ARTHRITIS & RHEUMATISM Vol. 46, No. 8, August 2002, pp 2095–2104 DOI 10.1002/art.10433 © 2002, American College of Rheumatology
Light-Activated Gene Transduction Enhances Adeno-Associated Virus Vector–Mediated Gene Expression in Human Articular Chondrocytes Michael Ulrich-Vinther,1 Michael D. Maloney,2 J. Jeffrey Goater,2 Kjeld Søballe,3 Mary B. Goldring,4 Regis J. O’Keefe,2 and Edward M. Schwarz5 Objective. To evaluate the effects of ultraviolet (UV) light as an adjuvant for recombinant adenoassociated virus (rAAV) transduction in human articular chondrocytes. Methods. Primary articular chondrocytes and immortalized chondrocytes (tsT/AC62) were exposed to various doses of UV light (0–1,000 J/m2) and infected at various multiplicities of infection (MOIs) with rAAV containing the enhanced green fluorescent protein (EGFP) gene. Cells were analyzed for viability and EGFP expression by fluorescence-activated cell sorting on days 2, 4, and 8 following infection. To evaluate the transduction efficiency in intact articular cartilage, fullthickness explants were exposed to UV light (0–200 J/m2), infected with rAAV-eGFP, and analyzed for transduction via immunohistochemistry. Results. Toxicity from UV exposure was observed at doses >500 J/m2 and >200 J/m2 in primary and
immortalized chondrocyte cultures, respectively. Transduction efficiency was dependent on the UV dose, MOI, and time. In the cell line, the adjuvant effect of UV on the percentage of cells transduced was modest, but 100 J/m2 increased the mean fluorescence intensity (MFI) of the transduced cells 4-fold. In contrast, UV treatment had a profound effect on the transduction efficiency of primary chondrocytes, which reached ⬃100% after exposure to 100 J/m2 of UV light and 103 MOIs for 8 days. Under the same conditions, 200 J/m2 of UV light enhanced the MFI 7-fold. In cartilage explants, there was no difference in the number of transduced chondrocytes at the edge of the explants in the superficial, intermediate, or basal zones; however, 200 J/m2 of UV light increased the transduction efficiency 2-fold at a low MOI. In the center of the explants, the superficial chondrocytes were efficiently transduced; those in the intermediate and basal zones could not be efficiently transduced under any condition. In the superficial chondrocytes, a low MOI and 200 J/m2 of UV light increased the transduction efficiency 3-fold (to 100%). Conclusion. UV light at doses of up to 200 J/m2 (which do not significantly affect cell viability) significantly enhances the transduction efficiency and expression of the transduced gene in cultures of rAAV-infected primary chondrocytes and in chondrocytes in the superficial zone of intact articular cartilage. These findings support the concept that UV-activated gene transduction could be used as an adjuvant for in vivo rAAV articular cartilage gene therapy with low viral titers to prevent and/or treat arthritis.
Dr. Goldring’s work was supported by NIH grant AR-45378. Dr. O’Keefe’s work was supported by NIH grant AR-46545. Dr. Schwarz’s work was supported by NIH grant AR-45971. 1 Michael Ulrich-Vinther, MD: The Center for Musculoskeletal Research, University of Rochester Medical Center, Rochester, New York, and Aarhus Amtssygehus, Aarhus University Hospital, Aarhus, Denmark; 2Michael D. Maloney, MD, J. Jeffrey Goater, MS, Regis J. O’Keefe, MD, PhD: The Center for Musculoskeletal Research, University of Rochester Medical Center, Rochester, New York; 3Kjeld Søballe, MD: Aarhus Amtssygehus, Aarhus University Hospital, Aarhus, Denmark; 4Mary B. Goldring, PhD: New England Baptist Bone and Joint Institute, Beth Israel Deaconess Medical Center, Boston, Massachusetts; 5Edward M. Schwarz, PhD: The Center for Musculoskeletal Research, University of Rochester Medical Center, Rochester, New York, and New England Baptist Bone and Joint Institute, Beth Israel Deaconess Medical Center, Boston, Massachusetts. Address correspondence and reprint requests to Edward M. Schwarz, PhD, The Center for Musculoskeletal Research, University of Rochester Medical Center, 601 Elmwood Avenue, Box 665, Rochester, NY 14642. E-mail:
[email protected]. Submitted for publication January 24, 2002; accepted in revised form April 16, 2002.
Articular cartilage injury (1–4), osteoarthritis (5– 7), and inflammatory cartilage degeneration such as that seen in rheumatoid arthritis (8,9) remain serious clinical problems and, collectively, are among the most preva2095
2096
lent diseases that affect humans. Our inability to cure these conditions is further complicated by the fact that articular cartilage has a very limited ability to respond to damage. The repair tissue produced by articular cartilage lacks normal structural, biomechanical, and biochemical properties. Currently, there is no drug treatment that can regenerate cartilage in a manner that restores the normal function of the articular surface (10). As a result, nonconventional forms of therapy are under investigation, including implantation of synthetic matrices and polymers (11) and stimulation of the residual cartilage with anabolic factors in order to initiate regeneration of the deteriorated tissue (12–14). Available antiinflammatory treatment options are inadequate for preventing the cartilage destruction associated with inflammatory joint diseases (15). Despite the diversity of the etiologies and pathologies of the diseases affecting articular cartilage, they unite in the absence of a therapeutic system that could deliver the necessary drugs or cytokines in a controlled and continuous manner and with a minimum of side effects. Increasing effort is being applied to the development of new therapeutic approaches involving bioengineering and genetic technology (14). Gene therapy is a novel approach that holds great promise for these incurable musculoskeletal diseases (16,17). Recently, the first “proof-of-principle” trials for human gene therapy directed against arthritis were completed (18). These trials have highlighted the practical limitations of ex vivo gene therapy in terms of costs in time, money, and labor. As a result, investigators have focused their efforts on developing in vivo gene therapies, which also have significant limitations that vary depending on the vector used. Previous studies have indicated that naked DNA, retroviral, adenoviral, and herpesvirus-based vectors are, in their present forms, less than ideal for use in human gene therapy (17,19). The deficiencies of these vectors have led us and others to test the recombinant adenoassociated virus (rAAV) vector, which is derived from an endemic, nonpathogenic, parvovirus (20–24), for its potential use in gene therapy for joint diseases (25–31). Indeed, rAAV vectors have several properties that make them attractive for use in musculoskeletal diseases (24). These properties include the absence of a host inflammatory, cytotoxic, or cell-mediated immune response that clears transduced cells and threatens the host; a broad tropism, including skeletal muscle (32,33), synovial cells (25,27), and articular chondrocytes (28); the ability to infect nondividing cells (32,34,35); the ability
ULRICH-VINTHER ET AL
to deliver life-long gene expression in some cell types (36–39); and high-titer (⬎1013/ml) production with facile ultra high-grade purification (20,40). In contrast, the only remaining major limitations to in vivo rAAVmediated gene therapy are the packaging constraints (⬍5 kb), long-term gene expression in dividing cells, and the infidelity of transduction. Following infection, the single-stranded DNA AAV genome rapidly translocates to the nucleus (41,42), where it remains in an inert state until a yet-to-be-identified host DNA polymerase synthesizes the second strand to generate the transducing episome. This second-strand synthesis is the rate-limiting step in rAAV transduction (43,44), often requiring days to complete following in vivo administration. Interestingly, the host DNA polymerase is believed to be a DNA repair enzyme that can be induced by a wide array of environmental stimuli. Among these are gamma and ultraviolet (UV) radiation (22,27,38,45). This finding led to the idea that rAAV gene therapy could be selectively targeted to tissues or even cells by irradiating the selected target prior to infection. Koeberl et al (38) were able to demonstrate that exposure of rAAV to gamma irradiation prior to administration into the tail vein increased its transduction rate in the mouse liver by up to 900-fold. Unfortunately, that protocol required high doses of radiation, which significantly limited its clinical utility. However, we have shown that as little at 30 J/m2 of UV irradiation results in a significant increase in the transduction of fibroblastlike synoviocytes (27). Irradiation of the injured host tissue with nontoxic doses of UV light might be useful clinically, since the expression of the transgenic protein could be magnified, thus permitting a decrease in the titer of vector required to induce an adequate biologic effect. Furthermore, focused UV irradiation of the target tissue could be used to direct rAAV transduction to a limited anatomic area. Together, these techniques would enhance the safety for the patient. With a goal of developing an rAAV-targeted anabolic gene therapy for articular cartilage repair, we examined the effects of light-activated gene transduction (LAGT) using rAAV as a vector for transduction of the enhanced green fluorescent protein (EGFP) gene in cultures of immortalized and primary human articular chondrocytes, as well as articular cartilage explants. We found that LAGT increases both the transduction efficiency and the expression of the transduced gene in rAAV-infected chondrocytes in both isolated cultures and intact cartilage.
LIGHT-ACTIVATED GENE TRANSDUCTION IN HUMAN ARTICULAR CHONDROCYTES
MATERIALS AND METHODS Preparation of rAAV-eGFP. The rAAV transfer vector plasmid we used contains the gene for EGFP under the transcriptional control of the TRUFR promoter, which is flanked by inverted terminal repeats. The rAAV-eGFP vector was prepared by the Gene Core Facility of the University of North Carolina at Chapel Hill, using the adenovirus-free system in which the transfer vector was cotransfected with pXX2 and pXX6 plasmid into 293T cells, as described previously (40). The rAAV-eGFP vector was titered on human embryonic kidney 293 cells, as we have done previously, in order to determine a relative concentration of infectious units, which was ⬃3.5 ⫻ 1010/ml. Because articular chondrocytes are more difficult to transduce, we found that the effective viral titer in the articular chondrocyte cultures was ⬃10 times lower than the viral titer in 293 cell cultures. Human articular chondrocyte cell line. The adult human articular chondrocyte cell line tsT/AC62, immortalized with a temperature-sensitive mutant of SV40 large T antigen (46), was cultured routinely in monolayer culture at 32°C in a humidified sterile atmosphere of 95% air and 5% CO2. The cells were passaged every 7–10 days at a split ratio of 1:2. For experiments, the tsT/AC62 chondrocytes were plated in 24well tissue culture plates (Corning, Corning, NY) at a density of 105 chondrocytes per well (⬃80% confluence) in 300 l/well of Dulbecco’s modified Eagle’s medium (DMEM)/Ham’s F-12 (1:1), containing 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (Gibco BRL, Gaithersburg, MD). The cultures were infected with rAAV-eGFP 48 hours after plating. Primary human articular chondrocyte cultures. All clinical samples were collected after obtaining the patients’ informed consent. The University of Rochester Research Subjects Review Board approved the study protocol. Samples of normal articular cartilage were harvested from the joints of 6 patients undergoing surgery because of an acute fracture or amputation because of osteosarcoma. The sexes of the patients were equally distributed between groups, and their ages ranged from 15 years to 54 years. The articular cartilage specimens were brought to the research laboratory from the operating room immediately after harvesting and maintained in humid and sterile conditions. Articular cartilage was then carefully dissected from the subchondral bone, and finely chopped fragments were washed twice with sterile phosphate buffered saline (PBS; pH 7.4). A 2-step enzymatic digestion was then performed with 1% testicular hyaluronidase (Sigma, St. Louis, MO) in DMEM/Ham’s F-12 for 1 hour, and a prolonged collagenase digestion with 1% clostridial collagenase A (Sigma) in DMEM/Ham’s F-12 (1:1) containing 10% FBS and 1% penicillin/streptomycin (all from Gibco BRL) for 24 hours in order to digest the extracellular matrix. These digestions were performed under vigorous shaking at 37°C. The isolated chondrocytes were filtered (Swinnex filter with 40-m pores; Millipore, Bedford, MA), and resuspended in slightly alkaline DMEM/Ham’s F-12 (1:1) containing 10% FBS and 1% penicillin/streptomycin. The chondrocytes were seeded onto 24-well plastic plates (Corning) at a density of 105
2097
cells per well and cultured under standard conditions at 37°C in a humidified sterile atmosphere of 95% air and 5% CO2. To confirm that the articular chondrocytes did not undergo dedifferentiation in long-term cultures of isolated monolayer cultures, we performed reverse transcription– polymerase chain reaction for markers of articular chondrocyte maturity (aggrecan and types I, II, and X collagen) using GAPDH as a control, before and after experiments. In all cases, the cells maintained characteristics of mature articular chondrocytes. We found high expression of type II collagen and aggrecan, but no detectable expression of either type X or type I collagen in the primary articular chondrocyte cultures or in the chondrocyte cell line before and after experiments. The culture periods did not exceed a maximum of 8 days. Human articular cartilage explants. Healthy fullthickness cartilage explants were cut into square cubes (⬃8 ⫻ 8 mm square), and the subchondral bone was carefully removed from the calcified subchondral tidemark. The specimens were covered with DMEM/Ham’s F-12 (1:1), 10% FBS, 1% penicillin/streptomycin in 24-well plates, and incubated under standard conditions as described above. Great care was taken not to penetrate the calcified layer separating the articular cartilage from the bone. In vitro and in situ transduction of rAAV-eGFP. For transduction of chondrocyte cultures, the growth medium was changed on the second day after plating, and rAAV-eGFP was added to the cultures at 10, 102, 103, and 104 multiplicities of infection (MOIs). Cultures were continued under standard conditions, and growth medium was changed every second day. Transduction of the articular cartilage explants was performed immediately after harvesting at either high concentration (1.1 ⫻ 107 transducing particles/ml) or low concentration (3.5 ⫻ 106 transducing particles/ml) of rAAV-eGFP. Growth medium was changed every second day. UV irradiation of human articular chondrocytes. On the second day after plating, isolated chondrocyte monolayer cultures were exposed to UV irradiation (Stratalinker; Stratagene, La Jolla, CA) at 4 different energy levels (25 J/m2, 50 J/m2, 100 J/m2, and 200 J/m2). Immediately afterward, the cultures were infected with low, intermediate, or high (10, 100, or 1,000) MOI of rAAV-eGFP in 300 l of media. The cartilage explants were irradiated with either 50 J/m2 or 200 J/m2 of UV light and subsequently transduced at either low or high MOI as described above. To determine the lethal dose of UV irradiation, monolayer cultures of tsT/AC62 cells and primary chondrocytes were grown to 80% confluence in 3-cm dishes and then exposed to increasing UV energies. The percentage of dead chondrocytes was determined by trypan blue exclusion, by counting 500 cells, immediately and after 24 hours of exposure. Any changes in the phenotypic appearance of the chondrocytes or the number of mitoses were noted in order to determine UV-induced toxicity. Observation period. On days 2, 4, and 8 after infection, fluorescence-activated cell sorting (FACS) of live chondrocytes was performed to determine intracellular fluorescence expression. The transduction efficiency in the articular explants was determined 8 days after infection.
2098
ULRICH-VINTHER ET AL
FACS analysis. For EGFP detection, chondrocytes were trypsinized and resuspended in PBS and sorted for fluorescence expression with FACSCalibur (Becton Dickinson, Mountain View, CA); 104 cells were counted per acquisition. Both the percentage of live EGFP-expressing cells and the intensity of fluorescence in this population were evaluated. The data were further analyzed with CellQuest software (Becton Dickinson). A maximum level of 5% was set as the background autofluorescence in live, uninfected chondrocytes. Articular cartilage explant immunohistochemistry. Full-thickness cartilage explants were embedded in paraffin and sectioned with a microtome into 3-m sections parallel to the articular surface. Three serial sections were prepared at increments of 200 m from the articular surface to the osteochondral tidemark. This method enabled a 3-dimensional assessment of the transduction efficiency in the explants, which we divided into the superficial, intermediate, and basal zones. Due to significant in situ autofluorescence, cellular EGFP expression was determined by immunohistochemistry using antibodies specific for EGFP according to the manufacturer’s protocol (Clontech, Palo Alto, CA). Transduction efficiency was assessed by quantifying the percentage of chondrocytes immunostaining positive for EGFP at 40⫻ magnification. A total of 300 chondrocytes were counted for each sample. In the various layers, the transduction efficiency was determined in both the periphery and the center of the explant. This allowed us to investigate the ability of rAAV-eGFP to penetrate normal human articular cartilage. As a negative control, EGFP immunohistochemistry was performed on uninfected human articular explants. As a positive control, direct green fluorescence microscopy of cryosections from the same explants was performed. The microscopy and cell counting were performed blindly. Statistical analysis. All data acquisition and analyses were performed blindly. Data were calculated as the mean and SD, and the groups were compared using 2-tailed analysis of variance (ANOVA). Statistical significance was set at P ⬍ 0.05.
RESULTS Chondrocyte sensitivity to UV irradiation. In order to determine the cytotoxic effect of UV irradiation on human articular chondrocytes, cultures of immortalized or primary chondrocytes were exposed to various doses of UV light (Figure 1). The number of dead chondrocytes was determined either immediately (representing the background cell death from manipulation) or 24 hours after irradiation (representing the background plus UV toxicity) by trypan blue exclusion. We found a significant, dose-dependent cytotoxic response to UV irradiation at doses of 200 J/m2 and higher in the immortalized chondrocyte cultures. The primary chondrocyte cultures were more resistant to UV; dose-dependent cytotoxic responses were observed at energies ⬎200 J/m2. Furthermore, we did not observe any change in phenotypic appearance, mitosis, or pat-
Figure 1. Chondrocyte sensitivity to ultraviolet (UV) irradiation. The tsT/AC62 cells (A) and primary human articular chondrocytes (B) were grown in monolayer cultures as described in Materials and Methods. Cultures were then exposed to the indicated doses of UV irradiation and stained with 0.2% trypan blue either immediately (instant cell death; control) or after 24 hours. Values are the mean ⫾ SEM percentage of trypan blue–positive cells of 300 randomly counted cells in 3 independent experiments. ⴱ ⫽ P ⬍ 0.05 versus control.
tern of collagen synthesis (data not shown), in chondrocyte cultures irradiated with ⬍200 J/m2. These results indicate that the level of UV exposure predicted to induce LAGT is well below the cytotoxic threshold. Effects of UV light on rAAV-eGFP gene transduction in human articular chondrocytes in vitro. To evaluate the effects of UV light on rAAV-mediated transduction of human articular chondrocytes in vitro, monolayer cultures of immortalized chondrocytes (tsT/ AC62) and primary articular chondrocytes were exposed to various doses of UV light and infected at various
LIGHT-ACTIVATED GENE TRANSDUCTION IN HUMAN ARTICULAR CHONDROCYTES
2099
Figure 2. Ultraviolet (UV) light enhancement of transduction efficiency and target gene expression in tsT/AC62 cells transduced with a recombinant adeno-associated virus (rAAV) enhanced green fluorescent protein (EGFP) vector (rAAV-eGFP). The tsT/AC62 cells were grown in monolayer cultures and exposed to UV irradiation at doses of 0, 50, or 100 J/m2. They were then infected with rAAV-eGFP at 10 (A and C) or 100 (B and D) multiplicities of infection and analyzed on the indicated days for EGFP by fluorescence-activated cell sorting, as described in Materials and Methods. Values are the mean ⫾ SEM percentage of EGFP-positive cells (A and B) or the mean fluorescence intensity (MFI) (C and D) of 10,000 cells analyzed in 3 independent experiments. ⴱ ⫽ P ⬍ 0.05 versus unexposed controls.
MOIs with rAAV-eGFP immediately afterward. The transduction efficiency and the relative level of EGFP gene expression in the cell line (Figure 2) and primary articular chondrocyte (Figure 3) cultures on days 2, 4, and 8 after infection were determined by FACS. Overall, we found that the transduction efficiency (percentage of EGFP-positive cells) and gene expression (mean fluorescence intensity [MFI] of EGFP-positive cells) in all of the cultures were dependent upon the UV dose, rAAV-eGFP MOI, and time. In the cell line, UV light had a small effect on the percentage of EGFPpositive cells, but a significant effect on the MFI. In contrast, UV irradiation enhanced both the transduction efficiency and the MFI in the primary chondrocytes. Both the MOI and time had predictable effects on transduction efficiency and EGFP gene expression in both the immortalized and primary chondrocytes. Also of note, no significant difference was observed among chondrocytes derived from different patients within the
same treatment group. Thus, it appears that the age and sex of the donor are not factors that contribute to rAAV transduction of isolated human chondrocytes. Some of the results of LAGT on the transduction of the primary chondrocytes are particularly noteworthy. For example, the number of EGFP-positive cells increased from 6% to 12% two days after infection with 10 MOIs at 200 J/m2 of UV. Remarkably, consistent increases in the number of EGFP-positive cells were also observed in the other cultures, which were proportional to the incubation period (4 days and 8 days) and the MOI (102 and 103). In terms of enhancing the expression of the transduced gene, 200 J/m2 of UV light increased the MFI from 200 to 1,300 on day 8 after infection with 103 MOIs (Figure 3F). Collectively, these results indicate that UV could be used to achieve the same enhancement of transduction efficiency and gene expression in rAAV-infected articular chondrocytes at a significantly lower MOI.
2100
ULRICH-VINTHER ET AL
Figure 3. UV light enhancement of transduction efficiency and gene expression in primary human articular chondrocytes transduced with rAAV-eGFP. Primary human articular chondrocytes were isolated and grown in monolayer culture and exposed to UV irradiation at doses of 0, 50, 100, or 200 J/m2. They were then infected with rAAV-eGFP at 10 (A and B), 100 (C and D), or 1,000 (E and F) multiplicities of infection and analyzed on the indicated days for EGFP by fluorescence-activated cell sorting, as described in Materials and Methods. Values are the mean ⫾ SEM percentage of EGFP-positive cells (A, C, and E) or the MFI (B, D, and F) of 10,000 cells analyzed in 3 independent experiments. ⴱ ⫽ P ⬍ 0.05 versus unexposed controls (n ⫽ 6). See Figure 2 for definitions.
Effects of UV light on rAAV-eGFP–transduced human articular chondrocytes in situ. To determine the ability of rAAV to penetrate the extracellular matrix of human cartilage, we studied articular explants from tissues obtained at surgery. The explants were cultured and exposed to various amounts of UV light and infected with rAAV-eGFP at either low concentration (3.5 ⫻ 106 transducing particles/ml) or high concentration (1.1 ⫻ 107 transducing particles/ml). The expression of the transduced EGFP was detected throughout the cartilage 8 days after infection, as determined by immunohistochemistry. An example of this is presented in Figure 4A, which shows the vertical distribution of transduced chondrocytes at the center of a cartilage specimen following exposure to 200 J/m2 of UV light and
infection with 1.1 ⫻ 107 transducing particles/ml of rAAV-eGFP. In the center of the explant, chondrocytes in all layers of the full-thickness explants were transduced, although transduction declined from the superficial layer to the basal layer (Figures 4A–C). This result was also confirmed by fluorescence microscopy, in which similar numbers of EGFP-positive chondrocytes could be detected even in the presence of a high background of autofluorescence (data not shown). At the periphery of the explant, LAGT had significant effects on the transduction of the chondrocytes, where transduction in all zones could be significantly facilitated by UV irradiation at intensities ⬎50 J/m2. Furthermore, the number of EGFP-expressing cells could be increased from an average of 38%
LIGHT-ACTIVATED GENE TRANSDUCTION IN HUMAN ARTICULAR CHONDROCYTES
2101
Figure 4. Immunohistochemistry of rAAV-eGFP–transduced human articular chondrocytes in cartilage explants. Surgically retrieved human articular cartilage explant cultures were exposed to 200 J/m2 of UV irradiation and infected with rAAV-eGFP at high concentration (1.1 ⫻ 107 transducing particles/ml), as described in Materials and Methods. After 8 days of culture, the explants were processed for analysis of EGFP by immunohistochemistry. Micrographs are representative of the findings in A, a full-thickness cartilage section or the center of the cartilage sections, B, the superficial zone, and C, the intermediate zone. Dark brown staining indicates positive EGFP immunoreactivity. (Original magnification ⫻ 10.) See Figure 2 for definitions.
throughout all 3 zones to an average of 84% by 200 J/m2 of UV light at low MOI (Figure 5A), and from 64% to 93% at high MOI (Figure 5B). However, no difference in transduction efficiency was observed among the different zones (Figure 5A and B). Hence, superficial chondrocytes are as transducible as intermediate and basal chondrocytes when they are in the peripheral region of the explant. Quantitation of the sections originating from the center of the cartilage explant, 3–4 mm away from the sides of the cube, revealed that only chondrocytes in the superficial zone were transduced at high efficiencies (Figures 5C–D). The number of transduced chondrocytes in the intermediate and basal zones in the center was significantly decreased compared with all zones in the periphery. There was no significant difference between the numbers of EGFP-expressing chondrocytes in the intermediate and basal zones in the central region (Figure 5D). UV light at 200 J/m2 induced a 3-fold increase in the number of EGFP-expressing chondrocytes in the superficial layer following infection at a low vector concentration (Figure 5C). However, UV light at any dose had no effect on the transduction of chondrocytes in either the intermediate or basal layers. Although we did not measure the ability of UV to penetrate the cartilage directly, the data from these studies indicate that as predicted, the extracellular ma-
trix inhibits UV penetration. This is suggested by the fact that we failed to detect a significant increase in the transduction of chondrocytes in the explants following irradiation with 50 J/m2 of UV (Figure 5A–D), while this dose had a significant effect on chondrocytes when they were removed from the matrix (Figure 3E). Furthermore, the MOI was also an important determinant in the transduction of chondrocytes within the penetrable zones. Another interesting finding is the high transduction efficiency of chondrocytes in the intermediate and basal layers in the periphery compared with the center of the explant, even in the absence of UV irradiation (Figure 5). This result indicates that under conditions of passive diffusion (absence of fluid force), rAAV can effectively penetrate the extracellular matrix of superficial cartilage but may have difficulty penetrating the matrix in the intermediate or basal zones. Additionally, the chondrocytes in the intermediate or basal zones may be transduced as efficiently as the chondrocytes in the superficial zone in regions of cartilage degeneration and surgical debridement. DISCUSSION While current therapies improve symptoms of articular cartilage deterioration secondary to trauma,
2102
ULRICH-VINTHER ET AL
Figure 5. UV light enhancement of transduction efficiency and gene expression in rAAVeGFP–infected chondrocytes in cartilage explants. Human articular cartilage explants were cultured, exposed to the indicated doses of UV irradiation, and infected with rAAV-eGFP at low (3.5 ⫻ 106 transducing particles/ml) (A and C) or high (1.1 ⫻ 107 transducing particles/ml) (B and D) concentration, as described in Materials and Methods. After 8 days of culture, the explants were processed for analysis of EGFP by immunohistochemistry, and the percentages of EGFP-immunostained chondrocytes in the superficial, intermediate, and basal zones were determined at the periphery (A and B) or the center (C and D) of the explant. ⴱ ⫽ P ⬍ 0.05 versus unexposed controls (n ⫽ 6). See Figure 2 for definitions.
osteoarthritis, and erosive arthritis, none is able to restore the structural and biomechanical characteristics of mature hyaline articular cartilage. Vectors based on naked DNA, retrovirus, adenovirus, herpesvirus, and rAAV have been used in the joint in order to transduce genes in vivo (14,19). In their present forms, each of these vectors has shortcomings that render it suboptimal for gene therapy for articular cartilage regeneration. However, based on all of these studies, it is our belief that rAAV vectors have the greatest potential for musculoskeletal gene therapy (24). It has been shown previously that rAAVmediated transduction of human articular chondrocytes is relatively inefficient, requiring an MOI ⬎103 (28). There are several potential reasons for this relatively low transduction efficiency. These include the distribution of the AAV tropic heparan sulfate proteoglycan receptor and coreceptors (fibroblast growth factor receptor 1 [47]
and ␣v5 [48]), which are known to decrease uptake of the virus, increased intracellular degradation (41), and limited second-strand synthesis due to a low level of DNA synthesis (43,44). Our hypothesis is that the major impediment to rAAV-mediated transduction of human articular chondrocytes is second-strand synthesis, which may be overcome by UV irradiation. Our studies demonstrate that UV irradiation markedly enhances the transduction efficiency of human articular chondrocytes at doses of up to 200 J/m2, while not significantly affecting cell viability. At these exposures, we also failed to see changes in the phenotypic appearance, increases in mitosis, or alterations in collagen and aggrecan messenger RNA synthesis (data not shown). However, it is still possible that there are more-subtle changes that could affect cellular metabolism. These changes could result from tautomerization of DNA bases and the generation of reactive species such as oxygen free radi-
LIGHT-ACTIVATED GENE TRANSDUCTION IN HUMAN ARTICULAR CHONDROCYTES
cals, which are produced downstream of UV irradiation and may result in neoplasms. UV treatment also increases the expression of the transduced gene in chondrocytes, probably by increasing the number of rAAV episomes per cell, as we have demonstrated for fibroblast-like synoviocytes (27). Within cartilage, UV irradiation can also significantly increase rAAV-mediated transduction of chondrocytes in the superficial zone of articular cartilage, particularly at lower concentrations of rAAV. However, chondrocytes in the intermediate and basal zones cannot be transduced effectively, which is likely due to decreased virus penetration through the cartilage matrix. This is demonstrated by the dramatic enhanced transduction of cells at the periphery compared with those at the center of the explants. It is difficult to say exactly how these penetration issues will affect the efficacy of treatment of a diseased joint. However, in the case of articular cartilage repair following trauma, one might see the penetration boundaries as an advantage, since this will focus and limit gene therapy to the chondrocytes in immediate proximity to the defect regardless of their zone. One of the goals of gene therapy research is to develop an approach by which genes can be selectively expressed only in specific cells. The rationale for this goal is that the target gene may have detrimental effects if expressed in other cell populations. In the case of articular cartilage injury, the expression of chondrogenic factors (i.e., transforming growth factor , insulin-like growth factor, and bone morphogenetic protein) or intracellular signaling molecules (i.e., SMADs, SOX9) may be useful to initiate a repair response in the defect, but could also potentially induce cartilage abnormalities if expressed in normal tissue. We have proposed that rAAV vectors have several empirical advantages that could make them useful for selective arthritis gene therapy (24,27). One of these is the potential use of tissue-specific promoters to selectively target cells within articular defects. First, the expression of the transduced gene could be restricted to articular chondrocytes by using a cartilage-specific promoter (i.e., type II collagen [49,50] or cartilage-derived retinoic acid–sensitive protein [51]). Then, intraarticular UV irradiation could be delivered, as described by Laing et al (52), to selectively render the desired chondrocytes in the articular defect susceptible to LAGT. Finally, a dose of rAAV that cannot efficiently transduce unstimulated articular chondrocytes (i.e., ⬍103 MOIs [28]) could be injected into the articular defect. Under these conditions, while many different cell types in the joint may be
2103
infected (i.e., synoviocytes [25,27]), only articular chondrocytes will express the target gene. Furthermore, few infected chondrocytes outside the irradiated defect will be transduced due to decreased UV or rAAV penetration (Figure 5) and incompetence for second-strand synthesis. Future in vivo studies to test this await the development of a UV arthroscope for use in small animals, which is currently in progress. In the present study, we found that rAAVmediated gene transduction of articular chondrocytes is a promising strategy that may allow for direct in vivo gene therapy through intraarticular injection or indirect ex vivo approaches that may combine various matrices. Whether targeting chondrocytes with therapeutic genes can be successfully used to treat articular defects awaits studies with in vivo models. However, we believe that the use of LAGT as a tool to direct and focus rAAVmediated gene therapy to articular cartilage has great potential. ACKNOWLEDGMENTS We thank Dr. P. Keng for help with initiating the FACS analysis and Dr. R. J. Samulski for assistance in the rAAV preparation.
REFERENCES 1. Alleyne KR, Galloway MT. Management of osteochondral injuries of the knee. Clin Sports Med 2001;20:343–64. 2. Browne JE, Branch TP. Surgical alternatives for treatment of articular cartilage lesions. J Am Acad Orthop Surg 2000;8:180–9. 3. Kish G, Modis L, Hangody L. Osteochondral mosaicplasty for the treatment of focal chondral and osteochondral lesions of the knee and talus in the athlete: rationale, indications, techniques, and results. Clin Sports Med 1999;18:45–66. 4. Martinek V, Fu FH, Lee CW, Huard J. Treatment of osteochondral injuries: genetic engineering. Clin Sports Med 2001;20: 403–16. 5. Frenkel SR, Di Cesare PE. Degradation and repair of articular cartilage. Front Biosci 1999;4:D671–85. 6. Malemud CJ, Goldberg VM. Future directions for research and treatment of osteoarthritis. Front Biosci 1999;4:D762–71. 7. Vangsness CT Jr. Overview of treatment options for arthritis in the active patient. Clin Sports Med 1999;18:1–11. 8. Hauselmann HJ. Mechanisms of cartilage destruction and novel nonsurgical therapeutic strategies to retard cartilage injury in rheumatoid arthritis. Curr Opin Rheumatol 1997;9:241–50. 9. Jorgensen C, Noel D, Apparailly F, Sany J. Stem cells for repair of cartilage and bone: the next challenge in osteoarthritis and rheumatoid arthritis. Ann Rheum Dis 2001;60:305–9. 10. Ulrich-Vinther M, Maloney MD, Schwarz EM, Rosier RN, O’Keefe RJ. Articular cartilage biology and regeneration. J Am Acad Orthop Surg. In press. 11. Jackson DW, Scheer MJ, Simon TM. Cartilage substitutes: overview of basic science and treatment options. J Am Acad Orthop Surg 2001;9:37–52. 12. Buckwalter JA. Articular cartilage: injuries and potential for healing. J Orthop Sports Phys Ther 1998;28:192–202.
2104
13. Coutts RD, Sah RL, Amiel D. Effects of growth factors on cartilage repair. Instr Course Lect 1997;46:487–94. 14. Evans CH, Ghivizzani SC, Smith P, Shuler FD, Mi Z, Robbins PD. Using gene therapy to protect and restore cartilage. Clin Orthop 2000;379 Suppl:S214–9. 15. Lee DM, Weinblatt ME. Rheumatoid arthritis. Lancet 2001;358: 903–11. 16. Evans CH, Ghivizzani SC, Kang R, Muzzonigro T, Wasko MC, Herndon JH, et al. Gene therapy for rheumatic diseases. Arthritis Rheum 1999;42:1–16. 17. Pfeifer A, Verma IM. Gene therapy: promises and problems. Annu Rev Genomics Hum Genet 2001;2:177–211. 18. Evans CH, Ghivizzani SC, Palmer GD, Gouze JN, Robbins PD, Gouze E. Gene therapy for rheumatoid arthritis. Expert Opin Biol Ther 2001;1:971–8. 19. Ghivizzani SC, Oligino TJ, Glorioso JC, Robbins PD, Evans CH. Gene therapy approaches for treating rheumatoid arthritis. Clin Orthop 2000;379 Suppl:S288–99. 20. Monahan PE, Samulski RJ. Adeno-associated virus vectors for gene therapy: more pros than cons? Mol Med Today 2000;6: 433–40. 21. Rabinowitz JE, Samulski RJ. Building a better vector: the manipulation of AAV virions. Virology 2000;278:301–8. 22. Samulski RJ, Srivastava A, Berns KI, Muzyczka N. Rescue of adeno-associated virus from recombinant plasmids: gene correction within the terminal repeats of AAV. Cell 1983;33:135–43. 23. Samulski RJ, Chang LS, Shenk T. A recombinant plasmid from which an infectious adeno-associated virus genome can be excised in vitro and its use to study viral replication. J Virol 1987;61: 3096–101. 24. Schwarz EM. The adeno-associated virus vector for orthopaedic gene therapy. Clin Orthop 2000;379 Suppl:S31–9. 25. Pan RY, Xiao X, Chen SL, Li J, Lin LC, Wang HJ, et al. Disease-inducible transgene expression from a recombinant adeno-associated virus vector in a rat arthritis model. J Virol 1999;73:3410–7. 26. Pan R-Y, Chen S-L, Xiao X, Liu D-W, Peng H-J, Tsao Y-P. Therapy and prevention of arthritis by recombinant adeno-associated virus vector with delivery of interleukin-1 receptor antagonist. Arthritis Rheum 2000;43:289–97. 27. Goater J, Muller R, Kollias G, Firestein GS, Sanz I, O’Keefe RJ, et al. Empirical advantages of adeno associated viral vectors in vivo gene therapy for arthritis. J Rheumatol 2000;27:983–9. 28. Arai Y, Kubo T, Fushiki S, Mazda O, Nakai H, Iwaki Y, et al. Gene delivery to human chondrocytes by an adeno associated virus vector. J Rheumatol 2000;27:979–82. 29. Cottard V, Mulleman D, Bouille P, Mezzina M, Boissier MC, Bessis N. Adeno-associated virus-mediated delivery of IL-4 prevents collagen-induced arthritis. Gene Ther 2000;7:1930–9. 30. Watanabe S, Imagawa T, Boivin GP, Gao G, Wilson JM, Hirsch R. Adeno-associated virus mediates long-term gene transfer and delivery of chondroprotective IL-4 to murine synovium. Mol Ther 2000;2:147–52. 31. Zhang HG, Xie J, Yang P, Wang Y, Xu L, Liu D, et al. Adeno-associated virus production of soluble tumor necrosis factor receptor neutralizes tumor necrosis factor alpha and reduces arthritis. Hum Gene Ther 2000;11:2431–42. 32. Fisher KJ, Jooss K, Alston J, Yang Y, Haecker SE, High K, et al. Recombinant adeno-associated virus for muscle directed gene therapy. Nat Med 1997;3:306–12. 33. Herzog RW, Hagstrom JN, Kung SH, Tai SJ, Wilson JM, Fisher KJ, et al. Stable gene transfer and expression of human blood coagulation factor IX after intramuscular injection of recombinant adeno-associated virus. Proc Natl Acad Sci U S A 1997;94:5804–9. 34. Xiao X, Li J, Breese GR, Samulski RJ. Differential and persistent expression patterns of CNS gene transfer by an adeno-associated virus (AAV) vector. Brain Res 1996;713:99–107.
ULRICH-VINTHER ET AL
35. Ozawa K, Fan DS, Shen Y, Muramatsu S, Fujimoto K, Ikeguchi K, et al. Gene therapy of Parkinson’s disease using adeno-associated virus (AAV) vectors. J Neural Transm Suppl 2000;58:181–91. 36. Xiao X, Li J, Samulski RJ. Efficient long-term gene transfer into muscle tissue of immunocompetent mice by adeno-associated virus vector. J Virol 1996;70:8098–108. 37. Kaplitt MG, Leone P, Samulski RJ, Xiao X, Pfaff DW, O’Malley KL, et al. Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat Genet 1994;8:148–54. 38. Koeberl DD, Alexander IE, Halbert CL, Russell DW, Miller AD. Persistent expression of human clotting factor IX from mouse liver after intravenous injection of adeno-associated virus vectors. Proc Natl Acad Sci U S A 1997;94:1426–31. 39. Wang L, Nichols TC, Read MS, Bellinger DA, Verma IM. Sustained expression of therapeutic level of factor IX in hemophilia B dogs by AAV-mediated gene therapy in liver. Mol Ther 2000;1:154–8. 40. Xiao X, Li J, Samulski RJ. Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus. J Virol 1998;72:2224–32. 41. Douar AM, Poulard K, Stockholm D, Danos O. Intracellular trafficking of adeno-associated virus vectors: routing to the late endosomal compartment and proteasome degradation. J Virol 2001;75:1824–33. 42. Seisenberger G, Ried MU, Endress T, Buning H, Hallek M, Brauchle C. Real-time single-molecule imaging of the infection pathway of an adeno-associated virus. Science 2001;294:1929–32. 43. Ferrari FK, Samulski T, Shenk T, Samulski RJ. Second-strand synthesis is a rate-limiting step for efficient transduction by recombinant adeno-associated virus vectors. J Virol 1996;70: 3227–34. 44. Fisher KJ, Gao GP, Weitzman MD, DeMatteo R, Burda JF, Wilson JM. Transduction with recombinant adeno-associated virus for gene therapy is limited by leading-strand synthesis. J Virol 1996;70:520–32. 45. Alexander IE, Russell DW, Spence AM, Miller AD. Effects of gamma irradiation on the transduction of dividing and nondividing cells in brain and muscle of rats by adeno-associated virus vectors. Hum Gene Ther 1996;7:841–50. 46. Robbins JR, Thomas B, Tan L, Choy B, Arbiser JL, Berenbaum F, et al. Immortalized human adult articular chondrocytes maintain cartilage-specific phenotype and responses to interleukin-1. Arthritis Rheum 2000;43:2189–201. 47. Qing K, Mah C, Hansen J, Zhou S, Dwarki V, Srivastava A. Human fibroblast growth factor receptor 1 is a co-receptor for infection by adeno-associated virus 2. Nat Med 1999;5:71–7. 48. Summerford C, Bartlett JS, Samulski RJ. AlphaVbeta5 integrin: a co-receptor for adeno-associated virus type 2 infection. Nat Med 1999;5:78–82. 49. Yamada Y, Miyashita T, Savagner P, Horton W, Brown KS, Abramczuk J, et al. Regulation of the collagen II gene in vitro and in transgenic mice. Ann N Y Acad Sci 1990;580:81–7. 50. Zhou G, Garofalo S, Mukhopadhyay K, Lefebvre V, Smith CN, Eberspaecher H, et al. A 182 bp fragment of the mouse pro alpha 1(II) collagen gene is sufficient to direct chondrocyte expression in transgenic mice. J Cell Sci 1995;108:3677–84. 51. Xie WF, Zhang X, Sandell LJ. The 2.2-kb promoter of cartilagederived retinoic acid-sensitive protein controls gene expression in cartilage and embryonic mammary buds of transgenic mice. Matrix Biol 2000;19:501–9. 52. Laing TJ, Ike RW, Griffiths CE, Richardson BC, Grober JS, Keroack BJ, et al. A pilot study of the effect of oral 8-methoxypsoralen and intraarticular ultraviolet light on rheumatoid synovitis. J Rheumatol 1995;22:29–33.
