Baculovirus has emerged as a novel vector for in vitro and in vivo gene delivery due to its low cytotoxicity and non-replication nature in mammalian cells, but the applications of baculovirus in the genetic modification of human mesenchymal stem cells (hMSCs) and tissue engineering are yet to be reported. In this study, we genetically engineered hMSCs with a baculovirus (Bac-CB) expressing bone morphogenetic protein-2 (BMP-2). Bac-CB transduction of hMSCs at a multiplicity of infection of 40 triggered effective differentiation of hMSCs into osteoblasts. Supertransduction at day 6 after initial transduction enhanced the BMP-2 expression and further accelerated the in vitro osteogenesis, as confirmed by alkaline phosphatase assay, Alizarin red staining and reverse transcription-polymerase chain reaction analysis of osteoblastic genes. Implantation of the supertransduced cells at ectopic sites in the nude mice resulted in efficient cell differentiation into osteoblasts at week 2 and induced progressive mineralization and partial bone formation at week 6, as confirmed by hematoxylin and eosin, immunohistochemical and Alizarin red staining. These data collectively demonstrated, for the first time, the potential of baculovirus in hMSCs engineering and implicated its use in bone tissue engineering.
Successful management of fracture nonunions and segmental bone defects often requires bone grafting, which however, is often hindered by the limited supply of autologous bone and significant morbidity associated with its harvest.1 Therefore, orthopedic surgeons have investigated osteoinductive growth factors and tissue engineering as possible solutions for bone healing problems. Among the numerous growth factors that potentially expedite bone healing, bone morphogenetic protein-2 (BMP-2) is potently osteoinductive and has received FDA approval to induce fusion of the lumbar spine and acute tibial fractures.1 However, direct administration of exogenous BMP-2 requires large amounts of proteins to induce biologic effects in humans.2 Therefore, gene therapy in combination with cell therapy, for increasing local production of osteoinductive BMP-2 within the defect, evolves to be an attractive option to speed up bone repair.3 Ex vivo gene/cell therapy approaches for delivering BMP-2 to improve bone healing have been reported using murine bone marrow stromal cells,4 human muscle-derived cells,5 pluripotent mesenchymal cells6 and murine osteoprogenitor cell line.7 In addition to these cell sources, mesenchymal stem cells (MSCs) are gaining growing interest as a promising cell therapy platform, thanks to their extensive capability of self-renewal and multi-lineage differentiation into adipocytes, chondrocytes and osteoblasts under appropriate environmental cues.8 Bone marrow-derived MSCs have been shown to heal critical-size femoral defects in an athymic rat model when implanted with a hydroxyapatite carrier.9 Furthermore, MSCs have been genetically modified to express BMP-2 to induce in vivo osteogenic differentiation and bone formation using adenoviral,10, 11 adeno-associated viral,12 retroviral,7 lentiviral13 and plasmid14 vectors.
Baculovirus (Autographa californica multiple nucleopolyhedrovirus (AcMNPV)) is a DNA virus that infects insects, and has been widely employed for recombinant protein production in insect cells.15 Since the finding that baculovirus is capable of transducing liver cells in 1995,16 the list of permissive cells, including primary and non-dividing cells, has been steadily expanding (for review, see Kost et al.,15 and Hu17, 18). More recently, baculovirus is also shown to transduce efficiently human embryonic stem cells.19 Therefore, increasing efforts have been directed towards developing baculovirus as a gene delivery vector. Prompted by the potential of MSCs for cell therapy and baculovirus for gene delivery, we have demonstrated that human MSCs (hMSCs) derived from bone marrow can be transduced by baculovirus at an efficiency of up to 76%.20 Furthermore, adipogenic, chondrogenic and osteogenic progenitors originating from hMSCs can be transduced by baculovirus, and the transgene (EGFP) expression level and duration rest heavily on the differentiation states at which the committed progenitors are transduced. Among these three lineages, the permissiveness of osteogenic progenitors to baculovirus transduction is relatively independent of the differentiation status and the transduction efficiency remains nearly 60% for hMSCs differentiating along the osteogenic lineage for 4 weeks.20 Very importantly, baculovirus transduction itself does not obstruct the differentiation capacity of hMSCs.20 These data indicate that hMSCs and hMSCs-derived osteogenic progenitors are highly susceptible to baculovirus transduction, thus implicating the possibilities of genetically modifying these cells for tissue engineering.
In this study, we explored the possibility of genetically modifying hMSCs with a baculovirus expressing BMP-2 (Bac-CB), in an attempt to trigger the spontaneous hMSCs differentiation into osteoblasts and stimulate in vivo bone formation. The in vivo ectopic bone formation was evaluated by implanting the baculovirus-transduced cells into the back subcutis of nude mice for hematoxylin and eosin (H&E), immunohistochemical and Alizarin red staining.
Effects of Bac-CB dosage on BMP-2 expression and in vitro osteogenesis
To examine whether the recombinant baculovirus (Bac-CB) expressed BMP-2 and determine the optimal virus dosage, the hMSCs (5 × 105 cells/well) were transduced by Bac-CB at multiplicity of infection (MOI) of 10, 20 and 40, respectively. Regardless of MOI, the enzyme-linked immunosorbent assay (ELISA) measurements (Figure 1a) showed that the BMP-2 concentration reached the maximum at day 3 and then gradually decreased thereafter, indicating the transient expression mediated by baculovirus. The maximum attainable BMP-2 concentration was dosage-dependent, with the highest approaching 9.0 ng/ml at MOI 40. The BMP-2 expression by mock-transduced hMSCs was undetectable at all times.
Whether the BMP-2 expression stimulated hMSCs differentiation towards osteoblasts was assessed by the alkaline phosphatase (ALP) expression and Alizarin red staining. ALP is an early osteoblast marker signaling the onset of matrix maturation and its expression increases during the transition from MSCs to osteoblasts while decreases during the mineralization phase.21 As expected, the mock-transduced hMSCs exhibited virtually no increase in ALP expression in the absence of osteoinductive supplements (data not shown). With higher BMP-2 expression at MOI 40, the ALP expression (Figure 1b) gradually increased over time, reaching the peak at day 9, which was ≈3.7-fold that of mock-transduced cells at the same time, and then decreased. At MOI 10 and 20, the lower BMP-2 expression resulted in a slower increase in ALP activities, which continued to rise until day 14. As such, the ALP expression profile suggested that MOI 40 induced a faster differentiation through osteoblast towards mineralization phase than MOI 20 and 10. Accordingly, the Alizarin red staining (Figure 1c) revealed more calcium spots and a denser calcium deposition on the culture plates at MOI 40 than at MOI 20 and 10. These data indicated that the osteogenic differentiation of hMSCs was guided by Bac-CB-mediated BMP-2 expression in a dose-dependent manner.
Supertransduction elevated BMP-2 expression and accelerated osteogenesis
Now that BMP-2 concentration declined over time, supertransduction using the same virus dosage as the initial transduction was performed at day 6 hoping to elevate BMP-2 concentration (Figure 2a). For the non-supertransduction group (MOI 40), BMP-2 concentration dropped to 1.6 ng/ml at day 9 and further leveled off to 0.7 ng/ml at day 14. Supertransduction (MOI 40) elevated the BMP-2 concentration to 2.9 ng/ml at day 9 and 1.8 ng/ml at day 14. Supertransduction at MOI 20 similarly elevated the BMP-2 concentration, but at a lower level than at MOI 40. Supertransduction at MOI 10 caused a minimal increase in BMP-2 expression.
Interestingly, the increase in BMP-2 expression by supertransduction at MOI 40 did not further elevate ALP expression at days 9 and 14 (Figure 2b), probably because the ALP activities approached saturation. Nonetheless, the Alizarin red staining (Figure 2c) revealed that supertransduction at MOI 40 led to a pronounced increase in calcium spots and calcium deposition compared to supertransduction at MOI 10 and 20, and the non-supertransduction counterpart shown in Figure 1c.
To characterize the osteogenic differentiation, the supertransduced (MOI 40) and non-supertransduced (MOI 40) cells were harvested at days 9 and 14 for reverse transcription-polymerase chain reaction (RT-PCR) analysis of osteopontin and osteocalcin expression. Osteopontin is an early bone marker that is expressed bimodally, with an early peak during the matrix secretion phase and another after initial mineralization of the extracellular matrix.21 Osteocalcin is a late bone marker secreted only by osteoblasts and also signals terminal osteoblast differentiation.21 As shown in Figure 3, the mock-transduced cells cultured for 14 days expressed neither osteopontin nor osteocalcin, indicating no spontaneous osteogenic differentiation of hMSCs under the culture condition. Comparing with the non-supertransduced cells, supertransduced cells expressed similar levels of osteopontin at day 9, suggesting that both groups contained comparable amounts of osteoblasts. However, the higher levels of osteocalcin expressed by the supertransduced cells at day 9 suggested that more supertransduced cells entered the late differentiation phase. The decreased osteopontin expression by supertransduced cells at day 14 compared to non-supertransduced cells further attested that supertransduced cells entered the later osteogenic differentiation stage at a faster rate. Figures 2 and 3 collectively confirmed that supertransduction at MOI 40 accelerated the in vitro osteogenesis.
Bac-CB transduction induced mineralization and ectopic bone formation
Now that Bac-CB transduction/supertransduction induced apparent in vitro osteogenesis, hMSCs were similarly transduced, and supertransduced (MOI 40) at day 6, followed by implantation into the nude mouse models. One million supertransduced (experimental group) or mock-transduced (control group) cells suspended in the alginate solution were co-injected subcutaneously (s.c.) with CaCl2 solution into either side of the back of nude mice.
Immediately after the cell implantation, the specimens were removed for hematoxylin and eosin (H&E) and Alizarin red staining (Figure 4a). The H&E staining showed that the cells in both the control (n=5) and experimental groups (n=5) were uniformly distributed and maintained round morphology (they were detached from the culture plates before implantation) in a similar way. Accordingly, Alizarin red staining revealed no difference in calcium deposition in both groups.
At 2 weeks post-implantation, the control group deposited nearly no calcium and osteocalcin in the matrix as judged by the faint Alizarin red and immunohistochemical staining (left panel, Figure 4b), indicating no osteogenesis for the mock-transduced hMSCs. In sharp contrast, abundant calcium and osteocalcin deposited densely in the matrix of the experimental groups (right panel, Figure 4b), demonstrating the osteoid formation. Notably, the calcium and osteocalcin deposition surrounding the cells were particularly dense, suggesting the formation of matrix vesicles.
To evaluate the mineralization and bone formation for a longer term, the supertransduced cells were implanted into mice as described above for 6 weeks. The Alizarin red staining (Figure 5) revealed accumulation of calcium over time as judged by the growing area that was stained more and more densely by Alizarin red, evidently illustrating the progressive mineralization and ectopic bone formation mediated by Bac-CB transduction. In contrast, very little mineralization was observed for the control group (data not shown).
Since the finding that baculovirus was capable of transducing mammalian cells, baculovirus-mediated gene delivery has been exploited for the development of cell-based assays, delivery of small interfering RNA, identification of MHC class I mimotopes, surface display of eucaryotic proteins, study of gene functions (for review, see Kost et al.,15 Hu17, 18), and even delivery of vaccine immunogens.22 Another appealing application of baculovirus is in vivo gene delivery, but most previous in vivo studies delivered only reporter genes into animal models.23, 24, 25 Although Huser et al.26 developed a baculovirus for in vivo expression of human Factor IX in the neonatal Wistar rats and Pieroni et al.27 demonstrated long-term baculovirus-mediated expression of erythropoietin in the quadriceps of BALB/c and C57BL/6 mice, no therapeutic effects were reported. Only recently did Wang et al.28 show that a baculovirus expressing diphtheria toxin A inhibits the glioma xenograft growth in the rat brain, suggesting the potentials of baculovirus in cancer therapy. However, the applications of baculovirus in mesenchymal stem cell engineering and bone repair have yet to be reported.
In this study, we demonstrated, for the first time, that baculovirus-mediated BMP-2 expression directed the in vitro commitment of naïve hMSCs into osteoblasts at a rate dependent on virus dosage (Figure 1). The BMP-2 expression level mediated by Bac-CB was considerably lower in the hMSCs (<10 ng/ml) than in the rabbit articular chondrocytes (≈500–1000 ng/ml, unpublished data), indicating that BMP-2 expression using the same baculovirus could vary depending on cell type. Despite the transient and low-level BMP-2 expression, Bac-CB transduction at MOI 40 apparently induced the hMSCs to differentiate into late osteoblast stage, as evidenced by ≈3–4-fold ALP expression stimulation at day 9 and declining ALP expression thereafter (Figures 1b and 2b). This rate of differentiation was faster than the in vitro differentiation of hMSCs transduced by BMP-2-expressing adenovirus (MOI 100), which resulted in rising ALP expression from 14 to 21 days post-infection.29 The Bac-CB supertransduction at day 6 further accelerated the differentiation progression as evidenced by the Alizarin red staining (Figure 2c) and RT-PCR analysis of osteopontin and osteocalcin (Figure 3).
Remarkably, the mock-transduced hMSCs failed to differentiate spontaneously into osteoblasts in vivo without the stimulation of BMP-2 at 2 weeks post-implantation (Figure 4b). This was because s.c. implantation model is thought to be a poor situation for bone formation.30 Despite the adverse environment for bone formation and transient BMP-2 expression, the cells supertransduced by Bac-CB exhibited considerably faster osteogenesis in 2 weeks post-implantation than the mock-transduced cells, and enabled progressive mineralization and partial ectopic bone formation (Figure 5). These data agreed with previous findings that transient BMP-2 expression is sufficient to induce irreversible bone formation in vivo.7, 31 Furthermore, in vivo data have established that 0.5–115 mg of single dose of recombinant BMP-2 is required to produce ectopic bone.32 Since our BMP-2 production was in the nanogram scale, baculovirus-mediated BMP-2 expression reduced the required in vivo BMP-2 concentration by roughly 1000-fold.
Taken together, this study confirmed the potential of baculovirus-mediated ex vivo gene therapy in the context of hMSC engineering and bone tissue engineering. Although combined applications of stem cells and BMP-2 gene delivery by various vectors have been reported previously, these vectors, however, possess various drawbacks of their own. For example, the plasmid transfection efficiency is generally low.1 Retrovirus and lentivirus vectors integrate the transgene into the host genome, which could result in insertional mutation and raise safety concerns. Moreover, the long-term expression mediated by these vectors could lead to heterotopic bone formation.13 Adeno-associated virus also confers long-term gene expression, but the cloning capacity is limited to 3.5–4 kb. Moreover, the manufacture of aforementioned viral vectors is cumbersome and production to high titers is generally difficult.33 By far, the majority of gene therapy treatments for bone engineering have used adenovirus vectors because they can be produced to high titers and can transduce dividing and non-dividing cells. The delivered gene remains episomal and does not replicate.34 However, the significant immune responses elicited as a result of adenoviral protein expression raises safety concerns and limits its application.33 Although gutless vectors have been developed to minimize the immune response,35 the propagation of gutless vectors requires the co-infection of helper viruses, making it difficult to purify gutless vector away from the helper virus. Furthermore, the majority of human population has been exposed to adenovirus infection, thus pre-existing immunity may limit the application of adenovirus.
In contrast to these vectors, baculovirus neither replicates nor is toxic inside the transduced cells.15 Baculoviral DNA degrades in the cells over time36, 37 and no evidence currently exists that baculoviral DNA integrates into host chromosomes in the absence of selective pressure.38 These attributes reduce the potential side effects and ease the safety concerns. Furthermore, since baculovirus is an insect virus in nature, it is unlikely that humans are exposed to baculovirus and develop pre-existing immunity. The baculovirus genome is large, thus conferring a huge cloning capacity of up to 38 kb.39 Moreover, recombinant baculovirus construction is easy and can be achieved in 7–10 days, and baculovirus can be propagated to high titers easily by infecting its natural host insect cells.18 One general disadvantage associated with baculovirus is that the virus confers only transient expression.17 However, unlike the treatment of chronic diseases, it is neither necessary nor desirable for transgene expression to persist beyond the few weeks or months needed to achieve healing3 and a short-term expression of BMP-2 is sufficient to induce irreversibly bone formation, suggesting that a stable genetic modification of hMSCs is not required for bone repair.7, 31
One major concern regarding the use of baculovirus in bone tissue engineering is whether baculovirus triggers in vivo immune responses in immunocompetent animals or humans. Recent studies revealed that baculovirus can induce antiviral effects in transduced mammalian cells40, 41 and innate immune responses in macrophages.42 Baculoviral promoters such as early to late (ETL)43 are also found to be functional in mammalian cells. In addition, a DNA microarray study has identified the transcription of at least 12 viral genes (for example, orf149, ie-0 and gp64) in baculovirus-transduced mammalian cells.44 These studies suggest the possibility that small quantities of baculoviral polypeptides may be synthesized in the transduced cells. However, baculoviral protein production is undetectable in these studies and there is no evidence of serious side effects in previous in vivo gene therapy studies. Furthermore, it is recently discovered that baculovirus strongly potentiates adaptive immune responses by inducing type I interferon.45 In this study, immunodeficient nude mice were used to avoid the possible immune responses against the xenogenic hMSCs; thus, whether baculovirus-engineered hMSCs-derived osteogenic progenitors provoke host immune response was not evaluated. Further studies to explore this question will be assessed in immunocompetent rabbits to ensure the safe use of baculovirus vectors.
In summary, this is the first report demonstrating that baculovirus-mediated gene therapy is capable of regulating the differentiation status of hMSCs in vitro. The animal study further confirms that baculovirus-mediated gene transfer is able to accelerate in vivo osteogenic differentiation and partial ectopic bone formation, demonstrating its potential in bone tissue engineering.
Materials and methods
Preparation and culture of bone marrow-derived hMSCs
Bone marrow-derived hMSCs were obtained from Cambrex Co. (Walkersville, MD, USA) and the subsequent hMSCs selection, enrichment and immunotyping by antibody labeling and flow cytometry were performed as described previously.46 The hMSCs were cultured using α-modified minimal essential medium (α-MEM, Hyclone, Ogden, UT, USA) containing 20% fetal bovine serum (FBS, GIBCO, Grand Island, NY, USA), 4 ng/ml basic fibroblast growth factor (R&D Systems, Minneapolis, MN, USA), 100 U/ml penicillin and 100 mg/ml streptomycin in a 37°C, 5% CO2 incubator. The hMSCs were expanded to passage 10 (P10) for all subsequent experiments.
Recombinant baculovirus construction, amplification and transduction
The recombinant baculovirus expressing BMP-2 under the control of cytomegalovirus immediate-early (CMV-IE) promoter was constructed and designated Bac-CB. Specifically, the CMV-IE promoter was PCR-amplified from pcDNA3.1 (Invitrogen, Carlsbad, CA, USA) and cloned into the SmaI site on pFastBac Dual vector (Invitrogen). The full-length DNA encoding human BMP-2 was PCR-amplified from the lambdagt10 vector (ATCC number, 40345) and subcloned into the XhoI site on pFastBac Dual downstream of the CMV-IE promoter. The recombinant plasmid was designated pBac-CB. The recombinant virus was subsequently constructed using Bac-To-Bac system (Invitrogen) according to the manufacturer’s instructions. The virus was propagated by infecting Sf-9 cells (cultured in TNM-FH medium containing 10% FBS) at MOI 0.1 and harvested at 4 days post-infection. Viral titers (pfu/ml) were determined by end-point dilution method using Sf-9 cells as the host.47
The hMSCs were transduced according to a transduction protocol developed recently.46, 48 In brief, the hMSCs were seeded onto six-well plates (5 × 105 cells/well) and cultured overnight. Before transduction, the spent α-MEM was aspirated and the hMSCs were washed with phosphate-buffered saline (PBS). Depending on the MOI, a certain volume of virus supernatant was mixed with PBS (pH 7.4) to adjust the final volume to 500 μl (per well). The transduction was initiated by directly adding the virus mixture to the cells and continued by gently shaking the six-well plates on a rocking plate for 4 h at 25–27°C. For mock transduction, the cells were incubated with 100 μl TNM-FH medium and 400 μl PBS for 4 h at 25–27°C. After the incubation period, the virus solutions were aspirated and the cells were washed with PBS again. The cells were replenished with 2 ml α-MEM containing 20% FBS without any osteoinductive supplements (for example, dexamethasone, β-glycerol phosphate or ascorbic acid-2 phosphate) and incubated at 37°C. Supertransduction was performed at 6 days after initial transduction in a similar way.
Quantification of BMP-2 concentration by ELISA
The hMSCs were either mock-transduced or transduced with Bac-CB at various MOI (40, 20 and 10). The culture medium (0.2 ml) was aspirated at days 1, 2, 3, 6, 9 and 14 after transduction. BMP-2 concentration was measured using an ELISA kit (R&D Systems).
Alkaline phosphatase assay
ALP activity was measured at 3, 6, 9 and 14 days post-transduction as described previously.49 Transduced and mock-transduced cells were trypsinized and lysed in deionized water by freeze/thaw five times, followed by the addition of substrate solution (50 mM glycine, 0.05% Triton X-100, 4 mM MgCl2 and 5 mM p-nitrophenol phosphate, pH 10.3). The reaction continued for 20 min at 37°C and was stopped with 0.1 M NaOH.7 The absorbance at 405 nm for the transduced cells was measured and normalized to that for the mock-transduced hMSCs at each time point to calculate the relative ALP activity.
Reverse transcription-polymerase chain reaction
To examine the osteogenic differentiation stages of transduced hMSCs, the expression of β-actin (control), osteopontin (an early osteoblast marker) and osteocalcin (a late osteoblast marker) was analyzed by RT-PCR. Total RNA was extracted from the cells using RNeasy Mini Kit (Qiagen, Valencia, CA, USA) and quantified spectrophotometrically. Approximately 1 μg of total RNA was used as the template for cDNA synthesis and subsequent PCR (30 cycles) using AccuPower RT/PCR PreMix kit (Bioneer, Daejeon, Korea). The primers specific for human osteopontin were designed as follows: 5′-AAGCTTCCATGGGAATTGGAGTGATTTGCTTTTGCCTC-3′ (forward) and 5′-GGATCCTTAATTGACCTCAGAAGATGCACTATCTAA-3′ (reverse). The primers specific for human osteocalcin were designed as follows: 5′-ATGAGAGCCCTCACACTCCTC-3′ (forward) and 5′-GCCGTAGAAGCGCCGATAGGC-3′ (reverse). The primers specific for human β-actin were designed as follows: 5′-GCACTCTTCCACCTTCCTTCC-3′ (forward) and 5′-TCACCTTCACCGTTCCAGTTTT-3′ (reverse). The PCR products were subjected to 2% agarose gel electrophoresis.
Alizarin red staining
The deposition of calcium phosphate was stained by Alizarin red, which selectively binds to calcium. The cells were washed with PBS, fixed in formalin for 1 h, washed with water again, and stained with Alizarin red (40 mM, pH 4.2, Sigma, St Louis, MO, USA) with gentle shaking at room temperature for 10 min. The cells were then washed with water and PBS (three times).
The in vivo ectopic bone formation by transduced and mock-transduced cells was evaluated by implanting the cells at the s.c. sites of nude mice (6-week-old female mice purchased from National Laboratory Animal Center, Taipei, Taiwan) using a dual-syringe mixing technique.50 The hMSCs were transduced at MOI 40 and cultured as described above. At 6 days post-transduction, the cells were supertransduced at MOI 40, trypsinized the following day, centrifuged and resuspended in PBS (1 × 107 cells/ml). In parallel, mock-transduced hMSCs were cultured for 7 days and processed similarly. For each mouse, 0.1 ml cell suspension (containing 106 cells) was mixed with 0.05 ml 2.4% sterile sodium alginate (Sigma) in PBS and the 0.15 ml alginate/cell suspension was aliquoted into a 1-ml syringe fitted with a three-way stopcock. A 1-ml syringe containing 0.05 ml 0.1 M CaCl2 solution was attached to the opposite side of the stopcock. The cell/alginate and CaCl2 solutions were co-injected s.c. into the back of the nude mice. The supertransduced (experimental groups) and mock-transduced (control groups) cells were injected at either side of the back of the mice.
H&E and immunohistochemical staining
The mice were killed at different time points post-implantation (n=5 for each time point) in compliance with the regulations of Guide for the Care and Use of Laboratory Animals prepared by the Institute of Laboratory Animal Resources, National Research Council. The tissue specimens were fixed, and the sections were de-paraffined and rehydrated using xylene and a gradient of ethanol (100, 90, 70, 50% and PBS), respectively. The sections were subjected to H&E, Alizarin red or immunohistochemical staining for osteocalcin. For osteocalcin staining, the sections were treated with 3% H2O2 and 0.2% hyaluronidase (Sigma) and 4% normal goat serum (Jackson Laboratories, West Grove, PA, USA) was added for blocking. The primary mouse anti-human MAb against osteocalcin was added (1:100 dilution, Abcam, Cambridge, UK) and incubated at room temperature for 2 h. After washing, the sections were incubated with biotinylated goat anti-rabbit/mouse secondary antibody (DAKO, Hamburg, Germany). One hour later, HRP-conjugated streptavidin (DAKO) was added for 30 min, followed by the addition of diaminobenzidine substrate-chromogen solution. The sections were observed using a light microscope and photographed.
We gratefully acknowledge the financial support from the National Health Research Institutes (NHRI-EX96-9412EI), National Science Council (NSC 95-2221-E-007-215) and Ministry of Economy Affairs (MOEA 95A0283P2), Taiwan.
About this article
Molecular Therapy (2009)