemic ROCK inhibition, we developed a bioreductive prodrug of a ROCK inhibitor, fasudil, that functions selectively under hypoxic conditions. By masking fasudil’s active site with a bioreductive 4-nitrobenzyl group, we synthesized a prodrug of fasudil that is inactive in normoxia. Reduction of the protecting group initiated by hypoxia reveals an electron-donating substituent that leads to fragmentation of the parent molecule. Under normoxia the fasudil prodrug displayed significantly reduced activity against ROCK compared to its parent compound, but under severe hypoxia the prodrug was highly effective in suppressing ROCK activity. Under hypoxia the prodrug elicited an antiproliferative effect on disease-afflicted pulmonary arterial smooth muscle cells and pulmonary arterial endothelial cells. The prodrug displayed a long plasma half-life, remained inactive in the blood, and produced no drop in systemic blood pressure when compared with fasudil-treated controls. Due to its selective nature, our hypoxia-activated fasudil prodrug could be used to treat diseases where tissue-hypoxia or hypoxic cells are the pathological basis of the disease.

Graphical Abstract

Introduction

Rho-associated protein kinases (ROCKs), also known as Rho-kinases, control numerous cellular processes, including smooth muscle cell (SMC) contraction, cell proliferation, adhesion, migration, and inflammation [1, 2]. ROCK-I and ROCK-II serine/threonine kinases play critical roles in the regulation of cytoskeleton organization via phosphorylating myosin light chain (MLC), myosin phosphatase targeting protein (MYPT1), and LIM kinases [3–5]. Rho kinases contribute to the adhesion and migration properties of diseased cells by increasing their cytoskeleton contractility and cellular tension. Rho kinases also downregulate endothelial nitric oxide synthase (eNOS) under hypoxic conditions [6, 7]. Thus, ROCK inhibitors can potentially be used in the treatment of diseases such as cancer, asthma, chronic obstructive pulmonary disease, inflammatory bowel disease, eye diseases, hypertension, and pulmonary arterial hypertension (PAH) [8, 9]. Fasudil, the most widely used ROCK inhibitor, occupies and blocks the adenine pocket of the ATP-binding site of the enzyme, resulting in disorganization of actin stress fibers, inhibition of cell migration, and vasodilation [10, 11]. This drug is currently approved for the treatment of cerebral vasospasm in Japan and China, and reported to be efficacious in the treatment of PAH patients and animal models of PAH [12–14]. However, systemic exposure to fasudil or other Rho-kinase inhibitors induces an undesirable drop in blood pressure, reduces the lymphocyte count, and thus causes a change in the therapeutic window of the drug [8, 15, 16]. As such, we assume that by avoiding systemic inhibition of Rho kinase, many of the adverse effects of fasudil can be minimized to render it safe for treatment of PAH.

Indeed, attempts have been made to reduce the systemic effects of ROCK inhibitors by formulating them for topical administration. Topical ROCK inhibitors have been used for the treatment of various eye diseases. ROCK inhibitors such as ripasudil, Y-27632, Y-39983 and AR-12286 have been applied as ophthalmic solutions for the treatment of glaucoma and elevated intraocular pressure. Topical preparations have also shown promise for treatment of diabetic retinopathy, corneal wound healing, and several cases of Fuchs corneal dystrophy [17–20]. In PAH, pulmonary delivery of fasudil showed low systemic availability. However, none of the formulations, topical or inhalational, are without side effects. Even ophthalmic solutions of ROCK inhibitors are reported to produce mild to severe conjunctival hyperemia [21]. Fasudil or liposomal fasudil, when administered intratracheally to PAH rats, reduced the mean systemic arterial pressure [22, 23]. Furthermore, multiple injections or continuous infusion of fasudil caused loss of consciousness and cardiovascular collapse due to a rapid fall in blood pressure [24].

In this study, we propose a prodrug strategy with a goal to reduce the systemic effects of bioactive compounds. However, the transformation of a prodrug into an active drug involves an inherent or induced enzymatic process at the site of action. Hypoxia has been an important driver in deploying prodrug strategies in cancer treatment [25–27]. Differences in the chemical environments between hypoxia and normoxia have been successfully utilized to activate drugs in the hypoxic region. Hypoxia-targeted anti-cancer prodrugs showed improved efficacy against solid tumors but reduced side effects. These prodrug-based strategies can be applied to a wide range of diseases where hypoxia or hypoxic cells are important in the disease pathology, including PAH. In PAH, alveolar hypoxia expedites pulmonary vasoconstriction, muscularization (presence of SMCs in the intimal layer) and arterial remodeling (thickening of the intimal/medial layers), which are major pathological features of the disease [28].

For targeting drugs to hypoxic tissues or cells, the parent compounds are conjugated with a bioreductive protecting group such as 4-nitrobenzyl, which is sensitive to enzymatic bioreduction and undergoes biotransformation in the hypoxic microenvironment to release the active drug [29]. The nitro group undergoes nitroreductase-mediated one-electron reduction to a radical anion in vivo, which rapidly undergoes oxidation in the presence of physiological levels of oxygen, rendering this pathway unproductive [30, 31]. However, under hypoxic conditions, the nitro group is reduced to an electron-donating hydroxylamine then to an amine, which is not re-oxidized but has a much longer half-life and initiates the release of the active compound [29]. This reversal in reactivity under normoxic versus hypoxic conditions results in activation of prodrugs. Several hypoxia-activated anticancer prodrugs have been tested in preclinical and clinical settings, whose activation depends on a quinone moiety (E09, RH1, and AQ4N), nitroaromatic group (CH-01 and TH-302) or nitroxide moiety (TPZ and CEN-209) [32–37].

TPZ and TH-302 are furthest along in their development and now in clinical trials. These two agents are used in combination with other compounds or radiotherapy for the treatment of various cancers such as head and neck cancers, metastatic pancreatic cancer, and soft-tissue sarcoma. The clinical trials reported no benefits or overlapping toxicities when used in combination with standard therapies. However, a biomarker-driven approach has been proposed to increase the likelihood of clinical success of hypoxia-activated prodrugs in properly defined subpopulations [38].

To date, the use of the bioreductive prodrug strategy has been limited to the application of DNA-damaging cytotoxins and a few tyrosine kinase inhibitors, but has never been employed for ROCK inhibitors. The Rho/ROCK pathway is also highly activated in hypoxic cells. It upregulates the expression of hypoxia-inducible factor-1α (HIF-1α), a major player in the pathogenesis of many diseases [39]. Thus, a hypoxia-targeted ROCK inhibitor would have its inhibitory effect restricted to hypoxic cells, remain inactivated in the systemic circulation, and protect normal tissues. In this study, we report the synthesis and characterization of a bioreductive prodrug of fasudil metabolite hydroxy-fasudil (FasOH), FasPRO. We posit that, under hypoxic conditions, FasPRO shows a therapeutic effect similar to that of its parent compound, but circumvents the systemic adverse effects associated with fasudil.

Materials and Methods:

Chemicals and characterization:

All solvents and chemicals were of reagent grade. Unless otherwise stated, all reagents and solvents were purchased from commercial vendors and used as received. The purity and characterization of compounds were established by a combination of methods, including TLC, HPLC, mass spectrometry, and NMR analysis. ¹H and ¹³C NMR spectra were recorded on a Bruker 400 MHz Avance III HD spectrometer using chloroform-d or DMSO-d6 with tetramethylsilane (TMS) (0.00 pm) or solvent peaks as the internal standard.

Tert-Butyl 4-(isoquinolin-5-ylsulfonyl)-1,4-diazepane-1-carboxylate (2):

0.30 g of Fasudil hydrochloride (0.93 mmol) was converted to a free base by treating with a saturated solution of NaHCO₃. The organic portion, extracted using DCM (3 × 50.00 mL), was combined and dried over anhydrous Na₂SO₄, producing fasudil as a pale-yellow oil. Further, a solution of fasudil (0.27 g, 0.93 mmol) in DCM was treated with triethylamine (1.39 g, 13.7 mmol), and di-tert-butyl dicarbonate (4.49 g, 20.60 mmol) was added in a dropwise manner. The reaction was stirred for 4 h at room temperature (RT), concentrated under reduced pressure, and the resulting residue was purified using flash chromatography (ethyl acetate: hexanes, 1:9 to 1:1) to prepare compound 2 as a yellow viscous liquid (0.28 g, 69.50%). ¹H NMR (400 MHz, CDCl₃) δ_H 1.43 (s, 9 H), 1.97 (t, 2 H, J = 6.24 Hz), 3.38 – 3.44 (m, 4 H), 3.51 – 3.58 (m, 4 H), 7.69 (t, 1 H, J = 1.00 Hz), 8.19 (d, 1 H, J = 8.56 Hz), 8.54 (m, 1 H), 8.41 (d, 1 H, J = 6.11 Hz), 8.70 (d, 1 H, J = 5.87 Hz), 9.35 (br. s, 1 H).

Tert-Butyl 4-((1-hydroxyisoquinolin-5-yl)sulfonyl)-1,4-diazepane-1-carboxylate (4): 5-((4-(tert-butoxycarbonyl)-1,4-diazepan-1-yl)sulfonyl)isoquinoline 2-oxide (3; intermediate compound):

To a solution of Boc-protected fasudil (0.82 g, 2.11 mmol) in DCM (50.00 mL), meta-chloroperoxybenzoic acid (mCPBA, 0.50 g, 2.94 mmol) was added gradually over a period of 15 min at 0°C. The reaction mixture was allowed to return to RT and then stirred overnight. Then the reaction solvent was removed under reduced pressure, and the formed residue was purified by recrystallization using ethyl acetate and used in the next step without further purification (0.78 g, 91%).

Tetramethylammonium chloride (0.016 g, 0.15 mmol) and sodium acetate (0.12 g, 1.47 mmol) in 3.00 mL of water were added to a solution of compound 3 (4.20 g, 0.49 mmol) in DCM (5.00 mL). After the suspension was formed, the reaction mixture was treated with benzoyl chloride (0.137 g, 0.98 mmol) by adding dropwise at RT. After completing the reaction, the organic portion was separated, dried over anhydrous magnesium sulfate, and the solvent was removed under reduced pressure. The target compound 4 was obtained from the recrystallization using ethyl acetate (0.16 g, 80%).¹H NMR (400 MHz, CDCl₃) δ_H 1.44 (s, 9 H), 1.96 (d, 2 H, J = 1.00 Hz), 3.36–3.43 (m, 4 H), 3.49–3.58 (m, 4 H), 7.35 (d, 1 H, J = 7.34 Hz), 7.46 – 7.62 (m, 2 H), 8.08 – 8.19 (m, 1 H), 8.22 (d, 1 H, J = 7.34 Hz), 8.68 (d, 1 H, J = 8.31 Hz), 10.81 (br. s., 1 H).

Tert-Butyl 4-((1-((4-nitrobenzyl)oxy)isoquinolin-5-yl)sulfonyl)-1,4-diazepane-1-carboxylate (5):

4-Nitrobenzyl bromide (0.13 g, 0.60 mmol) was added to a reaction mixture containing compound 4 (0.1 g, 0.25 mmol), cesium carbonate (0.19 g, 0.60 mmol) and toluene (2.00 mL). The resulting reaction mixture was heated to reflux (110°C) in a pressure vessel and stirred at that temperature for 5 h. The mixture was then cooled to RT and the solvent was removed using a rotary evaporator. The crude product was purified using a CombiFlash system (ethyl actetate: hexane, 1:9 to to 2:3) to obtain product 5 (0.06g) in 48% yield. Fig. S1: ¹H NMR (400 MHz, CDCl₃) δ_H 1.44 (s, 9 H, t-Boc), 1.95 (d, 2 H, J = 1.00 Hz, -CH₂-) 3.27 – 3.47 (m, 4 H, -CH₂-CH₂-CH₂-) 3.47 – 3.66 (m, 4 H, -CH₂-CH₂-) 5.29 (s, 2 H, -CH₂Ph) 7.16 – 7.30 (m, 1 H, ArH) 7.34 (d, 1 H, J = 7.83 Hz, ArH) 7.50 (d, 1 H, J = 8.80 Hz, ArH) 7.58 (t, 1 H, J = 7.95 Hz, ArH) 8.12 – 8.29 (m, 3 H, ArH) 8.68 (d, 1 H, J = 7.82 Hz, ArH). Fig. S2: ¹³C NMR (100 MHz, CDCl₃) δ_C 28.40 (-C(CH₃)₃), 45.53 (-CH₂-CH₂-CH₂-), 45.95 (-N-CH₂CH₂Nboc-), 47.80 (-N-CH₂CH₂CH₂), 49.28 (-NBoc-CH₂CH₂), 50.22 (-N-CH₂CH₂CH₂), 51.83 (-CH₂Ph), 79.99 (-C(CH₃)₃), 103.36 (-CPy), 124.16 (-CPh), 126.18 (-CPh), 128.15 (-CPh), 128.66 (-CPh), 133.29 (-CPh), 134.53 (-(-NCPy)Ph), 143.28 (-NCPy), 147.71 (-C(NO₂)Ph), 155.18 (-COO^tBu), 161.25 (-C(O)Py).

5-((1,4-diazepan-1-yl)sulfonyl)-1-((4-nitrobenzyl)oxy)isoquinoline (6):

To a solution of compound 5 (0.03g, 0.046 mmol) in DCM (2.00 mL), trifluoroacetic acid (0.50 mL) was added dropwise at 0°C. After stirring at RT for 2 h, the reaction solvent was removed under reduced pressure, and the resulting residue was treated with a saturated aqueous solution of NaHCO₃. A neutralized product was suspended in water, extracted with dichloromethane (2 × 5.00 mL), the organic portions were collected, dried over anhydrous magnesium sulfate, and the solvent was removed under reduced pressure to obtain the crude compound. The final compound 6 was obtained by using flash column chromatography (methanol:DCM, 1:20) (0.01 g) in 49% yield as a cream-colored powder. Fig. S3: ¹H NMR (400 MHz, DMSO-d6) δ_H 2.02 (br. s, 2 H, -CH₂-), 3.20 (br. s, 4 H, -CH₂-, -CH₂-), 3.48 (t, 2 H, J = 6.11 Hz, -CH₂-, -CH₂-), 3.65 (br. S., 2 H, -CH₂-CH₂-,), 5.37 (br. s, 2 H, -CH₂Ph-), 7.22 (d, 1 H, J = 7.82 Hz, ArH), 7.58 (d, 2 H, J = 8.80 Hz, ArH), 7.70 (t, 1 H, J = 7.83 Hz, ArH), 7.85 (d, 1 H, J = 7.83 Hz, ArH), 8.19–8.29 (m, 3 H, ArH), 8.56 (d, 1 H, J = 8.07 Hz, ArH), 9.09 (br. s, 2H). Fig. S4: ¹³C NMR (100 MHz, DMSO-d6) δ_C 34.63 (-CH₂-CH₂-CH2), 44.37 (-N-CH₂CH₂N), 44.72 (-N-CH₂CH₂N-), 46.67 (-NCH₂CH₂-), 47.07 (-N-CH₂CH₂CH₂), 65.39 (-CH₂Ph), 102.19 (-CPy), 124.27 (-CPh), 126.85 (-CPh), 127.84 (-Cph), 129.16 (-CPh), 133.18 (-CPh), 133.92 (-CPh), 134.28 (-CPh), 135.97 (-(-NCPy)Ph), 145.23 (-NCPy), 147.36 (-C(NO₂)Ph), 160.80 (-C(O)Py). Calculated [M+H]⁺ 443.49, obtained [M+H]⁺ 443.7; purity 95.20%.

Cell lines and reagents:

Cells used in this study were primary pulmonary arterial cells from healthy human male and female donors or patients with idiopathic PAH (IPAH), which we obtained from the Pulmonary Hypertension Breakthrough Initiatives (PHBI) specimen bank (www.ipahresearch.org) under an approved protocol and listed here as normal smooth muscle cells (N-SMCs), PAH-afflicted endothelial cells (PAH-ECs) and PAH-afflicted smooth muscle cells (PAH-SMCs).

Cell-free Rho-associated kinase activity:

The activity of a kinase (ROCK-II) was tested over a broad range of inhibitory concentrations to generate a dose-response curve according to the manufacturer’s instructions (CSA001, EMD Millipore, MA). Reactions containing 1 mU/μL ROCK-II with 10 μM ATP and different concentrations of fasudil, FasOH, and FasPRO were incubated in a MYPT-1-coated plate for 30 min at 30°C. After repeated washing, an anti-phospho-MYPT1 (Thr696) antibody solution was added and incubated at RT for 1 h. The binding of primary antibody to MYPT-1-coated wells was detected using an HRP-conjugated secondary antibody followed by incubation with TMB/E substrate. The absorbance was measured at 450 nm in a Synergy II microplate reader (BioTek Instruments Inc., VT). The IC₅₀ values were calculated from the concentrations that produce 50% inhibition of the enzyme activity. The results were analyzed by non-linear curve fitting with a sigmoidal dose-response and variable slope.

Reduction of FasPRO by zinc (Zn) dust:

To a dimethyl formaldehyde (DMF; 2 mL) solution of compound 6 (1 mg, 0.0021 mmol), aqueous ammonium chloride (20 μL, 10% w/v) and Zn powder (5 mg, 0.0765 mmol, 36 equivalent) were added. The resulting mixture was stirred at ambient temperature for 16 h. An aliquot of 200 μL was taken at different time points (T = 0, 1, 4, 12, 24 h, where T = 0 refers to the initial time before the addition of Zn powder). At each time point, an aliquot (5 μL) of Zn reduction assay was further added to 95 μL of acetonitrile. Collecting the supernatant after centrifugation, the samples were subjected to LC-MS/MS to confirm the conversion of inactive prodrug into its active form. All FasPRO was consumed within 1 h and the remaining samples were discarded.

Enzymatic reduction assay of FasPRO:

For the bioreduction of compound 6, bactosomal human NADPH-CYP reductase (Cypex, 18.8 mg/mL, 24300 nmol/min/mL) was used in combination with an NADPH-regenerating system (BD Biosciences, CA) that contains 50 μL of solution A, 10 μL of solution B, 726.6 μL of deionized H₂O, and 200 μL of 0.5 M KHPO4 (pH 7.4) buffer. The assay was carried out according to the manufacturer’s instructions. The final concentrations of FasPRO and enzymes were kept at 10 μg/mL and 100 pmol/mL, respectively, in the reaction mixture. Each vial was deoxygenated prior to the addition of the bactosomal enzyme and transferred into a normobaric hypoxic C-chamber (Biospherix, NY) set at 0.1% O₂, 5% CO₂, 94.9% N₂ [40–42]. Another set of samples was also kept under normoxic conditions. Samples were taken at different time points, aliquoted, stored at −80 °C until analysis, and then analyzed by LC-MS/MS.

Antiproliferative effects:

We grew human pulmonary arterial normal cells (PACs) or PAH-afflicted ECs and SMCs in gelatin-coated dishes using EC (PromoCell, Germany) and SMC (VascuLife®, Lifeline Cell Technology, CA) growth media, respectively, for 3–5 days until confluency. The cells in this study were used between passage P3 to P5 with population doublings up to 20. PAH-ECs, PAH-SMCs, and N-SMCs were seeded at a density of ~5,000 cells per well in 100 μL of medium and grown for a day. On day one of the seeding, we treated the cells with different concentrations (1000, 100, 10, 1, 0.1, 0.01, 0.001, 0 μM) of fasudil, FasOH, and FasPRO, and incubated them in a humidified incubator under normoxia (37 °C, 5% CO2) or hypoxia (0.1% O₂, 5% CO₂, 94.9% N₂). After 24 h of treatment, we incubated the cells with 10 μL of Cell Counting Kit-8 (Sigma-Aldrich Inc., MO) for 0.5~2.0 h at 37°C and recorded the absorbance at 450 nm. Each experiment was replicated 3–4 times.

Myosin phosphatase-targeting protein phosphorylation assays:

PAH-SMCs were cultured in SMC growth medium. Before the treatment, PAH-SMCs were starved for 21 h in basal medium supplemented with 1% antibiotics. Serum-starved PAH-SMCs were then treated with 1 μM of fasudil or FasPRO for 24 h, either under normoxic or hypoxic conditions. Whole-cell lysates from the treated cells were collected using lysis buffer [20 mM Tris-HCl (pH 7.4), 15% glycerol, 1% Triton X-100, 8 mM MgSO₄, 150 mM NaCl, 1 mM EDTA, supplemented with β-glycerophosphate, NaF, protease inhibitor, and phosphatase inhibitor]. After centrifuging (14000 x g for 10 min at 4ºC) the whole-cell lysates, we determined the protein concentration using a BCA protein assay kit (Pierce Biotechnology, IL). The cell lysates were separated by SDS-PAGE and transferred to nitrocellulose membranes. Membranes were blocked with 5% bovine serum albumin and incubated with primary antibodies against phospho-MYPT1, MYPT1, and β-Actin (Cell Signaling, MA) followed by incubating with goat anti-rabbit and goat anti-mouse IgG HRP-conjugated secondary antibodies. Immunoreactivity of the proteins was visualized using a ChemiDoc™ XRS molecular imager system (Bio-Rad, CA) after incubating the membranes with WesternSure^® chemiluminescence reagent (LI-COR Biosciences, NE) for 1 min.

Measurement of cyclic guanosine monophosphate (cGMP):

To determine the influence of fasudil and FasPRO on cGMP levels, we treated PAH-SMCs, N-SMCs, and PAH-ECs with various concentrations of fasudil and FasPRO. After 6 h of treatment, we prepared cell lysates and determined cGMP and nitrate levels in the lysate using a cGMP ELISA kit (R&D Systems, MN).

Pharmacokinetics and assay of fasudil, FasOH, and FasPRO in rat plasma:

For pharmacokinetic experiments, Sprague Dawley rats (male, weighing 250–300 g) were randomly divided into two groups (n = 4 for each group). Fasudil and FasPRO (3 mg/kg) were injected through the tail vein and blood was collected at 15 and 30 min, 1, 2, 4, 8, 12, 24, and 48 h. Plasma was isolated by centrifugation at 4,000 x g for 20 min. The plasma concentrations of fasudil and its active metabolite, FasOH, were analyzed in the plasma of fasudil-treated rats. We also analyzed the plasma concentration of FasPRO and its parent compound, fasudil, from the FasPRO-treated rats. The pharmacokinetics parameters of fasudil, FasOH, and FasPRO were estimated using non-compartmental analysis with WinNonlin (Certera, NJ). To estimate the kinetics parameters, below the lower limit of quantitation (LLOQ) values were set to zero. We calculated K_el (from the terminal segment of the natural concentration time curve using the last three points of the plasma-concentration time curve. Plotting the natural log of the last three concentrations against time, we calculated the K_el from the slope of that curve. Using that K_el, we calculated AUC_0-inf, t_1/2 and V_d. This method of calculation of various PK parameters is standard and reported in many pharmacokinetics textbooks (e.g. Shargel et al., Seventh Edition, p. 521) [43].

Fasudil concentrations were quantified by a tandem mass spectrometer (AB QTRAP 5500; SCIEX, Framingham, MA) attached to an Ultra Performance Liquid Chromatography (UPLC) system (Shimadzu; Agilent Technology, Santa Clara, CA) via an electrospray ionization (ESI) interface. Using ranitidine (GlaxoSmithKline, PA) as an internal standard, calibration curves were prepared in rat plasma after extraction with methanol. For drug extraction, 500 μL of cold methanol was added to 100 μL of plasma, vortexed for 30 seconds, centrifuged at 11,000 rpm for 5 min, then 500 μL of the supernatant was transferred to glass vials and evaporated to dryness under nitrogen at 40°C. Later, the residue was reconstituted in 100 μL of 50% methanol, vortexed again, and 20 μL of reconstituted solution injected into the LC-MS/MS instrument.

For chromatographic separation, a 100 mm × 2.1 mm XB-C18 column (Phenomenex, Torrance, CA), attached to a 12.5 mm × 2.1 mm guard column, was used with a gradient elution (2 mM ammonium formate in water, modified with 0.1% formic acid as eluent A and methanol as eluent B) at a flow rate of 0.45 mL/min. The gradient of eluent B was 10% at 0–0.5 min, 35% at 0.5–1.2 min, 75% at 1.2–1.6 min, 40% at 1.6–2.2 min, and 10% at 2.2–2.8 min. After a single run, the column was washed with 10% eluent B at 2.8–4.5 min. The drugs were quantified via a multiple reaction monitoring method, keeping the ESI source at a positive ionization mode, for the following transitions m/z 292.2 → 99.2, m/z 308.7 → 99.2, m/z 443.7 → 128.2, and m/z 315.3 → 176.2 for fasudil, FasOH, FasPRO, and internal standard ranitidine, respectively, with a dwell time of 150 milliseconds per transition. The amount of drug was read from a calibration curve prepared by plotting m/z area against drug concentration.

Assessing pulmonary hemodynamics in healthy rats:

After anesthetizing Sprague Dawley rats by an intramuscular injection of ketamine and xylazine, we shaved the ventral neck area and inserted a polyvinyl catheter (PV-1, Tygon®, OH) into the left ventricle via the right carotid artery to measure the mean systemic arterial pressure (mSAP). Using a PowerLab™ 16/30 system, connected to a computer loaded with LabChart Pro™ 7.0 software (ADInstruments Inc., CO) and equipped with a Millar^® MPVS ultra system (Millar Inc., TX), we recorded the real-time mSAP. Initially, the baseline pressure was monitored until we obtained a stable reading. Dividing the rats into two groups (n = 3), we injected fasudil and FasPRO (3 mg/kg) and monitored the mSAP for 6 h. Assuming the baseline pressure as 100%, we calculated the percentage reduction of mSAP from the baseline.

Statistical analysis:

We performed statistical analysis at α = 0.05 (95% confidence interval) with GraphPad Prism and used the t-test, one/two-way ANOVA, Tukey’s post hoc, and repeated measures ANOVA, as applicable.

Results

Synthesis of a bioreductive prodrug of the Rho-kinase inhibitor FasPRO:

The X-ray crystal structure of fasudil-ROCK complex revealed that the isoquinoline ring of fasudil occupies the ATP-binding site of ROCK [44]. The isoquinoline nitrogen of fasudil is the hydrogen bond acceptor that interacts with Met156 of ROCK, whereas the fused aromatic ring is responsible for the hydrophobic interaction with ROCK. The higher potency and selectivity of FasOH stems from a new binding mode caused by the difference in orientation of isoquinoline versus isoquinolinone: the carbonyl group accepts a hydrogen bond and the NH-group donates a hydrogen bond to ROCK. For both fasudil and FasOH, a mild interaction between the secondary amine of the homopiperazine ring with the side chain of ROCK at position Asp160 was also observed. In the ROCK ligand binding site, isoquinoline occupies the same space and has the same interactions as the adenine of ATP, which points to the role of the isoquinoline ring in fasudil as an active functional group to inhibit ROCKs. Based on these observations, we posited that substitution of the hydroxyl group of fasudil (FasOH) disrupts binding interactions between the inhibitor and ROCK.

The secondary amine of the homopiperazine of fasudil was protected with t-Boc using a reported scheme [45] (Fig. 1A). The oxide 3 was synthesized as previously reported by Valdivia et al., which was used in the next step without further purification to obtain the desired Boc-protected FasOH 4 [46]. Compound 4 was condensed with p-nitrobenzyl bromide to obtain compound 5. The fasudil prodrug (6) was obtained by cleavage of t-Boc using trifluoracetic acid. The final compound (6) was confirmed by ¹H and ¹³C NMR, which showed the characteristic peak for the O-C ether link formation at δ_H 5.37 ppm for 2 protons from the benzyl moiety, which corresponds to δ_C 65.39 ppm for the associated carbon (Figs. S3 and S4). We observed that the chemical shift at 5.37 ppm represents the O-benzyl -CH₂-, which is in line with previous reports [47, 48] but contradicts the regioselective O-benzylation reported by Ferrer et al. [49]. As we can see, there is a wide range of chemical shifts for the benzyl moiety in comparison with the reported O-benzylation derivatives, making it difficult to ascertain -CH₂-Ph precisely. But comparison of the chemical shift of the compound 6 with the similar scaffolds [50] indicate that O-benzylation is the correct assignment.

Figure 1:

Synthetic scheme and bioreduction process of FasPRO: (A) Introduction of the 4-nitrobenzyl group into the isoquinoline portion of fasudil: (i) (Boc)₂O, DCM, RT, 2 h; (ii) mCPBA, DCM, 0 °C, 15 min; RT, 18 h; (iii) sodium acetate, water, tetramethylammonium chloride, room temperature (RT), 2 h; (iv) cesium carbonate, toluene, RT to 110 °C, 5 h; (v) TFA, DCM, RT, 2 h. (B) Under hypoxic conditions, the nitro group is reduced to form an electron-donating substituent [29, 36], which induces fragmentation and ejects the hydroxy-fasudil (7) and parent compound, 1.

In-vitro assessment of the chemical- and bioreduction of FasPRO:

To determine the mechanism of FasPRO activation, chemical and enzymatic reduction methods were used [29]. Using a Zn reduction method, which reduces 4-nitrobenzyl to a 4-aminobenzyl group, we demonstrated a sequence of reductions to assess the properties of the hypoxia-activated prodrug. First, an LC-MS/MS method was developed and validated for the detection of fasudil, FasOH, and FasPRO from the same mixture (Fig. S5). After dissolving FasPRO in DMF, in the presence of Zn powder and NH4Cl, for a given time period, we analyzed the fragmentation of the prodrug using the LC-MS/MS method. We observed that the nitroaryl groups were completely reduced under these conditions after 1 h and FasPRO was fragmented into its parent compound, FasOH (Fig. 2A).

Figure 2:

Oxygen-dependent release of hypoxia-activated fasudil prodrug, FasPRO. (A) The reduction of FasPRO at different time points after incubation with Zn powder. An analysis of the supernatant revealed that the reduction of FasPRO was completed in 1 h. The statistical significance was calculated in relation to the quantity at t = 0. (B) LC-MS/MS extracted ion chromatograms reveal the loss of FasPRO (red) and the formation of hydroxy-fasudil (FasOH; blue) and fasudil (gray) after incubation with bactosomal human NADPH-CYP reductases at 0 to 24 h. (C) The propensity of FasPRO to undergo bioreduction when treated with bactosomal human NADPH-CYP reductases under normoxia or hypoxia. p value vs time 0 under hypoxia. (D) The bioreduction of FasPRO in PAH-afflicted endothelial cells (PAH-ECs) and smooth muscle cells (PAH-SMCs) under both normoxia and hypoxia. FasPRO content was measured in cell lysates. p value vs normoxia. (E) LC-MS/MS chromatograms of FasPRO, FasOH, and fasudil after 24 h of incubation with PAH-SMCs. The bioreduction of FasPRO → FasOH (main pathway) or FasPRO → fasudil (unwanted pathway) was prompted under hypoxia, but not in normoxia. * p < 0.05, ** p < 0.01, *** p < 0.001; data represent mean ± SD.

A nitroaryl-based prodrug can be reduced in vivo in different pathways that are likely to vary between cell lines. Thus, it is not feasible to assess all modes of enzymatic reduction in vitro. Treating FasPRO with bactosomal human NADPH-CYP reductase that reduces the nitroaryl groups in an oxygen-dependent manner [51, 52], similar to an in-vivo condition, we assessed the bioreduction propensity of the prodrug. The LC-MS/MS analysis of the reductase reaction mixture at the given time points (2.0–24 h) confirmed the reduction and fragmentation of FasPRO into FasOH and fasudil over time (Fig. 2B). The enzymatic reaction gradually fragmented FasPRO into its parent compound, FasOH. More than 60% of the prodrug was converted into FasOH or fasudil after 24 h of reaction under hypoxia (Fig. 2C). The reduction did not occur under normoxia at any of the time points, suggesting that the bioreduction was oxygen-dependent and enzyme-mediated.

Bioreduction of FasPRO in cultured cells:

To demonstrate hypoxia-driven enzymatic fragmentation of FasPRO, we evaluated the bioreduction process of the prodrug-cultured cells. In PAH, fasudil showed inhibitory and anti-migratory effects on both ECs and SMCs [53]. Thus, we used diseased PAH-ECs and PAH-SMCs as representative cell lines. These cell lines were obtained from the Type I pulmonary arteries of patients with WHO-classified Group I idiopathic PAH. After treating PAH-ECs and PAH-SMCs with FasPRO under both normoxic and hypoxic conditions, the cells were lysed, and the amounts of prodrug and its active metabolite were analyzed by LC-MS/MS. After 24 h of incubation, FasPRO accumulated in both cell types, but did not undergo significant fragmentation under normoxic conditions, as evidenced by the presence of negligible amounts of fasudil and FasOH in both cell types (Fig. 2D). In contrast to normoxia, the content of FasPRO decreased in PAH-ECs and PAH-SMCs under hypoxic conditions, as evidenced by the increase in the amount of FasOH or fasudil inside the cells (Fig. 2E). These results indicate that FasPRO was relatively stable under normoxic conditions but underwent reduction and fragmentation under hypoxic conditions inside the cells. This latter condition may lead to the conversion of inactive FasPRO into its parent compound.

To determine the major fragments of FasPRO after bioreduction, we analyzed the amount of fasudil or FasOH in the cell lysate. We treated PAH-SMCs with FasPRO for 12 h and analyzed the amount of fasudil in the cell lysate under normoxic and hypoxic conditions. LC-MS/MS data revealed the presence of negligible amounts of fasudil or FasOH in the cell lysate under normoxia. In contrast, under hypoxia, the amount of FasOH in the cell lysate was greater than that of fasudil (Fig. S8). This result suggests conversion of FasPRO to either FasOH (major fragment) or fasudil (minor fragment). This unusual fragmentation may be explained by the protonation of nitrogen at the isoquinoline that could lead to nucleophilic displacement of 4-nitrobenzyl alcohol [54]. The acidic environment of the hypoxic site [55] may also protonate the FasPRO and thus trigger the unusual biotransformation to produce fasudil.

In-vitro kinase activity of FasPRO:

The actions of Rho are mediated by the downstream Rho effectors ROCK and ROCK-II. ROCK controls Rho signaling and the phosphorylation of several substrates that alter the assembly of actin filaments and their contractility. ROCK phosphorylates MYPT1 at threonine residue 696 and thus causes an increase in the phosphorylated form of MLC [4]. Therefore, we tested the activity of ROCK-II in the presence of fasudil, FasOH, and FasPRO. The IC₅₀ values for inhibiting ROCK-II were obtained from dose-response titration curves. Reactions containing 1.0 mU of ROCK-II with 10 μM ATP were incubated with different concentrations of inhibitors. FasPRO differs in its potency to inhibit ROCK kinase activity relative to fasudil or its active metabolite, FasOH. Fasudil and FasOH inhibited the kinase activity of ROCK-II with equal potency, whereas a clear shift in IC₅₀ values was observed for the prodrug (Fig. 3). The addition of the 4-nitrobenzyl group to the hydroxyl group of FasOH resulted in a 15-fold reduction in ROCK-II kinase activity.

Figure 3:

Inhibitory concentrations determined for ROCK-II (n = 3 repeated experiments).

Under hypoxic conditions, FasPRO inhibits proliferation of PAH-afflicted ECs and SMCs:

To assess the hypoxia-driven activation and antiproliferative effect of the prodrug, PAH-ECs and PAH-SMCs were incubated with different concentrations of the inhibitors under both normoxic and hypoxic conditions. In cells incubated with fasudil or FasOH, an oxygen-independent antiproliferative effect was observed. In normoxia, the estimated IC₅₀ values of fasudil and FasOH for PAH-SMCs were 4.9 and 6.9 μM, whereas the IC₅₀ values in hypoxia were 3.9 and 10.3 μM (Table 1 and Fig. S9). In contrast, FasPRO showed only a modest antiproliferative effect under normoxia. The IC₅₀ of FasPRO under hypoxia was similar to that of fasudil and FasOH. The modest antiproliferative response elicited by FasPRO in normoxia compared to hypoxia may be due to the diseased phenotype of the cells. Thus, we also examined the antiproliferative effect of fasudil and FasPRO in normal pulmonary artery SMCs (N-SMCs). Consistent with our hypothesis, fasudil exhibited similar antiproliferative effects on N-SMCs under both normoxia (IC₅₀ = 2.8 μM) and hypoxia (IC₅₀ = 1.6 μM) (Table 1 and Fig. S10). On the other hand, FasPRO had the lowest antiproliferative effect for N-SMCs under normoxia (IC₅₀ = 247.9 μM) but was very effective under hypoxia (IC₅₀ = 1.04 μM).

Table 1:

Antiproliferative activities of the compounds on healthy and PAH-afflicted PASMCs and PAECs.

Tested compounds	PAH-SMC (IC50)		N-SMC (IC50)		PAH-EC (IC50)
	Normoxia	Hypoxia	Normoxia	Hypoxia	Normoxia	Hypoxia
Fasudil (μM)	4.95 ± 1.3	3.9 ± 2.8	2.8 ± 0.6	1.68 ± 0.5	16 ± 1.2	34.1 ± 5.9*
FasOH (μM)	6.9	10.3 ± 2.6	-	-	-	-
FasPRO (μM)	53.5	9.4 ± 1.2 ***	247.9 ± 28.9	1.04 ± 0.1 ***	2725 ± 50.9	25.4 ± 4.8 ***

^*p < 0.05 and

^***p < 0.001 versus normoxia of the same compound.

Data presented as mean ± SD.

Pulmonary vascular remodeling is primarily caused by EC apoptosis and proliferation, which subsequently causes recruitment, proliferation and differentiation of SMCs [28]. Fasudil reversed lung EC functions and vascular remodeling in a rat model of PAH with left heart disease [56]. Therefore, we investigated the effect of fasudil and FasPRO on PAH-ECs under both normoxic and hypoxic conditions. The results of the assay indicated that fasudil treatment inhibited the proliferation of PAH-ECs in a dose-dependent manner after 24 h, whereas the IC₅₀ for fasudil was almost 20 times greater in the case of PAH-SMCs (Table 1 and Fig. S11). Interestingly, FasPRO was almost inactive under normoxia, but under hypoxia showed an IC₅₀ value similar to that of fasudil.

FasPRO attenuates cellular Rho-kinase activity in hypoxia but not in normoxia:

In pathological conditions, ROCK downregulates eNOS expression, deactivates the soluble guanylyl cyclase (sGC) pathway, decreases the production of cyclic guanosine monophosphate (cGMP), and causes an imbalance in the cGMP-sGC pathway [57]. Thus, cGMP can be used as a biomarker for the measurement of ROCK inhibitors. Since fasudil can inhibit the protein kinases at higher concentrations, in this experiment we used low concentrations: 0.1–0.001 μM. The basal level of cGMP was very low in normoxia but undetectable in hypoxia in both PAH-SMCs and PAH-ECs, suggesting that the Rho kinase activities are high in the presence of hypoxia in these diseased cells (Fig. 4A, B). Preincubation of fasudil for 24 h dose-dependently increased the intracellular level of cGMP in both cell types under normoxic and hypoxic conditions; PAH-SMCs showed a 5-fold increase in cGMP levels compared to PAH-ECs (Fig. 4A, B). In normoxia, FasPRO-treatment did not improve cGMP secretion at any of the tested concentrations. Treating the cells with FasPRO under hypoxia dose-dependently increased the intracellular level of cGMP (Fig. 4A, B).

Figure 4:

Targeting Rho-kinase inhibition in hypoxic cells. The levels of intracellular cGMP in (A) PAH-SMCs and (B) PAH-ECs treated with various concentrations of fasudil and FasPRO under both normoxia and hypoxia. PAH-ECs were treated with 0.1 μM fasudil and FasPRO. p-value vs. control at normoxia (Nor-Con). Western blot analysis of phosphorylation of Myosin Phosphatase (pMYPT), total MYPT, and actin in (C) PAH-SMCs and (D) PAH-ECs treated with 1.0 μM fasudil or FasPRO under normoxia and hypoxia. * p < 0.05, ** p < 0.01, *** p < 0.001; data represent mean ± SD.

Upon activation, the Rho-kinase pathway inhibits the dephosphorylation of MLC by MYPT-1, which initiates smooth muscle relaxation. So, we investigated whether FasPRO induces changes in the phosphorylation of myosin phosphatase (phospho-MYPT1/MYPT1) in normoxia and hypoxia. Using western blot analysis, we determined whether the presence of fasudil and FasPRO causes a decrease in the hypoxia-driven phospho-MYPT1 signal. Hypoxia alone induced higher phosphorylation of MYPT1 than normoxia for both cell types, PAH-SMCs and PAH-ECs (Fig. 4C, D). Fasudil and FasPRO reversed the process under hypoxic conditions, but FasPRO did not cause any changes in the phospho-MYPT-1 level in normoxia. These data indicate that FasPRO inactivates and attenuates the activity of ROCK only under hypoxic conditions.

FasPRO is stable in the blood after intravenous administration:

The pharmacokinetic profiles of fasudil and FasPRO were obtained after intravenous administration of the drugs to Sprague-Dawley rats at a dose of 3 mg/kg. Fasudil was detectable in the blood up to 1 h after administration, showing a short elimination half-life (t_1/2 = 0.2 ± 0.05 h; Table 2). After intravenous administration of fasudil, the maximum plasma concentration (C_max) of FasOH was approximately 37~52% of that of the parent drug. The C_max of FasOH was much lower than that of the parent drug (Fig. 5A). Similar results were also obtained for other groups, where ~40% fasudil converted to FasOH after intravenous injection into rats. True to our assumption, FasPRO showed a long circulation time and a significantly extended terminal half-life of more than 7 h, which was 36-fold greater than that of fasudil (Fig. 5B and Table 2). To evaluate whether FasPRO undergoes fragmentation in the blood during its circulation, we determined the plasma concentration of fasudil and FasOH in FasPRO-treated rats. Of the four rats tested, one showed 13~14% of fasudil at the initial time points and another showed only 9.2% fasudil in the plasma 1 h after injection. In the remaining two rats, fasudil was not detectable in the blood (Fig. 5B). The trend for FasOH concentration was similar to that of fasudil in rat plasma (Fig. 5B). The rate-limiting step for disposition of fasudil is its elimination, because there is no significant difference between the half-life of fragmented-fasudil and the half-life of fasudil (Fig. 5C). If the disposition of fasudil and its fragmented counterpart were not elimination-rate limited, we would have observed a so-called “flip-flop phenomenon”, where the K_el and T_1/2 for the two compounds would be different (e.g. Shargel et al., Seventh Edition, p. 521)[43]. Therefore, FasPRO was not converted into its parent compound by blood esterases, but remained stable after administration.

Table 2:

Pharmacokinetic parameters of fasudil and FasPRO.

	C0 (ng/mL)	t1/2 (hr)	AUC 0-inf (ng.hr/mL)	Vd (L/Kg)	Cl (L/h/Kg)
Fasudil	1359.9 ± 368	0.2 ± 0.05	362.9 ± 54.1	0.9 ± 0.2	2.8 ± 0.4
FasPRO	471 ± 122.1	7.3 ± 3.4	951.2 ± 137.1	10.1 ± 4.0	1.1 ± 0.2

Data presented as mean ± SD.

Figure 5:

Pharmacokinetic profile of fasudil and its prodrug FasPRO. (A) The plasma-time versus concentration plot in a semi-log scale of intravenous fasudil at a dose of 3 mg/kg. The concentration of FasOH (the active metabolite of fasudil) was also calculated from the plasma of the same rat. (B) The plasma concentration of FasPRO, its fragmented compound fasudil (fragmented-fasudil), and FasOH from fragmented-fasudil, when given at an intravenous dose of 3 mg/kg. (C) The half-life (t_1/2) of fasudil, FasPRO, and fragmented fasudil. N = 4. Data are presented as mean ± SD.

FasPRO does not lower the systemic blood pressure:

To evaluate the effect of systemic exposure, we measured the mean systemic arterial pressure (mSAP) after treatment of Sprague-Dawley rats with fasudil and FasPRO, and calculated the reduction in mSAP. Intravenous administration of fasudil produced an immediate drop in mSAP followed by a maximum reduction of 38.6±8.4% within 30 min (Fig. 6A). The reduction in mSAP lasted for 90 min after fasudil injection. However, intravenous FasPRO did not produce any changes in mSAP until 6 h after its injection (Fig. 6B), suggesting that FasPRO was inactive in the systemic circulation.

Figure 6:

Changes in mean systemic arterial pressure (mSAP) in healthy Sprague-Dawley rats when a single dose of (A) fasudil and (B) FasPRO was administered intravenously at 3 mg/kg. Fasudil decreased the mSAP whereas FasPRO had no effect on mSAP. * p < 0.05 and ** p < 0.01 versus baseline (before drug administration); data represent mean ± SD.

Discussion

Physiological oxygen levels decline in a diverse range of disorders, a condition which contributes to the pathogenesis of many vascular diseases and tumors. In tumor cells, for example, oxygen levels fall to a level that cells become hypoxic (~0.1–1%) and even anoxic (~0.02%) [58]. Blood circulation may be impaired in many vascular diseases and lead to end-tissue/cell hypoxia. Hypoxic cells become difficult to treat, resistant to conventional therapies, and thus represent the most aggressive pathology of a disease that needs to be treated to improve patient outcome. Hypoxia-activated prodrugs selectively target the hypoxic microenvironment or cells of a disease and represent an appealing concept for the development of novel therapies [59]. This approach utilizes the coupling of nitroaromatic bioreductive groups to the functional group of a drug molecule that is important for its biological activity. The unreduced and protected form of the drug will remain inactive or show substantially reduced activity until it reaches hypoxic cells, wherein the protecting group will undergo biotransformation and release the active compound.

In the present study, we prepared 4-nitrobenzyl-conjugated fasudil, evaluated the structure activity relationship, and showed that modification of the hydroxyl group attenuates the activity of FasOH. Since the 4-nitrobenzyl group is too large to be accommodated in the ATP-binding pocket of ROCK, the substitution of the hydroxyl group is thought to reduce the activity of FasOH by disrupting an intramolecular hydrogen bond. To test whether the addition of a 4-nitrobenzyl group reduces the affinity of fasudil for ROCKs, we determined the cell-free kinase activity in the presence of the prodrug. Our supposition was largely validated, as the resulting prodrug lowered the potency against ROCK-II, with a IC₅₀ of 10.8 μM. Using reduction assays, we determined whether fragmentation of the prodrug had occurred and an active Rho kinase inhibitor was released. The Zn dust reduction method appears to yield amine-derived intermediate products, which are later reduced to FasOH. However, further studies are required to confirm the fragmentation of FasPRO into anilino-FasPRO and/or hydroxylamino-FasPRO. Enzymatic reduction by bactosomal human NADPH-CYP reductase under hypoxic conditions also yielded active FasOH, whereas under normoxic conditions the prodrug was stable. We also demonstrated that FasPRO was taken up and activated only under hypoxia in the case of PAH-afflicted pulmonary arterial cells. However, negligible amounts of fasudil and FasOH were detected when the cells were treated with FasPRO under normoxia. This result can be explained by the phenotype of PAH-afflicted cells. PAH-cells are considered to be hypoxic by nature. The phenotype of PAH-cells may not be entirely reversible when they are cultured under normoxia. Thus, the hypoxia-mediated bioreduction process may remain activated to some extent in these cells. Moreover, cell-based assays demonstrated the functional activities of FasPRO in hypoxia, and the difference between the functional activities of fasudil and FasPRO under normoxic conditions was significant. These results validate oxygen-dependent activation as a viable strategy for the design of ROCK inhibitor prodrugs.

This study suggests that the one-electron reductase-mediated activation of FasPRO may require severe/chronic hypoxia. FasPRO was activated in PAH-SMCs at 0.1% O₂, but not at 1% and 10% O₂. This result is consistent with the activation of CH-01, a hypoxia-activated prodrug of a Chk1/Aurora kinase inhibitor, where the 4-nitrobenzyl group of the inactive prodrug reduces and generates the active drug below 0.1% O₂ [60]. Targeting of severe pathologic hypoxia by FasPRO may also cover the moderate hypoxic regions found in certain normal tissues (3% O₂ and above). To some extent, FasPRO showed an antiproliferative effect when the cells were treated at normoxia, although the antiproliferative effect was observed at relatively high concentrations (>100 μM). Thus, we cannot exclude the possibility of hypoxia-independent activation of FasPRO at high concentrations. PR104H, a hypoxia-activated prodrug of nitrogen mustard, is activated by a hypoxia-independent pathway that utilizes a two-electron reduction process by aldo-keto reductase 1C3 (AKR1C3) [61, 62]. However, the expression of AKR1C3 in PAH-SMCs or PAH-ECs should have been investigated to rule in or rule out the involvement of a two-electron reduction process in the activation of FasPRO. The sensitivity of fasudil prodrug toward a particular hypoxic microenvironment can be tuned by using an appropriate protecting group. The propensity of the bioreduction process, termed the reduction potential, depends on the oxygen concentration at which the bioreductive groups are reduced [63]. The 4-nitrobenzyl group has a more negative reduction potential (−0.49 V) than the nitrothiophene- (−0.39 V) or nitrofuran-based group (−0.33 V), according to the observation that a smaller negative reduction potential indicates an increased ease of reduction at a reduced oxygen concentration [29]). Thus, the use of nitrothiophene- or nitrofuran-based protecting groups would allow fasudil to be activated under mild to moderate hypoxic conditions, although the reduction potential of a nitrobenzyl group is relatively small in cellular reductases. Thus, further studies are warranted to determine if other activation pathways play any roles in the reduction of FasPRO.

The development of ROCK inhibitors for indications that require systemic exposure has been largely hindered due to their adverse effects on systemic blood pressure and heart rate. The pharmacokinetics data (Fig. 5A–C) suggest that FasPRO was intact in the blood circulation and prolonged the circulation half-life of the drug by ~36-fold when compared with native fasudil. The absence of NADPH-CYP reductase in the blood indicates that FasPRO should remain in its intact form in the blood. When FasPRO was administered to healthy rats, systemic blood pressure did not show any noticeable changes. In contrast, fasudil treatment produced a large drop in systemic blood pressure in the rats. The reduction in the systemic pressure was reflective of the plasma half-life of fasudil in the blood. Even though the systemic exposure to the prodrug was longer than that of its parent compound, FasPRO was inactive. Thus, in addition to circumventing systemic effects, FasPRO is likely to reduce the dosing frequency and minimize the fluctuations in mSAP due to repeated injections or continuous infusion. Our results indicate that the prodrug may require hypoxic cells/tissues to be activated, as we have observed in in vitro studies. However, once FasPRO is activated by local hypoxia, the resulting fragmented fasudil might circulate rapidly in the blood and consequently elicit systemic side effects. The presence of a negligible amount of fragmented-fasudil in the cell lysates before or after treatment recovery points to the fact that fragmented-fasudil is unlikely to circulate in the blood stream and cause a major drop in peripheral blood pressure. Overall, this proof-of-concept study revolves around the use of hypoxia-triggered FasPRO for the treatment of PAH, a rare and devastating disease. FasPRO is very likely to reduce the mean pulmonary arterial pressure (mPAP) in animal models of PAH without significantly impacting the systemic pressure. Further studies are required to demonstrate the efficacy of FasPRO in disease models of PAH or cancers.

The study is not without limitations. Further chemical characterization might explain the unusual fragmentation of FasPRO to fasudil. The plasma-concentration-time profile data do not provide conclusive evidence as to whether the C_max has resulted from the fragmentation of FasPRO or from the metabolite of fasudil. Since the fragmentation of FasPRO to FasOH (the main pathway) is hypoxia-dependent, the fragmentation of FasPRO to FasOH in healthy rats is unlikely to occur. But the unusual fragmentation of FasPRO to fasudil (unwanted pathway) in healthy rats may lead to the formation of its active metabolite, FasOH. Further, the equal mass dose (3 mg/kg) does not translate into equal molar doses due to the difference in the molecular weights of FasPRO (443.7) and fasudil (292.2), which would be 6.8 μmol/kg and 10.3 μmol/kg, respectively. However, these differences in dose are unlikely to change the conclusion of the pharmacokinetics data because the drug half-life and elimination rate constant are independent of drug dose, whereas drug dose is more relevant to the bioavailability and pharmacodynamics of the drug [43].

Rho-kinase plays roles in the contraction of peripheral as well as pulmonary arteries. However, the use of the Rho-kinase inhibitor fasudil as a viable therapy in PAH has not come to fruition because of its pronounced peripheral vasodilatory effect. Inhaled fasudil also showed a lack of pulmonary selectivity in animal and human studies, and caused systemic hypotension and exerted cardio-depressive effects [23]. FasPRO is devoid of systemic effects and is likely to show pulmonary selectivity in diseases like PAH. A prolonged circulation time is likely to facilitate penetration of the drug from the vascular bed into hypoxic regions and consequently increase its therapeutic efficacy. In tumors, hypoxia increases the expression of immune regulatory proteins such as PD-1/PD-L1, which inhibit proliferation and activation of cytotoxic CD8+ T-lymphocytes. Studies have shown that ROCK blockade in combination with immunogenic therapies increases the activation and access of immune cells to tumors [64–66]. Thus, FasPRO could be an attractive therapy for cancer where drug accumulation at the tumor site is desirable. We envision to assess the therapeutic efficacy of the prodrug in various types of cancer cells, including radio-sensitive and radio-resistant cell lines. Importantly, FasPRO can potentially be used in pathological conditions where hypoxia is implicated in the genesis of the disease such as cerebral ischemia, neuroblastoma, and myocardial infarction.