Engineering Ex Vivo–Expanded Marrow Stromal Cells to Secrete Calcitonin Gene–Related Peptide Using Adenoviral Vector

Authors

  • Weiwen Deng,

    1. Departments of Pharmacology, Tulane University Health Sciences Center, New Orleans, Louisiana, USA
    Search for more papers by this author
  • Trinity J. Bivalacqua,

    1. Departments of Pharmacology, Tulane University Health Sciences Center, New Orleans, Louisiana, USA
    2. Urology, Tulane University Health Sciences Center, New Orleans, Louisiana, USA
    Search for more papers by this author
  • Natasha N. Chattergoon,

    1. Departments of Pharmacology, Tulane University Health Sciences Center, New Orleans, Louisiana, USA
    Search for more papers by this author
  • James R. Jeter Jr.,

    1. Structural and Cellular Biology, Tulane University Health Sciences Center, New Orleans, Louisiana, USA
    Search for more papers by this author
  • Philip J. Kadowitz Ph.D.

    Corresponding author
    1. Departments of Pharmacology, Tulane University Health Sciences Center, New Orleans, Louisiana, USA
    • Department of Pharmacology, SL83, Tulane University Health Sciences Center, 1430 Tulane Avenue, New Orleans, Louisiana 70112, USA. Telephone: 504-584-2637; Fax: 504-588-5283
    Search for more papers by this author

Abstract

Calcitonin gene–related peptide (CGRP) is a target for cardiovascular gene therapy. Marrow stromal cells (MSCs) hold promise for use in adult stem cell–based cell and gene therapy. To determine the feasibility of adenoviral-mediated CGRP gene transfer into ex vivo–expanded MSCs, rat MSCs were isolated, ex vivo expanded, and transduced with adenoviruses. Adprepro-CGRP and AdntlacZ, adenoviral vectors containing prepro-CGRP or nuclear-targeted β-galactosidase reporter gene ntlacZ under the control of Rous sarcoma virus promoter, were used. In this study, it can be shown that transduction efficiency of adenoviral-mediated gene transfer into ex vivo–expanded MSCs is dose dependent, transgene expression persists for more than 21 days in culture, and adenoviral transduction does not alter the proliferation or viability of MSCs. Transduced MSCs retain multipotentiality and transgene expression after cell differentiation. The expression and secretion of CGRP by Adprepro- CGRP–transduced MSCs was confirmed by Western blot analysis and enzyme immunoassay. The secretion of CGRP by Adprepro-CGRP–transduced MSCs is dose dependent, and the transduced cells release as much as 9.5 ± 0.4 pmol CGRP/1 × 106 cells/48 hours (mean ± standard error of mean, n = 3) into culture medium at a multiplicity of infection of 300. Furthermore, culture supernatant from Adprepro-CGRP–transduced MSCs increases intracellular cyclic AMP levels in pulmonary artery smooth muscle cells in culture. These findings suggest that replication-deficient recombinant adenovirus can be used to gene engineer ex vivo–expanded MSCs and that high-level secretion of biologically active CGRP can be achieved, underscoring the clinical potential of using this novel adult stem cell–based cell and gene therapy strategy for the treatment of cardiovascular diseases.

Introduction

Calcitonin gene–related peptide (CGRP), a 37-amino-acid neuropeptide derived from alternative splicing of RNA from the calcitonin gene [1], is an important molecule with multiple biological effects, including a potent vasodilator activity [2, 3]. CGRP is extensively localized in the perivascular or periadventitia nerves throughout the body, and its release from nerve endings is associated with the dilation of blood vessels [4]. CGRP exerts its vasodilator effect through interaction with CGRP receptors that are present on both endothelial cells and vascular smooth muscle cells via endothelium-dependent or endothelium-independent mechanisms, depending on vessel type and species [57]. CGRP deficiency has been suggested to be involved in cardiovascular diseases such as pulmonary hypertension [810], erectile dysfunction [11], and cerebral vasospasm after subarachnoid hemorrhage (SAH) [12, 13]. Therefore, the enhancement of local CGRP delivery is a promising approach for the treatment of these cardiovascular disorders.

Intravenous, intracavernosal, or intrathecal administration of exogenous CGRP has been shown to have a beneficial effect in pulmonary hypertension [1416], erectile dysfunction [17, 18], and cerebral vasospasm after SAH [1921]. However, the half-life of CGRP in plasma is only 10 minutes, and chronic infusion of CGRP to patients is not feasible [22]. An alternative approach such as a gene therapy strategy should be developed to deliver CGRP over long periods of time. Although direct in vivo injection of adenovirus containing prepro-calcitonin gene-related peptide (prepro-CGRP) has been shown to be effective for the treatment of pulmonary hypertension [23, 24], erectile dysfunction [25], and cerebral vasospasm [2628], major disadvantages, such as inflammation and random transgene expression in almost all cell types, would limit the clinical application of this strategy in human diseases, and an improved therapy should be developed [24, 25, 27, 28].

Gene-engineered ex vivo–expanded adult stem cells are attractive for use in gene therapy, because some disadvantages associated with direct in vivo delivery of viral vectors, nonviral vectors, or gene-modified ex vivo–expanded differentiated cells are avoided [29]. Marrow stromal cells (MSCs) are nonhematopoietic adult stem cells from bone marrow, are relatively easy to isolate and expand ex vivo, and have multipotential differentiation capability [3035]. These cells can be used as a vehicle for gene delivery in an adult stem cell–based gene therapy strategy [36, 37]. For therapeutic gene transfer, adenoviral vectors have major advantages over other vectors such as high-level transgene expression, a broad host range, the ability to infect quiescent cells, and ease of preparation of high titer viral stock [38]. Therefore, the aim of this study was to determine whether ex vivo–expanded rat MSCs (rMSCs) can be transduced with adenovirus containing CGRP and to ascertain whether the cells still retain their multipotentiality after adenoviral-mediated gene transfer.

In this study, data are presented showing that adenoviral vectors can be used to gene engineer ex vivo–expanded rMSCs and that high-level functional CGRP secretion by adult stem cells can be achieved, pointing out the potential clinical application of this novel method for adult stem cell–based cell and gene therapy.

Materials and Methods

AdenoviralVectors

Adprepro-CGRP, a replication-deficient recombinant adenovirus carrying the human prepro-CGRP gene under the control of Rous sarcoma virus (RSV) promoter, and AdntlacZ, a replication-deficient recombinant adenovirus carrying the nuclear-targeted β-galactosidase reporter gene ntlacZ under the control of RSV promoter, were purchased from University of Iowa Gene Transfer Vector Core (Iowa City, IA) [2325, 27, 28, 39]. The Adprepro-CGRP virus has a concentration of 1.0 × 1012 viral particles (vp) per milliliter (ml) and a titer of 1.6 × 1010 plaque-forming units (pfu) per milliliter. The AdntlacZ virus has a concentration of 1.4 × 1012 vp/ml and a titer of 1.0 × 1010 pfu/ml.

Isolation and Ex Vivo Expansion of rMSCs

rMSCs were isolated as previously described [4043]. Briefly, 6-week-old male brown Norway rats (Harlan, San Diego) were euthanized with CO2. Under sterile conditions, femurs and tibias were removed and placed in the culture medium for rMSCs (α-minimal essential medium [MEM] [GIBCO Invitrogen, Grand Island, NY], 20% fetal bovine serum [FBS] [lot-selected for rapid growth of rMSCs, GIBCO Invitrogen], 100 U/ml penicillin, 100 μg/ml streptomycin, 250 ng/ml amphotericin B [Atlanta Biologicals, Nor-cross, GA], and 2 mM L-glutamine [GIBCO Invitrogen]). Both ends of femurs and tibias were removed, and the bone marrow was flushed out using a 21-gauge needle connected to a 10-ml syringe filled with the culture medium for rMSCs. The bone marrow cells were filtered through a cell strainer with 70-μm nylon mesh (BD Biosciences, Bedford, MA), and the cells from one rat were plated in a 75-cm2 tissue culture flask. The cells were incubated at 37°C with 5% humidified CO2, and rMSCs were isolated by their adherence to tissue culture plastic. Fresh culture medium for rMSCs was added and replaced to remove nonadherent cells every 2–3 days. The adherent rMSCs were grown to confluency (defined as rMSCs at passage 0), harvested with 0.25% trypsin and 1 mM EDTA for 5 minutes at 37°C, diluted 1:3, replated in culture flasks, and again grown to confluency (passage 1). The cells were harvested with trypsin/EDTA, suspended at 1 to 2 × 106 cells/ml in culture medium for rMSCs containing 10% dimethylsulfoxide and 50% FBS, and frozen as 1-ml aliquots in liquid nitrogen for storage. To expand a culture, a frozen stock of rMSCs was thawed, plated in a 75-cm2 tissue culture flask, and grown to 70%–90% confluency over approximately 3–5 days. The cells were then harvested with trypsin/EDTA and diluted 1:3 per passage for additional ex vivo expansion. rMSCs at passages 1 through 3 were used for all of the experiments.

Transduction with Adenoviral Vectors

rMSCs were plated at a density of 10,000 cells/cm2 in six-well plates and incubated overnight. The cells were counted and then exposed to fresh culture medium containing adenovirus at various multiplicities of infection (MOI; defined as pfu/cell) for 48 hours. Three separate experiments, each in triplicate, were carried out. Cell viability was determined using trypan blue exclusion method.

X-Gal Histochemistry for β-Galactosidase Activity

Cells in six-well plates were washed with phosphate-buffered saline (PBS), fixed for 5 minutes in a PBS solution containing 2% formaldehyde and 0.2% glutaraldehyde (Sigma, St. Louis), washed with PBS twice, and incubated in the X-gal staining solution (1 mg/ml X-gal, 5 mM K ferricyanide, 5 mM K ferrocyanide, and 2 mM MgCl2 [Sigma]; prepared in PBS) at 37°C in the dark overnight. Cells were washed with PBS, and the expression of ntlacZ transgene in rMSCs was evaluated by light microscopy scoring of cells expressing the nuclear-targeted β-galactosidase activity. The β-galactosidase–positive blue cells found in three microscopic fields (× 25) were then counted and expressed as a percentage of the total number of cells in those fields.

In Vitro Differentiation of rMSCs into Osteoblasts and Adipocytes

In vitro differentiation of rMSCs into osteoblast and adipocyte lineages was conducted as previously described [4043]. Briefly, cells in six-well plates were treated with culture medium for rMSCs plus either osteogenic supplement (1 × 10−5 mM dexamethasone, 0.2 mM ascorbic acid, and 10 mM β-glycerol phosphate [Sigma]) or adipogenic supplement (0.5 μM hydrocortisone, 500 μM isobutylmethylxanthine, and 60 μM indomethacin [Sigma]). The differentiation medium was changed every 3 days until day 21. To assess mineral deposition, cells were washed with PBS, fixed with cold methanol (–20°C) for 10 minutes, washed with dH2O twice, stained with 2% Alizarin red S (pH 4.1, Sigma) for 15 minutes, and washed with dH2O five times. To assess lipid droplet formation, cells were washed with PBS, fixed with 10% formalin (Sigma) for 1 hour, washed with dH2O twice, stained with a freshly prepared Oil red O solution for 15 minutes, and washed with dH2O. The Oil red O solution was prepared by mixing three parts of an Oil red O stock solution (0.5%, prepared in isopropanol, Sigma) with two parts of dH2O and filtering through a 0.45-μm pore size filter.

Western Blot Analysis for CGRP

Western blot analysis for CGRP transgene expression in rMSCs was carried out using the whole-cell lysate. Briefly, cells in six-well plates were rinsed with cold PBS (4°C); drained; lysed with a buffer containing 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 0.1 mg/ml leupeptin, and 574 μM phenylmethylsulfonyl fluoride (Sigma; prepared in PBS); and scraped into a 1.5-ml centrifuge tube. The sample was incubated for 45 minutes on ice and centrifuged at 15,000 g for 20 minutes, and the supernatant was collected. Protein content of the whole-cell lysate was quantified colorimetrically using the BCA protein assay kit (Pierce, Rockford, IL). Fifteen micrograms of whole-cell lysate were loaded onto a 4%–20% Tris-Tricine gel (Jule Biotechnology Inc., Milford, CT). After electrophoresis, the protein was transferred onto a nitrocellulose membrane with 0.2-μm pore size by electroelution. Immunodetection was performed with the rabbit anti-CGRP polyclonal antibody (Peninsula Laboratories Inc., San Carlos, CA; 1:2,500 dilution). The secondary antibody was horseradish peroxidase–conjugated to anti-rabbit immunoglobulin G (IgG) (Santa Cruz Biotechnology, Santa Cruz, CA; 1:4,000 dilution). The nitrocellulose membrane was then processed using enhanced chemiluminescence Western blotting detection reagents (Amersham Pharmacia Biotech, Piscataway, NJ). Western blot analysis for CGRP secretion by rMSCs was carried out in the same way as above, except that 3 μl of culture supernatant per lane was loaded. Western blot analysis for β-tubulin expression was conducted in the same way as above, except that 15 μg of whole-cell lysate or 3 μl of culture supernatant per lane was loaded onto a 4%–20% Tris-Glycine gel (ICN Biomedicals, Aurora, OH), the protein was then transferred onto a nitrocellulose membrane with 0.45-μm pore size, and the immunodetection was performed using rabbit anti-β-tubulin polyclonal antibody (Santa Cruz Biotechnology, 1:2,500 dilution) and horseradish peroxidase conjugated to anti-rabbit IgG (Santa Cruz Biotechnology, 1:4,000 dilution).

Enzyme Immunoassay for CGRP

For measurement of CGRP concentration in the culture supernatant of rMSCs, the enzyme immunoassay (EIA) was used. Briefly, rMSCs were transduced with Adprepro-CGRP at MOI 50, 150, and 300 or AdntlacZ at MOI 300 for 2 days. The virus-containing culture medium was removed, the cells were washed with PBS three times, and fresh culture medium was added. The cells were cultured for 48 hours, and the culture supernatant was collected. The supernatant was then assayed for its content of CGRP peptide using a competitive EIA kit (Peninsula Laboratories Inc.). The EIA data were expressed as the mean picomole of peptide/1 × 106 cells/48 hours ± standard error of the mean (SEM), with n = 3 per group.

Measurement of Intracellular Cyclic AMP

Intracellular cyclic AMP (cAMP) of rat pulmonary artery smooth muscle cells (PASMCs) was measured using a competitive EIA kit (cAMP Biotrak EIA System, Amersham Biosciences, Piscataway, NJ). Rat PASMCs were isolated as previously described [44]. Rats were euthanized with CO2, and the pulmonary artery was removed. The artery was excised and treated with a collagenase solution (200 U/ml type I collagenase and 0.4 mg/ml trypsin inhibitor [Sigma], prepared in dH2O) for 30 minutes at 37°C. The adventitia was then removed, the artery was cut longitudinally, and the endothelial lining was disrupted using a sterile cotton swab. The vessel was minced into small pieces and treated with a collagenase/elastase solution (200 U/ml type I collagenase and 15 U/ml type III elastase [Sigma], prepared in dH2O) for 2 hours at 37°C. The tissue pieces were then plated in a 25-cm2 tissue culture flask with culture medium for PASMCs (medium 199 [Sigma], 10% FBS [GIBCO Invitrogen], 100 U/ml penicillin, 100 μg/ml streptomycin, 250 ng/ml amphotericin B [Atlanta Biologicals], and 2 mM L-glutamine [GIBCO Invitrogen]) at 37°C with 5% humidified CO2. The tissue pieces were allowed to attach for 5–7 days, and the culture medium was changed. The adherent cells were grown to 70% confluency (defined as passage 0) and harvested with 0.25% trypsin and 1 mM EDTA for 5 minutes at 37°C, diluted 1:3, replated in culture flasks, and again grown to 70% confluency (passage 1). The identity of PASMCs was confirmed by the typical hill-and-valley appearance and the positive immunostaining for α-smooth muscle actin (data not shown). Rat PASMCs (passage 1) were plated at a density of 10,000 cells/cm2 in a 96-well culture plate, incubated for 40 hours, and exposed to 100 μl fresh culture medium containing 1 mM isobutylmethylxanthine, a nonselective inhibitor of cyclic nucleotide phosphodiesterase, for 15 minutes at 37°C. The medium was then replaced with 100 μl culture supernatant from rMSCs, AdntlacZ-rMSCs (MOI 300), or Adprepro-CGRP-rMSCs (MOI 300) for 15 minutes at 37°C. The cells were then lysed, and intracellular cAMP was measured using nonacetylation EIA procedure.

StatisticalAnalysis

Data were expressed as mean ± SEM and analyzed statistically using a t-test or a one-way analysis of variance followed by post-hoc analysis with Tukey's test.

Results

Adenoviral Transduction of ExVivo–Expanded rMSCs

The effectiveness of adenoviral-mediated gene transfer into ex vivo–expanded rMSCs was examined using rMSCs transduced with AdntlacZ at MOI 50, 100, 150, 200, and 300. After 48 hours, the expression of nuclear-targeted β-galactosidase was assessed by X-gal staining, and, as shown in Figure 1, transduction efficiency was observed to be >40% at MOI 50 and >90% at MOI 300. Cell viability was determined to be >95% at all MOIs studied.

Figure Figure 1..

(A–D): Photomicrographs showing β-galactosidase–positive blue nuclei in AdntlacZ-transduced rMSCs. (A): Control rMSCs. (B): rMSCs transduced at MOI 50. (C): rMSCs transduced at MOI 150. (D): rMSCs transduced at MOI 300. Magnification, ×250. (E): Efficiency of adenoviral-mediated ntlacZ gene transfer into rMSCs. rMSCs were transduced with AdntlacZ at the indicated MOI for 48 hours. The cells were X-gal stained for the nuclear-targeted β-galactosidase activity, and the transduction efficiency was determined. Each value represents mean ± standard error of the mean (n = 3). Abbreviations: MOI, multiplicity of infection; rMSC, rat marrow stromal cell.

Differentiation of AdntlacZ-Transduced rMSCs In Vitro

To ascertain whether rMSCs retain multipotential differentiation capability after adenoviral-mediated ntlacZ gene transfer, rMSCs were first transduced with AdntlacZ at MOI 300 for 48 hours and then incubated in the presence of differentiation media. As seen in Figures 2A–2D, after exposure to osteogenic medium, mineral deposition was observed. Moreover, after exposure to adipogenic medium, the cells exhibited lipid droplets. Therefore, the osteogenic potential and adipogenic potential of AdntlacZ-transduced rMSCs were retained.

Figure Figure 2..

Photomicrographs showing that AdntlacZ-transduced rMSCs retain multipotentiality in culture and the differentiated cells express β-galactosidase. rMSCs were transduced with AdntlacZ at a multiplicity of infection of 300 for 2 days. The cells were then induced to differentiate into osteoblast or adipocyte lineage for 21 days and stained with Alizarin red S for mineral deposition, Oil red O for lipid droplet formation, or X-gal for the nuclear-targeted β-galactosidase activity. (A):AdntlacZ-transduced rMSCs stained with Alizarin red S. (B):AdntlacZ-transduced rMSCs first treated with osteogenic medium for 21 days and then stained with Alizarin red S. (C):AdntlacZ-transduced rMSCs stained with Oil red O. (D):AdntlacZ-transduced rMSCs first treated with adipogenic medium for 21 days and then stained with Oil red O. (E): AdntlacZ-transduced rMSCs first treated with osteogenic medium for 21 days and then X-gal stained for β-galactosidase activity. (F): AdntlacZ-transduced rMSCs first treated with adipogenic medium for 21 days and then X-gal stained for β-galactosidase activity. Arrows indicate ntlacZ+ blue differentiated osteoblasts, and arrowheads indicate ntlacZ+ blue differentiated adipocytes. Magnification, × 500. Abbreviation: rMSC, rat marrow stromal cell.

The percentage of differentiated cells in both control rMSCs and AdntlacZ-transduced rMSCs was counted. The percentage of differentiated osteoblasts in control rMSCs and AdntlacZ-transduced rMSCs was 59 ± 2% and 57 ± 2% (mean ± SEM, n = 3; p > .05, t-test), respectively. The percentage of differentiated adipocytes in control rMSCs and AdntlacZ-transduced rMSCs was 37 ± 1% and 39 ± 1% (mean ± SEM, n = 3; p > .05, t-test), respectively. Therefore, there is no significant difference between the differentiation potential of control rMSCs and AdntlacZ-transduced rMSCs either in osteogenic medium or adipogenic medium.

The expression of ntlacZ in these differentiated cells was also assessed. The percentage of β-galactosidase–positive cells at day 21 in osteogenic medium–treated cells was 6 ± 1% (mean ± SEM, n = 3). The percentage of β-galactosidase–positive cells at day 21 in adipogenic medium–treated cells was 29 ± 1% (mean ± SEM, n = 3). As seen in Figures 2E and 2F, the differentiated osteoblasts and adipocytes were still positive for β-galactosidase.

Persistence of Adenoviral-Mediated ntlacZ Transgene Expression In Vitro

To study the persistence of ntlacZ transgene expression in vitro, rMSCs were transduced with AdntlacZ at MOI 300 for 48 hours. The virus-containing culture medium was then removed. The cells were washed with PBS three times and additionally incubated in low-serum medium (α-MEM with 2% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, 250 ng/ml amphotericin B, and 2 mM L-glutamine) for 21 days. The low-serum medium was changed every 3 days, and ntlacZ transgene expression was assessed at various time intervals after transduction. As shown in Figure 3, the number of cells expressing β-galactosidase was >90% at day 2 and >50% at day 21.

Figure Figure 3..

(A–D): Photomicrographs showing AdntlacZ-transduced rMSCs at various time intervals after transduction. (A): Control at day 2. (B): MOI 300 at day 2. (C): Control at day 21. (D): MOI 300 at day 21. Magnification, × 250. (E): Effect of time on ntlacZ transgene expression in AdntlacZ-transduced rMSCs. rMSCs were transduced with AdntlacZ at MOI 300 for 2 days. The virus-containing culture medium was removed, and the cells were washed with phosphate-buffered saline three times and further incubated in low-serum medium. The low-serum medium was changed every 3 days until day 21. The cells were X-gal stained for ntlacZ expression at days 2, 4, 7, 14, and 21, and the percentage of cells expressing ntlacZ was quantified. Each value represents mean ± standard error of the mean (n = 3). *p > .05 versus day 2. **p > .05 versus day 2 or day 4. ***p < .01 versus day 2, day 4, or day 7. ****p < .001 versus day 2, day 4, or day 7 and p > .05 versus day 14 (analysis of variance). Abbreviations: MOI, multiplicity of infection; rMSC, rat marrow stromal cell.

Western Blot Analysis for CGRP

To determine whether rMSCs can be gene engineered with CGRP, rMSCs were transduced with Adprepro-CGRP at MOI 300. CGRP expression was assessed in the whole-cell lysate from control rMSCs, AdntlacZ-transduced rMSCs, and Adprepro-CGRP–transduced rMSCs using Western blot analysis. No constitutive CGRP expression was detected in either control rMSCs or AdntlacZ-transduced rMSCs. Expression of CGRP was detected in Adprepro-CGRP–transduced rMSCs (Fig. 4A). The culture supernatants from control rMSCs, AdntlacZ-transduced rMSCs, and Adprepro-CGRP–transduced rMSCs were also analyzed for CGRP secretion. As seen in Figure 4A, high-level CGRP secretion in culture supernatant from Adprepro-CGRP–transduced rMSCs was detected, whereas there was no constitutive CGRP secretion from either control rMSCs or AdntlacZ-transduced rMSCs.

Figure Figure 4..

(A): Western blot analysis for the expression and secretion of CGRP by Adprepro-CGRP–transduced rMSCs. rMSCs were transduced with Adprepro-CGRP (MOI 300) or AdntlacZ (MOI 300) for 48 hours, and the whole-cell lysates were used for the analysis of CGRP transgene expression. Lane 1: control rMSCs; lane 2: AdntlacZ-transduced rMSCs; lane 3: Adprepro-CGRP–transduced rMSCs. Furthermore, after being transduced with either Adprepro-CGRP (MOI 300) or AdntlacZ (MOI 300) for 2 days, the cells were washed with PBS three times and incubated in fresh culture medium for 48 hours. The culture supernatant was then collected and analyzed for CGRP secretion by Western blot analysis. Lane 4: supernatant from control rMSCs; lane 5: supernatant from AdntlacZ-transduced rMSCs; lane 6: supernatant from Adprepro-CGRP–transduced rMSCs. The analysis for β-tubulin was carried out to ensure that sample loading was similar in all lanes. (B): EIA analysis for CGRP secretion by Adprepro-CGRP–transduced rMSCs. rMSCs were transduced with Adprepro-CGRP (MOI 50, 150, and 300) or AdntlacZ (MOI 300) for 2 days, the virus-containing culture medium was removed, and the cells were washed with PBS three times and further incubated in fresh culture medium for 48 hours. The culture supernatant was then collected and analyzed for CGRP secretion by EIA. Each value represents mean ± standard error of the mean (n = 3). Abbreviations: CGRP, calcitonin gene-related peptide; EIA, enzyme immunoassay; MOI, multiplicity of infection; PBS, phosphate-buffered saline; rMSC, rat marrow stromal cell.

Enzyme Immunoassay for CGRP

CGRP concentration in the culture supernatant from control rMSCs, AdntlacZ-transduced rMSCs, and Adprepro-CGRP–transduced rMSCs was measured using the EIA. As seen in Figure 4B, there was no constitutive secretion of CGRP to culture medium by either control rMSCs or AdntlacZ-transduced rMSCs. However, as much as 9.5 ± 0.4 pmol CGRP/1 × 106 cells/48 hours (mean ± SEM, n = 3) was secreted to culture medium by Adprepro-CGRP–transduced rMSCs at MOI 300, and the secretion of CGRP is dose dependent.

CGRP Secreted by Adprepro-CGRP-Transduced rMSCs Is Biologically Active

To determine whether CGRP secreted by Adprepro-CGRP-transduced rMSCs is biologically active, rat PASMCs were treated with supernatant from control rMSCs, AdntlacZ-transduced rMSCs, or Adprepro-CGRP–transduced rMSCs, and intracellular cAMP levels were measured. As shown in Figure 5, culture supernatant from Adprepro-CGRP–transduced rMSCs increased intracellular cAMP levels in PASMCs, whereas culture supernatant from control rMSCs or AdntlacZ-transduced rMSCs did not increase intracellular cAMP levels in PASMCs. Therefore, CGRP secreted by Adprepro-CGRP–transduced rMSCs is biologically active [57, 24, 25, 27].

Figure Figure 5..

Effect of culture supernatant from Adprepro-CGRP–transduced rMSCs on intracellular cAMP levels of rat PASMCs. rMSCs were transduced with Adprepro-CGRP (MOI 300) or AdntlacZ (MOI 300) for 2 days. The virus-containing culture medium was removed, and the cells were washed with phosphate-buffered saline three times. The cells were further incubated in fresh culture medium for 48 hours, and the culture supernatant was collected. Rat PASMCs were then treated with culture supernatant from control rMSCs, AdntlacZ-transduced rMSCs, or Adprepro-CGRP–transduced rMSCs, and intracellular cAMP levels of the cells were measured. Each value represents mean ± standard error of the mean (n = 3). *p > .05 versus supernatant from control rMSCs. **p < .01 versus supernatant from control rMSCs or AdntlacZ-transduced rMSCs (analysis of variance). Abbreviations: CGRP, calcitonin gene-related peptide; MOI, multiplicity of infection; PAMSC, pulmonary artery smooth muscle cell; rMSC, rat marrow stromal cell.

Persistence of Adenoviral-Mediated CGRP Transgene Expression In Vitro

To study the persistence of CGRP transgene expression in vitro, rMSCs were transduced with Adprepro-CGRP at MOI 300 for 48 hours. The virus-containing culture medium was removed, and the cells were washed with PBS three times. The cells were cultured in fresh culture medium for 48 hours, and the culture supernatant was collected. The cells were further incubated in low-serum medium (α-MEM with 2% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, 250 ng/ml amphotericin B, and 2 mM L-glutamine), and the low-serum medium was changed every 2–3 days until day 21. The culture supernatant at various time intervals after transduction was collected and analyzed for CGRP secretion by EIA. As shown in Figure 6, Adprepro-CGRP–transduced rMSCs (MOI 300) secreted 9.5 ± 0.4 and 0.02 ± 0.02 pmol CGRP/1 × 106 cells/48 hours (mean ± SEM, n = 3) to culture medium at days 2 and 21, respectively.

Figure Figure 6..

Effect of time on CGRP secretion by Adprepro-CGRP–transduced rMSCs in culture. rMSCs were transduced with Adprepro-CGRP at a multiplicity of infection of 300 for 2 days. The virus-containing culture medium was removed, and the cells were washed with phosphate-buffered saline three times. The cells were cultured in fresh culture medium for 48 hours, and the culture supernatant was collected. The cells were further incubated in low-serum medium, and the low-serum medium was changed every 2–3 days until day 21. The culture supernatant at various time intervals was collected and analyzed for CGRP secretion by enzyme immunoassay. CGRP secretion by Adprepro-CGRP–transduced rMSCs in culture at days 2, 4, 7, 14, and 21 was then determined. Each value represents mean ± standard error of the mean (n = 3). *p < .01 versus day 2. **p < .001 versus day 2 or day 4. ***p < .001 versus day 2 or day 4 and p > .05 versus day 7. ****p < .001 versus day 2 or day 4 and p > .05 versus day 7 or day 14 (analysis of variance). Abbreviations: CGRP, calcitonin gene-related peptide; rMSC, rat marrow stromal cell.

Differentiation of Adprepro-CGRP–Transduced rMSCs In Vitro

To ascertain whether rMSCs retain multipotentiality after adenoviral-mediated CGRP gene transfer, rMSCs were transduced with Adprepro-CGRP at MOI 300 for 48 hours and additionally cultured in the presence of differentiation medium for osteoblast or adipocyte lineage. After exposure to osteogenic medium for 21 days, the cells were stained with Alizarin red S and mineral deposition was observed. After exposure to adipogenic medium for 21 days, the cells were stained with Oil red O and lipid droplet formation was confirmed. Thus, both the osteogenic potential and adipogenic potential of Adprepro-CGRP–transduced rMSCs were retained.

The percentage of differentiated cells in both control rMSCs and Adprepro-CGRP–transduced rMSCs was also counted. The percentage of differentiated osteoblasts in control rMSCs and Adprepro-CGRP–transduced rMSCs was 66 ± 5% and 69 ± 9% (mean ± SEM, n = 3; p > .05, t-test), respectively. The percentage of differentiated adipocytes in control rMSCs and Adprepro-CGRP–transduced rMSCs was 22 ± 2% and 22 ± 4% (mean ± SEM; n = 3; p > .05, t-test), respectively. These data indicate that there is no significant difference between the differentiation potential of control rMSCs and Adprepro-CGRP–transduced rMSCs in either osteogenic or adipogenic medium.

The secretion of CGRP by the differentiated Adprepro-CGRP–transduced rMSCs was also assessed. The cells treated with osteogenic medium for 21 days secreted 0.02 ± 0.01 pmol CGRP/1 × 106 cells/48 hours (mean ± SEM, n = 3) in culture. The cells treated with adipogenic medium for 21 days secreted 1.47 ± 0.03 pmol CGRP/1 × 106 cells/48 hours (mean ± SEM, n = 3) in culture.

Proliferation and Viability of rMSCs during and afterAdenoviral Infection

To determine whether adenoviral transduction can alter the proliferation and viability of rMSCs during adenoviral infection, rMSCs were transduced with AdntlacZ or Adprepro-CGRP at MOI 300 for 48 hours. The proliferation and viability of transduced rMSCs were then compared with control rMSCs. As seen in Figure 7A, there is no significant difference among control rMSCs, AdntlacZ-transduced rMSCs, and Adprepro-CGRP–transduced rMSCs for either cell proliferation or viability. To ascertain whether adenoviral transduction alters proliferation and viability of rMSCs after adenoviral infection, rMSCs were transduced with AdntlacZ or Adprepro-CGRP at MOI 300 for 48 hours. The virus-containing culture medium was then removed and the cells were washed with PBS three times and additionally incubated in fresh culture medium for 48 hours. The proliferation and viability of transduced rMSCs were then compared with control rMSCs. As shown in Figure 7B, there is no significant difference among control rMSCs, AdntlacZ-transduced rMSCs, and Adprepro-CGRP–transduced rMSCs for either cell proliferation or viability. Therefore, the proliferation and viability of rMSCs were not altered by either AdntlacZ or Adprepro-CGRP transduction at MOI 300.

Figure Figure 7..

Lack of influence of AdntlacZ or Adprepro-CGRP transduction at MOI 300 on the proliferation and viability of rMSCs. (A): AdntlacZ or Adprepro-CGRP transduction at MOI 300 does not alter the proliferation and viability of rMSCs during adenoviral infection. rMSCs were transduced with AdntlacZ or Adprepro-CGRP at MOI 300 for 48 hours. Cell proliferation was assessed by counting the cells and expressing the data as percentage of control, and cell viability was determined by trypan blue exclusion method. (B):AdntlacZ or Adprepro-CGRP transduction at MOI 300 does not alter the proliferation and viability of rMSCs after adenoviral infection. rMSCs were transduced with AdntlacZ or Adprepro-CGRP at MOI 300 for 48 hours. The virus-containing culture medium was removed, and the cells were washed three times with phosphate-buffered saline and further incubated in fresh culture medium for 2 days. Cell proliferation was assessed by counting the cells and expressing the data as percentage of control, and cell viability was determined by trypan blue exclusion method. For A and B, each value represents mean ± standard error of the mean (n = 3) and p > .05 by analysis of variance. Abbreviations: CGRP, calcitonin gene-related peptide; MOI, multiplicity of infection; rMSC, rat marrow stromal cell.

Discussion

In this study, it is shown for the first time that ex vivo–expanded rMSCs can be transduced with a replication-deficient recombinant adenoviral vector containing the CGRP gene, an important candidate for cardiovascular gene therapy [24, 25, 28, 4547]. The secretion of CGRP into culture medium by Adprepro-CGRP–transduced rMSCs is dose dependent, and the transduced cells release as much as 9.5 ± 0.4 pmol CGRP/1 × 106 cells/48 hours at MOI 300. The present data also demonstrate that CGRP secreted by Adprepro-CGRP–transduced rMSCs is biologically active. The present data show that rMSCs retain multipotential differentiation capability after adenoviral-mediated CGRP gene transfer and that proliferation and viability of rMSCs were not altered by Adprepro-CGRP transduction at MOI 300. Using a replication-deficient recombinant adenoviral vector containing the reporter gene ntlacZ, it can be shown that transduction efficiency of adenoviral-mediated gene transfer into ex vivo–expanded rMSCs is dose dependent, that transgene expression persists for at least 21 days in vitro, and that the cells retain their multipotentiality after adenoviral-mediated gene transfer and express transgene even after differentiation.

It was previously reported that ex vivo–expanded human MSCs (hMSCs) can be transduced with an adenoviral vector containing the reporter gene lacZ [48, 49]. However, these two reports have conflicting results. In one study, only 20% of hMSCs were transduced with the adenoviral vector, and neither MOI (range, 250 to 2,000) nor period of time the cells were exposed to the virus (beyond 6 hours) caused a variation in the percentage of transduced cells [48]. In the other study, adenoviral-mediated gene transfer into hMSCs was shown to be dose dependent. The transduction efficiency at MOI 1,000 was greater than 90%, and the RSV promoter was more active than the cytomegalovirus promoter in expressing lacZ in hMSCs [49]. In the present study, it was shown that adenoviral-mediated gene transfer into rMSCs is dose dependent and that the RSV promoter can be used to express both the reporter gene ntlacZ and the therapeutic gene CGRP, which released CGRP in amounts sufficient to have biologic activity.

A possible explanation for the reduction in transgene expression in vitro during the 21-day culture period is that cell proliferation occurred, reducing the percentage of CGRP-positive cells. Although a low-serum medium is used to culture rMSCs, the cells still retain the ability to proliferate. Therefore, the number of positive cells expressing the transgene will decrease as the cells divide. A possible explanation for the reduction of transgene expression in vitro after differentiation medium is applied is that ongoing cell proliferation reduces the number of positive cells expressing the transgene. Because osteogenic medium does not inhibit cell proliferation, rMSCs divide during the osteogenic differentiation process. Although adipogenic medium can greatly inhibit cell proliferation, it cannot completely prevent cell proliferation. Therefore, rMSCs still can slowly proliferate during the adipogenic differentiation process. Thus, the cell proliferation rate in osteogenic medium is much faster than in adipogenic medium. This may explain the observation that the percentage of β-galactosidase–positive cells at day 21 in osteogenic medium–treated cells was only 6 ± 1%, whereas the percentage of β-galactosidase–positive cells at day 21 in adipogenic medium–treated cells was 29 ± 1%. It also may explain the observation that Adprepro-CGRP–transduced rMSCs treated with osteogenic medium for 21 days only secreted 0.02 ± 0.01 pmol CGRP/1 × 106 cells/48 hours, whereas Adprepro-CGRP–transduced rMSCs treated with adipogenic medium for 21 days secreted 1.47 ± 0.03 pmol CGRP/1 × 106 cells/48 hours into the culture medium.

The characteristic feature of MSCs is their potential for osteogenic, chondrogenic, and adipogenic differentiation. However, a chondrogenic medium for rMSCs has not yet been developed. Therefore, to establish a phenotype for the transduced rMSCs, only osteogenic and adipogenic media are used. In one study, rMSCs isolated from rat femur and tibia by adherence to tissue culture plastic were transduced with retroviruses containing tyrosine hydroxylase and guanosine triphosphate (GTP) cyclohydrolase I to produce L-DOPA. To determine if the transduced rMSCs maintained their multipotentiality and to establish a MSC phenotype, the transduced cells were treated with osteogenic or adipogenic medium and were shown to differentiate into osteoblasts or adipocytes [41]. In another study, rMSCs were isolated by adherence to tissue culture plastic and were transduced with retrovirus containing green fluorescent protein. The transduced cells were then treated with osteogenic medium and differentiated into osteoblasts [43]. In a third study, rMSCs were isolated from rat femur and tibia and were transduced with adenoviral vector containing endothelial nitric oxide synthase (eNOS) so that the cells could release nitric oxide. To establish a MSC phenotype of the transduced rMSCs, the cells were treated with osteogenic or adipogenic medium and differentiated into osteoblasts or adipocytes [40]. In the present study, to establish a MSC phenotype of the Adprepro-CGRP–transduced rMSCs, the transduced cells were shown to differentiate into osteoblasts or adipocytes in culture and secrete CGRP.

For gene therapy of pulmonary hypertension, erectile dysfunction, and cerebral vasospasm, a therapeutic gene such as CGRP can be injected locally into the lung, corpus cavernosum, or brain [2328]. Previous studies have shown that direct injection of the adenovirus containing CGRP attenuated pulmonary hypertension, augmented erectile function, and attenuated responses to vasoconstrictor stimuli [2325, 28]. However, the disadvantages of this strategy include a local inflammatory response and random expression of the transgene.

An ex vivo gene therapy approach is to inject a gene such as CGRP gene–modified carrier cells into an organ. The carrier cells serve as a vehicle and locally deliver the gene product CGRP. The major advantage of this strategy is that the carrier cells can be transduced ex vivo. The carrier cells should be easily isolated and ex vivo expanded, and it is important that the cells express the CGRP transgene and secrete biologically active amounts of the peptide. Furthermore, the CGRP gene–modified carrier cells should persist for long periods in vivo and not elicit an inflammatory response.

MSCs can be easily isolated from bone marrow, readily ex vivo expanded, and efficiently gene engineered [30, 4850]. As multipotent adult stem cells, MSCs are capable of differentiating into osteoblasts, chondrocytes, adipocytes, myocytes, and other cell types; can survive for months in vivo; and do not elicit inflammation after autologous transplantation [3135, 41]. In one study, rMSCs transduced with retroviruses containing tyrosine hydroxylase and GTP cyclohydrolase I synthesized L-DOPA and retained multipotentiality in vitro. The cells were intracerebrally injected into a rat model of Parkinson's disease and produced a significant reduction in apomorphine-induced rotation in the rat. The transgene expression persisted for 9 days, but the cells engrafted and survived for at least 87 days in the rat brain [41]. In another study, swine MSCs were implanted into the infarcted area in a swine model, and microscopic and sonomicrometry analyses demonstrate that implantation of autologous MSCs results in sustained engraftment, myogenic differentiation, and improved cardiac function [35]. Therefore, MSCs fulfill the criteria for carrier cells in ex vivo gene therapy, and CGRP gene–modified ex vivo–expanded MSCs should be able to locally deliver the therapeutic gene product CGRP after autologous transplantation. The cells, as adult stem cells, may also be able to repair injured tissue or replace injured host cells after in vivo differentiation under the influence of a local tissue microenvironment.

To develop an improved therapy for cardiovascular diseases, the use of ex vivo–expanded MSCs as carrier cells for CGRP in ex vivo gene therapy is proposed. It is hypothesized that injection of CGRP gene–modified MSCs will attenuate pulmonary hypertension, improve erectile function, and attenuate cerebral vasospasm. It is also hypothesized that this novel adult stem cell–based ex vivo gene therapy strategy should induce less inflammation, release factors beneficial to the host, and regenerate damaged tissue after cell differentiation.

In a previous study, it was shown that rMSCs transduced with an adenoviral vector containing eNOS survived in rat corpus cavernosum after intracavernosal injection, expressed high levels of eNOS in vivo, and improved erectile function in aged rats [40]. In our recent work, we have shown that rMSCs transduced with adenoviral vector containing the reporter gene ntlacZ survived in the lung of both normal rats and rats with monocrotaline-induced pulmonary hypertension after intratracheal injection (unpublished data). In the present study, to demonstrate biologic activity, we treated rat PASMCs with culture supernatant from Adprepro-CGRP–transduced rMSCs and observed that smooth muscle intracellular cAMP levels were increased. This suggests that CGRP secreted by Adprepro-CGRP–transduced rMSCs is biologically active [57, 24, 25, 27]. Therefore, it is possible that in vivo administration of rMSCs transduced with adenoviral vector containing therapeutic gene such as CGRP may have a beneficial effect. To achieve a therapeutic effect in vivo, rMSCs transduced with Adprepro-CGRP at MOI 300 will be injected intracavernosally, intratracheally, or intrathecally into the rat in future studies. The CGRP gene–modified rMSCs should be able to release CGRP to neighboring host cells, and the concentration of CGRP in the area near the cell injection site in vivo should be high. Future studies will be undertaken in our laboratory to determine whether CGRP-secreting rMSCs attenuate pulmonary hypertension and improve erectile function.

Recent studies also suggest that the time course of transgene expression in MSCs may be altered when the cells are injected into an organ. In one study, rMSCs transduced with retroviruses containing tyrosine hydroxylase and GTP cyclohydrolase I secreted L-DOPA for months in vitro, but transgene expression only persisted for 9 days after being injected into rat brain, whereas rMSCs survived for over 87 days in rat brain [41]. In another study, hMSCs transduced with adenoviral vector containing lacZ expressed the transgene for months in the murine heart after intraventricular injection [49]. Therefore, the persistence of transgene expression in MSCs can be different in in vitro and in vivo settings. Our future studies will examine how long rMSCs can survive in vivo and how long CGRP transgene expression in Adprepro-CGRP–transduced rMSCs can persist after the cells are intracavernosally, intratracheally, or intrathecally injected into the rat.

In summary, the results of the present study show that adenoviral gene transfer of CGRP into ex vivo–expanded MSCs can be accomplished. The Adprepro-CGRP–transduced MSCs secreted high levels of biologically active CGRP into the culture medium. The cells retained multipotential differentiation capability after adenoviral-mediated CGRP gene transfer, and the proliferation and viability of MSCs were not altered by Adprepro-CGRP transduction at MOI 300. These data suggest that this novel adult stem cell–based cell and gene therapy strategy may represent a new form of therapy for the treatment of disorders in which CGRP activity is reduced or in which CGRP will have a beneficial effect.

Acknowledgements

This work was supported by NIH NHLBI grant HL-62000 and NCI grant CA-65600.

Ancillary