The α7β1-integrin is an adhesion molecule highly expressed in skeletal muscle that can enhance regeneration in response to eccentric exercise. We have demonstrated that mesenchymal stem cells (MSCs), predominantly pericytes, accumulate in muscle (mMSCs) overexpressing the α7B-integrin (MCK:α7B; α7Tg) and contribute to new fiber formation following exercise. Since vascularization is a common event that supports tissue remodeling, we hypothesized that the α7-integrin and/or mMSCs may stimulate vessel growth following eccentric exercise. Wild-type (WT) and α7Tg mice were subjected to single or multiple (3 times/wk, 4 wk) bouts of downhill running exercise. Additionally, 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI) -labeled mMSCs were intramuscularly injected into WT recipients. A subset of recipient mice were run downhill before injection to recapitulate the exercised microenvironment. While total number of CD31+ vessels declined in both WT and α7Tg muscle following a single bout of exercise, the number of larger CD31+ vessels with a visible lumen was preferentially increased in α7Tg mice following eccentric exercise training (P < 0.05). mMSC transplantation similarly increased vessel diameter and the total number of neuron-glial antigen-2 (NG2+) arterioles postexercise. Secretion of arteriogenic factors from mMSCs in response to mechanical strain, including epidermal growth factor and granulocyte macrophage-colony stimulating factor, may account for vessel remodeling. In conclusion, this study demonstrates that the α7-integrin and mMSCs contribute to increased vessel diameter size and arteriolar density in muscle in response to eccentric exercise. The information in this study has implications for the therapeutic treatment of injured muscle and disorders that result in vessel occlusion, including peripheral artery disease.
integrins are transmembrane adhesion proteins comprised of noncovalently bound α- and β-subunits. Integrins form clusters and recruit cytoskeletal and cytoplasmic proteins to the cell membrane to provide a mechanism of communication between the outside and inside of the cell (10). In skeletal muscle, the α7β1-integrin is the predominant heterodimer based on high levels of protein expression, and its primary function is to ensure adhesion by providing a stabilizing link between the actin cytoskeleton and the external environment, specifically, laminin (6). Transgenic expression of the integrin in mouse muscle can suppress macrophage infiltration, sarcolemmal damage, and reductions in force following eccentric exercise (3, 4, 13). Our recently published work also suggests α7BX2-integrin overexpression (MCK:α7BX2; α7Tg) can increase myogenesis, hypertrophic signaling, myofibrillar content, fiber cross-sectional area, muscle cross-sectional area, and maximal isometric force following single or multiple bouts of eccentric exercise (13, 34), a type of exercise that results in muscle lengthening during load and ultrastructural damage in the form of sarcomere disruption and sarcolemma damage (23). The precise cellular mechanisms underlying enhanced myogenesis and growth in muscle overexpressing the α7-integrin with eccentric exercise are not completely understood.
Multipotent mesenchymal stem cells (MSCs) reside within the perivascular niche within a variety of tissues, directly repairing injured tissue or indirectly, facilitating regeneration by secreting a repertoire of cytokines and growth factors that can stimulate resident progenitor cells (5, 15). We have recently established that MSCs accumulate in muscle (mMSCs) in a α7-integrin-dependent manner following eccentric exercise (29). The enhanced appearance of mMSCs in skeletal muscle expressing high levels of α7-integrin does not seem to depend on migration of the cells from a distal location but rather due to proliferation and/or dedifferentiation of preexisting resident cells as a result of increased sensitivity to mechanical strain. We have thoroughly characterized mMSCs isolated from α7Tg muscle as pericytes (Sca-1+CD146+NG2+PDGFRβ+CD45−CD56−CD34−CD31−) that facilitate new fiber formation, likely as a result of paracrine factor release and satellite cell activation (29). Thus these data demonstrate that an accumulation of mMSCs in muscle may provide the basis for accelerated myogenesis observed in α7Tg mice posteccentric exercise. Since vascularization is essential for optimal tissue healing and growth and there is some suggestion that pericytes can also regulate angiogenesis (7, 18, 28), we questioned whether the α7-integrin and/or mMSCs increase capillary number following eccentric exercise.
Sprouting angiogenesis is a well-described adaptation that occurs in skeletal muscle in response to endurance exercise training, a response necessary to facilitate the delivery of oxygen and nutrients necessary to maintain long-duration exercise (21, 32). Current studies evaluating the angiogenic response to endurance training have focused on molecular mechanisms, including increased release of vascular endothelial growth factor (VEGF) from contracting muscle fibers (20) and decreased release of anti-angiogenic factors (21). In contrast, minimal information exists regarding the structural changes that occur within the capillary bed of the muscle following eccentric contractions, commonly performed during resistance training. Mechanical strain, damage, and/or increased muscle fiber size associated with eccentric exercise may necessitate remodeling of the vasculature for optimal repair and support of muscle tissue.
Only one study to our knowledge has assessed the effect of eccentric contractions on capillary growth and structure. In this study, eccentric contractions of the rat gastrocnemius muscle elicited by electrical stimulation resulted in no change in capillary-to-fiber ratio but increased capillary density (7 days poststimulus) (12). Increases in capillary luminal area and distensibility (1 and 3 days poststimulus) were also observed, suggesting that vascular adaptations other than angiogenesis may occur following eccentric exercise. Chronic mechanical loading models have demonstrated the ability for mechanical strain to increase capillary diameter and lumen size, specifically as a result of increased transformation of capillaries into arterioles in skeletal muscle or arterialization (9, 25). Capillary lumen enlargement and enhanced arteriolar density have implications for increased blood flow and muscle tissue oxygenation. Therefore, understanding the mechanisms that contribute to these changes can influence our ability to optimally regenerate injured skeletal muscle and may assist in the development of therapeutic strategies for chronic obstructive disorders, such as peripheral artery disease.
In this study, we evaluated the extent to which the α7-integrin and/or mMSCs contribute to eccentric exercise-induced increases in capillary number and/or growth in skeletal muscle. We predicted that accumulation of mMSCs in α7Tg muscle would accelerate increases in capillary number and size following eccentric exercise, creating an environment that would be optimal for tissue health. Additionally, we hypothesized that mMSCs would indirectly contribute to beneficial vascular adaptations observed following eccentric exercise. The results from this study provide strong evidence that mMSCs constitute a therapeutic strategy for vascularization of injured or diseased muscle, a method that may provide an advantage over current treatments, including application of individual growth factors.
Protocols for animal use were approved by the Institutional Animal Care and Use Committee of the University of Illinois at Urbana-Champaign. Animals were fed standard laboratory chow and had access to water ad libitum. α7Tg mice (SJL/C57BL6: MCK-α7BX2) were produced at the University of Illinois Transgenic Animal Facility as prevously described (3). Wild-type (WT) littermates were used as controls. A total of 55 mice were assessed in this study, all of which were female to decrease the variability due to sex and to remain consistent with our previously published data.
Downhill running exercise.
Female WT and α7Tg mice remained at rest or completed single or multiple bouts of exercise. Five-week-old mice were used for acute and training studies. Eccentric exercise protocols consisted of one downhill running bout (−20°, 17 m/min, 60 min) or downhill running 3 times/wk (Monday, Wednesday, Friday) for 4 wk (−20°, 17 m/min, 30 min) for a total of 12 exercise bouts, respectively. Treadmill speed was gradually increased from 10 to 17 m/min over the first 2 min as a warm-up period during each experiment (3, 29). All running was performed on an Exer-6M treadmill (Columbus Instruments, Columbus, OH).
Tissue collection and preservation.
All mice were euthanized via carbon dioxide asphyxiation. To assess the time course of adaptation from a single bout of exercise, samples were collected 1, 2, and 7 days following the running session. Nonexercised controls (basal) were euthanized at 5 wk and 4 days. Samples collected from the multiple-bout mice were collected 1 day following the last running session. Gastrocnemius/soleus muscle complexes were collected from basal and exercised mice and immediately frozen in precooled isopentane. Muscle samples were dissected as a complex to preserve stem cells surrounding large vessels in the interstitium between the gastrocnemius and soleus.
Muscle complexes were divided at the midline along the axial plane and embedded in optimum cutting temperature (Tissue-Tek; Fischer Scientific). Three transverse cryosections per sample (8-μm nonserial sections, each separated by a minimum of 40 μm) were cut for each histological assessment using a CM3050S cryostat (Lecia, Wezlar, Germany). Sections were placed on frozen microscope slides (Superfrost; Fischer Scientific, Hanover Park, IL) and stored at −80°C before staining.
To assess skeletal muscle vascularity, tissue sections were stained with rat monocolonal anti-CD31 (390) (eBioscience, San Diego, CA), a marker for endothelial cells. Samples were costained with mouse monoclonal anti-dystrophin (MANDRA1) (Sigma-Aldrich, St. Louis, MO) to outline myofibers, necessary for measuring individual fiber cross-sectional areas. Briefly, frozen sections were fixed in ice-cold acetone for 10 min and blocked with PBS containing 5% bovine serum albumin (BSA), followed by 70 μg/ml AffiniPure anti-mouse fab fragments diluted in 5% BSA (Jackson ImmunoResearch, West Grove, PA) for 20 and 30 min, respectively. Both primary antibodies were diluted to a concentration of 1:100 in PBS with 1% BSA and were independently applied to the tissue sections for 60 min at room temperature. Fluorescein isothiocyanate (FITC)-labeled donkey anti-rat (1:250) and tetramethyl rhodamine isothiocyanate (TRITC)-labeled goat anti-mouse (1:100) secondary antibodies (Jackson ImmunoResearch) were used to detect the CD31 and dystrophin antibodies, respectively.
Tissue samples immunostained for neuron-glial antigen 2 (NG2), a proteoglycan expressed in arterioles and not venules, were processed as described above for CD31 (17). The rabbit polyclonal anti-NG2 (Millipore, Billerica, MA) was diluted to 1:200 in PBS with 1% BSA. FITC-labeled donkey anti-rabbit (1:200) was used for detection (Jackson ImmunoResearch).
Muscle sections were costained with rabbit polyclonal anti-Ki67 (1:100) (Abcam, Cambridge, MA) and rat monoclonal anti-stem cell antigen 1 (anti-Sca-1; D7; 1:100; BD Pharmingen, San Jose, CA) antibodies to determine the average number of proliferating cells within the vascular niche. This stain followed the same steps as the previously described above except for the addition of a 0.1% Triton X-100 permeabilization step before blocking, and a 4,6-diamidino-2-phenylindole (DAPI) stain was added to the second to last wash step (1:20,000) (Sigma Aldrich). Nuclei were visualized using DAPI to reliably assess cellular colocalization of Ki67 and Sca-1-specific staining.
To assess changes in capillarization, a combination of Adobe Photoshop and ImageJ plug-ins and tools were used to quantitate images acquired with a Ziess AxioCam digital camera and Axiovision software (Zeiss, Thornwood, NY). For evaluation, a total of five digital images acquired at ×20 magnification were captured at random from each sample. Images were acquired in separate color channels to allow for independent analysis of FITC (green) and TRITC (red) images. The CD31-FITC images were first analyzed for total number of transversely cut capillaries, as indicated by punctate, single-cell staining with the ImageJ template matching plug-in. Each image was then converted to an eight-bit image, and the longitudinally cut capillaries were manually traced using the NeuronJ plug-in. Finally, the dystrophin-TRITC images were imported into Adobe Photoshop (CS5 Extended) where an average of 500 fibers per sample were manually circled using the magnetic lasso tool that grabs the positively stained pixels and decreases subjectivity and interassessment error. Data from each fiber were recorded in the measurement log and included cross-sectional area as well as myofiber perimeter. Measures of capillary-to-fiber ratio (number of capillaries per number of muscle fibers), capillary density (number of capillaries per muscle fiber area), and tortuosity index (length of capillaries in contact with each fiber) were extrapolated from the data. Although it is common to count the number of capillary contacts per fiber as a way to estimate O2 supply, this measure does not take into consideration the tortuous path that vessels typically take in relation to the long axis of a myofiber (11, 27). For this reason, the tortuosity index, which is a measure of physical contact between capillaries and myofibers, was reported. Tortuosity index was calculated as follows: [total number of transversely cut capillaries × section thickness (8 μm) + sum of the lengths of longitudinally cut capillaries]/total myofiber perimeter.
For the evaluation of vessel growth, the entire muscle section was analyzed for the number of CD31-positive vessels with a visible lumen. The size of vessels with a lumen ranged from 7 to 50 μm, with most vessel diameters measuring from 7 to 20 μm. Data are reported as mean number of vessels per muscle section, where three tissue sections were evaluated per sample. NG2 analysis was performed using the same criteria with the exception of the vessels displaying NG2-specific staining rather than CD31.
Stem cell proliferation within the vascular niche was determined by the percentage of Sca-1+ cells expressing Ki67. Ki67+Sca-1+ cells were only counted if a nucleus was present. Large vessels and nerves are bundled together in the interstitium and have a characteristic structure that is easily identified by background staining. An average of three vascular niche areas per tissue section were evaluated. Data are reported as the average percentage of colocalization of all areas assessed.
Isolation of Sca-1+CD45− cells from skeletal muscle.
Gastrocnemius-soleus complexes were dissected, and Sca-1+CD45− cells were isolated from 5-wk-old α7Tg mice as previously described (29). Transgenic mice were subjected to a single bout of eccentric exercise before isolation of stem cells to increase yield for expansion and injection. However, mice remained sedentary before isolation of mMSCs for in vitro analyses to optimize detection of mMSC response to mechanical strain. After mechanical and enzymatic digestion of the dissected muscle tissue, filtered samples were incubated on ice with anti-mouse CD16/CD32 (1 μg/106 cells) (eBioscience) for 10 min to block nonspecific Fc-mediated interactions. Following the block, cells were stained in a cocktail of monoclonal anti-mouse antibodies Sca-1-phycoerythrin (600 ng/106 cells) and CD45-allophycoctanin (300 ng/106 cells) (eBioscience), diluted in 2% FBS in PBS. Fluorescence-activated cell sorting (FACS) was performed using an iCyt Reflection System (Champaign, IL), located at Carle Hospital (Urbana, IL). Negative and single-stained controls were used to establish gates. Sca-1+CD45− cells were collected in medium for culture [high glucose Dulbecco's modified Eagle's medium (DMEM), 10% FBS, 5 μg/ml gentamycin] and seeded on uncoated tissue culture dishes at 2.5 × 104 cells/cm2. Cultures were incubated at 37°C and 5% CO2. Media was replaced every 3 days.
Sca-1+CD45− cells isolated from α7Tg mice 24 h post-eccentric exercise were cultured on uncoated tissue culture dishes for 4 days before injection without passage. On the day of injection, cells were labeled with lipophilic dye (DiI) (Invitrogen, Grand Island, NY) and suspended in HBSS. Intramuscular injections at a concentration of 4 × 104 cells/ 50 μl HBSS were administered to 9-wk-old WT mice (n = 7), and contralateral legs were injected with 50 μl of HBSS. Four of the recipients were exercised 1 h before injection to recapitulate the postexercise microenvironment. Seven days following injection, the mice were euthanized via CO2 asphyxiation and gastrocnemius-soleus complexes were collected and preserved as described above.
In vitro mechanical strain.
Sorted Sca-1+CD45− cells (10 × 104) were expanded to 80–90% confluence on uncoated tissue culture dishes and then seeded on laminin (Tyr-Ile-Gly-Ser-Arg)-coated six-well flexible silicone elastomer membrane plates (9.62 cm2) for a total of 10.4 × 103 cells/cm2 (Flexcell International, McKeesport, PA). Cells were incubated in high-glucose DMEM at 37°C and 5% CO2 for 24 h to allow for sufficient cell attachment. Before mechanical strain, cells were washed with warm PBS and switched to serum-free high-glucose DMEM. Biaxial mechanical strain (10%, 1 Hz) was applied to the cells for 5 h using a FX-4000 Flexercell strain unit (Flexcell International) (29). Cells maintained under static conditions were used as nonstrained controls. Twenty four hours following the start of mechanical strain, conditioned media was collected from the nonstrained and strained cultures.
Mechanically strained and nonstrained cells were immunostained for levels of α-smooth muscle actin (α-SMA) expression. Twenty four hours following the start of mechanical strain, the cells were fixed in 4% paraformaldehyde solution for 15 min at room temperature. The cells were permeabilized in 0.1% Triton X-100 in PBS for 5 min, followed by a 60-min block in a 5% BSA in PBS solution. Mouse monoclonal anti-α-SMA (1A4; 1:100; ScyTek, Logan, UT) and FITC-labeled secondary (1:200; Jackson ImmunoResearch) antibodies were applied to the cells. Representative images were immediately acquired with a Zeiss AxioCam digital camera and Axiovision software (Zeiss).
Conditioned media from unstrained and strained cell cultures was analyzed with Proteome Profiler Antibody Arrays (ARY015, ARY013; R&D Systems, Minneapolis, MN), and the recommended protocol was followed. Briefly, 250 μl of media from each well (n = 6) was pooled and applied to a membrane precoated with multiple capture antibodies. A single array was run for each condition using the pooled media. Samples were incubated at 4°C overnight. Following multiple washes, a horseradish peroxidase-conjugated secondary was applied for 30 min. Blots were detected using SuperSignal West Dura extended duration substrate (Thermo Scientific, Rockford, IL) and a Bio-Rad ChemiDoc XRS system (Bio-Rad, Hercules, CA). Quantification was completed using Quantity One software (Bio-Rad).
All averaged data are presented as means ± SE. To determine significance, comparisons between groups were performed by two-way ANOVA. When main effects were detected, Tukey's post hoc analyses were performed. All analyses were completed with SPSS statistical software (20.0, Chicago, IL). Differences were considered significant at P < 0.05.
Number of CD31+ capillaries decrease in WT and α7Tg muscle following single and multiple bouts of eccentric exercise.
To accurately address the capillary response to exercise, several measures of angiogenesis were assessed, including capillary density, capillary-to-fiber ratio, and capillary tortuosity. All measures of capillarization were significantly decreased in the days following a single bout of downhill running exercise (Fig. 1, A–C). Capillary density was significantly decreased 1 and 7 days following a single bout of eccentric exercise compared with nonexercise controls (basal) in both WT and α7Tg skeletal muscle (P < 0.01) (Fig. 1A). Capillary-to-fiber ratio and tortuosity index were significantly suppressed 7 days in both WT and α7Tg skeletal muscle (P < 0.01) (Fig. 1, B and C). Four weeks of eccentric exercise training similarly decreased capillary density and tortuosity index compared with sedentary controls in WT (P < 0.05) (Fig. 2, A–C).
CD31+ vessel diameter is increased in α7Tg muscle following both single and multiple bouts of eccentric exercise.
With the use of the same samples that were evaluated for capillary number, CD31+ vessels with a visible lumen were counted following single and multiple bouts of eccentric exercise (Fig. 3A). Muscle sections were also costained for NG2 proteoglycan, a protein exclusively expressed in arterioles (17) (Fig. 3B). The number of CD31+ vessels with a visible lumen was increased in α7Tg muscle at 7 days following a single exercise bout compared with nonexercised WT muscle and WT muscle 1 and 2 days postexercise (P < 0.05) with no significant rise in capillary growth in WT at any day postexercise (Fig. 3C). A trend toward an increase in CD31+ vessels with a visible lumen was detected in α7Tg 7 days compared with α7Tg basal (P = 0.064) and α7Tg 7 days compared with WT 7 days (P = 0.057). A significant interaction was detected between group and treatment (P < 0.01) for vessel growth in response to exercise training (Fig. 3E). Eccentric exercise training significantly increased the quantity of CD31+ vessels with a visible lumen in both WT (2.1-fold) and α7Tg (2.7-fold) muscle following 4 wk of eccentric exercise training, with 30% more growth observed in α7Tg compared with WT (P < 0.05) (Fig. 3E). NG2+ vessels were not increased in WT or α7Tg muscle 7 days following a single bout of eccentric exercise (Fig. 3D) or 4 wk of eccentric exercise training (Fig. 3F).
Sca-1+ cells proliferate in α7Tg skeletal muscle following a single bout of eccentric exercise.
Ki67+Sca-1+ cells with positive nuclear staining were present in the vascular niche of α7Tg muscle at 1-day postexercise (21% of all Sca-1+ cells) compared with sedentary WT (9%) (P < 0.05) (Fig. 4). Trends for increases in Ki67+Sca-1+ cells were noted in WT muscle at 1-day postexercise compared with sedentary WT muscle and in sedentary α7Tg compared with sedentary WT.
mMSCs contribute to vascular adaptations following intramuscular transplantation.
Sca-1+CD45− cells were extracted from α7Tg muscle 24 h post-eccentric exercise, labeled with DiI and transplanted into WT recipient muscle (Fig. 5A). A subset of recipient mice were exercised 1 h before injection to recapitulate the exercise microenvironment. Consistent with observations presented in Fig. 1, eccentric exercise before injection decreased the capillary-to-fiber ratio and tortuosity index in both saline- and cell-injected muscle (P < 0.05) (Fig. 5, B–D). Interestingly, mMSC transplantation elevated the number of CD31+ vessels with a visible lumen only in muscle that had been exposed to eccentric exercise before injection compared with all other groups (P < 0.05) (Fig. 5E). The quantity of NG2+ vessels, or arterioles, was also increased in transplanted muscle following eccentric exercise compared with both saline-treated controls (P < 0.05) (Fig. 5F).
DiI labeling of the mMSCs allowed us to determine whether increased vessel growth occurred as a result of direct fusion with existing vessels or via paracrine factor release. While DiI+ mMSCs were found in the vascular niche and did appear to fuse with large vessels (Fig. 5G), this event was not common (<1% incorporation). Although the cells could migrate away from the site of injection, DiI+ mMSCs did not consistently appear in close proximity to CD31+ vessels and were randomly associated with a variety of structures in the muscle, including individual muscle fibers and the myotendinous junction.
Mechanical strain can stimulate the release of arteriogenic factors from mMSCs.
Pericytes provide structural support for vessels and are hypothesized to prevent damage in the presence of increased blood flow and/or mechanical strain (1, 33). The ability for pericytes to wrap and extend themselves around vessels suggests that these cells may be particularly sensitive to changes in mechanical strain that may occur during lengthening contractions. To test this hypothesis, we isolated mMSCs, previously characterized as pericytes (29), from α7Tg muscle and strained the cells in the presence of laminin, a substrate that we have found can maintain MSC marker expression. Strain resulted in decreased ability to visualize α-SMA protein in mMSCs in culture (Fig. 6A).
Arteriogenic (Fig. 6B) and anti-angiogenic factors (Fig. 6C) were detected in the conditioned media from mMSCs subjected to mechanical strain in the presence of laminin (Fig. 6B). While antiangiogenic factors were modestly increased (∼20–40%), several arteriogenic factors were markedly increased (40% to 4-fold) in the media following strain compared with media collected from unstrained cells. Specifically, epidermal growth factor (EGF) was elevated 4.2-fold, and granulocyte macrophage-colony stimulating factor (GM-CSF) was increased 3.1-fold. Protein delta homolog 1 (DLK1/Pref-1), a factor that inhibits MSC adipogenesis, was increased 2.4-fold (data not shown). Endoglin (CD105), a soluble TGF-β receptor, was also increased 3.4-fold (data not shown).
In this study, we provide novel insights regarding the ability for the α7-integrin and mMSCs to regulate the vascular response to eccentric exercise. First, we demonstrate that the α7-integrin can positively influence the rate and degree of vessel growth following eccentric exercise. Second, we establish the ability for mMSCs to rapidly and markedly stimulate vessel growth and arteriolar density following eccentric exercise. Finally, our work provides evidence that mMSC paracrine factor release may contribute to vascular growth post-eccentric exercise. Ultimately, the information provided in this study may contribute to the development of strategies aimed at improving skeletal muscle vascularization, which, in turn, has the potential to improve muscle function.
Angiogenesis is important for wound healing and is an adaptive response to endurance exercise. Advantages to capillarization postexercise include creation of a larger surface area for oxygen diffusion, decreased oxygen diffusion distance, and reduced blood flood velocity necessary for increased metabolic gas exchange (32). A paucity of data exists regarding the angiogenic response following eccentric exercise. In this study, we demonstrate that capillary density, capillary-to-fiber ratio, and tortuosity index decrease following a single bout of eccentric exercise. New fiber synthesis in the muscle in response to eccentric contractions might account for a reduction in the capillary-to-fiber ratio. If this were true, we would expect to observe a preferential decline in CD31+ vessels in α7Tg compared with WT muscle given the marked rise in newly synthesized fibers previously observed in these samples (13), yet the number of CD31+ vessels was similarly decreased in WT and α7Tg samples post-eccentric exercise.
We hypothesize that endothelial cells in capillaries and/or arterioles are remodeling in response to increased stretch or tension associated with lengthening contractions (22, 26). Previous studies based on morphological assessment using electron microscopy suggest overt capillary damage does not occur in response to eccentric exercise (12). However, a more recent study has provided evidence that apoptosis occurs in skeletal muscle at 7 days postexercise, almost exclusively in endothelial cells (26). Microvascular remodeling observed by caspase-3 and CD31 coexpression is also commonly observed in rapidly growing tumors and in patients with peripheral artery disease (16, 31). We suspect that a dynamic endothelial degradation and reformation process stimulated by myofiber strain is responsible for the decline in CD31+ vessels and that this event is important for increasing lumen diameter post-eccentric exercise.
Arteriogenesis refers to growth of preexisting collateral arterioles and transformation into fully functioning large diameter conductance arteries, an adaptation that occurs to compensate for reduction in tissue blood supply caused by arterial obstruction (30). In the 1970s, it was determined that the increase in size of arterioles was not due to simple vasodilation, but due to increased proliferation of endothelial and smooth muscle cells, resulting in a dramatic increase in vessel diameter (24). Arteriogenesis has not been reported in muscle response to eccentric exercise. However, vascular adaptations such as capillary growth and capillary-to-arteriole transformation, or arterialization (14), are commonly observed in rodent models following eccentric exercise or chronic loading (9, 12, 25). During the course of our evaluation, larger vessels were visually apparent in α7Tg samples postexercise. Larger vessels with an open lumen were quantitated and found to be higher in α7Tg muscle compared with WT with a single bout of eccentric exercise and following 4 wk of training. It is interesting to note that acute loss of CD31 expression in microvessels coordinated with the increased appearance of larger vessels in α7Tg 7 days following acute eccentric exercise, but not WT. These data suggest that the α7-integrin positively influences vessel growth in response to eccentric exercise.
We previously have established that mMSCs accumulate in skeletal muscle following a single bout of eccentric exercise in WT mice and that this effect is dependent on the presence of the α7-integrin (29). Specifically, we have shown that mMSCs are significantly higher in α7Tg muscle and are minimally observed in α7−/− muscle compared with WT postexercise. The absence of CD45 expression would suggest mMSCs are tissue resident and not migrating from an exogenous tissue source, such as bone marrow. We have not yet elucidated the mechanism by which the integrin stimulates the presence of mMSCs in muscle but previously suspected that the α7-integrin and/or eccentric exercise could increase proliferation of resident mMSCs (29). In the current study, the percentage of Sca-1+ cells positive for Ki67 in the vascular niche was increased in α7Tg muscle 1 day following eccentric exercise compared with sedentary WT muscle. A statistically significant rise in Ki67+Sca-1+ cells was not detected in WT muscle 1-day postexercise compared with WT basal in this study, yet a positive trajectory was noted. Thus cell proliferation of Sca-1+ cells in the vascular niche may contribute to the total number of mMSCs (Sca-1+CD45−) previously observed in WT and α7Tg postexercise as analyzed by FACS and immunohistochemistry analyses (29). Given the potential for mMSC expansion in WT muscle, it is not clear why an increase in vessel growth is only observed in α7Tg postexercise and not WT. The possibility exists that mMSCs in WT muscle are susceptible to apoptosis as a result of damage and inflammation and greater turnover is necessary to reestablish the population. We predict that mMSC survival and accumulation in α7Tg muscle may account for the rapid and sustained vessel growth observed in these mice.
mMSC transplantation can stimulate satellite cell expansion and new fiber synthesis in WT muscle (29). We now demonstrate that mMSCs may also underlie rapid vascular adaptations in α7Tg muscle postexercise. Introduction of mMSCs into WT recipient muscle 1) dramatically increased larger CD31+ vessels with a visible lumen and 2) enhanced the number of NG2 proteoglycan-positive arterioles. These observations occurred within 1 wk and were optimal when WT recipient mice were subjected to eccentric exercise before injection. We do not believe that damage or inflammation associated with eccentric exercise is responsible for mMSC contribution to CD31+ vessel and arteriolar growth given the capacity for these events to occur in α7Tg muscle which is protected from injury and macrophage infiltration (3, 13). Secretion of a factor from the muscle fiber in response to lengthening contractions may allow for an increase in the expansion of the injected mMSCs, but this event was not evaluated in the current study since sufficient tissue was not available to quantitate mMSCs by flow cytometry. An addition of conditioned media from stretched myotubes to mMSCs and an evaluation of the mMSC response might be the appropriate experimental design to address this question. Regardless, the results from this experiment suggest that mMSC-mediated vascular growth may be an integral part of the regeneration and repair process in muscle that is critical for revitalization of function. Our current studies focus on the ability for mMSCs to increase blood flow and enhance muscle function in both young and old mice.
It is becoming increasingly clear that MSCs repair damage via secretion of paracrine factors (5, 15). Our recent work suggests that mMSCs, predominantly pericytes that reside around large vessels and the capillary vascular niche, upregulate RNA expression of MSC cell surface markers (Sca-1, CD90, PDGFR-β) following mechanical strain on silicone membranes incorporating a laminin peptide (29). We used these in vitro conditions to determine whether mMSCs are sensitive to mechanical strain and whether the strain stimulus could facilitate the release of factors that promote vessel growth. GM-CSF, EGF, VEGF, and heparin-binding EGF were found to be elevated in media from stretched mMSCs compared with nonstretched controls. GM-CSF (3.1-fold) and EGF (4.2-fold) were dramatically higher compared with 80 other proteins assessed. Both appear to contribute to vessel growth by different mechanisms. Whereas GM-CSF can promote arteriogenesis, or arteriole-to-artery transformation, following occlusion by increasing monocyte function (8), EGF can enhance MSC adhesion to endothelial cells (2). These data are in agreement with other studies demonstrating that MSCs secrete factors to support vessel growth without the need for engraftment (19), yet extend this concept to suggest that mMSCs optimally secrete these factors as a result of mechanical strain. Further studies are necessary to determine whether mMSCs dedifferentiate and/or secrete factors important for vessel growth following exposure to tension associated with eccentric contractions. The degree to which mMSC phenotype dictates growth factor release in response to strain should also be explored.
In conclusion, we have demonstrated that the α7-integrin and mMSCs stimulate vessel growth and arteriolar density in skeletal muscle following eccentric exercise. Our data suggest that increased mechanical strain associated with lengthening contractions may facilitate mMSC release of factors necessary to confer beneficial vascular adaptations. The results from this study justify further evaluation of α7-integrin- and mMSC-based therapeutic strategies to circumvent the loss of oxygenation with peripheral artery disease and enhance the regeneration and repair of muscle following injury.
This work was supported by a grant from the Ellison Medical Foundation (to M. D. Boppart). H. D. Huntsman was supported by a National Science Foundation Integrative Graduate Education and Research Traineeship in Cellular and Molecular Mechanics and BioNanotechnology.
No conflicts of interest, financial or otherwise, are declared by the author(s).
H.D.H., K.Z., M.C.V., M.D.L., and M.D.B. conception and design of research; H.D.H., N.Z., K.Z., P.R., and M.C.V. performed experiments; H.D.H., N.Z., K.Z., and P.R. analyzed data; H.D.H., K.Z., M.D.L., and M.D.B. interpreted results of experiments; H.D.H., K.Z., and M.D.B. prepared figures; H.D.H. and M.D.B. drafted manuscript; H.D.H., N.Z., K.Z., M.C.V., M.D.L., and M.D.B. edited and revised manuscript; H.D.H., N.Z., K.Z., P.R., M.C.V., M.D.L., and M.D.B. approved final version of manuscript.
We would like to thank Cassie Drummond and Danielle Mae Weech for their assistance with data analysis.
- Copyright © 2013 the American Physiological Society