If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Schroeder Arthritis Institute, Krembil Research Institute, University Health Network, Toronto, Ontario, CanadaInstitute of Medical Science, Department of Medicine, University of Toronto, Ontario, CanadaDepartment of Rheumatology, University of Toronto, Ontario, Canada
Schroeder Arthritis Institute, Krembil Research Institute, University Health Network, Toronto, Ontario, CanadaDepartment of Surgery, University of Toronto, Ontario, Canada
Schroeder Arthritis Institute, Krembil Research Institute, University Health Network, Toronto, Ontario, CanadaInstitute of Medical Science, Department of Medicine, University of Toronto, Ontario, CanadaDepartment of Rheumatology, University of Toronto, Ontario, Canada
Schroeder Arthritis Institute, Krembil Research Institute, University Health Network, Toronto, Ontario, CanadaDepartment of Surgery, University of Toronto, Ontario, Canada
Schroeder Arthritis Institute, Krembil Research Institute, University Health Network, Toronto, Ontario, CanadaDepartments of Medical Biophysics and Computer Science, University of Toronto, Toronto, ON, CanadaInstitute of Neuroimmunology, Slovak Academy of Sciences, Bratislava, Slovakia
Address correspondence and reprint requests to: M. Kapoor, Schroeder Arthritis Institute, University Health Network, Department of Surgery and Department of Laboratory Medicine and Pathobiology, University of Toronto; Krembil Research Institute, 60 Leonard Avenue, Toronto, Ontario, M5T 2S8, Canada. Tel.: 1-416-603-5800 ext. 4796.
Schroeder Arthritis Institute, Krembil Research Institute, University Health Network, Toronto, Ontario, CanadaDepartment of Surgery, University of Toronto, Ontario, CanadaDepartment of Laboratory Medicine and Pathobiology, University of Toronto, Ontario, Canada
To investigate the role of zinc finger protein 440 (ZNF440) in the pathophysiology of cartilage degeneration during facet joint (FJ) and knee osteoarthritis (OA).
Methods
Expression of ZNF440 in FJ and knee cartilage was determined by immunohistochemistry, quantitative (q)PCR, and Western blotting (WB). Human chondrocytes isolated from FJ and knee OA cartilage were cultured and transduced with ZNF440 or control plasmid, or transfected with ZNF440 or control small interfering RNA (siRNA), with/without interleukin (IL)-1β. Gene and protein levels of catabolic, anabolic and apoptosis markers were determined by qPCR or WB, respectively. In silico analyses were performed to determine compounds with potential to inhibit expression of ZNF440.
Results
ZNF440 expression was increased in both FJ and knee OA cartilage compared to control cartilage. In vitro, overexpression of ZNF440 significantly increased expression of MMP13 and PARP p85, and decreased expression of COL2A1. Knockdown of ZNF440 with siRNA partially reversed the catabolic and cell death phenotype of human knee and FJ OA chondrocytes stimulated with IL-1β. In silico analysis followed by validation assays identified scriptaid as a compound with potential to downregulate the expression of ZNF440. Validation experiments showed that scriptaid reduced the expression of ZNF440 in OA chondrocytes and concomitantly reduced the expression of MMP13 and PARP p85 in human knee OA chondrocytes overexpressing ZNF440.
Conclusions
The expression of ZNF440 is significantly increased in human FJ and knee OA cartilage and may regulate cartilage degenerative mechanisms. Furthermore, scriptaid reduces the expression of ZNF440 and inhibits its destructive effects in OA chondrocytes.
. Increased levels of inflammatory mediators, such as interleukin (IL)-1β, IL-6 and monocyte chemoattractant protein-1 (MCP-1), lead to increased expression of cartilage-degrading matrix metalloproteinases (MMPs) and reduced expression of components of the cartilage extracellular matrix (ECM), such as type II collagen (COL2A1), resulting in articular cartilage destruction
. ZNFs are divided into subclasses based on the amino acid sequence in the folded DNA-binding domain. Among subclasses, a two cysteine and two histidine residue (C2H2) unit is the most common and suitable for DNA binding. Evidence suggests that certain ZNF proteins may play key pathophysiological roles in a wide spectrum of diseases, including OA
Expression profiling and functional implications of a set of zinc finger proteins, ZNF423, ZNF470, ZNF521, and ZNF780B, in primary osteoarthritic articular chondrocytes.
ZNF440 is a C2H2-type protein found in primates including humans, chimpanzees, and monkeys. We previously showed that two microRNAs (miRNAs), namely miR-181a-5p and miR-4454, are increased in spine facet joint (FJ) cartilage obtained from patients with moderate to severe FJ OA compared to control FJ cartilage obtained from patients with no detectable or mild FJ OA
. We also found that ZNF440 expression is increased by IL-1β stimulation whereas reducing ZNF440 expression levels results in downregulation of phospho-NF-κB-p65 (p-p65). However, the overall contribution of ZNF440 to articular cartilage degeneration during OA pathobiology is largely unknown.
In this study, we determined the expression of ZNF440 and its contribution to degenerative mechanisms using both human FJ and knee OA cartilage. In silico drug transcription profile analysis identified scriptaid as a small molecule drug capable of downregulating ZNF440 expression in human chondrocytes and reducing ZNF440-induced markers of cartilage destruction and apoptosis. Overall, our results demonstrate that ZNF440 is involved in promoting cartilage degenerative mechanisms during OA.
Material and methods
Acquisition and grading of human FJ and knee cartilage
Human FJ and knee cartilage were obtained as previously described
. Patient characteristics of FJ and knee cartilage samples used are provided in Supplemental Table 1.
Histopathology and immunohistochemistry (IHC)
The degree of cartilage degeneration and the expression of ZNF440 in FJ and knee cartilage were determined by histological and IHC analysis, respectively, as previously described
. All primers were designed using Primer3Plus (http://primer3plus.com; Supplemental Table 3). RNA expression was normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH).
Chondrocyte culture and treatments with siRNA or compounds
Chondrocytes were extracted from FJ and knee OA cartilage as previously described
. To test the effect of ZNF440 knockdown, chondrocytes (1 х 104 cells/cm2) were treated with or without recombinant human IL-1β (10 ng/ml) for 18 h (to mimic an OA-like phenotype) followed by transfection with ZNF440 or control siRNA (50 nM) for 48 h. For compound validation, knee OA chondrocytes were treated with scriptaid or H7-dihydrochloride (0–10,000 nM) for 72 h following treatment with vehicle or IL-1β. Total RNA and proteins were isolated for further analysis. Details of siRNA, reagents and compounds used are provided in Supplemental Table 2.
ZNF440 plasmid transduction in human chondrocytes
Extracted and cultured human knee OA chondrocytes were transduced with lentivirus carrying either control-GFP (CTL-GFP) or ZNF440-GFP plasmid (Applied Biological Materials: ABM) as per manufacturer's instructions (vector details are available in Supplemental Table 4). Briefly, resuspended human knee OA chondrocytes (1 × 105 cells/ml) were transduced with lentivirus (MOI 2.5) and cells were transferred into 6 well plates and incubated for 24 h. Culture media was replaced after 3 days with 2 ml complete media containing puromycin and cells were grown to confluence. GFP signals were visualized using the EVOS FL Imaging System (Life Technologies).
. Briefly, protein-transferred membranes were incubated with primary antibodies for 1.5 h followed by overnight incubation at 4°C with secondary antibodies in 5% skimmed milk-TBS. Protein bands were visualized with enhanced chemiluminescent substrate (SuperSignal West Pico, Thermo Science) and signal intensity was quantified using Image J. Details of antibodies used are shown in Supplemental Table 2.
In silico drug transcription profile analysis
Compounds that could potentially down-regulate ZNF440 expression were identified through a literature search and analysis of gene expression data from Connectivity Map 02 (CMAP)
. CMAP data was processed to determine the fold-change in ZNF440 expression due to individual compounds in CMAP: a mean fold-change was calculated across all instances of a compound. Compounds identified to reduce mean ZNF440 expression at least 1.5-fold were selected, which included H7 and scriptaid. Importantly, scriptaid was predicted to substantially reduce ZNF440 expression, while modifying fewer additional genes.
Flow cytometry (FCM)
FCM was used to assess GFP positive cells after ZNF440-GFP transduction or to assess apoptosis and cell death markers (Annexin V/7-AAD). Data were acquired on a FACSCanto II (BD) and analyzed with FlowJo.
Statistical analysis
Data are presented as scatter dot plots with error bars (mean ± standard deviation). All data was log-transformed prior to analysis. Normality was assessed per group by D'Agostino & Pearson test (n ≥ 8 per group) or Shapiro–Wilk test (n < 8 per group). All experimental sample sets had displayed no significant deviation from Gaussian distribution (P > 0.05), except for 4 groups across three experiments. However, normality quantile-quantile plots were visually inspected to determine deviation from predicted normal distributions across all datapoints in each experiment and confirmed by inspecting residual quantile-quantile (QQ) plots. QQ plots did not show an observable deviation from a normal distribution in any experimental set. Thus, parametric tests were used. For comparison of two groups, unpaired t-tests with Welch's correction or paired t-tests were used. For comparisons of three or more treatments, log-transformed data was modelled using one- or two-way analysis of variance (ANOVA), as appropriate, with the Geisser-Greenhouse correction to correct for differences in variance, while accounting for matched samples. To correct for multiple comparisons, the two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli post-hoc tests were used to control the false discovery rate (FDR) and assess differences between treatment groups. For all comparisons, two-tailed tests were used and individual P or FDR-corrected P (q) < 0.05 was considered statistically significant.
Study approval
The use of human FJ and knee cartilage was approved by the Institutional Research Ethics Board (REB), University Health Network, Toronto (approval#: 14-8174 and 14-7592-AE) and University of Calgary REB, Calgary (approval#: 15-0005).
Results
ZNF440 expression in FJ and knee OA cartilage
We first observed that the number of cells expressing ZNF440, as assessed by IHC, was significantly increased in FJ OA cartilage (MRI grade 2–3
; moderately to severely degenerated FJ cartilage) compared to control FJ cartilage (MRI grade 0–1; non-discernible or mildly degenerated FJ cartilage) [Fig. 1(A)–(F)]. WB analysis also showed increased ZNF440 expression in FJ OA cartilage compared to control FJ cartilage [Fig. 1(G)]. Similar to FJ OA cartilage, we observed a significant increase in ZNF440 expression in knee OA cartilage compared to control (non-degenerated and microscopically undamaged knee cartilage), as assessed by IHC, WB, and qPCR [Fig. 1(H)–(O)].
Fig. 1Increased expression of ZNF440 in facet joint (FJ) and knee OA cartilage enhances expression of cartilage catabolic and apoptosis markers. (A, B) Histological analysis using Safranin O/fast green staining showing FJ cartilage with non-discernible degeneration (L5 superior cartilage; grade 0; control) and severe degeneration (L5 superior cartilage; grade 3; FJ osteoarthritis [OA]). (C–F) Immunohistochemistry (IHC) shows a higher ratio of cells positive for ZNF440 in FJ OA cartilage compared to that of control FJ cartilage (n = 4 patients/group). Scale bars = 100 μm. Statistical analysis comparing two groups with parametric data was performed by unpaired t-test with Welch's correction. ∗P < 0.05. (G) Western blot analysis shows increased protein expression of ZNF440 in FJ OA cartilage compared to control FJ cartilage. (H, I). Histological analysis using Safranin O/fast green staining showing knee cartilage with non-discernible degeneration (control) and severe degeneration (knee OA). (J-M) IHC shows a higher ratio of cells positive for ZNF440 in knee OA cartilage compared to control knee cartilage (n = 4 patients/group). Statistical analysis comparing two groups with parametric data was performed by unpaired t-test with Welch's correction. ∗∗P < 0.01. Western blot analysis (N) and qPCR analysis (n = 5 patients/group; O) show increased gene and protein expression of ZNF440 in knee OA cartilage compared to control knee cartilage. Statistical analysis comparing two groups with parametric data was performed by unpaired t-test with Welch's correction. ∗∗P < 0.01. (P–W) Microscopic images [P–S, bright field; T-W, immunofluorescence (IF); scale bar, 100 μm] of human knee chondrocytes transduced with PBS (phosphate-buffered saline; non-transduced control), control-blank (CTL) lentivirus, control-GFP (CTL-GFP) plasmid lentivirus, or ZNF440-GFP plasmid lentivirus. (X, Y) Expression of ZNF440 in human OA chondrocytes transduced with ZNF440-GFP compared to controls (n = 4/group), as assessed by flow cytometry (X), or qPCR (Y). Statistical analysis was performed by one-way analysis of variance (ANOVA) with Geisser-Greenhouse correction, followed by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli post-hoc tests. ∗∗q < 0.01. (Z-BB) Increased protein expression of ZNF440, MMP13, and poly (ADP-ribose) polymerase (PARP) p85 in human OA chondrocytes transduced with CTL-GFP or ZNF440-GFP, assessed by Western blotting. Statistical analyses comparing two groups with parametric data were performed by unpaired t-test with Welch's correction. ∗P < 0.05, ∗∗P < 0.01. (CC) Gene expression of IL6, MCP1, MMP13 and COL2A1 in human knee OA chondrocytes transduced with CTL-GFP or ZNF440-GFP, with or without IL-1β (n = 4/group). Statistical analyses were performed by two-way ANOVA with Geisser-Greenhouse correction, followed by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli post-hoc tests. ∗q < 0.05, ∗∗q < 0.01, CTL-GFP without IL-1β treatment vs CTL-GFP with IL-1β treatment, or CTL-GFP with IL-1β treatment vs ZNF440-GFP with IL-1β treatment. #q < 0.05, ##q < 0.01, CTL-GFP without IL-1β treatment vs ZNF440-GFP without IL-1β treatment. †q < 0.05, ††q < 0.01, ZNF440-GFP without IL-1β treatment vs ZNF440-GFP with IL-1β treatment. (DD, EE) Protein expression of MMP13 assessed by Western blotting (n = 4/group). Statistical analyses were performed by two-way ANOVA with Geisser-Greenhouse correction, followed by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli post-hoc tests. ∗∗q < 0.01, CTL-GFP without IL-1β treatment vs CTL-GFP with IL-1β treatment, or CTL-GFP with IL-1β treatment vs ZNF440-GFP with IL-1β treatment. ##q < 0.01, CTL-GFP without IL-1β treatment vs ZNF440-GFP without IL-1β treatment. ††q < 0.01, ZNF440-GFP without IL-1β treatment vs ZNF440-GFP with IL-1β treatment. Data presented as individual dot plots with standard deviation.
Effect of lentivirus-mediated overexpression of ZNF440 on the expression of catabolic, anabolic, inflammatory, and apoptosis markers
We next determined the role of ZNF440 in cartilage degeneration during OA. ZNF440 overexpression using lentivirus-mediated transduction was performed in human knee OA chondrocytes isolated from knee OA patients (KL grade 3 or 4) undergoing total knee replacement (TKR). Chondrocytes were cultured with phosphate-buffered saline (PBS), control-blank lentivirus, CTL-GFP-plasmid lentivirus or ZNF440-GFP-plasmid lentivirus and assessed by immunofluorescence [IF; Fig. 1(P)–(W)]. Quantification of cellular transduction with ZNF440-GFP in human knee OA chondrocytes was determined by FCM, qPCR and WB [Fig. 1(X)–(Z)].
We next evaluated the effect of ZNF440 overexpression on inflammatory, cartilage catabolic and cartilage anabolic markers in the presence/absence of IL-1β stimulation. While no significant change in the expression of IL6 and MCP1 was observed, a significant increase in MMP13 expression (assessed by qPCR and WB), and a significant or moderate (q = 0.0591) decrease in COL2A1 expression was found in ZNF440-GFP-transduced chondrocytes with or without IL-1β stimulation, as compared to those transduced with CTL-GFP [Fig. 1(CC)–(EE)].
Since chondrocyte apoptosis is closely associated with cartilage degeneration
, we also evaluated the effect of ZNF440 overexpression on poly (ADP-ribose) polymerase (PARP) p85, a marker of apoptosis. We observed that protein expression of PARP p85 was significantly increased in ZNF440-GFP-transduced human OA knee chondrocytes compared to CTL-GFP-transduced chondrocytes [Fig. (1Z), (BB)]. In addition, we observed that after 24 h in culture, the total number of chondrocytes transduced with ZNF440-GFP was significantly reduced compared to CTL-GFP-transduced cells [Supplemental Fig. 1(A)]. To assess cell death through apoptosis, we tested the chondrocytes for double-positive expression of AnnexinV and 7-AAD post ZNF440-GFP or CTL-GFP transduction by fluorescence-activated cell sorting (FACS) analysis. Chondrocyte cultures transduced with ZNF440-GFP had a greater percentage of cells double-positive for AnnexinV/7-AAD compared to those transduced with CTL-GFP [Supplemental Fig. 1(B) and (C)].
Effect of ZNF440 knockdown on the expression of catabolic, anabolic, inflammatory, and apoptosis markers
We next investigated the effect of ZNF440 knockdown on cartilage degeneration during FJ and knee OA. FJ OA chondrocytes and knee OA chondrocytes were treated with or without IL-1β in the presence of ZNF440- or control-siRNA. We first confirmed that ZNF440 expression was significantly reduced by transfection of chondrocytes with ZNF440-siRNA [Fig. 2(A), (G)]. We next observed that IL-1β-induced increases in MMP13, IL6, and MCP1 RNA levels, and PARP p85 protein levels were partially attenuated in both FJ OA and knee OA chondrocytes transfected with ZNF440-siRNA compared to those transfected with control-siRNA [Fig. 2(B)–(F) and (H)–(L)]. In contrast, ZNF440-siRNA treatment did not rescue IL-1β-induced decreases in COL2A1 expression (Supplemental Fig. 2). Of note, protein expression of PARP p85 in FJ chondrocytes transfected with control-siRNA and treated with IL-1β were only moderately increased compared to cells transfected with ZNF440-siRNA in the absence of IL-1β (Fig. 2(F), q = 0.0654).
Fig. 2Effect of ZNF440 inhibition and identification of scriptaid as a compound that decreases the expression of ZNF440 in human OA chondrocytes. (A) Expression of ZNF440 in facet joint (FJ) OA (n = 6/group) chondrocytes transfected with ZNF440 small interfering RNA (siRNA) or control siRNA. (B–E) qPCR analysis of the expression of cartilage catabolic (MMP13) and inflammatory (IL6 and MCP1) markers in FJ OA chondrocytes (n = 8/group), treated with (+) or without (−) IL-1β, and control-siRNA or ZNF440-siRNA. (E, F) Poly (ADP-ribose) polymerase (PARP) p85 expression relative to β-actin expression by Western blotting in FJ OA chondrocytes (n = 4/group). (G) Expression of ZNF440 in knee OA (n = 6/group) chondrocytes transfected with ZNF440-siRNA or control-siRNA. (H-J) qPCR analysis of the expression of cartilage catabolic (MMP13) and inflammatory (IL6 and MCP1) markers in knee OA chondrocytes (n = 8/group), treated with (+) or without (−) IL-1β, and control-siRNA or ZNF440-siRNA. (K, L) PARP p85 expression relative to β-actin expression by Western blotting in knee OA chondrocytes (n = 4/group). (A, G) Statistical analyses comparing two groups with parametric paired data were performed by paired t-test. ∗P < 0.05, ∗∗P < 0.01. (B–F, H-L) Differences in the levels of RNA or protein expression between control- and ZNF440 siRNA-treated cultures, with or without IL-1β treatment, were determined by two-way analysis of variance (ANOVA) with Geisser-Greenhouse correction, followed by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli post-hoc tests. ∗q < 0.05, ∗∗q < 0.01, control-siRNA without IL-1β treatment vs control-siRNA with IL-1β treatment, or control-siRNA with IL-1β treatment vs ZNF440-siRNA with IL-1β treatment. #q < 0.05, ##q < 0.01, control-siRNA without IL-1β treatment vs ZNF440-siRNA without IL-1β treatment. †q < 0.05, ††q < 0.01, ZNF440-siRNA without IL-1β treatment vs ZNF440-siRNA with IL-1β treatment. (M) Schematic flow showing the identification of potential compounds that regulate the expression of ZNF440 via in silico compound screening and validation. (N) Expression of ZNF440 in knee OA chondrocytes in response to scriptaid (Scr 0, 10, 100, 1000 nM; n = 7/group). Statistical analysis was performed by one-way analysis of variance (ANOVA) with Geisser-Greenhouse correction, followed by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli post-hoc tests. ∗q < 0.05, ∗∗q < 0.01. (O–Q) qPCR analysis of the expression of MMP13, IL6 and MCP1 in knee OA chondrocytes (n = 8/group), treated with (+) or without (−) IL-1β, and dimethyl sulfoxide [DMSO (Scr 0 nM)] or Scr (1000 nM). (R, S) PARP p85 expression relative to β-actin expression by Western blotting in knee OA chondrocytes treated with DMSO (Scr 0 nM) or Scr (1,000 nM) in the presence or absence of IL-1β (n = 4/group). (N–S) Differences in the levels of expression between DMSO (Scr 0 nM) and Scr (1,000 nM), with or without IL-1β treatment, were determined by two-way ANOVA with Geisser-Greenhouse correction, followed by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli post-hoc tests.∗q < 0.05, ∗∗q < 0.01. DMSO (Scr 0 nM) without IL-1β treatment vs DMSO (Scr 0 nM) with IL-1β treatment, or DMSO (Scr 0 nM) with IL-1β treatment vs Scr (1000 nM) with IL-1β treatment. ##q < 0.01, DMSO (Scr 0 nM) without IL-1β treatment vs Scr (1000 nM) without IL-1β treatment. †q < 0.05, ††q < 0.01, Scr (1000 nM) without IL-1β treatment vs Scr (1000 nM) with IL-1β treatment. (T, U) Gene expression of MMP13 and COL2A1 in knee OA chondrocytes transduced with either control (CTL)-GFP or ZNF440-GFP plasmid and treated with Scr (1,000 nM) (n = 4/group). (V, W) PARP p85 expression relative to β-actin expression by Western blotting in knee OA chondrocytes transduced with either control (CTL)-GFP or ZNF440-GFP plamid and treated with DMSO (Scr 0 nM) or Scr (1,000 nM) (n = 4/group). (T-W) Differences in the levels of RNA and protein expression between DMSO (Scr 0 nM) and Scr (1,000 nM) with or without IL-1β treatment were determined by one-way ANOVA with Geisser-Greenhouse correction, followed by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli post-hoc tests. ∗q < 0.05, ∗∗q < 0.01. (X) Schematic image of the proposed mechanisms of ZNF440 in regulating the expression of cartilage catabolic and apoptosis markers.
Identification of compounds that inhibit ZNF440 expression in OA chondrocytes
Since we observed that ZNF440, at least partially regulates the expression of cartilage catabolic, inflammatory, and apoptosis markers in both FJ and knee OA chondrocytes, we next sought to identify compounds that may suppress ZNF440 expression in OA chondrocytes. Using in silico transcription profile analysis that tests all compounds from Connectivity Map (total of 1,309 compounds)
, we identified two candidate compounds, namely scriptaid and H7-dihydrochloride, which were predicted to downregulate the expression of ZNF440 [Fig. 2(M); scriptaid was predicted to achieve a larger fold-change in ZNF440 expression, and modified fewer other genes compared to H7-dihydrochloride]. We next treated knee OA chondrocytes with these two compounds to validate their effects on ZNF440 expression. Following pilot compound concentration testing (Supplemental Table 5), we used chondrocytes from independent donor samples to verify that treatment of cells with scriptaid (10–1,000 nM) significantly reduced ZNF440 expression compared to knee OA chondrocytes treated with vehicle control [dimethyl sulfoxide (DMSO); Fig. 2(N)]. Pilot testing of H7-dihydrochloride indicated that ZNF440 expression was not decreased, but rather increased, with compound treatment (Supplemental Table 5) and was not tested further.
Effect of scriptaid treatment on the expression of cartilage catabolic, inflammatory, and apoptosis markers in OA chondrocytes
Since scriptaid suppressed the expression of ZNF440 in knee OA chondrocytes; we next explored the effect of scriptaid on the expression of aforementioned markers in knee OA chondrocytes. qPCR analysis showed that treatment of cells with scriptaid (1,000 nM) partially suppressed IL-1β-induced increases in MMP13, IL6 and MCP1 RNA [Fig. 2(O)–(Q)]. Furthermore, the protein level of PARP p85 was significantly decreased in knee OA chondrocytes treated with scriptaid compared to non-treated chondrocytes, as determined by WB [Fig. 2(R), (S)].
Although we observed a partial suppression of IL-1β-induced increases in markers of cartilage catabolic, inflammatory and apoptosis markers in response to scriptaid, it was unclear whether cellular outcomes related to scriptaid treatment were a direct result of reducing ZNF440 expression. To test this, we isolated human OA knee chondrocytes and transduced the cells to express ZNF440-GFP or CTL-GFP, under treatment with/without scriptaid. Our results showed that scriptaid significantly suppressed ZNF440-induced increases in the expression of MMP13 and PARP p85 [Fig. 2(T), (V), (W)], suggesting that the function of scriptaid is linked to ZNF440-mediated catabolic and apoptosis-inducing activities. However, there was no such effect observed in the expression of COL2A1 [Fig. 2(U)], consistent with our observation that ZNF440-siRNA did not rescue IL-1β-induced decreases in COL2A1 expression.
Discussion
In this study, we demonstrated that ZNF440 expression was increased in both FJ OA and knee OA cartilage compared to control (non-degenerated) cartilage, and, for the first time, revealed a potential contribution of ZNF440 to cartilage degeneration during OA. Although no previous studies have reported a role for ZNF440 in cartilage degeneration, other types of ZNF proteins have been shown to play an important role in cartilage degeneration and chondrogenesis
Zinc-finger protein 145, acting as an upstream regulator of SOX9, improves the differentiation potential of human mesenchymal stem cells for cartilage regeneration and repair.
Given that ZNF440 may be a regulator of catabolic processes associated with OA, we took advantage of in silico compound screening followed by validation by qPCR, and identified scriptaid, which reduced ZNF440 expression in human knee OA chondrocytes. Scriptaid is a broad-spectrum histone deacetylase inhibitor (HDACi), chiefly inhibiting class II (a) HDACs
. Our in vitro study determined that treatment with scriptaid significantly reduced the expression of catabolic and apoptosis makers in ZNF440-overexpressing human knee OA chondrocytes. Consistent with our results, Trichostatin A, another broad-spectrum HDACi, protects cartilage from degeneration in trauma-induced OA in mice, in part, by reducing MMP13 expression
. Although this compound may also target other genes, it is possible that it alleviates the expression of catabolic and apoptosis markers, at least partially, through downregulation of ZNF440 expression in OA chondrocytes. We previously identified that ZNF440 regulates, at least partially, p-p65 in human FJ chondrocytes
. Thus ZNF440 may regulate apoptosis and inflammatory marker expression through modification of p-p65 expression. In line with this, Ji et al. show that IL-1β-induced upregulation in the expression of apoptosis and inflammatory markers are attenuated by reducing levels of p-p65, albeit in the mouse chondrocyte cell line ATDC5
Isoliquiritigenin suppresses IL-1β induced apoptosis and inflammation in chondrocyte-like ATDC5 cells by inhibiting NF-κB and exerts chondroprotective effects on a mouse model of anterior cruciate ligament transection.
. However, the exact roles ZNF440 plays in promoting inflammation and apoptosis needs to be further elucidated in future studies.
There are limitations to our current study. First, our FACS analysis of cell apoptosis did not include a control for GFP-transduced chondrocytes alone. Thus increased expression of cell death markers in ZNF440-GFP-transduced chondrocytes compared to those transduced with CTL-GFP could be due to differences in the transduction efficacy between ZNF440-GFP and CTL-GFP. Furthermore, determining the contribution of ZNF440 to OA pathogenesis in vivo is an important next step. However, ZNF440 is only identified in humans, chimpanzees and monkeys, limiting investigations in small animal models of OA, or the use of genetic deletion or silencing techniques of ZNF440 in vivo.
In summary, this is the first study to show that ZNF440 is increased in both FJ OA and knee OA cartilage compared to control cartilage, and regulates the expression of inflammatory, catabolic, and apoptosis markers in OA chondrocytes [Fig. 2(X)]. These finding suggest that ZNF440 contributes to the destructive mechanisms associated with cartilage degeneration in OA.
Author contributions
All authors were either involved in conception and design, or analysis and interpretation of data. Each author was involved in drafting the article and revising it critically for important intellectual content.
Conflicts of interest
A US provisional patent application (62/299,305, filed on 24 February 2016) and a PCT international patent application (PCT/CA2017/000019, filed 31 January 2017) have been filed with respect to ZNF440 as a potential therapeutic target of OA.
Funding
This study is supported by grants to MK by the Krembil Foundation, the Canadian Institute of Health Research (CIHR; #156299) and the Arthritis Program, University Health Network. MK is the recipient of a Tier 1 CRC from the Canada Research Chairs Program. IJ is, in part, supported by the Canada Research Chairs Program (CRC #203373 and #225404), Canada Foundation for Innovation (CFI #12301, #203373, #29272, #225404, #30865) and IBM. AN is a recipient of a CIHR fellowship and Krembil postdoctoral salary award.
Acknowledgement
Authors would like to thank Kim Perry, Amanda Wetson, Sarah Gabrial, Erdeta Prifti, Luis Montoya and Daniel Antflek for their help with collecting FJ cartilage and patient data. Authors would also like to acknowledge the help of members of the Schroeder Arthritis Institute, University Health Network (Toronto).
Appendix A. Supplementary data
The following is the Supplementary data to this article:
Expression profiling and functional implications of a set of zinc finger proteins, ZNF423, ZNF470, ZNF521, and ZNF780B, in primary osteoarthritic articular chondrocytes.
Zinc-finger protein 145, acting as an upstream regulator of SOX9, improves the differentiation potential of human mesenchymal stem cells for cartilage regeneration and repair.
Isoliquiritigenin suppresses IL-1β induced apoptosis and inflammation in chondrocyte-like ATDC5 cells by inhibiting NF-κB and exerts chondroprotective effects on a mouse model of anterior cruciate ligament transection.
31220-6&id=gr1.jpg){kind=link}
31220-6&id=gr2.jpg){kind=link}