^CD146+Mesenchymal Stem Cells Treatment Improves Vascularization, Muscle Contraction and VEGF Expression, and Reduces Apoptosis in Rat Ischemic Hind Limb

Abstract

Peripheral arterial disease (PAD) is an increasingly common narrowing of the peripheral arteries that can lead to lower limb ischemia, muscle weakness and gangrene. Surgical vein or arterial grafts could improve PAD, but may not be suitable in elderly patients, prompting research into less invasive approaches. Mesenchymal stem cells (MSCs) have been proposed as potential therapy, but their effectiveness and underlying mechanisms in limb ischemia are unclear. We tested the hypothesis that treatment with naive MSCs (nMSCs) or MSCs expressing CD146 (^CD146+MSCs) could improve vascularity and muscle function in rat model of hind-limb ischemia. Sixteen month old Sprague-Dawley rats were randomly assigned to 4 groups: sham-operated control, ischemia, ischemia+nMSCs and ischemia+^CD146+MSCs. After 4 weeks of respective treatment, rat groups were assessed for ischemic clinical score, Tarlov score, muscle capillary density, TUNEL apoptosis assay, contractile force, and vascular endothelial growth factor (VEGF) mRNA expression. ^CD146+MSCs showed greater CD146 mRNA expression than nMSCs. Treatment with nMSCs or ^CD146+MSCs improved clinical and Tarlov scores, muscle capillary density, contractile force and VEGF mRNA expression in ischemic limbs as compared to non-treated ischemia group. The improvements in muscle vascularity and function were particularly greater in ischemia+^CD146+MSCs than ischemia+nMSCs group. TUNEL positive apoptotic cells were least abundant in ischemia+^CD146+MSCs compared with ischemia+nMSCs and non-treated ischemia groups. Thus, MSCs particularly those expressing CD146 improve vascularity, muscle function and VEGF expression and reduce apoptosis in rat ischemic limb, and could represent a promising approach to improve angiogenesis and muscle function in PAD.

Graphical Abstract

1. INTRODUCTION

Peripheral arterial disease (PAD) is an increasingly common atherosclerosis of the peripheral arterial beds such as the carotid, subclavian, visceral, splanchnic and lower extremity arteries. Atherosclerosis and narrowing of the lower extremity peripheral arteries cause limb ischemia, decreased blood supply and inadequate fulfilment of the metabolic demands of the lower limb muscles and other tissues. PAD and lower limb ischemia are often manifested as lower limb intermittent claudication and muscle pain at rest. PAD and lower limb ischemia affect approximately 8.5 million individuals in the United States and more than 200 million individuals worldwide [1]. PAD is the third leading cause of atherosclerotic cardiovascular morbidity after coronary artery disease and stroke [2, 3]. If not corrected, PAD of the lower limb could lead to loss of muscle mass, decreased muscle contractile force, tissue necrosis, and gangrene. The mortality from complications of critical limb ischemia exceeds that of other occlusive cardiovascular diseases including symptomatic coronary artery disease, with mortality rates as high as 20% within 6 months, 10% to 40% within 1 year, and over 50% within 5 years from diagnosis [4–6]. Surgical interventions using vein or arterial grafts and endovascular revascularization are the most effective and immediate approaches to restore blood flow to the ischemic limb. However, many of the lower limb ischemia patients are elderly individuals, and up to 30% patients may not be good candidates for surgical interventions due to potentially high risks of invasive surgical approaches [7]. This has prompted research into alternative less invasive approaches to improve lower limb vascularization in elderly subjects with PAD and at high surgical risks.

Stem cell therapy has been introduced to improve angiogenesis and vascularity in the ischemic heart [8, 9], kidney [10, 11], and brain [12, 13]. Growing pieces of evidence have demonstrated that mesenchymal stem cells (MSCs) could differentiate into vascular cells and in turn increase tissue vascularization [14]. A randomized controlled clinical trial has shown that bone marrow-derived MSCs therapy could increase lower limb perfusion and promote healing of diabetic foot ulcer in diabetic patients [15]. In support, our previous studies have shown that MSCs derived from peripheral blood could promote angiogenesis in aged rat model of chronic limb ischemia [16]. Increased release of vascular endothelial growth factor (VEGF) has been suggested as a major factor in promoting angiogenesis in tumors [17–19] and in response to critical limb ischemia [20–23], and MSCs differentiation into vascular cells could increase vascularization of ischemic tissues in part through the release of VEGF.

MSCs comprise multiple cells that play different roles and express various biomarkers such as Sca-1, CD14, CD29, CD34, CD44, CD45, CD90, CD105 and CD146. Specifically, CD146 has been reported as a major biomarker of vascularization and tumor growth both in vivo and in vitro [24–26]. Also, tissue culture studies have shown that CD146 is expressed in capillary pericytes during endothelial cell proliferation, migration and tube formation [24, 27]. Additionally, ^CD146+MSCs have been shown to be more efficient than nMSCs in improving myocardial regeneration in ischemic hearts, and in extending survival in muscular dystrophy mouse models [28, 29]. However, little is known regarding the in vivo angiogenic benefits of CD146, and the mechanisms underlying CD146-mediated benefits in limb ischemia. Here we hypothesize that ^CD146+MSCs are superior to nMSCs and have potentially greater angiogenic effects in lower limb ischemia. We used aged rat model of hind-limb ischemia to test whether: 1) ^CD146+MSCs improve the clinical scores of the ischemic limb, 2) ^CD146+MSCs improve the ischemic limb muscle vascularization, capillary density and contractile force, 3) The ^CD146+MSCs improved vascularization and muscle function are related to decreased muscle apoptosis and increased VEGF mRNA expression.

2. MATERIALS AND METHODS

2.1. Rationale for the experimental animal groups

PAD is an atherosclerotic disease of the peripheral arterial beds that is more common in elderly human individuals [1], and therefore it is ideal to study the mechanisms of the disease and potential therapies in an aged animal model. On the other hand, elderly human individuals may not be good candidates for bone marrow MSC isolation because they do not have much bone marrow within their bones. Also, previous studies and our preliminary experiments have shown that it is difficult to harvest healthy high-quality MSCs from aged rats [16]. Therefore, we used aged rats to simulate elderly individuals with arteriosclerosis obliterans and treated them with MSCs isolated from young rats. Pathogen free male Sprague-Dawley rats (n=40) from the animal core facility of Nanjing Medical University (Nanjing, China) were used. Normal young 12 w old rats (n=5) were used to harvest nMSCs and isolate ^CD146+MSCs from the bone marrow. Aged 16 mo old rats (n=35) were used in the different experimental animal models and groups. All animal procedures followed the guidelines of the American Physiological Society regarding the Humane Care and Use of Animals for Research, and were approved by the Ethics Committee of Ganzhou People’s Hospital (Protocol Number: 20180325, Date of Approval: August 22, 2018).

2.2. Harvest of rat nMSCs and CD146+MSCs

Young 12 w old rats (n=5) were euthanized and nMSCs were collected from the bone marrow. Briefly, the femur and tibia bones were isolated, the bone marrow cavity was open using plyers, and the bone marrow cavity was scraped and rinsed with phosphate-buffered saline (PBS). Density gradient centrifugation (Ficoll–Paque PLUS, 1.077 ± 0.001 g/mL; GE Healthcare Life Sciences, Fairfield, CT, USA) was used to collect the mononuclear cells. The collected bone marrow preparation was centrifuged at 1000 ×g for 5 min, resulting in 4 distinct fractions. The first top fraction containing PBS and dead cells was discarded. The second fraction containing mononuclear cells was collected and cultured in Gibco minimum essential medium (Thermo Fisher Scientific, Waltham, MA USA) containing 15% fetal bovine serum, 1% penicillin, 1% streptomycin and glutaMAX-I. The culture medium was changed and non-adherent cells were removed every 3 days. The adherent MSCs were harvested and subcultured for 5-passages, then characterized and separated by analysis of their surface markers using flow cytometry (EPICS ALTRA; Beckman Coulter Inc, Brea, CA). For identification of nMSCs surface markers, specific antibodies for Sca-1, CD14, CD29, CD34, CD44, CD45, CD90, CD105 were used (Abcam, Cambridge, MA). For isolation of ^CD146+MSCs, nMSCs were further purified by flow cytometry using surface marker for ^CD146+MSCs and specific antibody for CD146 (Abcam) [16, 30]. The expression of CD146 in ^CD146+MSCs was confirmed using RT-PCR as described below.

2.3. Animal models

To simulate aging individuals with atherosclerotic lesions and arteriosclerosis obliterans, aged 16 mo old rats (n=35) were maintained on a diet containing 20% lard (Jiajie Grease Co. LTD, Hebei, China), 3.5% (wt/wt) cholesterol (Tianqi Chemical Technology Co. LTD, Anhui, China), 2% sodium cholate (Tianqi Chemical Technology Co. LTD, Anhui, China), and 0.2% propylthiouracil (South China Pharmaceutical Group Co. LTD, Guangdong, China) for three months [31]. Rats are resistant to induction of atherosclerosis by high cholesterol diet alone, but become more susceptible in the hypothyroid state achieved by administration of propylthiouracil [32]. To further intensify atherosclerosis, the rats also received 35000 unit per 100 g vitamin D3 (General Pharmaceutical Co. LTD, Shanghai, China) i.m. in the affected limb, once per month for three months. The rats were randomly divided into 4 groups: 1) sham-operated control group (n=5), 2) ischemia group treated with saline (n=10), 3) ischemia+nMSCs group treated with nMSCs (n=10), 4) ischemia+^CD146+MSCs group treated with ^CD146+MSCs (n=10) (Fig. 1A).

Fig. 1.

A) Schematic diagram of the surgical procedure, treatment and treatment periods in different rat groups. Aged rats were randomly assigned to four groups: sham-opertaed control group (n=5), ischemia group treated with saline (n=10), ischemia+nMSCs group treated with nMSCs (n=10), and ischemia+^CD146+MSCs group treated with ^CD146+MSCs (n=10). Following the respective treatment, Clinical score, Tarlov score, muscle capillary density, apoptosis, contractile force, and VEGF mRNA expression were assessed in the ischemia limb. B) RT-qPCR and mRNA expression of CD146 in nMSCs and ^CD146+MSCs isolated and cultured from bone marrow of young rats. Data represent means±SD (n=5). **P < 0.01 vs nMSCs.

All ischemia rats underwent a surgical procedure to induce chronic hind-limb ischemia as previously described [16]. Briefly, the rats were anesthetized by isoflurane inhalation and the right femoral artery was exposed and freed of fat and connective tissue under sterile and aseptic conditions. A small stainless-steel clip was applied to the femoral artery cranially near its origin and another clip placed 3 mm caudally. To induce limb ischemia, a small 2 mm incision was made in the femoral artery near the top clip, and a small sterile and heparin-coated silicone tube (0.5 mm OD, 2 mm long, Yukang Medical Device Co. LTD, China) was implanted in the femoral artery and secured to the arterial wall using polypropylene (Prolene) sutures. The stainless-steel clips were removed, lack of bleeding was confirmed, the wound was closed using silk sutures, and the animals were observed for 30 days. Our previous studies have shown that following the hind limb ischemia surgery, the limb would show a decrease in blood flow that reaches a steady-state 3–4 weeks after surgery [16, 33, 34]. After 30 days, all animals undergoing surgery showed manifestation of limb ischemia including pale foot, gait abnormalities, and toe(s) gangrene. Control animals were sham-operated.

After 30 days of limb ischemia surgery, ischemia rats were divided randomly into 3 different groups, and injected on two successive days intramuscular in the ischemic limb at 6 sequential injection sites 5 mm apart with either saline (ischemia group), or 1×10⁶ nMSCs (ischemia+nMSCs), or 1×10^{6 CD146+}MSCs (ischemia+^CD146+MSC). This muscle injection protocol would allow a more symmetrical and uniform distribution of stem cells throughout the muscle tissue as previously described [35, 36]. The animals were observed daily, and measurements were taken 28 days after the respective experimental treatment, a duration that was shown in previous studies and with other treatment modalities to be sufficient to improve blood flow in the ischemic limb relative to that in the contralateral normal limb when measured simultaneously with Doppler ultrasound [16, 33, 34].

None of the ischemia animals died in this study (mortality rate = 0), and treatment of the ischemia animals with nMSCs or ^CD146+MSCs did not affect the survival rate. Also, the contralateral hind-limb was not made ischemic, and induction of ischemia and injection of MSCs in the ipsilateral hind-limb were not associated with any detectable changes in the contralateral limb, which appeared normal, functional, and without any ischemic manifestations.

2.4. Ischemic clinical score and Tarlov score

After 4 weeks of the respective treatment, the stage of limb ischemia in the different groups was determined by two blinded investigators using two different and previously established ischemic scores, namely the clinical score [37] and Tarlov scale [38] (Table 1).

Table 1.

Ischemic scores in limb ischemia model

Score	Clinical Presentation
Clinical Score
0	Auto-amputation of > half lower limb
1	Gangrenous tissue > half foot
2	Gangrenous tissue < half foot, with lower limb necrosis
3	Gangrenous tissue < half foot, without lower limb necrosis
4	Pale foot or gait abnormalities
5	Normal
Tarlov score
0	No movement
1	Barely perceptible movement, no weight bearing
2	Frequent and vigorous movement, no weight bearing
3	Supports weight, may take one or two steps
4	Walks with only mild deficit
5	Normal, but slow walking
6	Full and fast walking

2.5. Measurements of capillary vessel density

After 4 w of respective treatment and determination of the ischemic score, rats were euthanized by CO₂ inhalation and sacrificed by exsanguination from the carotid artery. The adductor longus muscle was isolated, fixed in 4% paraformaldehyde fixative, embedded in paraffin and sliced into sections using a microtome. The muscle sections were incubated at 4°C overnight in fluorescein isothiocyanate (FITC)-labeled monoclonal antibody to the endothelium marker CD31 (JC/70A, Abcam, Cambridge, MA), FITC-labeled monoclonal anti α-smooth muscle actin antibody (1A4, Abcam) and tetramethylrhodamine-isothiocyanate (TRITC)-labeled lectin Kit for blood cells (EY Laboratories, San Mateo, CA) as previously described [16, 33, 34, 39]. The stained muscle sections were examined by two blinded investigators using Axio Scope A1 microscope (Carl Zeiss, Heidenheim, Germany). The number of capillaries was assessed in five different randomized fields of each slide at higher magnification using SlideBook-6 digital microscopy software (Intelligent Imaging Innovations, Denver, CO). The average number of capillaries was calculated for each photomicrograph of the different rat groups.

2.6. TUNEL apoptosis assay

The muscle sections were stained using TUNEL apoptosis assay kit (Beyotime Biotechnology, Shanghai, China). Briefly, paraffin-embedded sections were deparaffinized with xylene, absolute ethanol, 90% ethanol, 70% ethanol and distilled water. The tissue section was treated with protease K for 30 min at 25°C, then washed with PBS three times. The tissue sections were incubated in 3% H₂O₂ PBS for 20 min, then washed again with PBS three times. Tissue sections were incubated in 50 ml TUNEL reaction solution in a dark environment at 37°C for 60 min. The tissue sections were washed in PBS three times, 50 μl streptavidin-HRP was added according to the manufacturer’s instructions and the sections were incubated at room temperature for 30 min. Sections were washed in PBS, mounted on slides and examined under DWLB2 light microscope (Leica microsystems, Wetzlar, Germany). Apoptotic TUNEL positive muscle cells appeared with dark brown nuclei. The average number of TUNEL positive muscle cells was counted from three randomized fields per section. Results were presented as the percentage TUNEL positive cells relative to total cells in the field.

2.7. Measurements of muscle contractile force

In euthanized rats, the hind limb skin was open, the subcutaneous tissue was separated and the surface of gastrocnemius muscle was exposed. The thigh muscles were carefully dissected layer by layer, the sciatic nerve was localized between the semimembranosus and biceps femoris muscles, and approximately 1.5 cm of the sciatic nerve was dissected. The gastrocnemius muscle and its tendon, vessels and sciatic nerve were excised, and the tendon of gastrocnemius was ligated to make the gastrocnemius muscle-sciatic nerve preparation. One muscle preparation was prepared from each rat from each group (n=5–10 per group). The sciatic nerve was connected to the electrodes of Master-8 stimulator (A.M.P.I., Jerusalem, Israel), and the gastrocnemius muscle tendon was hooked to a force transducer. The frequency of nerve stimulation was increased until the gastrocnemius muscle contracted tetanically, and the maximum isometric contractile force was calculated and recorded using ASB240U biological signal acquisition system (Aosheng electronics Co, Ltd, Chengdu, China). The frequencies of nerve stimulation, and time to reach tetanic contraction were not significantly different between groups

2.8. VEGF and CD146 mRNA expression analysis

The mRNA expression of VEGF was measured in the affected limb muscle, and the mRNA expression of CD146 was measured in cultured nMSCs and ^CD146+MSCs using a standard quantitative reverse transcription polymerase chain reaction (RT-qPCR). Briefly, the gastrocnemius muscle from the right hind limb, and cultured MSCs were prepared for RNA isolation. Total RNA was isolated following TRIzol reagent user guide (Thermo Fisher Scientific, Waltham, MA), and reverse transcribed using High-Capacity cDNA Reverse Transcription Kit with RNase Inhibitor (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA).

The primers included:

VEGF	Forward	5’-CCTGGCTTTACTGCTGTACCT-3’
	Reverse	5’-GATGTCCACCAGGGTCTCAAT-3’ [40]
CD146	Forward	5’-GGGTACCCCATTCCTCAAGT-3’
	Reverse	5’-CAGTCTGGGACGACTGAATG-3’ [41]
GAPDH	Forward	5’-TGATTCTACCCACGGCAAGTT-3’
	Reverse	5’-TGATGGGTTTCCCATTGATGA-3’ [42]

PCR was carried out with 1 cycle for 10 min at 95°C then 40–45 cycles of 30 sec denaturation at 95°C, 45 sec of annealing at 56°C, and 30 sec of extension at 72°C, followed by 1 min of final extension step at 95°C. The number of PCR cycles varies according to the expression level of the target gene. An appropriate primer concentration and number of cycles were determined to ensure that the PCR is taking place in the linear range and thereby guarantees a proportional relationship between input RNA and the cycles readout. The relative gene expression was calculated by comparing cycle thresholds relative to the housekeeping gene reference glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The relative mRNA expression levels were calculated using the 2^−ΔΔCt method as previously described [43].

2.9. Statistical analysis

Results were presented as means±standard deviation (SD). Statistical analysis was performed using IBM SPSS Statistics 24.0 sof43tware (International Business Machines Corporation, New York, NY, USA). The means in different groups were compared by analysis of variance (ANOVA) followed by least significant difference test (for normally distributed data) or Kruskal-Wallis test (for not normally distributed data). A value of P < 0.05 was considered statistically significant.

3. RESULTS

Isolated and subcultured nMSCs were characterized using flow cytometry, surface markers antibodies and matching reagents. Among the different MSCs surface markers tested, Sca1, CD29, CD44, CD90, CD105 were positive and CD14, CD34, CD45 were negative. This is consistent with previously reported and well-established MSCs phenotype [44]. ^CD146+MSCs were further purified from nMSCs, using flow cytometry, surface marker for ^CD146+MSCs and specific CD146 antibody. RT-qPCR analysis confirmed that CD146 mRNA expression was ~30 fold higher in ^CD146+MSCs compared with nMSCs (Fig. 1B).

All rats survived throughout the surgical procedure, and showed different manifestations after 2 weeks and 4 weeks of respective treatment (Table 2). After 2 weeks of the respective treatment, the ischemia group treated with saline had one rat with necrotic lower extremity, one rat with auto-amputation, and three rats with toe(s) gangrene. In comparison, in the ischemia+nMSCs group, two rats had toe(s) gangrene, and in the ischemia+^CD146+MSC group only one rat had toe(s) gangrene. After 4 weeks of the respective treatment, the rats in the ischemia group did not show any improvement. In comparison, the ischemia+nMSCs group showed improvement and had only one rat with toe(s) gangrene, while all rats in the ischemia+^CD146+MSCs group showed further improvement and normal appearance and activity.

Table 2.

Number and % of rats presenting ischemic complications after respective treatment for 2 and 4 weeks

Treatment Period	Ischemia+Saline (n=10)		Ischemia-nMSCs (n=10)		Ischemia+CD146+ MSCs (n=10)
	Number	%	Number	%	Number	%
After 2 weeks
Auto-amputation	1	10	0	0	0	0
Necrotic lower extremity	1	10	0	0	0	0
Toe(s) gangrene	3	30	2	20	1	10
Normal appearance	5	50	8	80	9	90
After 4 weeks
Auto-amputation	1	10	0	0	0	0
Necrotic lower extremity	1	10	0	0	0	0
Toe(s) gangrene	3	30	1	10	0	0
Normal appearance	5	50	9	90	10	100

3.1. Ischemia+CD146+MSC group scored highest in clinical score and Tarlov score

The different rat groups were examined for their ischemic clinical score and Tarlov score (Fig. 2). The clinical score was in sham control group > ischemia group < ischemia+nMSCs group < ischemia+^CD146+MSCs group with a score of 5.0, 2.0±0.94, 3.1±0.57 and 3.8±0.42, respectively. The clinical score was markedly reduced in the ischemia group versus sham control group, and was significantly improved in ischemia+nMSCs and ischemia+^CD146+MSCs versus the ischemia group. Furthermore, the clinical score was significantly improved in ischemia+^CD146+MSCs group versus ischemia+nMSCs group. Using a different scale for ischemia, Tarlov score was in sham control group > ischemia group < ischemia+nMSCs group < ischemia+^CD146+MSCs group with a score of 6.0, 2.1±0.57, 3.1±0.74 and 3.9±0.88, respectively. Tarlov score was significantly decreased in the ischemia versus sham control group, and was significantly improved in ischemia+nMSCs and ischemia+^CD146+MSCs versus the ischemia group. Additionally, Tarlov score was significantly improved in ischemia+^CD146+MSCs versus ischemia+nMSCs group (Fig. 2).

Fig. 2.

Ischemic clinical score (A) and Tarlov score (B) in sham control, ischemia, ischemia+nMSCs and ischemia+^CD146+MSCs rats. The scores were evaluated for normality. Data are presented as means±SD, n=5–10/group. **P < 0.01 vs Ischemia, *P < 0.05 vs Ischemia+nMSCs.

3.2. Ischemia+CD146+MSC group showed higher muscle capillary density

Muscle tissue sections and photomicrographs were quantified for measurement of muscle capillary density in the different groups (Fig. 3). The muscle capillary density was in sham control group (868.6±13.01/mm²) > ischemia group (360.5±13.57/mm²) < ischemia+nMSCs group (533.6±23.05/mm²) < ischemia+^CD146+MSCs group (552.6±13.34/mm²). Muscle capillary density was significantly decreased in the ischemia group versus sham control group, and was significantly greater in ischemia+nMSCs and ischemia+^CD146+MSCs versus the ischemia group. Importantly, the muscle capillary density was significantly greater in ischemia+^CD146+MSCs group versus ischemia+nMSCs group (Fig. 3).

Fig. 3.

Muscle capillary staining and micrographs in sham control, ischemia, ischemia+nMSCs and ischemia+^CD146+MSCs rats (A-D). Capillary numbers and density were measured and presented as means±SD (E). n=5–10/group. **P < 0.01 vs Ischemia, *P < 0.05 vs Ischemia+nMSCs.

3.3. CD146+MSCs suppressed skeletal muscle apoptosis more remarkably than nMSCs

Muscle tissue sections and photomicrographs from the different rat groups were assessed for apoptosis using the TUNEL assay (Fig. 4). The percentage of TUNEL positive muscle cells was in sham control group (10.40±1.12%) < ischemia group (25.70±1.85%) > ischemia+nMSCs group (19.67±1.65%) > ischemia+^CD146+MSC group (17.52±1.31%). The sham control group showed the least TUNEL positive apoptotic cells when compared to other groups. Apoptotic muscle cells were significantly more abundant in the ischemia group than sham control group, and were significantly less in ischemia+nMSCs and ischemia+^CD146+MSCs than the ischemia group. Furthermore, the apoptotic muscle cells were less in ischemia+^CD146+MSCs group than the ischemia+nMSCs group (Fig. 4).

Fig. 4.

TUNEL apoptosis assay and micrographs in sham control, ischemia, ischemia+nMSCs and ischemia+^CD146+MSCs rats (A-D). Arrows illustrate apoptotic nuclei stained brown by TUNEL. The number of TUNEL positive cells was measured as % of total cells in the field (E). Data are presented as means±SD, n=5–10/group. **P < 0.01 vs Ischemia, *P < 0.05 vs Ischemia+nMSCs.

3.4. Ischemia+CD146+MSCs group gained most powerful muscle contractile force

Muscle contractile force was compared in the nerve-muscle preparation from the different groups (Fig. 5). The muscle contractile force was in sham control (0.88±0.03 N) > ischemia (0.50±0.04 N) < ischemia+nMSCs (0.60±0.03 N) < ischemia+^CD146+MSCs group (0.63±0.03 N). The muscle contractile force was significantly decreased in ischemia group versus sham control group, and was significantly improved in ischemia+nMSCs group and ischemia+^CD146+MSCs group. The muscle contractile force was also significantly greater in ischemia+^CD146+MSCs group compared with ischemia+nMSCs group (Fig. 5).

Fig. 5.

Muscle contractile force in sham control, ischemia, ischemia+nMSCs and ischemia+^CD146+MSCs rats. Data are presented as means±SD, n=5–10/group. **P < 0.01 vs Ischemia, *P < 0.05 vs Ischemia+nMSCs.

3.5. Ischemia+CD146+MSCs group expressed higher VEGF than ischemia+nMSCs group

Muscle VEGF mRNA expression was relatively and significantly higher in the ischemia group than the sham control group, and was further greater in ischemia+nMSCs group and ischemia+^CD146+MSCs group than the ischemia group and sham control group. Muscle VEGF mRNA expression was also significantly greater in ischemia+^CD146+MSCs than ischemia+nMSCs group (Fig. 6).

Fig. 6.

Muscle VEGF mRNA expression in sham control, ischemia, ischemia+nMSCs and ischemia+^CD146+MSCs rats. Data are presented as means±SD, n=5–10/group. **P < 0.01 vs Ischemia, *P < 0.05 vs Ischemia+nMSCs.

4. DISCUSSION

The main findings of this study are: 1) Treatment with nMSCs and particularly ^CD146+MSCs improved ischemia and clinical and Tarlov scores in rat ischemic limb, 2) Treatment with nMSCs and particularly ^CD146+MSCs improved muscle capillary density and contractile force in ischemic limb, 3) Treatment with nMSCs and particularly ^CD146+MSCs was associated with decreased apoptotic cells and increased VEGF mRNA expression in muscle of ischemic limb. The nMSCs and ^CD146+MSCs induced improvement of vascularity, muscle function and VEGF expression and reduction in apoptosis could be underlying mechanisms for the improved clinical and Tarlov scores in the ischemic limb.

Data from the peripheral arterial disease awareness risk and treatment: new resources for survival program (PARTNERS) have shown that PAD is more common in high-risk 50–69 y old individuals with a history of cigarette smoking or diabetes, or over 70 y old without additional risk factors, with a prevalence rate of approximately 29% in this group of the population [1]. Because of the high surgical risks in elderly individual, extensive research has been conducted to develop alternative less-invasive approaches in elderly subjects with PAD. Different bone marrow, peripheral blood, and placenta-derived cells have been tested and have shown variability in their differentiation into vascular cells in different models of limb ischemia. For instance, bone marrow-derived stromal cells were found to promote arteriogenesis in a murine hind-limb ischemia model [45]. Also, bone marrow mononuclear cells transplantation was shown to augment neovascularization in rat hind-limb ischemia [46], and to induce functional angiogenesis in mouse ischemic hind-limb [47]. Transplantation of human placenta-derived MSCs was also found to alleviate critical limb ischemia in diabetic nude rats [48]. However, in the mouse model of hind-limb ischemia, bone marrow-derived cells may not promote vascular growth by incorporating into vessel walls, and instead function as supporting cells such as fibroblasts, pericytes, and leukocytes [49]. Stem cells therapy has been proposed as potential therapeutic approach in ischemic conditions of the heart [8, 9], kidney [10, 11], and brain [12, 13]. In a previous study, we have shown that MSCs derived from peripheral blood could promote angiogenesis in aged rats with chronic limb ischemia [16]. MSCs derived from bone marrow have also been extensively and frequently studied. MSCs phenotype is often characterized by being positive for specific biomarkers such as Sca-1, CD13, CD29, CD44, CD90, CD105, CD146 and CD166 and negative for CD10, CD11b, CD14, CD34, CD45, CD59b [44, 50, 51]. In the present study, the isolated and cultured nMSCs were positive for Sca1, CD29, CD44, CD105, CD90 and negative for CD14, CD34, CD45, which are generally in accordance with most of the reported markers [52]. Importantly, higher cell surface expression of CD146 is often associated with greater differentiation potential of nMSCs [52]. Lee and coworkers have used the fluorescence-activated cell sorting enrichment method and found that the expression level of CD146 is positively correlated with the growth rate and angiogenic capability of nMSCs [53]. CD146⁺ cells are also critical for the development of kidney vasculature, and depletion of CD146⁺ cells abolishes endothelial cell regeneration in the embryonic kidney [54]. The ^CD146+MSCs used in the present study showed the highest CD146 mRNA expression, supporting their significant differentiation potential.

We examined the potential benefits of treatment with ^CD146+MSCs versus nMSCs in a rat model of hind-limb ischemia. We utilized two well established scales to assess the different signs of limb ischemia in non-treated and MSCs-treated limb ischemia model, namely the clinical and Tarlov scores. The ischemic model showed successfully severe ischemic manifestations as determined by the low clinical and Tarlov scores. We compared the effectiveness of nMSCs and ^CD146+MSCs in improving the manifestations of hind-limb ischemia. The clinical score and Tarlov score were both higher in nMSCs and ^CD146+MSCs treated than non-treated ischemia group, suggesting that both MSCs therapies were effective. Importantly, the ischemia+^CD146+MSCs group scored higher than ischemia+nMSCs group in both the clinical and Tarlov scores, suggesting that the ischemia+^CD146+MSC group may have better vascularization and blood supply to the ischemic limb. This is consistent with the present observation that the muscle capillary density was markedly reduced in ischemia group versus sham control rats, and that treatment with nMSCs and ^CD146+MSCs improved capillary density in the rat ischemic muscle. Importantly, ^CD146+MSCs caused greater increases in muscle capillary density than nMSCs, suggesting that ^CD146+MSCs are more beneficial than nMSCs in promoting vascularization of the ischemic limb.

Adequate blood supply to the lower limb is critical for the muscles to perform their functions during standing, walking and exercise. Some of the common manifestations of PAD and limb ischemia are intermittent caludications and muscle pain at rest. Chronic limb ischemia and hypoxia in mice impair vascular reactivity and functional vasodilation of feed arteries in response to metabolic demands [55]. Consistent with these reports the present study showed that muscle contraction was reduced in the ischemic limb, supporting a relationship between muscle ischemia and reduced muscle function. Importantly, treatment with nMSCs and ^CD146+MSCs significantly improved contraction in the ischemic limb muscle. Notably, ^CD146+MSCs caused greater improvement in contraction of the ischemic limb muscle when compared to nMSCs, suggesting that ^CD146+MSCs induced improvement in muscle vascularization is more beneficial than nMSCs in improving muscle function. MSCs have shown the capacity to differentiate into vascular cells and tissues [56–59], which could enhance angiogenesis and muscle vascularity and in turn improve functional vasodilation in response to metabolic demands. In this respect, the ^CD146+MSCs subset appears to be more efficient than nMSCs in promoting muscle vascularity and function, consistent with reports that higher cell surface expression of CD146 is often associated with greater differentiation potential of MSCs [52]. These findings should be interpreted with caution, as force measurement was taken right after tetanic contraction occurred, when the contractile force could also be determined by other factors such as high energy phosphate concentration, glycogen storage and mitochondrial function in the muscle, and the differential effects of ^CD146+MSCs versus nMSCs on these parameters should be examined in future studies.

An important question is how MSCs improve muscle vascularization and function in the ischemic limb. Studies have shown increases in VEGF in critical limb ischemia [20–23], likely through ischemia/hypoxia-induced release of hypoxia inducible factor-1α (HIF-1α) and induction of VEGF expression [60]. Consistent with these reports, we found that VEGF expression was enhanced in ischemia vs sham control group. Of note, depending on the specific tissue environment, MSCs have shown the potential of differentiation into tissue-specific functional cells when stimulated by fibroblast growth factor-2, hepatocyte growth factor, brain derived-neurotrophic factor, or VEGF [61]. For instance, during angiogenesis, perivascular cells secrete pro-angiogenic VEGF, and MSCs differentiate into endothelial cells upon stimulation with VEGF [61–63]. The expression of VEGF and its receptors can also be regulated by microvascular endothelial cells [64]. In the present study, the expression of VEGF was upregulated after hind-limb ischemia and further upregulated following treatment with nMSCs and ^CD146+MSCs. This is likely because limb ischemia is believed to increase the release of VEGF [20–23, 60], which would stimulate MSCs differentiation into vascular cells, increase the capacity of the ischemic limb to produce VEGF, and in turn promote further MSCs differentiation into vascular cells and VEGF production, thus creating a VEGF positive loop that leads to further increases in VEGF expression. Importantly, VEGF mRNA expression was also higher in the ischemia+^CD146+MSCs group than the ischemia+nMSCs group. The critical role of CD146 is supported by reports that the ability for sprouting, migration and tube formation in response to VEGF treatment is impaired in endothelial cells of CD146 knockout mice [65]. Knockdown of CD146 has also been shown to reduce the migration and proliferation in human endothelial cells [24]. Interestingly, CD146 has been suggested as a co-receptor for VEGFR-2 to enhance VEGF-induced signal transduction and angiogenesis [66]. Taken together, the present findings and other reports suggest that the significantly higher expression of VEGF in the ischemia+^CD146+MSCs could promote further MSCs differentiation into vascular endothelial cells, resulting in higher capillary vessel density, and improved muscle function.

Patients with PAD often show muscle weakness and atrophy, and a decline in muscle strength and stamina in the affected limb [67–70]. Also, mouse model of chronic limb ischemia and hypoxia shows decline of skeletal muscle fibers and strength [71]. Consistent with these reports, the present study showed decreased muscle contraction in the ischemic limb, and improved muscle contraction in the ischemic limb treated with nMSCs and ^CD146+MSCs. This could be partly related to the ability of MSCs to differentiate into skeletal muscle fibers and myofibrils [72–75], and in turn improve muscle function and strength. Of note, ^CD146+MSCs caused better improvement in muscle contraction than nMSCs, consistent with the greater capacity of ^CD146+MSCs to differentiate into tissue-specific cells when compared to nMSCs [52]. We should note that the muscle vascularization and skeletal muscle function were measured at one time point 8 weeks following induction of hind-limb ischemia in rats. While the observed beneficial effects of treatment of ischemic limb with nMSCs and ^CD164+MSCs can be ascribed to improvement in both muscle vascularization and skeletal muscle function, the current data could not identify or estimate which effect was primary and more predominant. Future time course studies at different time points of ischemia and treatment with MSCs should help to determine if the improved vascularization or muscle function is the primary and predominant factor in the improved clinical scores observed in rats with hind-limb ischemia and treated with MSCs.

Ischemia has also been associated with cell apoptosis in different tissues and cells [76–81]. Consistent with these reports, the present TUNEL measurements suggest increased muscle apoptosis and apoptotic nuclei in the ischemic limb. We should note that TUNEL assay measures DNA fragmentation, and as several types of cell death cause DNA fragmentation, the use of the TUNEL DNA fragmentation assay as a measure of cell apoptosis should be interpreted with caution. Bone marrow-derived stromal cells were found to improve limb function, reduce the incidence of autoamputation, and attenuate muscle atrophy in a murine hindlimb ischemia model [45]. Also, MSCs have been shown to reduce cell apoptosis in different tissues under ischemia and other stresses [82–85]. Paracrine effects mediated by the transplanted MSCs may account for their angiogenic and anti-apoptotic effect rather than the actual cells. For instance, it is shown that angiogenic and anti-apoptotic capacities of MSCs are mediated through secretion of chemokine C motif ligand (XCL1) and activation of nuclear factor-κB (NFκB) [86, 87]. Hence, an alternative explanation of the improved muscle function in the MSCs-treated ischemic limb could be related to changes in muscle apoptosis. This is supported by the present observation that muscle apoptosis was reduced in nMSCs and ^CD146+MSCs treated ischemic limb. Consistent with the observation that ^CD146+MSCs caused better improvement in muscle function than nMSCs, ^CD146+MSCs caused greater reduction in muscle apoptosis in the ischemic limb than nMSCs, which would be translated into better protection of muscle fibers and enhancement of their strength and function. This is consistent with reports that ^CD146+MSCs are more efficient in protecting injured hearts and increasing the life span of muscular dystrophy mouse model than nMSCs [28, 29]. Of note, studies have shown muscle apoptotic response to denervation, disuse, and aging [88, 89]. Therefore, hind-limb ischemia and consequent reduction in the limb movement could lead to skeletal muscle atrophy, increased TUNEL positive cells, and decreased muscle strength. Therefore, the observed reduction in TUNEL positive cells and the improvement in muscle strength could be due to an indirect effect of MSCs-induced decrease in muscle atrophy. As we observed marked muscle apoptosis in the ischemic limb, muscle atrophy is expected to be in direct proportion to muscle apoptosis. Future morphological and histological studies should determine quantitatively the effects of ischemia and treatment with nMSC and ^CD146+MSCs on muscle atrophy in the hind-limb ischemia model.

Atherosclerotic PAD is a major cause of morbidity, and likewise could be contributing to cause of death. Earlier reports have shown that PAD is more prevalent in men than in women [90, 91]. These reports were based on the presence of symptoms, such as intermittent claudication, to diagnose PAD [90]. However, PAD could be asymptomatic or present with atypical symptoms in many subjects [92]. In effect, the use of a more sensitive noninvasive tests, such as the ankle-brachial index, has shown a 5-fold increase in the detection of cases than those detected by a history of intermittent claudication [90]. Of note, these more sensitive diagnostic approaches have reported a prevalence of PAD among women between 3% and 29% [90], at least the same as in men, or even higher in some cases [93]. Other population-based studies showed that the mean prevalence of PAD was 15.6% in women and 13.4% in men [94]. The present study was performed on male rats with induced limb ischemia. Whether nMSCs and specifically ^CD146+MSCs would improve angiogenesis and muscle function in female rats with limb ischemia should be examined in future investigations.

In conclusion, treatment with nMSCs and particularly ^CD146+MSCs reduces apoptosis, and improves vascularization, muscle contraction and VEGF expression in rat ischemic hind-limb. Some clinical trials have shown promising effects of autologous transplantation of bone-marrow cells in the treatment of critical limb ischemia. [95]. Also, peripheral blood mononuclear cells harvested and transplanted directly into the human ischemic limb improved walking distance and pain scale [47]. Autologous transplantation of peripheral blood stem cells was also proposed as an effective approach for severe arteriosclerosis obliterans of lower extremities [96]. A phase I/II clinical trial has shown that allogeneic bone marrow-derived MSCs are safe when injected IM at a dose of 2 million cells/kg body weight, with some efficacy parameters such as ankle-brachial pressure index and ankle pressure showing positive trend in critical limb ischemia [97]. The observed beneficial effects of nMSC and ^CD146+MSCs on improving vascularization and muscle function in the hind-limb ischemia model lay the ground for future clinical studies to test the effects of nMSC and ^CD146+MSCs in human critical limb ischemia and PAD. ^CD146+MSCs therapy may be a promising alternative approach to invasive surgery in PAD patients at high risk of surgical intervention.

List of Abbreviations:

MSCs

Mesenchymal Stem Cells

nMSCs

naive MSCs

PAD

Peripheral Arterial Disease

VEGF

Vascular Endothelial Growth Factor