| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,1
* Unité Mixte de Recherches 7561, Centre National de la Recherche Scientifique-Université Henri Poincaré Nancy 1, Faculté de Médecine, Vandoeuvre lès Nancy, France;
Unité Mixte de Recherche 8121 "Vectorologie et transfert de gènes" Centre National de la Recherche Scientifique, Institut Gustave-Roussy, Villejuif Cedex, France;
Unité Mixte de Recherches 7563 "Mécanique et Ingénierie cellulaire et tissulaire" (LEMTA), Centre National de la Recherche Scientifique, Université Henri Poincaré Nancy 1, Faculté de Médecine, Vandoeuvre lès Nancy, France
2Correspondence: Unité Mixte de Recherches 7561, Centre National de la Recherche Scientifique-Université Henri Poincaré Nancy 1, Faculté de Médecine, Ave. de la Forêt de Haye, BP184, F-54505 Vandoeuvre lès Nancy, France. E-mail: Pierre.Gillet{at}medecine.uhp-nancy.fr
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: gene delivery • patellar cartilage • apoptosis • MIA
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
These changes result from poorly understood events (trauma, mechanical and biochemical injuries) occurring over a long period that leads to both cartilage matrix degradation and inhibition of matrix components synthesis. Cartilage hypo-cellularity also contributes to the development of clinical (1)
or experimental (2)
OA due to chondrocyte death by either apoptosis or necrosis (3)
. In fact, chondrocytes are critical to the OA process in that the progression of OA can be judged by the viability of chondrocytes and their ability to resist apoptosis. Thus, therapeutic strategies designed to modulate the imbalance between anabolic and catabolic pathways in OA may include inhibition of signaling pathways which result in apoptosis dependent on mature caspase activity (4)
.
One family of molecules, stress proteins, also called heat shock proteins (HSPs), are a highly conserved family of protein that promote cell survival during the apoptotic process (5)
. Among them, HSP70 is induced under various forms of stresses, e.g., thermal, biochemical (energy deprivation, free radical stress...) or biomechanical (6)
. HSPs play a protective role as molecular chaperones in cells by facilitating the folding, intracellular transport, assembly, and disassembly of other proteins. In addition, HSPs protect cells from oxidative damage both in vivo and in vitro and protect cells from either necrosis or apoptosis (7)
.
Concerning cartilage, some experiments have demonstrated that inducible HSP70 is overexpressed in early stages of OA and could be regarded as a protective phenomenon by the articular cartilage (8)
. In vitro, overexpression of HSP70 via gene delivery prevents nitric oxide-induced apoptosis in articular chondrocytes (9)
. In vivo, a preventive induction of chondrocytic HSP70 via intra-articular injection of MG132, a reversible proteasome inhibitor, diminishes the severity of mono-iodoacetate (MIA) -induced OA in rat knee (10)
. However, this protective effect may be contaminated by factors other than HSP70, because MG132 induces the increase of other anti cell death factor, e.g., HSP27.
To dissect direct role of HSP70 on cyto- and/or chondroprotection during experimental OA, we considered that a method to alter genetic information might be suitable. In this study, we first investigated in vitro the cytoprotective properties of HSP70 gene transfer against mono-iodoacetate (MIA) -induced chondrocyte cell death. Iodoacetate, a glycolytic inhibitor, is widely used to induce apoptotic and necrotic phenomena both in vitro (11)
and in vivo (12)
, MIA destabilizing the functionality of mitochondria in chondrocytes by a decrease in the mitochondrial membrane potential, thus leading to the loss of the main source of ATP and energy of cartilage cells. Second, we investigated whether the in vivo electric gene transfer with HSP70 in rat patella could introduce overexpression of HSP70 in the cartilage and thus confer chondroprotection during MIA-induced experimental OA (13
, 14)
. Previous work has established that EP provides an efficient approach for tissue targeted local cartilage expression, e.g., in rat patellar cartilage, with a relatively high and stable level of expression in spite of the density of extracellular matrix (15)
.
|
MATERIAL AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Isolation and culture of rat chondrocytes
Normal articular cartilage was obtained from Wistar male rats (175–200 g, Charles River Laboratories, l’Arbresle, France) killed under anesthesia (ketamine 50 mg·kg–1 and acepromazine (1.25 mg·kg–1). After joint surgery, articular cartilage pieces were aseptically dissected from femoral head caps and chondrocytes were obtained by sequential digestion with pronase (2 mg/mL) and collagenase B (1.5 mg/mL) (Roche, Meyland, France). The cells were washed two times in phosphate-buffered saline (PBS) and cultured to confluence in 75 cm2 flasks at 37°C in a humidified atmosphere containing 5% CO2. The medium used was DMEM/Ham’s F-12 supplemented with L-glutamine (2 mM), gentamicin (50 µg/mL), amphotericin B (0.5 µg/mL), and heat-inactivated fetal calf serum (FCS, 10%) (Invitrogen). In all experiments, chondrocytes were cultured under low FCS conditions (5%, v/v) and used at confluence.
Transfection of rat chondrocytes
Transfection of rat chondrocytes was performed with polyethylenimine (PEI) reagent (Exgen 500, Euromedex, Mundolsheim, France). Briefly, chondrocytes were seeded in 24-well plates in DMEM/F12 with 5% FCS at 37°C in 5% CO2. For each well, 500 ng of plasmid DNA was layered onto the cells for 3 h according to the manufacturer’s recommendations. The mixture was removed and chondrocytes were seeded in DMEM/F12 for 24 h before MIA treatment.
Assessment of cytotoxicity by MTT assay
Cells were seeded in 96-well plates at a density of 5 x 104 cells per well. Cell stress was induced using different concentrations of MIA (2, 5, and 10 µM). The assay is based on the cleavage of the soluble yellow tetrazolium salt MTT (3-(4, 5-dimethylthiazole-2-yl)-2, 5-diphenyltetrazolium bromide) (Sigma, Saint Quentin-Fallavier, France) by dehydrogenase activity in intact mitochondria, forming insoluble purple formazan crystals. After washing, the formazan crystals were solubilized and absorbance of the resulting colored solution was measured at 550 nm (MR 5000 DYNATECH, Guyancourt, France). Cells were incubated with 100 µL of DMEM/F12 and 25 µL of MTT solution (5 mg/mL). After 4 h, the supernatant was removed and 100 µL of solubilization solution (sodium-dodecyl-sulfate/di-methyl-formamide) was added into each well. The plate was incubated overnight before measurement.
Assay for lactate dehydrogenase activity.
Cells were seeded in 24-well plates at a density of 2 x 105 cells per well. "Cytotoxicity Detection Kit" (Roche) measures the lactate dehydrogenase (LDH) activity released from the cytosol of damaged cells into the supernatant and allows to evaluate the percentage of cell death. One hundred microliters of supernatant were incubated with 100 µL of the reaction mixture. After 10 min incubation in the dark, the reaction was stopped by 50 µL of HCl (1N). Absorbance was measured using a plate reader at 490 nm (MR 5000, DYNATECH). Cytotoxicity was determined in comparison using a positive test with a Triton X100 (2%, v/v) diluted in DMEM/F12 solution.
Analysis of transgene expression by RT-PCR
Cells were seeded in 6-well plates at a density of 5 x 105 cells per well, then transfected with the plasmid containing HSP70 as described previously. Chondrocytes were washed twice with PBS, and total RNAs were extracted using TrizolTM reagent (Invitrogen). The amount of total RNA was quantified at 260 nm. cDNA was obtained by reverse transcription using 1 µg of total RNA and oligodT primers. Then, PCR amplification was performed using specific primers for HSP70i (1946-bp product), HSF-1 (615 pb) and GAPDH (816-bp product). Samples were first heated at 94°C for 15 min to activate the HotStarTMTaq polymerase (Qiagen) and amplification was performed for 30 cycles (94°C for 45 s, 60°C for 45 s and 72°C for 1 min). Amplification products were resolved by electrophoresis in a 1% (w/v) agarose gel, stained by ethidium bromide, and photographed under UV light by a computer-assisted camera (WV-CL350, Panasonic, Osaka, Japan).
In vivo studies
Preventive DNA injection and electric pulses delivery
Experiments with animals were conducted in accordance with institutional guidelines following the recommendations of the NIH for the animal experimentation. Six-wk-old male Wistar rats were kept in a 12:12 light-dark cycle (light-on period, 6:00 AM–6.00 PM) in a controlled temperature chamber (24±1°C). Before the electric pulses delivery, rats were anaesthetized using ketamine hydrochloride (50 mg·kg–1) mixed with acepromazine (1.25 mg·kg–1) injected intraperitoneally. As described (15)
, rat legs were shaved, and DNA (30 µg) resuspended in 50 µL of NaCl 0.9% was injected intra-articularly (ia) into both shaved rat knees through the infra-patellar ligament (day 0). After 1 min, transcutaneous electric pulses were applied by means of two custom-made stainless steel parallel plate electrodes (2 cm long, 2 cm large, 1 mm wide and 12.5 mm apart) placed at each side of the knee, previously covered with a conductive gel. Square-wave electric pulses were generated by a PS-15 Jouan electropulsator (GHT 1281, Jouan, St Herblain, France). Pulses were controlled by means of a digital oscilloscope (VC 6025, Hitachi, Tokyo, Japan). As described for mice and rat skeletal muscle (16)
, we applied 8 pulses of 20 ms delivered at a repetition frequency of 1 Hz.
Induction of MIA-induced OA
One dose of MIA (0.03 mg, Merck-Clevenot, Nogent-sur-Marne, France), dissolved in 50 µL of sterile physiologic saline, was injected through the infra-patellar ligament (13)
, 48 h after the electrotransfer of the plasmid coding for the GFP alone or for the GFP fused to the HSP70 coding sequence. Controls animals received an injection of 50 µL of saline (0.9% NaCl) instead of pDNA and were submitted or not to electro gene transfer.
Histological assessment
Rats were killed 10 or 15 days after the MIA or saline injections and the whole knee joints were fixed immediately in 4% paraformaldehyde, decalcified, and embedded in paraffin. Sagittal sections were stained with hematoxylin-eosine-safran (HES), Toluidine blue or Sirius red. The severity of the OA lesions was graded according to Mankin’s score (17)
with minor modifications. This scale evaluates the severity of OA lesions based on proteoglycan integrity (scale 0–4), cellular changes (scale 0–3), collagen fibers integrity (scale 0–3), invasion of the tidemark by blood vessels (scale 0–1), bone modifications (scale 0–2) and structural changes (scale 05, where 0=normal cartilage structure and 5=erosion of the cartilage down to the subchondral bone).
Measurement of proteoglycan synthesis
Proteoglycan synthesis was assessed as described (13)
. Briefly, 10 or 15 days after intra-articular injection of MIA or saline, rats were killed and the patellae from the treated (MIA) or control (saline) knees were carefully dissected out. The amount of 35S-sulfate incorporated in the central part of the patella was counted using scintillation fluid (Hionic Fluor, Packard) in a ß counter (Packard Bell, Paris, France). Results are expressed as the differences between untreated patellae and the MIA-treated ones (i.e., only electroporated patellae and the MIA-electroporated patellae) as the mean percentage of 35S incorporated.
Detection of HSP70 and GFP expression by immunohistochemistry
Patellae and knees were isolated and fixed in 4% paraformaldehyde, decalcified in 10% EDTA, and embedded in paraffin. Sections (5 µm) of paraffin-embedded samples were deparaffinized in toluene, rehydrated in a graded series of ethanol, and incubated with chondroitinase ABC in PBS, pH 8.0 for 1 h 30 min at 37°C. Slices were washed in PBS, then incubated in Triton X 100 (0.3%) in PBS, pH 7.4. Endogenous peroxidase activity was blocked by incubating the sections with 3% (v/v) hydrogen peroxide in water for 30 min. They were further incubated with blocking serum and overlaid with antibodies (rabbit polyclonal anti-HSP70; 1: 300 or rabbit polyclonal anti-GFP, 1/50) overnight at 4°C in a humidified chamber. Biotin-labeled goat anti-rabbit was used as a secondary antibody (45 min at room temperature).
A biotin-streptavidin-peroxidase detection system was used according to the manufacturers’ recommendations. Color was developed with 3,3'-diaminobenzidine (DAB, Sigma) containing hydroxide peroxide and slides were counterstained for 1 min in 1% methyl green. The counterstaining gives green-stained nuclei and specific staining appears as brown-stained cytoplasm. The percentage of positive chondrocytes was measured by counting the positive cells and the total number of cells in the different cartilage layers. For each specimen, 3 microscopic fields were examined in each layer of the cartilage (the superficial, intermediate and deep layers). Each slide was subjected to a double-blind evaluation by trained investigators (CCH, SE) and a maximal variation of only 5% was obtained.
Detection of caspase dependent apoptosis
Caspase-dependent apoptosis was quantified histologically in patellar cartilage by the detection of active caspase-3 into patellar chondrocytes (18)
. We used the same technique of immunostaining as described above (except that the primary antibody was a rabbit polyclonal anti-active caspase-3; 1:300, R&D System, Euromedex). The percentage of positive chondrocytes was measured by counting the positive cells and the total number of cells in the different cartilage layers. For each specimen, 3 microscopic fields were examined in each layer of the cartilage (the superficial, intermediate and deep layers). Each slide was subjected to a double-blind evaluation by trained investigators (CCH, SE) and a maximal variation of only 5% was obtained.
Statistical analyses
All the results in the figures are expressed as mean and SD. All analyses and figures editing were performed using GraphPad Prism (release 4, GraphPad Software, Inc. San Diego, CA, USA). Student’s t test was used to compare a batch to its own control and ANOVA with post hoc Bonferroni when required (i.e., when >2 groups are involved).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
No HSPs induction (HSP27, HSP40, HSP47, HSP90, data not shown) was revealed after gene transfer in all conditions, meaning that the pDNA delivery approach was not stressful for chondrocytes (data not shown). Theses results were confirmed by RT-PCR analysis of HSP70 gene expression in chondrocytes transfected with either the GFP alone or the HSP70/GFP plasmids. A very faint PCR product was detected in wild-type cells and in cells expressing the GFP, due to the presence of a slow "constitutive" expression of HSP70i mRNA in chondrocytes. When cells transfected with HSP70/GFP vector were submitted to RT-PCR analysis, an overexpression of HSP70 mRNA was observed, as early as 12 h after gene transfer (Fig. 1
A). After a maximal expression at 48 h, expression rate of HSP70 mRNA returned to the basal level in 72 h (data not shown). These findings clearly indicated that our plasmid construct was able to generate a transient overexpression of HSP70 in rat chondrocytes.
|
Functionality of the construct
To confirm the functionality of the recombinant protein, we also studied by RT-PCR analysis, the expression of a cytoplasmic co-factor of this protein, HSF1, which is bound to HSP70 in the absence of stress conditions. The results depicted in Fig. 1C
demonstrated that an induction of HSF1 in HSP70/GFP transfected cells when compared with nontransfected cells.
After these prerequisites, we assessed the functionality of recombinant HSP70/GFP fusion protein, by exposing rat chondrocytes to a cytotoxic compound, mono iodo-acetate (MIA). Thus, cells were transfected with plasmid carrying HSP70/GFP sequence and, 24 h after gene delivery, cells were exposed to MIA treatment for 6 and 24 h. MIA toxicity was determined by two independent methods measuring either the percentage of lactate dehydrogenase (LDH) release or the reduction of tetrazolium salt (MTT). The results are presented in Fig. 2
A, B, respectively. In untransfected cells, after 6 h of exposure to MIA, toxicity was only detected, by LDH analysis (Fig. 2A
), with a dose of 10–5 M while no deleterious effects were found with the lower concentrations of MIA.
|
In cells expressing the HSP70 construction, MIA treatment was unable to induce a detectable level of cellular death by LDH analysis, even at the highest concentration (10–5 M). MTT experiments confirmed these results: a weak toxicity was observed with the doses of 5 x 10–6 M (20%) in unmodified cells but only with 10–5 M of MIA in HSP70 expressing cells (35%) (Fig. 2B
). After 24 h of MIA exposure, LDH and MTT analyses revealed a strong cytotoxic effect of MIA in all cells, except for the lowest dose (2x10–6 M). Cells transfected with HSP70 expression vector seemed to be less susceptible to MIA cytotoxicity but overexpression of recombinant protein was not sufficient to protect the cells in the case of a prolonged exposure to MIA.
In vivo experiments
Gene delivery into rat patellae
After these in vitro preliminary studies, which demonstrated the chondroprotective potential of HSP70 overexpression vs. MIA cytotoxicity, in vivo experiments were performed to 1) demonstrate the feasibility of electroporation to achieve pDNA delivery in rat patella and 2) validate the "chondroprotective" potential of HSP70 expression in cartilage cells.
Safety
Rats were submitted to electro gene transfer either with the empty vector coding for the GFP alone or with the plasmid containing the fusion sequence HSP70/GFP. Initially, we evaluated the possible damages that could be caused by external electrical pulses and pDNA injection on rat patellae. To ensure that electric pulses generated no deleterious side effects, a histological study was performed to evaluate the chondrotoxicity of the electric pulses on cartilage. Rats were killed at different times (2 days, 1 wk, 1, 2, and 3 months) after gene delivery and patellar cartilage was carefully removed. Histological stainings were performed to visualize the proteoglycan content (Toluidine blue), the cell viability (HES) and collagen network (Picrosirius red) (15)
.
No deleterious effects were observed at 2 days and at 1 wk (data not shown), for rats submitted to electric pulses alone and for those who received pDNA (either GFP or HSP70/GFP) before the electric pulses delivery. Proteoglycan biosynthesis has been shown to be one of the earlier parameters that could be modified during cartilage alteration. We thus evaluated chondrocyte capacity to synthesize this key component of the ECM: neither the electric pulses alone nor the pDNA injection led to a decrease in the biosynthetic activity of articular chondrocytes (15)
.
Time course expression of HSP70/GFP in patellar cartilage
Considering these results, a follow-up study was thus designed to evaluate transgene persistency and recombinant protein expression after in vivo pDNA electrotransfer (Fig. 3
). Preliminary immunohistological studies demonstrated that GFP expression could be detected up to 2 months after the gene transfer, mainly in the deep zone of the cartilage (containing hypertrophic chondrocytes). In an another set of experiments, the expression of both proteins was studied (Fig. 4
). Immunostaining of HSP70 (Fig. 3)
revealed a basal expression of the protein, localized principally in the deep layer of patellar cartilage in control rats and EGT rats with the reporter gene. No visible variation was observed between both groups. For HSP70/GFP construction-treated rats, a significant overexpression was depicted as soon as 48 h (50% of chondrocytes were positively stained (with GFP antibody) after HSP/GFP gene transfer, P<0.05, Fig. 3
), in all areas of cartilage and this sustained expression was maintained for 3 months, mainly in the deep layer.
|
|
To confirm that overexpression of HSP70 was related to the plasmid transfer into chondrocytes, the same patellae sections were submitted to another immunostaining analysis with a specific GFP antibody (data not shown). Examination of the slides revealed a positive staining of patellar chondrocytes from day 2 to month 3 and positively stained cells were localized in concordant zones of patellar cartilage. We could notice that the number of GFP positive chondrocytes was lower (15–20%) than the number of GFP positive cells in rats transfected with the reporter gene alone (30%, data published in (15)
), at the same time point. To ensure that HSP70 overexpression was consistent with electric pulses mediated gene transfer, an in situ hybridization with a digoxigenin-DNA labeled probe was performed and the results clearly demonstrated the persistency of plasmid DNA in patellar cartilage (data not presented).
Examination of the sections revealed that the maturation-related chondrocytic hypocellularity was lessened, in HSP70/GFP vs. GFP immunostaining positive chondrocytes. In fact, at month 1, cells localized in the superficial layer no more expressed the transgene but ISH staining demonstrated the presence of plasmid DNA. Studies on time course expression revealed a difference between patellae electroporated with GFP expressing vector and those submitted to EP transfer with HSP70/GFP. The number of chondrocytes (localized in the middle and deep layers) was increased (due to a lesser decrease) in animals treated with the HSP70/GFP fusion protein (Fig. 4)
, and this increase was detectable 1 month after transfection (NS) but was only statistically significant at month 3 (P<0.05 Student’s t test). When performing a 2-way ANOVA, the treatment significantly affected the result, whereas there was only a trend (P=0.0651) concerning the time effect.
Cytoprotective effect of HSP70 gene transfer in MIA-induced osteoarthritis
In vitro studies have demonstrated that overexpression of HSP70 by gene transfer was able to protect chondrocytes in culture from the cytotoxic effects of MIA treatment. Previous experiments performed by various groups have established that a single or repeated intra-articular injections of MIA could induce OA-related lesions in a rat model. Experimental OA lesions are very reproducible and dose and time dependant: there have been extensively studied by our group and others (13
, 19
20
21)
. Previous experiment (data not shown) have demonstrated that EGT pulses with GFP vector do not alter the pathological response to MIA exposure.
To determine whether HSP70 overexpression could protect cartilage cells from MIA toxicity, rats were first electroporated with GFP or HSP70/GFP constructs. Forty-8 h after in vivo gene delivery, rats were injected intra-articularly (day 0) with MIA (0.03 mg). HSP70-mediated protection against OA lesions induced by MIA treatment was evaluated by both macroscopic and microscopic examinations, 10 and 15 days after MIA injection (time required for the development of lesions).
A significant change was histologically detected only at 15 days: the severity of OA lesions was more pronounced in control and in rats electroporated with reporter gene alone than in rats transfected with the HSP70/GFP (Fig. 5
A). Indeed, HES staining revealed a strong remodeling of subchondral bone, disappearance of hypertrophic chondrocytes layer and a decrease in cellularity in animals submitted to EP with pcDNA3.1/GFP followed by MIA injection. In rats expressing recombinant HSP70/GFP fusion protein, the structure of patellar cartilage was quite similar to controls, with sufficient cellularity, presence of hypertrophic chondrocytes and lack of bone remodeling. The Toluidine blue staining revealed a decrease in proteoglycan content in animals submitted to EP (pcDNA3.1/GFP and pcDNA3.1HSP70/GFP) exposed to MIA, which was more significant in animals transfected only with the reporter gene. A similar pattern of coloration was observed with picrosirius red staining, which demonstrated a decrease in collagen content. As for Toluidine blue, this reduction of collagen content was more pronounced in animal expressing the GFP only than in animal transfected with pcDNA3.1-HSP70/GFP.
|
The slices were then submitted to active caspase 3 immunostaining and the percentage of positively stained cells (Fig. 5B
) was markedly reduced in animals expressing HSP70 rather than in rats expressing GFP or naive animals injected with MIA. To confirm these results, we evaluated proteoglycan biosynthesis by measuring radiosulfate intake after MIA treatment (Fig. 6
). Proteoglycan biosynthesis was markedly reduced in control rats as in GFP transfected animals (decrease of 45%) whereas rats expressing HSP70/GFP protein displayed only a slight decrease in biosynthetic activity (15%) on day 10. Anabolism study performed 15 days after intra-articular MIA injection confirmed the cytoprotective effect of HSP70 gene transfer, because rats expressing HSP70 recombinant protein presented the lowest decrease of proteoglycan biosynthesis, when compared with control group or animals treated with reporter construct.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Apoptosis is the process in which each cell systematically inactivates its function and disassembles its structure. Recently, the occurrence of apoptotic events was confirmed in the articular chondrocytes of OA patients (23)
. In parallel, numerous in vivo experimental studies have confirmed that chondrocytic apoptotic events are linked with the severity of experimental OA (24
, 25)
. Understanding the interactions that promote chondrocyte apoptosis and cartilage hypocellularity is essential for developing appropriate targeted therapies for inhibition of chondrocyte apoptosis and the treatment of OA. With this in mind, experimental OA models are crucial to establish the linkage between chondrocyte hypocellularity and apoptosis, and to assess that controlled pharmacological inhibition of key steps in cell death process could be targeted for novel therapeutic strategies for the treatment of OA.
Experimental OA and apoptosis
Among experimental OA models (26)
, MIA-induced OA in the rat easily and quickly reproduces dose, time, and site-dependant OA-like lesions, and, in certain circumstances, functional impairment, similar to that observed in human disease (13)
. These parameters, as well as proteoglycan metabolism and cytokine expression in weight bearing areas (20)
, could serve as indicators for studying chondroprotective drugs (27
, 28)
or for evaluating the ability of imaging techniques to detect and evaluate chondral lesions (29)
. Proapoptotic properties of MIA, an ATP depleting agent inhibiting glycolysis, are well established in vitro on various cell lines, like neurons (30)
, Jurkat cells (31)
, cardiomyocytes (32)
, macrophages (33)
. Necrotic and proapoptotic properties of MIA have also been demonstrated in rat chondrocytes in vitro (10)
, this anaerobic cell line being especially sensitive to MIA-induced mitochondrial dysfunction, MIA destabilizing the functionality of mitochondria in chondrocytes by a alteration in the mitochondrial membrane potential (changes in the 
m as measured by JC1 staining) thus leading to the loss of the main source of ATP and energy of cartilage cells (data not shown). Preliminary data have established that caspase-dependant apoptosis occurs in MIA-induced OA in the rat (34)
, as recently confirmed in human OA (23)
. Caspase 3 in its active form is one of the key mediators of apoptosis in its execution phase and its expression may herald imminent apoptosis better than TUNEL assay (35)
and may act as a surrogate specific marker for early chondrocyte apoptosis (22
, 36
37
38)
.
HSP70 and chondroprotection
Consequently, therapeutic modulation of apoptotic caspases could be of great benefits during early phases of OA, and may hold the key for new therapies that prevent progressive hypocellularity and cartilage matrix loss in animals and humans. Among natural caspases inhibitors, HSP 70 prevents the activation and activity of caspases and interact with the caspases recruitment domain of Apaf-1, preventing its oligomerization and the formation of a functional apoptosome. The role of HSP70 is not limited to caspases, and can prevent apoptosis in a caspases independent but mitochondrial dependant manner by a direct interaction with the AIF (7)
.
Numerous in vitro studies (reviewed in (39)
) have demonstrated that transfection increases the synthesis of cytoprotective HSP70 may prove to be of clinical use in critical situations like organ transplantation or ischemia. Concerning chondrocytes, HSP70 induction by proteasome inhibitor MG132 protects articular chondrocytes from cellular death in vitro (10)
and it has been recently demonstrated that HSP70 transfection protects in vitro chondrocytes from NO-induced apoptosis (9)
. Several authors have recently demonstrated the effectiveness of in vivo gene transfection of HSP70 in various models, like ischemia reperfusion injury of rat lungs (40)
, rat (41)
or rabbit heart (42)
, and rat brain (43)
. On the basis of these encouraging results, we transfected the HSP70 plasmid construct into rat chondrocytes to study in vitro and in vivo its effect on subsequent MIA chondrotoxicity.
EGT of HSP70 in rat patella
With this in mind, we had evaluated a strategy of gene transfer by means of the delivery of external electric pulses to the rat patella after intra-articular injection of a reporter gene (15)
. The present study demonstrated in vivo that the fusion protein developed herein was correctly expressed because GFP expression was dependent from the CMV promoter, localized upstream of the HSP70 sequence. Time course expression of GFP revealed a decrease in the GFP staining, probably due to a loss of efficiency in transcription/translation processes, when compared with the expression level obtained with pcDNA3.1-GFP vector. Histological studies of EGT patellae with HSP70/GFP plasmid also revealed an increase (more precisely a lesser decrease) in cell density vs. controls, detectable after 1 month and statistically different after 3 months. This increase in cell density is probably inherent to the overexpression of HSP70, which is able to inhibit apoptosis inherent to the maturation process.
Chondroprotection and HSP70 in rat patella
Histologically and metabolically, the cytoprotective properties of HSP70 gene transfer in cartilage are concordant with those already observed experimentally in vivo with viral gene transfer HSP70 in various models (see above) and with our previous results obtained in the MIA rat model via the induction of HSP70 via MG132 (21)
. However, these protective effects seen in our previous study after MG132 pretreatment could potentially be caused by factors other than HSP70, thus reinforcing the interest of our gene-specific and site-targeted study. Furthermore, the design of a plasmid coding for the GFP fused to the HSP70 coding sequence permits to separate the physiological roles of HSP70, especially in chondrocytes of growing plate in the rat tibia (44)
. With this in mind, overexpression of HSP70 may account for the relative hypercellularity localized in the middle and deep layers in control animals transfected with the HSP70/GFP fusion protein, this increase being detectable 1 month after transfection, but statistically significant at month 3.
As observed in the clinics, caspase-dependant chondrocyte death, either necrotic or apoptotic, is observed in this particular rat model of OA after intra-articular MIA injection, as earlier shown in ACLT and meniscectomized rats (45)
. Consequently, therapeutical modulation of apoptotic caspases could be of great benefits during early phases of OA (46)
. Previous studies have demonstrated that the chondrocytic expression of HSP70 is positively correlated with the clinical severity of OA (8)
and that HSP70 could play a role in cell protection from stress, especially in the early events. Our results suggest that the protective effect of HSP70 in our experimental conditions might be the result of anti-apoptotic mechanisms via interfering with apoptosis downstream of caspase activation and activity, as suggested by our results with active caspase 3 immunostaining. We routinely use this index in evaluating experimental OA (18
, 34)
, because TUNEL is less specific and apoptosis is overestimated by TUNEL assay compared with caspase 3 detection.
Limitations of the study
We have thus evaluated the chondroprotective potential of a preventive electric gene transfer of HSP70/GFP in rat patella during MIA-induced OA in the rat, this model leading to OA-like histological and metabolic changes in all knee compartments, i.e., medial and lateral femoro-tibial compartments as well as femoro-patellar compartment. Hence, rat patella, as an anatomical entity, appears, in this peculiar model, as the archetypal candidate for studying the potential of "purpose-made" targeted electric gene transfer during a "diffuse" metabolically induced OA in the rat knee. To the best of our knowledge, the lack of such studies using non-viral gene transfer to the cartilage has largely been a result of the difficulty of transferring foreign genetic materials to postmitotic, nondividing chondrocytes in hyaline cartilage. Our laboratory recently took advantage of plasmid DNA electrotransfer for intracellular and long-lasting secreted protein expression in patellar cartilage (15)
, thus avoiding the extracellular matrix "barrier" negative influence, as described earlier in cornea, tendon, and even the brain (47)
. As a laboratory tool, electrotransfer appears to be, in vivo, an interesting method since it is chondroneutral, nonimmunogenic, easy and efficient enough to allow postgenome functional studies and gene validation, especially HSP70, particularly in patellar cartilage. Nevertheless, the influence of anti-apoptotic properties of HSP70 on physiologic maturation as well on ageing of normal patellar articular cartilage remains to be established in rat knee.
In summary, our study provides strong evidence for a possible role of HSP70 overexpression in cytoprotection of rat chondrocytes toward MIA toxicity in vitro and in vivo. A possible role for overexpression of HSP70 in chondroprotection is supported by the following: 1
) gene transfer prior to MIA exposure in vitro is sufficient to enhance chondrocyte survival and 2
) preventive electroporation of a functional HSP70 plasmid in rat patella before an intra-articular injection of MIA significantly ameliorates ancillary chondral lesions. Electroporation appears "chondroneutral" and allows a specific plasmid targeting in rat patella, thus limiting putative site effects of HSP dissemination (e.g., in synovium) and avoiding the use of viral mediated gene transfer approach. Although these findings suggest that HSP70 gene transfer may confer "chondroprotection" in rat patella during MIA-induced OA, further work is necessary to understand the molecular mechanisms, for instance, at the mitochondrial level, by which HSP70 afford cytoprotection and avoid possible unwanted side effects.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication November 18, 2004. Accepted for publication October 3, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. Fujii, Y. Isaka, Y. Takabatake, M. Mizui, C. Suzuki, S. Takahara, T. Ito, and E. Imai Targeting of interstitial cells using a simple gene-transfer strategy Nephrol. Dial. Transplant., October 1, 2006; 21(10): 2745 - 2753. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
0.05, **P
