Mechanism Of Pro-inflammatory Cytokines Induced Expression Of ADAMTS-4: Implications In Intervertebral Disc Degeneration

Part I: Mechanism of TNF-and IL-1Regulation of ADAMTS-4in NPcellsBackground: The intervertebral disc (IVD) is a unique tissue that that permitsrotation, as well as flexion and extension of the spine. It consists of a gel-like nucleuspulposus (NP) surrounded circumferentially by a fibrocartilagenous annulus fibrosus. Cellsof the NP are derived from the notochord, an embryonic tissue with limited blood supply.In common with chondrocytes, NP cells secrete a complex extracellular matrix thatcontains fibrillar collagens and the proteoglycan aggrecan. Assembly of thesemacromolecules provides a robust hydrodynamic system that accommodates appliedbiomechanical forces to the spine. It has been reported that during disc degeneration andherniation, in addition to infiltrating immune cells, resident NP and annulus fibrosus cellsproduce high levels of the cytokines TNF-and IL-1. These cytokines stimulateproduction of NGF, BDNF and VEGF, molecules associated with nerve ingrowth andangiogenesis by NP cells. Moreover, both the cytokines upregulate expression of catabolicmatrix metalloproteinases (MMPs) and two major aggrecanases (ADAMTS-4andADAMTS-5) by NP cells. Among several members of the ADAMTS family that possessability to cleave aggrecan in vitro, ADAMTS-4(aggrecanase1) and ADAMTS-5(aggrecanase2) are the most likely to play a role in aggrecan degradation and subsequentdisc degeneration as in the pathogenesis of osteoarthritis. despite the importance ofADAMTS-4in pathogenesis of osteoarthritis and disc disease, The expression andregulation of ADAMTS-4transcription in NP cells is still not clear.Objective: The major objective of the investigation was to examine thatpro-inflammatory cytokines regulate expression of A Disintegrin and Metalloproteinasewith Thrombospondin motif-4(ADAMTS-4) by nucleus pulposus cells and to elucidate themechanism of this regulation.Methods: Expression of ADAMTS-4in mature rat tissues was studied using real-timePCR and Western blot analysis. To explore the premise that cytokines concerned with discdegeneration regulate ADAMTS-4expression, NP cells were treated with TNF-and IL-1and expression of ADAMTS-4analyzed. In addition, we measured the level of ADAMTS-4protein in conditioned medium of treated NP cells by Western blot analysis. To investigate if the regulation of expression is at the transcriptional level, we measured the activity of a3.5kb ADAMTS-4promoter following cytokine treatment. To determine if MAPK and/orNF-B signaling is required for the cytokine dependent induction of ADAMTS-4in NP cells,we first evaluated activation of these signaling pathways following treatment with TNF-and IL-1. We also examined levels of phosphorylated MAPK isoforms p38, ERK1/2andJNK. To ascertain if the cytokine induced ADAMTS-4and-5expressions requires NF-Band/or MAPK signaling, NP cells were pretreated with specific pathway inhibitors. Toinvestigate the mechanism of MAPK regulation of ADAMTS-4expression, we transfectedNP cells with dominant-negative DN-P38, p382, p38or DN-ERK1or DN-ERK2expression plasmids and measured ADAMTS-4promoter activity. Since CCAATenhancer-binding protein beta (C/EBP-) or LAP2has been shown before to control IL-1and TNF-dependent transcription in chondrocytes, we examined if a similar regulatorysystem exists in cells of the NP. To investigate the role of NF-B in the transcriptionalregulation of ADAMTS-4, we first used the JASPAR database analysis for evidence ofputative NF-B binding motifs in promoter. To confirm that ADAMTS-4promoter activityis responsive to NF-B signaling, we performed loss of function studies NP cells are treatedwith the NF-B inhibitor SM7368or co-transfected with DN-NF-B/I B M. To furthervalidate the role of p65/RelA and to determine if there is cell type specificity, we measuredADAMTS-4promoter activity in RelA null and wild type fibroblasts. We then examinedthe effect of overexpression of NF-B subunits on ADAMTS-4promoter activity.Results: The basal expression of ADAMTS-4mRNA in healthy NP as well as annulusfibrosus tissue is very low. Our result shows that NP tissue shows weak ADAMTS-4bandsat approximately58and73kDa. To explore the premise that cytokines concerned with discdegeneration regulate ADAMTS-4expression, NP cells were treated with TNF-and IL-1and expression of ADAMTS-4analyzed. Our result indicates that treatment with bothTNF-and IL-1results in dose-dependent increase in the levels of ADAMTS-4mRNAlevels. In addition, we measured the level of ADAMTS-4protein in conditioned medium oftreated NP cells by Western blot analysis. Moreover, cytokine treatment significantlyincreases ADAMTS-4protein expression in both rat and human NP cells. To investigate ifthe regulation of expression is at the transcriptional level, we measured the activity of a3.5kb ADAMTS-4promoter following cytokine treatment. The result shows that both thecytokines significantly increase the promoter activity. Following treatment with TNF-andIL-1treatment, there is a rapid increase in phosphorylated p65protein levels. Activation is maximal at5to30min and then declines rapidly. As expected, there is no appreciablechange in the level of total p65during the treatment period. We also examined levels ofphosphorylated MAPK isoforms p38, ERK1/2and JNK. Again, there is a rapid increase inp-p38, pERK1/2and pJNK; among the MAPK isoforms, ERK evidences more sustainedlevel of phosphorylation. To ascertain if the cytokine induced ADAMTS-4and-5expressions requires NF-B and/or MAPK signaling, NP cells were pretreated with specificpathway inhibitors. Pretreatment causes a significant suppression in TNF-and IL-1induction of both ADAMTS-4and-5mRNA levels. Similarly, a pronounced decrease incytokine mediated increase in levels of ADAMTS-4protein is seen in the presence ofMAPK and NF-B pathway inhibitors. DN-p38and DN-ERK2did not suppress promoteractivity. In contrast, cytokine dependent induction in ADAMTS-4promoter activity issignificantly suppressed by co-transfection with DN-P38p382, p38and DN-ERK1.We transfected NP cells with liver-enriched inhibitory protein (LIP), a functional LAPantagonist and measured cytokine dependent ADAMTS-4promoter activity. Surprisingly,suppression of C/EBP-function results in further induction of the promoter activity byTNF-. On the other hand, co-transfection with LAP2results in suppression of the basalpromoter activity. We used the JASPAR database analysis for evidence of putative NF-Bbinding motifs in promoter. Sequence analysis revealed four putative binding sites; thesequence, their location in the promoter relative to transcription start site, and relativescore is as follows: GGCAAGTCCC;-159/-150bp;0.90, GGGGATTCTC;-1042/-1033bp;0.88, GGGATTCTCC;-1043/-1034bp;0.88, GGGGATTTCC;-1394/-1385bp;0.98.We then examined the effect of overexpression of NF-B subunits on ADAMTS-4promoter activity. Co-transfection with p65results in a dose-dependent increase inADAMTS-4promoter activity. On the other hand neither RelB nor c-Rel influencedADAMTS-4promoter activity. While, p50alone has no effect on ADAMTS-4promoteractivity, even at a low dose, it blocked the inductive effect of p65. Noteworthy, p50completely suppresses inductive effect of both the cytokines on ADAMTS-4promoter. Toconfirm that ADAMTS-4promoter activity is responsive to NF-B signaling, weperformed loss of function studies. When cells are treated with the NF-B inhibitorSM7368or co-transfected with DN-NF-B/I B M, cytokine mediated induction inADAMTS-4promoter activity is completely abolished. Specificity of NF-B inhibitor,SM7368and I B M is validated by measuring the activity of a well characterized NF-Bresponsive reporter. To further validate the role of p65/RelA and to determine if there is cell type specificity, we measured ADAMTS-4promoter activity in RelA null and wildtype fibroblasts. Only in wild type cells is the promoter activity cytokine inducible.Conclusions:1.ADAMTS-4is expressed by tissue of the nucleus pulposus andannulus fibrosus, the basal expression of ADAMTS-4mRNA in healthy NP as well asannulus fibrosus tissue is very low.2. TNF-and IL-1treatment results in elevatedADAMTS-4expression in a dose dependent manner.3. Both the cytokines significantlyincrease the promoter activity.4. By controlling the activation of MAPK and NF-Bsignaling, TNF-and IL-1modulate the expression of ADAMTS-4in NP cells. MAPKregulation of ADAMTS-4promoter activity clearly showed isoform specificity. As forNF-B signaling pathway, the inductive effect is restricted to p65; while p50suppressed theinductive effect of p65on ADAMTS-4promoter.Part II: The role of ADAMTS-4in aggrecan degradationBackground: Intervertebral disc degeneration is characterized by increasedexpression of catabolic enzymes, decreased proteoglycan synthesis, and an overall shifttowards synthesis of a fibrotic matrix. When this occurs, the water-binding capacity of thetissue is compromised resulting in a failure to resist compressive forces and a reduction indisc height. Although a great deal is known about importance of proteoglycan secretionand function, molecular mechanisms controlling aggrecan turnover in cells of both thenormal and degenerated disc, are not well understood. It has been reported that during discdegeneration and herniation, in addition to infiltrating immune cells, resident NP andannulus fibrosus cells produce high levels of the cytokines TNF-and IL-1. Thesecytokines stimulate production of NGF, BDNF and VEGF, molecules associated withnerve ingrowth and angiogenesis by NP cells. Moreover, both the cytokines upregulateexpression of catabolic matrix metalloproteinases (MMPs) and two major aggrecanases(ADAMTS-4and ADAMTS-5) by NP cells. ADAMTS-4and-5produce fragments ofaggrecan usually found in synovial fluid and cartilage by cleaving the protein followingGlu373, Glu1545, Glu1714, Glu1819, and Glu1919. Unlike cartilage, both ADAMTS-4and ADAMTS-5expression is elevated in the NP in human degenerative disc disease.Surprisingly, despite the importance of these aggrecanases in pathogenesis of osteoarthritisand disc disease, a few studies have investigated regulation of ADAMTS transcription in NP cells. A clue to the mechanism was the recent finding that in NP cells, TNF-andIL-1regulate ADAMTS-4and ADAMTS-5mRNA expression, and modulateADAMTS-5enzymatic activity through syndecan-4. This observation begs the question:how does TNF-and IL-1control the expression of ADAMTS-4and what is the relativecontribution of ADAMTS-4and ADAMTS-5in NP cells in terms of aggrecan degradation.Objective: The major objective of the investigation was to examine the role ofADAMTS-4in aggrecan degradation, to elucidate the significance of ADAMTS-4in discdegeneration and might provide a therapeutic target to mitigate degenerative disc disease.Methods: We silenced the expression of individual NF-B signaling components(p65, p52, Ikk and Ikk) and measured the expression of ADAMTS-4expression inhuman NP cells. To determine if silencing of individual NF-B signaling componentspromoted ADAMTS activity, we also measured generation of aggrecan fragments usingneoepitope recognizing anti-ARGSVIL antibodies. We then examined the effect ofsilencing of ADAMTS-4and ADAMTS-5expression on aggrecan degradation in human NPcells. We then used neo-epitope specific antibodies anti-ARGSVIL and anti-NITEGE tomeasure aggrecan degradation in ADAMTS silenced cells. The level of cellassociated/pericellular matrix aggrecan degradation was also evaluated.Results：We silenced the expression of individual NF-B signaling componentsand measured the expression of ADAMTS-4expression in human NP cells. Our resultshows that there is a robust YFP expression by the virally transduced cells, indicating highlevel of transduction efficiency and transgene expression. In those cells transduced withShp65and Shp52there is a significant decrease in the protein levels of p65and p52respectively compared to cells transduced with control shRNA. Similarly, transduction ofNP cells with ShIkk and ShIkk results in decreased levels of Ikk and Ikk proteinrespectively. Importantly, suppression of individual NF-B signaling componentssignificantly blocked the inductive effect of TNF-on the expression of ADAMTS-4and-5protein levels as well as aggrecan degradation as measured by neoepeptide generation.We examined the effect of silencing of ADAMTS-4and ADAMTS-5expression onaggrecan degradation in human NP cells. The results show that ADAMTS protein levels inthe silenced NP cells or the condition medium is significantly reduced. Furthermore, thereis no compensatory increase in ADAMTS-5or-4when either ADAMTS-4or-5issilenced. We then used neo-epitope specific antibodies anti-ARGSVIL and anti-NITEGEto measure aggrecan degradation in ADAMTS silenced cells. Our result clearly shows that suppression of ADAMTS-4and-5expression results in a significant inhibition in TNF-mediated aggrecan degradation in NP cells. Densitometric analysis validates these findings.As expected, cells transduced with Control-ShRNA exhibit an increase in aggrecanneoepitope generation following TNF-treatment. The level of cell associated/pericellularmatrix aggrecan degradation was also evaluated. The result clearly shows that in cellassociated protein fraction, compared to control cells, silencing of ADAMTS-4andADAMTS-5blocks aggrecan degradation following TNF-treatment.Conclusions：1. Suppression of individual NF-B signaling componentssignificantly blocked the inductive effect of TNF-on the expression of ADAMTS-4and-5protein levels.2. Both ADAMTS-4and ADAMTS-5contribute to aggrecan degradationin human NP cells.