CTLA4-Ig reduces proliferation and inflammatory gene expression in muscle fibroblasts, corresponding to less fibrosis and inflammation in mdx muscular dystrophy

Abstract

Muscle pathology in Duchenne muscular dystrophy (DMD) is greatly amplified by the immune response to dystrophic muscle, which provides the rationale for targeting the immune system in DMD therapies. Much of immune-driven pathology in the mdx mouse model of DMD is caused by T-lymphocytes and macrophages; thus, preventing T-cell activation by blocking pathways that cause their activation has the potential to reduce muscle damage and fibrosis in muscular dystrophy. CTLA4-Ig is a recombinant protein that blocks costimulatory signaling between T-cells and antigen-presenting cells, such as macrophages. In this investigation, we tested whether treatment of mdx mice until they reach advanced, fibrotic stages of the disease reduces pathology. Our findings show that CTLA4-Ig treatments reduced muscle damage and inflammation and also reduced numbers of fibrogenic cells and fibrosis of muscles in aging, mdx mice. However, the treatments did not reduce numbers of CD8 + T-cells or activated CD25 + cells, suggesting that the reduced pathology was not mediated by affecting T-cell activation. Complete blood counts and clinical histopathological scoring also showed that the treatments had little effect on hematopoietic tissues. In vitro, CTLA4-Ig acted directly on muscle fibroblasts, reducing their expression of proinflammatory genes without affecting the expression of genes encoding connective tissue proteins, assayed by quantitative PCR (qPCR). However, CTLA4-Ig-treated fibroblasts were less proliferative in vitro. Collectively, these findings show that CTLA4-Ig acts directly on fibroblasts to produce changes in proliferation and gene expression that are consistent with the reductions in muscle pathology that occurs in aging, mdx mice treated with CTLA4-Ig.

Graphical Abstract

NEW & NOTEWORTHY

This investigation shows that CTLA4-Ig, which blocks costimulatory signaling between immune cells, is effective at reducing fibrosis and inflammation in aging, dystrophic muscle. CTLA4-Ig acts directly on muscle fibroblasts, causing reductions in their proliferation and expression of proinflammatory genes.

INTRODUCTION

Potential strategies for treating Duchenne muscular dystrophy (DMD) have advanced tremendously, adding to hopes that elevating expression of dystrophin or functional fragments of dystrophin will be clinically available one day (1). However, obstacles to obtaining sufficient systemic and persistent expression of therapeutic constructs remain; a reasonable expectation is that genetic interventions may reduce but not prevent DMD for patients with some more-common mutations (1-3). Expression of dystrophin or dystrophin fragments in DMD patients can also produce immune responses to cells that express the foreign protein, which may become a major hurdle to successful application of gene therapy approaches (4). Thus, alternative strategies that target downstream events in the pathology of DMD to reduce muscle damage offer advantages (5). First, they have the potential for rapid development if they are based on available pharmaceutical agents. In addition, therapeutic manipulations that do not rely on introduction of a foreign protein will not generate an acquired immune response to dystrophic muscle. Finally, gene-based strategies are likely to improve but not prevent muscular dystrophy and there may still be cycles of muscle injury and repair that can be ameliorated by managing downstream events.

Because the immune response to muscle pathology in DMD and in the mdx mouse model of DMD greatly amplifies muscular dystrophy, manipulation of the immune system provides a feasible approach to target downstream events (6). Discoveries over the past 25 yr have demonstrated a chronic, complex immune response to dystrophic muscle that includes a cellular immune response, in which CD8⁺ cytotoxic T-lymphocytes (CTLs) promote pathology (7, 8), and an innate response, in which muscle damage is attributable in part to invading myeloid cells, especially macrophages, neutrophils, and eosinophils (9-11). Each of those lymphoid and myeloid populations exacerbate muscle fiber damage in muscular dystrophy (7-11), indicating that each is a potentially useful, therapeutic target.

Therapeutic interventions that affect the activation of T-cells provide an attractive strategy for controlling the immune response to muscular dystrophy because activated T-cells influence both the cellular and innate immune responses to diseased muscle. Activation of T-cells that promote muscular dystrophy requires two signals (12). The first signal occurs when T-cell receptors on CD8 + T-cells bind antigens associated with major histocompatibility complex (MHC) class I antigens. Most cells can express MHC class I on their surface. CD4 + T-cells receive their first signal from macrophages and other professional antigen-presenting cells (APCs) that present antigen with MHC class II on their surfaces. This first stage of activation via either MHC class I or class II must be accompanied by a second signal, called a costimulatory signal, to fully activate T-cells. In the best-characterized, costimulatory pathway, APCs provide a second signal by expressing CD80 and CD86 on their surfaces, which bind CD28 on T-cell surfaces. T-cell activation following costimulation leads to their proliferation and production of immunomodulatory molecules, including proinflammatory cytokines (13, 14). However, T-cell activation via CD28 ligation also leads to elevated expression of cytotoxic T-lymphocyte-associated antigen 4 (CTLA4) (15). CTLA4 is structurally similar to CD28 and provides an immune checkpoint that downregulates the immune response (16). Although more than one mechanism contributes to CTLA4’s capacity to attenuate the immune response, its release from immune cells leads to its ligation of CD80/86, preventing their binding of CD28, thereby disabling the costimulatory pathway (17, 18). Thus, CTLA4 provides a negative feedback system for immune responses.

We have recently shown that a recombinant fusion protein composed of the extracellular domain of CTLA4 linked to a modified Fc portion of immunoglobin G (IgG) reduces pathology in the mdx mouse model of DMD following a 1 wk period of treatment (19). CTLA4-Ig, marketed as “abatacept,” is approved by the Federal Drug Administration (FDA) for treating rheumatoid arthritis (RA) and psoriatic arthritis, in which it reduces chronic inflammation that increases tissue damage (20). Importantly, abatacept is also FDA approved for treating juvenile idiopathic arthritis for patients as young as 2 yr of age, reflecting its safety and efficacy in pediatric populations (20). Our findings showed that much of the reduction of muscle pathology in young mdx mice after a short treatment period was attributable to its direct action on macrophages, leading to downregulation of their expression of proinflammatory genes and reduced inflammation and reduced T-cell activation (19). These immunosuppressive effects reduced muscle damage, which suggested that CTLA4-Ig is a promising immunotherapeutic for DMD. These beneficial findings in muscular dystrophy resemble treatment effects in RA. For example, preclinical studies demonstrated that CTLA4-Ig treatments in a collagen-induced, rodent model of arthritis were highly effective in reducing clinical and histological features of joint disease (21). Similarly, in humans with RA, CTLA4-Ig treatments reduced joint inflammation and damage (22). Importantly, beneficial effects were maintained over long periods of treatment. RA patients treated with CTLA4-Ig for 5 yr showed consistent safety and long-term sustained clinical efficacy over the course of the trial (22). Thus, the safety and efficacy of CTLA4-Ig for long-term, therapeutic immunomodulation in young RA patients support the possibility that it may be similarly useful for long-term, beneficial treatment of DMD.

Although our findings showed that brief, CTLA4-Ig treatments of young, mdx mice reduced muscle inflammation and damage (19), muscular dystrophy is a chronic, life-long disease, in which progressive, muscle fibrosis coincides with increased muscle pathology and weakness (23). We have shown previously that the innate immune response to dystrophic muscle drives muscle fibrosis (11, 24); thus, the immunosuppressive effects of CTLA4-Ig have the potential to reduce mdx muscle fibrosis. However, whether long-term CTLA4-Ig treatments can reduce muscle fibrosis is unknown and whether prolonged treatments are safe or instead produce unknown, detrimental, off-target effects in muscular dystrophy have not been examined.

In the present investigation, we test whether CTLA4-Ig is effective in treating mdx muscular dystrophy when administered for long periods that include the late, progressive stages of mdx dystrophy, when muscle fibrosis is an increasingly important pathological feature. We also assay by necropsy whether prolonged treatments produce off-target effects that may be detrimental, particularly in the immune system, which is the primary target of CTLA4-Ig. Addressing these questions will help establish whether CTLA4-Ig may be a promising candidate for the long-term treatment of DMD.

MATERIALS AND METHODS

Mice

C57BL/10ScSn-Dmd^mdx/J mice (RRID:IMSR_JAX:001801) were purchased from The Jackson Laboratory (Jax Labs, Bar Harbor, ME) and bred and housed at University of California, Los Angeles (UCLA) in a specific pathogen-free barrier facility with ventilated cages, a 12-h light/dark cycle and ad libitum access to a veterinary diet and water. Animal care and welfare were overseen by a licensed veterinary staff. All procedures related to animal handling, experimentation, and euthanasia were performed by trained staff according to protocols approved by the UCLA Institutional Animal Care and Use Committee (Animal Welfare Assurance number, D16-00124 A3196-01). Mice were euthanized by isoflurane inhalation and weighed immediately before tissue collection. Muscles were weighed and samples for RNA isolation were frozen and stored in liquid nitrogen. Muscles for histological analyses were embedded in Tissue-Tek O.C.T compound (Sakura Finetek, Torrance, CA) and rapidly frozen in liquid nitrogen-cooled isopentane and stored at −80°C. Only male mice were used for this study.

In Vivo CTLA4-Ig or Control IgG Treatment

Male, mdx mice received intraperitoneal injections of 25 μg/g body mass recombinant CTLA-4-Ig (BioXCell, Lebanon, NH, Cat. No. BE0099) or control, human IgG (BioXCell, Cat. No. BE0092) in sterile, Dulbecco’s phosphate-buffered saline, pH 7.4 (DPBS, MilliporeSigma, Rockville, MD) at 23, 25, and 27 days of age and monthly thereafter, with the final injection at 15 mo of age. Mice were euthanized, and tissues were collected at 16 mo of age. We selected this treatment schedule because our previous investigation using short-term, CTLA4-Ig treatments of young mdx mice showed that 25 μg/g body mass produced significant reductions in muscle pathology (19) and previous investigators demonstrated that monthly injections of human RA patients with CTLA4-Ig for 5 yr maintained reductions in RA pathology (22). A total of 17 mice treated with control IgG and 18 mice treated with CTLA4-Ig were used in the investigation.

Muscle Fiber Injury Assay

Muscle membrane damage was assessed by labeling frozen, 10-μm thick cross sections of mdx quadriceps muscles with antibodies to mouse IgG, which is normally localized outside the muscle membrane. The presence of mouse IgG in the sarcoplasm indicates the presence of large lesions of the sarcolemmal membrane, providing an established assay for injured muscle fibers (19, 25-29). Frozen sections were air-dried for 30 min, fixed in cold acetone for 10 min, air-dried again, and washed in phosphate-buffered saline (PBS, pH 7.4). Nonspecific binding was inhibited by incubating sections in blocking buffer [3% bovine serum albumin (BSA), 2% gelatin, and 0.05% Tween-20 in 50 mM Tris-HCl, pH 7.6 containing 150 mM NaCl] for 1 h. Anti-mouse IgG conjugated to DyLight 594 (Vector Laboratories, Cat. No. DI-2594, RRID: AB_2336412) was diluted 1:100 in antibody buffer (3% BSA and 0.05% Tween-20 in 50 mM Tris-HCl, pH 7.6 containing 150 mM NaCl) and applied to sections overnight at 4°C in a humidified chamber. Following three washes in PBS, the sections were mounted with ProLong Gold anti-fade reagent (Invitrogen, Waltham, MA, Cat. No. P36934). The damaged fibers, indicated by intracellular mouse IgG labeling, were manually quantified using an Olympus-BH2 microscope (Central Valley, PA) with fluorescent and Nomarski optics. The number of damaged fibers in each section was expressed relative to the total number of fibers in that section. CTLA4-Ig and control Ig that were injected into treated mice were human IgG and were not detected by the anti-mouse IgG that was used to identify injured fibers.

Immunohistochemistry

Cross sections 10 μm thick were cut from the midbelly of frozen quadriceps muscles using a Microm HM550VP cryostat (Thermo Fisher Scientific, Kalamazoo, MI). Sections were air-dried for 30 min and fixed in cold acetone for 10 min. Endogenous peroxidase activity was quenched by immersing sections in 0.3% H₂O₂ in PBS for 10 min followed by a 5 min PBS wash. The sections were incubated overnight at 4°C in a humidified chamber with primary antibodies diluted in antibody buffer as follows: rat anti-mouse F4/80 (1:50, Thermo Fisher Scientific, Cat. No. 14-4801, RRID: AB_2314387), rabbit anti-mouse CD163 (1:100, Santa Cruz Biotechnology, Cat. No. sc-33560, RRID:AB_2074556), Armenian hamster anti-mouse CD80 (1:50, BioLegend, Cat. No. 104702, RRID:AB_313123), rat anti-mouse CD25 (1:50, Thermo Fisher Scientific, Cat. No. MA5-36078, RRID:AB_2866690), rat anti-mouse CD8A-Ly2 + (1:100, SouthernBiotech, Cat. No. 1550-01, RRID:AB_2794870), goat anti-type I collagen (1:50, SouthernBiotech, Cat. No. 1310-01, RRID:AB_2753206), goat anti-type III collagen (1:50, SouthernBiotech, Cat. No. 133001, RRID:AB_2794734), goat anti-type V collagen (1:50, SouthernBiotech, Cat. No. 1350-01, RRID:AB_2794740), rabbit anti-HSP47 (1:100, Abcam, Cat. No. ab109117, RRID: AB_10888995), rat anti-FoxP3 (1:50, Thermo Fisher Scientific, Cat. No. 14-5773-82, RRID:AB_467576), or mouse anti-developmental myosin heavy chain (dMyHC) (1:50, Leica, Cat. No. 50-255-2212, RRID:AB_563901). After a series of three, 10-min washes in PBS, the tissues were probed with species-specific, biotin-conjugated secondary antibodies (Vector Laboratories) diluted in PBS to 1:200 for 30 min at room temperature (horse anti-goat IgG, Cat. No. BA-9500, RRID:AB_2336123; rabbit anti-rat IgG, Cat. No. BA-4001, RRID:AB_10015300; horse anti-rabbit IgG, Cat. No. BA-1100, RRID:AB_2336201; goat anti-hamster IgG, Cat. No. BA-9100, RRID:AB_2336137). Washes were repeated, and sections were incubated with avidin D-conjugated horseradish peroxidase (HRP) (1:1,000 in PBS, 30 min at room temperature, Vector Laboratories, Cat. No. A-2004, RRID:AB_2336507). The peroxidase substrate, 3-amino-9-ethylcarbazole (AEC), was applied following the manufacturer’s protocol (Vector Laboratories, Cat. No. SK-4200, RRID:AB_2336076), and staining was visualized as a red reaction product. Specificity of staining was verified by excluding the primary antibody from a negative control section. Tissues were mounted using Vectamount AQ (Vector Laboratories, Cat. No. H-5501).

Tissues probed for Foxp3 were fixed in a 4% paraformaldehyde solution for 10 min, after which an antigen-retrieval procedure (30) was performed before proceeding with the standard protocol. The tissues were blocked and probed using the M.O.M immunodetection kit (Vector Laboratories, Cat. No. PK-2200, RRID:AB_2336835) as directed, with 0.3 M glycine added. Tissues probed for dMyHC were blocked with an avidin/biotin blocking kit (Vector Laboratories, Cat. No. SP2001, RRID:AB_2336231) before incubation with the secondary antibody.

Clinical Analyses

Hematopoietic tissues were collected after euthanasia to assay for pathology. Following dissection, the tissues were immersed in 10% neutral-buffered formalin and then placed in plastic cassettes. Cassettes containing the tissues were loaded onto a Sakura VIP 5 Tissue Processor (Sakura Finetek) and processed through graded alcohols for dehydration, xylene for clearing, and paraffin for infiltration. Processed tissues were transferred to a Tissue-Tek Embedding Center (Sakura Finetek) and embedded in paraffin blocks. Sections with 4–5μm thick were collected from each tissue using a HM 355S Microtome (Microm) and mounted onto glass slides. Sections were stained with hematoxylin and eosin (H&E) using a ST5020 automated slide stainer (Leica Biosystems, Nussloch, Germany).

Histopathological analyses of hematopoietic tissues were performed by clinical pathologists blinded to the samples’ treatment group (IDEXX Bioanalytics, Columbia, MO) to assess whether pathological changes occurred with long-term CTLA4-Ig treatment. Changes in histology from ten, 16-mo-old mdx mice (5 mice treated with control IgG and 5 mice treated with CTLA4-Ig) were evaluated using the International Harmonization of Nomenclature and Diagnostic Criteria for Lesions in Rats and Mice (INHAND) nomenclature. A 4-point scoring system was used to grade severity of lesions with “0” indicating the tissue is “within normal limits” and “4” indicating “severe” changes that occupy most of the tissue (31). Whole blood was collected postmortem, and serological analyses from 5 mice treated with control IgG and 5 mice treated with CTLA4-Ig were performed by IDEXX.

RNA Isolation

RNA from cultured cells was extracted with chloroform and precipitated with isopropanol. The resulting pellets were washed with ethanol, and RNA was cleaned and DNase-treated on Clean and Concentrate-5 columns (Zymo, Irvine, CA). Frozen muscles were mechanically homogenized (Dupont, Wilmington, DE) in TRIzol reagent (Invitrogen), and RNA was extracted as described earlier. The resulting pellets were cleaned, and DNase was treated using RNeasy spin columns per manufacturer’s instructions (Qiagen, Hilden, Germany). RNA samples were electrophoresed, and quality was verified by confirming integrity of 28S and 18S ribosomal RNA bands. The RNA concentration of each sample was quantified at 260 nm absorbance, and purity was validated by confirming that all samples had a 260/280 nm absorbance ratio greater than 1.8. cDNA for quantitative PCR was generated by reverse-transcribing 2 μg of each RNA sample using Super Script Reverse Transcriptase II (Invitrogen) and Oligo (dT)_12-18 to prime extension (Invitrogen).

Quantitative PCR

Quantitative PCR (qPCR) was performed on a QuantStudio 5 thermal cycler (Thermo Fisher Scientific, Waltham, MA) using iTaq Universal SYBR Green Supermix (Bio-Rad, Hercules, CA). We empirically identified reference genes that did not vary between our experimental groups using geNorm 3.5 software (geNORM (RRID:SCR_006763) (32) The normalization factor for each sample was calculated by geometric averaging of the Ct values of the reference genes. The expression of each target gene in control samples was set to “1,” and the experimental samples were scaled to that value. Primers used for qPCR were synthesized by MilliPoreSigma and are listed in Table 1.

Table 1.

Primers used for qPCR

Gene Name	Accession No	Sequence (5′-3′)
Ccl2	NM_011333.3	Fwd GCTCAGCCAGATGCAGTTAAC Rev CTCTCTCTTGAGCTTGGTGAC
Ccl3	NM_011337.2	Fwd TTCTCTGTACCATGACACTCTGC Rev CAACGATGAATTGGCGTGGAA
Ccl5	NM_013653.3	Fwd GCAGTCGTGTTTGTCACTCGAAG Rev GAGCAAGCAATGACAGGGAAG
Ccl7	NM_013654.3	Fwd CAACCAGATGGGCCCAATG Rev GATAACAGCTTCCCAGGGACAC
CD80	NM_001359898.1	Fwd CTCTTTGTGCTGCTGATTCGTC Rev GTTTCCCAGCAATGACAGACAG
CD86	NM_019388.3	Fwd CTCTTTCATTCCCGGATGGTG Rev GAGGGCCACAGTAACTGAAGCTG
Coll1	NM_007742.4	Fwd TGTGTGCGATGACGTGCAAT Rev GGGTCCCTCGACTCCTACA
Coll3	NM_009930.2	Fwd ATCCCATTTGGAGAATGTTGTGC Rev GGACATGATTCACAGATTCCAGG
Coll5	NM_001317388.1	Fwd CGGGGTACTCCTGGTCCTAC Rev GCATCCCTACTTCCCCCTTG
Cxcl1	NM_008176.3	Fwd GGCTGGGATTCACCTCAAGAAC Rev GTGGCTATGACTTCGGTTTGG
Cxcl2	NM_009140.2	Fwd CTCTCAAGGGCGGTCAAAAAG Rev CTTTGGTTCTTCCGTTGAGGGAC
Fn	NM_001276412.1	Fwd GCTCAGCAAATCGTGCAGC Rev CTAGGTAGGTCCGTTCCCACTG
Gpr84	NM_030720.2	Fwd CCTTTCTCCGTGGACACATACC Rev CCCTTGGCACTGAAAACCTG
Ifng	NM_008337.3	Fwd GACAATCAGGCCATCAGCAAC Rev CGGATGAGCTCATTGAATGCTT
Pax7	NM_011039	Fwd CTCAGTGAGTTCGATTAGCCG Rev AGACGGTTCCCTTTGTCGC
Rnps1	NM_009070.2	Fwd AGGCTCACCAGGAATGTGAC Rev CGATAGTGCATCTTGGCCTTT
Spp1	NM_009263.3	Fwd CAGCCTGCACCCAGATCCTA Rev GCGCAAGGAGATTCTGCTTCT
Srp14	NM_009273.4	Fwd GAGACGAGCAGTTCCTGAC Rev CGGTGCTGATCTTCCTTTTC
TGFβ	NM_011577.2	Fwd CTCCACCTGCAAGACCAT Rev CTTAGTTTGGACAGGATCTGG
Tnf	NM_013693.3	Fwd CTTCTGTCTACTGAACTTCGGG Rev CACTTGGTGGTTTGCTACGAC
Tpt1	NM_009429.4	Fwd GGAGGGCAAGATGGTCAGTAG Rev CGGTGACTACTGTGCTTTCG

Primary Fibroblast Cultures

Fibroblasts were aseptically isolated from hindlimb muscles of 14-mo-old male mdx mice, using a modification of previously described protocols (33-36). Muscles were collected postmortem in sterile conditions and minced into a coarse slurry and digested in Dulbecco’s modified Eagles medium (DMEM, MilliporeSigma) containing 2.5 U/mL dispase II (Invitrogen), 1% collagenase B (Worthington Biochemical, Lakewood, NJ), and 2.5 mM CaCl₂ dihydrate at 37°C for 45 min with gentle trituration through a serological pipet each 15 min. The digest was passed through a 100-μm filter, and the cells were collected by centrifugation at 416 g at room temperature. The cells were resuspended in growth medium [DMEM containing 100 μ/mL penicillin and 100 μg/mL streptomycin (MilliporeSigma)] with 10% fetal bovine serum (FBS, Omega Scientific, Tarzana, Cat. No. 218211) and plated on gelatin and collagen-coated tissue culture dishes for 1 h at 37°C and 5% CO₂ to allow the fibroblasts to attach. The supernatant containing the unattached cells was removed, and the remaining, attached fibroblasts were cultured in growth medium with changes on alternate days.

CTLA4 Stimulation of Fibroblasts

Muscle fibroblasts were grown in vitro to 80%–85% confluence, after which they were stimulated with 100 μg/mL CTLA-4-Ig or control IgG in growth medium for 3 h or 24 h at 37°C and 5% CO₂. At the end of the stimulation period, the cultures were washed twice with ice-cold DPBS, and cells were collected in TRIzol reagent (Invitrogen, Waltham, MA, Cat. No. 15596026) and frozen for RNA isolation and subsequent experimentation.

Fibroblast Proliferation Assay

Muscle fibroblasts were isolated from 14-mo-old male mdx mice as described earlier and cultured on cover slips coated with gelatin and collagen for use in immunocytochemical analyses. The cultures were grown to ~70% confluency, and the fibroblasts were treated with 100 μg/mL of either control IgG or CTLA4-Ig for 3 h or 24 h. Following stimulation, the plates were washed with DPBS, and the coverslips were fixed in cold methanol for 10 min and air-dried. After washing in PBS, the cells were treated with 0.3% solution of H₂O₂ in PBS and incubated in blocking buffer as described earlier. The fibroblasts were incubated overnight with a mouse-anti-Ki67 (1:50, Santa Cruz Biotechnology, Cat. No. sc-23900, RRID:AB_627859) at 4°C in a humidified chamber and washed as before. Biotinylated anti-mouse antibody (1:200, Vector Laboratories, Cat. No. BA-9200, RRID:AB_2336171) was applied for 30 min at room temperature, the cells washed again, and incubated with avidin-HRP for 30 min at room temperature. After a final series of washes, the staining was visualized using AEC substrate, and coverslips were mounted with Vectamount AQ. Specificity of staining was verified by excluding the primary antibody from a negative control sample. The fibroblasts with Ki67+ nuclei in each sample were manually counted and expressed as the proportion of total fibroblasts.

CD80 Protein Expression in Fibroblasts and Macrophages

The relative expression levels of CD80 protein in bone marrow-derived macrophages (BMDMs) and muscle fibroblasts isolated from adult male, mdx skeletal muscle were measured by Western blot analysis. BMDMs were prepared by aseptically flushing bone marrow from isolated femurs and tibiae and culturing the bone marrow cells (BMCs) for 6 days, as previously described (37). The cells were differentiated into BMDMs in vitro by stimulating with 10 ng/mL macrophage colony-stimulating factor (MCSF, R&D Systems, Minneapolis, MN, Cat. No. 416-ML) in DMEM containing 0.25% heat-inactivated FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin for 24 h. Fibroblasts for protein analysis were isolated from 14-mo-old mdx males and cultured as described earlier.

The cultured BMDMs and fibroblasts were washed with ice-cold DPBS three times and collected in reducing sample buffer [80 mM Tris-HCl pH 6.8, 0.1 M dithiothreitol, 70 mM sodium dodecyl sulfate, 10% glycerol, protease inhibitor cocktail (1:100, Sigma, St. Louis, MO, Cat. No. P8340), 0.2 M Na₃VO₄, and 5 M NaF]. The cells were lysed by passing the suspension through a 23-gauge needle and then boiled for 3 min and centrifuged at 12,000 g for 1 min at 4°C to clear the lysate. Protein concentration of the supernatant was quantified as previously described (38), and 10 μg total protein per lane was used for SDS-PAGE. Proteins were electrophoretically transferred to nitrocellulose membranes, and equal loading of samples was confirmed by staining the blots with 0.1% Ponceau S solution (MilliporeSigma, St. Louis, MO, Cat. No. P-7170). The membranes were blocked with PBS containing 0.1% Tween-20, 0.2% gelatin, and 3% nonfat, dry milk overnight at 4°C to reduce nonspecific binding. The membranes were then probed with rabbit anti-mouse CD80 (1:200, Abcam, Cat. No. ab238481, RRID:AB_3717859) for 2 h at room temperature and washed in PBS containing 0.1% Tween-20 (PBS-T) for 1 h with buffer changes each 10 min followed by a 1-h incubation with HRP-conjugated donkey anti-rabbit IgG (1:10,000, Cytiva, Cat. No. NA934, RRID: AB_772206). The series of PBS-T washes was repeated, and CD80 was visualized with FemtoGlow Western Plus chemiluminescent substrate (Michigan Diagnostics, Royal Oaks, MI, #FWPD02) using a Syngene PXi fluorochem imaging system (Discovery Scientific Solutions, Phoenix, AZ, GeneSys RRID: SCR_015770) running GeneSys V1.5.4.0 software.

Muscle Fiber Cross-Sectional Area and Minimum Feret Diameter Measurements

Muscle cross sections of 10 μm thick were cut from the midbelly of quadriceps muscles, air-dried for 30 min, and then stained with hematoxylin solution (Vector Laboratories, Cat. No. H-3401, RRID:AB_2336842) for 15 min followed by three washes in water. Images of the sections were collected using a BIOQUANT Life Science digital imaging system (RRID:SCR_016423) and an Olympus BH2 microscope equipped with Nomarski optics. The cross-sectional area (CSA) and minimum Feret diameter of 500 randomly selected fibers per section were measured using Fiji software platform (ImageJ2 V2.3.0/1.53q) (RRID:SCR_00285). The classification for large or small fibers was determined by setting three standard deviations from the mean CSA for the control group, as previously described (39). Fibers that were cut at an oblique angle or obscured by artefacts were excluded.

dMyHC Expression

The proportion of quadriceps muscle fibers expressing dMyHC in each treatment group was determined by manually counting the numbers of positively stained and unstained fibers and expressing that as a proportion of the total number of fibers in an entire muscle section. The fiber counts were obtained manually by an investigator using an Olympus BH2 microscope.

Muscle Function Testing

The effects of CTLA4-Ig treatment on functional outcomes in 16-mo-old mice were assessed by measuring the time to exhaustion from treadmill running and performing a modified wire-hang test. Mice were acclimated to the treadmill (Exer 3/6, Columbus Instruments, Columbus, OH) by running at low speed (4 m/min) and later challenged to run to exhaustion at 8 m/min at a 10° incline. Exhaustion was identified when a mouse stopped running and remained on a grid at the end of the treadmill belt that delivered a shock stimulus at an intensity of 3.4 mA with a maximum repetition rate of three, 200 ms pulses per second rather than continuing to run. Mice were also assessed using a modified, wire-hang test (32), during which they performed multiple trials of gripping a horizontal wire with their forelimbs for as long as possible. There were 1-min rest intervals between trials, and the longest time recorded was used for analysis.

Statistical Methods

For all in vivo experiments, each sample represents a separate mouse. Each sample in in vitro experiments consisted of a separate biological replicate. Data are presented as the means ± standard error. The Grubb’s outlier test (P < 0.05) was used to identify statistical outliers prior to data analysis. Statistical significance of experimental outcomes consisting of two treatment groups was evaluated using unpaired Student’s t tests. Clinical scores were tested for significant differences between groups using Fisher’s exact test. Differences with a P value < 0.05 were considered statistically significant. Statistical analyses were performed using GraphPad Prism (RRID:SCR_002798).

RESULTS

Long-Term CTLA4-Ig Treatment Reduces Muscle Fiber Damage and Muscle Inflammation in Aging, Dystrophic Muscle

Previous research showed that depletion of macrophages from young, mdx mice at the early stages of muscular dystrophy greatly reduced muscle fiber membrane lesions and muscle pathology (9) and that treatment of young, dystrophic mice with CTLA4-Ig also reduced numbers of intramuscular macrophages and necrotic fibers (19). Because those findings showed that CTLA4-Ig could reduce muscle pathology at early stages in muscular dystrophy, we tested whether those beneficial effects were also evident in older mdx mice at later, progressive stages of disease. Mice received an injection of CTLA4-Ig (25 μg of per gram mouse mass) once per month for 15 mo, and tissues were examined 1 mo later at 16 mo of age. This treatment produced significant reductions in the frequency of muscle fibers showing detectible membrane lesions (63% reduction compared with controls; P = 0.0001) (Fig. 1, A-C). This CTLA4-Ig injection schedule also significantly reduced numbers of F4/80 + macrophages in dystrophic muscles (71%; P = 0.0001) (Fig. 1, D-F). In addition, macrophages that expressed CD163, a hemoglobin scavenger receptor expressed by a subpopulation of macrophages, were reduced by CTLA4-Ig treatments in 16-mo-old mice (29% reduction; P = 0.0134) (Fig. 1, G-I).

Figure 1.

Prolonged CTLA4-Ig treatments reduce muscle fiber damage and inflammation in aging, mdx dystrophic mice. A and B: microscopic image of a cross section of 16-mo-old quadriceps muscle from an mdx mouse treated with control IgG (A) or CTLA4-Ig (B). The section was labeled with antibodies to mouse IgG. Red fluorescent fibers (arrows) contain endogenous IgG, indicating fiber injury. Scale bars = 70 μm. C: quantification of the % total fibers in quadriceps muscle cross sections that contained endogenous IgG, indicating fiber damage in control IgG or CTLA4-Ig-treated, 16-mo-old mdx mice. D and E: immunohistochemistry of F4/80 + macrophages (examples indicated by arrows) in cross sections of quadriceps muscles from 16-mo-old, control IgG-treated (D) or CTLA4-Ig-treated (E) mdx mice. Scale bars = 20 μm. F: quantification of numbers of F4/80 + macrophages in cross sections of muscles of 16-mo-old mdx mice. G and H: immunohistochemistry of CD163 + macrophages in cross sections of quadriceps muscles from 16-mo-old, control IgG-treated (G) or CTLA4-Ig-treated (H) mdx mice. Scale bars = 20 μm. I: quantification of numbers of F4/80 + macrophages in cross sections of muscles of 16-mo-old mdx mice. Error bars in histograms show the standard error. Each symbol in the histograms represents data from a single mouse. *Significantly different from control IgG group (p < 0.05 using unpaired Student’s t test).

Long-Term CTLA4-Ig Treatment Reduces Numbers of CD80 1 APCs without Affecting Numbers of CD25 1 Cells or CD8 1 T-Cells

Because activated T-cells promote the pathology of mdx muscular dystrophy (7, 8) and CD80 expression on the surface of APCs contributes to T-cell activation through costimulatory pathways, we assayed whether long-term CTLA4-Ig treatment of mdx mice affected numbers of CD80 + cells in aging, dystrophic muscle. Immunohistochemistry showed that CTLA4-Ig treatments caused a significant reduction in CD80 + cells in 16-mo-old mdx muscles (49% reduction; P = 0.0086) (Fig. 2, A and B). However, that reduction in APCs was not accompanied by a reduction in cells expressing the activation marker CD25 (Fig. 2, C and D) or reduction in the number of CD8 + cytotoxic T-cells that can promote pathology in mdx muscle (Fig. 2, E and F).

Figure 2.

Long-term CTLA4-Ig treatment reduces numbers of antigen-presenting, CD80 + cells but does not affect numbers of activated CD25 + cells or CD8 + T-cells in aging, dystrophic muscle. A: immunohistochemistry of CD80 + cells (example indicated by arrow) in cross section of quadriceps muscle from 16-mo-old, control IgG-treated mdx mouse. B: quantification of numbers of CD80 + cells in cross sections of quadriceps muscles of 16-mo-old mdx mice. C: immunohistochemistry of CD25 + cells in cross section of quadriceps muscle from 16-mo-old, control IgG-treated mdx mouse. D: quantification of numbers of CD8 + cells in cross sections of muscles of 16-mo-old mdx mice. E: immunohistochemistry of CD8 + cells in cross sections of quadriceps muscle from 16-mo-old, control IgG-treated mdx mouse. F: quantification of numbers of CD8 + cells in cross sections of muscles of 16-mo-old mdx mice. Scale bars = 50 μm. Error bars in histograms show the standard error. Each symbol in the histograms represents data from a single mouse. *Significantly different from control IgG group (P < 0.05 using unpaired Student’s t test).

Long-Term CTLA4-Ig Treatment Reduces Collagen Content in Aging, Dystrophic Muscle but Does Not Affect Expression of Fibrogenic Genes in Muscle Fibroblasts

We also tested whether CTLA4-Ig treatments affected muscle fibrosis in 16-mo-old mdx mice because fibrosis is a significant pathological feature in muscular dystrophy in aging, mdx mice (23). Immunohistochemical observations showed that CTLA4-Ig significantly reduced the proportion of mdx muscle that was occupied by the major connective tissue proteins, collagen type I (19% reduction; P = 0.0001), collagen type III (18% reduction; P < 0.0001), and collagen type V (19% reduction; P = 0.011) (Fig. 3, A-I). We next tested whether mdx muscle fibroblasts expressed CTLA4 ligands that would enable them to be direct targets of CTLA4-Ig because previous studies demonstrated that fibrogenic cells can express CTLA4 ligands including CD80 (40, 41). Our qPCR data confirmed that mdx muscle fibroblasts express CD80 at levels equivalent to bona fide antigen-presenting macrophages (Fig. 4A). However, mdx muscle fibroblasts expressed CD86 at less than 0.1% (99.98% reduction; P < 0.0001) of the expression levels in macrophages (Fig. 4B). Western blots showed that mdx muscle fibroblasts expressed CD80 protein at higher levels than macrophages (Fig. 4C). We were unable to reliably detect CD86 in Western blots of fibroblasts (data not shown). These data indicate that CD80 on the surface of muscle fibroblasts is the primary ligand for CTLA4-Ig.

Figure 3.

Long-term CTLA4-Ig treatment reduces collagen content in aging, dystrophic muscle. A and B: cross sections of quadriceps muscles from 16-mo-old, mdx mouse treated with control IgG (A) or CTLA4-Ig (B) and labeled with anti-collagen type I (reddish brown color). C: quantification of the percentage of the entire muscle cross section that contained collagen type I. D and E: cross sections of quadriceps muscles from 16-mo-old, mdx mouse treated with control IgG (D) or CTLA4-Ig (E) and labeled with anti-collagen type III (reddish brown color). F: quantification of the percentage of the entire muscle cross section that contained collagen type III. G and H: cross sections of quadriceps muscles from 16-mo-old, mdx mouse treated with control IgG (G) or CTLA4-Ig (H) and labeled with anti-collagen type V (reddish brown color). I:. quantification of the percentage of the entire muscle cross section that contained collagen type V. Scale bars = 40 μm. Error bars in histograms show the standard error. Each symbol in the histograms represents data from a single mouse. *Significantly different from control IgG group (P < 0.05 using unpaired Student’s t test).

Figure 4.

Treatment of muscle fibroblasts with CTLA4-Ig in vitro reduces the expression of proinflammatory cytokines but does not affect expression of connective tissue proteins. A and B: qPCR data showing relative expression levels of CD80 (A) or CD86 (B) in bone marrow-derived macrophages (BMDMs) (Macs) or mdx muscle fibroblasts (Fbs). Expression level in fibroblasts was set at “1.” *Significantly different from control IgG group (P < 0.05 using unpaired Student’s t test). C: upper panel shows Western blot of CD80 protein in extracts of BMDMs and muscle fibroblasts. Mass differences in BMDM and fibroblast samples reflect differences in glycosylation of CD80. Arrows indicate the locations of 70 kDa and 60 kDa molecular mass markers. Bottom: the same blot stained with Ponceau solution before treatment with antibodies to show uniformity of sample loading and transfer to the nitrocellulose membrane. D: qPCR data showing relative expression levels of CD80 and CD86 in muscle fibroblasts after 3 ho or 24 h of treatment with control IgG or CTLA4-Ig. E: qPCR data showing relative expression levels of genes encoding proinflammatory cytokines and chemokines in muscle fibroblasts after 3 h of treatment with control IgG or CTLA4-Ig. F: qPCR data showing relative expression levels of connective tissue genes in muscle fibroblasts after 3 h of treatment with control IgG or CTLA4-Ig. G: qPCR data showing relative expression levels of genes encoding proinflammatory cytokines and chemokines in muscle fibroblasts after 24 h of treatment with control IgG or CTLA4-Ig. H: qPCR data showing relative expression levels of connective tissue genes in muscle fibroblasts after 24 h of treatment with control IgG or CTLA4-Ig. In all histograms, error bars indicate standard error, and each sample represents a biological replicate for which the cells were isolated from a different mouse. For histograms D to H, expression levels in control IgG samples were set at “1” and * indicates significantly different from control IgG group (P < 0.05 using unpaired Student’s t test). In D to H, no comparisons were made between multiple groups; comparisons were only made between the CTLA4-Ig group and the control IgG group for a given outcome, using unpaired Student’s t tests. Each symbol in the histograms represents data from cells isolated from a separate mouse.

We then tested whether 3-h-treatments of mdx muscle fibroblasts with CTLA4-Ig affected gene expression. Because our previous findings showed that direct treatment of macrophages with CTLA4-Ig reduced their expression of proinflammatory cytokines and chemokines (19), we assayed whether CTLA4-Ig had similar effects on muscle fibroblasts and found that treated cells showed significantly reduced expression of Ccl2, Ccl3, Ccl5, Ccl7, Cxcl1, Cxcl2, Tnf, and Gpr84 (Fig. 4E). However, the treatment had no effect on the expression of genes encoding connective tissue proteins that were downregulated in muscles of CTLA4-Ig treated mice (collagen types I, III, and V) or other major connective tissue proteins (fibronectin) or growth factors associated with fibrosis of dystrophic muscle [transforming growth factor-β (TGFβ)] (42, 43) (Fig. 4F). Similar outcomes occurred following longer-term treatments (24 h). CTLA4-Ig-treated fibroblasts showed significant reductions in the expression of Ccl3, Ccl5, Cxcl1, Tnf, and Gpr84 without affecting expression of collagen types I, III, V, fibronectin, or TGFβ (Fig. 4, G and H).

Long-Term CTLA4-Ig Treatment Reduces Numbers of Fibrogenic Cells in Aging, Dystrophic Muscle and Reduces Proliferation of Aging Muscle Fibroblasts In Vitro

Our findings that CTLA4-Ig treatments of aging mdx mice reduced connective tissue accumulation in vivo and could act directly on fibroblasts to reduce proinflammatory gene expression without affecting profibrotic gene expression suggested that the antifibrotic effect that we observed in vivo could result from a CTLA4-Ig-mediated reduction in the numbers of fibroblasts in muscle. Immunohistochemistry observations showed that CTLA4-Ig treatments significantly reduced by 21% (P < 0.0001) the number of fibrogenic cells in the muscle of 16-mo-old mdx mice, identified by their expression of the collagen chaperone, HSP47 (Fig. 5, A-C), suggesting that CTLA4-Ig treatments reduced fibrosis by reducing the numbers rather than activity of fibroblasts. Furthermore, our qPCR data show that CTLA4-Ig treatments reduced the expression of Spp1 in 16-mo-old mdx muscles (43% reduction; P = 0.021, unpaired Student’s t test; n = 6 for each group). Spp1 is a strongly profibrotic gene that is expressed by muscle fibroblasts and immune cells (44, 45). We also tested whether CTLA4-Ig could act directly on fibroblasts to affect their proliferation and found that CTLA4-Ig stimulation of fibroblasts in vitro reduced the proportion of cells that expressed detectible levels of Ki67, which is expressed by proliferative cells (48% reduction; P = 0.0043) (Fig. 5, D-F). Collectively, these results support the interpretation that CTLA4-Ig treatments of aging, mdx mice acts directly on fibroblasts to reduce their proliferation, leading to fewer connective tissue-producing cells in mdx muscles and less fibrosis.

Figure 5.

Long-term CTLA4-Ig treatment reduces numbers of fibrogenic cells in aging, dystrophic muscle and reduces fibroblast proliferation in vitro. A and B: cross sections of quadriceps muscles from 16-mos-old mdx mice treated with control IgG (A) or CTLA4-Ig (B). Sections were immunolabeled with anti-HSP47 to identify fibrogenic cells (examples indicated by arrows). Scale bars = 50 μm. C: quantification of numbers of HSP47 + cells in cross sections of quadriceps muscles of 16-mo-old mdx mice. Each symbol in the histograms represents data from a single mouse. D and E: immunocytochemistry of muscle fibroblasts grown in vitro and treated with control IgG (D) or CTLA4-Ig (E) before labeling with anti-Ki67 to identify nuclei in proliferative cells. Examples of nuclei in proliferative cells are indicated with arrows. Scale bars = 25 μm. F: quantification of the percentage of total fibroblast nuclei in cultures that were Ki67 +. Each symbol in the histogram represents data from fibroblasts isolated from a separate mouse. In all histograms, error bars indicate standard error and * indicates significantly different from control IgG group (P < 0.05 using unpaired Student’s t test).

Long-Term CTLA4-Ig Treatments Have Little Effect on Hematological Data from Aging, Dystrophic Mice

We also assayed whether long-term, CTLA4-Ig treatment affected the numbers or proportion of circulating lymphoid cells or myeloid cells, which can also express CTLA4 ligands on their surfaces. Complete blood counts (CBCs) of mdx mice treated with CTLA4-Ig or control IgG for 15 mo showed that the absolute numbers of all myeloid and lymphoid cell populations were in the normal range for both the CTLA4-Ig and control IgG groups (Table 2). However, CTLA4-Ig treatment produced a large, significant decrease in the numbers of eosinophils in circulation (83% reduction; P = 0.001), although the numbers remained in the normal range. CTLA4-Ig treatment also increased the numbers of circulating reticulocytes (16% increase; P = 0.03) and tended to increase neutrophil numbers (204%; P = 0.07), although the numbers remained in the normal range. All red blood cell parameters [hematocrit (HCT), mean corpuscular volume (MCV), and hemoglobin concentration] were also normal.

Table 2.

Peripheral blood counts from control IgG- and CTLA4-Ig treated, 16-mo-old, mdx mice

Parameter Assayed	Normal Range (95% Interval)	Control IgG (SE)	CTLA4-Ig (SE)	P Value
CBC parameters
WBC, 103/μL	4.45–13.9§	7.0 (0.96)	7.98 (2.63)	0.46
RBC, 106/μL	6.1–10.7*; 7.14–12.2§	9.04 (0.64)	8.89 (1.11)	0.80
HGB, g/dL	10.8–19.2§	12.26 (0.80)	12.30 (1.24)	0.95
HCT, %	33.5–47.8*; 37.3–62§	42.02 (2.56)	42.08 (3.61)	0.98
MCV, fL	43.4–47.8*; 42.7–56.0§	46.60 (1.52)	46.60 (48.48)	>0.99
NRBC (100 WBC)	0	0 (0)	0 (0)	>0.99
MCH, pg	14.8–17.6*; 11.7–16.3§	13.58 (0.41)	13.88 (0.40)	0.28
MCHC, g/dL	29.3–35.9*; 24.6–34.9§	29.18 (0.38)	29.18 (0.56)	>0.99
Absolute data
Absolute neutrophils, /μL	530–3009§	836.2 (411.3)	1702 (841.9)	0.07
Absolute band cells, /μL	0	86.8 (90.3)	235.4 (120.2)	0.06
Absolute lymphocytes, /μL	3240–11150§	5073 (675.9)	4580 (1376)	0.49
Absolute monocytes, /μL	150–940§	722.0 (194.0)	874.8 (304.3)	0.38
Absolute basophils, /μL	0	0 (0)	0 (0)	>0.99
Absolute eosinophils, /μL	10–420§	282.0 (72.6)	48.0 (60)	0.001
Absolute reticulocytes, 103/μL		639.4 (14.8)	740.6 (84.5)	0.03
Absolute platelets, 103/μL	841–2159§	1775 (1080)	1739 (774.9)	0.98
Absolute promyelocytes, /μL	0	0 (0)	0 (0)	>0.99
Absolute myelocytes, /μL	0	0 (0)	0 (0)	>0.99
Absolute metamyelocytes, /μL	0	0 (0)	0 (0)	>0.99
Absolute promyelocytes, /μL	0	0 (0)	0 (0)	>0.99
Differential data
Neutrophils, %	8–20*	11.60 (4.83)	21.20 (8.14)	0.05
Band cells, %	0–5	1.20 (1.10)	3.20 (1.92)	0.08
Lymphocytes, %	76–91*	72.60 (4.04)	58.60 (13.35)	0.06
Monocytes, %	0–4*	10.40 (2.61)	15.40 (6.95)	0.17
Basophils, %	0–130§	0(0)	0 (0)	>0.99
Eosinophils, %	0–3§	4.20 (1.64)	1.60 (2.51)	0.09
Myelocytes, %	0	0 (0)	0 (0)	>0.99
Metamyelocytes, %	0	0 (0)	0 (0)	>0.99
RBC morphology
Reticulocytes, %	3.8–4.2*	7.10 (0.53)	7.65 (0.24)	0.10
Heinz bodies	0	0 (0)	0 (0)	>0.99
Poikilocytosis	None	Not observed	Not observed

Data for normal range values were obtained from Santos et al. (46; indicated by *) or from Charles River labs (47; indicated by §). n = 5 for each group. HCT, hematocrit; HGB, hemoglobin; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; MCV, mean corpuscular volume; NRBC, nucleated red blood cell; RBC, red blood cell; WBC, white blood cell.

Differential blood cell counts showed that there was no significant effect of CTLA4-Ig treatment on any red or white blood cell population (Table 2). However, there were strong trends for decreased proportions of lymphocytes (20% reduction; P = 0.06) and eosinophils (62% reduction; P = 0.09) and increased proportions of neutrophils (83% increase; P = 0.05) and band cells (67% increase; P = 0.08).

Long-Term CTLA4-Ig Treatments Did Not Increase Histopathology in Hematopoietic Tissues in Aging, Dystrophic Mice

Hematopoietic cells can also express CTLA4 ligands on their surfaces (48), which suggests that they could be direct targets of long-term treatments with CTLA4-Ig. We assayed whether CTLA4-Ig treatments were associated with pathological changes in hematopoietic tissues by histopathological assessment of tissues harvested postmortem and graded with a clinical score (31). Each tissue showed some pathological features, such as increased occurrence in lipofuscin pigment containing cells and autolysis in spleens (Fig. 6, A and B), increased occurrence of autolysis and necrosis in liver (Fig. 6, A and C), increased occurrence of dyshematopoiesis and lipofuscin pigment in bone marrow (Fig. 6, A and D), and increased neoplasia in lymph nodes (Fig. 6A). However, all of those pathological features were observed in control IgG-treated mice and were consistent with age-related changes that occur in aging tissues. No significant, CTLA4-Ig treatment effects were observed for any pathological feature that was assessed (Table 2).

Figure 6.

Long-term CTLA4-Ig treatments did not increase histopathology in hematopoietic tissues. A: graded clinical scores for selected histopathological features for hematopoietic tissues from control IgG and CTLA4-Ig treated 16-mo-old mdx mice (5 mice/group). No score differed between the 2 groups for any pathological feature in any tissue assayed. Significance of differences was assessed using Fisher’s exact test (P < 0.05). Error bars show standard errors. B: hematoxylin and eosin (H&E)-stained section of spleen from a control IgG-treated, 16-mo-old mdx mouse showing cells with accumulations of lipofuscin pigment (arrows). Scale bars = 50 μm. C: H&E-stained section of liver from a control IgG-treated, 16-mo-old mdx mouse showing an area with autolysis and necrosis (between brackets). Scale bars = 60 μm. D: H&E-stained section of bone marrow from a control IgG-treated, 16-mo-old mdx mouse. Large cells showing dyshematopoiesis are indicated (arrows). Scale bars = 80 μm. E: H&E-stained section of a rhabdomyosarcoma identified in an intercostal muscle in a CTLA4-Ig-treated, 16-mo-old mdx mouse. The arrow indicates a nucleus in a central-nucleated muscle fiber located in the rhabdomyosarcoma. Scale bars = 60 μm.

Other, nonhematopoietic tissues were also examined by necropsy, which identified no significant, pathological differences between the two groups. Although rhabdomyosarcomas were identified in two of five CTLA4-Ig treated mice (Fig. 6E) and no rhabdomyosarcomas were observed in control-Ig treated mice, the frequency of occurrence of rhabdomyosarcomas did not differ significantly between the two groups (Fisher’s exact test; P = 0.444).

Long-Term CTLA4-Ig Treatment Improves Motor Function in Aging, Dystrophic Mice without Affecting Muscle Mass or Growth

The reduction in muscle damage, inflammation, and fibrosis in CTLA4-Ig treated mdx mice suggested that the treatments could also produce improvements in gross, motor function in dystrophic mice. Although CTLA4-Ig-treated mice showed a significant increase in hang time when the mice clasped a wire by their forelimbs (65% increase; P = 0.0038) (Fig. 7A), the length of time that CTLA4-Ig treated mice ran on a treadmill before exhaustion did not differ from IgG-treated, control mice (Fig. 7B). The functional improvement in the wire-hang test that was not observed in the treadmill running time may reflect that the treatment affects anaerobic but not aerobic function in aging, dystrophic mice, but this is not addressed by the investigation.

Figure 7.

Long-term CTLA4-Ig treatment improves motor function in aging, dystrophic mice without affecting muscle mass or growth. A: hang-time for individual, 16-mo-old mdx mice after treatment with control IgG or CTLA4-Ig. *Significantly different from control IgG group (P < 0.05 using unpaired Student’s t test). B: treadmill running time to exhaustion for individual, 16-mo-old mdx mice after treatment with control IgG or CTLA4-Ig. C: masses of entire bodies for individual, 16-mo-old mdx mice after treatment with control IgG or CTLA4-Ig. D: quadriceps masses from 16-mo-old mdx mice after treatment with control IgG or CTLA4-Ig. E and F: hematoxylin-stained cross sections of entire quadriceps muscles from 16-mo-old mdx mice after treatment with control IgG (E) or CTLA4-Ig (F). Scale bars = 80 μm. G and H: quantification of muscle fiber cross-sectional areas (CSAs) (G) and minimum Feret diameters (H) measured in hematoxylin-stained cross sections of entire quadriceps muscles from 16-mo-old mdx mice after treatment with control IgG or CTLA4-Ig. I: frequency distribution of fiber CSAs of control IgG or CTLA4-Ig treated mouse muscles (n = 5 for each group). J: qPCR data showing relative expression level of Pax7 in quadriceps muscles from 16-mo-old mdx mice after treatment with control IgG or CTLA4-Ig. K and L: cross sections of quadriceps muscles from 16-mo-old, mdx mice treated with control IgG (K) or CTLA4-Ig (L) and labeled with anti-developmental myosin heavy chain (dMyHC) (reddish brown color). M: quantification of the percentage of total fibers that contained detectible dMyHC. In all histograms, error bars indicate standard error. Each symbol in the histograms represents data from a single mouse.

Because CTLA4-Ig treatments significantly reduced numbers of CD163 + macrophages (Fig. 1I), which can promote muscle growth and regeneration (19), we also assayed for treatment effects on mouse mass and muscle mass, but observed no treatment effects on masses of 16-mo-old mice (Fig. 7C) or the masses of quadriceps muscle (Fig. 7D) or other muscles that primarily contain fast muscle fibers (tibialis anterior; P > 0.999, n = 6 per group) or primarily slow muscle fibers (soleus, P = 0.775, n = 6 per group). CTLA4-Ig treatments also did not affect the CSAs of muscle fibers (Fig. 7, E-G) or the minimum Feret diameter of muscle fibers (Fig. 7H) or the relative proportions of small and large fibers in 16-mo-old mdx muscles (Fig. 7I), indicating an absence of effect on muscle growth and regeneration. CTLA4-Ig treatments also had no effect on levels of expression of Pax7 (Fig. 7J) or proportion of muscle fibers that expressed dMyHC (Fig. 7, K-M) in 16-mo-old mice.

DISCUSSION

The primary finding in our investigation is that long-term treatment of mdx mice with CTLA4-Ig reduces muscle damage, inflammation, and fibrosis. Because the treatments began at 3 wk of age and continued until 15 mo of age, which corresponds to approximately early childhood to midadulthood in humans, these results indicate that CTLA4-Ig can provide a long-term therapy for treating DMD. In addition, the long-term CTLA4-Ig treatments did not increase histopathology in hematopoietic tissues and had little effect on peripheral blood counts; this is a significant outcome because hematopoietic cells would be the primary targets of CTLA4-Ig and could potentially experience unpredicted, negative consequences. Because the half-life of CTLA4-Ig in serum following injection into mice is ~19 to 70 h (49), the in vivo treatment effects that we observed at 30 days following the final CTLA4-Ig injection reflected prolonged effects that did require persistent drug exposure.

The effects of long-term treatments with CTLA4-Ig on the immune response to mdx muscular dystrophy reported in the present study differ from the effects of short-term treatments on young, mdx mice (19). Although both treatments produced significant reductions of CD80 + APCs in dystrophic muscle, short-term treatments of young mdx mice reduced the numbers of cytotoxic CD8 + T-cells and cells expressing the CD25 activation marker, but long-term treatments did not affect numbers of those cells in muscle. This suggests that the primary mechanism through which CTLA4-Ig reduces muscle damage and inflammation in dystrophic muscle may differ between short versus long-term treatments, in which the short-term effect involves reductions in costimulatory activation of intramuscular T-cells, whereas long-term treatments do not involve detectible reductions in T-cell activation. Previous investigators have also found that the response of young and old mice to CTLA4-Ig differs. Young mice that were treated with CTLA4-Ig following skin transplantation experienced significant reductions in specific subsets of CD4 + and CD8 + T-cells and lower levels of IFNγ expression, but CTLA4-Ig treatment of old transplant recipients did not produce those outcomes (50). However, an alternative explanation for the different outcomes in young and old mdx mice is that the effects may reflect the different schedules for CTLA4-Ig delivery in the two studies. Young mice in the short-term treatment investigation received CTLA4-Ig injections on days 5, 3, and 1 before the muscles were collected for analysis (19). In the present, long-term study, muscles were collected after 15 mo of monthly injections. We were unable to resolve the question of whether more frequent injections of CTLA4-Ig during long-term treatments would produce a different outcome because of the high cost of CTLA4-Ig.

Although blockade of costimulatory signals required for T-cell activation is the best characterized mechanism through which CTLA4-Ig treatments can affect the immune response in autoimmune diseases, CTLA4-Ig can affect gene expression and cell function by acting directly on CD80 + APCs, independent of the presence of T-cells. For example, application of CTLA4-Ig to human monocytes or macrophages in vitro reduces their production of proinflammatory and profibrotic cytokines (51, 52). Direct action of CTLA4-Ig on BMDMs from mdx mice also reduces their expression of proinflammatory genes and inhibits their chemotactic response to chemoattractive chemokines, such as CCL2, that are expressed in dystrophic muscle and reduces mdx macrophage cytotoxicity in vitro (19). Thus, the reduction in inflammation and muscle injury observed in the current study that occurs in the absence of increased T-cell activation is attributable, in part, to direct actions of CTLA4-Ig on myeloid cells in the aging, dystrophic mice.

Our findings also show that fibroblasts from aging, dystrophic muscle express CD80, which indicates that they can also be direct targets of CTLA4-Ig treatments for muscular dystrophy. Previous investigators have shown that fibrogenic cells can express costimulatory molecules that are ligands for CTLA4, which potentially makes them direct targets for CTLA4-Ig treatments. For example, fibrogenic cells from the synovia of mice with chronic inflammatory arthritis express CD80 on their surfaces, and their stimulation with CTLA4-Ig in vitro causes CD80 internalization although no change in the expression of inflammation-related genes occurred (41). In contrast, our findings show that CTLA4-Ig stimulation reduced the expression of proinflammatory genes by fibroblasts isolated from mdx mouse muscles, indicating that reduction in inflammation of dystrophic muscle may reflect direct actions CTLA4-Ig on fibroblasts, in addition to myeloid cells.

The ability of CTLA4-Ig to act directly on mdx muscle fibroblasts to modify their expression of proinflammatory genes suggested that the reduction in connective tissue accumulation in the muscles of CTLA4-Ig-treated, aging mdx muscle could also result from the downregulation of connective tissue gene expression. Previous findings supported the possibility by showing that stimulating fibrogenic cells from patients with systemic sclerosis with CTLA4-Ig in vitro reduced their expression of collagen type 1 and TGFβ (40). However, we observed no treatment effect of CTLA4-Ig on the expression of major connective tissue genes by mdx muscle fibroblasts and also found no significant effect on the expression of TGFβ, which is associated with increased fibrosis in mdx muscular dystrophy (42, 43). Instead, we found that CTLA4-Ig reduced proliferation of mdx muscle fibroblasts, supporting our interpretation that the antifibrotic effects on CTLA4-Ig on muscles in aging mdx mice resulted from a reduction in their numbers rather than affecting their profibrotic activities.

We observed that more CTLA4-Ig treated, mdx mice had rhabdomyosarcomas than occurred in control IgG-treated mice, although the increase was not statistically significant. Nevertheless, the observation provides a caution before advancing toward clinical treatment of DMD with CTLA4-Ig. Although rhabdomyosarcomas are rare in humans, they occur in ~6% of aging, mdx mice (53). In contrast, a recent matched-cohort study that included 2,355 patients with muscular dystrophy reported no occurrence of sarcomas in the patients, although tumors of the central nervous system increased (54); this shows that the risk of sarcomas associated with muscular dystrophy varies between humans and mouse models of the disease. Other findings suggest that long-term treatment with CTLA4-Ig increases occurrence of cancer in humans, at least in some chronic diseases other than muscular dystrophy. RA has been best studied in this regard. Clinical assessment of 317 RA patients who received CTLA4-Ig at 10 mg/kg every 4 wk for 5 yr showed that 21 patients had malignant neoplasms at the end of the study, although no control group receiving placebo for during the last 4.5 yr of the study was included for comparison (22). The most frequently observed neoplasms identified in the study were basal cell carcinoma (n = 6) and squamous cell carcinoma (n = 6) (22). No sarcomas were reported. A subsequent, observational study in which the risk of cancer in RA patients was assessed using the World Health Organization’s global database compared the frequency of reports of cancer in RA patients treated with CTLA4-Ig compared with patients treated with other biologic, disease-modifying-anti-rheumatic-drugs (bDMARDs), such as TNFα inhibitors (55). Analysis of data from 15,846 CTLA4-Ig-treated RA patients and 290,568 RA patients treated with other bDMARDs identified an increased risk of reporting malignant melanomas in patients treated with CTLA4-Ig, although the occurrence of sarcomas was not addressed in the report (55). Those findings of unexpected, pathological effects of CTLA4-Ig treatments for RA provide a caution for similar pathological outcomes if the drugs were used to treat DMD. However, the occurrence of serious, adverse effects can also be greatly influenced by the pathological context in which the drug is administered. This was highlighted by the history of medical treatments of DMD patients. Some children with DMD who received safe, FDA-approved, commonly used anesthetic agents (e.g., halothane) for unremarkable medical procedures, such as tonsillectomy, experienced dangerous and, in some cases, fatal drug reactions that were attributable to DMD pathology (56-59).

Overall, the significant reductions in injury, inflammation, and fibrosis that occurred in aging, mdx mice after prolonged treatment with CTLA4-Ig that we have identified provide substantial evidence that CTLA4-Ig can be therapeutically useful for treating DMD. However, additional preclinical studies with a larger sample size over a prolonged period of treatment are needed to determine whether CTLA4-Ig increases tumor formation in muscular dystrophy, preferably using an animal model of muscular dystrophy in which there is not an elevated occurrence of rhabdomyosarcomas.

DATA AVAILABILITY

The datasets that were generated and analyzed in the current investigation are available from the corresponding author upon reasonable request.