Evaluation of ADAMTS-9 Expression in Post-Mortem Brain Tissue

Background: Extracellular Matrix (ECM) modifications have been reported in the Central Nervous System (CNS) of multiple sclerosis (MS) within post mortem brain tissue due to both the increased synthesis of ECM proteoglycans, and release activation of matrix metalloproteinase (MMPs). Multiple sclerosis (MS) is chronic inflammatory demyelinating disease written off as inflammation and demyelination disease, mainly located in central nervous system (CNS) white matter (WM). This study aims to clarify the potential pathophysiologic role of adisintegrin And metalloproteinase with thromboSpondin motif-9 (ADAMTS-9) in MS. Materials and methods: Immunohistochemistry was performed to study the expression of ADAMTS-9 in both normal and MS white matter by confocal microscopic using specific antibodies. Fifty frozen blocks of brain tissue were obtained from the UK MS Society Tissue. All tissues blocks were marked by immunohistological material including antibodies to Human leukocyte antigen (HLA-DR) to assess the macrophage activation, Glial fibrillary acidic protein (GFAP) for astrocyte, neurofilaments (NF) for neurons and Von Willbrand factor (VWF) for endothelial. Results: In MS lesions, ADMTS-9 expression was increased in comparison to control samples. The expression of ADAMTS-9 was increased in active lesions as well as was associated with different cells from neuron, endothelial covering blood vessels astrocyte and microglia. Conclusion: The expression of ADAMTS-9 at the protein level was increased in active inflammatory lesions with evidence of myelin breakdown, suggesting that up-regulation of ADAMTS-9 may be a general phenomenon induced by CNS injuries.


Introduction
The ADAMTS enzyme family contains 20 individual gene products (Apte, 2004).Certain members of this family (ADAMTS-1, -4, -5, -8, -9 and 15) can proteolytically process aggrecan within the interglobulin domain (IGD) separating its globular G1 and G2 domains at a specific Glu 373 -Ala 374 bond or at one or more sites within the C-terminal glycosaminoglycan (GAG)bearing region (Kuno et al., 2000;Tortorella et al., 1999).ADAMTS-9 was originally cloned by Clark et al., (2000) which localized the gene to chromosome 3p14.2-p14.3(Clark et al., 2000).Of all ADAMTSs, ADAMTS-9 comprises the highest number of C-terminal TSP-1 like domains (Gottschall & Howell, 2015).This is indicative of a protein with the ability to bind to heparan sulphate proteoglycans (HSPGs) of the blood brain barrier (BBB) basement membrane and glial scars as well as glycosphingolipid sulphatide, a lipid predominantly expressed in the brain (Zeng et al., 2006).ADAMTS-9 is widely expressed throughout the adult human body including brain, heart, placenta and others (Somerville et al., 2003).Levels of pro-inflammatory cytokines TNF and IL-1 are up-regulated in the CNS in MS (Di Penta et al., 2013) and ADAMTS-9 has been shown to be increased in response to these in chondrocytes and chondrosarcoma cells (Demircan et al., 2013;Yaykasli et al., 2009).Though, anti-inflammatory mediators such as TGFβ1 downregulated ADAMTS-9 expression in prostate cells (Cross et al., 2005).
Unlike other ADAMTSs, such as ADAMTS-1 and ADAMTS-4, pro-domain cleavage of ADAMTS-9 leads to reduced versicanase activity rather than enhancing the catalytic function, and pro-domain processing occurs outside the cytoplasm on the cell surface (Kumar, Rao, & Ge, 2012).It has been suggested that ADAMTS-9 degradation of Chondroitin Sulphate Proteoglycans (CSPGs) may increase access of inflammatory cells to the CNS and lead to axonal damage.Alternatively, breakdown of CSPGs, by ADAMTS-9 may enable axonal regeneration and neurite outgrowth, which is inhibited by CSPGs (Tauchi et al., 2012).Several ADAMTSs have been shown to be elevated in human neurodegenerative disease and animal models of brain injury.ADAMTS-1, appears to be up-regulated in Down's syndrome, Pick's disease and Alzheimer's disease (Miguel, Pollak, & Lubec, 2005) and ADAMTS-4 in MS (Haddock et al., 2006b).
Enzymatic degradation of aggrecan and versican was observed in MS lesions, the aim of this study was to determine whether ADAMTS-9 is implicated in this ECM degradation; to develop an understanding of its activity by assessing its substrates ex vivo, which may then subsequently be exploited therapeutically through the design and application of molecules to increase or inhibit its activity.This study was designed to establish a reproducible protocol for both (1) the examination of ADAMTS-9 expression together with (2) phenotypic markers of specific CNS cells.

Immunohistochemistry
Serial cryostat sections of thirty MS active lesion (AL) and twenty normal brains were obtained from MS Tissue Bank in London under ethics number 08/MRE09/31.All specimens were cut and mounted onto polysine coated glass slides and fixed in ice-cold acetone for 10 mins.To determine the cellular origin of ADAMTS-9 expression, the dual-label immunofluorescence protocol was carried out, Briefly, sections were incubated with polyclonal ADAMTS-9 antibody overnight (1: 100) at -4°C and then detected by incubating with Alexa 488 or Alexa 568-conjugated donkey anti-goat immunoglobulins (1: 1000, Invitrogen) at room temperature for 90 mins.Following three 5 min washes in PBS, sections were incubated in the mouse monoclonal antibodies to human GFAP (1: 1000), HLA-DR (1:50), NF-L (1:100) or vWF (1; 500) for 2 h.Sections were washed with PBS and incubated with secondary antibody 488 or 568-conjugated rabbit anti-mouse.
Human brain was removed from the deceased donor, specimens were prepared by cutting a whole brain into anterior and posterior halves by a single cut through the mammillary bodies.Each block was one centimetre thick coronal slices.Cutting slices were numbered according to their position, as anterior (frontal pole) or posterior (occipital pole) to the mammillary bodies.They were numbered A1, A2 etc and the latter were numbered P1, P2 etc respectively.The coronal slices were laid on a gride, cut into 2cm brain tissue blocks stored at -80°C.Furthermore, each patient already given an ID number related to the coronal section.For instance, MS027 A2D3, the patient number is 027, and A2 meaning coronal slice was taken from the second section anterior to the mammillary bodies, and the another brain tissue block derived from gride co-ordinate D3.
MS cases include seventeen females, mean ages 54.52 years (ranges 39-77) and thirteen males with a mean age of 60.54 years (range 38-75), while control cases include eight males, mean ages of 67.8 (range 35-88) and twelve females, mean age of 68.3 (range 50-77).Tissue blocks were stored at -80°C until required.All MS cases had a confirmed diagnosis of secondary progressive SPMS.

Image Capture and Data Analysis
Images of immunofluorescence on section to assess expression of ADAMTS-9 were taken by using confocal microscope.Immunofluorescence image to determine co-localization of ADAMTS-9 were captured using a Zei 510 confocal scanning laser microscope equipment with krypton/ argon laser.
Fluophores were excited at wavelengths of either 488 or 568 nm.Individual pixels were scanned for each channel with set intensity threshold, which were consistent for all analysis data.Co-localization pixels were represented as yellow in the composite image.Statistical analysis and any significance of any increase or decrease of ADAMTS-9 levels was determine via using image J software.All statistical analysis of data was performed via kruskal-wallis test, where, *p >0.01 and *p<0.001.

Results
ADAMTS-9 expression appeared to be associated with the cell body and the elongated processes of parenchymal astrocytes in active MS lesion.There were also increased levels of blood vessel associated ADAMTS-9 expression compared to normal appearing white matter (NAWM) and normal control (NC) white matter, which appeared to be both endothelial and astrocytic-end-feet in origin.To determine the exact cellular distribution of ADAMTS-9, dual label immunofluorescence was carried out with polyclonal goat anti-ADAMTS-9 antibody and rabbit anti-NF or mouse anti-HLA-DR or mouse anti-vWF or mouse anti-GFAP antibodies for identification of neurons, microglia/macrophages, endothelial cells and astrocytes respectively.Immunofluorescence was carried out following the application of three different commercially

R E T R A C T E D
‫األسمرية‬ ‫الجامعة‬ ‫مجلة‬ : ‫والتطبيقية‬ ‫األساسية‬ ‫العلوم‬ Journal of Alasmarya University: Basic and Applied Sciences available anti-GFAP antibodies to determine the cellular location of ADAMTS-9 protein expression within astrocytes cells, and to ensure there were no cross reactions (anti-GFAP, Abcam, UK; anti-GFAP, Millipore, UK; anti-GFAP, Invitrogen, UK).Following initial optimisation, mouse anti-GFAP antibody from Millipore was carried forward for use in this experiment.
Utilisation of the Zeiss 510 software enabled individual pixels to be scanned and viewed as a yellow colour if true co-localization existed between the two channels of interest.Colocalisation was observed with ADAMTS-9 and cerebral vascular endothelium in the lesion (Figure 1).Co-localisation was also observed in the foamy macrophages and activated microglia in MS active lesion (Figure 2).Co-localisation of the Neuron phenotypic marker (NF) with ADAMTS-9 provided definitive evidence of ADAMTS-9 expression by astrocyte cells within MS active lesions as shown in figure 3.In addition, expression of ADAMTS 9 was associated with astrocyte cells confirmed with GFAP proteins as showed figure 4. All these results were compared using image J software analysis and showed that much greater expression of ADAMTS-9 was observed in active lesion compared to NC and MS NAWM as shown in figure 5.

Discussion
The results presented here validate the spectrum of ADAMTS-9 distribution in CNS tissue of MS by IHC, along with confocal microscope analysis of the relative level of ADAMTS-9 expression in normal control white matter, NAWM and the active lesions.Tissue was classified based on evidence of recent myelin breakdown and extent of cellular activation as determined by oil red O (ORO) staining and HLA-DR immunoreactivity, classification parameters as described by other authors (Sanders, Conrad, & Tourtellotte, 1993;Van der Valk & De Groot, 2000).The first objective of this study was to determine whether ADAMTS-9 was expressed and then to establish a reproducible protocol for identification of ADAMTS-9 by immunofluorescence in snap frozen human autopsy brain material.
High levels of ADAMTS-9 were presented in regions of demyelination, which has not previously been reported in MS by IHC.In contrast, much lower levels of ADAMTS-9 was consistently found in normal appearing white matter (NAWM) of MS and normal control samples tissue where there was no evidence of demyelination.Previous educations using qRT-PCR shown that ADAMTS-9 was expressed by differentiated neuroblastoma cells lines and significantly up-regulated following transient middle cerebral artery occlusion (tMCAo) in the rat (Reid et al., 2009).The ADAMTS-9 expression was associated with the endothelium of blood vessels and with the cell body and elongated processes of parenchymal astrocytes and the astrocytic end feet that encompass the cerebral vasculature.Co-localisation of ADAMTS-9 with the astrocyte phenotypic marker, GFAP, or the endothelial marker, vWF or neuronal marker, NF-L following dual labelled immunohistochemistry, confirmed the morphological observations and clearly demonstrated that these three cell types are responsible for ADAMTS-9 expression.These observations confirm ADAMTS-9 expression within MS active lesions similar to that for other matrix metalloproteinase informed previously e.g.ADAMTS-4 and ADAM 17 (Goddard, Bunning, & Woodroofe, 2001;Haddock et al., 2006b).ADAMTS-9 was also established to be expressed by activated macrophages/microglia cells in MS tissue and was up-regulated in MS. principally in active lesions where ADAMTS-9 immunreactivity is evident in numerous large foamy macrophages, implicating a possible role in MS pathology.
Changes in ADAMTS-9 reported in human chondrocytes in rheumatoid arthritis (RA) results from transcriptional changes in ADAMTS-9 or changes in protein activity due to upregulation of aggrecan within injury sites (Demircan et al., 2005).ADAMTS-9 may be involved in the loss of ECM by cleaving proteoglycans, as demonstrated for other ADAMTSs and other tissues (Lemarchant et al., 2013;Sandy et al., 2001) and as described here.Sobel and Ahmed, (2001) demonstrated that in active plaque centre and active lesions, CSPGs were decreased in the ECM and accumulated in foamy macrophage which is similar to our findings in a previous paper.This loss of CSPGs is related to the high levels of ADAMTS-9 observed in active MS plaques (Sobel & Ahmed, 2001).
Transcriptional induction of ADAMTSs e.g.ADAMTS-1 and 9 in neurons and astrocytes following physical or toxic injury in the CNS has been reported and it was suggested that this expression at the site of injury may favour neurite outgrowth (Cua et al., 2013;Lemons, Sandy, Anderson, & Howland, 2001).It has been reported that ADAMTS-4 cleaves versican in vivo, in a manner similar to the ADAMTS-mediated cleavage of aggrecan and versican in cartilage (Demircan et al., 2005;Tortorella, Malfait, Deccico, & Arner, 2001) , 2001;Rodrıguez-Manzaneque et al., 2002) and brain tissue (Matthews et al., 2000;Westling et al., 2004).It was also reported that ADAMTS-4 was abundant in MS active lesions (Haddock et al., 2006b), in agreement with our reported observations for ADAMTS-9.Thus, ADAMTS-9 together with ADAMTS-4 expression in the CNS, may contribute to remodelling of the ECM in the brain in MS patients.On the other hand, others also suggest that the level of endogenous ADAMTSs is not sufficient to recover plasticity and functional recovery after human brain injury.The cleavage of CSPGs induced by glial cells may be achieved not only by ADAMTS-9 but also by other ADAMTSs and MMPs, which are also up-regulated after neuronal injuries (Nakamura et al., 2000;Tauchi et al., 2012).MMPs also cleave other matrix proteins, such as laminin and collagen, leading to loss of blood brain barrier integrity; enhancing the infiltration of immune cells which may worsen the pathology of MS (Rosenberg, Estrada, & Dencoff, 1998).
ADAMTS-9 may play a role in a number of ways in the development of MS; proteolytic cleavages of ECM Heparan sulphate proteoglycan (HSPGs) of the BBB basement membrane and glial scar as well as cleavage of HGSPs into their soluble form, may create a chemotatic gradient for inflammatory cells to enter the CNS (Chen, Herndon, & Lawler, 2000;Zeng et al., 2006).ADAMTSs are reported to cleave CSPGs and adhesion molecules from the luminal surface of blood vessels; however in MS, increased expression of CSPGs has been reported, suggesting an imbalance in enzyme/inhibitor activity.ADAMTS-9 expression has been demonstrated to be associated with astrocytes and activated macrophages/microglia and expressed at high levels in lesions with high level of cellular activity as gauged by HLA-DR and ORO expression.ADAMTS-9 cleaves CSPGs thereby possibly subjecting neurons to its cytotoxic effects (Hoffmann et al., 2015).CSPGs have been reported to inhibit myelin repair (Lau, Cua, Keough, Haylock-Jacobs, & Yong, 2013) neurite outgrowth (Barritt et al., 2006) and axon regeneration (Brown et al., 2012;Shen et al., 2009).
Following up-regulation, ADAMTS-9 could be contributing to neuronal regeneration by two potential mechanisms; 1) promotion of neuronal differentiation/proliferation by enzymatic activity in the nucleus (Somerville et al., 2003) or 2) operating as a peptidase on the cell surface of migrating CNS cells and clearing a path through the ECM or glial scar by degradation of CSPGs (Somerville et al., 2003), enabling migration of oligodendrocyte precursor cells.Alternatively, the breakdown of CSPGs by ADAMTS-9 could have a damaging effect on the CNS by allowing the migration of inflammatory cells and activated macrophages to susceptible neurons.The efficiency by which ADAMTSs cleaves CSPGs does not appear to be as high as MMPs, perhaps limiting the impact of the peptidase as proteoglycanase in vivo (Somerville et al., 2003).Therefore the up-regulation of ADAMTS-9 expression following stroke is not necessarily a bad thing because it is unlikely that the normal ECM structure will be significantly degrade by the peptidase.
The identification of the cells responsible for the synthesis of ADAMTS-9 is important when analysing the functional aspects of ADAMTS-9.ADAMTS-9 could be the sheddase responsible not only for CSPGs breakdown, but also for release of growth factors and adhesion molecules (Luan et al., 2008).Moreover, astrocytes and macrophages can contribute to immune regulation through their role in resealing of the BBB and their direct effects on immune cell modulation molecules, such as TGFβ, and TNF and proteoglycans (Faulkner et al., 2004 , 2006).Proteoglycan, and principally CSPGs, are known for their immune-related activity in peripheral tissue.Owing to their adhesiveness to chemo-attractive agents and growth factors that are required for recruiting and activating immune cells, proteoglycans can capture these factors increasing their focal concentration and thereby targeting the immune response to the injured areas (Nandini & Sugahara, 2006;Rolls et al., 2006).Reid et al., (2009) provided evidence that IL-1β and TNF-α potentially up-regulated ADAMTS-9 in an astrocytic cell line (U373-MG) by activating transcription factor NFkB.This study provided strong evidence that ADAMTS-9 expression is modulated by pathological conditions that occurs in response to CNS injury (Reid et al., 2009).The ADAMTS-9 data described here are consistent with previous studies of ADAMTS-1 and -4 expressions, which were both up-regulated in MS.In contrast to ADAMTS-5 and -8, levels which were not significantly raised in response to experimental stroke or in MS (Chen et al., 2000;Haddock et al., 2006a;Tian et al., 2007).However, Luan et al., (2008) indicated that specific blocking antibodies against IL-1β and TNF-α dramatically inhibited ADAMTS-7 and ADAMTS-12 induction in chondrocytes in vitro, and the suppression of TNF-α and IL-1β expression by siRNA in human chondrocytes prevented ADAMTS-7 and ADAMTS-12mediated degradation of cartilage oligomeric matrix protein.Regardless of cell source, ADAMTS-9 expression by astrocytes and macrophages/microglia would allow cleavage of CSPGs, resulting in promotion of pro-inflammatory after acute CNS injury.These suggestions are in agreement with other reports indicating that, in the periphery, CSPGs regulate the motility and activation of macrophages (Hayashi, Kadomatsu, & Muramatsu, 2001;Rolls et al., 2006), dendritic cells (Kodaira, Nair, Wrenshall, Gilboa, & Platt, 2000) and other immune cell types (Rolls et al., 2006).

CONCLUSION
All previous studies have been reported that CSPGs, including aggrecan was up regulated in MS lesions, these increasing were associated with breakdown of ECM proteins including CSPGs.Here we report the cellular origin and distribution of ADAMTS-9 expression within clinically and neuropathologically confirmed MS and normal control white matter, assessed by IHC and confocal microscope.ADAMTS-9 expression was associated predominantly with activated macrophages/microglia, parenchymal astrocytes and to lesser extent with blood vessel endothelium and axons in MS white matter.We demonstrated here that ADAMTS-9 expression at the protein level was also increased in active inflammatory lesions with evidence of myelin breakdown, suggesting that upregulation of ADAMTS-9 may be a general phenomenon induced by CNS injuries.ADAMTS-9 could degrade aggrecan and versican and reverse the neurite outgrowth inhibition mediated by human brain proteoglycans allowing OPCs to cross the glial scar to repair axons.However, as repair in MS is limited; this degradation of ECM proteins deposited in lesions does not allow the repair process to be complete.Further studies on this enzyme are now required to define its function in normal CNS physiology and in the pathological process of MS.

Figure 1 :Figure 2 :
Figure 1: Dual label immunofluorescence of ADAMTS-9 and vWF in MS AL brain tissue sections: Dual label immunofluorescence for (a) vWF (green) and (b) ADAMTS-9 (red) in MS active lesion white matter (MS062 P1C3) with high levels of parenchymal ADAMTS-9 immunoreactivity.(C) Co-localisation is represented by the pixels in quadrant 3 on the graph and is demonstrated as yellow pixels in the composite image (d) following Zeiss 510 CSLM software reading of individual pixels for each fluorophore.V= vessel.R E T R A CT E D

Figure 3 :
Figure 3: Dual label immunofluorescence of ADAMTS-9 and NF in MS AL brain tissue: Dual label immunofluorescence for (a) NF (green) and (b) ADAMTS-9 (red) in MS active lesion white matter (MS062 P1C3) with high levels of parenchymal ADAMTS-9 immunoreactivity.(c) Co-localization is represented by the pixels in quadrant 3 on the graph and is demonstrated as yellow pixels in the composite image (d) following Zeiss 510 CSLM software reading of individual pixels for each fluorphore.Arrows indicate co-localisation of ADAMTS-9 and NF on a few axons.V= vessel.R E T R A CT E D

Figure 4 :
Figure 4: Dual label immunofluorescence of ADAMTS-9 and GFAP in MS AL brain tissue: Dual label immunofluorescence for (a) GFAP (red) and (b) ADAMTS-9 (green) in MS active lesion white matter (MS062 P1C3) with high levels of parenchymal ADAMTS-9 immunoreactivity.(C) Co-localization is represented by the pixels in quadrant 3 on the graph and is demonstrated as yellow pixels in the composite image (d) following Zeiss 510 CSLM software reading of individual pixels for each fluorphore.V= vessel.

Figure 5 :
Figure 5: Comparison of expression of ADAMTS-9 in MS and NAWM brain tissue by IHC.Image J software analysis shows expression of ADAMTS-9 normal control and active lesion.Low level of HLA-DR were observed in areas of intact myelin and no level recorded of ADAMTS-9.In contrast, high level of ADAMTS-9 in going demyelination along with activated microglia.Data represented as mean using kruskal-wallis, n=50 five reading for each region.R E T R A CT E D Journal of Alasmarya University: Basic and Applied Sciences al.