Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Cells require adequate nucleotide levels for normal cellular metabolism and proliferation, including the synthesis of RNA and DNA. Nucleotides can be made available via two distinct mechanisms, the salvage pathway and by de novo synthesis. Using the salvage pathway, cells recycle available nucleosides and nucleobases, whereas with de novo synthesis the purine or pyrimidine ring systems of the nucleotides are assembled in a stepwise manner (for review, see Allison and Eugui, 2000). Different cell types rely on these two pathways of nucleotide biosynthesis to varying degrees. Proliferation of lymphocytes relies more on the de novo nucleotide synthesis pathway compared with other cell types. This observation makes enzymes of the de novo pathway an attractive target for pharmacological intervention aimed at inhibiting lymphocyte proliferation (Allison and Eugui, 2000).

The enzyme inosine 5'-monophosphate dehydrogenase (IMPDH) catalyzes an essential step in the de novo biosynthesis of guanine nucleotides, namely, the conversion of inosine 5'-monophosphate to xanthosine 5'-monophosphate. Using specific inhibitors of IMPDH, it has been shown that the proximal biochemical effect of blocking IMPDH in sensitive cell types such as lymphocytes is a decrease in cellular guanine nucleotide levels. This in turn blocks RNA and DNA synthesis, and hence proliferation, for which adequate nucleotide levels are required. IMPDH inhibition resulting in a decrease of guanine nucleotides causes reversible antiproliferative, antiviral, and antiparasitic effects (for review, see Franchetti and Grifantini, 1999).

The pharmacological effects of IMPDH inhibition have been exploited by a number of marketed products. For example, mycophenolic acid (MPA), a potent, uncompetitive IMPDH inhibitor, has been demonstrated to inhibit the proliferation of lymphocytes (Eugui et al., 1991a; Allison and Eugui, 2000). An ester prodrug of MPA, mycophenolate mofetil (CellCept), has been developed and approved for the prevention of acute rejection in kidney, heart, or liver transplantation (for review, see Mele and Halloran, 2000) when used in combination with steroids and cyclosporine A (CsA). Mizoribine (Bredinin) and ribavirin (Virazole, Rebetol) are nucleoside analogs and after intracellular phosphorylation are competitive IMPDH inhibitors (Franchetti and Grifantini, 1999). Mizoribine is approved in Japan for multiple indications, including prevention of rejection after renal transplantation, idiopathic glomerulonephritis, lupus nephritis, and rheumatoid arthritis (Ishikawa, 1999). Ribavirin is approved as an inhaled antiviral agent for treatment of respiratory syncytial virus and, orally in combination with interferon-, for the treatment of chronic hepatitis C viral infection (Davis et al., 1998; McHutchison et al., 1998; Poynard et al., 1998).

Therapeutic uses of currently available IMPDH inhibitors are limited by unfavorable tolerability profiles. Furthermore, nucleoside analogs may have pharmacological activity other than inhibition of IMPDH. Given the limitations of available drugs, we and our colleagues have designed novel, non-nucleoside IMPDH inhibitors using a rational drug design approach (for review, see Sintchak and Nimmesgern, 2000). The first crystal structure of the target enzyme was solved at Vertex Pharmaceuticals, Inc. (Sintchak et al., 1996). Several X-ray structures of inhibitor-bound IMPDH have since been published (for review, see Sintchak and Nimmesgern, 2000). This article summarizes information regarding the in vitro and in vivo pharmacological activity of VX-148. VX-148 is a selective, highly potent, reversible, and uncompetitive inhibitor of both isoforms of human IMPDH. VX-148 is the second lead IMPDH compound to be selected by us for clinical development. Based on the rapid progress in preclinical evaluation, VX-148 has already entered phase I clinical trial. The first IMPDH inhibitor selected from this program was VX-497, which is currently in phase II trials for the treatment of hepatitis C (Markland et al., 2000; Jain et al., 2001).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents. Most cell culture reagents, unless specified otherwise, were obtained from Invitrogen (Carlsbad, CA) or JRH Biosciences (Lenexa, KS). Complete RPMI 1640 medium was prepared by adding up to a final of 10% fetal bovine serum, 55 µM -mercaptoethanol, 50 units/ml penicillin with 50 µg/ml streptomycin, 300 µg/ml L-glutamine, and 10 mM HEPES, pH 7.5, to RPMI 1640 medium. Stock solutions of guanosine (Sigma-Aldrich, St. Louis, MO) and phytohemagglutinin (PHA)-P (Difco, Detroit, MI) were prepared in fetal bovine serum-free RPMI 1640 medium and stored at 20°C. Staphylococcal protein A immobilized on Sepharose CL-4B (SPAS; Amersham Biosciences, Piscataway, NJ) was freshly prepared on the day of the assay in complete RPMI 1640 medium. VX-148 (synthesized in-house) and MPA (Sigma-Aldrich; Calbiochem, La Jolla, CA) were dissolved in DMSO at a concentration of 20 mM and stored at 20°C. CsA was purchased from Novartis (Basel, Switzerland). Tritiated thymidine and leucine were purchased from PerkinElmer Life Sciences (Boston, MA).

IMPDH Type I and II Enzyme Assays. The expression and purification of recombinant IMPDH type I and type II proteins was done as described previously (Fleming et al., 1996; Nimmesgern et al., 1999). Other dehydrogenases were purchased from Sigma-Aldrich and used as per their recommendation. The IMPDH assay was performed in 200-µl assay volume containing a final concentration of 100 mM KH₂PO₄, 0.5 mM EDTA pH 8, 5% DMSO, 200 µM inosine 5'-monophosphate, 200 µM NAD, 2 mM dithiothreitol, and 50 nM IMPDH. Initially, 10 µl of compound dissolved in DMSO was added to 180 µl of the reaction buffer. The reaction was incubated at 37°C for 10 min. The reaction was initiated with 10 µl of enzyme (1 µM enzyme in buffer with 40 mM dithiothreitol) and read at 340 nm at 37°C for 10 min.

Isolation of Peripheral Blood Mononuclear Cells (PBMCs) and T Cells. Human venous blood was drawn from normal healthy volunteers using heparin as an anticoagulant. PBMCs were isolated from blood by centrifugation over Ficoll-Paque gradient or CPT tubes (BD Biosciences, San Jose, CA) using standard conditions. PBMCs were harvested, washed, and resuspended in complete RPMI 1640 medium, counted, and diluted to 1 × 10⁶ cells/ml. T cells were purified from PBMCs using Lymphokwik reagents and protocol (One Lambda, Inc., Canaga Park, CA). The resulting preparation was stained with anti-CD3-FITC (BD Biosciences) to confirm >95% purity of T cells. Isolation and preparation of mouse and rat lymphocytes was done as per Jain et al. (2001).

PBMC and Splenocyte Proliferation Assays. Cells (5 × 10⁴ for human PBMC T and purified T cells), 1 × 10⁵ cells (for human PBMC B cells), or 1 × 10⁵ cells (for rat PBMC T or B cells) were added per well to 96-well plates. For mouse and rat splenocytes, 5 × 10⁵ mononuclear cells were plated. PHA was added to a final concentration of 10 to 20 µg/ml/well. The mitogens SPAS (final concentration of 2 mg/ml/well; for human) and lipopolysaccharide (final concentration of 4 µg/ml/well; for mouse and rat) were used to stimulate B cells.

Plaque-Forming Cell (PFC) Assay. Adult CD-1 mice (Charles River Laboratories, Inc., Wilmington, MA), sheep red blood cells (SRBCs; Charles River Pharm Services, Margate, Kent, UK), guinea pig complement (Invitrogen), agar (Difco), and DEAE-Dextran (Amersham Biosciences) were obtained from specified sources. Six (VX-148 or MPA dose groups) or 10 (vehicle) mice were tested in each treatment group. The vehicle used for VX-148 and MPA comprised 10% Cremophor-EL (Sigma-Aldrich), 10% Labrasol (GatteFosse, Saint-Priest Cedex, France), 25% polyethylene glycol 400 (Sigma-Aldrich), 25% propylene glycol (Sigma-Aldrich), and 30% water. The PFC assays were done as per Jain et al. (2001); discrete areas of hemolysis or plaques were counted and the number of PFCs per 10⁶ spleen cells calculated from duplicate platings.

Mouse Skin Transplant Model. The mouse trunk skin transplant model was established using published methods (Gardner, 1995). Donor (BALB/c) trunk skin was removed and kept cold in saline before grafting on recipient C57BL/6 mice. Male mice (n = 10) were dosed orally for 11 days (where day 0 is the day of transplant surgery, and the mice are administered vehicle or drug once, 30 min before surgery. CsA was administered at 50 mg/kg b.i.d. in 1% CMC in water, and VX-148 was administered at 25, 50, or 100 mg/kg b.i.d. in the vehicle used for the PFC assay at a 10-ml/kg dose volume. For each study, appropriate vehicle-treated control groups were run concurrently. One mouse each in the vehicle and 100-mg/kg VX-148 dose group died during the anesthesia process. Graft rejection was quantified as the number of days to reach R4 rejection (>75% of graft scabbed). Statistical analyses were performed using Kaplan Meier log-rank test comparisons with vehicle-treated group.

Pharmacokinetics. Male and female CD-1 mice (Harlan, Indianapolis, IN) were used in this study. VX-148 was administered orally via gavage in the vehicle used for PFC and skin transplant studies. Dose volumes of 10 ml/kg were used. Blood samples were collected at 0.25, 0.5, 1, 1.5, 2, 3, 4, 6, and 8 h after compound administration. Blood samples were centrifuged at 12,000 rpm to obtain plasma, which was stored at 70°C pending HPLC analysis. Noncompartmental pharmacokinetic analytical method was used for the analysis of the plasma concentration data (Gibaldi and Perrier, 1982). The maximum measured plasma concentration (C_max) and the corresponding time (T_max) were recorded from the observed data. The area under the plasma concentration-time curve from zero to the time of last measurable plasma concentration (AUC_0-t), and from zero to infinity (AUC_0-) and the terminal elimination half-life were calculated. The terminal phase elimination rate constant (z) was estimated by a log-linear regression, using the last three to four nonzero plasma concentration data. AUC was computed using the linear-trapezoidal rule from the measured plasma concentration data. The last observed plasma concentration was used to calculate the extrapolated AUC.

Quantitation of VX-148 by HPLC-UV Assay. Plasma (180 µl) was precipitated with 200 µl of acetonitrile and 10 µl of a saturated zinc chloride solution. A Hewlett Packard 1090 was used for HPLC analysis. Supernatant (80 µl) was injected onto a Zorbax SB-CN column (250 × 4.6 mm, 5 µm; Agilent Technologies, Palo Alto, CA), and the analyte was chromatographically separated with a linear gradient. The linear gradient contained an initial mobile phase (0.1% trifluoroacetic acid, 75% water, 2.5% methanol, and 2.5% acetonitrile) to a final mobile phase (0.1% trifluoroacetic acid, 40% water, 54% methanol, and 6% acetonitrile), which was initiated 3 min postinjection and maintained for the next 3 min. The column temperature was maintained at 60°C and the mobile phase flow rate was maintained at 1 ml/min. The analyte was monitored at 305 nm.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The chemical structure of VX-148 is novel, with a molecular weight of 435.48 (Fig. 1). VX-148 inhibits IMPDH type I enzyme with a K_i value of 14 nM and type II enzyme with a K_i value of 6 nM. The IC₅₀ values of VX-148 are very similar to the potency of MPA (6-10 nM) and VX-497 (7-10 nM) against the purified recombinant human IMPDH type II enzyme (Eugui et al., 1991a; Jain et al., 2001). The mechanism of inhibition of IMPDH by VX-148 has been confirmed as being uncompetitive, as is the case for VX-497 (Saunders and Raybuck, 2000). VX-148 binds at the NAD⁺ cofactor binding site (Sintchak et al., 1996; Sintchak and Nimmesgern, 2000). VX-148 did not inhibit other NAD or NADPH dehydrogenases (glucose dehydrogenase, glucose-6-phosphate dehydrogenase, 6-phosphogluconic acid dehydrogenase, and isocitrate dehydrogenase) at concentrations up to 50 µM.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Structural formulae of VX-148 and VX-497.

The immunosuppressive activity of VX-148 was evaluated in several cell-based assays. VX-148 was found to potently inhibit the uptake of [³H]thymidine by human PBMCs stimulated with either T- or B-cell mitogens (Table 1). The IC₅₀ values for VX-148 obtained in T (PHA-stimulated) and B (SPAS-stimulated) lymphocytes were 82 and 73 nM, respectively. These IC₅₀ values are slightly lower than those obtained previously with MPA (90 and 127, respectively), or VX-497 (104 and 132, respectively) (Jain et al., 2001). The addition of 50 µM guanosine to the cell culture incubations blocked the inhibition of proliferation by VX-148 substantially (IC₅₀ values of 8514 and 4668 nM in T and B cells, respectively), demonstrating that VX-148 is a specific inhibitor of IMPDH. A reversal of inhibition produced by MPA is also observed in the presence of 50 µM guanosine with IC₅₀ values being >20 µM (Table 1).

IC₅₀ values for VX-148 were also determined in PBMC proliferation assays where [³H]leucine uptake was measured instead of [³H]thymidine uptake. These assays were performed to confirm that the compounds tested in the [³H]thymidine uptake assay indeed demonstrated inhibition of cell proliferation rather than any nonspecific effect on the uptake of thymidine. The IC₅₀ values obtained with [³H]leucine were almost identical to the values obtained in the [³H]thymidine uptake assays (data not shown). To confirm that VX-148 inhibited a late step in T- and B-cell proliferation, as expected for an IMPDH inhibitor, VX-148 was added at 48 h instead of at the beginning of the proliferation assay. The IC₅₀ values for both T and B cells were comparable whether VX-148 was added immediately or 48 h after the addition of mitogens in the 3-day proliferation assay (data not shown). The IC₅₀ value for VX-148 obtained in purified T cells stimulated with PHA was comparable with that of MPA (Table 1). VX-148 was also able to inhibit the proliferation of lymphocytes in 25% concentration of human whole blood, but the IC₅₀ value was increased ~10-fold over that observed with purified PBMCs (Table 1), presumably due to the presence of serum proteins and other blood cells that may bind VX-148. The IC₅₀ value of MPA, however, was similar to that in purified lymphocytes and did not seem to be altered by serum proteins. The reason for this difference between VX-148 and MPA is not known.

VX-148 inhibited the proliferation of lymphocytes obtained from mice and rats (Table 2). The IC₅₀ values of VX-148 obtained in mouse and rat splenocytes, stimulated with T- or B-cell mitogens, were similar to those obtained for mitogen-stimulated human PBMCs (Table 1). Unlike human lymphocytes where equivalent activity of MPA and VX-148 was observed, MPA was often 2- to 3-fold more potent than VX-148 in a side-by-side comparison in murine cells. However, in a previous study comparing VX-497 with MPA, 2- to 3-fold higher IC₅₀ values for MPA were obtained (Jain et al., 2001), which are similar to the ones observed for VX-148 in this study.

To determine the range of specific cell types that are sensitive to VX-148, several immortalized lymphoid and nonlymphoid cell lines were evaluated (Table 3). VX-148-mediated inhibition of proliferation of the L1210 and Jurkat T cell lines, Raji B cell line, and U937 monocytic cell line was assessed by the use of XTT dye, which is dependent on the activity of cellular mitochondrial dehydrogenases. With the exception of Jurkat T cells, VX-148 was observed to be equipotent to MPA in the lymphoid cell lines tested (Table 3). In the Jurkat T cells, VX-148 was more potent than MPA. The antiproliferative effect of VX-148 was evaluated in two human fibroblast cell lines, WI38 and Hs-68 (Table 3). The IC₅₀ values for VX-148 were greater than 20 µM in both fibroblast cell lines, demonstrating specific inhibition of proliferation of cell types such as lymphocytes that are most dependent on de novo guanine nucleotide synthesis. In summary, VX-148 is a potent inhibitor of several lymphocytic cell lines, as expected for an IMPDH inhibitor, without demonstrating any general cytotoxicity in nonlymphocytic cell types such as fibroblasts.

Based on the potent and selective inhibition of lymphocytes in vitro, and good oral bioavailability in mice (Table 4), VX-148 was evaluated in vivo in a murine PFC assay, a model of primary antibody production in response to antigenic stimulation. Female CD-1 mice were immunized with SRBCs on day 1. Oral gavage treatment with VX-148 was begun immediately after SRBC injection and continued twice daily through day 4. On the morning of day 5, spleen cells were harvested and plated together with complement and SRBCs in an agar solution. Plaques formed by IgM type antibody-secreting splenocytes were enumerated.

VX-148 was found to produce a significant reduction in plaque formation (Fig. 2). At 25, 50, and 100 mg/kg b.i.d., VX-148 reduced plaque formation by 25, 66, and 86%, respectively, relative to vehicle-treated controls (Fig. 2A). In the same experiment MPA inhibited plaque formation by 99% when administered at 50 mg/kg b.i.d. (Fig. 2A). VX-148 was evaluated in two independent PFC experiments, and the ED₅₀ value for VX-148 was calculated to be 38 mg/kg from the two experiments (Fig. 2B). The 50- and 100-mg/kg doses of VX-148 were statistically significant in both experiments with the p value being <0.01 using the Student's t test. These studies indicated that VX-148 had the ability to suppress T-cell-dependent antibody production after an antigenic challenge. All doses of VX-148 seemed to be well tolerated by mice dosed twice daily for five consecutive days in that mice in all dose-groups gained weight similarly to the vehicle-dosed group and showed no apparent distress. The ED₅₀ values for MPA and VX-497 previously calculated from similar PFC assays have been 30 to 35 mg/kg (Eugui et al., 1991b; Jain et al., 2001). Hence, VX-148 demonstrated efficacy in this mouse model of immunosuppression at doses similar to that of VX-497 and MPA.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Efficacy of VX-148 in murine plaque-forming cell assays. A, VX-148, MPA, or vehicle was dosed orally twice daily at the indicated concentrations in mice challenged with SRBC antigens. Six mice were dosed for each VX-148 or MPA dose group, and 10 mice were dosed in the vehicle group. Antibody-secreting cells were enumerated as plaques, as described under Materials and Methods, and normalized as number of PFCs per million spleen cells. The error bars represent standard deviations. Values that are statistically significant compared with the vehicle group are indicated by asterisks and were calculated using the Student's t test. B, number of plaques, expressed as the percentage of plaques observed in vehicle-dosed mice, are shown from two independent experiments.

Skin transplant rejection is a strong immune response and serves as a very sensitive test of the immunosuppressive potential of drugs in organ transplantation and graft rejection. VX-148 and CsA were compared in two MHC mismatched strains (donor, BALB/c; and recipient, C57BL/6) for skin transplant rejection performed as described by Gardner (1995). Graft rejection was quantified as the number of days to reach R4 rejection (>75% of graft scabbed). The mean rejection time for the vehicle was 11.2 ± 1.92 days and 13.7 ± 1.64 (p = 0.015) for CsA. VX-148 treatment was found to increase graft survival to 12.8 ± 0.92 (p = 0.16) at 25 mg/kg, to 12.7 ± 1.25 at 50 mg/kg (p = 0.19) and to 13.4 ± 1.59 (p = 0.051) at 100 mg/kg b.i.d. (Fig. 3). The immunosuppressive effects of VX-148 were dose-dependent and indistinguishable from those of CsA at 50 mg/kg b.i.d., which is the maximum tolerated dose of CsA in mice. These results are comparable with those obtained with MPA and VX-497 in similar skin transplantation experiments where administration of 50 mg/kg b.i.d. MPA (12.4 ± 1.4; p < 0.05; n = 7) or VX-497 (13.2 ± 1.2; p < 0.001; n = 17) significantly prolonged skin graft survival (Decker et al., 2001). All doses of VX-148 seemed to be well tolerated by mice dosed twice daily for up to 10 consecutive days.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Graft survival in a mouse skin transplantation model. Vehicle, VX-148, or CsA was dosed orally twice at the indicated concentrations in C57BL/6 mice grafted with trunk skin from BALB/c mice, as described under Materials and Methods. Statistical analyses were performed using the Kaplan Meier log-rank test comparisons and are shown on the figure.

Oral administration of VX-148 to male and female CD-1 mice at 100 mg/kg resulted in comparable plasma AUCs (38-48 µg × h/ml), C_max values (22 µg/ml for males, 17 µg/ml for females), and dose-proportional exposures (Table 4). In the PFC study (Fig. 2A), a significant inhibition of plaque formation is observed in both the 50- and 100-mg/kg dose groups, but not in the 25-mg/kg dose group. In addition, the response at 100 mg/kg is greater than that observed at 50 mg/kg, indicating that the maximal effect is not achieved at 50 mg/kg. In the skin transplant study (Fig. 3), a significant response is achieved only at the 100-mg/kg dose. These findings indicate that although exposures of approximately 14 µg × h/ml result in an immunosuppressive effect in the PFC assay, higher exposures of VX-148 may be required to observe a significant response in the more stringent skin transplantation model.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

MPA or MMF was the first successful IMPDH inhibitor to be used for the prevention of kidney, heart, and liver transplant rejection. However, there remain a number of other immune cell-mediated, chronic, inflammatory disorders that may benefit from treatment with immunosuppressive drugs such as MMF (for review, see Allison and Eugui, 2000). These include rheumatoid arthritis, myasthenia gravis, many renal nephropathies such as lupus nephritis; several skin diseases such as psoriasis, dyshidrotic eczema, and pemphis vulgaris; Crohn's and other inflammatory bowel diseases; and ocular inflammation. Even within transplantation-related applications, prevention of chronic Graft versus Host disease remains a challenge. Hence, there is a vast unmet clinical need for potent, reversible, lymphocyte-selective inhibitors that are suitable for long-term dosing, particularly with respect to lack of hepatotoxicity and gastrointestinal side effects.

We have developed several new, non-nucleoside, orally bioavailable IMPDH lead compounds that are structurally unrelated to other IMPDH inhibitors (Saunders and Raybuck, 2000; Sintchak and Nimmesgern, 2000). VX-497 was the first chemically distinct IMPDH inhibitor advanced into clinical trials by us since MMF. Illustrating the applicability of IMPDH inhibitors in a number of indications, VX-497 was simultaneously evaluated in two phase II trials for two different indications, psoriasis and hepatitis C. It is now being advanced to a second 12-month phase II triple combination trial for the treatment of hepatitis C. Hepatitis C viral patients will be treated either with VX-497, pegylated interferon  and ribavirin, or with placebo in combination with pegylated interferon  and ribavirin. We are, therefore, developing a second generation IMPDH inhibitor, VX-148, for the treatment of autoimmune and immune-mediated inflammatory disorders. As described in this study, VX-148 has an in vivo and in vitro immunosuppressive potency similar to that of VX-497 and MPA. It is a potent, specific, and reversible IMPDH inhibitor in vitro. The cytostatic effect of VX-148 is most pronounced on lymphocytes, correlating well with the known dependence of lymphocytes on the de novo pathway for the synthesis of guanine nucleotides. VX-148 is slightly more potent than MPA or VX-497 in its ability to limit the proliferation of human, mouse, and rat lymphocytes. VX-148 does not inhibit the proliferation of fibroblast cells, demonstrating its lack of general cytotoxicity. In in vivo studies, VX-148 had favorable pharmacokinetic properties and excellent oral bioavailability. In two different animal models of immunosuppression, VX-148 was as effective as clinically relevant drugs such as MPA (in the PFC assay) and cyclosporin (in the skin transplantation study), making it a promising drug candidate for the treatment of immune disorders.