Amelioration of muscular dystrophy phenotype in mdx mice by inhibition of Flt1

Duchenne muscular dystrophy (DMD) is an X-linked recessive genetic disease in which the dystrophin coding for a membrane stabilizing protein is mutated. Recently, the vasculature has also shown to be perturbed in DMD and DMD model mdx mice. Data-mining DMD transcriptomics revealed the defects were correlated to a vascular endothelial growth factor (VEGF) signaling pathway. To reveal the relationship between DMD and VEGF signaling, mdx mice were crossed with constitutive (CAG/CreERTM:Flt1LoxP/LoxP) and endothelial cell-specific conditional gene knockout mice (Cdh5CreERT2:Flt1LoxP/LoxP) for Flt1 which is a decoy receptor for VEGF. Previous work demonstrated that heterozygous global Flt1 knockout mice increased vascular density and improved DMD phenotypes when crossed with DMD model mdx and mdx:utrn-/- mice. Here, we showed that while constitutive deletion of Flt1 is detrimental to the skeletal muscle function, endothelial cell-specific Flt1 deletion resulted in increased vascular density and improvement in the DMD-associated phenotype in the mdx mice. These decreases in pathology, including improved muscle histology and function, were recapitulated in mdx mice given anti-FLT1 peptides or monoclonal antibodies, which blocked VEGF-FLT1 binding. The histological and functional improvement of dystrophic muscle by FLT1 blockade provides a novel pharmacological strategy for the potential treatment of DMD.


Introduction
Duchenne muscular dystrophy (DMD) is an X-linked muscle disease affecting one in 5,000 newborn males, in which the gene encoding the dystrophin protein is mutated. It is a progressive neurodegenerative disease with clinical symptoms manifesting at 2-3 years of age, loss of ambulation in early teen years and death by either respiratory insufficiency or cardiac failure in their 20s. A disease model for DMD is the mdx mouse, which lacks functional dystrophin expression due to a point mutation in the dystrophin gene. The mdx mouse has been extensively characterized and contributed to the understanding of the disease pathology (1).
Although the role of dystrophin in the skeletal muscle is widely appreciated, endothelium and vascular smooth muscle cells also express dystrophin (2). The absence of dystrophin in these cells induced vessel dilation and abnormal blood flow, resulting in a state of functional ischemia, worsening the muscle pathology in mdx mice (3). Restoration of dystrophin specifically in the smooth muscle of the vasculature rescued some aspects of the skeletal muscle pathology associated with the mdx mice (4). Disruption of the dystrophin-associated sarcoglycan complex in vascular smooth muscle perturbed vascular function resulting in exacerbation of muscular dystrophic changes (5). Dystrophin is responsible for anchoring neuronal nitric oxide synthase (nNOS) to the cell surface, which is crucial for exercise-induced increases in blood supply in muscle via NOmediated vasodilation (6). Administration of a phosphodiesterase-5  inhibitors to mdx mice, which increased NO production, rescued the muscle from this state of functional ischemia, and improved muscle function in mdx mice (7,8). Similarly in humans, PDE-5 inhibitors given to both DMD boys and adult patients with Becker muscular dystrophy (BMD), a milder form of muscular dystrophy, alleviated functional ischemia during muscle contraction (9,10) . More recent data showed that mdx skeletal muscle was less perfused and displayed marked microvessel 4 alterations compared to wild-type C57BL6 mice (11,12). While current studies support the importance of NO-mediated vasodilation in DMD, the relationship between DMD and angiogenesis is not well understood.
Vascular endothelial growth factor (VEGF) signaling is one of the strongest modulators of angiogenesis and includes the ligands VEGFA, VEGFB, VEGFC and PlGF. VEGFA is the most well studied ligand of the system and acts through its two receptors, VEGF receptor-1 (VEGFR1/FLT1) and VEGF receptor-2 (VEGFR2/FLK1/KDR). Although FLK1 possesses stronger signaling capabilities, FLT1 has considerably higher affinity for VEGF but weaker signaling capabilities. In normal tissue, FLT1 acts as a sink trap for VEGF thereby preventing excessive pathological angiogenesis. In addition, soluble FLT1 (sFLT1) functions as an endogenous VEGF trap (13). Despite the known angiogenic defect in DMD and mdx mice, it is not known whether VEGF and its receptors are implicated in this disease process. Previous data from our laboratory demonstrated that heterozygous Flt1 gene knockout (Flt1 +/-) mice were viable and displayed developmentally increased capillary density in the skeletal muscles (14).
Importantly, when crossed Flt1 +/with mdx or mdx:utrn -/mice, these mice displayed both histological and functional improvements of the dystrophic pathologic phenotype. However, it remained unknown whether postnatal Flt1 gene deletion and pharmacological blockage of FLT1 could recapitulate these improvements in mdx mice.
In this report, we compared adult mdx mice with a constitutive conditional knockout and an endothelial cell-specific conditional knockout of Flt1. We showed that endothelial cell-specific Flt1 deletion increased the capillary density in skeletal muscle and improved the DMD-associated muscle pathology. In addition, we showed that intravenous administration of anti-FLT1 peptides and monoclonal antibodies (MAbs) in mdx mice recapitulated the reduction in DMD-associated 5 pathology seen after Flt1 deletion in mdx mice, validating Flt1 as a therapeutic target for the treatment of DMD.

Postnatal Flt1 gene deletion in mice display increase in capillary density
We previously found that mdx mice developmentally lacking one copy of the Flt1 allele have increased muscle angiogenesis and improved muscle pathology (14). To investigate whether postnatal deletion of Flt1 gene could affect the vasculature density, we crossed CAG CreERTM mice carrying a constitutively expressed CreER TM gene (15) with Flt1 LoxP/LoxP mice (16) to generate conditional Flt1 gene knockout (CAG CreERTM :Flt1 LoxP/LoxP ) mice ( Figure 1A). Upon treatment with tamoxifen (TMX), which leads to global Flt1 gene and FLT1 protein deletion (Supplemental Figure 1A, B), Flt1 Δ/Δ mice displayed significantly increased CD31+ vascular density compared to the Flt1 +/+ (Flt1 LoxP/LoxP ) mice ( Figure 1B, C). The increase in capillary density following TMXmediated Flt1 gene deletion was rapid (within 8 days) and long lasting (more than 207 days) ( Figure 1D). This allowed us to be confident that we were able to phenotype late term changes following deletion of Flt1 gene. Postnatal global loss of Flt1 resulted in a reduction in body mass (17) without reduction of tibialis anterior (TA) muscle mass (Supplemental Figure 1C, D). mdx:Flt1 Δ/Δ mice display worse muscle pathology compared to control mdx:Flt1 +/+ mice As we did not see any gross changes in the skeletal muscle except for increased vascular density in the Flt1 Δ/Δ mice, we crossed the mdx mice to the CAG CreERTM :Flt1 LoxP/LoxP to obtain mdx:CAG CreERTM :Flt1 LoxP/LoxP mice. We obtained these mice in expected mendelian ratios (Supplemental Figure 2A). Our original goal was to induce Flt1 gene deletion prior to the onset of muscle pathology, thus before postnatal day 21 (p21), by treatment with TMX or its active form, 4-hydroxy tamoxifen (4-OHT). However, perinatal loss of Flt1 resulted in lethality when TMX or 4-OHT treatment was initiated at p3 or p5 and partial lethality at p16 in 7 mdx:CAG CreERTM :Flt1 LoxP/LoxP (mdx:Flt1 Δ/Δ ) mice (Supplemental Figure 2B), indicating that Flt1 is required in the perinatal stage for survival. The mice displayed no lethality when recombination was induced on or after p21. Importantly, the increase in capillary density by loss of the Flt1 gene ( Figure 1A) was maintained in the mdx background in the mdx:Flt1 Δ/Δ mice ( Figure 2A). This was accompanied by a physiological increase in skeletal muscle perfusion as shown by laser Doppler imaging ( Figure 2B). The mdx:Flt1 Δ/Δ mice showed a shift in fiber type composition toward increases in oxidative type I fibers (Supplemental Figure 3A, B). This was more pronounced in the EDL compared to the soleus, which is already predominantly type I.
Endothelial cell-specific loss of Flt1 in mdx improves muscle phenotype in mdx mice.
Flt1 is expressed in several cell types including endothelial cells, myeloid cells and some neurons.
Thus, we hypothesized that Flt1 may be indispensable in one of these other compartments. Since endothelial cell-specific Flt1 deletion resulted in increased capillary density in heart and adipose 8 tissue (16,22), we hypothesized that deletion of endothelial cell-specific Flt1 would be sufficient to increase angiogenesis and improve muscle pathology in the mdx mice.
We attempted to increase the angiogenesis in skeletal muscle using an endothelial cellspecific VE-cadherin (Cdh5)-CreERT2-mediated Flt1 deletion in mice. Cdh5 is an endothelial cell-specific cadherin gene used for lineage tracing and conditional deletion of endothelial cells (23). Goel. et al. recently reported the presence of Cdh5 in satellite cells questioning the validity of using Cdh5 CreERT2 in skeletal muscle tissue (24). We verified the endothelial cell specificity of the Rosa26R mTmG reporter, and Cre-mediated excision resulted in the mGFP expression in the endothelial cells but no other cell types (Supplemental Figure 4A, B, C). We confirmed that the Cdh5 CreERT2 was not present in the Pax7+ satellite cells using single muscle fiber immunostaining (data not shown). We saw no difference in the body mass or muscle mass in endothelial cellspecific Flt1 deleted mice compared with the control mice (Supplemental Figure 4D, E).
We crossed the mdx:Flt1 LoxP/LoxP mice to the Cdh5 CreERT2 mice to yield the mdx:Cdh5-Flt1 Δ/Δ mice ( Figure 3A). Upon TMX treatment, capillary density and laser Doppler flow were increased in the skeletal muscle in mdx:Cdh5-Flt1 Δ/Δ mice compared with mdx:Cdh5-Flt1 +/+ or mdx:Cdh5-Flt1 +/Δ mice, indicating that endothelial cell-specific deletion of Flt1 was sufficient to increase capillary density in the skeletal muscle ( Figure 3B, C). This was accompanied by a physiological increase in skeletal muscle perfusion using laser Doppler ( Figure 3D). Moreover, the mdx:Cdh5-Flt1 Δ/Δ did not show any significant changes in body and muscle mass loss (Supplemental Figure 5). Signs of DMD-associated pathology such as increased EBD uptake and fibrosis were significantly reduced in the mdx: E,F). The muscle fibers the mdx:Cdh5-Flt1 Δ/Δ mice had decreased centrally located nuclei and maintained larger myofibers fibers compared with the mdx:Cdh5-Flt1 +/+ mice ( Figure 4A, B, C). The mdx:Cdh5-Flt1 Δ/Δ mice showed increased grip strength compared with the mdx:Cdh5-Flt1 +/+ mice ( Figure   4D). Taken together, these data indicate that endothelial cell-specific Flt1 loss was sufficient to increase capillary density and result in the histological improvements correlated with a functional improvement in the mdx mice.

Pharmacological inhibition of FLT1 improved mdx mice.
The genetic model of Flt1 deletion showed an ameliorated phenotype in the mdx mice. To translate our genetic results into therapeutic approaches for DMD model mice, we utilized a previously reported anti-FLT1 hexapeptide (Gly-Asn-Gln-Trp-Phe-Ile or GNQWFI) that inhibits VEGFbinding to FLT1. Intramuscular administration of the anti-FLT1 peptide in TA muscle of perinatal mdx mice (Supplemental Figure 6A) increased capillary density and decreased muscle pathology in the treated muscle (Supplemental Figure 6B-D). We assessed the diaphragm muscle after systemic (IP) injection of the anti-FLT1 peptide at a low (10 mg/kg body weight) and a high dose (100 mg/kg body weight) to test for therapeutic potential in mdx mice ( Figure 5A). While treatment with a low dose of anti-FLT1 peptide had no effect on capillary density, skeletal muscle perfusion, membrane permeability or fibrosis, the high dose increased capillary density, increased skeletal muscle perfusion, decreased EBD+ fibers, and decreased fibrosis ( Figure 5B-F). Consequently, anti-FLT1 peptide-treated mdx mice increased grip strength compared with the control mdx mice ( Figure 5F, G). There was no significant body mass alteration following anti-FLT1 peptide treatment (Supplemental Figure 6E, F). To increase stability and the hydrophobicity (25), we tested a D-isoform anti-FLT1 peptide attached to polyethylene glycol (PEG). However, systemic administration of the PEG-D-form anti-FLT1 peptide did not increase capillary density or improve muscle pathology or function in the mdx mice (Supplemental Figure 7). Taken together, these proof-of-concept experiments showed that postnatal inhibition of FLT1 by anti-FLT1 peptide could ameliorate the pathology associated with DMD in the mdx mice. However, the functional dose (100 mg/kg body weight) was orders of magnitude higher than the generally acceptable pharmacological standards of body weight dosage for small molecule drugs (26). A more potent or alternative strategy is required for further translational studies.

Screening for antibody against FLT1
As postnatal gene deletion and pharmacological inhibition of FLT1 decreased the muscular dystrophy-associated pathology in the mdx mice, we next sought to examine whether it could do this in a more translational manner using biologics to block FLT1. To establish proof of principle, we screened 8 commercially available MAbs that could block VEGF-FLT1 binding. We first screened for the ability of the MAbs to block chimeric FLT1-FC binding to PLGF2, a VEGF family protein, using ELISA (Supplemental Figure 8A, B), since both PLGF2 and VEGFA occupy the same binding sites on the extracellular domain of FLT1 (27). Three MAbs showed higher binding affinities: EWC (Novus Biologicals), MAB0702 Mab (Angio-Proteomie) and EWC (Acris GmbH) blocked binding by 65.4%, 64.8%, and 60.6%, respectively, while the control polyclonal anti-FLT1 antibody could block 83.1% of binding (Supplemental Figure 8B). We selected two MAbs (MAB0702 from Angio-Proteomie and EWC from Novus Biologicals) for further analyses based on their blocking efficiency and their availability for large in vivo studies. The MAB0702 and EWC were raised against the extracellular domains of the human FLT1, and likely target both membrane-bound (mFLT1) and sFLT1 due to similar homology of epitopes. We validated the blocking affinity to FLT1-VEGFA binding, and found that MAB0702 and EWC blocked binding by 40.1% and 19.9%, respectively, compared to the 94.3% for the polyclonal control (Supplemental Figure 8B). We further determined the antibodies' affinities against mouse and human FLT1-FC protein using Biacore analysis (Supplemental Table 1). MAB0702 and EWC had similar association rate/binding constants for both mouse and human FLT1-FC. By contrast, MAB0702 had significantly higher dissociation constants for mouse FLT1-FC compared with EWC, while MAB0702 had lower dissociation constants for human FLT1-FC compared with EWC. Based on these affinity studies, we decided to utilize both MAB0702 and EWC MAbs for in vivo experiments.

Testing Anti-FLT1 antibody treatment for DMD
We IV injected MAB0702 at a dose of 20 mg/kg body weight into mdx mice, and measured free sFLT1 and VEGFA in the serum. We found a significant decrease in free serum sFLT1 (Supplemental Figure 8C) and an increase in serum VEGFA levels (Supplemental Figure 8D) following MAB0702 injection, suggesting the efficient blocking of MAB0702 to sFLT1 and FLT1 in vivo. Based on our screening results, mdx mice were injected with IV MAbs or isotype control IgG dosing at 2 mg/kg or 20 mg/kg body weight every 3 days for four weeks beginning at 3 weeks of age ( Figure 6A). While capillary density was not altered by treatment with either MAB0702 or EWC at 2 mg/kg body weight (data not shown), the treatment with MAB0702 at 20 mg/kg body weight but not EWC or isotype control significantly increased capillary density and skeletal muscle  Table 1), we were surprised to find that in vivo administration of the EWC did not induce changes in capillary density, skeletal muscle perfusion or muscle pathology ( Figures 6B, C-G, 7A, B). MAB0702-treated mdx mice generated increased grip strength compared with the mdx mice ( Figure 7D). Skeletal muscle endurance, as assessed by treadmill running as an indicator of maximal muscle capacity, showed running duration and distance of MAB0702-treated mdx mice significantly increased compared to mdx mice ( Figure 7E, F). Thus, anti-FLT1 antibody administration depleted free serum FLT1 levels and increased free serum VEGFA levels, which led to increased angiogenesis and reduced muscle pathology in mdx mice, providing a potential new pharmacological strategy for treatment of DMD.

VEGFA signaling is perturbed in DMD animal models and DMD patients.
While angiogenic defects have been reported in the mdx mice, it is not known whether VEGF family and its downstream targets are implicated in dystrophinopathies. We probed the VEGF ligands and receptors in microarrays from skeletal muscles from mdx mice and the golden retriever muscular dystrophy (GRMD) canine model of DMD. VEGFA was downregulated in both models ( Figure 8A). Flt1 was downregulated in GRMD but not mdx muscles. To examine whether VEGF signaling is altered in DMD patients, we performed gene expression analysis on previously available data from microarrays and RNA-seq from patients with DMD. We also aggregated and probed microarray data from muscle biopsies of patients with various neuromuscular diseases or of healthy individuals after exercise. In the microarray data, VEGFA expression was increased after an acute bout of exercise, and VEGFA expression was reduced in ALS muscle, BMD muscle, as well as both early and late phases of DMD muscle (28) (Figure 8B). This was corroborated by 13 RNA-seq data ( Figure 8C). Angiogenic genes downstream of VEGF, such HRAS, KRAS and NRAS, were downregulated in DMD and BMD muscle despite an increase in VEGF pathways such as HIF1A ( Figure 8D). By contrast, the expression of VEGF receptors (Flt1 and Flk1) and the coreceptors (Nrp1 and Nrp2) was not significantly changed in DMD muscle ( Figure 8B, C). These data indicate that VEGFA expression is decreased in dystrophinopathy, and thus may benefit people with DMD by either increasing VEGFA and/or decreasing sFLT1 as a therapeutic target.

Discussion
In this report, we show that Flt1 is important postnatally as conditional Flt1 deletion in mdx:Flt1 Δ/Δ mice in the perinatal stage results in lethality in mice. While deletion of Flt1 in neonates increased capillary density in both C57Bl6 and mdx mice, it led to the worsening of the skeletal muscle phenotype. Flt1 is expressed in several cell types including endothelial cells, myeloid cells and some neurons (23,(29)(30)(31). Thus, Flt1 may be indispensable in one of these other compartments when perinatally deleted. For example, Flt1 is expressed in motor neurons, where is not simply acting as a VEGFA sink-trap. In motor neurons, its tyrosine kinase activity it is responsible for their survival (29). Thus, the observed increased angiogenesis but worse pathological alterations in mdx:Flt1 Δ/Δ muscle may be due to motor neuron-associated changes. By contrast, loss of Flt1 in endothelial cells in the postnatal stage increased the capillary density and blood perfusion in skeletal muscle without significant decreases in weight or muscle mass. Increased angiogenesis and blood perfusion was observed in the global Flt1 heterozygous knockout mice (14), indicating that loss of endothelial cell-specific Flt1 in mdx:Cdh5-Flt1 Δ/Δ mice was sufficient to produce increased capillary density and vascular perfusion in the skeletal muscle. This also led to an improved mdx-associate muscle pathology, confirming that postnatal deletion of Flt1 is able to rescue the dystrophinopathy related muscle pathology.
Importantly, for the first time, we demonstrated that administration of both anti-FLT1 peptide and anti-FLT1 MAb increased angiogenesis, which led to an improved the pathology associated with DMD in mdx mice, and is a phenocopy of our genetic models (mdx:Cdh5-Flt1 Δ/Δ mice). We screened commercially available MAbs for blocking ligands-FLT1, and demonstrated that administration of MAB0702 MAbs was able to phenocopy our genetic model in a manner suited for translational studies to reduce muscle pathology in mdx mice. Improvement of dystrophic muscle function by FLT1 blockade may provide a novel pharmacological strategy for treatment of diseases associated with DMD via increased serum and tissue VEGFA levels, which induce increased vascular density and blood perfusion.
After MAB0702 administration, we showed a small increase in VEGFA levels in serum. It should be noted that a mere 2-fold increase of VEGFA during development is incompatible with life in transgenic mice (32). Computation and experimental models showed that local VEGF gradients are more important than the total concentration (33)(34)(35)(36). We recently demonstrated that muscle satellite cells express abundant VEGFA, which recruits endothelial cells and capillaries to the proximity of satellite cells on muscle fibers (37). Taken together, these data strongly suggest that hot-spots with high levels of VEGFA in Flt1 knockout or anti-FLT1 treated mice may result in increased capillary density and vascular perfusion, which is predicted to result in decreased associated DMD-type pathological changes in the skeletal muscles.
While VEGFA binds to both FLT1 and FLK1, VEGFB, PlGF1 and PlGF2 only bind to FLT1 (38). This creates a scenario where PlGF1/2 and VEGFB binding can sequester FLT1, increasing serum and tissue VEGFA availability for VEGFA-FLK1 binding-mediated angiogenic induction. While PlGF is dispensable for normal development and health and not expressed in the normal adult tissues outside the female reproductive organs (39), VEGFB is expressed in the muscle tissue and muscle fibers (40). VEGFB may play a role in diet induced obesity and free fatty uptake by the endothelial cells in skeletal muscle (41), but this remains controversial (22). However, VEGFB overexpression does not result in an angiogenic response in ischemic skeletal muscle (42), indicating that blocking VEGFB-FLT1 signaling is not likely to be responsible for the angiogenic changes seen in this study.

16
The present study shows that increased angiogenesis may be a novel avenue to improve some of pathology associated with loss of dystrophin, including function, and could be used in conjunction with other treatment strategies. For example, intramuscular injection of VEGF containing recombinant adeno-associated viral vectors resulted in both functional and histological improvements in both ischemic and mdx muscle (19,49,50). This improvement was seen in combination with increased in angiogenesis in the muscle. Increased capillary density in the mdx muscle work through paracrine stimulation, by protecting muscle fiber damage and promoting satellite cell proliferation, survival and self-renewal in the vascular niche (37,43). Christov et al. and our group showed that satellite cells were preferentially located next to capillaries (37,44).
Since our approaches increased the number of capillaries found in the examined muscles, it may effectively increase the amount of vascular niche that houses the satellite cell compartment in the muscle. Recently, we showed that genes encoding for endothelial and satellite cells were highly correlated across muscle groups, and endothelial cells could mediate satellite cell self-renewal via Notch activation (37) .Therefore, increase in the vascular niche may increase satellite cell selfrenewal and the number of myogenic precursor cells, which we hypothesize may be responsible for the improved phenotype seen in the mdx mice with Flt1 deletion or functional blockage.
MAb-based therapeutics for the treatment of DMD have been developed to increase muscle mass via targeting myostatin and to decrease fibrosis via targeting fibroadipogenic progenitors (45). Further optimization using with humanized antibodies are required for future translation to humans (46). Taken together, we have gathered evidence for the first time that FLT1-targeted MAbs may be an effective therapeutic approach for the treatment of DMD.
Cre recombination was induced using tamoxifen (TMX) (T5648, MilliporeSigma) dosed as 75 mg/kg body weight x 3 time over one week at 3-4 weeks of age unless otherwise specified. We also injected 4-hydroxy tamoxifen (4-OHT) (H6278, MilliporeSigma) dosed as 25 mg/kg body weight. Control mice contained the wild-type (WT) CreER allele or were injected with the vehicle (corn oil or 10% ethanol). All animal studies were approved by the IACUC at University of Minnesota.

Anti-FLT1 peptides and antibodies
The anti-FLT1 peptide was synthesized from Peptide 2.0 Inc. based on the sequence (Gly-Asn-Gln-Trp-Phe-Ile or GNQWFI) as previously described (49). DMSO was used to dissolve the peptide. Twenty µg of peptide diluted in 2% DMSO in PBS solution for intramuscular injection per day in the TA muscle. Ten mg/kg body weight and 100 mg/kg body weight were used for the systemic treatment. The second generation anti-FLT1 peptide (PEG-G D N D Q D W D F D I D ) was synthesized with the following modifications: The polyethylene glycol moiety was attached to the peptide to improve solubility in polar solvents and the D isomeric form was used instead of the L to enhance stability of the peptide (25). PBS was used as a vehicle. The peptides were commercially synthesized (LifeTein, LLC). Commercially available anti-FLT1 antibodies were obtained from the manufacturer in carrier free and preservative free form (AF471 from R&D Systems, D2 from Santa Cruz Biotechnology, EWC from Acris GmbH and Novus Biologicals, LS-C6855 from LifeSpan BioSciences, MAB1664 from MilliporeSigma and MAB7072 from Angio-Proteomie, Abcam and Santa Cruz Biotechnology). Isotype IgG (Santa Cruz Biotechnology) was used for control experiment. Two or 20 mg/kg body weight was used for the systemic treatment. Retro-orbital IV injections were performed for systemic treatment for both the peptides and the antibody treatment.

RNA and genomic DNA isolation and qPCR
Mouse TA muscle was homogenized in TRIzol TM reagent (15596026, ThermoFisher Scientific) for RNA isolation. RNA was isolated using the Direct-zol TM RNA Microprep Kit (R2062, Zymo Research) with on-column DNase digestion followed by cDNA synthesis using the Transcriptor First Strand cDNA synthesis kit (04379012001, Roche Molecular Diagnostics) using random primers. Genomic DNA was isolated from mouse tail snips with lysis buffer containing Protenase K (P2308, MilliporeSigma). Genotyping was performed by agarose gel electrophoresis-mediated detection following PCR reaction by Taq polymerase (M0273, New England Biolabs). qPCR was performed using GoTaq® qPCR Master Mix (A6001, Promega). Primer sequences are listed in Table S1. All primers were synthesized as custom DNA oligos from Integrated DNA technologies (IDT).

Muscle perfusion
RBC flux was evaluated using the moorLabTM laser Doppler flow meter as previously described (14). with the MP7a probe that allows for collecting light from a deeper tissue level than standard probes according to the manufacturer's instructions (Moor Instruments). The fur from the right hind leg was removed using a chemical depilatory. Readings were taken using the probe from at least 10 different spots on the TA muscle. The AU was determined as the average AU value during a plateau phase of each measurement.

Grip strength test
Forelimb grip strength test was performed following a previously published procedure (50).
Briefly, mdx mice were gently pulled by the tail after fore limb-grasping a metal bar attached to a force transducer (Columbus Instruments). Grip strength tests were performed by the same blinded examiner. Five consecutive grip strength tests were recorded, and then mice were returned to the cage for a resting period of 20 minutes. Then, three series of pulls were performed each followed by 20 min resting period. The average of the three highest values out of the 15 values collected was normalized to the body weight for comparison.

Treadmill running
Exer-3/6 Treadmill (Columbus Instruments) was used for treadmill running test as previously described (51). Briefly, for acclimation, mice were placed in each lane and forced to run on a treadmill for 5 minutes at a speed of 10 m/min on a 0% uphill grade for 3 days. And then, mice were forced to run on a treadmill with a 10% uphill grade starting at a speed of 10 m/min for 5 minutes. Every subsequent 2 minutes, the speed was increased by 2 m/min until the mice 20 underwent exhaustion which was defined as the inability of the mice to remain on the treadmill.
The time of running as well as the distance run were recorded.

Biacore Surface Plasmon Resonance (SPR) Binding Assay for Anti-FLT1 MAbs
The single cycle kinetics method was used for sFLT1 binding assay by Biacore (GE Healthcare Bio-Science). A CM5 series S sensor chip (GE Healthcare Bio-Science) with mouse and human soluble FLT1-FC chimeric protein (471-F1-100 and 321-FL-050/CF, R&D Systems) immobilized to about 1,000 RU was used as ligand for the analyte binding of anti-FLT1 MAbs. An analyte range of 0-5 nM was used for kinetic experiments. A 5 min association step was used for each dilution followed by a 40 min dissociation. Chip surface was regenerated using pH 2.0 glycine between experiments.
Blood was collected for biomaker analysis. Serum concentrations of sFLT1 and VEGFA were measured by ELSA kits following company recommended protocols (DY471 and DY493, R&D Systems). For a colormetric detection, TMB was used as described above.

Microarray and RNA-seq analysis
Microarray analysis was performed using the Affymetrix Transcriptome Analysis Console (TAC).
Samples in each experiment were RMA normalized and the expression was acquired using the Affeymetrix Expression analysis console with gene level expression. Heatmaps were generated in the Graphpad 7.1 (Prism). The code for generating each graph is listed in the following table, along with the link to the data in tabular format. All the data was obtained from NCBI GEO: Exercise, ALS, DMD, BMD, FSHD GSE3307, Early DMD GSE465, mdx GSE466, GRMD GSE69040, Satellite cells GSE15155. All arrays were normalized to their respective controls.

Statistics
Statistics and graphs were calculated using Graphpad 7.1(Prism). Students t-test or ANOVA was used to compare two or more groups. Multiple comparison adjustment was performed with comparisons of 3 groups or more. * indicates p<0.05, ** indicates p<0.01, *** indicates p<0.001, *** indicates p<0.0001. Figure 1: Postnatal deletion of Flt1 can increase capillary density in skeletal muscle.

Figures legends
A. Experimental scheme for assessing angiogenic response from conditional Flt1 deletion.
B. Representative images of CD31-stained cryosections from the skeletal muscle from      C. Anti-FLT1 peptide injection increases capillary density in the mdx mouse muscle at high dose.
D. Anti-FLT1 peptide injection is sufficient to increase skeletal muscle perfusion in mdx mice at high dose.
E. Anti-FLT1 peptide injection decreases EBD+ area in the mdx mouse muscle at high dose.
F. Anti-FLT1 peptide injection decreases fibrotic area in the mdx mouse muscle at high dose.
G. Grip strength is improved by anti-FLT1 peptide injection at high dose in mdx mice normalized to body weight. C. Capillary destiny is increased in MAB0702 antibody treated mice but not EWC antibody in mdx mice compared with isotype control. D. MAB0702 but not EWC injection is sufficient to increase skeletal muscle perfusion in mdx mice.
E. EBD+ area is decreased in MAB0702 antibody treated mice but not EWC antibody in mdx mice compared with isotype control.

33
F. Fibrotic area is decreased in MAB0702 antibody treated mice but not EWC antibody in mdx mice compared with isotype control.
G. Calcification is decreased in MAB0702 antibody treated mice but not EWC antibody in mdx mice compared with isotype control.