LIPID COMPOSITION OF MICROGLIAL CELLS AND EFFECTS OF A REMYELINATION-PROMOTING ANTIBODY ON THE SPHINGOLIPID PATTERNS

Multiple sclerosis (MS) is one of the most prevalent demyelinating diseases among young adults (Lemus et al., 2018). Myelin damage and oligodendrocyte death are caused by the autoimmune attack of T cells that cross the blood-brain barrier resulting in the formation of sclerotic plaques (Pearce, 2005, Lubetzki and Stankoff, 2014). In the early stages of the disease, damaged myelin is to some extent repaired, however the process is inefficient due to several factors, including impairment in OPCs recruitment and/or OLs differentiation likely due to the inflamed environment, and the inhibitory effect of myelin debris present at the lesion sites (Barres, 2008, Lassmann, 2014, Domingues et al., 2016). There are currently no cures for multiple sclerosis, only treatments that aim at alleviating its symptoms or slowing down its progression. Therefore, interventions that stimulate and promote endogenous myelin repair mechanisms may be a feasible treatment option (Lemus et al., 2018). In animal models of demyelination, the ability of certain monoclonal antibodies to stimulate myelin sheath synthesis has been proven. In this regard, the rHIgM22 monoclonal antibody has showed particularly promising results. This antibody seems to act mainly on cells of the oligodendrocyte lineage, in particular on oligodendrocyte precursor cells promoting their proliferation and inhibiting their apoptosis (Warrington et al., 2000). However, experimental evidence suggests that not only OPCs and OLs are targeted by rHIgM22, but also other glial cells, e.g., astrocytes and microglia, In particular, it has been shown that rHIgM22 stimulates an increase in the phagocytic activity in microglia (Zorina et al., 2018). The phagocytosis of myelin debris by microglia is one of the most delicate and limiting steps in the myelin repair mechanism (Neumann et al., 2009, Domingues et al., 2016, Napoli and Neumann, 2009, Lloyd and Miron, 2019). The protective effects exerted by rHIgM22, at least towards the primary targeted glial cells, appear to be mediated by the reorganisation of membrane lipids leading to the recruitment of a multimolecular complex that results in the development of numerous signalling mechanisms (Watzlawik et al., 2010, Watzlawik et al., 2013a). Indeed, data obtained in our laboratory showed that different glycosphingolipids were involved in rHIgM22 binding at the cell surface but also that rHIgM22 treatment induced changes in the sphingolipid pattern of oligodendroglial cells (Grassi et al., 2015; S. Grassi et al., 2021). Since microglia seem to be a crucial player in the processes that are at the center of myelin repair, they have been the focus of this thesis. We used as an experimental model the BV-2 microglial cell line. The BV-2 cell line is widely used because it retains the morphological, phenotypical, and functional characteristics described for freshly isolated primary microglia and has functional markers of macrophages (Blasi et al., 1990). BV-2 is a semi adherent cell line, where adherent and floating cells may represent two different sub-populations. Due to the paucity of information in the literature regarding the lipid composition of microglia, as the first step, we analysed the lipid composition of BV-2 cells using a combination of traditional chemical analysis and lipidomic using ESI-mass spectrometry. We analysed lipids in different culture conditions since it is well known that culture conditions affect the composition of certain lipid classes (Frechin et al., 2015). We identified differences in the lipid pattern depending on cell culture conditions but also some differences in the lipid composition among adherent and floating cells. Specifically, we observed key differences between adherent and floating cells at 80–90% confluence. In this experimental condition, ceramide was present at higher levels in the floating cells while sphingomyelin was more enriched in the adherent cells; in the cells collected at over confluence, these distinctions were absent. In addition, we identified the absence of specific lipids, including galactosylceramide and sulfatide, and a generally low amount of gangliosides in both sub-populations, irrespective of confluence. Then, we studied the sphingolipid turnover in more detail using a metabolic labeling approach. Our results showed that BV-2 cells are characterised by a rapid sphingolipid turnover rate. Finally, based on our previous findings indicating that rHIgM22 was able to modify sphingolipid patterns in OPCs and OLs, we evaluated the effects of rHIgM22 on the sphingolipid pattern of BV-2 cells using different experimental settings and culture conditions. Our findings showed that rHIgM22 could influence sphingolipid patterns in microglia as well. Comparing different experimental approaches, the effects of the two antibodies, rHIgM22 and the human IgM used as the negative control, are distinct. Specifically, rHIgM22 decreases GlcCer, SM, GM1, and PE in adherent cells relative to control IgM, whereas increases GlcCer and GD1a in floating cells. Moreover, in another experimental approach, rHIgM22 decreases GlcCer, SM, GM1, and PE in adherent cells relative to control IgM, whereas GlcCer and GD1a increase in floating cells. Therefore, it is plausible to hypothesise that also in microglia the effects of the treatment with rHIgM22 might be due to the reorganisation of the membrane lipid microenvironment, possibly affecting the function of membrane-associated signalling complexes.