Cyclin-dependent kinase inhibition limits glomerulonephritis and extends lifespan of mice with systemic lupus

Authors


  • Presented in part at the Annual Meeting of the American Society of Nephrology, St. Louis, MO, October 27–November 1, 2004.

Abstract

Objective

To examine whether the cyclin-dependent kinase (CDK) inhibitor seliciclib ameliorates autoimmune nephritis in (NZB × NZW)F1 mice.

Methods

In experiment 1, NZB × NZW mice received seliciclib (100 mg/kg or 200 mg/kg) or vehicle by gavage, beginning at age 2 months and ending at 8 months of age. In experiment 2, seliciclib (200 mg/kg) was administered alone or combined with low-dose methylprednisolone, starting at age 5 months, when immune complex deposition in the kidney had already occurred. Animals were followed up until all vehicle-treated mice died. In 2 additional groups of NZB × NZW mice treated with seliciclib or vehicle from 2 months of age until 5 months of age, splenocytes were isolated and tested ex vivo for T cell and B cell activity.

Results

Seliciclib, given at an early phase of disease, prolonged survival, delayed the onset of proteinuria and renal function impairment, and protected the kidney against glomerular hypercellularity, tubulointerstitial damage, and inflammation. Combining seliciclib with low-dose methylprednisolone in mice with established disease extended the lifespan and limited proteinuria and renal damage more than treatment with either agent alone. Seliciclib limited immunologic signs of disease, reducing glomerular IgG and C3 deposits and levels of serum anti-DNA antibodies. Moreover, it inhibited ex vivo T cell and B cell proliferative responses to polyclonal stimuli. T cell production of interferon-γ and interleukin-10 and B cell release of IgG2a were reduced by treatment with seliciclib.

Conclusion

These findings suggest that CDK activity may be a useful target in the treatment of systemic lupus erythematosus. A direct immunomodulatory action of seliciclib on T cells and B cells may be one of the mechanisms underlying the beneficial effects.

Systemic lupus erythematosus (SLE) is a chronic autoimmune disease characterized by activation and proliferation of autoreactive T cells and B cells with production of autoantibodies against nuclear and endogenous antigens, which results in the development of immune complex–mediated glomerulonephritis and renal failure (1–3). Circulating autoantibodies form complexes that are retained in the kidney. Complement activation and recruitment of T cells and macrophages follow. Cytokines generated by inflammatory and renal cells, in turn, trigger mesangial cell proliferation and enhance manifestations of renal disease. Treatment of SLE includes immunosuppressants and cytostatic agents, with extensive use of steroids; these medications have side effects (4). Murine models of SLE have been of considerable help in clarifying disease mechanisms. As in humans, proliferation of T cells and B cells and glomerular hypercellularity manifest in mouse SLE and contribute to disease pathogenesis (5).

Cell proliferation is dependent on the coordinated activation of specific cell-cycle regulatory proteins; cyclins and cyclin-dependent kinases (CDKs) act as positive regulators, while CDK inhibitors are their negative counterpart (6). Cyclins and CDKs form heterodimers that phosphorylate key substrates such as the retinoblastoma protein, triggering the release of the transcription factor E2F1, which stimulates the expression of genes involved in DNA replication. Perturbations in the cyclin–CDK complexes have been shown to alter immune system function, such that mice deficient for the CDK inhibitor p21 exhibited excessive T cell activation and proliferation, loss of tolerance to nuclear antigens, and increased susceptibility to lupus (7). Conversely, administration of a p21Waf1/Cip1 peptide mimic halted disease progression in lupus mice (8), suggesting that inhibition of CDK activity may be a useful strategy for therapeutic intervention in lupus.

Seliciclib (R-roscovitine; CYC202) is a 2,6,9-substituted purine analog that competes with ATP for binding to the active site of CDKs and potently inhibits CDK-2–cyclin E, CDK-7–cyclin H, and CDK-9–cyclin T (9). Seliciclib has been shown to inhibit in vitro the growth of different tumor cell types (10–12). Seliciclib is currently in phase II clinical trials in patients with solid tumors and B cell malignancies (13), offering a strong advantage over other molecular approaches that are yet to achieve translation into the clinic.

In the present study, we examined whether seliciclib interferes with T cell and B cell activity in vitro, as a preliminary step to its use in autoimmune lupus nephritis. We then investigated whether seliciclib affects the course of the disease and animal lifespan in the NZB × NZW SLE-prone mouse model (14, 15).

MATERIALS AND METHODS

Experimental design.

Female (NZB × NZW)F1 mice (Harlan, Milan, Italy) and CD-1 mice (Charles River, Calco, Italy) were used. Animal care and treatment were conducted to conform with the institutional guidelines that are in compliance with national (Decreto Legislativo n.116, Gazzetta Ufficiale suppl 40, 18 febbraio 1992, Circolare n.8, Gazzetta Ufficiale 14 luglio 1994) and international laws and policies (EEC Council Directive 86/609, OJL358-1, December 1987; Guide for the Care and Use of Laboratory Animals, US National Research Council, 1996). Animals were housed in a constant-temperature room, with a 12-hour dark/12-hour light cycle, and fed a standard diet.

In experiment 1 (prevention study), oral administration of seliciclib (Cyclacel, Dundee, UK) (100 mg/kg or 200 mg/kg) or vehicle (HCl, 50 mM) by gavage began at 2 months of age, and the experiment was terminated at 8 months. In experiment 2 (therapy study), seliciclib (200 mg/kg) was administered alone (n = 15 mice) or combined with low-dose methylprednisolone (MP; 1.5 mg/kg intraperitoneally) (16, 17) (n = 16 mice) starting at 5 months, when immune complex deposition in the kidney has already occurred (18). Twelve mice received MP alone, and 10 mice received vehicle. Animals were followed up until all vehicle-treated mice died. Seliciclib dosages were chosen on the basis of data from other experimental settings (9).

Two additional groups of NZB × NZW mice (n = 5 mice in each group) were treated with seliciclib (200 mg/kg) or vehicle, from 2 months to 5 months of age. When the mice were killed, their spleens were removed, and splenocytes were isolated and tested ex vivo for T cell and B cell activity. Age-matched control CD-1 mice (n = 5) were studied in parallel.

In vitro T cell proliferation and cytokine release.

Spleen cells were isolated by stainless steel screen filtration followed by hypotonic lysis of erythrocytes. Spleen cells were cultured in complete RPMI medium supplemented with 10% heat-inactivated fetal calf serum (FCS) and antibiotics (Invitrogen, Paisley, UK). The proliferation assays were conducted in 24-well plates in which spleen cell suspensions (5 × 105) were cultured for 72 hours under resting conditions or with concanavalin A (Con A; 10 μg/ml) (Calbiochem-Novabiochem, La Jolla, CA) alone or in the presence of seliciclib. T cell proliferation in response to anti-CD3/anti-CD28 antibodies was assessed by coating 96-well flat-bottomed plates with anti-CD3 monoclonal antibody (mAb) (0.5 μg/well; hamster anti-mouse CD3, clone 7D169; Serotec, Oxford, UK) overnight at 4°C. Wells were then washed, and spleen cells (5 × 105/well) were added to the wells with 0.2 μg/well anti-CD28 mAb (hamster anti-mouse CD28, clone 37.51.1; Caltag, Burlingame, CA) alone or in the presence of seliciclib. Cell proliferation was determined by pulsing the cells with 3H-thymidine during the last 14–16 hours of culture and measuring the radioactivity incorporated by liquid scintillation counting. Results of the proliferative response were expressed as counts per minute.

Further aliquots of the spleen cell suspension (4 × 106/ml) were incubated in 24-well plates under resting conditions or with 10 μg/ml Con A, alone or in the presence of seliciclib. Cultures were then stopped after 24 hours and 72 hours, respectively, for assessment of interferon-γ (IFNγ) and interleukin-10 (IL-10) production. Cytokine levels were measured in supernatants, using commercially available enzyme-linked immunosorbent assay (ELISA) kits (Endogen-Pierce, Rockford, IL).

In vitro B cell proliferation and IgG2a production.

Splenocytes (5 × 105) were resuspended in complete RPMI supplemented with 10% FCS and incubated for 4 days under resting conditions or with 10 μg/ml lipopolysaccharide (LPS) from Escherichia coli serotype 026:B6 (Sigma, St. Louis, MO), alone or in the presence of seliciclib (1, 10, or 30 μM). During the last 18 hours, cultures were pulsed with 1 μCi 3H-thymidine and harvested. The results of proliferative response were expressed as counts per minute. Supernatants were collected, and IgG2a levels were measured by isotype-specific ELISA (Bethyl Laboratories, Montgomery, TX).

Flow cytometric analysis of splenocyte subsets.

Splenocyte cell-surface phenotype analysis was performed by flow cytometric analysis using FACSsort (Becton Dickinson, Mountain View, CA). The following antibodies were used: rat anti-mouse immunoglobulin κ-chain (clone Hb58, used as tissue culture supernatant) (American Type Culture Collection, Rockville, MD) followed by fluorescein isothiocyanate (FITC)–conjugated F(ab′)2 goat anti-rat IgG as secondary antibody (Southern Biotechnology, Birmingham, AL), FITC-conjugated rat anti-mouse CD4 (clone CT-CD4; Caltag, Burlingame, CA), R-phycoerythrin (R-PE)–conjugated anti-mouse CD44 (clone IM7; BioLegend, San Diego, CA), and R-PE–conjugated anti-mouse CD86 (clone GL1; BD Biosciences Europe, Erembodegem, Belgium). To block nonspecific binding, a 30-minute preincubation with 5% rat serum was performed. All stainings included negative controls from which the primary antibodies were omitted or with isotype antibodies as appropriate. Light scattering parameters were set to exclude dead cells and debris.

Biochemical parameters.

Urinary protein concentrations were determined with the Coomassie blue G dye binding assay, using bovine serum albumin (BSA) as standard. Baseline values for urinary protein excretion, as measured in 2-month-old NZB × NZW mice, were <4 mg/day. Thus, levels exceeding 4 mg/day during the subsequent followup period were considered abnormal. Renal function was assessed as the concentration of blood urea nitrogen (BUN) in heparinized blood, using the Reflotron test (Roche, Indianapolis, IN). BUN levels exceeding 30 mg/dl were considered abnormal (in our laboratory, normal range 14–29 mg/dl). Serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels were measured using the Reflotron test. ELISA kits were used for evaluation of serum anti–double-stranded DNA (anti-dsDNA) autoantibodies (Diastat anti-dsDNA kit, Bouty Laboratory, Milan, Italy) (17) and serum IgG levels (Bethyl Laboratories).

Renal morphology.

Fragments of renal cortex were fixed in Duboscq-Brazil solution, dehydrated in alcohol, and embedded in paraffin. Sections (3 μm) were stained with hematoxylin and eosin, Masson's trichrome, and periodic acid–Schiff (PAS). Glomerular intracapillary hypercellularity was scored on a scale of 0 to 3+, where 0 = absent, 1+ = mild, 2+ = moderate, and 3+ = severe. Extracapillary hypercellularity was graded according to the percentage of glomeruli involved, as follows: 0 = absent, 1+ = <25%, 2+ = 25–50%, and 3+ = >50%. Glomerular immune deposits were also graded according to the percentage of glomeruli involved, as follows: 0 = no deposits, 1+ = <25%, 2+ = 25–50%, and 3+= >50%. At least 100 glomeruli were examined for each biopsy specimen. Tubular changes (atrophy, casts, and dilatation) and interstitial inflammation were graded according to the percentage of the sample affected, as follows: 0 = no changes, 1+ = <25%, 2+ = 25–50%, and 3+ = >50%. At least 10 fields per sample were examined at low magnification (× 10) for scoring of the interstitium. Biopsy specimens were analyzed by the same pathologist, in a single-blind manner.

Immunofluorescence analysis.

Studies were performed on frozen tissue embedded in OCT compound. Sections (3 μm) were fixed in acetone, washed with phosphate buffered saline (PBS) (pH 7.4) for 10 minutes, blocked with PBS/1% BSA for 15 minutes, and stained with fluorescein isothiocyanate–conjugated antibodies to mouse IgG (Sigma-Aldrich, St. Louis, MO), C3 (Cappel, West Chester, PA), or CD4 (Caltag, Burlingame, CA). Glomerular staining intensity for IgG and C3 was graded on a scale from 0 to 3+, as follows: 0 = absent, 1+ = faint, 2+ = intense, and 3+ = very intense. CD4+ T cells were counted in 20 randomly selected high-power (400×) fields per animal.

Immunohistochemical analysis.

Three-micrometer paraffin sections were processed as previously described (18) for staining with rat mAb against a cytoplasmic antigen of mouse monocyte/macrophages (F4/80; 2.5 μg/ml) (Caltag). Primary antibody was incubated overnight at 4°C, followed by the addition of biotinylated goat anti-rat IgG (Vector, Burlingame, CA) and avidin–biotin–peroxidase complex solution, and development with diaminobenzidine. Negative controls were obtained by omitting the primary antibody. F4/80-labeled cells were counted in 20 randomly selected high-power fields (400×) per animal.

Statistical analysis.

Data are expressed as the mean ± SEM. Statistical analysis was performed by repeated-measures or factorial one-way analysis of variance for analysis of the in vitro and ex vivo effects of seliciclib, respectively. The cumulative survival curves were compared by log rank test. Data on the frequency of proteinuria and BUN levels were analyzed by Fisher's exact test. All other parameters were analyzed by the Kruskal-Wallis or Mann-Whitney U test, as appropriate. P values less than 0.05 were considered significant.

RESULTS

Effect of seliciclib on polyclonal mitogen-induced activation of T cells and B cells in vitro.

We assessed the effect of seliciclib on T cell and B cell activity in vitro. Adding seliciclib (1–30 μM) to splenocytes isolated from normal CD-1 mice dose-dependently inhibited T cell proliferation induced by polyclonal stimulation with Con A (Figure 1A) or anti CD3-anti CD28 mAb, a condition that mimics antigen-presenting cell stimulation (data not shown). The production of IFNγ and IL-10 was decreased by seliciclib in Con A–stimulated spleen cells (Figures 1B and C). Seliciclib abrogated T cell–independent proliferation of splenic B cells in response to LPS (Figure 1D) and reduced the release of IgG2a in the cell supernatant (Figure 1E).

Figure 1.

Seliciclib-induced inhibition of T cell and B cell activity in vitro. A, Splenocytes from control CD-1 mice (n = 5) were incubated under resting conditions or with concanavalin A (Con A; 10 μg/ml), a selective T cell mitogen. Seliciclib was added at the indicated concentrations in the presence of Con A. The proliferative response was evaluated at the end of 72 hours of incubation. Release of B, interferon-γ (IFNγ) and C, interleukin-10 (IL-10) in cell supernatant was evaluated by enzyme-linked immunosorbent assay after 24-hour and 72-hour incubations of splenocytes, respectively. D and E, Splenocytes were incubated under resting conditions or in the presence of 10 μg/ml of lipopolysaccharide (LPS), a T cell–independent polyclonal B cell activator. Seliciclib was added as described above. Proliferation (D) and release of IgG2a in the supernatant (E) were assessed after 72 hours of incubation. Seliciclib did not affect cell viability (70–80%, by trypan blue dye exclusion in each experimental condition). Results are the mean and SEM. ∗ = P < 0.01 versus Con A and Con A plus seliciclib 1 μM; ○ = P < 0.01 versus Con A and Con A plus seliciclib 1 μM and 10 μM; # = P < 0.05 versus Con A plus seliciclib 10 μM; $ = P < 0.01 versus LPS and LPS plus seliciclib 1 μM; Δ = P < 0.05 versus LPS plus seliciclib 10 μM; • = P < 0.05 versus LPS.

Effect of seliciclib on survival in NZB × NZW mice.

The long-term effect of seliciclib treatment was evaluated in female (NZB × NZW)F1 mice. In these mice, a lupus-like disease characterized by production of IgG anti-dsDNA antibodies begins to develop at 3–4 months of age, which results in rapidly progressive glomerulonephritis, causing death of the animal within 12 months (14, 15). No weight loss, signs of liver dysfunction (serum AST and ALT levels were within normal range), or other apparent adverse effects were observed during seliciclib treatment of lupus mice.

The survival rate of NZB × NZW mice was remarkably prolonged by seliciclib. At 8 months, in experiment 1 (prevention study), most seliciclib-treated mice remained alive (71% and 77% of those treated with 100 mg/kg and 200 mg/kg, respectively), as compared with 31% of vehicle-mice (Figure 2, top). At this time point in experiment 2, 67% of mice treated with seliciclib were alive, versus 30% of mice given vehicle (Figure 2, bottom). Combining seliciclib with MP was more effective, resulting in 94% survival. At 12 months, when all vehicle-treated mice had died, 62% of animals receiving combined therapy were still alive (P < 0.0001).

Figure 2.

Effect of seliciclib on lifespan of NZB × NZW mice. Top, Survival rate in NZB × NZW mice administered seliciclib (100 mg/kg or 200 mg/kg; n = 13 mice for each dose) or vehicle (n = 13) daily by gavage, from 2 months of age to 8 months of age (experiment [exp.] 1). Survival curves were analyzed by log rank test. Bottom, Survival rate in NZB × NZW mice treated at a phase of established disease (experiment 2). Mice received daily seliciclib (200 mg/kg by gavage; n = 15), seliciclib plus low-dose methylprednisolone (MPS; 1.5 mg/kg intraperitoneally [n = 16]), methylprednisolone alone (n = 12), or vehicle (n = 10), beginning at 5 months of age. The study terminated at 12 months, when all vehicle-treated mice had died. NZB × NZW mice receiving combined therapy survived significantly longer than vehicle-treated mice. Survival rates were analyzed as in experiment 1.

Effect of seliciclib on proteinuria and renal function and structure.

Proteinuria, a direct sign of renal dysfunction, was delayed by preventive treatment with seliciclib, in a dose-dependent manner. At 8 months, the cumulative percentages of mice with proteinuria (>4 mg/day) were 43% in the group receiving 100 mg/kg of seliciclib and 23% in the group receiving 200 mg/kg of seliciclib, compared with 85% in the vehicle-treated group (P < 0.05 and P < 0.01, respectively) (Figure 3A). Late administration of seliciclib in experiment 2 reduced the percentage of mice with proteinuria only modestly when the compound was administered alone, whereas it significantly (P < 0.05) reduced the proportion of mice with proteinuria when combined with MP (Table 1). At 8 months, impairment of renal function, as evaluated by serum BUN measurement, was less severe in NZB × NZW mice administered seliciclib (as either a preventive or therapeutic regimen) compared with that in vehicle-treated mice (percentage of mice with abnormal BUN levels, e.g., >30 mg/dl: experiment 1, 30–33% versus 50%; experiment 2, 43% versus 80%). Combining the CDK inhibitor with MP resulted in only 27% mice with abnormal BUN values (P < 0.01 versus vehicle).

Figure 3.

Effect of seliciclib on onset of proteinuria. A, Seliciclib-induced reduction in the cumulative percentage of NZB × NZW mice with proteinuria (>4 mg/day) at 8 months (experiment 1). Renal morphologic parameters were examined in biopsy specimens obtained from 8-month-old NZB × NZW mice. B–D, Seliciclib-induced reduction of glomerular intracapillary and extracapillary hypercellularity and immune deposits, as quantitated by the score on a scale of 0 to 3+ (see Materials and Methods). E, IgG and C3 deposits detected by immunofluorescence in glomeruli of NZB × NZW mice administered seliciclib 200 mg/kg or vehicle. F, Representative images of periodic acid–Schiff–stained kidney specimens from 8-month-old NZB × NZW mice administered vehicle or seliciclib 200 mg/kg (original magnification × 40). G, Tubular and interstitial changes. H, Interstitial accumulation of F4/80+ monocyte/macrophages and CD4+ T cells, examined by immunostaining. Values are the mean and SEM. hpf = high-power field.

Table 1. Proteinuria, glomerular and tubular damage, and interstitial accumulation of monocyte/macrophages and CD4+ T cells in NZB × NZW mice treated with seliciclib or low-dose MP, alone or in combination*
TreatmentMice with proteinuria >4 mg/day, cumulative %Intracapillary hypercellularity scoreExtracapillary hypercellularity scoreGlomerular immune deposits scoreTubular/ interstitial damage scoreF4/80+ monocyte/ macrophages, cells/hpfCD4+ T cells, cells/hpf
8 mos.12 mos.
  • *

    Except where indicated otherwise, values are the mean ± SEM. Morphologic analysis was performed using renal biopsy specimens obtained from NZB × NZW mice that were terminally ill or were killed at age 12 months. Intracapillary hypercellularity was scored as follows: 0 = absent, 1+ = mild, 2+ = moderate, and 3+ = severe. Extracapillary hypercellularity was scored as follows: 0 = absent, 1+ = <25%, 2+ = 25–50%, and 3+ = >50% of glomeruli involved. Deposits were scored as follows: 0 = no deposits, 1+ = <25%, 2+ = 25–50%, and 3+ = >50% of glomeruli involved. Tubular changes and interstitial inflammation were scored as follows: 0 = no changes, 1+ = <25%, 2+ = 25–50%, and 3+ = >50% of the sample. MP = methylprednisolone; hpf = high-power field.

  • P < 0.05 versus vehicle.

  • P < 0.05 versus vehicle and versus MP.

  • §

    P < 0.01 versus vehicle.

  • P < 0.01 versus vehicle; P < 0.05 versus MP.

Vehicle90901.60 ± 0.241.00 ± 0.002.20 ± 0.371.70 ± 0.0961.5 ± 4.620.6 ± 1.2
Seliciclib, 200 mg/kg66731.14 ± 0.260.86 ± 0.141.43 ± 0.301.29 ± 0.2341.4 ± 8.610.8 ± 2.7
MP, 1.5 mg/kg75751.40 ± 0.400.80 ± 0.201.80 ± 0.201.20 ± 0.1457.2 ± 10.516.8 ± 5.1
Seliciclib plus MP37440.70 ± 0.300.30 ± 0.150.90 ± 0.230.45 ± 0.21§21.8 ± 4.86.2 ± 1.5§

Histologic analysis of vehicle-treated mice, performed at 8 months (experiment 1), showed glomerular changes consisting of intracapillary hypercellularity associated with focal extracapillary proliferation (Figures 3B, C, and F). Immune-type deposits were observed in the mesangium and in the glomerular capillary walls, with subendothelial distribution (Figure 3D). Tubular damage and interstitial inflammation were detected (Figures 3G and H). In contrast, mice administered seliciclib had milder glomerular and tubular damage and reduced interstitial accumulation of F4/80-positive monocyte/macrophages and CD4+ T cells as compared with vehicle-treated mice (Figure 3H). After receiving seliciclib, NZB × NZW mice had fewer glomerular deposits of IgG and complement C3 in the mesangium and in the glomerular capillary wall (Figure 3E). In experiment 2, because we performed survival studies, morphologic analysis was made on renal biopsy specimens obtained from terminally ill mice or mice that were killed at age 12 months. Seliciclib or MP administered alone mildly affected renal morphology, whereas combined treatment limited glomerular hypercellularity, immune deposits, and tubulointerstitial inflammation (Table 1).

Effect of seliciclib on signs of autoimmune disease in NZB × NZW mice.

A striking finding in our study was the effect of seliciclib on circulating anti-DNA antibodies. In NZB × NZW mice, levels of anti-dsDNA antibodies averaged 11 ± 1.3 units/ml at 2 months of age (baseline) and increased significantly (P < 0.01) over time, averaging 181 ± 74 units/ml at 5 months of age. Levels reached 367 ± 69 and 354 ± 40 units/ml at 8 months in mice given vehicle, in experiments 1 and 2, respectively. Anti-dsDNA antibody levels were significantly (P < 0.05) reduced in NZB × NZW mice treated with seliciclib compared with vehicle, in both the prevention study (for the 100-mg/kg dose, 73 ± 21; for the 200-mg/kg dose, 77 ± 21 units/ml) and the therapy study (for the 200-mg/kg dose, 148 ± 27 units/ml), which reflects the immunoregulatory potential of seliciclib in NZB × NZW mice. Similarly, seliciclib affected serum IgG levels, which were significantly (P < 0.05) lower than those in the vehicle-treated group, in both the prevention study (for 100 mg/kg, 244 ± 29 μg/ml; for 200 mg/kg, 235 ± 38 μg/ml; for vehicle, 549 ± 109 μg/ml) and the therapy study (for 200 mg/kg, 263 ± 20 μg/ml; for vehicle, 488 ± 74 μg/ml).

Splenomegaly is a hallmark of disease in NZB × NZW mice. In both experiments, seliciclib at a dose of 200 mg/kg substantially decreased the splenocyte number as compared with vehicle, although not to a significant extent (Table 2). In contrast, in the prevention study, the highest dose of seliciclib significantly (P < 0.05) reduced the total number of B cells and activated B cells (CD86+) and the total number of CD4+ T cells and memory CD4+ T cells (CD4+,CD44high) as compared with vehicle (Table 2). In the therapy study, a significant decrease in the absolute numbers of total and CD86+ B cells and memory CD4+,CD44high T cells was observed in animals treated with seliciclib plus MP compared with vehicle-treated mice (Table 2).

Table 2. Absolute number of spleen cells (×106) in NZB × NZW mice*
 Total splenocytesB cellsCD86+ B cellsCD4+ T cellsCD4+,CD44high T cells
  • *

    Values are the mean ± SEM. Lymphocyte subsets were analyzed by flow cytometry. MP = methylprednisolone.

  • P < 0.05 versus vehicle.

  • P < 0.05 versus vehicle and versus seliciclib 100 mg/kg.

  • §

    P < 0.05 versus MP.

  • P < 0.05 versus vehicle and versus MP.

Prevention study
 Vehicle (n = 4)135.48 ± 56.5047.13 ± 14.6336.57 ± 10.3332.83 ± 10.1423.66 ± 7.58
 Seliciclib 100 mg/kg (n = 9)132.28 ± 31.0934.68 ± 3.9322.57 ± 2.0032.37 ± 3.3619.39 ± 2.95
 Seliciclib 200 mg/kg (n = 10)119.05 ± 19.7127.60 ± 2.8221.27 ± 2.3218.38 ± 2.6011.53 ± 1.14
Therapy study
 Vehicle (n = 3)135.83 ± 9.8059.31 ± 7.6455.38 ± 8.1137.24 ± 8.0931.56 ± 2.15
 Seliciclib (n = 3)92.00 ± 4.0045.52 ± 1.2441.56 ± 0.6421.68 ± 2.3217.21 ± 0.76
 MP (n = 3)166.33 ± 60.5879.34 ± 29.1370.44 ± 26.8938.09 ± 13.1930.78 ± 10.71
 Seliciclib plus MP (n = 9)120.00 ± 20.0833.18 ± 2.63§25.77 ± 1.5824.36 ± 3.1418.45 ± 1.51

Effect of seliciclib on T cell and B cell activation in NZB × NZW mice.

T cell and B cell activity was evaluated ex vivo at the end of the prevention study. Splenocytes from 8-month-old NZB × NZW mice were unresponsive to polyclonal stimuli in vitro (data not shown), consistent with evidence that T cells from aged lupus mice were exhausted and underwent a shift to an activated/memory subset in vivo (19). T cell and B cell activity was therefore assessed in an additional group of NZB × NZW mice administered seliciclib (200 mg/kg) or vehicle, beginning at 2 months of age. Mice were killed at 5 months of age and studied in parallel with age-matched control CD-1 mice. The proliferative response to Con A (Figure 4A), anti-CD3/anti-CD28 mAb (Figure 4D), as well as LPS (Figure 4E) was significantly higher in splenocytes from lupus mice receiving vehicle compared with mice receiving seliciclib. Seliciclib normalized the splenocyte proliferative response to polyclonal stimuli (Figures 4A, D, and E).

Figure 4.

Effect of seliciclib on T cell and B cell activation in NZB × NZW mice. Splenocytes from 5-month-old NZB × NZW mice treated with vehicle (n = 5) or seliciclib beginning at age 2 months (n = 5) were isolated and tested ex vivo. A, Splenocytes were stimulated with 10 μg/ml concanavalin A (Con A), and proliferation was assessed by 3H-thymidine incorporation after 72 hours of incubation. Further aliquots of splenocytes were incubated with Con A for evaluation of B, interferon-γ (IFNγ) and C, interleukin-10 release after 24 hours and 72 hours of incubation, respectively. Spontaneous release of IL-10 from splenocytes under resting conditions was also assessed. D, T cell proliferation in response to anti-CD3/anti-CD28. E and F, Splenocytes isolated from each animal were activated with 10 μg/ml lipopolysaccharide (LPS). The proliferative response (E) and the release of IgG2a in the supernatant (F) were evaluated after 72 hours. The spontaneous release of IgG2a from splenocytes under resting conditions was also assessed. Shaded horizontal bars represent control ranges. Values are the mean and SEM. # = P < 0.05 versus control CD-1 mice; ∗ = P < 0.01 versus NZB × NZW plus vehicle; § = P < 0.01 versus control CD-1 mice; ○ = P < 0.05 versus NZB × NZW plus vehicle.

Levels of IFNγ in supernatant of splenocytes isolated from seliciclib-treated mice and stimulated in vitro by Con A were approximately half the levels in vehicle-treated mice (Figure 4B). We also observed that the production of IL-10 was partially reduced in splenocytes from lupus mice receiving seliciclib compared with vehicle-treated mice, under both resting and Con A–stimulated conditions (Figure 4C). B cell activity, as evaluated by IgG2a release, was suppressed by seliciclib (Figure 4F).

DISCUSSION

Seliciclib (R-roscovitine; CYC202) is a substituted purine analog that inhibits formation of CDK–cyclin complexes by interfering with the ATP-binding pocket located in the cleft between the small and large lobes of the kinase (20), and as such is a member of a large class of small-molecule inhibitors of CDKs. Although the CDK inhibitors were originally developed as antiproliferative agents for cancer therapy, it has been recognized that they may play a role in proliferative nonmalignant disease (21, 22). In SLE, the production of autoantibodies is secondary to the activation of T cells. We began our study by determining whether seliciclib affected the activity of normal mouse T cells in vitro. Seliciclib caused a concentration-dependent inhibition of proliferation induced by polyclonal stimulation with Con A or anti-CD3/anti-CD28 mAb in T cells. In addition, the drug interfered with the production of Th1 and Th2 cytokines, as indicated by decreased release of IFNγ and IL-10 in Con A–stimulated spleen cells. Seliciclib also exerted inhibitory effects on B cell proliferation in response to LPS and reduced the release of IgG2a in the cell supernatant.

The immunoregulatory potential exhibited by seliciclib in T cells and B cells in vitro provided the rationale for testing the effects of the drug in the NZB × NZW murine model of spontaneous SLE. Our data demonstrated that seliciclib extends the lifespan of female (NZB × NZW)F1 mice. Specifically, administration of seliciclib as part of a prevention regimen beginning at age 2 months significantly improved animal survival compared with vehicle-treated mice, as evaluated at 8 months of age. At 8 months, the CDK inhibitor also improved survival of NZB × NZW mice that began receiving seliciclib at 5 months of age, when disease is established. Later, a significant effect on survival persisted when seliciclib was administered in combination with low-dose MP.

Seliciclib attenuated the manifestations of renal disease in NZB × NZW mice to the extent that, in the prevention study, the drug significantly delayed the development of proteinuria and renal function impairment and limited glomerular hypercellularity, immune deposits, tubular damage, and interstitial accumulation of monocyte/macrophages. Late administration of seliciclib alone affected renal function and structure parameters to a partial degree, whereas a striking effect was achieved by combining seliciclib with low-dose MP. The latter finding may be of particular relevance to the clinical situation, because it suggests that the CDK inhibitor could help “sparing” steroids, which cause major problems, even today, in patients with SLE. Our data are consistent with those from a recent study showing that a cell cycle inhibitor peptide, the p21Waf1/Cip1 mimic, which prevents the formation of cyclin-D–CDK-4 complexes, inhibited the proliferation of autoreactive T cells and B cells, and interrupted the progression of disease in NZB × NZW mice (8).

The renoprotective effects of CDK antagonists have been previously reported in rats with experimental mesangial proliferative glomerulonephritis. Early treatment with roscovitine (a less potent form of seliciclib) decreased glomerular hypercellularity and matrix production, resulting in improved renal function (21). Similarly, in a mouse model of crescentic glomerulonephritis, treatment with roscovitine limited podocyte proliferation and improved renal function (22). Other studies showed that targeting aberrant cell-cycle progression in mice with collapsing glomerulopathy by treatment with seliciclib resulted in attenuation or even reversal of renal disease (23).

A major finding of the present study is the striking amelioration of the immune component of lupus disease afforded by seliciclib. Anti-DNA antibodies are considered instrumental in the pathogenesis of the immune glomerulonephritis in SLE, in both mice and humans (24, 25), and hold promise as a key link between the onset of disease and tissue damage. Our data demonstrate that NZB × NZW mice treated with seliciclib, as both preventive and therapeutic regimens, had lower levels of anti-DNA antibodies and total IgG in their sera and less IgG deposition within glomeruli. Mouse SLE involves abnormal activation of CD4+ T cells that accumulate as activated memory cells and help trigger polyclonal B cell activation and expansion (26). As a result of its immunoregulatory property, seliciclib significantly reduced activated/memory T cell and B cell accumulation in NZB × NZW mice. Furthermore, as observed in vitro, seliciclib administered in an early phase of the disease resulted ex vivo in the inhibition of both T cell and B cell proliferative responses to polyclonal stimuli. Blocking B cell proliferation by seliciclib also inhibited the B cell differentiation process, as indicated by the decreased release of IgG2a.

Among cytokine abnormalities observed in lupus mice, one of the most consistent has been high expression of IFNγ, a Th1 cytokine that exacerbates disease by promoting the switch of IgM to the IgG2a subclass, the predominant nephritogenic autoantibodies in NZB × NZW mice (27). Here, we documented that the release of IFNγ by splenocytes isolated from NZB × NZW mice given seliciclib and stimulated in vitro by Con A was substantially decreased compared with that in splenocytes from vehicle-treated mice. Reduced production of IL-10, a Th2 cytokine that accelerates B cell proliferation and differentiation (28), was also observed. Taken together, these findings provide evidence for modulation by seliciclib of T cell and B cell effector functions critical to the pathogenesis of lupus autoimmune disease, leading to improvement in disease and prolonged survival.

In conclusion, the present data showing that seliciclib limited autoimmune renal damage and extended the lifespan of NZB × NZW mice indicate that targeting CDK activity is effective in halting lupus nephritis. A direct immunomodulatory action of seliciclib on T cells and B cells may be one of the mechanisms underlying the beneficial effects.

AUTHOR CONTRIBUTIONS

Dr. Zoja had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Zoja, Casiraghi, Remuzzi, Benigni.

Acquisition of data. Casiraghi, Conti, Corna, Rottoli, Cavinato.

Analysis and interpretation of data. Zoja, Casiraghi, Conti, Corna, Rottoli, Cavinato, Remuzzi, Benigni.

Manuscript preparation. Zoja, Casiraghi, Corna, Remuzzi, Benigni.

Statistical analysis. Casiraghi, Rottoli, Cavinato.

Acknowledgements

We thank Drs. Borislav Dimitrov and Annalisa Perna for performing statistical analysis and Drs. Marina Noris and Mauro Abbate for helpful discussion. Marta Todeschini helped to perform fluorescence-activated cell sorting analysis. We greatly acknowledge Drs. Anna Barnett, Sheelag Frame, and Athos Gianella-Borradori for their suggestions and continuous support.

Ancillary