Reactive astrocytes in multiple sclerosis impair neuronal outgrowth through TRPM7‐mediated chondroitin sulfate proteoglycan production

Abstract Multiple sclerosis (MS) is a chronic inflammatory disorder of the central nervous system (CNS), characterized by inflammation‐mediated demyelination, axonal injury and neurodegeneration. The mechanisms underlying impaired neuronal function are not fully understood, but evidence is accumulating that the presence of the gliotic scar produced by reactive astrocytes play a critical role in these detrimental processes. Here, we identified astrocytic Transient Receptor Potential cation channel, subfamily M, member 7 (TRPM7), a Ca2+‐permeable nonselective cation channel, as a novel player in the formation of a gliotic scar. TRPM7 was found to be highly expressed in reactive astrocytes within well‐characterized MS lesions and upregulated in primary astrocytes under chronic inflammatory conditions. TRPM7 overexpressing astrocytes impaired neuronal outgrowth in vitro by increasing the production of chondroitin sulfate proteoglycans, a key component of the gliotic scar. These findings indicate that astrocytic TRPM7 is a critical regulator of the formation of a gliotic scar and provide a novel mechanism by which reactive astrocytes affect neuronal outgrowth.


| INTRODUCTION
Multiple sclerosis (MS) is a progressive inflammatory and demyelinating disease of the central nervous system (CNS). It is estimated that more than 2 million people worldwide suffer from MS, making it one of the most common chronic neurological disease of young adults. MS is generally considered as an auto-immune disease in which autoreactive T-lymphocytes and infiltrating monocytes induce focal inflammation and demyelination. Although demyelination is the most prominent histopathological feature of MS lesions, axonal damage has been shown to be a more accurate predictor of long-term clinical outcome and disease progression (Lubetzki & Stankoff, 2014).
In the last decades, great progress has been made in understanding and controlling the inflammatory components of MS, but the pathophysiological mechanisms that contribute to neurodegeneration and hamper neuroregeneration remain largely elusive (Lassmann, van Horssen, & Mahad, 2012). Astrocytes play an essential role in the regulation of CNS homeostasis by supporting neuronal function and metabolism (Finsterwald, Magistretti, & Lengacher, 2015). In MS lesions, astrocytes gain a reactive phenotype due to dysfunction of the blood-brain barrier, ongoing inflammation and chronic demyelination. Reactive astrocytes can facilitate blood-brain barrier repair, secrete immunosuppressive molecules and exert neuroprotective properties. However, they can also exacerbate inflammation and blood-brain barrier leakage by secreting proinflammatory molecules, thereby facilitating immune cell influx 3 Jack van Horssen and Helga E. de Vries contributed equally to this study. (Mizee et al., 2014;Sofroniew & Vinters, 2010;van Doorn et al., 2012). Furthermore, reactive astrocytes generate a glial scar, marked by interwoven astrocytic processes accompanied by an excessive accumulation of extracellular matrix (ECM) components. This glial scar may form a physical barrier around areas of demyelination to prevent widespread tissue damage, but also inhibits remyelination and axonal outgrowth (Nair, Frederick, & Miller, 2008;Sofroniew & Vinters, 2010). Thus, reactive astrocytes play a dual role in the pathogenesis of MS (for review see [Brosnan & Raine, 2013;Miljkovi c, Timotijevi c, & Mostarica Stojkovi c, 2011]).
Recent evidence shows that phospholipase C (PLC), which subsequently activates the inositol 1,4,5-trisphosphate (IP 3 )-Ca 2+ signaling cascade, is involved in pathophysiological functions of astrocytes (Kanemaru et al., 2013). It has been suggested that G protein-coupled receptors, which can be activated by a wide array of ligands, are primarily responsible for mediating (anti)inflammatory responses via PLC.
Therefore, a better understanding of their downstream signaling pathways may lead to the development of novel therapeutics, aimed to dampen the inflammatory responses in astrocytes. In this regard, altered expression and/or activation of IP 3 -dependent ion channels on astrocytes may play an important role in MS pathogenesis. Within this class of ion channels, transient receptor potential (TRP) channels have been recognized to be involved in numerous pathological conditions, including neurological disorders (Jordt & Ehrlich, 2007). Interestingly, previous reports identified Transient Receptor Potential Melastatin 7 (TRPM7), a member of the TRP channel family with inherent kinase activity, to be involved Alzheimer's disease, Parkinson disease, amyotrophic lateral sclerosis and stroke via both Ca 2+ and Mg 2+ dependent and independent mechanisms (Sun, Sukumaran, Schaar, & Singh, 2015). In addition, comparative gene expression analysis identified TRPM7 as one of the common upregulated genes in Alzheimer's disease and MS compared with healthy controls (Tseveleki et al., 2010). Besides a role for TRPM7 in neurodegenerative diseases, TRPM7 is involved in fibrotic diseases, such as liver fibrosis, pulmonary fibrosis, and cardiovascular fibrosis (Du et al., 2010;Fang et al., 2013;Yu et al., 2013). These disorders are characterized by excessive accumulation of ECM, similar as seen in MS lesions (Lau, Cua, Keough, Haylock-Jacobs, & Yong, 2013;van Horssen, Bö, Dijkstra, & de Vries, 2006;Van Horssen, Dijkstra, & De Vries, 2007). Here, we provide evidence for increased expression of TRPM7 in astrocytes within MS lesions and show that astrocytic TRPM7 impairs neurite outgrowth by enhanced production of ECM proteins.

| Brain tissue
Brain samples were obtained from 9 MS patients and 4 nonneurological controls, in collaboration with the MS Centrum Amsterdam and the Netherlands Brain Bank. Detailed clinical data are summarized in Table 1. All donors, or their next of kin, had given informed consent for brain autopsy and use of their brain material and clinical information for research purposes.
All sections were counterstained with hematoxylin after which they were analyzed with a light microscope (AXIO Scope A1, Carl Zeiss, Germany). For cellular localization studies, sections were incubated overnight with appropriate antibodies followed by incubation with Alexa-488-labeled goat antimouse IgG1 (1:200; Molecular Probes, USA) and analyzed by confocal microscopy (Leica DMI 6000 SP8, Leica, Germany).
Dissociated tissue was passed through a 100 μm cell strainer and centrifuged at 500g for 5 min. After which the pallet was resuspended in plating medium consisting of DMEM containing pyruvate, glucose and glutamine (Invitrogen, USA), penicillin (100 U/ml), streptomycin (100 mg/ml) (Lonza, Switzerland) and 10% FCS (Invitrogen, USA). Cells were cultured for 7-10 days. Microglia was removed from the mixed glia cell culture using an orbital shaker at 230 rpm for 3 hr.

| RNA isolation and real-time quantitative PCR
Gene expression analysis was performed on sub-confluent primary human astrocytes or U373 astrocytoma cells. Messenger RNA was isolated using Trizol (Invitrogen, USA) according to manufacturer's protocol. mRNA concentration and quality were measured using Nanodrop (Nanodrop Technologies, USA). cDNA syntheses was performed using the Reverse Transcription System kit (Promega, USA) following manufacturers guidelines. RT-PCR was performed as described previously (García-Vallejo et al., 2004). Primer sequences used are listed in supplementary Table 3.

| Intracellular Ca 2+ recording
Ca 2+ recordings were performed as described previously (Langeslag, Clark, Moolenaar, van Leeuwen, & Jalink, 2007;Visser et al., 2013). In short, cells were grown in 24 wells plate and incubated with Oregon Green 488 BAPTA-1-AM (Molecular Probes, USA) followed by further incubation in 2 ml DMEM F/12. The plate was mounted on an IX81 inverted epifluorescence microscope (Olympus, Japan). Recordings were made at 37 C in 5% CO 2 and 80% humidity. Excitation of Oregon Green-488 was performed at 480/25 nm and emission was detected between 502 and 538 nm. All Ca 2+ recordings were normalized by setting the response to ionomycin (Sigma-Aldrich, USA) at 100%.   Taken together, our data indicate that TRPM7 expression is markedly increased in reactive astrocytes within active and chronic active MS lesions and that in vitro, TGF-β1 induces gene expression of TRPM7.

| Astrocytes in vitro express functionally active TRPM7
U373 astrocytoma cells in which TRPM7 cDNA was stably transduced (TRPM7 + astrocytes) were used to further elucidate the role of TRPM7 in astrocytes. TRPM7 overexpression resulted in a six-fold induction of endogenous TRPM7 levels (Figure 2a). Overexpression of TRPM7 did not alter viability or proliferation as determined by MTS assay (Supporting Information Figure S1). To reveal whether TRPM7 protein is functionally expressed at the cell surface we used a wellaccepted Ca 2+ imaging approach. We examined the Ca 2+ influx via TRPM7 by stimulating astrocytes with bradykinin, a PLC-coupled receptor agonist leading to an IP 3 mediated Ca 2+ flux from internal stores. This initial Ca 2+ flux was followed by a prolonged Ca 2+ influx, likely via TRPM7, in cells overexpressing TRPM7 compared with mock astrocytes (Figure 2b). These data indicate that TRPM7 is functionally active at the cell surface of TRPM7 + astrocytes.

| Astrocytic TRPM7 expression inhibits neurite outgrowth
Since astrocytes play a central role in neuronal homeostasis and reactive astrocytes hamper neuronal function, we aimed to determine the effect of astrocytic TRPM7 overexpression on neuronal morphology.  Figure S2).

| DISCUSSION
In this study, we showed that TRPM7 is strikingly upregulated in reactive astrocytes in both active and chronic active MS lesions and that TGF-β1 treatment is able to enhance astrocytic TRPM7 expression (e) CSPG protein expression in mock astrocytes, chondroitinase ABC-treated TRPM7 + astrocytes, and TRPM7 + astrocytes. Quantification showed that CSPG protein production is significantly increased in TRPM7 overexpressing astrocytes compared with mock astrocytes expression highest in (Student's t test, n = 3) *p < .05 [Color figure can be viewed at wieyonlinelibrary.com] in vitro. We demonstrated that astrocytic TRPM7 negatively impacts neuronal outgrowth possibly via increased production of CSPGs.
Taken together, we identified TRPM7 as a novel player in MS pathophysiology involved in neuronal outgrowth, which might contribute to impaired neuroregeneration in MS patients.
TRPM7 is weakly expressed in astrocytes within control white matter and normal appearing white matter and strongly expressed in reactive astrocytes within active and chronic active MS lesions. Analysis of publicly available gene expression datasets (GEO accession number GSE38010) revealed that TRPM7 expression was increased in tissue containing active, chronic active or chronic MS lesions in four out of the five cases compared with control cases (Han et al., 2012).
TRPM7 is known to be involved in a variety of pathologies, including different cancer types and numerous fibrotic diseases (Middelbeek et al., 2012(Middelbeek et al., , 2015Visser, Middelbeek, van Leeuwen, & Jalink, 2014;Xu et al., 2015). Yet, little is known about a putative role of TRPM7 in neurological diseases. Dysregulation of neuronal TRPM7 activity has been associated with familial Alzheimer's disease, Parkinson's disease and stroke (Landman et al., 2006;Sun et al., 2015).
Overactive neuronal TRPM7 leads to calcium-induced cytotoxicity during cerebral ischaemia (Aarts et al., 2003;Hermosura et al., 2005) and changes in the function of TRPM7 may either be caused by genetic alterations or changes in regulatory pathways of TRPM7 channel activity (Sun et al., 2015). Enhanced astrocytic TRPM7 protein expression may be the result of the inflammatory environment and the production of TGF-β1 by activated glia and infiltrated macrophages (Miljkovi c et al., 2011;van Horssen et al., 2006). Exposure of astrocytes to TGF-β1 induced gene expression of TRPM7 and a recent study described that TGFβ-induced epithelial-to-mesenchymal transition in prostate cancer is mediated via TRPM7 expression (Sun, Schaar, Sukumaran, Dhasarathy, & Singh, 2018). Additionally, fibroblasts exposed to TGF-β1 promote extracellular matrix formation via upregulation of TRPM7 (Fang et al., 2014). Furthermore, it has previously been shown that treatment of astrocytes with TGF-β1 induces specific alterations that are consistent with astrogliosis (Cullen, Simon, & LaPlaca, 2007;Logan et al., 1994;Silver & Miller, 2004). Other inducers of a reactive astrocyte phenotype, such as Il-1α, TNF-α, and C1q, failed to induce TRPM7 expression in astrocytes in our hands. Interestingly, these mediators are known to induce a so called A1 astrocyte phenotype that is thought be neurotoxic whereas TGF-β1 on the other hand has been shown to limit the A1 reactive astrocyte phenotype (Liddelow et al., 2017). Based on these observations, we conclude that TRPM7 expression is not associated with the A1 neurotoxic astrocyte phenotype.
To allow in-depth investigation on the role of astrocytic TRPM7 expression, we generated a U373 astrocyte cell line with functionally active TRPM7. Manipulation of TRPM7 expression in vitro has been proven to be challenging, since strong upregulation, as well as downregulation of TRPM7 induces calcium toxicity and ultimately cell death (Nadler et al., 2001). Furthermore, long-term incubation with TRPM7 channel blockers also reduces cell viability (Zierler et al., 2011). However, in agreement to our data, it has been shown that a moderate increase in TRPM7 expression does not influence cell viability (Middelbeek et al., 2015).
Using TRPM7 + astrocytes as a feeder layer for primary neurons, we observed reduced neuronal outgrowth compared with neurons grown on top of mock-transfected astrocytes without affecting neuronal number. Despite co-culture of human U373 astrocytes together with primary rat neurons, we are able to obtain reliable results with this model. It would be interesting to inhibit TRPM7 using Waixenicin-A, a known specific TRPM7 inhibitor (Zierler et al., 2011), however, due to the ubiquitous expression pattern of TRPM7, it is impossible to selectively target astrocytes.
Reactive astrocytes are known to play a Janus-faced effect on neuronal outgrowth, by production of neurotrophic factors that facilitates neurite outgrowth and concomitant secretion of neurotoxic factors that hamper neuroregeneration (Landman et al., 2006;Liddelow et al., 2017). We demonstrated that that difference in neurite outgrowth was not mediated via secreted factors suggesting that physical interaction between the TRPM7+ astrocytes and neurons is needed for the observed changes in neuron morphology. Such a mechanism is via production of a glial scar, a physical barrier formed after demyelination to prevent widespread tissue damage. Astrocytes in glial scar tissue produce proteoglycans, such as aggrecan, brevican, neurocan, and versican, which reduce neurite outgrowth (Sobel & Ahmed, 2001). Our immunocytochemical analysis revealed that TRPM7 + astrocytes produce abundant levels of CSPGs, well-known outgrowth inhibitory and permissive molecules (Anderson et al., 2016). Future studies are needed to unravel which specific CSPG species are induced in TRPM7+ astrocytes and the mechanism by which TRPM7 modulates CSPG production.
CSPGs are also known to negatively affect oligodendrocyte maturation and myelination (Keough et al., 2016;Lau et al., 2012). In addition, recent data suggest that CSPGs boost the migratory potential of leukocytes to cross the glia limitans into the CNS, thereby contributing to inflammation (Stephenson et al., 2018). It is therefore possible that enhanced astrocytic expression of TRPM7 contributes to several pathological processes involved in MS lesion development and progression.
Complete prevention or removal of the glial scar causes increased inflammation and tissue damage thereby worsening functional outcome of several neuroinflammatory and neurodegenerative disorders (Anderson et al., 2016;Haroon et al., 2011;Robel, Berninger, & Götz, 2011). Yet, inhibition of CSPGs expressed by astrocytes after CNS insult is linked to improved axonal regeneration after trauma (Cregg et al., 2014;Sharma, Selzer, & Li, 2012). Therefore, a subtle approach, in which only the detrimental effects of the glial scar are targeted, is needed. Interestingly, FTY720, also known as fingolimod, a clinically approved drug to treat MS (Brinkmann et al., 2010) has been shown to be able to block the activity of TRPM7 (Qin et al., 2013). The inhibitory potential of FTY720 is however restricted to the bio-inactive, non-phosphorylated form of FTY720. When FTY720 is phosphorylated, it becomes the active compound of fingolimod and loses it inhibitory effect on TRPM7. Future studies are needed to address the question whether part of fingolimods mechanism of action is via inhibiting astrocytic TRPM7 mediated CSPG production. Finally, it will be interesting to study whether astrocyte specific knockout of TRPM7 would alter the clinical symptoms in mice suffering from experimental autoimmune encephalomyelitis (EAE), the animal model for MS. Global deletion of TRPM7 in mice is lethal (Jin et al., 2008), studies investigating the viability of mice with an astrocyte specific knockout of TRPM7 are needed to open up the possibility for an EAE study.
In conclusion, we show that TRPM7 is highly expressed in reactive astrocytes within MS lesions and that enhanced astrocytic TRPM7 levels impair neurite outgrowth by increased production of CSPGs, a key component of the gliotic scar. These findings indicate that astrocytic TRPM7 is a critical regulator of the formation of a gliotic scar and provide a novel mechanism by which reactive astrocytes affect neuronal outgrowth.

ACKNOWLEDGMENTS
We are grateful for Jeroen Middelbeek for helpful and inspiring discussions. Expert technical support by the AO|2 M facility (Advanced Optical Microscopy facility in O|2, VU Medical Center, Amsterdam) was highly appreciated. This work was supported by the MS research foundation (MS-14-358e, AK), which had no role in study design, data collection and analyses, decision to publish, or preparation of the article.