Bioengineered constructs combined with exercise enhance stem cell-mediated treatment of volumetric muscle loss

Abstract

Volumetric muscle loss (VML) is associated with loss of skeletal muscle function, and current treatments show limited efficacy. Here we show that bioconstructs suffused with genetically-labelled muscle stem cells (MuSCs) and other muscle resident cells (MRCs) are effective to treat VML injuries in mice. Imaging of bioconstructs implanted in damaged muscles indicates MuSCs survival and growth, and ex vivo analyses show force restoration of treated muscles. Histological analysis highlights myofibre formation, neovascularisation, but insufficient innervation. Both innervation and in vivo force production are enhanced when implantation of bioconstructs is followed by an exercise regimen. Significant improvements are also observed when bioconstructs are used to treat chronic VML injury models. Finally, we demonstrate that bioconstructs made with human MuSCs and MRCs can generate functional muscle tissue in our VML model. These data suggest that stem cell-based therapies aimed to engineer tissue in vivo may be effective to treat acute and chronic VML.

Figure 1. Mouse model of VML results in irrecoverable loss of function and structure.

(a) VML surgical procedure. From left to right, a photographic sequence shows tattooing and surgical ablation with a micro-blade (indicated by a yellow arrowhead) of a defined mass from a TA muscle. (b) Model of TA muscle with ablation injury. A rendition of the rectangular pocket created within the TA by the surgical excision is shown. (c) Representative H&E staining of whole TA muscle cross-sections. An uninjured TA muscle (left panel) and a TA muscle 30 days following VML injury (right panel) are shown. The tibia and the extensor digitorum longus (EDL) muscle have been drawn for each image to facilitate the orientation and the anatomy of the TA muscles. A dashed line indicates the area corresponding to the muscle area removed in the surgical procedure. For the VML model image, the boxed areas are shown in higher magnification to the right (scale bar=2 mm). Left box: peripheral fibrotic scarring is observed in place of the excised muscle (indicated by arrow bars). Right box: a fibrotic scar can be seen extending into the belly of the TA muscle (indicated by arrow head) (scale bar=500 μm). (d) Quantification of TA muscle tissue masses. TA muscles were weighed immediately following dissection. For muscles subjected to VML injuries, the ablated muscle tissues were immediately weighed (white bar) and added to the muscle mass remaining 30 days after the ablation (red bar). Comparisons were made to unablated muscles. (e) Force production, measured through a force transducer, of TA muscles 30 days after VML injuries compared with uninjured control TA muscles. Contractions were induced in vivo through direct sciatic nerve stimulation (left graph) or ex vivo by inducing contraction directly through electrical stimulation in a culture bath (right graph) (n=6). Data are±s.e.m. For statistical analysis, t-tests were used. *P<0.05; **P<0.001; ****P<0.00001.

Figure 2. MRCs support MuSCs in de novo myofibre formation.

(a) Representative FACS plot of MRCs. Lower limb muscles were dissected and digested to obtain a mononucleated cellular suspension. These cells were marked using cell specific surface antigens as described in the ‘Results’ and in the ‘Methods’ sections and were analysed using FACS. Five populations were isolated: muscle stem cells (MuSCs), hematopoietic cells (HCs), endothelial cells (ECs), fibro-adipogenic progenitor cells (FAPs) and fibroblast-like cells (FLCs). The relative percentages of each cell population are 10%, 25%, 39%, 8% and 18%, respectively (n=6). (b) Quantified results of in vitro bioluminescence generated from cultured bioconstructs containing Luc⁺ MuSCs, either alone (MuSC⁺/MRC⁻) or in combination with Luc⁻ MRCs (MuSC⁺/MRC⁺). Bioconstructs were cultured for three days, and bioluminescence was measured each day (n=4). (c) Representative images of bioluminescence measured from mice 10 days after transplantation of bioconstructs in left TA muscles immediately following VML injury (d) Quantified results of non-invasive imaging of transplanted bioconstructs. Bioconstructs with no cells (MuSC⁻/MRC⁻), Luc⁺ MuSCs (MuSC⁺/MRC⁻), or Luc⁺ MuSCs in addition to Luc⁻ MRCs (MuSC⁺/MRC⁺) were transplanted into TA muscles that had received VML injuries. Bioluminescence was measured 10 days following transplantation (n=4). Data are±s.e.m. For statistical analysis, t-tests were used. *P<0.05; **P<0.001.

Figure 3. Perfusion of bioconstructs in a bioreactor improves MuSC efficacy.

(a) Perfusion of bioconstructs sustains MuSC viability in vitro. Bioconstructs were reconstituted with Luc⁺ MuSCs and Luc⁻ MRCs and then cultured for 24 h. Bioluminescence was measured at different time points, as indicated (n=6). (b) Model of the bioreactor designed to perfuse cultured bioconstructs. An autoclavable plexiglass chamber, capable of holding, in parallel, five independent tubing lines, is shown (top panel). These lines were connected to an external pump. Each line contained one tubular culture chamber capable of hosting one bioconstruct (bottom panel). (c) Representative bioluminescence and fluorescence images of bioconstructs before and after transplantation. Bioconstructs were generated with Luc⁺/GFP⁺ MuSCs and Luc⁻/GFP⁻ MRCs, cultured under either perfused or static conditions, and imaged non-invasively before transplantation (Pre Op, left panels). Bioconstructs that had been imaged previously in vitro were imaged non-invasively in vivo immediately following transplantation (Post Op, right panels). (d) Quantitative results comparing bioluminescence of perfused and static bioconstructs at periodic intervals following transplantation into VML injuries (n=4). (e) (Left panels) Representative H&E-stained cross-sections of muscles 10 days after receiving perfused or static bioconstruct transplantation to treat VML injuries. Bioconstructs (BC) are shown adjacent to unablated muscle (UM) (scale bar=500 μm). (Right panels) The regions outlined by the yellow boxes in the left panel are magnified here (scale bar=200 μm Data are±s.e.m. For statistical analysis, t-tests were used. *P<0.05; **P<0.001; ***P<0.0001.

Figure 4. Treatment of acute VML restores mass and force production.

(a) Schematic model of our VML treatment procedure using bioconstructs. TA muscles are isolated from donor mice (1). These muscles are used to obtain either: decellularized scaffolds, MuSCs, MRCs or ECM proteins to generate hydrogels (2). Scaffolds are reconstituted with isolated cells and hydrogel to generate bioconstructs (3). Bioconstructs are cultured in a bioreactor while the hydrogel cured, allowing media to perfuse across the bioconstructs (4). Once ready, bioconstructs are transplanted into VML injured TA muscles (5). (b) In vivo force production measurements of TA muscles treated with different bioconstructs following VML injury. After 30 days, the distal tendons were attached to a force transducer and contractions were induced through sciatic nerve stimulation (n=8). (c) Ex vivo force production measurements. The same muscles measured in b were then dissected and cultured in a chamber. The distal tendons were attached to a transducer and contractions were induced electrically in the culture bath (n=8). (d) The mass of each TA muscle was measured following ex vivo force measurements (n=6). In (b–d) average values of muscles that did not received VML injuries are labelled ‘uninjured’ and are indicated by a blue dotted line; average values of muscles that received a VML injury without any treatment are labelled ‘VML’ and are indicated by a red dotted line. Data are±s.e.m. For statistical analysis, t-tests were used. **P<0.001; ****P<0.00001, n.s.: not significant.

Figure 5. Treatment of VML improves and restores tissue structure.

(a) Representative immunofluorescence (IF) images of cross-sections of VML-injured TA muscles treated with different bioconstructs. Three of the same muscles used for force productions in Fig. 4 were sectioned and immunostained (scale bar=1 mm). (b) Measurements of whole muscle tissue areas for all muscles assessed in Fig. 5 (n=6). (c) Measurements of fibrotic tissue areas within total muscle areas quantified in b (n=6). (d) Representative IF cross-sectional images showing donor-derived eYFP⁺ myofibres following treatment. Muscles treated with bioconstructs generated using eYFP⁺ MuSCs and eYFP⁻ MRCs were harvested after taking force measurements. The panel shows a region in which a bioconstruct was implanted within a VML-injured TA muscle. The yellow arrow indicates peripheral fibrotic scarring as illustrated in Fig. 1. The white arrow indicates residual scaffold surrounded by de novo myofibres (scale bar=500 μm). (e) Quantification of the number of eYFP⁺ myofibres in muscles treated with bioconstructs generated with either eYFP⁺ MuSCs alone or together with eYFP⁻ MRCs (n=6). (f) Quantification of cross-sectional areas of the myofibres counted in e (n=6). (g) Representative IF longitudinal images showing donor-derived eYFP⁺ myofibres in transplanted bioconstructs (scale bar=500 μm). The left panel shows a low magnification image of a TA muscle centred on the transplanted bioconstruct region (scale bar=500 μm). The yellow arrow indicates the peripheral fibrotic scarring adjacent to the unablated muscle (UM), which is overlying de novo muscle fibres within the bioconstruct (BC). The orange arrows indicate two stiches securing the BC within the tissue. The white arrow indicates a region of residual scaffold. The right panel shows at higher magnification image of donor-derived eYFP⁺ myofibres. The white arrows again indicate regions of residual scaffold (scale bar=50 μm). (h) Representative IF images showing blood vessels structures formed by endothelial cells (CD31⁺) within regions of transplanted bioconstructs (scale bar=100 μm). (i) Quantification of blood vessels from regions of transplanted bioconstructs in muscles with VML injuries (n=6). Data are±s.e.m. For statistical analysis, t-tests were used. **P<0.001; ***P<0.0001.

Figure 6. Exercise improves innervation of de novo myofibres and improves forces in vivo.

(a) Representative image of a mouse during a gait analysis (top) and the gait footprints collected during the analysis (bottom). Mice were positioned in a transparent treadmill and a camera was positioned underneath to record the gait. (b) Quantification of the gait ‘disability score’ resulting from the analysis of 47 parameters (see Methods) (n=6). (c) In vivo force production measurements of TA muscles treated with bioconstructs following VML injury in non-exercised or exercised mice. After 30 days, the distal tendons were attached to a force transducer and contractions were induced through sciatic nerve stimulation (n=6). (d) Ex vivo force production measurements from non-exercised or exercised mice. The same muscles measured in c were then dissected and cultured in a chamber. The distal tendons were attached to a transducer and contractions were induced electrically in the culture bath (n=6). (e) (Left) Representative IF image of transplanted bioconstruct. Yellow arrows indicate donor-derived (eYFP⁺) myofibres with NMJs (αBTX⁺) within regions of the transplanted bioconstruct. (Right) Higher magnification of an NMJ associated with a donor-derived myofibre (scale bars=50 μm). (f) Quantification of NMJs in whole cross-sections of transplanted bioconstructs along 3 mm lengths of TA muscles. Muscles were either uninjured and exercised (‘-VML, +Ex’) or subjected to VML injury and bioconstruct treatment (‘+VML, +Tx’) without (‘-Ex’) or with (‘+Ex’) exercise (n=5). (g) (Larger panels) Representative IF images of myofibres with mature NMJs (αBTX⁺ and also stained positive for Synaptophysin (SynPh) and Neurofilament (NFL)) within regions of transplanted bioconstructs (scale bar=100 μm). (Smaller panels) Higher magnification of a mature NMJ (scale bar=10 μm). (h) Quantification of mature NMJs in whole cross-sections of uninjured muscles or of injured muscles with transplanted bioconstructs along 3 mm lengths of TA muscles characterized as in f (n=4). Data are±s.e.m. For statistical analysis, t-tests were used. **P<0.001; ****P<0.00001.

Figure 7. Human MuSCs and MRCs generate de novo human muscle tissue in VML injuries.

(a) Representative histological images of murine TA muscles following VML injury and treatment with hMuSC/hMRC bioconstructs. The yellow dashed lines indicate the regions where bioconstructs were transplanted (scale bar=2 mm). (b) Representative IF staining of mouse and human proteins in the same muscles as in a (scale bar=50 μm). (c) Quantification of the number of human Integrin α7β1⁺ myofibres in muscles treated with bioconstructs generated with either human MuSCs alone or together with human MRCs (n=4). (d) Bioconstructs were generated with: no cells, freshly isolated mouse MuSCs and MRCs, or freshly isolated human MuSCs and MRCs. These bioconstructs were transplanted in VML injuries of TA muscles of immunocompromised mice. Muscles were analysed for force production 30 days later. The graph shows quantification of in vivo force production comparing the three conditions. (e) The same muscles analysed in d were analysed and quantified for ex vivo force production. Data are±s.e.m. For statistical analysis, t-tests were used. **P<0.001.