Receptor tyrosine kinases play a significant role in human oligodendrocyte inflammation and cell death associated with the Lyme disease bacterium Borrelia burgdorferi

Lyme neuroborreliosis (LNB), caused by the spirochete Borrelia burgdorferi, is a morbid form of Lyme disease with complex neurological sequelae. LNB manifests in about 15% of patients with primary erythema migrans, and may be evidenced as a lymphocytic meningo-radiculitis, lymphocytic meningitis, cranial neuritis, radiculoneuritis, radiculoneuropathy, and/or, rarely, encephalitis and myelitis [1]. The pathogenesis of LNB is not clearly understood. We have established a rhesus macaque model and several in vitro and ex vivo models to help delineate mechanisms of acute LNB. Intrathecal inoculation of live B. burgdorferi in rhesus macaques elicited in these animals signs of acute LNB, including leptomeningitis, radiculitis, inflammatory lesions in dorsal root ganglia (DRG), accompanied by glial and neuronal apoptosis in the DRG [2]. Stereotaxic intraparenchymal inoculations in rhesus macaques have similarly shown oligodendrocyte and neuronal cell death by apoptosis in the central nervous system (CNS) [3]. In either study, apoptosis of neuroglial cells occurred in an inflammatory environment, leading us to hypothesize that inflammation elicited by B. burgdorferi is a leading cause of LNB pathogenesis. In proof of principle, administration of dexamethasone, an anti-inflammatory agent, in vivo in rhesus macaques, not only downregulated the pleocytosis of lymphocytes and monocytes in the cerebrospinal fluid (CSF) but also glial and neuronal cell death [4]. In separate in vitro studies, B. burgdorferi has been documented to induce inflammation and cell death by apoptosis in neurons in the presence of microglia [5], and without any other cell involvement in the case of oligodendrocytes [6]. Downregulation of inflammation by pretreatment with dexamethasone, also inhibited oligodendrocyte cell death. In addition, a study of the mechanisms of inflammation in oligodendrocytes has shown that the MEK/ERK pathway plays a predominant role in inducing not only the production of several key inflammatory mediators but also the expression of the transcription factor p53 [7]. Inhibition of the ERK pathway suppressed p53 levels, and suppression of either the MEK/ERK pathway or the mitochondrial p53 downregulated apoptosis in oligodendrocytes, indicating that this mitogen-activated protein kinase (MAPK) pathway plays an important role in both inflammation and apoptosis of these glial cells. In continuation of this study, the current manuscript explores the roles of several surface receptors in mediating such downstream events, and we show that the signaling mechanisms in human oligodendrocytes in response to B. burgdorferi in vitro are part of a multi-factorial complex process with non-traditional receptors playing key roles.

Lipoproteins and flagellin are major components of B. burgdorferi that can initiate signaling events through TLR2 and TLR5, respectively, in many cell types [10, 15, 16], with TLR2-mediated events being the most widely studied. With respect to Lyme neuroborreliosis, TLR2, TLR4, and TLR5 expression was shown to be upregulated in primary rhesus microglia upon B. burgdorferi exposure [9] with TLR2 and TLR5 subsequently shown to play a significant role in the induction of immune responses to the Lyme disease bacterium in these glial cells [10]. Primary rhesus astrocytes showed an inflammatory response to L-OspA, a major TLR2-binding protein of B. burgdorferi as well as to FliC, a TLR5-binding ligand [9]. Hence, we decided to investigate whether these receptors similarly played a significant role in induction of inflammation in human oligodendrocytes, which were previously demonstrated to undergo apoptosis under an inflammatory milieu upon B. burgdorferi exposure [2, 6]. We show here that unlike with microglia or astrocytes, where these TLRs had a significant role in the induction of innate immune responses, oligodendrocyte TLR2 (and TLR5) had mainly an inhibitory role, with a modest role in IL-6 induction (Tables 1 and 2). TLR4 also had only a minor role in the induction of CCL5 as confirmed with both TLR4 siRNA and inhibitor. This was evidence to suggest that other receptors, either other TLRs, integrins, or novel receptors, likely had a significant role in the innate immune responses of oligodendrocytes to B. burgdorferi. It also indicated that although receptors such as TLR2 or TLR5 are common factors in B. burgdorferi-mediated signaling events in multiple cell types, differences do exist, and every cell type needs to be evaluated before a consensus can be reached, or treatment targets devised.

We next decided to look at non-TLR receptors that might play a pivotal role in mediating inflammation and apoptosis. In other studies, EGFR, PDGFR, and other receptor tyrosine kinases were shown to mediate neuroinflammation in various models of neurological disorders [11‐13]. We found that inhibition of signaling through EGFR/FGFR/PDGFR downregulated not only the inflammation associated with B. burgdorferi exposure but also the accompanying cell death via apoptosis (Table 3 and Fig. 1). This indicated that in MO3.13 human oligodendrocytes, these specific receptor tyrosine kinases play a key role in the neuroinflammation and apoptosis that is associated with exposure to B. burgdorferi, similar to TLR2- and TLR5-mediated neuroinflammation in microglia [10].

In our previous work, we demonstrated that the ERK pathway plays an important role in both MO3.13 oligodendrocyte neuroinflammation and apoptosis via mitochondrial p53 [7]. In this study, we show that EGFR/FGFR/PDGFR-mediated signaling similarly directs both phenomena upon B. burgdorferi exposure. As these receptors are known to trigger the ERK pathway [17‐20], it is likely that B. burgdorferi exposure triggers activation of these receptors, which in turn stimulates the ERK MAPK pathway predominantly, leading to inflammatory molecule production and p53-mediated apoptosis in these cells. We also undertook to investigate the role of these receptors in primary human oligodendrocyte cells (ScienCell Laboratories Carlsbad, CA), which have also been characterized with respect to inflammation, apoptosis, and the role of the MEK/ERK pathway in Lyme neuroborreliosis [6, 7]. Unfortunately, the concentrations of the pertinent reagents that were used with MO3.13 cells proved too toxic for the primary cells (data not shown). Due to the rare availability of these cells, currently, we are unable to continue experiments with primary human cells. We plan to resume this approach once the cells become available.

The RTK inhibitor, according to the manufacturer, inhibits EGFR, FGFR1, and PDGFRβ receptors. We next looked at these receptors’ expression pattern in order to discern any variability in expression upon B. burgdorferi exposure in comparison to cells containing medium alone. We did not see any visible signal for EGFR in a pilot study using a mouse monoclonal antibody. It is possible that this was a technical problem as all other antibodies used were polyclonal. EGFR expression has also been documented to decrease from embryonic to adult stages in the mouse midbrain, with higher expression in oligodendrocyte precursors. It is possible that this might be a factor wherein the expression was punctate and drowned out by background fluorescence [21]. Additionally, it is also possible that a mutated/truncated EGFR is present, not clearly detectable by the monoclonal. PDGF has been shown to increase myelin gene expression and number of mature oligodendrocytes [22]. As PDGF signals through its receptor PDGFR, it is not surprising to see very high expression of PDGFRβ in the mature MO3.13 cells (PDGFRα expression has been shown to be primarily in oligodendrocyte progenitors and pro-oligodendrocytes and absent in mature cells [23]). FGFRs particularly 1 and 2 are required for the generation of oligodendrocyte precursors [24], and FGFR3 for the terminal differentiation of oligodendrocytes [25], with their expression pattern controlled over the developmental stages. Although their levels vary, all three receptor forms are expressed by mature oligodendrocytes in rats [26]. The expression patterns of FGFR1 and FGFR3 are similar to those shown in some previously published studies [27‐30]. FGFR2 was also expressed in these cells; however, the pattern of expression was not consistent, which might be attributed to technical issues with the antibody (not shown). Also, we could not establish significant differences in expression levels between medium control and those exposed to B. burgdorferi through receptor staining (Fig. 2). It is possible that differences might be established by staining for phosphorylated forms of the receptors or through Western blots.

We next attempted to delineate which of the three receptors involved primarily mediate(s) the neuroinflammation and cell death associated with B. burgdorferi. However, as seen in Table 4, the individual inhibition of each of these receptors did not elicit a major effect on suppression of multiple chemokines/cytokines. Neither did the combination of EGFR/FGFR, EGFR/PDGFR, or FGFR/PDGFR (data not shown). It is possible that inhibition of signaling through one receptor is compensated with increased signaling through the other two (or others) resulting in the increased production of inflammatory mediators as we observed (Table 4). Therefore, it is likely that all three receptors need to be inhibited simultaneously to achieve the multiple chemokines/cytokines suppression phenotype.

C-src, an intracellular tyrosine kinase, is upstream of ERK activation but below the RTKs in signaling sequence of events. It has been known to be associated with RTK such as EGFR, FGFR, and PDGFR among others [31]. While src kinase family inhibition had some significant effect on chemokine and cytokine production (Table 4), it did not achieve the level of inhibition seen with ERK inhibition alone [7]. It also did not produce conclusive effects on cell death (not shown). It is likely that another molecule associated with receptors, like phospholipase Cγ for instance [32] is also needed to initiate downstream events. Therefore, a cumulative contribution may be needed to achieve a threshold of ERK activation so as to generate the level of inflammatory output and p53 levels to initiate apoptosis. In view of these results, our tentative conclusion is that exposure of MO3.13 human oligodendrocytes to B. burgdorferi engages TLR2, 4, 5, EGFR/FGFR/PDGFRs, with the latter three being the main inducers of all five chemokines and cytokine; TLR2 and 5 perhaps contributing towards IL-6 output, while TLR4 effecting CCL5 production. On this basis, we hypothesize that activation of RTKs recruits c-src and others to their intracellular domains leading to increased activation of the MEK/ERK pathway via multiple receptors, yielding increased levels of inflammatory mediators and p53 levels to initiate apoptosis.

We show here that receptors that have no known PAMP associated with B. burgdorferi significantly mediate inflammation and apoptosis of MO3.13 oligodendrocytes. These cells do not produce significant EGF or FGF2 upon exposure to B. burgdorferi but have not been evaluated for PDGF expression [7]. However, since inhibition of PDGFR did not elicit a significant downregulation of most chemokines/cytokine (Table 4), it is possible that these receptors are transactivated through ligand-independent mechanisms. G-protein-coupled receptors (GPCR) and integrins have both been demonstrated to transactivate RTKs through multiple mechanisms, including activation of src/src family kinases, which in turn can activate RTK intracellular domain [31, 33‐36]. B. burgdorferi has been shown to bind and activate integrins, while activation of GPCRs would likely occur upon ligand binding of soluble factors induced through as yet unknown receptor/mechanisms [37, 38]. The precise mechanism of their transactivation is currently unknown. A summary of signaling events along with hypothetical activation mechanisms is described in Fig. 3. Please note that this figure is not exhaustive. As an additional point of interest, the profound increase in chemokines and cytokines that occurred after suppressing FGFR alone is also worthy of follow-up studies. While it is possible that this is due to overexpression of FGFR1 in rhabdomyosarcoma cells, as documented in one study [39], nevertheless, it might reveal a novel role in B. burgdorferi-mediated pathogenic mechanisms. Finally, all of the experiments involving spirochetes were performed at an MOI of 10:1. This is the concentration that is used in in vitro experiments by most investigators in the field but it is conceivable that at lower MOIs the effects we observed might not be as marked.

Unlike previous data, the overall immune response from human oligodendrocytes, upon exposure to B. burgdorferi, results from activation of primarily three receptor tyrosine kinases, with as yet unknown activation mechanisms. Such activation also contributes to cell death processes, further underscoring the observation that inflammation in the nervous system due to B. burgdorferi results in apoptosis of neuroglial cells. The key receptors that were identified in this study may lead to the uncovering of novel pathogenic mechanisms in LNB.