SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

Objective

To investigate the role of autophagy in the regulation of cell death in rheumatoid arthritis synovial fibroblasts (RASFs).

Methods

RASFs and osteoarthritis synovial fibroblasts (OASFs) were treated with thapsigargin (TG), an inducer of endoplasmic reticulum (ER) stress, and MG132, a proteasome inhibitor. Then, 3-methyladenine was used as an autophagy inhibitor and bafilomycin A1 as a lysosome inhibitor. Polyubiquitinated proteins, p62, and autophagy induction were evaluated by immunoblotting, immunofluorescence microscopy, and immunohistochemistry, respectively. OASFs were transfected with small interfering RNA targeting autophagy-linked FYVE protein (ALFY). Cell death was evaluated by flow cytometry and a caspase 3 activity assay.

Results

In RASFs, the induction of autophagy by TG and MG132 was increased compared to that in OASFs. Whereas autophagy promoted a caspase 3–independent induction of cell death under ER stress, autophagy had a protective role in apoptosis induced by proteasome inhibition. Treatment of RASFs with 3-methyladenine blocked TG-induced cell death. ER stress induced a strong accumulation of p62-positive polyubiquitinated protein aggregates, accompanied by the formation of large vacuoles in RASFs but not OASFs. Furthermore, TG-induced p62 protein expression was increased, whereas TG-induced ALFY expression was reduced, in RASFs compared to OASFs. ALFY knockdown promoted the accumulation of p62, the formation of polyubiquitinated protein aggregates, and cell death.

Conclusion

Our data provide the first evidence of a dual role of autophagy in the regulation of death pathways in RASFs. A reduced expression of ALFY and the formation of p62-positive polyubiquitinated protein aggregates promote cell death in RASFs under severe ER stress.

Rheumatoid arthritis (RA) is characterized by chronic joint inflammation and progressive destruction of cartilage and bone, which leads to severe joint pain and, ultimately, loss of function. RA synovial fibroblasts (RASFs) residing in the joint have emerged as key players in the pathogenesis of RA. RASFs are capable of producing a large set of inflammatory cytokines, chemokines, and matrix-degrading enzymes, thereby actively contributing to the inflammatory and joint-destructive state in RA ([1]).

Since RA synovial cells produce large amounts of cytokines and enzymes, as much as 30% of all newly synthesized, endoplasmic reticulum (ER)–sorted proteins are unfolded ([2]). When the level of unfolded proteins exceeds the capacity of this organelle, defective proteins are eliminated by a ubiquitin/proteasome-degrading process called the unfolded protein response (UPR) ([2]). The UPR is the main nonlysosomal degradative pathway for ubiquitinated proteins. In contrast, autophagy is a highly regulated and evolutionary conserved process of lysosome-mediated degradation of organelles and cellular components that is activated by various cellular stress conditions, such as ER stress, hypoxia, starvation, heat shock, and microbial infection ([3]). During cellular stress, large quantities of proteins are damaged, resulting in their unfolding/misfolding, polyubiquitination, aggregation, and possibly, the induction of cell death. The robust and efficient removal of these toxic factors by autophagy can help to relieve the cell of stress and reinstate homeostasis ([4]).

Although autophagy constitutes a cytoprotective response activated by cells to cope with stress and is rather protective of cell death, induction of autophagy has also been reported to lead to cell death, generally called autophagic cell death or cell death associated with autophagy ([5]). Increased induction of autophagy in RASFs as compared to osteoarthritis synovial fibroblasts (OASFs) was recently described ([6]). However, the role of autophagy and its connection to death pathways in RASFs is incompletely understood. In addition, we asked whether autophagy and UPR could compensate for each other. In the present study, we investigated the functional role of autophagy activation by the accumulation of polyubiquitinated proteins in RASFs and the role of this pathway in the regulation of cell death pathways.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

Patient samples and cell preparation

SFs were derived from synovial tissue specimens obtained from RA and OA patients during joint replacement surgery (Department of Orthopedic Surgery, Schulthess Clinic Zurich, Zurich, Switzerland). All patients fulfilled the American College of Rheumatology criteria for classification of RA ([7]) or for OA ([8]). All patients provided informed consent. Cells were cultured as described elsewhere ([8]) and used between passages 4–8 for all experiments.

Treatment of cells

RASFs and OASFs were treated with 5 μM thapsigargin (TG; Sigma-Aldrich) or 50 μM MG132 (Merck) for the indicated times, in the presence or absence of 5 mM of the autophagy inhibitor 3-methyladenine (3-MA; Enzo Life Sciences). Controls were treated with matched amounts of DMSO. Where indicated, 100 nM of bafilomycin A1 (Sigma-Aldrich), which inhibits the fusion of autophagosomes with lysosomes, was added to cell cultures for the last 4 hours of treatment ([9]).

Western blotting

Cells were lysed in Laemmli buffer (62.5 mM Tris HCl, 2% sodium dodecyl sulfate [SDS], 10% glycerol, 0.1% bromphenol blue, 5 mM β-mercaptoethanol). Whole cell lysates were separated on 10% SDS–polyacrylamide gels and electroblotted onto nitrocellulose membranes (Whatman). Membranes were blocked for 1 hour in 5% (weight/volume) nonfat milk in TBS-T (20 mM Tris base, 137 mM sodium chloride, 0.1% Tween 20, pH 7.6). After blocking, the membranes were probed with antibodies against light chain 3B (LC3B; Cell Signaling Technology), ubiquitin Lys48 (Merck), or p62 (Abcam) or with α-tubulin (Abcam) as an endogenous control. As secondary antibodies, horseradish peroxidase–conjugated goat anti-rabbit or goat anti-mouse antibodies (Jackson ImmunoResearch) were used. Signals were detected using enhanced chemiluminescence Western blot detection reagents (GE Healthcare) and the Alpha Imager software system (Alpha Innotech). Expression analysis of specific proteins was performed by pixel quantification of the electronic image.

Analysis of cell death

After treatment, cells were detached with trypsin, washed twice with phosphate buffered saline (PBS), and resuspended in annexin V binding buffer (BD Biosciences) at a concentration of 1 × 106 cells/ml. Next, cells were incubated for 15 minutes at room temperature in the dark with fluorescein isothiocyanate (FITC)–annexin V (BD Biosciences) and propidium iodide (PI; Sigma-Aldrich), and then analyzed by flow cytometry (FACSCalibur; BD Biosciences).

Analysis of caspase 3 activity

Following treatment, cells were detached with trypsin, resuspended in cell culture medium at a concentration of 1 × 106 cells/ml, and incubated for 20 minutes at room temperature in the dark with caspase 3 substrate (NucView 488, Biotium). Caspase 3 activity was analyzed by flow cytometry (FACSCalibur).

Immunocytochemistry

Cells were cultured in chamber slides (Lab-Tek; Nunc), fixed with 4% paraformaldehyde, and permeabilized with Tris buffered saline containing 0.1% Triton X-100. Nonspecific binding was blocked with 1% bovine serum albumin (BSA)/5% goat serum for 40 minutes. Slides were incubated overnight at 4°C with primary antibodies. After washing, slides were incubated for 30 minutes with FITC- or Texas Red–conjugated secondary antibodies (Thermo Scientific) and nuclei were stained with DAPI (Sigma-Aldrich). Slides were covered with fluorescent mounting medium (Dako) and analyzed with a fluorescence microscope (AxioImager; Carl Zeiss).

Immunohistochemistry

After deparaffinization, tissue sections obtained from RA and OA patients were pretreated with citrate buffer (10 mM sodium citrate, pH 6.0). Endogenous peroxidase activity was disrupted with 3% H2O2. Nonspecific protein binding was blocked for 40 minutes with 1% BSA/5% goat serum. Monoclonal rabbit anti–ubiquitin Lys48 antibodies (Merck) and rabbit IgG1 (isotype control) were applied (10 μg/ml) overnight at 4°C. Slides were washed in PBS-T (0.05% Tween 20 in PBS) and incubated with biotinylated goat anti-rabbit antibodies (1:1,000; Jackson ImmunoResearch). The signal was amplified with ABC reagent and detected with 3,3′-diaminobenzidine (both from Vector).

Real-time quantitative reverse transcription–polymerase chain reaction (qRT-PCR).

Total RNA was isolated from cells using an RNeasy Mini Kit (Qiagen) including on-column DNase I (Qiagen) digest, and reverse transcribed. Real-time PCR was performed using SYBR Green (Applied Biosystems) and primers specific for p62 and autophagy-linked FYVE protein (ALFY). The primer sequences were as follows: for p62, 5′-AAGCCGGGTGGGAATGTTG-3′ (forward) and 5′-GCTTGGCCCTTCGGATTCT-3′ (reverse) and for ALFY, 5′-GAATCAGCTGGAGCCCAGA-3′ (forward) and 5′-TCTTGGCTGGTTGGTGAGA-3′ (reverse). Constitutively expressed human GAPDH was measured for internal standard sample normalization, using primers as described elsewhere ([10]). Relative expression of messenger RNA (mRNA) was calculated by the comparative threshold cycle method (ΔΔCt).

Transfection experiments

OASFs (2.5 × 105) were transfected with 1 μM small interfering RNA (siRNA) targeting ALFY or with scrambled siRNA (both from Qiagen) as a control, using an Amaxa basic Nucleofector kit for primary mammalian fibroblasts (Lonza). Twenty-four hours after transfection, cells were treated as indicated and then harvested for RNA isolation, Western blotting, or flow cytometry. Knockdown of ALFY was verified by qRT-PCR.

Statistical analysis

Mean ± SD values were calculated. Unpaired t-tests were used for statistical evaluation of the data (with GraphPad Prism 5.0 software). P values less than 0.05 were considered significant.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

Induction of autophagy in RASFs and OASFs via ER stress and proteasome inhibition

RASFs and OASFs were treated with TG, an inducer of ER stress, or MG132, a proteasome inhibitor, for 3, 6, 12, 24, or 48 hours. LC3-I, an autophagic protein, was converted to LC3-II in a time-dependent manner after treatment with TG (Figures 1A and B) and MG132 (Figures 1C and D), indicating the induction of autophagy. The induction of autophagy was more pronounced in RASFs than in OASFs, both after TG treatment and after MG132 treatment. The most pronounced differences in autophagy induction were observed after 48 hours. Interestingly, in addition to an increased level of conversion from LC3-I to LC3-II in RASFs as compared to OASFs, the onset of autophagy induction was earlier in RASFs than in OASFs, both after TG stimulation and after MG132 stimulation.

image

Figure 1. Determination of autophagy induction by monitoring the conversion of light chain 3 type I (LC3-I) to LC3-II in whole protein extracts derived from osteoarthritis synovial fibroblasts (OASFs) and rheumatoid arthritis synovial fibroblasts (RASFs). OASFs and RASFs were left untreated or were treated with 5 μM thapsigargin (TG) (A and B) or 50 μM MG132 (C and D) for the indicated times. Autophagic flux was monitored in RASFs after 24 hours of treatment with TG or MG132 in the presence or absence of 100 nM bafilomycin A1 (Baf) (E). Results are representative of 3 experiments using cells from different patients. α-tubulin was used as a loading control in all experiments. Values in B and D are the mean ± SD. = P < 0.05 versus OASFs.

Download figure to PowerPoint

Conversion of LC3-I to LC3-II, which is indicative of autophagy induction, reflects only the number of autophagosomes formed, but gives no information about the overall autophagic flux ([11]). Therefore, we stimulated RASFs with TG or MG132 in the presence and absence of the lysosomal inhibitor bafilomycin A1 in order to block the autophagic pathway at a late stage. This experiment showed that both TG and MG132 treatment also increased the autophagic flux in RASFs, as demonstrated by an increased amount of LC3-II in the presence of bafilomycin A1 (Figure 1E).

Induction of cell death by ER stress or proteasome inhibition

Since autophagy was described to have both protective and promotive effects on cell death pathways ([5]), we used flow cytometry to investigate the numbers of dead cells after treatment with TG or MG132 in a time course experiment. The amount of annexin V– or PI-positive cells induced by TG was increased in RASFs as compared to OASFs (Figures 2A and B). Whereas the amount of annexin V– or PI-positive RASFs increased in a time-dependent manner to a maximum of 37 ± 8.2% (mean ± SD), only a maximum of 12 ± 4.8% OASFs was observed. In contrast, RASFs were more resistant to MG132-induced cell death than were OASFs (Figures 2C and D).

image

Figure 2. Determination of cell death in OASFs and RASFs under conditions of endoplasmic reticulum stress or proteasome inhibition. Cells were left untreated or were treated with 5 μM TG (A and B) for the indicated times or with 50 μM MG132 (C and D) for 24 hours. Numbers of dead cells were determined by flow cytometry following annexin V/propidium iodide (PI) staining. Dead and damaged cells were also evaluated microscopically in OASFs and RASFs treated with 5 μM TG for 48 hours or 50 μM MG132 for 24 hours (E). Original magnification × 100. Caspase 3 activity in RASFs treated with 5 μM TG for 96 or 120 hours (solid histograms) or 50 μM MG132 for 6 or 48 hours (solid histograms) was compared with DMSO-treated cells (open histograms) (F). Values in B and D are the mean ± SD; each data point in D represents a single sample. = P < 0.05 versus OASFs. See Figure 1 for other definitions.

Download figure to PowerPoint

These data were supported by the characteristics of the changes in phenotype of RASFs and OASFs induced by treatment. TG treatment (48 hours) induced the formation of large vacuoles in RASFs, but did so only to a lesser extent in OASFs. In contrast, MG132 treatment (24 hours) increased the amount of floating and damaged OASFs, but not RASFs (Figure 2E). Although TG significantly induced annexin V– or PI-positive cells in RASFs after 96 hours (Figures 2A and B), caspase 3 activity was not induced before a treatment period of 120 hours (Figure 2F). These data indicate that apoptosis is minimal and only potentiates late stage ER stress–induced cell death observed in RASFs. In contrast, MG132 treatment strongly induced caspase 3 activity from an early time point (6 hours), highlighting a role for apoptotic pathways after proteasome inhibition.

Effect of autophagy inhibition on cell death and caspase 3 activity after ER stress and proteasome inhibition

To further investigate the role of autophagy in cell death pathways after ER stress induction and proteasome inhibition, RASFs were treated with TG or MG132 in the presence or absence of the autophagy inhibitor 3-MA, an inhibitor of the class III phosphoinositide 3-kinase. As expected, 3-MA reduced the TG-induced and MG132-induced activation of autophagy, as indicated by a reduced conversion of LC3-I (Figure 3A). Interestingly, treatment of RASFs with 3-MA decreased the amount of TG-induced annexin V– or PI-positive cells (Figures 3B and C) and late-stage caspase 3 activation (Figure 3D), whereas it enhanced these pathways induced by MG132. Treatment with 3-MA alone did not affect these pathways (data not shown). These data were also supported by the characteristics of phenotypic changes of RASFs and OASFs after treatment. Treatment with 3-MA inhibited the formation of vacuoles in TG-treated cells but produced additional damage in MG132-treated cells (Figure 3E). These data indicated that autophagy can at least partially compensate for an impaired proteasome, since an active autophagic pathway was protective after MG132 treatment.

image

Figure 3. Effects of autophagy inhibition on cell death and caspase 3 activity under conditions of endoplasmic reticulum stress and proteasome inhibition. A, Inhibition of autophagy induction in protein lysates from RASFs, as determined by Western blotting with anti-LC3 antibodies. Cells were left untreated or were treated for 24 hours with 5 μM TG or 50 μM MG132 in the presence or absence of 5 mM 3-methyladenine (3-MA), an inhibitor of autophagy. α-tubulin was used as a loading control. B and C, Numbers of dead cells in RASFs treated for the indicated times with 5 μM TG or 50 μM MG132 in the presence or absence of 3-MA, as determined by flow cytometry following annexin V/propidium iodide (PI) staining (B). Numbers in each compartment are the percentage of dead cells. Data obtained on 3 individual RASF samples in the presence versus absence of TG (left) and MG132 (right) are also shown (C). = P < 0.05. D, Caspase 3 activity in RASFs treated for the indicated times with 5 μM TG or 50 μM MG132 in the presence (open histograms) or absence (solid histograms) of 5 mM 3-MA. E, Microscopic analysis of OASFs and RASFs treated as in B. Original magnification × 100. See Figure 1 for other definitions.

Download figure to PowerPoint

Induction of polyubiquitinated proteins by ER stress and proteasome inhibition

Both ER stress induction and proteasome inhibition induced polyubiquitinated proteins in RASFs and OASFs (data available upon request from the corresponding author). The amounts of polyubiquitinated proteins were similar in RASFs and OASFs. Only a small amount of polyubiquitinated proteins accumulated with TG treatment, presumably since those were cleared by the active proteasome. Interestingly, immunocytochemistry revealed profound differences in the distribution of polyubiquitinated proteins in RASFs and OASFs after TG treatment (Figure 4A). In RASFs, ER stress induced a strong, dot-like aggregation of polyubiquitinated proteins, accompanied by the formation of large vacuoles, which was rarely observed in OASFs, where ubiquitin was evenly distributed in the cytoplasm. Treatment of RASFs with 3-MA reduced the formation of polyubiquitinated protein aggregates (data not shown). In contrast, MG132 treatment did not reveal any differences in the distribution of polyubiquitinated proteins in RASFs and OASFs. These results indicated that the observed differences in cell death induced by TG in RASFs and OASFs were not due to the amount of polyubiquitinated proteins, but rather, were due to differences in the aggregation of polyubiquitinated proteins under ER stress.

image

Figure 4. Determination of the induction of polyubiquitinated proteins by endoplasmic reticulum stress and proteasome inhibition. A, Immunocytochemical analysis of the distribution of polyubiquitinated proteins in RASFs and OASFs following TG or MG132 treatment. Cells were left untreated or were treated for 48 hours with 5 μM TG or 50 μM MG132, fixed, and then stained with antiubiquitin antibodies (green) and DAPI (blue). Arrow indicates large vacuoles. Original magnification × 400. B, Staining of synovial tissue from OA and RA patients with antiubiquitin antibodies (top) or isotype control (middle). Boxed area in RA sample at the top is shown at higher resolution in the bottom image. Arrows indicate dot-like polyubiquitin aggregates. See Figure 1 for definitions.

Download figure to PowerPoint

Ubiquitin staining revealed strong expression in the lining layer of both RA and OA synovial tissue (Figure 4B), possibly due to the multiple functions of polyubiquitin. This finding also reflects our in vitro data that showed similar amounts of polyubiquitinated proteins in RASFs and OASFs under conditions of ER stress. Under high magnification, we detected dot-like polyubiquitin aggregates in RA tissue; however, it is not clear whether these structures resemble the aggregates found in RASFs in vitro under conditions of severe ER stress.

ER stress–induced expression of p62.

To date, little is known about the mechanisms behind the aggregation of polyubiquitinated proteins. However, p62, one of the known targets of autophagic degradation, has been shown to contribute to the formation of polyubiquitinated protein aggregates through its ability to interact with ubiquitin and to self polymerize ([12, 13]). Although p62 mRNA was increased to a similar extent and in a time-dependent manner in RASFs and OASFs after TG treatment (Figure 5A), p62 protein expression (Figures 5B and C) was increased to a greater degree in RASFs than in OASFs over time, pointing to decreased p62 degradation or increased stability in RASFs during ER stress. Furthermore, p62 colocalized with polyubiquitinated protein aggregates in RASFs after TG treatment (Figure 5D), indicating a functional role of p62 in protein aggregation in RASFs under conditions of ER stress.

image

Figure 5. Expression of p62 induced by endoplasmic reticulum stress in OASFs and RASFs treated for the indicated times with 5 μM TG. A, Expression of p62 mRNA, as determined by quantitative reverse transcription–polymerase chain reaction analysis. Values are the mean ± SD. B, Verification of p62 protein expression by Western blotting. Results are representative of 4 experiments using cells from different patients. α-tubulin was used as a loading control. C, Quantification of the Western blot results. Values are the mean ± SD. = P < 0.05 versus OASFs. D, Immunocytochemical analysis of p62 and polyubiquitinated proteins in RASFs. Cells were treated for 48 hours with 5 μM TG, fixed, and stained with antiubiquitin antibodies (green), anti-p62 antibodies (red), and DAPI (blue), and the 3 images were merged. Arrow indicates a vacuole. Original magnification × 400. See Figure 1 for definitions.

Download figure to PowerPoint

In order to exclude the possibility that the accumulation of p62 was not due to impaired lysosomal function, we analyzed the expression of histone deacetylase 6 (HDAC-6), another known ubiquitin binding protein that can be degraded by autophagy ([14]). HDAC-6 mRNA expression was also induced to a similar extent in RASFs and OASFs by TG treatment. The autophagic turnover of HDAC-6, however, was more pronounced in RASFs than in OASFs, indicating that the autophagic flux after TG treatment was intact, and again confirming that autophagy is more active in RASFs than in OASFs. Moreover, treatment of RASFs with TG in the presence of bafilomycin A1 increased the accumulation of HDAC-6 protein and LC3-II, confirming that HDAC-6 degradation is dependent on a functional autophagic pathway (data available upon request from the corresponding author).

ER stress–induced expression of ALFY in OASFs but not RASFs

ALFY, one of the autophagy adaptors, has been shown to facilitate the autophagic degradation of p62 through its ability to interact both with p62 and with the autophagic membrane ([15, 16]). We therefore sought to determine whether ALFY expression was also altered by TG treatment. Whereas TG induced in a time-dependent manner the expression of ALFY in OASFs, ALFY induction in RASFs was very limited (Figure 6A). These data suggested that low expression of ALFY in RASFs might contribute to the reduced degradation of p62-positive polyubiquitin protein aggregates during ER stress. To verify this, we transfected OASFs with siRNA targeting ALFY (Figure 6B). ALFY knockdown resulted in increased expression of p62 protein (Figure 6C), the formation of polyubiquitinated protein aggregates and vacuoles (Figure 6D), as well as increased cell death after TG treatment (Figures 6E and F). These data showed that ALFY knockdown in OASFs and subsequent treatment with TG resembled the phenotype of RASFs under ER stress. MG132 altered the expression of p62 and ALFY; however, there were no differences between OASFs and RASFs (data not shown).

image

Figure 6. Expression of autophagy-linked FYVE protein (ALFY) induced by endoplasmic reticulum stress. A, ALFY expression in OASFs and RASFs, as determined by quantitative reverse transcription–polymerase chain reaction (qRT-PCR) analysis. Cells were treated for the indicated times with 5 μM TG. B, Small interfering RNA (siRNA)–mediated knockdown of ALFY in transfected OASFs, as verified by real-time qRT-PCR. Values in A and B are the mean ± SD. C, Expression of p62 in OASFs following ALFY knockdown, as determined by Western blotting following treatment with TG. α-tubulin was used as a loading control. D, Formation of polyubiquitinated protein aggregates and vacuoles in OASFs following ALFY knockdown. Cells were transfected with siRNA targeting ALFY and treated for 48 hours with 5 μM TG, fixed, and stained with antiubiquitin antibodies (green) and DAPI (blue). Arrow indicates a vacuole. Original magnification × 400. Dead and damaged cells were also evaluated microscopically (bottom images). Original magnification × 100. E, Numbers of dead cells in OASFs, as determined by flow cytometry using annexin V/propidium iodide (PI) staining. Cells were transfected with siRNA targeting ALFY and treated for 72 hours with 5 μM TG. Numbers in each compartment are the percentages of dead cells. F, Data obtained on 4 individual OASF samples treated as in E. = P < 0.05; ∗∗ = P < 0.01 versus RASFs in A. ∗∗∗ = P < 0.001 for the indicated comparison in B. = P < 0.05 for the indicated comparison in F. See Figure 1 for other definitions.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

Continued removal of unfolded and misfolded proteins by the proteasome pathway and by autophagy is essential for the survival of cells. Both pathways have been reported to be more active in RASFs as compared to control fibroblasts ([17]). Recently, the RA synovium was reported to exhibit a highly increased ER stress–associated gene signature ([18]), and TNFα was shown to further increase the expression of ER stress markers in RASFs ([17]). Consistent with previous data ([6]), we showed that autophagy induction by proteasome inhibition or ER stress is more pronounced in RASFs than in OASFs.

Although autophagy induction is mainly thought to play a protective role, induction of this pathway was also reported to be associated with autophagic cell death in a variety of cancer cells upon in vitro treatment with chemotherapeutic agents ([19-21]). Interestingly, we identified a dual role of autophagy in the regulation of cell death in RASFs. Whereas autophagy promotes cell death induced by ER stress, it plays a protective role in apoptosis induced by proteasome inhibition. We showed that OASFs were more susceptible to apoptosis induced by proteasome inhibition, a pathway that was dependent on caspase 3 activation. These data confirm those from previous studies showing an apoptosis-resistant phenotype in RASFs ([1]). In contrast, we showed that RASFs were more sensitive to an autophagy-associated cell death pathway that was mostly independent of caspase 3.

Autophagic cell death was recently defined as cell death that is accompanied by a massive cytoplasmic vacuolization and LC3 conversion and can be suppressed by the inhibition of the autophagic pathway with the use of chemicals or by genetic means ([5]). Such characteristics can be applied to the ER stress–induced cell death we observed in RASFs. A protective role of autophagy induction on TG-induced cell death in RASFs was previously described. However, the investigators evaluated the amount of dead cells by trypan blue exclusion and caspase 3 activity 60 hours after TG treatment ([6]). In the present study, we showed that caspase 3 is only mildly activated and only at late stages (120 hours) in RASFs treated with TG. We also observed increased numbers of dead RASFs, as assessed by annexin V/PI staining, occurring earliest at 72 hours after TG stimulation, with a peak at 144 hours, time points that were not investigated in other studies.

Proteasome inhibition in animal models of experimental arthritis suggested proapoptotic and antiinflammatory effects in RA ([22-24]). We showed that inhibition of autophagy made RASFs more susceptible to apoptosis induced by proteasome inhibition, indicating that autophagy can at least partially compensate for a reduced clearance of polyubiquitinated proteins in the presence of impaired proteasome function. In addition, autophagy induction was recently shown to compensate for amino acid scarcity during proteasome inhibition and to rescue cells from cell death ([25]). The modulation of autophagy activity combined with proteasome inhibition might provide new opportunities for less harmful therapies in RA. This combination was also shown to have synergistic effects in experimental anticancer therapies ([26, 27]).

Aggregation of polyubiquitinated proteins in RA has not previously been described. Aggregates of p62 containing polyubiquitinated proteins have been reported in neurodegenerative diseases, such as Alzheimer's disease and Parkinson's disease ([28, 29]), and in hepatic disorders, including alcoholic hepatitis and fatty liver disease ([30]). Our data provide the first evidence that an imbalance in the expression of p62 and ALFY plays a critical role in the formation of polyubiquitinated protein aggregates under conditions of ER stress in RASFs and promotes cell death. It has been reported that p62, one of the known targets of autophagic degradation, played an important role in the formation of polyubiquitinated protein aggregates through its ability to interact with ubiquitin and to self-polymerize ([12, 13]). Furthermore, ALFY, one of the autophagy adaptors, has been shown to facilitate the degradation of p62-positive polyubiquitinated protein aggregates through its ability to interact with both p62 and the autophagic membrane ([15, 16]). We showed that low induction levels of ALFY after ER stress contributed to the aggregation of p62-positive polyubiquitinated proteins and the induction of autophagic cell death in RASFs in vitro. Furthermore, we detected a strong staining for ubiquitin in the lining layer of RA and OA tissue. The in vivo relevance of our data is supported by a recent study showing increased expression of the autophagy markers LC3-II and beclin 1 in the lining layer of RA tissue as compared to OA tissue, which negatively correlated with apoptosis induction in RA ([31]).

In summary, our data provide the first evidence of a dual role of autophagy in the regulation of death pathways in RASFs. Autophagy activation exhibited a protective role in MG132-induced apoptosis and contributed to the apoptosis-resistant phenotype seen in RASFs. In contrast, RASFs were hypersensitive to autophagy under conditions of severe ER stress induced by TG, which was associated with an imbalance in the expression of p62 and ALFY, leading to the formation of polyubiquitinated protein aggregates and nonapoptotic cell death.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Klein had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Kato, R. Gay, S. Gay, Klein.

Acquisition of data. Kato, S. Gay, Klein.

Analysis and interpretation of data. Kato, Ospelt, R. Gay, S. Gay, Klein.

Acknowledgments

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

We thank Dr. Christoph Kolling for obtaining synovial tissues for our research project, Maria Comazzi and Peter Künzler for excellent technical assistance, and Prof. Dr. Beat A. Michel for his support.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES