In the present study, we demonstrate a potential role of the Gas6/TAM receptor system in the loss of white matter integrity in MS patients, using CSF biomarker measurements of proteins related to these pathways. Essentially all Gas6/TAM biomarkers were higher in PMS compared to HC, and Tyro3, Axl, and Gas6 were higher in PMS vs. RRMS. In multiple linear regression models, baseline CSF concentrations of soluble Tyro3 and Gas6 were associated with greater reduction in WM and MyC at 60 months compared to baseline. In addition, RRMS patients who exhibited higher MyC at 60 months compared to baseline had lower baseline Tyro3 and Gas6, signifying an association with ineffective remyelination. Moreover, we found correlations between concentrations of Tyro3 and Gas6, and biomarkers of neuro-axonal damage (CSF NfL) and astrocyte activation (serum and CSF GFAP).

The association with GFAP was particularly consistent, further emphasizing findings from animal models [28] and fresh-frozen human MS brain tissue [29], in which the Gas6/TAM system was associated with axonal damage, inferior ability of remyelination and higher degree of neurodegeneration. GFAP is a marker of astrogliosis that has previously been proposed as a reliable biomarker for disease severity in MS [30, 31]. Further, sGFAP reflects astrocytic damage in patients with primary progressive MS [32], emphasizing its potential role in reflecting neurodegenerative processes in MS. Moreover, sGFAP has been demonstrated to associate with microstructural damage in normally-appearing white matter, as assessed by diffusion tensor imaging [33]. Recently, sGFAP has been shown to associate with progression independent of relapse activity in MS [34, 35]. Hence, the association of Tyro3 and Gas6 with CSF and serum GFAP may reinforce the notion of cross-talk between astrocytes, oligodendrocytes, and microglia in driving the chronic neurodegenerative processes (myelin damage and failure of repair) that characterize progressive MS.

Tyro3 has been previously implicated in the regulation of myelination in the CNS [13]. Knock-out mouse models have shown that loss of Tyro3 leads to delayed myelination and altered myelin thickness in the CNS [13]. Gas6 exhibits binding affinity and activation capabilities across all three TAM receptors; however, the heightened expression of Tyro3 on oligodendrocytes designates this receptor as the primary candidate for mediating the pro-myelinating effects induced by Gas6. It has been previously demonstrated that the absence of Tyro3 results in the abrogation of Gas6's pro-myelinating impact, leading to a delay in developmental myelination and the production of thinner-than-normal myelin. Notably, this effect is confined to the myelination process and is not attributable to alterations in the proliferation or differentiation of oligodendrocyte precursor cells. The loss of Gas6 has been demonstrated to be associated with delayed remyelination after a demyelinating injury induced by cuprizone [15].

We demonstrate an association of Tyro3, Gas6, and Mer with higher WM volume at baseline, and Gas6 was weakly associated with baseline MyC as well. However, this association was lost at 12 months follow-up, and baseline concentrations of Tyro3 and Gas6 were instead associated with greater reduction in WM volume and MyC at 60 months follow-up compared to baseline. Hypothetically, this may be explained neuroanatomically by greater concentrations of oligodendrocytes, astrocytes, and microglia in areas with larger WM volume and MyC, accounting for higher concentrations of soluble Gas6 and TAM receptors that are being released into the extracellular space. However, over time, this association is reversed, and higher baseline soluble Tyro3 and Gas6 are rather associated with greater reduction in WM and MyC, probably due to defective remyelination, partly because of TAM receptor inactivity.

Gas6-Tyro3 interactions have been demonstrated to be important in regulating remyelination and myelin sheath growth [13]. Recently, it was demonstrated that the promyelinating effects of Gas6 are dependent on the presence of Tyro3 [36]. It is thus not surprising that it is specifically soluble Tyro3 and the ligand Gas6 that particularly stand out in our analysis, and that both associate with loss of myelin and white matter volume as determined by quantitative MRI.

We found higher levels of Tyro3 and Gas6 in patients with progressive MS, implying that the Gas6/TAM receptor system may play a role in the degenerative processes driving progressive MS. Since Gas6/TAM did not correlate with NfL at baseline, it is likely that the Gas6/TAM system is not involved in the acute focal inflammatory neuroaxonal injury in MS, but rather involved in other neurodegenerative processes, related to ongoing demyelination and failure of remyelination. The lack of association between Gas6/TAM concentrations and new/enlarging T2 and contrast-enhancing lesions also reinforces this notion.

We found an association between baseline and 12 months Tyro3 concentrations and greater change (reduction) in PASAT score, indicating a possible role for Tyro3 in (re)myelination of fibres important for auditory information processing speed and ability. This preliminary finding encourages further explorations into the interaction of the Gas6/TAM receptor system and cognitive decline in MS.

However, our analysis did not reveal any convincing influence of DMT on Gas6/TAM concentrations at 12 months follow-up, although the levels of Mer were indeed reduced on a group level. Moreover, Gas6/TAM did not show a treatment response when dichotomizing the cohort according to NEDA-3 criteria or when comparing low- and high efficacy DMT treatment. However, we cannot rule out that the effect of treatment on Gas6/TAM is considerably slower and takes place over longer periods of time.

Most of the evidence on the role of the Gas6/TAM system is provided from experimental animal models of MS. To date, there has been limited replication of the aforementioned findings in individuals with MS. One study involving autopsy material from MS patients revealed an up-regulation of Axl and Mer in homogenates derived from chronic silent and chronic active lesions, respectively. [11] Additional insights into the association of Gas6/TAM with MS in humans have surfaced through genome-wide association studies (GWAS). Numerous single nucleotide polymorphisms (SNPs) situated within the MerTK gene have been identified as linked to susceptibility to MS [37, 38]. A distinctive SNP of the MerTK gene, denoted as rs7422195, was reported, which displays a discordant association with MS contingent upon the human leukocyte antigen (HLA)-DRB1*15:01 status [39]. Notably, rs7422195 demonstrates a protective effect in DR15 homozygosity, while exacerbating the disease in the absence of DR15. Furthermore, the minor allele of rs7422195 is correlated with an augmented gene and protein expression of MerTK in monocytes and CD4 + cells. More recently, this finding was confirmed [40], showing that DR15 and MerTK genotype independently influence proportions of CD14 + MERTK + monocytes in MS.

One study has previously investigated concentrations of Gas6/TAM in body fluids from MS patients [41], although this study assessed CSF and plasma concentrations of only Gas6. Findings from this study indicate that MS patients do not exhibit substantial alterations in plasma Gas6 concentration compared to controls. However, a dissociation between CSF and plasma was observed, wherein CSF Gas6 levels were higher in MS patients than in those with other non-inflammatory neurological diseases. Intriguingly, individuals experiencing more severe or prolonged relapses demonstrated lower CSF Gas6 concentrations, comparable to controls. In contrast, those with briefer and milder relapses displayed higher concentrations (almost twofold). Notably, CSF Gas6 concentration did not vary in relation to the completeness of recovery. Furthermore, neither plasma nor CSF Gas6 demonstrated associations with relapse rates or EDSS progression in a follow-up cohort. This is similar to a potentially protective role of increased serum TAMs in Alzheimer’s disease [42]. Patients with higher serum Gas6/TAM levels also demonstrate stronger immune regulation exerted by the TAM receptor system and promotion of phagocytosis and cell survival, resulting in preserved brain structure and delayed cognitive decline [43]. A very recently study published demonstrated a weak association between higher baseline CSF Gas6 and EDSS < 3 at diagnosis, but no association with follow-up EDSS, and a correlation between serum Gas6 with lower MS severity score [44]. These findings further emphasize the initial protective role of the Gas6/TAM system in the short term. Further, that study did not find differences in CSF concentrations of Gas6/TAM biomarkers when comparing patients on high-efficacy, low-efficacy, or no DMT, although patients on high-efficacy DMT exhibited somewhat lower serum levels of Axl. However, this analysis was cross-sectional and did not compare concentrations at baseline and follow-up (after DMT) in the same patients, as has been done in our study.

Our investigation has several limitations. This was a proof-of-concept study with a relatively small sample size, which limits statistical power and generalizability. Notwithstanding, we still could find meaningful differences in outcomes that should be validated in larger cohorts. The sampling before and after treatment initiation is only 12 months apart. However, the primary focus of the study was not the evaluation of treatment response nor the effect of DMT on biomarkers. Nonetheless, the follow-up time for MyC and WM was 60 months, which is sufficient to capture meaningful changes. Second, all our patients with RRMS are newly diagnosed, treatment naïve and have short disease duration at baseline. It may be valuable to investigate Gas6/TAM biomarkers in RRMS patients with longer disease duration in future studies to examine if there is increasing/decreasing trend in Gas6/TAM during disease course and to include treatments that could potentially have remyelinating effects. Further, the volumetric measures were performed after contrast, which is known to affect their values. However, the exact same protocol was used throughout the study and therefore this should not change the main results. Even though different dilution factors have been used for the different biomarkers, which might lead to quantitative proportional differences, this is not expected to influence the results of this study, as biomarker concentrations were not compared with each other.

In conclusion, we provide much needed evidence of CSF concentrations of Gas6, Tyro3, Axl, and Mer in people with MS and HC. Our findings underscore the potential role of Gas6/TAM receptor system, particularly Tyro3 and Gas6, in enhancing remyelination within the CNS in patients with early MS.