Interplay of IKK/NF-kappaB signaling in macrophages and myofibers promotes muscle degeneration in Duchenne muscular dystrophy

Abstract

Duchenne muscular dystrophy (DMD) is a lethal X-linked disorder associated with dystrophin deficiency that results in chronic inflammation and severe skeletal muscle degeneration. In DMD mouse models and patients, we find that IkappaB kinase/NF-kappaB (IKK/NF-kappaB) signaling is persistently elevated in immune cells and regenerative muscle fibers. Ablation of 1 allele of the p65 subunit of NF-kappaB was sufficient to improve pathology in mdx mice, a model of DMD. In addition, conditional deletion of IKKbeta in mdx mice elucidated that NF-kappaB functions in activated macrophages to promote inflammation and muscle necrosis and in skeletal muscle fibers to limit regeneration through the inhibition of muscle progenitor cells. Furthermore, specific pharmacological inhibition of IKK resulted in improved pathology and muscle function in mdx mice. Collectively, these results underscore the critical role of NF-kappaB in the progression of muscular dystrophy and suggest the IKK/NF-kappaB signaling pathway as a potential therapeutic target for DMD.

Figure 1. NF-κB activity in dystrophic muscle is localized to both muscle and immune cells.

(A–C) Muscle nuclear extracts from 5-week-old WT C57BL/10 and 3- and 5-week-old mdx mice (A), 5-week-old WT and mdx mice (B), or 7-week-old WT, mdx, and DKO mice (C) were used in EMSA for NF-κB and Oct-1. Supershift assays were performed on mdx muscle extracts using antibodies against p65 and p50. Arrowheads denote shifted subunits. (D) IKK assays performed with IκBα WT and mutant (double serine to threonine [SS/TT]) or p65 WT and mutant (serine to alanine [S/A]) substrates using gastrocnemius muscle lysates from 7-week-old WT or mdx mice. Immunoprecipitates were probed for IKKγ as a loading control. Western blots are shown for p-IKK, p-IκBα, IκBα, p-p65, and p65. GST, glutathione-S-transferase. (E) Gastrocnemius muscles from 7-week-old WT or mdx mice were immunostained for p-p65. Scale bars: 50 μm. Black arrowheads denote immune cells, and blue arrowheads indicate regenerating fibers. (F) Muscles from either 4- or 7-week-old mdx mice were double stained with p-p65 (green) and CD68 (red) or p-p65 and E-MyHC (red), respectively. Scale bars: 20 μm. (G) H&E staining and p-p65 immunohistochemistry were performed on muscle biopsies from healthy controls and age-matched DMD patients (n = 4). Scale bar: 50 μm.

Figure 2. NF-κB activity is deregulated during postnatal development of mdx mice.

(A) Muscle extracts from 5-week-old WT;3xκB-Luc-Tg or mdx;3xκB-Luc-Tg mice were prepared for luciferase assays. Luciferase values were normalized to total protein. *P < 0.05; n = 5. (B) Gastrocnemius muscles were immunostained for luciferase expression in 7-week-old WT;3xκB-Luc-Tg and mdx;3xκB-Luc-Tg mice (n = 3). Scale bar: 50 μm. (C) EMSAs for NF-κB and Oct-1 performed in gastrocnemius muscles from WT and mdx mice during postnatal development. (D) Primary myoblasts isolated from 6-day-old WT;3xκ-Luc-Tg and mdx;3xκB-Luc-Tg mice were differentiated and subsequently switched to either medium alone or medium containing TNF-α (5 ng/ml) for 6 hours.

Figure 3. Heterozygous deletion of p65 rescues mdx pathology.

(A) Histopathology of muscle cryosections stained with H&E, comparing 5-week-old mdx;p65+/⁺ and mdx;p65^+/– mice or mdx;p50^+/+ and mdx;p50^+/– mice. Scale bars: 20 μm. (B) Gastrocnemius sections from WT C57BL/10, mdx;p65^+/+, and mdx;p65^+/– mice were immunostained with F4/80. Scale bar: 50 μm. (C) RNA was isolated from the remaining part of the muscle, and real-time PCR was performed for lysozyme and CD68, comparing 5-week-old WT C57BL/10, mdx;p65^+/+, and mdx;p65^+/– mice (n = 3). (D) Necrotic fibers from gastrocnemius muscles were quantitated based on IgG staining (filled fibers appear in red, nuclei in blue) (n = 5). Scale bar: 100 μm. (E) Calcifications were quantitated from mdx;p65^+/+ and mdx;p65^+/– muscles (n = 10). Data are plotted as mean ± SEM from 2 independent experiments. *P < 0.05.

Figure 4. Heterozygous deletion of p65 promotes regenerative myogenesis in mdx mice.

(A) Gastrocnemius cryosections from 7-week-old mdx;p65^+/+ and mdx;p65^+/– male mice were stained with H&E (top row) or immunostained for E-MyHC (bottom row). Scale bars: 15 μm (top) and 50 μm (bottom). (B and C) Quantitation of E-MyHC–positive fibers and CLN. (D and E) TA sections from 5-week-old cardiotoxin-treated p65^+/+ and p65^+/– mice were stained with H&E, and numbers of E-MyHC fibers were calculated. Scale bar: 20 μm. Data are plotted as mean ± SEM from 2 independent experiments (n = 4). *P < 0.05.

Figure 5. IKKβ deletion in myeloid cells reduces inflammation.

(A and B) Gastrocnemius sections from 4-week-old mdx;IKKβF/F and mdx;IKKβF/F;Lys-Cre male mice were analyzed either by double immunofluorescence staining against IKKβ (green) and CD68 (red) (A) or by immunohistochemistry for p-p65 (B). Dashed white lines in A mark field of macrophages. m, muscle fibers. Scale bars: 10 μm. (B) p-p65 immunohistochemical staining of immune infiltrates (top row) and regenerating muscle fibers (bottom row). Scale bars: 15 μm (top) and 10 μm (bottom). (C) Necrotic fibers were identified by IgG staining (filled fibers in red, nuclei in blue) (n = 6). Scale bar: 50 μm. Quantification of necrosis appears in the graph. (D) Real-time PCR analysis of CD68 in TA from sex- and age-matched WT, mdx;IKKβF/F and mdx;IKKβF/F;Lys-Cre mice. (E) Real-time PCR analysis was performed as described in D for TNF-α, IL-1β, and MCP-1. Graphs are plotted as mean ± SEM. *P < 0.05.

Figure 6. IKKβ deletion in muscle cells promotes regeneration.

Muscles harvested from 4-week- (A–F) or 12-week-old (F only) mdx;IKKβF/F and mdx;IKKβF/F;MLC-Cre mice were used for protein analysis (A and C) or histology (B and D). (A) Western blots probing for IKK subunits. (B) Muscles were immunostained for p-p65 expression. (C) Western blots probing for p-p65, p65, and IκBα. (D) H&E staining of mdx;IKKβF/F and mdx;IKKβF/F;MLC-Cre muscles. (E) Muscles used in D were stained for quantitation of E-MyHC–positive stained fibers or stained with H&E for counting of CLN. *P < 0.05. Scale bars: 15 μm (B) and 50 μm (D). Graphs are plotted as mean ± SEM. (F) Mean fiber distribution in mdx;IKKβF/F and mdx;IKKβF/F;MLC-Cre mice was determined from a minimum of 3,000 fibers from randomly chosen fields, obtained from multiple muscle sections from a minimum of 4 mice per group.

Figure 7. Muscle progenitor cells increase in mice lacking IKKβ.

Gastrocnemius or quadriceps were harvested from 4-week-old mdx;IKKβF/F and mdx;IKKβF/F;MLC-Cre mice and frozen for RNA analysis (A), cryosectioned (B), or homogenized for immunoblotting (C). (A) Real-time PCR analysis of CD68, TNF-α, IL-1β, MCP-1, RANTES, and MIP-1α (macrophage inflammatory protein 1α) obtained from 3–6 mice per group. (B) Sections as described in A or from 7-week-old mdx;p65^+/+ and mdx;p65^+/– mice were immunostained with MyoD and quantitated from a minimum of 20 randomly chosen fields per section in 3–4 animals per group. (C) Western blots probing for Pax7 and α-tubulin. (D and E) Satellite cells were isolated from pooled hind limb muscles from 4-week-old mdx;IKKβF/F and mdx;IKKβF/F;MLC-Cre littermates and either fixed and stained with Pax7 or differentiated for 3 days and subsequently stained with MyHC to determine myotube number normalized per mm² area. Scale bars: 200 μm. (F) Flow-cytometric analysis from freshly isolated cells of 4-week-old mdx;IKKβF/F and mdx;IKKβF/F;MLC-Cre mice were stained with cell-surface markers CD34 and Sca-1. Graph represents averages from 2 independent experiments. Data are plotted as mean ± SEM.

Figure 8. Pharmacological inhibition of IKK rescues the histopathology and function of dystrophic muscle.

(A) Amino acid sequence of WT or mutant (mut) NBD peptides. Underlined amino acids indicate changes from WT to mutant forms. (B) Soleus muscles from WT or mutant NBD–treated mice were stained with F4/80, and macrophages were quantitated. Scale bar: 300 μm. (C) Gastrocnemius muscles harvested from mdx mice treated for 4 weeks were sectioned and stained with either p-p65 (C) or H&E (E). Scale bars: 20 μm (C and E). (D) Lysates from mice were used for Western blots probing for p-p65, p65, and IκBα. (F) RNA was isolated from similar muscles as used for D including C57BL/10 and mdx controls, and real-time PCR was performed for lysozyme (n = 4). (G) Regeneration potential was measured by quantitating fibers with centronucleation and positive E-MyHC staining from mdx mice treated with saline or NBD peptides. (H) Force generation assessed by measuring active developed force comparing diaphragm muscles from mice treated with either WT or mutant NBD peptide for 4 weeks. Quantitative data are plotted as mean ± SEM from 3 independent experiments. *P < 0.05.