Functional expression of smooth muscle-specific ion channels in TGF-β₁-treated human adipose-derived mesenchymal stem cells

Materials.

Phosphate-buffered saline, trypsin, α-MEM, and fetal bovine serum were purchased from Invitrogen (Carlsbad, CA). Carbachol was purchased from BIOMOL (Plymouth Meeting, PA). Rat tail collagen (type I) was purchased from BD Biosciences (Bedford, MA). BayK 8644, nicardipine, and iberiotoxin were purchased from Tocris Bioscience (Ellisville, MO). Anti-α-SMA (A2547) and anti-calponin (C2687) antibodies were purchased from Sigma-Aldrich (St. Louis, MO). Antibodies against myocardin (ab22621), anti-smoothelin-B (ab8969), anti-h-caldesmon (ab50016), anti-SM-MHC (ab53219), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were purchased from Abcam (Cambridge, MA). Peroxidase-labeled secondary antibodies were purchased from Amersham Biosciences.

Culture of hASCs.

Adipose tissue was obtained from elective surgeries with patient's informed written consent as approved by the Institution Review Board of Pusan National University Hospital. Culture of hASCs was performed in the same way as described previously (42). Briefly, we washed liposuction tissues more than three times with phosphate-buffered saline and mixed them with an equal volume of collagenase (type I) suspension (1 g/l buffered saline solution buffer with 1% bovine serum albumin) for 60 min at 37°C with shaking. Floating adipocytes were separated from the stromal-vascular fraction by centrifugation (300 g for 5 min). The cellular pellet was resuspended in MEM supplemented with 100 U/ml penicillin, 10% fetal bovine serum, and 100 μg/ml streptomycin, and cells were plated in dishes for tissue culture at 3,500 cells/cm². The primary hASCs were cultured for 4 or 5 days until they reached confluence, defined as passage “0”. hASCs ware used between passages 3 and 10 in our experiments. The hASCs were positive for CD29, CD44, CD73, CD90, and CD105 and negative for CD31, CD34, and CD45. For TGF-β₁-induced differentiation of hASCs into SMCs, hASCs were seeded onto six-well culture plates at 70% confluence, and they were treated with serum-free α-MEM in the absence or presence of 2 ng/ml TGF-β₁ for 96 h. Differentiation of hASCs into SMCs was characterized by Western blotting and electrophysiological analysis. For RT-PCR analysis of TGF-β₁-induced expression of smooth muscle-specific genes, hASCs were treated with serum-free α-MEM in the absence or presence of 2 ng/ml TGF-β₁ for 24 or 96 h.

Electrophysiological recordings.

Electrophysiological recording was performed in a whole cell configuration using an Axopatch 1C amplifier (Axon Instruments, Union, CA) and digital interface (NI-DAQ 7; National Instruments, Union, CA). All experimental parameters were controlled by Patchpro software, which were developed by our group. The patch pipettes (Clark Electromedical Instruments, Pangbourne, UK) were pulled with two-step vertical puller (PP-83, Narishige, Tokyo, Japan) and had tip resistances of 3–4 MΩ with the filling of the pipette solution. The recorded signals were sampled at a rate of 1∼3 kHz and filtered at 0.5∼1.0 kHz.

The extracellular normal Tyrode solution for recordings of BK_Ca and K_v channels contained the following (in mM): 140 NaCl, 5.4 KCl, 1.8 CaCl₂, 0.33 NaH₂PO₄, 0.5 MgCl₂, 5 HEPES, and 16.6 glucose, adjusted to pH 7.4 with NaOH. The pipette-filled solution for recordings of BK_Ca channels contained the following (in mM): 115 K-aspartate, 5 NaCl, 25 KCl, 2 MgCl₂, 3 Mg-ATP, and 10 HEPES, adjusted to pH 7.25 with KOH. The pipette-filled solution for recordings of K_v channels contained the following (in mM): 105 K-aspartate, 5 NaCl, 25 KCl, 2 MgCl₂, 3 Mg-ATP, 10 BAPTA, and 10 HEPES, adjusted to pH 7.25 with KOH.

The Ba²⁺-sensitive and voltage-dependent Ca²⁺ currents were recorded using nystatin-perforated patch clamp technique. For the perforated-patch recordings of Ca²⁺ currents, the cells bathed in a solution contained the following (in mM): 120 NaCl, 5 CsCl, 10 BaCl₂, 0.5 MgCl₂, 5 TEA-Cl, 10 HEPES, 10 glucose, and 0.01 BayK 8644, adjusted to pH 7.4 with NaOH. The pipette-filled solution contained the following (in mM): 130 CsCl, 10 EGTA, 10 HEPES, and 5 Mg-ATP, adjusted to pH 7.2 with CsOH. Nystatin (200 μg/ml) was added to the pipette solution every 2 h.

RT-PCR.

Total cellular RNA was extracted using the mRNA isolation system (Novagen, Darmstadt, Germany), following the manufacturer's guideline. For the RT-PCR analysis, 2-μg RNA aliquots were subjected to cDNA synthesis with 200 U MMLV reverse transcriptase (Invitrogen) and 0.5 μg oligo (dT) 15 primer (Promega, Madison, WI). For the PCR reaction, 1 μl synthesized cDNA was used as a template. PCR primers were used to amplify human BK_Ca, Ca²⁺, and K_v channels. The thermal cycle profile was as follows: 30-s denaturation at 95°C, 45-s annealing at 46–65°C, and extension for 45 s at 72°C with 0.5 U of GoTaq DNA polymerase (Promega, Madison, WI). This step was followed by extension step of 72°C for 10 min. For semiquantitative evaluation of expression levels, each PCR was performed for 30 cycles. PCR products were analyzed by 1.2% agarose gel electrophoresis and ethidium bromide staining.

Collagen lattice contraction model.

Cell contractility was measured by trypsinized cells in monolayer cultures by treatment with trypsin-EDTA contained solution. Then, cells were counted and resuspended in α-MEM at a density of 1 × 10⁶ cells/ml. After the cell suspension was mixed with collagen gel solution to make an experimental concentration of 3 mg collagen/ml and 4 × 10⁵ cells/ml, we added the cell on the collagen mixture into 12-well culture plates (Nunc, Roskilde, Denmark). Under standard culture conditions, plates were incubated for 1 h to make collagen cell lattices and serum-free α-MEM was added into the plates. To initiate collagen gel contraction, lattices were added from the bottom of the culture dishes by gentle agitation at the lattice-dish interface after the addition of activators or inhibitors. The extent of gel contraction was analyzed by detecting the dimensions of the lattice using Scion Image software (compliments of Scion) (23).

Western blot.

To prepare membrane fractions, cells were homogenized in lysis buffer containing the following (in mM): 150 NaCl, 20 Tris·HCl, 50 NaF, 10 EDTA, and 25 NaVO₄, and centrifuged at 12,000 rpm for 10 min. Extract was ultra-centrifuged again at 100,000 g for 1 h to separate the plasma membrane fraction. The pellet was homogenized in lysis buffer containing the following (in mM): 150 NaCl, 20 Tris·HCl, 50 NaF, 10 EDTA, 25 NaVO₄, and 1% Triton X-100. Equal amounts of membrane proteins (30 μg) were separated by 8% SDS-PAGE and were electrophoretically transferred to nitrocellulose membranes. The blocked membrane was incubated with antibodies to Ca_v1.2 (Sigma), Ca_v3.1 (Alomone Laboratories), Ca_v3.2 (Alomone Laboratories), Ca_v3.3 (Chemicon International), Ca_vβ₁ (Santa Cruz Biotechnology, Santa Cruz, CA), BK_Caα (BD Biosciences, San Diego, CA), BK_Caβ (Abcam, Cambridge, MA), K_v1.2 (Abcam), K_v1.5 (Santa Cruz Biotechnology), K_v3.2 (Santa Cruz Biotechnology), and secondary antibodies, a goat anti-mouse IgG (Abcam) for BK_Caα and rabbit anti-goat IgG (Abcam) for Ca_v1.2, Ca_v3.1, Ca_v3.2, Ca_v3.3, Ca_vβ₁, BK_Caβ, K_v1.2, K_v1.5, and K_v3.2. Images were captured using a Fuji Imager LAS3000 at 75% of total resolution 10- to 60-s exposure and Multi-Gauge V2.3 system.

Hindlimb ischemia, cell transplantation, and blood flow measurement.

Athymic nude mice (age 8–10 wk and weighing 17–22 g) were anesthetized with 160 mg/kg pentobarbital injection. The femoral artery was excised from its proximal origin. After arterial ligation, mice were divided by following experimental groups: the TGF-β₁-treated hASCs, the serum-starved hASCs, and the vehicle group (saline buffer; n = 8 per each group). In each animal, 3 × 10⁶ cells (200 μl) or medium (DMEM) was injected intramuscularly into four points of the gracilis muscle located in the medial thigh. The extent of necrosis in ischemic hindlimb was recorded on day 28 after surgery. The scores of necrosis were evaluated as follows: 0, limb salvage; 1, toes amputation; 2, foot amputation; and 3, limb amputation. Blood flow of the normal and ischemic limb was measured by laser Doppler perfusion imaging (LDPI) analyzer (Moor instruments, Devon, UK) on days 0, 7, 14, 21, and 28 after hindlimb ischemia. The perfusion of the nonischemic and ischemic limb was presented as colored histogram pixels. Red and blue colors indicate high and low perfusion, respectively. Blood perfusion is expressed as the LDPI index representing the ratio of ischemic vs. nonischemic limb blood flow. A ratio of 1 before operation indicates equal blood perfusion of both legs.

Histological and immunofluorescence analyses.

After blood flow measurement was completed over 28 days, hindlimb muscles were removed, formalin-fixed, and paraffin-embedded. Tissues were sliced into 5-μm sections, and endothelial cells and SMCs were identified by immunostaining with rat anti-CD31 (550274; BD Pharmingen) and rabbit anti-α-SMA (ab5694; Abcam) antibodies and then followed by incubation with Alexa 488 goat anti-rat and Alexa 488 goat anti-rabbit secondary antibodies, respectively. The specimens were washed and mounted in Vectashield medium (Vector Laboratories) with DAPI for visualization of nuclei. The stained sections were visualized by laser scanning confocal microscopy (Olympus FluoView FV1000). Capillary density was assessed by counting the number of CD31-positive features per high-power field (×400). Nine mice from each experimental group were analyzed. Twelve randomly selected microscopic fields from three serial sections in each block were examined for the existence of capillary endothelial cells.

Data analysis and statistics.

Origin 6.0 software (Microcal Software, Northampton, MA) was used for data analysis. For statistical analysis of the data, Student's t-test, two-way ANOVA, and Scheffé's post hoc test were used to test for statistical significance. The results are expressed as means ± SE. We considered P < 0.05 to be statistically significant.

TGF-β₁-induced differentiation of hASCs to SMCs.

To test whether hASCs could be differentiated into SMCs by TGF-β₁, we exposed hASCs to TGF-β₁ for 4 days. As shown in Fig. 1, A and B, the expression levels of SMC-specific contractile proteins (α-SMA, calponin, SM-MHC, smoothelin-B, myocardin, and h-caldesmon) and smooth muscle-specific transcription factor myocardin dramatically increased after treatment with TGF-β₁. Although we detected most of the proteins in various SMCs, including visceral and vascular SMCs, vascular SMCs predominantly expressed smoothelin-B (102 kDa); visceral SMCs express a different isoform: smoothelin-A (59 kDa) (23, 59). We detected only smoothelin-B after the TGF-β₁-induced differentiation of hASCs into SMCs, suggesting that the differentiated SMCs were vascular SMCs.

Collagen lattice contraction assay is useful for measuring the contractile force of cells (23). As shown in Fig. 1, C and D, treating the hASCs with TGF-β₁ to differentiate them into vascular SMCs reduced the size of the collagen lattices in a time-dependent manner. Undifferentiated hASCs and/or pretreated with the TGF-β antagonist SB431542 (10 μM) did not contract. We also evaluated the contractility of TGF-β₁-induced, differentiated hASCs after treatment with a contractile agonist (1 μM carbachol) and membrane depolarization (60 mM KCl). As shown in Fig. 1E, carbachol or high KCl strongly shrank the collagen lattices of the TGF-β₁-induced, differentiated hASCs compared with undifferentiated hASCs. These results show that TGF-β₁-induced, differentiated hASCs exhibited the characteristics of SMCs.

Expression of voltage-dependent Ca²⁺ channels on TGF-β₁-induced differentiation of hASCs.

Increased Ca²⁺ influx through voltage-dependent Ca²⁺ channels plays a crucial role in the contractile response of SMCs (1). To test whether L-type Ca²⁺ channels were functionally expressed during the TGF-β₁-induced differentiation of hASCs, we applied the patch-clamp technique to record the L-type Ca²⁺ channels directly. Figure 2 shows the Ba²⁺-sensitive inward current in the presence of the L-type Ca²⁺ channel agonist BayK 8644 (10 μM) recorded during the TGF-β₁-induced differentiation of hASCs (Fig. 2, B and D) compared with the lack of inward current in the undifferentiated hASCs (Fig. 2, A and D). The inward current was almost abolished by the application of the L-type Ca²⁺ channel inhibitor nicardipine (1 μM; Fig. 2B). These results strongly suggest that L-type Ca²⁺ currents are highly expressed during the TGF-β₁-induced differentiation of hASCs. Furthermore, the recorded L-type Ca²⁺ channel amplitude and density were very similar to those of native vascular SMCs (Fig. 2, C and D), demonstrating that the TGF-β₁-induced, differentiated hASCs shared the characteristics of vascular SMCs. To further evaluate the molecular identity of the observed functional Ca²⁺ currents, we investigated the expression patterns in undifferentiated hASCs and TGF-β₁-induced, differentiated hASCs with specific primers (Table 1). Vascular SMCs express the major L-type (Ca_v1.X, specifically Ca_v1.2) Ca²⁺ channels (43) and a small portion of the T-type (Ca_v3.X) Ca²⁺ channels. Consistent with these facts, as shown in Fig. 3, A and B, the Ca_v1.2 and all Ca_v3 subtypes dramatically increased along with the Ca_vβ₁ and Ca_vβ₃ subtypes in TGF-β₁-induced, differentiated hASCs. Western blotting confirmed the increased representative Cav1.2, Ca_v3.1, Ca_v3.2, Ca_v3.3, and Ca_vβ₁ subtypes during the TGF-β₁-induced differentiation of hASCs (Fig. 3, C and D).

Expression of BK_Ca channels during TGF-β₁-induced differentiated hASCs.

In vascular smooth muscle, the major outward currents activated at positive voltage are composed of BK_Ca and K_v channels (48). To identify the expression of BK_Ca channels, we applied iberiotoxin (100 nM), a specific BK_Ca channel inhibitor, on recorded outward current in hASCs and TGF-β₁-induced, differentiated hASCs. Figure 4 shows the lack of iberiotoxin-sensitive currents in hASCs (Fig. 4, A and D) and the abundance of noisy iberiotoxin-sensitive currents in TGF-β₁-induced, differentiated hASCs (Fig. 4, B and D). The current amplitude and electrophysiological characteristics were very similar to those of native vascular SMCs (Fig. 4, C and D), suggesting that TGF-β₁-induced hASCs differentiate into vascular SMCs. To further evaluate the molecular identity of the observed BK_Ca currents, we investigated the expression patterns of BK_Ca channels with specific BK_Ca channel primers (Table 2). As shown in Fig. 5, A and B, the TGF-β₁-induced differentiation of hASCs dramatically increased the expression levels of the major BK_Ca channel α-subtypes (K_Ca1.1) but not the SK_Ca channel α-subtypes (K_Ca2.X) and IK_Ca channel α-subtypes (K_Ca3.1). The expression levels of β-subtypes were relatively various. The expression of BK_Caβ₁, and BK_Caβ_3a subtypes were slightly decreased in TGF-β₁-induced differentiation of hASCs. However, BK_Caβ_3e and BK_Caβ₃ all subtypes were slightly increased. Consistent with the RT-PCR data, the Western blots also confirmed that the TGF-β₁-induced differentiation of hASCs increased the expression levels of the α-subtypes of BK_Ca but did not change the expression of the β-subtypes of BK_Ca (Fig. 5, C and D).

Expression of K_v channels during the TGF-β₁-induced differentiation of hASCs.

K_v channels, which are activated by membrane depolarization and allow the efflux of K⁺ to repolarize the membrane to the resting potential, are highly expressed in most vascular SMCs (31). To identify the expression of K_v channels, BK_Ca channels were inhibited by pretreatment of iberiotoxin (100 nM) in the extracellular solution and inclusion of BAPTA (10 mM), a Ca²⁺ chelator, in the pipette solution. As shown in Fig. 6, compared with the lack of K_v currents in hASCs (Fig. 6, A and E), the large K_v currents were recorded in TGF-β₁-induced, differentiated hASCs with intrinsic inactivation (Fig. 6, B and E), which is similar to native vascular SMCs (25, 48) (Fig. 6D). The general K_v channel inhibitors TEA (10 mM) and 4-aminopyridine (4-AP; 10 mM) nearly inhibited the large K_v currents recorded in the TGF-β₁-induced, differentiated hASCs (Fig. 6, C and E). These data also support that the TGF-β₁-induced, differentiated hASCs had the characteristics of vascular SMCs. Expression patterns of the K_v channel subtypes in TGF-β₁-induced, differentiated hASCs were investigated with specific primers (Table 3). As shown in Fig. 7, the TGF-β₁-induced differentiation of hASCs dramatically increased the expression of the specific subtypes known to be expressed in vascular SMCs: K_v1.1, K_v1.2, K_v1.4, K_v1.5, and K_v9.3; the expression of the K_v3.2, K_v4.3, K_v5.1, and K_v6.1 subtypes also increased (Fig. 7, A and B). The representative Western blotting of the K_v1.2, K_v1.5, and K_v3.2 subtypes also showed increased expression levels in the TGF-β₁-induced, differentiated hASCs (Fig. 7, C and D).

Transplantation of TGF-β₁-treated hASCs stimulates blood perfusion of ischemic hindlimb through stimulating angiogenesis.

To establish the therapeutic impact of TGF-β₁-induced differentiation of hASCs into SMCs, TGF-β₁-treated hASCs, serum-starved hASCs, or saline buffer were intramuscularly injected into the ischemic hindlimb, and blood flow was measured over a period of 4 wk using the LDPI analyzer. Intramuscular injection of TGF-β₁-treated hASCs into the ischemic limb significantly improved blood perfusion, as judged by the LDPI. The LDPI ratio showed significant increase in mice transplanted with TGF-β₁-treated hASCs, compared with the control groups in which serum-starved hASCs or saline buffer was injected (Fig. 8, A and B). Moreover, transplantation of TGF-β₁-treated hASCs significantly inhibited tissue necrosis and amputation as shown by reduced necrosis score 4 wk after induction of ischemia and cell transplantation, compared with control groups (Fig. 8C).

To assess whether transplantation of TGF-β₁-treated hASCs can stimulate angiogenesis in vivo, we determined densities of capillary and arterioles/arteries in ischemic muscle by immunostaining. CD31-positive capillary density was significantly higher in ischemic limb transplanted with TGF-β₁-treated hASCs than the control groups, which were injected with either serum-starved hASCs or saline buffer (Fig. 9, A and B). The densities of α-SMA-positive blood vessels were increased in ischemic limb injected with TGF-β₁-treated hASCs, compared with the control groups (Fig. 9, A and C). CD-DiI-labeled hASCs were detected in α-SMA-positive blood vessels (Fig. 9D). These findings of increased densities of CD31-positive capillary and α-SMA-positive arterioles/arteries after TGF-β₁-treated hASCs transplantation are consistent with the increased blood perfusion and reduced necrosis in the TGF-β₁-treated hASC-injected ischemic limb.