4-week-old C57BL/6J and C57BL/10ScSn-Dmdmdx/J (also known as MDX or dystrophin-deficient) male and female breeding pairs were purchased from Jackson Laboratory (Bar Harbor, ME) and bred to establish colonies for this study. All animal use was in accordance with National Institutes of Health guidelines and conformed to the principles specified in a protocol approved by Western Michigan University Homer Stryker M.D. School of Medicine Institutional Animal Care and Use Committee.

Mouse MuSCs (3-month-old, C57BL/6J female) were isolated according to Lavasani et al. (2013) and the E. coli lacZ gene was integrated into MuSCs (LacZ + MuSC) according to Naderian et al. (2011) using a lentivirus vector. LacZ + MuSC were maintained on collagen coated substrates with Dulbecco’s modified Eagle’s medium, high glucose (DMEM, Gibco), supplemented with 20% fetal bovine serum (FBS, Gibco), 1.5% chick embryo extract (APS, United Kingdom), and 1% penicillin/streptomycin at 37°C and 5% CO₂. LacZ + MuSCs were subcultured with trypsin (0.25% trypsin-EDTA, Gibco BRL) at 50%–60% confluency and were used up to 30 passages.

Six biological replicates of LacZ + MuSCs were taken for RNA extraction after trypsinization. Total mRNA was isolated using the Monarch^® Total RNA Miniprep Kit (New England BioLabs, United States). The high- Capacity cDNA Reverse Transcription Kit (Applied Biosystems, United States) was used for the first strand cDNA synthesis. Quantitative PCR (qPCR) was performed with iTaq Universal SYBR Green Supermix (Bio Rad, United States) using the CFX Connect Real-Time System Thermal Cycler (Bio-Rad). All reactions were run in duplicate. After vortexing, 10 μL aliquots of the mixture were pipetted into each well of a 96-well thin-wall PCR plate (Bio-Rad). PCRs consisted of a denaturing cycle at 95°C for 3 min followed by 40 cycles of 10 s at 95°C, and 30 s at 60°C. Fluorescence was quantified during the 60°C annealing step and product formation was confirmed by melting curve analysis (65°C–95°C). Relative mRNA amounts of target genes were calculated after normalization to an endogenous reference gene, β-actin, and compared to the negative marker (CD45) with the arithmetic formula 1/ΔCt. Based on the cDNA sequences available at the EMBL database, the following specific murine primer pairs were designed by the software Primer3 (Whitehead Institute for Biomedical Research, Cambridge, MA, United States, http://www-genome.wi.mit.edu/cgibin/primer/primer3_www.cgi/) and confirmed by the sequences in the NCBI database (http://www.ncbi.nlm.nih.gov/). All primers shown in Supplementary Table S1 were synthesized by Integrated DNA Technologies (United States).

A hypoxia incubator (ThermScientific HERACellVIOS 160i, United States), set to 5% O₂ nitrogen injection, was used to treat LacZ + MuSCs. Cells were treated for 6, 12, 24, 48, or 72 h under 5% O₂ with 21% O₂ as the control group. Additionally, LacZ + MuSCs were treated with hypoxia in conjunction with 0.03 ng/μL mouse IFN-γ Recombinant Protein (BioLegend, United States) or 1 μM BMS-1 (small-molecule inhibitor of PD-L1, MedChemExpress, United States).

LacZ + MuSC cells were cultured in six-well plates and then lysed with 2x laemmli sample buffer solution (Bio Rad, United States) for the collection of total cellular protein. Each lane of a sodium dodecyl sulfate (SDS)-polyacrylamide gel was loaded with approximately 150 μg of extracted protein. SDS-Page was used to separate proteins by size which were then transferred to polyvinylidene fluoride (PVDF) membranes (Bio Rad, United States). The PVDF membranes were placed in a blocking buffer [TBS-T with 5% skimmed milk) then incubated at 25°C for 1 h. The same buffer was used to dilute the primary and secondary antibodies. Each PVDF membrane was incubated with primary and secondary antibodies sequentially. Finally, protein blots were developed using a SuperSignal West Pico PLUS Chemiluminesent Substrate ECL Plus kit (Thermo Scientific, United States)] and visualized using a chemiluminescence imager (LI_COR, United States). All Western blotting raw data in this paper are shown in Supplemental Materials. Antibodies used for Western blotting include: anti-PD-L1 (1:500, Abcam, ab213480, Boston, United States), anti-β actin (1:5,000, NB600-501). HRP-Goat Anti-Rabbit and HRP-Goat Anti-Mouse (Life Technologies, United States) were used as the secondary antibodies for anti-PD-L1 and anti-β actin, respectively.

All mice received intramuscular injections to the gastrocnemius thigh muscles (GM). To prevent variation due to injection technique, a single, experienced technician performed intramuscular injections for all mice. Mice were anesthetized with isoflurane (5% induction, 1%–2% maintenance) delivered in oxygen by using an anesthetic vaporizer (Isoflurane, funnel-fill vaporizer, VetEquip, Piney River, VA). Once anesthetized, mice were placed sternally on a tabletop with a nose cone supplied with 1%–2% isoflurane gas to maintain anesthetic depth. Once a mouse was prepared for injection, the technician advanced the needle through the skin, within the marked area, and into the target muscle to a depth of approximately 2–4 mm. The syringe plunger was aspirated to check for inadvertent placement within a blood vessel, and then the injection was delivered. 2.0 × 10⁶ treated LacZ + MuSCs suspended in 20 µL of Hank’s Balanced Salt Solution (HBSS, Gibco) were injected into each GM.

Animals were sacrificed using carbon dioxide (CO₂) inhalation followed by cervical dislocation as approved by the animal protocol. The GM was isolated according to Ionescu et al. (2016). Minced muscle tissue was digested using Collagenase Type II (ThermoFisher, United States) for 1.5 h at 37°C and 5% CO₂ followed by Dispase II (ThermoFisher, United States) digestion for 1 h. After centrifugation at 3,500 rpm for 5 min, digested muscle tissue was mechanically dissociated in complete media (DMEM, 20% fetal bovine serum, 1.5% chick embryo extract, 1% penicillin/streptomycin) and filtered through a 70 µm filter. Isolated cells were plated in 6-well plates to be LacZ stained for LacZ + MuSCs survival after injection to the GM. Additionally, isolated cells were used for flow analysis to assess CD4⁺ and CD8⁺ infiltration after 3, 5, and 7d in vivo after GM injection.

An Attune NxT flow cytometer (Invitrogen, United States) was used for flow cytometric analysis. Antibodies included phycoerythrin (PE)-conjugated anti-mouse MHC-I (Catalog #562004, BD Biosciences, United States), TLR3 (Catalog #141904, BioLegend, United States) and TLR7 (Catalog #160004, BioLegend, United States); allophycocyanin (APC)-conjugated anti-mouse PD-L1 (Catalog #124312, BioLegend, United States), ICAM (Catalog #116120, BioLegend, United States) and CD4 (Catalog #100412, BioLegend); and Fluorescein isothiocyanate (FITC)-conjugated anti-mouse MHC-II (Catalog #562015, BD Biosciences, United States) and CD8 (Catalog #100706, BioLegend, United States). For cell surface analysis of MHC-I, MHC-II, TLR3, TLR7, ICAM, and PD-L1, both C2C12 and LacZ + MuSCs monolayers were grown for 72 h under 5% O₂, 21% O₂, 5% O₂ with 0.03 ng/μL mouse IFN-γ, and 21% O₂ with 0.03 ng/μL mouse IFN-γ before assessment with flow cytometry. CD4⁺ and CD8⁺ assessment was done on digested GM samples collected after 3, 5, 7d in vivo on injected LacZ + MuSCs treated for 72 h with 5% O₂, 21% O₂, 5% O₂ with 1μM BMS-1, 21% O₂ with 1 μM BMS-1 and 5% O₂ with 0.03 ng/μL mouse IFN-γ, and 21% O₂ with 0.03 ng/μL mouse IFN-γ. Spleen was collected as a positive control for CD4⁺ and CD8⁺ assessment. Medium fluorescent intensity results were assessed by Offline analysis with FlowJo 10.9.

LacZ gene expression was examined after fixing adherent cells (at 80% confluency) isolated from the GM muscle with 100% methanol for 5 min as described previously (Verma and Asakura, 2011).

BMS-1 toxicity on the metabolic activity of LacZ + MuSC was analyzed using the MTT assay under normoxia conditions (21% O₂). To assess cell viability for each BMS-1 concentration, 1 × 10³ lacZ + MuSC cells in 200 μL media were incubated with 1, 2, or 4 μM BMS-1 in a 96-well plate coated with collagen. After a 24 h incubation, 8 μL of a 5 mg/mL MTT 1-(4,5-Dimethylthiazol-2-yl)-3,5-diphenylformazan, Sigma #M2003) solution was added to each well. After 4 h, the cell-free supernatant was discarded, and the formazan crystals were dissolved with 100 μL DMSO. The OD was detected at 570 nM. The % viability rates were calculated relative to NAD(P)H-dependent cellular oxidoreductase activity in non-treated (CTL) versus BMS-1 treated cells.

The data was presented as mean ± SD for flow cytometry analysis or mean ± SEM for western blot and MTT analysis. One-way ANOVA followed by Dunnett’s multiple comparisons test, and Unpaired or one-tailed t-test was performed (where applicable) using GraphPad Prism version 10.1.0. If not indicated otherwise, the criterion for significance was set at P ≤ 0.05. All **** represent P ≤ 0.0001 unless otherwise indicated.

Myoblast transfer therapy was the first form of quasi-gene therapy proposed for treating DMD (Partridge, 2002). However, skeletal muscle stem/progenitor cells are more valuable given their enhanced regenerative capacity and their improved engraftment potential (Hosoyama et al., 2012). Thus, we first isolated MuSC and integrated the E. coli lacZ gene to these cells (LacZ + MuSCs). We then used qPCR to quantify the levels of the following markers: CD34, a hematopoietic/progenitor marker; Sca1, a muscle derived stem cell marker (MDSC); Pax7, a MuSC maintenance marker; Myogenic factor 5 (Myf5), a satellite cell marker; myoblast determination protein 1 (MyoD), an active satellite cell marker; Myogenin, an early muscle skeletal muscle differentiation marker; Myosin Heavy Chain 7 (MyHC7), a skeletal muscle marker; and Ki67, a proliferation marker (Figure 1A). These markers were compared to the negative control, CD45, a marker of nucleated hematopoietic cells. The qPCR results showed significant (****P ≤ 0.0001) muscle stem/progenitor related transcriptome levels of CD34, Sca1, Pax7, Myf5, MyoD, and myogenin, compared to the negative marker, CD45, whose baseline was indicated with a dotted line. Low levels of CD45 expression confirm the population of MuSCs used for analysis was indeed composed purely of muscle-derived cells and not contaminated with immune cells of hematopoietic origin (Testa et al., 2020). Additionally, no significance in the MyHC7 marker indicated that our LacZ + MuSCs were not undergoing further differentiation during proliferation, as indicated by their significantly high transcriptome level of Ki67 (****P ≤ 0.0001). Next, we assessed the LacZ + MuSCs ability to upregulate PD-L1 using increasing exposure time to 5% hypoxia (HYP) compared to normoxia (CTL). Western blot analysis indicated increasing PD-L1 protein expression up to 72 h (Figure 1B). To determine if LacZ + MuSCs were able to upregulate PD-L1 at an earlier time point during hypoxia conditioning, we compared PD-L1 protein expression levels after 6h and 12 h (Figure 1C). Semi-quantitative analysis of PD-L1 relative protein expression normalized to the β-actin control indicated that MuSC started upregulating PD-L1 expression as early as 12 h and peaked close to 72 h. Given these findings, all further hypoxia conditioning was done for a minimum of 72 h to maximize PD-L1 expression for further investigation. Finally, to determine the immunological capacities of hypoxia conditioned LacZ + MuSCs in vitro, we investigated the cell surface expression with flow cytometry of MHCI/II, Toll-like receptors 3 and 7 (TLR3, TLR7), ICAM, and PD-L1 after a 72 h treatment with normoxia (CTL), 5% hypoxia (HYP), IFN-γ (INF) alone, and INF with 5% hypoxia (INF + HYP) (Figures 1D, E). We showed that cell surface expression of MHC-I, MHC-II, TLR3, and TLR7 in MuSC was similar under normoxia and hypoxia conditions (Figure 1D). However, MHC-II was significantly upregulated with IFN-γ with and without hypoxia (Figure 1D). Furthermore, a calculated 7.1 and 9.5-averaged fold change of PD-L1 cell surface expression was seen with IFN-γ and IFN-γ plus hypoxia treated LacZ + MuSCs, respectively, in comparison to hypoxia treatment alone.

Given that we saw PD-L1 upregulation in LacZ + MuSCs after 72 h with 5% hypoxia conditioning, we sought to discover how long PD-L1 would stay upregulated after the conditioned cells were placed back into normoxia conditions (Figure 2). First, we confirmed that PD-L1 relative protein expression was upregulated (****P ≤ 0.0001) after the 72 h hypoxia conditioning (Post 0 h) in comparison to the baseline cells conditioned in normoxia for 72 h (CTL). Furthermore, a calculated 0.62 average fold increase in PD-L1 protein expression was seen with 72 h hypoxia conditioning compared to 72 h normoxia conditioned LacZ + MuSCs. Interestingly, PD-L1 expression increased and plateaued from 2 to 8 h under normoxia conditions (Post 2h, Post 4h, and Post 8 h) before decreasing to baseline levels by 24 h. Average fold change in PD-L1 expression was 2.5, 2.25, and 2.22 for Post 2h, Post 4h, and Post 8 h in comparison to the 72 h normoxia conditioned LacZ + MuSCs, respectively. This data indicated that LacZ + MuSCs were able to maintain hypoxia induced PD-L1 expression even under normoxia, mitigating the need for a continuous hypoxic environment to harvest the cells for cell transplantation procedures.

Infiltration of CD4⁺ and CD8⁺ T cells play a major role in acute allograft rejection, as well as increase memory cells associated with long-term allograft dysfunction (Yap et al., 2015). Thus, we evaluated the percent CD4⁺ and CD8⁺ cell infiltration in C57BL/J6 (6–9 months, male and female, n = 13) with both 72 h hypoxia conditioned and normoxia conditioned LacZ + MuSCs implanted separately into the left and right GMs of each mouse. These treated muscles were harvested for cell isolation 3, 5, and 7 days and stained for CD4⁺ and CD8⁺ cells for Flow Cytometry (Figures 3A, B). A rigorous gating strategy of SSC-W vs. SSC-H, FSCH vs. FSC-A, and SSC-A vs. FSC-A was used to isolate singlets in the population (Supplementary Figure S4). This same gating strategy was applied to CD4⁺ and CD8⁺ cells isolated from hypoxia conditioned and hypoxia conditioning with IFN-γ injected muscle samples (Supplementary Figure S4). A spleen sample was used to confirm CD4⁺ and CD8⁺ positive cell populations while unstained samples were used as negative controls. Our results suggest that CD4⁺ and CD8⁺ cell infiltration was decreased across 3d, 5d, and 7d with hypoxia (HYP CD4⁺ or CD8⁺) conditioned injected muscles versus matched, normoxia (CTL CD4⁺ or CD8⁺) conditioned injected muscles (Figure 3B). The table shows the mean, standard deviation (STDV), and p-value from two tailed student t-tests for each group. Additionally, Cohen’s d effect size was calculated as the difference between the mean of two groups and then divided by the standard deviation of the control group (Lakens, 2013). The commonly used interpretation was to refer to effect sizes as weak effect (d = 0–0.20), modest effect (d = 0.21–0.50), moderate effect (d = 0.50–1.00), and strong effect (d = >1.00) (Cohen, 2013). HYP CD8⁺ injected muscles showed a similar decrease in CD8⁺ T-cells across 3d, 5d, and 7d compared to matched CTL CD8⁺ injected muscles. The table showed the mean, standard deviation (STDV), and p-value from two tailed student t-tests for each group. Additionally, Cohen’s d effect size was calculated as the difference between the mean of two groups and then divided by the standard deviation of the control group. Overall, we saw a statistically significant (p < 0.05) decrease and a moderate Cohen’s effect in CD4⁺ and CD8⁺ cells 3d after injection compared to 5d and 7d post injection. Thus, any further experiments involving CD4⁺ and CD8⁺ infiltration were limited to 3d post injection.

We then explored whether 72 h hypoxia and IFN-γ conditioned LacZ + MuSCs implanted separately to the GMs would have any further effects in CD4⁺ and CD8⁺ infiltration after 3d compared to the matched GM implanted with 72 h hypoxia conditioned LacZ + MuSCs. This was based on our previous observation that showed significantly high surface expression levels of PD-L1 in LacZ + MuSCs when conditioned with hypoxia and 0.03 ng/μL IFN-γ for 72 h. C57BL/J6 mice, 3 months, male and female (n = 6) were used for this experiment.

We observed a similar range of percent lymphocytes of HYP CD4⁺ and CD8⁺ cells (Figure 3D) compared to the batch of 3–9-month-old mice (Figure 3B) indicating that the age of the mice did not impact the percent CD4⁺ and CD8⁺ lymphocytes present in the GMs. Interestingly, we also saw a further decrease in CD4⁺ and CD8⁺ lymphocyte infiltration with the hypoxia and IFN-γ injected GMs (HYP + INF-γ CD4⁺ and CD8⁺) in comparison to the matched hypoxia conditioned GMs (Figure 3D). More specifically, CD8⁺ cells were significantly decreased (p = 0.05) with injected hypoxia and IFN-γ conditioned LacZ + MuSCs compared to CD4⁺ cells. Cohen’s effect size (d) indicated that injected hypoxia and IFN-γ conditioned LacZ + MuSCs had a moderate effect in reducing CD4⁺ and CD8⁺ infiltration. These results added further evidence that the higher levels of PD-L1 upregulation seen after hypoxia and IFN-γ conditioning have increased effects in reducing T-cell infiltration after cell transplantation.

BMS-1 is considered a small molecule inhibitor of PD-L1 due to its ability to bind to PD-L1, thereby blocking its interaction with PD-1. This binding is thought to alleviate the inhibitory effect of soluble PD-L1 on T-cell receptor-mediated activation of T-lymphocytes (Skalniak et al., 2017). Since there were no references on BMS-1’s effect or toxicity on skeletal muscle stem/progenitor cells, we sought to confirm an optimal concentration to use to treat our LacZ + MuSCs. First, we tested the toxicity of increasing concentrations of BMS-1 after 24 h treatment on the LacZ + MuSCs using the MTT assay under normoxia conditions (Figure 4A). The results showed that BMS-1 concentrations above 1 μM had dose dependent negative effects on the percent viability of our LacZ + MuSCs. Thus, 1 μM BMS-1 was considered a safe, maximum dose and was used for all experiments from this point forward. A previous publication recommended a 72 h treatment on natural killer cells and HepG2 cells with BMS-1 to disturb the interaction between PD-1 and PD-L1 (Zhao et al., 2019). Given that the 72 h BMS-1 treatment aligned with our 72 h hypoxia conditioning, we decided to utilize 72 h as our minimum treatment time for both in vivo and in vitro experiments. Thus, we investigated the ability of BMS-1 to compete with the anti-PD-L1 IgG antibody through western blot analysis (Figure 4B). The results showed that BMS-1’s inhibitory capability was almost half in both normoxia and hypoxia conditioned LacZ + MuSCs after 72 h. The average fold change of PD-L1 protein expression in the presence of BMS-1 was 43% and 42% less after 72 h normoxia and hypoxia conditioning, respectively. To prove that BMS-1 does not affect PD-L1 transcriptome expression, we measured the cell surface PD-L1 receptor expression (medium fluorescent intensity) after 72 h treatment with BMS-1 in normoxia and hypoxia conditions using Flow Cytometry (Figure 4C). We noted that surface expression of PD-L1 in LacZ + MuSCs after hypoxia conditioning was similar to our previous results in Figure 1D. Additionally, the results showed no statistical differences in PD-L1 surface expression between cells treated and not treated with BMS-1 under normoxia and hypoxia conditions, implying BMS-1’s antagonistic effect on the PD-L1 receptors themselves and not the PD-L1 expression at the transcriptional and translational levels.

Finally, to determine if PD-L1 upregulation was a factor of reduced CD4⁺ and CD8⁺ T-cell infiltration after cell implantation to the GMs, LacZ + MuSCs were conditioned for 72 h with 5% hypoxia. Prior to injection to one side of the GM per mouse, 1 μM BMS-1 was added to 2 × 10⁶ of the hypoxia treated cells while the other GM of the same mouse received 2 × 10⁶ hypoxia treated cells to act as a control. MDX (3 months, male and female, n = 6) were used for this experiment. After 3d in vivo the GMs of matched mice were isolated to evaluate CD4⁺ and CD8⁺ infiltration with and without BMS-1 on LacZ + MuSCs conditioned with hypoxia. The same rigorous gating strategy of SSC-W vs. SSC-H, FSCH vs. FSC-A, and SSC-A vs. FSC-A was used to isolate CD4⁺ and CD8⁺ singlets from each GM sample as mentioned before (Supplementary Figure S6). A spleen sample was used to confirm CD4 and CD8 positive cell populations while unstained samples were used as negative controls. We showed that the addition of 1 μM BMS-1 increased the presence of CD4⁺ and CD8⁺ cells in the GM 3d post injection (Figure 4D). Furthermore, CD4⁺ cells were significantly increased in the GM (p = 0.05) with the addition of BMS-1 compared to GMs injected with hypoxia conditioned LacZ + MuSCs alone. Finally, Cohen’s effect size (d) indicated a strong effect in the percent population of CD4⁺ and CD8⁺ cells in GMs injected with the BMS-1 small molecule inhibitor.

To estimate the ability of skeletal muscle stem/progenitor cell survival after cell transplantation, LacZ + MuSCs were conditioned with 5% hypoxia for 72 h. The same number (2 × 10⁶) of both LacZ + MuSCs were implanted separately into the GM, except that one set of cells was mixed with 1 μM of BMS-1 prior to the injection. MDX (2–3 months, male and female, n = 5) mice were used for this experiment. These muscles were then harvested for cell isolation 4 days after injection and subsequently evaluated for LacZ + cell survival by counting dark blue nuclei (Figure 4D). Counting was performed by capturing five microscopic images taken from random portions of each cell culture plate well. Our results showed that the hypoxia-conditioned MuSCs had an improved survival rate compared to the MuSCs injected with BMS-1 as indicated by a p-value = 0.04 using one-tailed t-test (Figure 4E).