SEARCH

SEARCH BY CITATION

Keywords:

  • Mesenchymal progenitor cells;
  • Cellular vehicles;
  • Gene delivery;
  • Adenovirus;
  • Thymidine kinase

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Summary
  8. Acknowledgements
  9. References

Mammalian cells represent a novel vector approach for gene delivery that overcomes major drawbacks of viral and nonviral vectors and couples cell therapy with gene delivery. A variety of cell types have been tested in this regard, confirming that the ideal cellular vector system for ex vivo gene therapy has to comply with stringent criteria and is yet to be found. Several properties of mesenchymal progenitor cells (MPCs), such as easy access and simple isolation and propagation procedures, make these cells attractive candidates as cellular vehicles. In the current work, we evaluated the potential utility of MPCs as cellular vectors with the intent to use them in the cancer therapy context. When conventional adenoviral (Ad) vectors were used for MPC transduction, the highest transduction efficiency of MPCs was 40%. We demonstrated that Ad primary-binding receptors were poorly expressed on MPCs, while the secondary Ad receptors and integrins presented in sufficient amounts. By employing Ad vectors with incorporated integrin-binding motifs (Ad5lucRGD), MPC transduction was augmented tenfold, achieving efficient genetic loading of MPCs with reporter and anticancer genes. MPCs expressing thymidine kinase were able to exert a bystander killing effect on the cancer cell line SKOV3ip1 in vitro. In addition, we found that MPCs were able to support Ad replication, and thus can be used as cell vectors to deliver oncolytic viruses. Our results show that MPCs can foster expression of suicide genes or support replication of adenoviruses as potential anticancer therapeutic payloads. These findings are consistent with the concept that MPCs possess key properties that ensure their employment as cellular vehicles and can be used to deliver either therapeutic genes or viruses to tumor sites.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Summary
  8. Acknowledgements
  9. References

One of the major requirements for successful gene therapy is efficient delivery of therapeutic genes to the target sites. Among the vector systems that have been used for gene delivery are viruses and plasmids [1]. An optimal gene delivery vector must show a high level of transduction of the desired tissue or cell population, efficient targeting, and the ability to withstand degradation by the immune system for prolonged periods and after readministration. To date, vectors meeting all three requirements are not available. On this basis, it is apparent that novel vector approaches are required to advance gene therapy.

Mammalian cells have been proposed as gene delivery vehicles with the potential for overcoming the physiological barriers to viral vectors, especially rapid degradation by the immune system and toxicity with escalating doses. The transfer of genetically modified cells into the body constitutes a novel therapeutic paradigm of coupling cell therapy with gene delivery [2, 3]. By virtue of native or introduced tropism, cellular vehicles, after reintroduction, may reside in selective tissues or organs and deliver engineered vectors or genes before they are released from cellular carriers. Therefore, expression or release of these therapeutic payloads can, thus, be achieved at a locoregional level in areas otherwise inaccessible to gene transfer, such as tumor sites. Several somatic cell types, such as fibroblasts [4, 5], hepatocytes [6], mesothelial cells [7], myoblasts [8, 9], neuronal cells [10], and others have been evaluated in a variety of applications as cellular vehicles. Mature endothelial cells have been used to deliver therapeutic genes into areas of angiogenesis [11, 12] as well as to deliver cytotoxic genes to tumors [13]. However, difficult isolation and in vitro expansion present major limitations for the use of mature terminally differentiated cells. Hematopoietic stem cells have also been evaluated as cellular vehicles [14, 15]. However, low levels of transduction also limit their applicability [16]. In the context of cancer treatment, human tumor infiltrating lymphocytes (TILs) were the first immune cells to be genetically modified and applied in a human cancer gene therapy clinical trial [17]. It was suggested that TILs could naturally home to tumor sites; however, in practice, it has been shown that they localize poorly into tumors after reinfusion [18, 19]. Utilization of TILs requires expensive expansion of cells in vitro or in vivo and applying retargeting strategies. Such studies demonstrate both the feasibility of utilizing different cell types as vectors for gene delivery and also the stringent criteria for the ideal cellular vehicle, providing necessities of research into alternative cell populations.

Mesenchymal progenitor cells (MPCs) were identified as a subpopulation of bone marrow stromal cells, which form an essential structural and functional component of the bone marrow microenvironment and are critical in hematopoiesis [20]. MPCs mainly have been employed in cell therapies based on the delivery of the cells themselves, either modified or not, in different models and in the clinical setting to replenish tissue defects or trigger regeneration of various mesenchymal tissues [2123]. Recent advances in studying MPC biology have shown that this cell population exhibits some properties that suggest the feasibility of their use as a cellular vehicle: A) a simple isolation method [24]; B) the ability to be cultured in vitro in minimal conditions and to expand to quantities required for therapy [25]; C) the ability for ex vivo transduction with viral vectors, although with varying levels of efficiency [26]; D) plasticity, the potential to differentiate under exogenous stimuli [27]; E) the ability to engraft after reintroduction [28]; F) high metabolic activity and efficient machinery to express therapeutic proteins in secretory form [29], and G) the ability to be delivered systemically or locally [30, 31]. Practical attempts to use MPCs as cellular vehicles, however, have been undertaken mainly for the delivery of therapeutic gene products (interleukin-3, growth hormone, factor IX) into systemic circulation [29, 32]. Research studies exploiting MPCs in the context of cancer are limited. However, recently, one study reported the very important fact that MPCs could selectively proliferate in tumors after systemic delivery and serve as precursors for stromal fibroblasts of solid tumors [33].

Given that the intrinsic properties of MPCs satisfy some of the criteria for the ideal cell vector, this cell population can be considered as an attractive candidate for a cellular vehicle. Our goal in this study was to investigate the utility of MPCs as novel cellular vehicles for gene delivery to tumor sites. Our present data show that MPCs possess key vector properties allowing them to function as cellular vehicles for cancer gene therapy.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Summary
  8. Acknowledgements
  9. References

Reagents

Mouse anti-CAR (coxsackievirus and adenovirus receptor) monoclonal antibody (RmcB) prepared as ascites fluid was obtained from Dr. R.L. Crowell (Hahnemann University; Philadelphia, PA). Antibodies to ανβ3 (LM609) and ανβ5 (P1F6) integrins were purchased from Chemicon International, Inc. (Temecula, CA; http://www.chemicon.com).

Cell Lines

The human ovarian carcinoma cell line SKOV3.ip1 was obtained from Dr. Janet Price (University of Texas M.D. Anderson Center; Houston, TX). Cells were maintained in Dulbecco's modified Eagle's medium/F-12, containing 10% fetal bovine serum (FBS) (HyClone Lab; Logan, UT; http://www.hyclone.com) and 2 mM glutamine at 37°C in a humidified atmosphere of 5% CO2. HeLa cells were obtained from the American Type Culture Collection (Manassas, VA; http://www.atcc.org) and cultured as recommended.

Isolation and Culture of MPCs

Leftover materials (screen filters with bone marrow cells remaining) were obtained from several individuals undergoing bone marrow harvest for allogeneic transplantation at the University of Alabama Birmingham (UAB) Stem Cell Facility. Mononuclear cells were separated by centrifugation in Ficoll-Hypaque gradients (density = 1.077 g/cm3; Sigma; St. Louis, MO; http://www.sigmaaldrich.com), suspended in α-minimum essential medium (MEM) containing 20% FBS and seeded at a concentration of 1 × 106 cells/cm2. After 3 days, nonadherent cells were removed by washing with phosphate-buffered saline (PBS), and the monolayer of adherent cells was cultured until confluence. The resultant monolayer of adherent cells was designated as MPC, passage 1. Cells were expanded by consecutive subcultivations in α-MEM with 10% FBS at densities of 5,000–6,000 cells/cm2 and used for experiments at passages 2–6. The initial seeding cell number (Ns), harvested cell number (Nh), and days in culture needed to reach 90% confluence (D) were used to calculate doubling time (DT) for each subculture using the formula DT = 2 × D × Ns/Nh. The mean doubling time was calculated as the average of all DTs for each subculture.

Assay for Cell Differentiation

Mesenchymal progenitor cells were cultured in medium containing either osteogenic (0.1 μM dexamethasone, 10 mM β-glycerophosphate, 50 μg/ml ascorbic acid) or adipogenic (1 μM dexamethasone, 100 μg/ml 3-isobutyl-1-methylxanthine, 5 μg/ml insulin, 60 mM indomethacin) supplements (Sigma) for up to 3 weeks. The differentiation potential was defined by the appearance of osteogenic, adipogenic, and chondrogenic phenotypes, which were revealed after Alizarin Red S, Oil Red, and Alcian Blue staining, respectively.

Recombinant Adenoviruses

Replication-incompetent recombinant adenoviral (Ad) vectors having wild-type and genetically modified Ad5 fibers were used for experiments to determine their transduction efficiencies on MPCs. Ad vectors having wild-type Ad5 fibers were: Ad5 expressing either green fluorescent protein (eGFP) (AdCMVGFP; a gift from Corey Goldman; Cleveland Clinic Foundation; Cleveland, OH) or Escherichia coli β-galactosidase (AdCMVlacZ; provided by Robert Gerard; The Center for Transgene Technology and Gene Therapy; Leuven, Belgium). Viruses expressing firefly luciferase (AdCMVluc) and herpes virus thymidine kinase (HSV-TK) (AdCMV-TK and AdRGD-TK, see below for arginine-glycine-aspartic acid modification) were constructed through homologous recombination in E. coli using the AdEasy system [34] at UAB Gene Therapy Center. All vectors used in these experiments contained transgene cassettes placed in the E1-deleted region of an Ad vector genome.

The luciferase and GFP-encoding Ad vectors, Ad5RGDluc and AdCMVGFP RGD [35], have a genetically modified fiber protein with an integrin-binding motif (CDCRGDCFC) inserted in the HI loop. Ad5/3Luc1 has a chimeric fiber, where the knob of Ad5 fiber is replaced by the Ad3 fiber knob [36]. Ad5luc3 [37] is a replication-competent adenovirus, having the E3 region replaced with the luciferase-expressing cassette (provided by F. Graham; McMaster University; Hamilton, Canada). Ad5/3luc3 is identical to Ad5luc3 except for a chimeric fiber where the knob of the Ad5 fiber is replaced by the Ad3 fiber knob.

In Vitro Ad-Mediated Gene Transfer

Recombinant Ad vectors were used to transduce cells as described previously [38]. In brief, cells were plated 24 hours before infection. The transduction was performed with the virus diluted in a small volume of appropriate cell-specific medium containing 2% FBS for 2 hours at 37°C. Media were then replaced with complete media containing 10% FBS.

In Vitro Analysis of Gene Expression

Twelve hours after plating (5.0 × 104 cells/well in a 24-well plate), HeLa cells and MPCs were infected with AdCMVlacZ, AdCMVGFP or AdCMVluc, Ad5RGDluc, and Ad5/3luc1 at ratios ranging from 50–5,000 plaque-forming units (pfu) per cell. Gene expression was estimated after 48 hours of incubation in complete media.

X-gal Staining of MPCs

LacZ expression in Ad-transduced cells was evaluated by staining with 5-bromo-4-chloro-3-indolyl-β-D-galactoside (X-gal) at 1 mg/ml followed by cell counting. AdlacZ-transduced MPCs were fixed in freshly prepared 2% formaldehyde/0.2% glutaraldehyde in PBS for 5 minutes at 4°C, washed, and stained in freshly prepared X-gal in 20 mM potassium ferrocyanide and 2 mM MgCl2 in PBS.

Luciferase Assay

Luciferase expression was determined in cell lysates using a luciferase assay system (Promega; Madison, WI; http://www.promega.com) according to protocol. The luciferase activities were measured in a Lumat LB 9501 luminometer (Lumat, Wallac, Inc.; Gaithersburg, MA; http://www.lifesciences.perkinelmer.com) for 15 seconds immediately after initiation of the light reaction and normalized by the protein concentration in cell lysates (Bio-Rad DC Protein Assay kit, Bio-Rad; Hercules, CA; http://www.bio-rad.com). All experiments were performed in triplicate.

Flow Cytometry

GFP expression and expression of Ad receptors were assessed by flow cytometry. Cultured cells were released with versene, pelleted by centrifugation, and washed with PBS. GFP-expressing cells after Ad-GFP transduction were directly used for flow cytometry. To test the expression of Ad receptor antibodies to CAR (RmcB), integrins ανβ3 (LM609) and ανβ5 (P1F6) were used. Cells (5 × 105) were incubated with the primary antibody diluted in PBS containing 0.1% bovine serum albumin for 1 hour at 4°C. Concentrations of the primary antibodies were 10 μg/ml, 2.5 μg/ml, and 10 μg/ml for RmcB, P1F6, and LM609, respectively. Cells were washed and incubated with fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG for 30 minutes at 4°C in the dark. After washing in PBS, the cells were resuspended in 500 μl of PBS containing 0.1% formaldehyde. Flow cytometry was performed on a FACScan flow cytometer (Becton Dickinson; Erembodegem, Belgium; http://www.bd.com).

Toxin Gene-Killing Experiments

In Vitro Analysis of the Cytocidal Effect of HSV-TK Expression

HSV-TK expression by MPCs and SKOV3.ip1 cells after exerting ganciclovir (GCV) was estimated by measuring tetrazolium salt (MTS) 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium cell-killing assay. Briefly, MPCs or SKOV3.ip1 cells were plated on 96-well plates at 2 × 104 cells/well. Cells were transduced with AdCMV-TK or AdCMV RGD-TK at the multiplicity of infection (MOI) of 5, 50, 100, and 500 viral particles (vp)/cell. After 24 hours, plates were treated with increasing concentrations of GCV. Four days later, the number of surviving cells was analyzed by MTS assay according to the manufacturer's protocol (Cell Titer 96 Aqueous Non-Radioactive Cell Proliferation Assay, Promega; Madison, WI; http://www.promega.com). The plates were processed in an automated E-max spectrophotometer plate reader (Molecular Device Corp.; Sunnyvale, CA; http://www.moleculardevices.com). All experiments were carried out in triplicate.

In Vitro Analysis of Cytotoxic Bystander Effect

Cell-mixing experiments were performed on MPCs or SKOV3.ip1 cells to assess the bystander effect as described elsewhere [38]. Briefly, MPCs were transduced with AdCMV-TK or AdCMV RGD-TK and then mixed with untransduced cells (MPCs or SKOV3.ip1 cells). The ratios of TK-expressing cells to untransduced cells were as follows: 0:100, 10:90, 50:50, 90:10, and 100:0. Cell mixtures were plated in 96-well plates at total cell densities of 2 × 104 cells/well. Twenty-four hours later, half the samples were treated with GCV (20 μM). Five days later, the number of surviving cells was analyzed by MTS assay. All experiments were carried out in triplicate.

Assays for Ad Replication in MPCs

Crystal Violet Cell Viability Assay

The cytopathic effect (CPE) of replication-competent Ad was estimated by crystal violet staining. Ad-permissive HeLa cells were used as a positive control for Ad replication and virus-induced cytopathic effect. HeLa cells and MPCs were plated on 6-well plates at 2 × 105 cells/well and infected with Ad5luc3 for 2 hours at 37°C at the indicated MOI, followed by PBS washes. Plates were incubated at 37°C until a prominent CPE developed. Cells were carefully washed with PBS, fixed with 10% buffered formalin, and stained with 0.2% crystal violet in 10% ethanol for 1 hour. The plates were washed with tap water, dried, and photographed with a Kodak DC260 digital camera. All experiments were done in triplicate.

Cytopathic Effect of Produced Adenovirus on Permissive Cell Line

MPCs or HeLa cells were infected with a replication-competent virus (Ad5luc3) at the designated MOI. After 1 hour of infection, all wells were washed twice with PBS, and complete medium was added, and cultures were incubated for 3 days. Cells were pooled at the end of the incubation period and freeze-thawed three times. The lysates were diluted in 10-fold steps and used for secondary infection of HeLa cells. The cytopathic effect of cell lysates was detected after 4 days with crystal violet staining. All experiments were done in triplicate.

Quantitative Polymerase Chain Reaction Analysis of Ad DNA

MPCs and HeLa cells were plated on 6-well plates at a density of 2 × 105 cells/well 24 hours before infection. Cells were infected with nonreplicative (Ad5luc) or replication-competent (Ad5luc3 and Ad5/3luc3) viruses at an MOI of 1. The infected cells and virus-containing supernatants were harvested 1, 3, 5, 7, and 9 days after infection and used for viral DNA and viral protein quantitation. Culture media and cells from three wells corresponding to each time point were collected, pooled, and used for viral and total cell DNA isolation. Total DNA from cells was purified using the QIAamp DNA Blood kit (QIAGEN; Valencia, CA; http://www.qiagen.com) according to the kit instructions. Total cell DNA was used for quantitative polymerase chain reaction (PCR) with E1 primers, to determine copy numbers of vDNA, and actin primers, to determine number of actin cDNA copies. To isolate only encapsidated viral DNA from culture media and to destroy naked viral DNA present in the supernatant, DNase1 treatment was performed. Then, EDTA, SDS, and proteinase K were added (final concentrations: 20 mM, 0.5%, 0.2 mg/ml, respectively) and samples were incubated at 56°C for 1 hour followed by phenol/chloroform extraction and ethanol precipitation. The purified viral DNA was dissolved in Tris-EDTA buffer, pH 8.0.

Real-time PCR (LightCycler; Roche; Indianapolis, IN; http://www.roche.com) analysis was used for quantitative evaluation of Ad DNA copy number. Oligonucleotides corresponding to the sense strand of the Ad E1 region (5′-AAC CAGTTGCCGT GAGAGTTG-3′: 1433–1453), the antisense strand of the E1 region (5′-CTCGTTAAGCAAGTCCTCG ATACA-3′: 1500-1476), and TaqMan fluorigenic probe (5′-CACAGCCTGGCGACGCCCA-3′: 1473-1455) were synthesized and used as primers and probe for real-time PCR analysis. To correct for differences in total DNA concentration for each sample, Ad E1A copy numbers were normalized by actin DNA copy number. Primer sequences to detect transcripts of actin DNA were as follows: sense strand 5′-CAGCAGATGTGGATCAGCAAG-3′; antisense strand 5′-CTAGAAGCATTTGCGGTGGAC-3′; and TaqMan probe 5′-AGGAGTATGACGAGTCCGGCCCCTC-3′.

Evaluation of Presence of Viral Protein in Culture Media

The presence of viral hexon in culture media was estimated with enzyme immunoassay for the detection of Ad in infected cell cultures (IDEA kit; DAKO; Carpinteria, CA; http://www.dako.dk) in accordance with kit instructions. OD450 was determined on an automated E-max spectrophotometer plate reader. Samples giving OD450 in the range of 2.0 were diluted, and the correct QD450 was obtained by multiplying by the dilution factor.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Summary
  8. Acknowledgements
  9. References

Characterization of Isolated Human MPCs

The practical realization of cells as vehicles for gene therapy requires a favorable combination of native properties of a chosen cell type, as well as the ability to be effectively loaded by genes, or other therapeutics, and to deliver such a payload to target sites. The theoretical considerations of utilizing MPCs as such a cellular vector were predicated primarily by their availability, propagating properties, and simple culturing conditions. Our goal was to test several vector properties of MPCs related to their potential application as delivery vehicles for gene therapy. First, we tested the ability of human MPCs to be maintained in culture and to be propagated in quantities required for in vivo applications. Human MPCs were isolated according to the protocol described elsewhere [39, 40] and plated at densities of 6–7 × 103 cells/cm2 in media supplemented with 10% FBS. Human MPC cultures, after two passages in vitro, represented a mainly homogeneous population of cells as assessed by cell morphology, with predominant fibroblast-shaped cells and large flat and star-shaped cells at various percentages (Fig. 1A). Several isolations of MPCs were accomplished according to the described protocol, and primary cultures were monitored for a number of passages (Fig. 1C). The doubling time was calculated for each passage of an individual culture under the described culture conditions and varied from 3 to 6 days. Some MPC cultures were monitored for as long as 12 passages with no noticeable significant changes in propagating properties (Fig. 1B). To confirm that the population of isolated human bone marrow cells contained MPCs or stem cells, we tested the abilities of primary cultures on passage 2–3 to undergo osteogenic, adipogenic, and chondrogenic differentiation after applying the corresponding conditioned media (Fig. 2A–D). The differentiation potentials of the individual cultures varied from having all three tested lineages being represented to being predominantly skewed toward osteogenic or adipogenic differentiation. Nevertheless, most of the tested cultures demonstrated good propagation and differentiation abilities, confirming the concept that this adherent cell population in fact contains progenitor cells and has sufficient propagation properties to be exploited for application as a cell vehicle.

thumbnail image

Figure Figure 1.. A) Representative phase-contrast photomicrograph of a human MPC culture.Typical cell phenotypes were observed: most cells were spindle shaped, whereas large flat cells and star-shaped cells were present at lower numbers. B) Expansion of MPCs in culture. The primary culture from a single bone marrow donor was followed for up to 12 passages; the doubling time for each passage and the mean doubling time for all passages in observation were calculated as described inMaterials and Methods. C) Mean doubling times for different MPC cultures.

Download figure to PowerPoint

thumbnail image

Figure Figure 2.. Assay for differentiation abilities of MPCs.Primary MPC cultures at passage 2–5 were subjected in vitro to conditions stimulating osteogenic, adipogenic, and chondrogenic differentiation followed by the appropriate staining. A) MPCs cultured in osteogenic media for 21 days. The accumulation of mineral deposits was detected by staining with Alizarin Red. B) MPC culture incubated in adipogenic media for 7 days. Fat droplets in the cells were stained with Oil Red. C) MPC culture incubated under chondrogenic conditions for 14 days. Cells formed a pellet, which stained blue with Alcian Blue. D) MPC culture incubated in nondifferentiating medium, stained with Alizarin Red. No mineral deposits were observed.

Download figure to PowerPoint

Accessibility of MPCs to Ad Vector Infection

Our next step was to test the ability of MPCs to be efficiently transduced by Ad vectors and then express the genetic payload. Previous studies have documented that MPCs could be transduced in vitro by Ad vectors, although in any conditions used, the efficiency of transduction did not exceed 20% [41]. For our studies, Ad vectors containing reporter genes (E. coli LacZ or eGFP) were used for transduction of MPCs at escalating MOI. Forty-eight hours post infection, MPCs and HeLa cells expressing LacZ were stained with X-gal, and transduction efficiency was assessed by microscopic scoring. Cells expressing eGFP were detected by flow cytometry. Infected with an Ad vector at an MOI of 500 vp/cell, MPC cultures consistently demonstrated a lower efficiency of transduction than did HeLa cells; transduced MPCs represented only 10%–20% of cells in MPC cultures (Fig. 3A). As detected by flow cytometry at the highest MOI tested (5,000 vp/cell), when almost 100% of HeLa cells were transduced, MPC transduction was only 40% (Fig. 3B). These results are consistent with previous observations by others [41, 42] reporting that MPCs are relatively refractive to Ad vector transduction.

thumbnail image

Figure Figure 3.. Transduction of MPCs with recombinant Ad5 vectors.A) MPCs and HeLa cells (4 × 105) were transduced with an Ad vector encoding the E. coli LacZ gene at an MOI of 500 vp/cell. Forty-eight hours postinfection, cells were fixed and stained with X-gal. B) MPCs and HeLa cells were transduced with an Ad vector encoding GFP reporter at an MOI of 500 or 5,000 vp/cell. Forty-eight hours postinfection, cells were treated with versene and the percentage of GFP-positive cells was determined by flow cytometry. Data are shown as the percent of gated (GFP-positive) cells at the corresponding MOI.

Download figure to PowerPoint

The observed relative resistance of MPC cultures to Ad vector transduction could be due to low expression of Ad attachment (CAR) and/or internalization receptors (ανβ integrins). By flow cytometry, we compared the expression profiles of Ad receptors on MPCs, HeLa (high CAR, relatively low integrins) cells, and RD (low CAR, high integrins) cells. As shown in Figure 4, MPCs showed very low, if any, expression of CAR but high expression levels of both of the integrins tested. Thus, the pattern of expression of Ad receptors on MPCs resembles that of RD cells and corresponds to a low CAR, high-integrin phenotype. These data suggest that CAR-independent transfer is required to improve the level of Ad-mediated transgene expression by MPC cultures. These considerations prompted our evaluation of the genetically modified adenoviruses for genetic transfer to MPCs.

thumbnail image

Figure Figure 4.. Cell surface expression of Ad attachment (CAR) and internalization (ανβ3,ανβ5) receptors.After treatment with the indicated primary antibody and FITC-conjugated secondary antibody, cells were analyzed by flow cytometry. Histograms show fluorescence intensity data for HeLa (A), RD (B), and MPC (C) cells. HeLa and RD cultures were taken as a positive control for CAR and integrin expression, respectively.

Download figure to PowerPoint

Increase in Gene Expression by MPCs Transduced with Modified Ad Vectors

It has been shown that significant augmentation of transduction efficiency can be successfully achieved by exploiting Ad vectors with genetically modified tropism. We have designed and characterized an Ad vector with the RGD-4C peptide incorporated in the HI loop of the fiber knob (AdRGD) as well as a vector having the Ad5 knob replaced with the knob of Ad3 (Ad5/3) [35, 36]. These structural modifications confer expanded tropism to the adenoviruses, enabling CAR-independent mechanisms of infection via binding to either cellular integrins or still unknown Ad3 receptors. In addition, these vectors have shown a remarkable improvement in gene transfer efficiency in cell lines normally refractory to CAR-dependent infection [43, 44].

We used genetically modified Ad vectors encoding the luciferase gene, Ad5RGDluc, and Ad5/3luc1, in which the expression of luciferase is driven by the CMV promoter, to evaluate the transduction of several primary MPC cultures. Improvement in transduction efficiency was evaluated by direct comparison with an unmodified control Ad5 vector, AdCMVluc. The application of AdRGDluc to MPCs repeatedly resulted in a tenfold augmentation of gene transfer compared with AdCMVluc (Fig. 5A). Ad5/3luc1 also showed an improvement in gene transfer, although with varied results on different primary MPC cultures (data not shown). Considering that an elevated level of luciferase expression in MPC cultures could result in either a greater number of transduced cells or enhanced protein expression by selective MPC subpopulations, we also tested MPC transduction efficiency with Ad vectors encoding GFP as the reporter gene. Transduction with CMVAd GFP RGD resulted in greater numbers of GFP-expressing cells detected by flow cytometry (Fig. 5B) and microscopy (Fig. 5C). These data indicate overall enhancement of MPC gene transduction by RGD-modified Ad vectors. Thus, these results indicate that an Ad vector retargeted to cellular integrins provides a means to overcome the CAR deficiency in MPCs and allows for an enhanced gene transfer to these cells.

thumbnail image

Figure Figure 5.. Enhanced transduction efficiency of MPCs with Ad vectors having genetically modified fibers.Comparison of the gene transfer efficiencies was performed employing nonreplicative Ad5 vectors: Ad5luc, Ad5RGDluc, and Ad5/3luc1 encoding the luciferase gene (A) or AdCMVGFP and AdCMVGFP RGD (B, C). Human MPCs and HeLa cells were infected at the designated MOI (pfu/cell) and analyzed for luciferase expression after 48 hours (A). Data are shown as relative light units (RLU)/μg of total cellular protein. All experiments were done in triplicate. Error bars show standard deviations. MPCs transduced with GFP-encoding vectors were analyzed for GFP expression by flow cytometry (B) and by microscope (C). Microphotographs represent MPC cultures transduced with AdCMVGFP RGD at the MOI designated on each picture (10, 100, or 1,000 pfu/cell). A bright-field microphotograph of a corresponding MPC culture is labeled BF.

Download figure to PowerPoint

MPC Expression of Prodrug-Converting HSV-TK Results in the Cytotoxic Effect of GCV

Next, we proceeded to confirm the ability of MPCs to express a therapeutic anticancer gene. For this study, we utilized the HSV-TK gene, whose product causes cell killing by activating GCV. HSV-TK was chosen with the intent to subsequently employ MPCs as vehicles delivering cytotoxicity to tumor sites in an animal model relevant to ovarian cancer. In this study, MPCs and SKOV3.ip1 cells, a human ovarian cancer cell line taken as a parallel control, were transduced with AdCMV-TK and AdRGD-TK at different MOIs (5, 50, 100, or 500 pfu/cell). Forty-eight hours postinfection, GCV was added to the media at different concentrations (0, 10, 100, or 1,000 μM), and 5 days later, the number of remaining viable cells was determined by an MTS assay. A similar killing effect on MPCs and SKOV3.ip1 cells was observed after AdCMV-TK transduction (Fig. 6A and C). For example, infection at an MOI of 50 pfu/cell resulted in 50% of cells of both types being killed in wells with 10 μM GCV added to the culture medium. In contrast, a dramatic enhancement of the toxic effect exerted by the prodrug was observed when AdRGD-TK was used for MPC transduction (Fig. 6B and D). Compared with AdCMV-TK, AdRGD-TK caused a killing effect at a lower MOI and a lower GCV concentration (Fig. 6B). This finding validated the aforementioned advantage of RGD-containing Ad vectors in introducing genes into MPCs. Results of this experiment also suggest that transient expression of the HSV-TK delivered by Ad vectors will allow achievement of sufficient levels of suicide gene expression suitable for a cellular vehicle strategy.

thumbnail image

Figure Figure 6.. Viability of MPCs transduced with AdCMV-TK or AdRGD-TK after GCV treatment.MPCs or SKOV3.ip1 cells were plated on 96-well plates at a density of 3,000 cell/well and transduced with AdCMV-TK or AdRGD-TK at an MOI of 5, 50, 100, or 500 pfu/cell for 2 hours. After infection, GCV was added at a final concentration of 0, 10, 100, or 1,000 μM. Cell viability was determined by MTS assay after 5 days of incubation. Data are expressed as percent of viable cells at corresponding GCV concentrations. All experiments were done in triplicate. Error bars represent standard deviations.

Download figure to PowerPoint

Bystander Effect of AdCMV-TK-Transduced MPCs on SKOV3.ip1 Cells

A key advantage embodied in the molecular chemotherapy strategy to kill cancer cells is based on the bystander effect, whereby the killing of untransduced neighboring cells can be induced by a toxic drug metabolite present in a transduced cell. We, therefore, investigated whether HSV-TK-expressing MPCs could accomplish such a bystander effect on the SKOV3.ip1 tumor cells in vitro. AdCMV-TK- or AdRGD-TK-transduced MPCs were mixed in various ratios with untransduced MPCs or SKOV3.ip1 cells. An MOI of 50 pfu/cell, for the transduction of cell populations, and a GCV concentration of 10 μM were chosen as the lowest viral and prodrug doses showing a killing effect in the previous experiment (Fig. 6). Our mixing experiment demonstrated that AdRGD-TK-transduced MPCs had a greater killing effect on nontransduced MPC populations (Fig. 7A) and on SKOV3.ip1 cell populations than AdCMV-TK-transduced MPCs (Fig. 7B). Moreover, an appreciable bystander effect was documented for AdRGD-TK-transduced MPCs, as the addition of only 10% of TK-expressing MPCs resulted in a total cell viability of 50%. AdCMV-TK-transduced MPCs did not show a noticeable bystander effect in these conditions, likely due to inefficient transduction of MPCs by the unmodified Ad vector, resulting in an insufficient number of TK-producer cells in the mixture. These mixing experiments establish that MPCs possess a capacity to accomplish a bystander effect using the HSV-TK/GCV enzyme/prodrug approach.

thumbnail image

Figure Figure 7.. Bystander effect exhibited in vitro by AdCMV-TK- and AdRGD-TK-transduced MPCs when mixed at various ratios with uninfected MPCs (A) and SKOV3.ip1 cells (B).MPCs were first transduced with AdCMV-TK or AdRGD-TK. Twenty-four hours after transduction, MPC-TK were mixed in different ratios with untransduced cells (MPCs or SKOV3ip1 cells) and plated on 96-well plates. Twenty-four hours after plating cell mixtures, half the cells were treated with 10 μM GCV. Five days later, cell killing was measured by MTS assay. Data are presented as a percent of viable cells in GCV-treated wells. All experiments were done in triplicate. Error bars represent standard deviations.

Download figure to PowerPoint

MPCs as Cellular Vehicles Delivering Replication-Competent Ad Vectors

We also considered the possibility of using MPCs to deliver an oncolytic virus as a therapeutic payload. For this approach to be successful, MPCs must be able to support Ad replication and produce viral particles after viral infection. We first determined if replication-competent Ad vectors could cause a cytopathic effect in MPC cultures, an indication of the ability of MPCs to support a productive Ad infection. For this study, MPCs and HeLa cells, used as a highly permissive cell line, were infected with replication-competent Ad5luc3 at an escalating MOI; the cytopathic effect was evaluated by crystal violet staining. Three days postinfection HeLa cells were completely killed in wells infected at MOIs of 100 and 10 vp/cell, and early cytopathic effects were observed at an MOI of 1 vp/cell. At this time, MPCs remained viable at all viral doses. However, 7 days postinfection, MPCs showed a cytopathic effect at the highest MOI tested (Fig. 8), which indicates that the cycle of Ad replication in MPCs is not inhibited completely, but may have different kinetics.

thumbnail image

Figure Figure 8.. Cytopathic effect of replication-competent adenovirus (Ad5luc3) on MPCs and HeLa cells.HeLa cells and MPCs were plated on 6-well plates with a plating density of 5 × 105cells/well and infected in duplicate with Ad5luc3 at the designated MOI. The viral cytopathic effect was estimated by crystal violet staining on day 3 and day 7 after infection.

Download figure to PowerPoint

To investigate the kinetics of Ad replication in MPCs, we infected the cells with Ad5luc3 and Ad5/3luc3 at an MOI of 1 pfu/cell and monitored the infection for 11 days. Real-time PCR was used to measure the amount of viral DNA in cells and culture media over time (Fig. 9B and 9C). We also used the detection of the viral hexon protein in culture media as an indication of viral replication (Fig. 9A). HeLa cells were included in the experiment as a cell line supporting a high level of Ad replication.

thumbnail image

Figure Figure 9.. Kinetics of Ad production by MPCs.HeLa cells and two primary cultures of MPC (MPC2 and MPC19) were infected with replication-competent Ad5luc3 and Ad5/3luc3 viruses at an MOI of 1 pfu/cell. Virus-containing media were collected at 1, 3, 5, 7, and 9 days after infection. Viral production was estimated as A) accumulation of viral protein in culture media by immunoenzyme assay with detection of Ad hexons. Data are presented as OD450 of 100 μl of culture media. B) Accumulation of viral DNA in Ad-infected cells. C) Accumulation of viral DNA in culture media measured by quantitative PCR. Data are presented as Ad vDNA copy number relative to actin DNA copy number.

Download figure to PowerPoint

In all MPC cultures tested (two of four shown), we observed an increase of viral DNA copy number over time in cell lysates as well as in medium samples. Quantitatively, copy numbers of viral DNA produced by MPCs were 10–100 times lower than with HeLa cells. Ad5/3 did not have an infectivity advantage over Ad5 in HeLa cells as judged by quantitation of viral DNA and viral protein production, while two of the four MPC cultures showed greater amounts of viral DNA and viral protein after infection with Ad5/3. This effect of replication-competent Ad5/3 on MPC cultures corroborates our earlier observation with Ad5/3 nonreplicative viruses resulting in differential efficiencies of gene transfer on different primary MPC cultures. Testing the culture media of infected cultures for the presence of viral hexons confirmed the data obtained with viral DNA. The amount of viral protein increased with progression of infection for all cell cultures. In HeLa cells, the hexon concentration in media increased 14-fold and peaked 5 days after infection. Hexon measurement in infected MPC cultures found an increase of only two- to threefold after 7 days postinfection. Elevation of viral DNA and viral protein during infection, as well as virus-induced cytolysis, clearly indicated viral replication in MPCs.

To prove that MPCs are able to complete the whole cycle of viral replication and produce the next generation of viral particles, we took cell lysates of MPC cultures infected with escalated doses of Ad5luc3 and applied tenfold dilutions of those lysates to the Ad-permissive cell line, HeLa. As shown in Figure 10, the positive control representing lysates of infected HeLa cells resulted in a second round of infection of HeLa cells. Lysates of infected MPCs also caused a cytopathic effect on HeLa cells, confirming the presence of viable Ad virions in the samples. Thus, these data prove that MPCs are able to support Ad replication and can, therefore, potentially serve as vehicles to deliver not only gene products but also viruses in vivo.

thumbnail image

Figure Figure 10.. Production of Ad particles by MPCs.HeLa cells (A) and MPCs (B) were plated on 6-well plates with a plating density of 5 × 105cells/well and infected with Ad5luc3 at an MOI of 1, 10, or 100 pfu/cell. After 3 days, cells in individual wells were lysed, and serial dilutions of lysates were applied onto HeLa cells. The cytopathic effect of produced virus on HeLa cells is shown by crystal violet staining on day 3.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Summary
  8. Acknowledgements
  9. References

There is a continuous intensive search for the optimal delivery vector for different gene therapy protocols. Studies employing mammalian differentiated cells, such as fibroblasts, endothelial cells, mesothelial cells, and others, revealed the feasibility of an ex vivo strategy in general. The progenitors (stem cells) have also been tested in this regard [45, 46], although only a few attempts have been undertaken to address their full potentials as cellular vehicles. This study was undertaken to establish key concepts of an MPC-based cellular vehicle strategy, with the intent to develop an autologous cellular vehicle to deliver therapeutic agents to tumor sites.

We demonstrated that isolation of adherent cells from bone marrow resulted in a cell population with the morphologic and functional characteristics of multipotential MPCs. Overall, isolated MPC cultures had sufficient proliferative and differentiation potentials and retained a multipotential phenotype in culture during several passages. Recently, the high proliferative potential of MPCs has been challenged even further after plating the cells at a low density [25]. In optimized culturing conditions, extensive and rapid cell expansion has been achieved, which will be of great importance for future cell and gene therapy applications.

In addition to intrinsic properties, several characteristics of delivery vehicles must be investigated to consider the MPC as a candidate for a cell-based strategy. First, it is prerequisite that these cells can be efficiently loaded with a gene of choice. In an attempt to achieve this, both retroviral and Ad vectors have been evaluated for gene transfer to human and nonhuman MPCs; however, the transduction had a relatively low efficiency in both cases [42, 45]. Our strategy is based on the short-term expression of a therapeutic gene; therefore, we considered Ad vectors as the vector of choice for ex vivo MPC loading. In our experiments, the efficiency of ex vivo transduction of MPCs with Ad vectors encoding two reporter genes, LacZ and GFP, did not exceed 40% at the highest MOI tested. It is widely understood that the main reason for Ad refractivity, especially in vitro, is the low level of expression of primary Ad receptors. The absence of CAR and the presence of αν-integrins on the surfaces of MPCs were confirmed by flow cytometry, providing a plausible explanation for Ad resistance. As a means to circumvent CAR deficiency, we used Ad vectors retargeted to alternative receptors. Ad5lucRGD, having an integrin-binding motif in the fiber protein, introduces an alternative Ad entry pathway, allowing viral binding directly to cell-surface integrins. Ad5/3luc1 also represents an Ad vector redirected to an, as yet, unidentified Ad3 receptor. We previously demonstrated that such retargeted vectors often augment gene transfer to a variety of primary cell types and cancer cell lines that are otherwise relatively refractory to Ad5 infection [35, 43, 44]. In agreement with previous observations, all primary MPC cultures showed substantial enhancement of gene transfer with Ad5lucRGD. Thus, we demonstrated that MPC genetic loading can be increased tenfold by ex vivo transduction with integrin-retargeted Ad vectors.

Transduction with Ad5/3luc1 on several primary MPC cultures resulted in variable levels of luciferase expression. It is possible that the Ad3 receptor is expressed only within certain subpopulations of MPCs, representing variable fractions in different primary cultures. Of note, MPCs infected with luciferase-encoding Ad vectors produced higher levels of luciferase than HeLa cells in the same experimental setting. One possible explanation for this observation is that each individual transduced MPC might produce more luciferase than each transduced HeLa cell. The ability of MPCs to produce large amounts of protein may represent a quality that could also be useful in the context of a cell-based strategy.

Selective and effective killing of tumor cells remains a major strategy for any of the cancer gene therapy applications. Suicide therapy employs transfer of genes responsible for converting nontoxic products to toxic drugs or genes sensitizing tumor cells to irradiation. To be applicable in the cancer context, our proposed cellular system was tested for the ability of MPCs to express an anticancer gene and to exert cytotoxicity on neighboring tumor cells, thereby providing a local bystander killing effect. We found that MPCs could effectively express a suicide gene and that modified MPCs themselves were sensitive to the toxic effect of GCV. We also explored the ability of MPC-TK to induce a bystander cytotoxic effect on the ovarian tumor cell line SKOV3.ip1 in vitro as a molecular chemotherapy approach on a model system of ovarian carcinoma. This approach was tested previously using human endothelial cells and was shown to accomplish an antitumor effect comparable with viral vector-mediated toxin delivery [13].

Here, we report that the efficiency of MPC gene loading can be substantially increased and the expression of payload genes reaches relevant levels, thereby meeting conditions for applications requiring a relatively short-term intervention. An enhanced bystander effect might potentially overcome the requirement to achieve quantitative transduction of the majority of tumor cells within tumor foci, a major challenge for gene therapy for cancer. Thus, we demonstrated that two important requirements for the use of MPCs as cellular vectors can be met: efficient gene transfer to the cells and high levels of desired protein production, which serve to achieve the biological effect.

We also hypothesized that cellular vehicles can potentially be exploited to deliver oncolytic viruses, which can be considered as promising suicidal tumor agents. For treating neoplastic diseases, conditionally replicative adenoviruses (CRADs) represent a novel and promising approach, employing the intrinsic cytopathic effect of the virus with additional specificity against tumor cells [47]. However, delivery of such a virus to tumor sites faces the same problems as nonreplicative Ad vectors. In this study, we explored the possibility of using MPCs as a means of delivering CRADs. To this end, the ability of MPCs to maintain Ad infection, which includes viral DNA replication and formation of new infectious viral particles, has to be examined. Nothing is known about the ability of MPCs to support Ad infection. Susceptibility of several committed hematopoietic cell lines to different Ad serotypes has been studied, and very low or no Ad5 production has been documented for the cell lines tested [48]. Nevertheless, low levels of virus production have been demonstrated for Ad11 and Ad35, which also showed enhanced binding to those cells. We proved, by several methods, that MPCs were able to support Ad5 replication, although the kinetics of viral production were different from those in a cell line highly permissive to Ad replication. We showed that Ad5 infection of MPCs developed at a slower rate and resulted in the production of a significantly lower number of viral particles than did HeLa cells. Nevertheless, there is no immediate obstacle to utilizing these cells as carriers for oncolytic viruses, and this strategy may have therapeutic applications. Further improvement of CRADs, in terms of oncolytic potency and tumor selectivity, will offer new promising anticancer agents, and the development of appropriate cell carriers for such viruses would represent an attractive line of investigation.

Tumor-specific trafficking of infused cells remains the major question of investigation for cell-based strategies. Apparently, it can be mediated either by biological properties of the tumor itself or by native tumor-related tropism of the chosen cell population. Although the use of circulating cellular vehicles was first proposed a decade ago, only a few cell types have been exploited in the context of tumors. Human TILs were considered as an enriched source of tumor-specific cytotoxic T lymphocytes, and based on their putative preferential localization to tumor sites, were used as vehicles for retroviral-mediated gene transfer [49]. In a recently proposed complex targeting strategy, T lymphocytes were recruited to both produce and deliver a retrovirus to the tumors in an animal model [50]. This approach showed that inefficient T-cell targeting could be overcome by introducing a complex regulatory mechanism. It also demonstrated, for the first time, that cells can be turned into a source of viral production and that those progeny viruses can serve as the delivery cargo. Several lines of investigation utilized the process of angiogenesis, which is highly activated in tumors, as an attractant of cellular vehicles naturally endowed with angiogenic tropisms [11].

Tumor-specific targeting of MPC as one of the essential vector characteristics for our cell-based strategy has not been included in this study but will represent a subject of our future investigations. Although native homing of MPCs after systemic or local infusion has been tested on different animal models in a variety of experimental settings [28, 31], there is still insufficient information about in vivo distribution and survival of infused MPCs. The possibility of MPCs homing to tumor sites could be theoretically envisioned. However, there are only a few very restricted studies investigating the relation of MPCs to tumor formation or growth. It has been confirmed that infused MPCs may reach tumor sites, selectively proliferate there, and participate in the formation of tumor stroma [33]. It has also been demonstrated that MPCs tend to home to sites of injury or damaged tissue irrespective of the damaged tissue type. This ability of MPCs has been demonstrated in experiments with brain injuries [51], wound healing, and tissue regeneration sites [52]. Apparently, signals from the site of damage are required for MPCs to travel and fill this niche by engrafting and proliferating. Fibroblasts have demonstrated a similar effect of accumulating at the site of injury [4]. Theoretically, a tumor can be considered as a site of “damage” or at least a site of production of cytokines and chemokines of a different nature. It remains to be tested, at least in a model system, which conditions are required for MPCs to see a tumor as an attractant. Another attractive approach that may seem relevant is to introduce an engineered affinity or tropism to MPCs, making them artificially attracted to a tumor.

Summary

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Summary
  8. Acknowledgements
  9. References

The current study demonstrated the ability of MPCs to foster expression of a suicide gene and to support replication of an adenovirus as potential anticancer therapeutic payloads. The potential of MPCs as cell-based vectors for delivery of therapeutic genes and viruses certainly warrants further investigation. Our observations to date establish some of the key properties of this cell population that allow them to function as cellular vectors and thus represent a rational strategy for cancer gene therapy.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Summary
  8. Acknowledgements
  9. References

We thank Ming Wang for performing quantitative PCR and Joel Glasgow for proofreading of the manuscript and helpful advice. We give special acknowledgment to the Department of Pathology, UAB Core Center for Musculoskeletal Disorders (NIH RCC grant P30AR46031) for obtaining MPCs and for their useful advice. This work was supported in part by grants from the National Institutes of Health (R01 CA94084, P50 CA83591, R01 CA83821) and from the Lustgarten Foundation and Susan B. Komen Foundation

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Summary
  8. Acknowledgements
  9. References