VBP15, a novel anti-inflammatory and membrane-stabilizer, improves muscular dystrophy without side effects

Abstract

Absence of dystrophin makes skeletal muscle more susceptible to injury, resulting in breaches of the plasma membrane and chronic inflammation in Duchenne muscular dystrophy (DMD). Current management by glucocorticoids has unclear molecular benefits and harsh side effects. It is uncertain whether therapies that avoid hormonal stunting of growth and development, and/or immunosuppression, would be more or less beneficial. Here, we discover an oral drug with mechanisms that provide efficacy through anti-inflammatory signaling and membrane-stabilizing pathways, independent of hormonal or immunosuppressive effects. We find VBP15 protects and promotes efficient repair of skeletal muscle cells upon laser injury, in opposition to prednisolone. Potent inhibition of NF-κB is mediated through protein interactions of the glucocorticoid receptor, however VBP15 shows significantly reduced hormonal receptor transcriptional activity. The translation of these drug mechanisms into DMD model mice improves muscle strength, live-imaging and pathology through both preventive and post-onset intervention regimens. These data demonstrate successful improvement of dystrophy independent of hormonal, growth, or immunosuppressive effects, indicating VBP15 merits clinical investigation for DMD and would benefit other chronic inflammatory diseases.

Keywords: anti-inflammatory; dystrophy; mdx; membrane injury; muscle.

© 2013 The Authors. Published by John Wiley and Sons, Ltd on behalf of EMBO.

Figures

Figure 1. VBP15 structure differentiates it from glucocorticoids and improves GR specificity

1. A–D. The chemical structure of VBP15 is provided (A). VBP15 was selected for clinical development by having the optimal profile from a library of Δ-9,11 compounds, whose general structure is provided in (B). Compound diversity for this library was generated by medicinal chemistry probing of the R1–R3 groups. These compounds are structurally related to glucocorticoids (C), but contain an essential Δ-9,11 double bond modification to the steroid C-ring, at the location depicted by the red box. This modification produces novel properties and clear differences in sub-activity profiles of these compounds. Prednisolone (D) is the active form of prednisone, a current glucocorticoid standard of care for DMD.

2. E–H. Receptor specificity was determined through competitive binding assays. Here, radiolabeled high-affinity ligands were incubated with extracted steroid receptors and increasing concentrations of unlabeled competitor (high-affinity control, VBP15, VBP1 or VBP3). Best-fit curves are provided. VBP15 showed increased GR specificity through increased binding to the (E) GR and decreased binding to the (F) MR in comparison to other Δ-9,11 compounds (VBP1 and VBP3). VBP15 also showed low (H) androgen receptor binding and no detectable binding to the (G) oestrogen receptor. (Triam, triamcinolone; Spiro, spironolactone; R1881, methyltrienolone; Estradiol, 17β-estradiol).

Figure 2. VBP15 inhibits inflammatory signaling and promotes membrane stability

1. A. In an NF-κB reporter assay, increasing concentrations of prednisolone or VBP15 were applied to TNFα-induced C2C12 myoblasts stably expressing a luciferase reporter under the control of an NF-κB driven promoter. Significant reporter inhibition was observed at all concentrations over 10 nM.

2. B. Inhibition was also observed for endogenous NF-κB activated inflammatory transcripts. Cox2, Irf1 and Nos2 expression was significantly reduced in TNFα-induced H2K myotubes, as determined by real time qPCR. Results are mean ± SEM from representative experiments performed in triplicate (ANOVA, *p < 0.05, **p < 0.005, ***p < 0.0005).

3. C,D. Laser injury and dye exclusion assays to determine effects of VBP15 and prednisolone on cell membrane integrity. (C) Images of C2C12 myoblasts exposed to the indicated drug 15 min prior to laser wounding, with fluorescent visualization of FM1-43 dye entry into cells (white arrowheads mark sites of injury, scale bars = 5 µm). VBP15 shows protection against laser-induced injury. (D) Quantitation of FM1-43 influx over time (laser injury at time 6 s). VBP15 shows reduced impact of initial injury and enhanced repair, whereas prednisolone shows greater impact from injury with elevated dye uptake. Representative data from one of four experiments are presented as mean ± SEM. (ANOVA, n ≥ 16 per treatment, *p < 0.05; Pred, prednisolone).

Figure 3. GR is required for anti-inflammatory activities of VBP15 and prednisolone, but VBP15 shows loss of GRE-mediated sub-activities associated with side effects

1. A. Application of a steroidal receptor antagonist (RU-486) ablated the NF-κB inhibitory activity of both prednisolone and VBP15, indicating that both drugs share steroidal anti-inflammatory pathways.

2. B,C. Assessment of GRE-mediated transcriptional (hormonal) activities differentiates prednisolone and VBP15. (B) Sgk1 gene expression is controlled by a positive GRE and showed reduced activation by VBP15 in comparison to glucocorticoids in AtT-20 pituitary cells, as measured by qRT-PCR. α-Tocopherol was included as a negative control compound that lacks any GRE activity. (C) ACTH expression is controlled by a negative GRE, and is considered a component of adrenal suppression (negative side effect of pharmacological glucocorticoids). VBP15 showed reduced effects on this side effect pathway via ELISA of treated AtT-20 cell media.

3. D. Western blot of GRnull fibroblasts and the control GR positive L929 line they were derived from, illustrating the absence of detectable GR protein in the GRnull cells. GAPDH was included as a loading control.

4. E,F.GRnull and GR positive fibroblasts were treated with drug, then induced with TNFα and transcript levels assayed by real time qPCR. (E) Sgk1 levels illustrate an absence of induction in GRnull cells, confirming Sgk1 dependence on the GR and the absence of GR function in this cell line. (F) Inhibition of the endogenous NF-κB activated inflammatory transcripts Irf1, Tnfα and Il1a was observed in GR positive cells but not in GRnull cells, indicating the GR is essential for this inhibition.

5. G,H. Spleens were harvested from GRdim/dim and wild type control mice. Splenocyte suspensions were treated with drug, then induced with TNFα and transcripts assayed by qPCR. (G) Nfkbia levels illustrate an absence of GRE induction by the mutant GR, as well as a lack of induction of NF-κB inhibiting gene products, in GRdim/dim splenocytes. (H) Inhibition of Irf1, Tnfα and Il6 inflammatory transcripts was observed in both wild type and GRdim/dim mutant splenocytes. This indicates GRdim isoforms, which maintain protein-protein interactions but lose receptor-DNA interactions, still maintain inhibition of endogenous NF-κB activated inflammatory transcripts. (Pred, prednisolone; Dex, dexamethasone; α-Toc, vitamin E; ANOVA, *p < 0.05, **p < 0.005, ***p < 0.0005).

Figure 4. VBP15 improves dystrophic phenotypes of mdx mice in two pre-clinical trials (pre-symptomatic and post-onset treatment regimens)

1. A–F. Prophylactic treatment of mdx mice beginning at 2 weeks of age showed dose-dependent improvement of clinical and histological endpoints. Mouse limb strength increased upon VBP15 treatment as measured by grip strength of 6 week old mice for both (A) forelimb and (B) hindlimb (n ≥ 12 mice/group). (C) Maximal force exerted by mouse forelimbs increased with VBP15 treatment but decreased with prednisolone treatment, due to prednisolone effects on mouse size (presented later). (D) Specific force of isolated EDL muscle increased with VBP15 treatment (n = 10 mice/group). (E) Live-animal imaging of cathepsin protease activity (ProSense680) shows reduced inflammation and necrosis of the hindlimbs in VBP15-treated mdx mice (E images; E′ quantitation of fluorescence; n ≥ 6 mice/group). (F) Histology of diaphragm muscle shows a decrease in inflammatory foci from VBP15 treatment at 15 and 30 mg/kg (F representative images, F′ quantitation; n = 6 mice/group).

2. G–I. A second pre-clinical trial was performed in exercised adult mdx mice to assay post-onset efficacy. (G) Live-animal imaging of inflammation (ProSense680) showed a significant decrease with VBP15 treatment (G representative images, (G′) quantitation; n ≥ 6 mice/group). (H) Specific force of isolated EDL muscle was measured ex vivo at trial conclusion with 15 mg/kg VBP15 showing an increase consistent with the neonate trial (n ≥ 7 mice/group). (I) Histology of adult diaphragm showed a significant reduction in inflammatory foci upon 45 mg/kg VBP15 treatment (n = 6 mice/group). Values are mean ± SEM. For treatments, the mean percentage of increase or decrease of mdx vehicle values towards WT is provided. (Pred, prednisolone; FL, forelimb; HL, hindlimb; data exceeding 2 SD's was removed from specific force values as an outlier but included in all statistical analyses; one-tailed t-test of single dose versus vehicle mdx†p < 0.05; ANOVA of dose-dependence groups versus vehicle mdx *p < 0.05, **p < 0.005, ***p < 0.0005).

Figure 5. VBP15 does not show immunosuppressive activities shown by prednisolone

1. Prednisolone significantly reduced the number of viable splenocytes per gram of spleen tissue, whereas VBP15 did not at any dose.

2. The percentage of B lymphocytes, as measured by FACS analysis of B220 positive cells, was reduced in spleens from prednisolone treated mdx mice, while VBP15 showed no decrease in B cells.

3. Spleen CD4+ T cell numbers were significantly decreased in prednisolone treated mdx spleens, but not by VBP15 treatment.

4. Activation of mdx splenocyte T cells by concavalin A was impaired by prednisolone, but not impaired by VBP15 treatment. Values are mean ± SEM. (Pred, prednisolone; (A) n ≥ 12, (B–D) n = 3–5; *p ≤ 0.05, **p < 0.005).

Figure 6. VBP15 lacks the side effects of current glucocorticoid regimens in vivo

1. A. Prednisolone treatment stunted the growth of developing mice in comparison to both vehicle and VBP15 groups. Representative photographs (A) and quantitation of body length (A′) are provided.

2. B. Bone lengths were reduced upon prednisolone treatment. X-rays (B) of mouse tibias illustrate size differences (scale bars = 2 mm). Quantitation shows a significant decrease in tibia length (B′).

3. C. MicroCT imaging analysis of femur revealed a significant decrease in trabecular thickness (C′) for prednisolone treated mice.

4. D–F. Increases in cardiac fibrosis and heart mass were detected in prednisolone treated mice, suggestive of cardiac damage as a side effect lacking for VBP15. Sirius red staining of cardiac muscle shows increased fibrosis in prednisolone-treated mice, but not VBP15 mice. Representative images (D) and digital quantitation of fibrosis (E) are provided. To the right of the image panel is a higher magnification image from the area outlined in box. (F) Heart mass ratios were increased by prednisolone but not by VBP15.

5. G,H. In adult mdx mice as well, increases in cardiac fibrosis (G) and heart mass (G) were observed with prednisolone treatment but not VBP15 treatment. In adult mdx vehicle mice, an expected disease- and age-related increase in fibrosis over WT is seen. Values are mean ± SEM. (n ≥ 12 per group for (A,B,E,F); n ≥ 5 for (C,G,H); *p < 0.05, **p < 0.005, ***p < 0.0005).

Figure 7. Working model of VBP15 and prednisone drug mechanism sub-activity profiles

Steroidal compounds such as glucocorticoids (prednisone) and Δ-9,11 compounds (VBP15) are multi-potent drugs. Through dissecting the sub-activities of these compounds, we find that: (1) VBP15 reduces inflammation but does not show the immunosuppressive impairment of lymphocyte viability and function observed for prednisone. (2) Within an environment of plasma membrane disruption, VBP15 helps to promote resistance to and repair of injuries, while prednisone can exacerbate membrane injury. (3) Inside cells, both compounds bind to and activate the GR to potently inhibit inflammatory NF-κB signaling through protein–protein interactions. (4) Though they both bind to the GR, prednisone causes strong induction of hormonal GRE controlled promoter elements, while VBP15 eliminates or greatly reduces these effects.