Abnormal Upregulation of GPR17 Receptor Contributes to Oligodendrocyte Dysfunction in SOD1 G93A Mice.

Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease characterized by progressive loss of motor neurons (MN). Importantly, MN degeneration is intimately linked to oligodendrocyte dysfunction and impaired capacity of oligodendrocyte precursor cells (OPCs) to regenerate the myelin sheath enwrapping and protecting neuronal axons. Thus, improving OPC reparative abilities represents an innovative approach to counteract MN loss. A pivotal regulator of OPC maturation is the P2Y-like G protein-coupled receptor 17 (GPR17), whose role in ALS has never been investigated. In other models of neurodegeneration, an abnormal increase of GPR17 has been invariably associated to myelin defects and its pharmacological manipulation succeeded in restoring endogenous remyelination. Here, we analyzed GPR17 alterations in the SOD1G93A ALS mouse model and assessed in vitro whether this receptor could be targeted to correct oligodendrocyte alterations. Western-blot and immunohistochemical analyses showed that GPR17 protein levels are significantly increased in spinal cord of ALS mice at pre-symptomatic stage; this alteration is exacerbated at late symptomatic phases. Concomitantly, mature oligodendrocytes degenerate and are not successfully replaced. Moreover, OPCs isolated from spinal cord of SOD1G93A mice display defective differentiation compared to control cells, which is rescued by treatment with the GPR17 antagonist montelukast. These data open novel therapeutic perspectives for ALS management.


Introduction
Amyotrophic lateral sclerosis (ALS) is the most common adult onset motor neuron (MN) disorder involving MN degeneration and causing serious and irreversible muscle weakness and atrophy, with death of patients within 3-5 years from diagnosis [1]. Approximately 10% of patients have a familial form of this disease and among these, about 20% of ALS cases are caused by mutations in the gene encoding for superoxide dismutase 1 enzyme (SOD1) [2].
For a long time, ALS has been only considered a disease of the grey matter (GM) involving MN degeneration; however, this 'neuron-centric' view has been increasingly challenged over the last years. Alterations in the white matter (WM) structure have been reported to be more pronounced

Time-Dependent Upregulation of GPR17 Expression in the Spinal Cord of SOD1 G93A Mice
To characterize disease-related changes of the GPR17 protein, western blot analysis was performed on tissue homogenates obtained from spinal cords of both wild-type SOD1 (wtSOD1) and SOD1 G93A mice, collected at different ages corresponding to different stages of disease progression (namely, pre-symptomatic stage P30, early symptomatic stage P90 and late symptomatic stage P120). Results indicate that GPR17 protein levels were significantly increased at the pre-symptomatic stage P30, but only in the lumbar spinal cord tract ( Figure 1A). Furthermore, a significant increase in GPR17 protein levels was also detected at P90, but only in the thoracic spinal cord tract ( Figure 1B). More relevant was the increase observed in all spinal cord tracts of SOD1 G93A mice at the late symptomatic stage P120 ( Figure 1A-C). To better characterize the pool of GPR17-expressing (GPR17 + ) cells, we performed immunohistochemical analysis focusing on the ventral lumbar spinal cord, that is the region most affected by ALS pathology [33][34][35], of both wtSOD1 and SOD1 G93A mice, in order to detect possible regional differences between WM and GM areas. As shown in Figure 1D, the majority of GPR17 + cells displayed a highly ramified morphology typical of the oligodendroglial lineage and co-expressed the oligodendrocyte transcription factor 2 (Olig2, Figure S1A). The density of GPR17 + cells was found to be already significantly increased in the WM of the ventral spinal cord at P30 and, more markedly, in both WM and GM at P120 in SOD1 G93A mice with respect to wtSOD1 mice ( Figure 1D,E). Accordingly, the percentage of Olig2 + oligodendroglial cells co-expressing the GPR17 receptor was significantly increased in SOD1 G93A mice as compared to wtSOD1 animals at both pre-symptomatic and late symptomatic stages ( Figure S1B).  In parallel, alterations of the oligodendrocyte differentiation process were assessed by analyzing the density of CC1-expressing (CC1 + ) cells, representing cells that have completed their differentiation process and reached a mature phenotype [36]. The number of mature oligodendrocytes was slightly but significantly reduced at P30 in the ventral lumbar spinal cord WM of SOD1 G93A mice. This condition was exacerbated during the late disease progression phase (P120) in the ventral lumbar spinal cord (both in the WM and GM) of SOD1 G93A mice compared to wtSOD1 mice (Figure 2A,B). A similar reduction at P120 has been observed also for the percentage of Olig2 + cells co-expressing CC1 ( Figure S1C).
In parallel, alterations of the oligodendrocyte differentiation process were assessed by analyzing the density of CC1-expressing (CC1 + ) cells, representing cells that have completed their differentiation process and reached a mature phenotype [36]. The number of mature oligodendrocytes was slightly but significantly reduced at P30 in the ventral lumbar spinal cord WM of SOD1 G93A mice. This condition was exacerbated during the late disease progression phase (P120) in the ventral lumbar spinal cord (both in the WM and GM) of SOD1 G93A mice compared to wtSOD1 mice (Figure 2A,B). A similar reduction at P120 has been observed also for the percentage of Olig2 + cells co-expressing CC1 ( Figure S1C). Representative images of CC1 staining in the ventral lumbar spinal cord of wtSOD1 and SOD1 G93A mice at pre-symptomatic stage P30, early symptomatic stage P90 and late symptomatic stage P120. Hoechst 33258 was used to label cell nuclei. Dashed line separates white matter (WM) and grey matter (GM). Scale bar: 50 µm. (B) Histograms show the quantitative analysis of the density of CC1 + cells in WM and in GM of the ventral lumbar spinal cord (n = 3). * p < 0.05, SOD1 G93A vs wtSOD1; Student's ttest.

Impaired Oligodendrogenesis in the Developing Spinal Cord of SOD1 G93A Mice
It is known that, in rodents, myelinogenesis occurs predominantly postnatally within the first 3 weeks [37]. During this process, GPR17 expression progressively increases between P0 and P14 before massive myelin production, occurring around P21 [17]. To assess whether, as already observed for MNs [38], oligodendrocytes may also be altered before ALS symptoms onset, western blot analysis was extended to total spinal cord collected at very early post-natal periods, i.e., P2 and P7-10. Interestingly, a significant increase in GPR17 protein levels was found in the spinal cord of P7-P10 SOD1 G93A mice compared to age-matched wtSOD1, whereas no statistically significant changes were found at P2 ( Figure 3A-C). The increase of GPR17 expression was also confirmed by immunohistochemical analysis which revealed that this upregulation mainly affected the GM of

Impaired Oligodendrogenesis in the Developing Spinal Cord of SOD1 G93A Mice
It is known that, in rodents, myelinogenesis occurs predominantly postnatally within the first 3 weeks [37]. During this process, GPR17 expression progressively increases between P0 and P14 before massive myelin production, occurring around P21 [17]. To assess whether, as already observed for MNs [38], oligodendrocytes may also be altered before ALS symptoms onset, western blot analysis was extended to total spinal cord collected at very early post-natal periods, i.e., P2 and P7-10. Interestingly, a significant increase in GPR17 protein levels was found in the spinal cord of P7-P10 SOD1 G93A mice compared to age-matched wtSOD1, whereas no statistically significant changes were found at P2 ( Figure 3A-C). The increase of GPR17 expression was also confirmed by immunohistochemical analysis which revealed that this upregulation mainly affected the GM of ventral spinal cord of SOD1 G93A mice compared to wtSOD1 mice ( Figure 3D,E). Interestingly, in SOD1 G93A mice, we also found an increase of neural glial antigen 2 (NG2) expression in the GM of ventral spinal cord, in parallel with a strong reduction of newborn mature CC1 + oligodendrocytes in the WM of ventral spinal cord, suggesting an impairment of the differentiation process in the spinal cord of SOD1 G93A mice ( Figure 3D,E) during early development. ventral spinal cord of SOD1 G93A mice compared to wtSOD1 mice ( Figure 3D,E). Interestingly, in SOD1 G93A mice, we also found an increase of neural glial antigen 2 (NG2) expression in the GM of ventral spinal cord, in parallel with a strong reduction of newborn mature CC1 + oligodendrocytes in the WM of ventral spinal cord, suggesting an impairment of the differentiation process in the spinal cord of SOD1 G93A mice ( Figure 3D,E) during early development. Since, in SOD1 G93A mice, GPR17 expression had been found to be already significantly increased in the very early post-natal period, we decided to isolate OPCs from the spinal cord of P7 SOD1 G93A , Since, in SOD1 G93A mice, GPR17 expression had been found to be already significantly increased in the very early post-natal period, we decided to isolate OPCs from the spinal cord of P7 SOD1 G93A , wtSOD1, and WT mice in order to identify potential alterations in the capacity of these cells to proliferate and differentiate. To perform a detailed characterization of the differentiation process of spinal cord OPCs, cells were fixed after 2, 3, 4, and 5 days in culture (or days in vitro, DIV) as shown in Figure A1A. At the indicated time points, immunocytochemistry analysis was performed to get more information on the morphological characteristics and immunophenotype of GPR17 + cells, as described in Appendix A. More than 99% of cells expressed the oligodendroglial marker Olig2, indicating that the OPC culture was pure ( Figure A1D,E). The experimental protocol described above was then employed to compare the proliferation rate and differentiation capabilities of OPCs isolated from spinal cords of P7 SOD1 G93A , wtSOD1, and WT mice ( Figure 4A). No significant differences concerning both proliferation and differentiation capabilities were found between WT and wtSOD1 mice ( Figure S2); thus, OPCs obtained from WT mice were used as control. For proliferation studies, OPCs in proliferation medium were cultured in presence of 5 µM 5-ethynyl-2 -deoxyuridine (EdU) for 2, 7, or 24 h, fixed and stained for EdU. Results indicate no significant variation in the number of cells incorporating EdU in OPCs from SOD1 G93A mice compared to OPCs from WT mice ( Figure 4B). For the differentiation studies, OPCs were cultured in proliferation medium for 2 days and then in differentiating medium. After 3 days, cells were fixed and stained for GPR17 and myelin basic protein (MBP). A significantly reduced number of OPCs expressing the mature marker MBP (5.33 ± 0.61%) was found in cultures obtained from spinal cords of SOD1 G93A mice compared to OPCs from WT mice (9.84 ± 0.74%) ( Figure 4C). wtSOD1, and WT mice in order to identify potential alterations in the capacity of these cells to proliferate and differentiate. To perform a detailed characterization of the differentiation process of spinal cord OPCs, cells were fixed after 2, 3, 4, and 5 days in culture (or days in vitro, DIV) as shown in Figure A1A. At the indicated time points, immunocytochemistry analysis was performed to get more information on the morphological characteristics and immunophenotype of GPR17 + cells, as described in Appendix A. More than 99% of cells expressed the oligodendroglial marker Olig2, indicating that the OPC culture was pure ( Figure A1D,E). The experimental protocol described above was then employed to compare the proliferation rate and differentiation capabilities of OPCs isolated from spinal cords of P7 SOD1 G93A , wtSOD1, and WT mice ( Figure 4A). No significant differences concerning both proliferation and differentiation capabilities were found between WT and wtSOD1 mice ( Figure S2); thus, OPCs obtained from WT mice were used as control. For proliferation studies, OPCs in proliferation medium were cultured in presence of 5 µM 5-ethynyl-2′-deoxyuridine (EdU) for 2, 7, or 24 h, fixed and stained for EdU. Results indicate no significant variation in the number of cells incorporating EdU in OPCs from SOD1 G93A mice compared to OPCs from WT mice ( Figure 4B). For the differentiation studies, OPCs were cultured in proliferation medium for 2 days and then in differentiating medium. After 3 days, cells were fixed and stained for GPR17 and myelin basic protein (MBP). A significantly reduced number of OPCs expressing the mature marker MBP (5.33 ± 0.61%) was found in cultures obtained from spinal cords of SOD1 G93A mice compared to OPCs from WT mice (9.84 ± 0.74%) ( Figure 4C). Of note, at this time points, 74.16 ± 3.74% of GPR17 + cells of WT cultures exhibited a highly ramified morphology ( Figure 5A,B). In SOD1 G93A cultures, the number of GPR17-expressing cells with branched processes was significantly reduced to 64.28 ± 2.44% ( Figure 5A,B). This result was corroborated by the observation that, at this time point, in SOD1 G93A cultures, there still was a higher percentage of GPR17 + cells (35.56 ± 2.47%) with simple bi-or tri-polar morphology ( Figure 5A,B), Of note, at this time points, 74.16 ± 3.74% of GPR17 + cells of WT cultures exhibited a highly ramified morphology ( Figure 5A,B). In SOD1 G93A cultures, the number of GPR17-expressing cells with branched processes was significantly reduced to 64.28 ± 2.44% ( Figure 5A,B). This result was corroborated by the observation that, at this time point, in SOD1 G93A cultures, there still was a higher percentage of GPR17 + cells (35.56 ± 2.47%) with simple bi-or tri-polar morphology ( Figure 5A,B), confirming that OPCs from SOD1 G93A mice do show impaired differentiation capabilities. Interestingly, significantly increased levels of GPR17 mRNA were detected at DIV3 in OPCs from SOD1 G93A mice compared to those in OPCs from WT animals ( Figure 5C), suggesting that the upregulation of GPR17 could be responsible for the block of OPC differentiation, as already demonstrated for brain OPCs [26]. confirming that OPCs from SOD1 G93A mice do show impaired differentiation capabilities. Interestingly, significantly increased levels of GPR17 mRNA were detected at DIV3 in OPCs from SOD1 G93A mice compared to those in OPCs from WT animals ( Figure 5C), suggesting that the upregulation of GPR17 could be responsible for the block of OPC differentiation, as already demonstrated for brain OPCs [26]. Data reported above demonstrate that OPCs from SOD1 G93A mice exhibit an altered differentiation program in vitro compared to both WT and wtSOD1 cells and suggest that a pathological GPR17 upregulation could be at the basis of the detected differentiation block.
To assess whether the pharmacological blockade of GPR17 could restore impaired SOD1 G93A OPC terminal maturation, we used the GPR17 antagonist montelukast (MTK) [18,19,39,40] to treat OPCs isolated from the spinal cords of WT and SOD1 G93A mice. As shown in Figure 6, MTK was able to significantly increase the percentage of mature MBP + cells in cultured OPCs from SOD1 G93A mice, while it had no effect on WT OPC cultures ( Figure 6A,B). No effects of the treatment have been detected on either proliferation or cell viability ( Figure S3). Importantly, the morphological characterization of cells revealed that the increased number of MBP + cells found after MTK treatment in SOD1 G93A cultures could be ascribed to cells displaying a bushy morphology, typical of cells capable of producing myelin sheaths [41], and not to cells with a ring-like structure, which is instead indicative of still immature phenotype ( Figure 6C,D). Globally, these results unveil the existence of intrinsic differentiation defects in SOD1 G93A OPCs and highlight the GPR17 blockade by its antagonist MTK as an effective approach to correct this dysfunction.

Restoration of the Differentiation Capabilities of OPCs from P7 SOD1 G93A Mice by Montelukast
Data reported above demonstrate that OPCs from SOD1 G93A mice exhibit an altered differentiation program in vitro compared to both WT and wtSOD1 cells and suggest that a pathological GPR17 upregulation could be at the basis of the detected differentiation block.
To assess whether the pharmacological blockade of GPR17 could restore impaired SOD1 G93A OPC terminal maturation, we used the GPR17 antagonist montelukast (MTK) [18,19,39,40] to treat OPCs isolated from the spinal cords of WT and SOD1 G93A mice. As shown in Figure 6, MTK was able to significantly increase the percentage of mature MBP + cells in cultured OPCs from SOD1 G93A mice, while it had no effect on WT OPC cultures ( Figure 6A,B). No effects of the treatment have been detected on either proliferation or cell viability ( Figure S3). Importantly, the morphological characterization of cells revealed that the increased number of MBP + cells found after MTK treatment in SOD1 G93A cultures could be ascribed to cells displaying a bushy morphology, typical of cells capable of producing myelin sheaths [41], and not to cells with a ring-like structure, which is instead indicative of still immature phenotype ( Figure 6C,D). Globally, these results unveil the existence of intrinsic differentiation defects in SOD1 G93A OPCs and highlight the GPR17 blockade by its antagonist MTK as an effective approach to correct this dysfunction.

Discussion
Through the last decades, purinergic signaling has been involved in modulation of OPC proliferation, migration, and myelination. The large number of purinergic receptors identified on both OPCs and mature oligodendrocytes, the different signaling pathways induced by their activation and the combined activity of ectonucleotidase enzymes make this system highly complex [32]. Of note, expression and function of some purinergic receptors were found to be altered under disease conditions characterized by neurodegeneration, aberrant inflammatory response, and oligodendrocyte dysfunction [32]. Moreover, a central role of P1 and P2 receptors in the pathogenesis of ALS has been also suggested [42,43], raising interest for this still unexplored system in view of its exploitation for novel therapeutic opportunities for this incurable disease. Previous studies from our

Discussion
Through the last decades, purinergic signaling has been involved in modulation of OPC proliferation, migration, and myelination. The large number of purinergic receptors identified on both OPCs and mature oligodendrocytes, the different signaling pathways induced by their activation and the combined activity of ectonucleotidase enzymes make this system highly complex [32]. Of note, expression and function of some purinergic receptors were found to be altered under disease conditions characterized by neurodegeneration, aberrant inflammatory response, and oligodendrocyte dysfunction [32]. Moreover, a central role of P1 and P2 receptors in the pathogenesis of ALS has been also suggested [42,43], raising interest for this still unexplored system in view of its exploitation for novel therapeutic opportunities for this incurable disease. Previous studies from our and other laboratories demonstrate that the P2Y-like GPR17 receptor represents a good pharmacological target to implement repair and remyelination under several neurodegenerative conditions including cerebral ischemia, multiple sclerosis, and traumatic brain injury [32,44]. However, so far, the involvement of GPR17 has never been studied in the context of ALS pathology. Understanding whether alterations of GPR17 expression are present in ALS is important to identify the disease stage at which a GPR17-based therapeutic approach would be more relevant.
Here we show, for the first time, an upregulation in GPR17 expression in the spinal cord of a murine model of ALS. A slight, but significant, increase of GPR17 protein levels was already detectable in the lumbar spinal cord of SOD1 G93A mice at pre-symptomatic stage P30 compared to wtSOD1 mice. Immunofluorescence analyses showed that GPR17 upregulation occurs specifically in cells belonging to the oligodendrocyte lineage. These results are in line with previous findings reporting oligodendroglial alterations prior to MN degeneration and before symptomatic manifestation of the pathology [13], and the major involvement of the lumbar tract as the region of the spinal cord being most affected by the pathology [33][34][35].
At P90, the difference in GPR17 levels was no longer present in the lumbar tract but was instead slightly detected in the close thoracic tract. This is consistent with the fact that the thoracic and cervical portions are affected by the pathology to a smaller extent, and probably at later stages after clinical onset, compared to the lumbar one. However, the most prominent increase in GPR17 protein expression was detected at late symptomatic stage (P120) in all tracts of the spinal cord. Interestingly, in the lumbar tract, GPR17 upregulation mainly affected the ventral region of spinal cord and was accompanied by loss of mature CC1 + oligodendrocytes. These data are in line with reactive changes of OPCs in this area, as already demonstrated previously [45].
Altogether, our results confirm that, in a similar way to other experimental models of neurodegenerative disorders, cells expressing GPR17 precociously react to damage also in SOD1 G93A mice. Our results also confirm that this reaction is already present at disease pre-symptomatic stage in the lumbar tract. We speculate that, at this early time point, the pool of OPCs expressing GPR17, which is maintained in the adult CNS for repair purposes [23], is rapidly mobilized in order to compensate for initial oligodendroglia dysfunction. This is nevertheless in agreement with already published data indicating that proliferation of NG2 + precursors is enhanced at pre-symptomatic stages [12,13]. However, despite this rapid reaction of the GPR17 + pool of precursors, these cells fail to properly maturate at late disease stages, as confirmed by the reduction of the number of mature CC1 + oligodendrocytes. Recently, mutant SOD1 expression in mature oligodendrocytes has been demonstrated to be sufficient to induce myelin defects driving MN degeneration [15]. On this basis, we speculate that GPR17 upregulation starts very early as a positive event to promote substitution of dysfunctional oligodendrocytes. However, when upregulation is maintained in differentiating OPCs for excessively long times, terminal maturation is impeded and myelination impaired (see also below and [26]). In this respect, several studies have highlighted how GPR17 correct expression timing is the result of the complex integration of intrinsic determinants regulating oligodendroglial differentiation with the extracellular stimuli acting on the Gpr17 gene [44], that could themselves be altered during ALS disease progression. This important issue still remains to be investigated.
Symptom onset in SOD1 G93A mice is a very controversial topic within the scientific community [6,34,[46][47][48]. However, although behavioral alterations become evident at adulthood, histological and biochemical modifications mainly affecting MNs are already detectable during embryonic [38] and postnatal development up to P10 [49]. Abnormalities in neuronal architecture, excitability, and axonal transport have been already described at very early stages of the embryonic development [50][51][52]. Interestingly, the first signs of alterations for glial cells, such as activation of astrocyte and microglia and increased number of OPCs have been also described in lumbar spinal cord of SOD1 G93A versus wtSOD1 mice around P30 [38,48,49,53], but no data are reported before this disease stage. Our immunohistochemical analysis revealed GPR17 upregulation in oligodendrocytes of the ventral spinal cord at very early developmental stages of the disease (P7-10), which is accompanied by a reduction of the number of CC1 + mature oligodendrocytes. Globally, the findings of our in vivo studies suggest that, during disease progression, oligodendrocytes exhibit alterations before disease onset in SOD1 G93A mice. Importantly, these early defects in oligodendrocytes may be linked to myelin abnormalities that have been previously observed prior to symptoms onset in both zebrafish [15] and murine mutant SOD1 models [54], thus contributing to progressive axonal loss. On this basis, restoring the proper OPC differentiation capability may help to preserve MN functionality and to counteract their degeneration.
To implement the analysis of OPC dysfunction in ALS and to study the effects of pharmacological compounds targeting GPR17, we moved to in vitro studies using primary cultures of spinal cord OPCs from SOD1 G93A mice. Results from immunocytochemical analysis showed that, while the proliferation rate is almost the same, the differentiation ability of SOD1 G93A OPCs appears to be altered compared to both WT and wtSOD1 controls. In fact, after 2 days in proliferation medium and 3 additional days in differentiation medium, the percentage of MBP + cells and the amount of GPR17 + OPCs with a more advanced multi-branched morphology were significantly lower in SOD1 G93A than in WT mouse cultures. The percentage of GPR17-expressing cells with an immature bi-or tri-polar morphology was instead still high in cultures from SOD1 G93A animals compared to WT controls, confirming that mutant cells are less differentiated at this time in culture. Thus, it can be hypothesized that expression of mutated SOD1 in differentiating OPCs may lead to accumulation of reactive species within these cells, which in turn can induce the release of damage signals able to over-stimulate GPR17 expression resulting in a maturation block [44]. These results are in line with previously published data reporting that, in SOD1 G93A mice, the newly generated OPCs with repair purposes fail to restore oligodendroglial dysfunction, since they do not reach the final stage of maturation to mature myelinating oligodendrocytes, leaving large areas of demyelination [12]. Very interestingly, our results obtained with spinal cord OPCs are different of those obtained by Ferraiuolo et al., who did not observe any difference in the differentiation capabilities of cortical OPCs obtained from SOD1 G93A mice compared to those obtained from WT littermates [14]. A possible explanation of these discrepancies could be found in the different intrinsic properties and heterogeneity of spinal cord and cortical OPCs which are strictly related to their CNS region location [55].
Based on previous studies highlighting the involvement of GPR17 in the regulation of OPC differentiation program [18,32], we assessed whether pharmacological targeting of GPR17 could help restoring the maturation program of OPCs from SOD1 G93A mice. As detailed above, both our in vivo and in vitro results clearly show that GPR17 is pathologically upregulated within these cells and that this is associated to a concomitant blockade of OPC terminal maturation. Under this condition, we reasoned that the pharmacological manipulation of GPR17 could be useful in re-establishing a correct receptor activity and in restoring OPC functions. In this contest, the use of either an agonist or an antagonist is still a debated issue likely because the conflicting results of in vitro experiments [28,[56][57][58][59][60][61][62][63][64]. Previous works had already demonstrated that the in vivo pharmacological inhibition of GPR17 is able to improve WM integrity, increasing the number of mature oligodendrocytes and promoting remyelination [30,31]. Accordingly, our in vitro results demonstrate that MTK, an already marketed drug whose ability to antagonize GPR17 has been previously shown in primary purified OPCs and in the murine Oli-Neu oligodendroglial cell line [18,19] and recently confirmed in radioligand binding studies [40], markedly increased the percentage of cells expressing MBP in comparison to vehicle-treated cells in SOD1 G93A cultures. These data confirm that blockade of GPR17 under pathological condition characterized by its abnormal upregulation can successfully induce OPCs to resume their normal differentiation program. Even though this drug is not a selective antagonist of GPR17, its use in primary OPC cultures from SOD1 G93A mice-in which CysLT1 receptor mRNA expression was undetectable (as previously described also for WT OPCs [18])-allows us to rule out the observed effects be due to blockade of this cysteinyl leukotrienes receptor subtype. Of course, in vivo administration of MTK in SOD1 G93A mice may provide beneficial effects not only by acting on GPR17-expressing OPCs, but also by managing other pathological features of ALS, including oxidative stress and neuroinflammation, as already described for different disease models [30]. Furthermore, being an already marketed orally available drug, MTK may represent a multi-target drug with high translational potential for repurposing strategies [65], thus deserving further evaluations in vivo.
In conclusion, from this study, GPR17 emerges as a critical player in ALS pathogenesis and a new potential pharmacological target to be exploited to develop novel therapeutic approaches to counteract oligodendrocyte dysfunction in ALS and to retard both MN degeneration and disease progression.

Animals and Genotyping
B6SJL-TgN SOD1/G93A (+)1Gur mice expressing high copy number of mutant human SOD1 with a Gly 93 Ala substitution (SOD1 G93A ), B6SJL-TgN (SOD1)2Gur mice expressing wild-type human SOD1 (wtSOD1) [66], and wild-type (WT) mice were originally obtained from Jackson Laboratories (Bar Harbor, ME, USA) and bred at the animal facility of the Pharmacology and Toxicology Unit, Department of Pharmacy in Genoa. Transgenic animals have been crossed with background-matched B6SJL wild-type female and selective breeding maintained each transgene in the hemizygous state. All transgenic (human SOD1 G93A or human wtSOD1) and non-transgenic mice (WT) were identified analyzing crude extracts obtained from tail tips. Tail tips were homogenized in phosphate-buffered saline (PBS) solution, lysed by two cycles of freezing and thawing, and centrifuged at 23,000× g for 15 min at 4 • C. The SOD1 level was evaluated by staining for its enzymatic activity after 10% non-denaturing polyacrylamide gel electrophoresis [46]. Animals were housed at constant temperature (22 ± 1 • C) and relative humidity (50%) with a regular 12 h/12 h light cycle (light 7:00 a.m.

Western Blot
Spinal cords have been collected from SOD1 G93A and wtSOD1 mice at specific disease stages (i.e., postnatal day-P-2, 7/10, 30, 90, 120) and divided in three different portions (corresponding to cervical, thoracic, and lumbosacral regions). Animal tissues were then lysed and mechanically homogenized in lysis buffer (20 mM Tris pH = 7.2, 0.5% DOC, 1% Triton, 0.1% SDS, 150 mM NaCl, 1 mM EDTA, Sigma Aldrich, Milan, Italy) added with phosphatase inhibitors (2 mM EGTA, Sigma Aldrich) and 1% of protease inhibitors (Sigma Aldrich) for each sample. A range of 25-30 µg aliquots from each protein sample were loaded on 8% sodium-dodecylsulphate polyacrylamide gel and electrophoretically transferred onto PVDF membranes (BioRad Laboratories, Segrate, Italy). Membranes were immunoblotted overnight at 4 • C with rabbit GPR17 antibody (1:300, custom antibody produced by PRIMM, Milan, Italy) and incubated with HRP-conjugated secondary antibodies, both diluted in 5% non-fat dry milk (BioRad Laboratories). Target proteins were detected by ECL detection kit (BioRad Laboratories) and analyzed by ImageJ program. For each sample, α-tubulin (Sigma Aldrich) was used as internal control to normalize GPR17 protein levels and normalized values were used for comparison, expressed as percentage of control lane values as previously described [26].

Total RNA Extraction, Retrotranscription, and Real-Time PCR
After 1 day in culture in differentiation medium, SOD1 G93A and WT cells were lysed with TRIZOL ® reagent (Life Technologies). Total RNA was extracted using Direct-zol™ RNA Micro-Prep (Zymo Research, Irvine, CA, USA) according to the manufacturer's instructions. RNA was then pre-treated with RQ1 DNase (Promega, Milan, Italy) for eliminating genomic DNA contamination. Retrotranscription of 400 ng RNA was performed with SensiFAST™ cDNA synthesis kit (Bioline, London, UK). For real-time PCR, several mixes were prepared according to the number of interested genes. Each mix included Master Mix 2x (Life techonologies), 250 nM probe (for GPR17 Mm02619401_s1, for Rpl13a Mm05910660_g1) and 20 ng of cDNA. Gene-expression was analyzed with TaqMan ® Gene Expression Assay and normalized to housekeeping gene Rpl13a expression using CFX96 real-time PCR system (BioRad Laboratories) following the manufacturer's protocol.

Pharmacological Treatment
After one day in differentiation medium, primary spinal cord WT and SOD1 G93A OPC cultures were treated with the non-selective GPR17 antagonist montelukast (MTK; 1 µM, Cayman, MI, USA) diluted in dimethylsulfoxide (DMSO, Sigma Aldrich). OPCs treated with the same amount of DMSO (indicated as vehicle) alone were used as controls. After 48-72 h, cells were fixed and stained for Olig2 and for the mature cell marker MBP (for details see above). In the same experimental conditions, Olig2 + cell viability was evaluated using the Click-iT™ Plus TUNEL Assay Alexa Fluor-488 Kit (Life Technologies) according to manufacturer's protocol.

Statistical Analisys
All results were expressed as mean ± standard error (SE). Statistical analysis was performed using the nonlinear multipurpose curve-fitting Graph-Pad Prism program (Graph-Pad). The statistical test used was chosen according to the type of experiment performed and was indicated in the legend of the figure. Four degrees of significance were considered: p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), p < 0.0001 (****).

Appendix A
We set up primary purified OPC cultures from spinal cord of P7 wild-type mice. Briefly, dissected spinal cords of P7 WT mice were enzymatically and mechanically dissociated using MACS technology (Miltenyi Biotec; see methods section for details); in order to obtain a pure OPC cell culture, cells expressing PDGFRα (also known as CD140a) were isolated from single-cell suspension, maintained in proliferation medium for 2 days and then switched to a differentiation medium supplemented with T3 for additional 3 days, to promote differentiation into highly ramified MBP-expressing (MBP + ) cells. The expression of the GPR17 protein in cultured wild-type OPCs was studied in parallel with NG2 and MBP during their spontaneous maturation process ( Figure A1). As shown in Figure A1B,C (red line), at DIV2 a 95.43 ± 3.46% of total cell population was positive for NG2 and most cells presented a typical bi-polar morphology with little secondary branching. GPR17 was already detectable in the majority of cells with a bi-polar or tri-polar morphology, even if the protein could be detected as a single intracellular spot close to the nucleus, reminding of localization of the protein into the Golgi apparatus, or also including the initial segments of few processes in tight contiguity with the Golgi apparatus, as previously demonstrated in OPCs isolated from the brain [17]. NG2-GPR17 double-positive cells (NG2 + /GPR17 + ) accounted for the 80.61 ± 8.02% of the total NG2 population. As OPCs began to differentiate spontaneously in vitro, an increased and strong GPR17 expression was found in NG2 + cells with many branched processes emerging from the cell body, reaching a percentage of 93.57 ± 5.72% of NG2 + /GPR17 + cells of the total NG2 population after 1 day in T3 (corresponding to DIV3). The number of GPR17 + OPCs continued to increase during differentiation ( Figure A1B,C; green line), reaching a maximum peak around DIV4 when they represented about the 97.42 ± 1.1% of the total cell population, and when a significant number of cells acquired a pre-oligodendrocyte phenotype. At this stage, 99% of GPR17 + OPCs were still slightly immunoreactive for NG2, whereas only 3.28 ± 0.04% of cells were immuno-positive for MBP. At later stages of maturation, the number of MBP + cells ( Figure A1B,C; blue line) progressively increased reaching a 9.87 ± 1.77% at DIV5 with a percentage of GPR17-MBP double-positive cells (GPR17 + /MBP + ) of 10.61 ± 2.23%. Nevertheless, MBP + oligodendrocytes that reach terminal differentiation downregulated GPR17, according to previously published data from our laboratory on brain OPCs [18].