In older studies using outbred animals (e.g. non-human primates and guinea pigs), where both sexes have been tested, males and females generally appear to be equally susceptible to development of EAE (54–57). Only some rabbit strains are able to develop EAE, but, in those susceptible strains, the incidence of disease is reported to be similar in males and females (36).

In inbred rat and mouse strains, however, there are many differences reported between males and females ( Table 1 – details of variables such as strain, age, antigen etc are given when this information is available). For example, adult (8-10 week old) Lewis rats of both sexes can develop EAE, but in males disease is generally monophasic, whereas females of the same age often develop relapses of disease (58–60); however, juvenile (6-week-old) male Lewis rats develop more severe relapsing EAE than juvenile females. In contrast, in adult DA rats, both males and females develop relapsing EAE with similar clinical severity; however, by histology, males show more severe inflammation, whereas females have more astrocyte activation (47). To round out the spectrum of responses, in the PVG strain of rats, males are much more susceptible than females to development of EAE and disease severity is greater in the males (63).

Similarly, in mice, there are strain-specific relationships in the effects of sex on EAE, but these can be quite complex ( Table 1 ). For example, in the SJL mouse strain with EAE induced with a variety of antigens (CNS tissue homogenates, PLP or MBP), females have generally been found to be significantly more susceptible to EAE than males (>80% vs <30%, respectively), have an earlier time of onset of disease, and have higher clinical scores compared to males (49, 52, 67). However, this female predominance in SJL mice is not seen when EAE is induced with MOG_92-106 peptide (49). In addition, the observed sex differences can be age dependent: 6-8 week adult SJL males are resistant to induction of PLP-induced EAE, however, the risk for development of EAE in males increases by 4% for each increasing week of age; in contrast, disease severity does not change in females with ageing (51, 68, 69).

It does not appear that the major histocompatibility complex (called the H-2 complex in mice) allele carried by the mice has a major role in the effects of biological sex on development of EAE, as other strains of mice carrying the H-2^s allele (same as SJL) either show no significant differences in disease incidence, clinical scores, or disease course between males and females (e.g. A.SW strain), or else both sexes are resistant to development of EAE (e.g. B10.S strain). The only other mouse strains in which females are reported to be significantly more susceptible to actively-induced EAE compared to males are (B10.S x SJL)F1 mice (H-2^s) and the NZW/LAC strain (H-2^d), although, in the latter strain, males that show clinical signs have more severe disease than females (49). In mouse strains such as C57BL/6 (H-2^b), NOD (H-2^g7), and PL/J and B10.PL (both H-2^u), the incidence of EAE in male and female mice does not differ significantly between the sexes (48–50, 70), but male PL/J and B10.PL mice show increased EAE-related mortality compared to females (49). Interestingly, a recent report suggests that a female bias in disease severity becomes more apparent in C57BL/6 mice when lower amounts of pertussis toxin are used during the induction phase of EAE, suggesting that a sex difference in the response to pertussis toxin could confound other underlying sexual dimorphism in this strain (53).

Several “spontaneous” EAE models have also been developed, in which there is transgenic expression of immunoglobulin heavy (IgH) chains and/or T cell receptors (TCR) specific for peptides of MOG on either C57BL/6, NOD, or SJL backgrounds (20, 71–73). Some of these models exhibit sexual dimorphism ( Table 2 ). Double transgenic (i.e. IgH X TCR) mice on a C57BL/6 background show no sex differences (72), however all the other models show a higher incidence of spontaneous disease in females. In spontaneous models on an SJL background, female mice are more likely to develop a relapsing-remitting type of disease, whereas males mostly develop progressive disease from the outset (20, 73).

Adoptive transfer studies have been done to try to determine if the sex of the donor of the encephalitogenic T cells affects the susceptibility to development of EAE, or if is the sex of the recipient that determines the outcome. These studies have yielded conflicting results ( Table 3 ). A confounder in the interpretation of these studies is that each has used a different mouse strain and/or antigen. The different studies variously suggest that :i) the sex of the recipient determines the differences (52: ii) that disease severity is related to the sex of the donor cells, with female cells giving greater disease, and suggesting that the sex differences play a greater role during the induction phase of the disease (74); or iii) that cells from male donors induce more severe disease, irrespective of whether they are transferred into male or female recipients (75). In vitro conditions under which the cells to be transferred are prepared may also affect the outcome, as differences in findings from adoptive transfer studies do not always reflect the sex bias seen in actively induced EAE (66).

Overall, we can conclude that some mouse strains show a female bias in disease, but this may either not occur in other strains, or be masked by sex-related effects of the antigens or adjuvants used. In most studies, females appear more likely to develop a relapsing-remitting disease course, whereas males are more likely to develop a chronic progressive disease course with greater inflammation than seen in females, which is somewhat reminiscent of what occurs in humans with MS.

When considering the effects of biological sex on the outcome of EAE, it is important to remember that the sex of an animal impacts the makeup of the immune cells and organs under physiologically normal conditions, as well as when they are pathologically challenged. This is true of humans, as well as experimental animals, where (for example) CD4⁺ cells from female healthy volunteers show a much higher levels of production of IFNγ following T cell activation with anti-CD3/anti-CD28 than do cells from males, whereas IL-17 production is higher in cells from males (76). In humans, elevated Th17 responses have been linked to male sex in some autoimmune diseases (95).

In mice, biological sex affects the numbers and activity of Th1, Th17 and Th2 T cells, regulatory T cells (Treg), and antibody levels (96, 97). However, these differences are not necessarily seen to an equal degree in all immune organs in an individual (e.g. there can be variation between the sex-related differences seen in the spleen and those seen in other secondary lymphoid organs) and they can occur in a mouse strain-specific manner (96).

Effects of biological sex on adaptive immune responses in EAE are summarized in Table 4 . Neuroantigen-specific CD4⁺ Th1 and Th17 cells are the main cells responsible for the induction of EAE. Although many papers state that Th17 cells are critical for disease development, there are quite a number showing that, at least in both the models of MOG-induced EAE in C57BL/6 mice and PLP-induced EAE in SJL mice, Th1 cells, Th17 cells, and cells producing both IFNγ and IL-17 appear to be equally able to induce EAE (76, 98–101). However, there do appear to be differences in the site of induction of antigen-specific T cells in male and female mice. For example, in PLP-induced EAE in SJL mice, lymph nodes cells from female mice produce copious amounts of both IFN-γ and IL-17 in response to a secondary challenge with the autoantigen, whereas lymph node cells from male mice produce neither of these. In contrast, spleen cells from the same mice produce predominantly IFN-γ if they are from females, but IL-17 if they are from males (76). These observations have been interpreted as providing an explanation for the ability (or not) of male mice to develop EAE, based on how much IL-17 they can produce.

However, the interpretation of these results may be more complicated. A recent study found (using adoptive transfer of MOG-specific transgenic Th17 cells from males and females into sex-matched NOD.scid recipients) that male Th17 cells are more encephalitogenic than female Th17 cells, but only if they rapidly switch (once they are transferred into the recipient) from producing IL-17 to producing large amounts of IFN-γ- cells from males that produce some IFN-γ, but also retain IL-17 production, as is seen in females, are less encephalitogenic (75). Relative levels of IFN-γ and IL-17 therefore appear to play different roles in the pathogenesis of EAE in males vs females, but much work remains to determine the exact relationship between these measures and the susceptibility of different strains of mice to EAE and how (or whether) this relates to humans with MS, in whom similar mixtures of myelin antigen-specific Th1, Th17 and mixed phenotype cells have also been reported (102).

There are several molecules that are of potential interest in these effects, including PPARγ, one of the central regulatory molecules in determining T cell fate, which selectively inhibits Th17 cell differentiation only in male cells and which modulates Th1, Th2 and Th17 differentiation in female T cells, based on the levels of estrogen (103), and Jarid1c, an X-linked immune regulator that represses severity of Th17-mediated EAE in males (75).

There also appear to be sex-based differences in the action of regulatory types of T cells in EAE. In PLP-induced EAE in SJL mice, male mice develop IL-10 producing cells that prevent relapses of disease (46) and this appears to be due to direct effects of testosterone on CD4⁺ T cells (104), whereas females have increased T cell proliferation and reduced expression of the inhibitory molecule CD152 (CTLA-4) compared to males (87). In B10.S mice, which are normally resistant to PLP-induced EAE, male mice show greater susceptibility to disease after depletion of Treg cells, but depletion of Treg cells has no effect on susceptibility of female mice (80). In MOG-induced EAE in C57BL/6 mice, where there are no sex differences in incidence of disease, there appear to be underlying sex differences in pathophysiology, as male mice have both increased inflammatory processes and increased regulatory processes, which cancel out, so that disease is similar to that in females (97). Another study using the same model found that young male mice develop less severe EAE than middle aged male mice, due to reduced numbers of CD4⁺CD25⁺ Treg cells in the middle-aged vs the younger males (64).

As with cells of the adaptive immune response, there are differences between numbers and location of dendritic cells (DC), macrophages, natural killer (NK) cells and other innate cells in male and female mice (96, 97). In actively-induced EAE studies, an important consideration when looking at innate immune responses is the effects of the adjuvants (typically complete Freund’s adjuvant and pertussis toxin) used in the induction of disease, as a recent study (97) found that there were sex differences in the induction of key innate inflammatory factors by these adjuvants alone.

Several sex-related differences in macrophages have been reported in EAE ( Table 4 ). Resistance of male SJL mice to induction of EAE is reportedly dependent on a subpopulation of macrophages (Mac-1⁺, Mac-2^-, Mac-3⁺) that play a regulatory role in the induction of encephalitogenic T cells (65). Additionally, several rodent models have associated higher levels of inducible nitric oxide synthase (iNOS) and nitric oxide (NO) production by macrophages with sexual dimorphism in EAE; however, in PVG rats, this is linked to resistance to EAE in females (63), whereas in SJL mice it is linked to increased severity of EAE in females (89). Other molecules produced or expressed by macrophages or monocytes that have been implicated in the sexual dimorphism seen in EAE include PPARα and PPARγ (78), and several Th2 cytokines, including IL-33. In addition, apolipoprotein E (apoE), which is produced by macrophages and DCs, has been reported to enhance anti-inflammatory macrophage phenotypes and downregulate Th1 and Th17 responses (105, 106). It is therefore not surprising to find that apoE deficiency in C57BL/6 mice with MOG-induced EAE leads to more severe EAE in females; however, it is surprising to find less severe EAE in apoE-deficient males (85). The mechanisms underlying these varying effects of apoE deficiency in males and females remain to be elucidated. Finally, inhibition of p38 MAPK activity in C57BL/6 mice with MOG-induced EAE ameliorates EAE in female but not male mice, and this is due specifically to affects in myeloid cells of the females, which have a greater dependence on p38 signaling for expression of proinflammatory genes than do male myeloid cells (91).

Other IL-13- and IL-33-expressing cells, including mast cells and innate lymphoid cell type 2 (ILC2) cells, are also of relevance for sex-related differences in EAE. In PLP-induced EAE in SJL mice, androgen receptor-expressing mast cells from males produce higher levels of IL-33 in response to testosterone and other inflammatory activators (79). IL-33 then activates ILC2 cells to promote and sustain a Th2 type of response (via production of IL-13) and prevent EAE. Treatment of male mice with anti-IL-33 antibody makes them more susceptible to EAE, whereas treatment of females with IL-33 leads to a reduction in EAE severity (79). Conversely, decreased accumulation of IL-13-secreting ILC2 cells in the brains of male, but not female, SJL–c-Kit mutant mice leads to increased susceptibility to development of EAE in the male mice, but resistance to the development of EAE in females (77), via a switch from a Th2- to a Th17-dominated T cell response that is regulated by the presence or absence of mature ILC2 cells in the draining lymph nodes and the CNS. In C57BL/6 mice, knockout of IL-13 reduces the incidence and severity of MOG-induced EAE in females but not in males (50).

Within the CNS, microglia and astrocytes are the main cell types thought to be involved in innate responses. They differ regionally and numerically between male and female mice (107–111) and there is sexual dimorphism in their responses to CNS assault in adult mice (107, 112). For microglia, these can include differences in inflammatory sensitivity and reactivity, their ability to affect cellular repair, and in their response to oxidative stress (113, 114). Studies in which microglia have been depleted from brains of mice using the CSF1R inhibitor, PLX5622, result in adverse sex-specific behavioral effects, with females developing long-term hyperactivity and anxiolytic-like behaviour (115). Astrocytes display sex differences in uptake of glutamate and their responses to fatty acids, and are known to express the estrogen receptor alpha (ER-α) (116–118). There are also sex differences in mitochondrial bioenergetics of astrocytes, but not microglia (119).

In EAE studies, some sex-specific effects on or by these CNS cell types have also been reported ( Table 4 ). In microglia, for example, iNOS can be induced by MBP-primed T cells from female and castrated male SJL mice, but not from intact males, and this is dependent on activation of a CCAAT-enhancer-binding protein (C/EBPβ) in the microglia (84). In MOG-induced EAE in microglial-depleted C57BL/6 mice, re-engraftment of microglia exacerbates disease in females but not males (83). Astrocytes are also involved in the pathology of EAE (120) and sex differences in the astrocyte response during EAE have been described. For example, there is expression of complement by astrocytes in optic nerves of C57BL/6 mice with MOG-induced EAE, and this is greater in females than males and correlates with worse retinal ganglion cell and axonal loss in females (82). Similarly, transgenic expression in astrocytes of C57BL/6 mice of the RNA regulator HuR, which plays a major role in regulating production of cytokine and growth factors, results in attenuation of MOG-induced EAE in females, but not males (81). Furthermore, this effect can be reversed by ovariectomy.

One molecule that is widely expressed in both peripheral and CNS innate immune cells, and which has previously been implicated in both MS and EAE, is cystatin C. However, cystatin C appears to have different functions in immune and CNS cells, and there has been much debate on whether it plays a protective or pathogenic role in MS and EAE. In MOG-induced EAE in C57BL/6 mice, deficiency of Cst3, the gene encoding cystatin C, leads to improvement of EAE in females, but has no effect in males. Furthermore, the sex-dependent effects of cystatin C are negated by ovariectomy of females, and are revealed by treatment of males with female hormones (88).

Immunometabolism, the metabolic reprogramming of anaerobic glycolysis, oxidative phosphorylation and metabolite synthesis that occurs when immune cells are activated, is an important regulator of immune responses and immune homeostasis. Immunometabolic dysfunction has been linked to an inflammatory phenotype of immune cells and has been implicated in development of autoimmune diseases (121), including EAE (122). There is an interaction of immunometabolism with sex, especially through the actions of reproductive hormones (123). Male DA rats with EAE have more severe disease and greater oxidative stress than females (90), consistent with studies that show sex differences in levels of pro-oxidant and anti-oxidant enzymes (124, 125). It has also been found that high sodium (126), high fat (127), or high carbohydrate (128) diets can all exacerbate EAE. In SJL mice fed a high sodium diet from prior to induction of PLP-induced EAE, there is an increase in disease severity in female mice, whereas male mice are not affected (129). In contrast, compared to mice fed a normal diet, both male and female C57BL/6 mice with MOG-induced EAE and fed a high sodium diet show some increase in disease severity, particularly during the peak of the acute disease (129). In vitro studies have suggested that the effects of high levels of salt appear to be due to stabilization of the Th17 cell phenotype, via the activation of several pathways, including p38 MAPK, nuclear factor of activated T cells 5 (NFAT5), and the salt-sensing serum glucocorticoid kinase 1 (SGK1) (126, 130). However, a high salt diet has also been reported to elevate serum sodium, leading to increased blood-brain barrier permeability and more severe brain pathology (129). Additionally, it cannot be ruled out of these studies that different effects of high salt diets in mice of different sexes may relate to changes to the gut microbiome. The gut microbiome appears to play an important role in both EAE and MS (131). Sex hormones influence the composition of the gut microbiota (132) and its interaction with the brain (“the gut-brain axis”) (133). In MOG-induced EAE in C57BL/6 mice, protection due to estrogen is associated with changes in the gut microbiota and the gut lymphoid tissue (134). In the same model, oral administration of 2.6% alcohol for 3 weeks prior to EAE induction is protective in both sexes, but the protective effect is greater in males, and also appears to be mediated though changes in the gut microbiota (135).

Several therapeutic interventions also show sex-specific effects. Supplementation with Vitamin D3 protects against EAE in female but not in male mice of several strains (94, 136), and this effect appears to be due to sex differences in vitamin D3 metabolism, with females having an enhanced ability to metabolize vitamin D3 to the active hormone 1,25-(OH)₂D3 (94). Ovariectomy abrogates this protective effect (94). Other therapeutic approaches show more of a male-specific effect, e.g. oral tolerance can be induced in male, but not female, B10.PL mice fed with an altered peptide ligand of MBP (137).

In several autoimmune diseases it is recognized that the severity of disease can be influenced by the resistance of the target tissue to damage (138, 139). In EAE and MS, target organ resistance could involve variability in the processes that lead to damage to the myelin-producing oligodendrocytes and neurons, and in the processes of repair. Since myelin is the target of immune attack in EAE and MS, several studies have investigated the effects of biological sex on oligodendrocytes. These studies have shown that there is sexual dimorphism in oligodendrocytes’ capacity for myelination and their responses to stress (140, 141). There are reports of sex differences in expression of mRNA for MBP and PDGF alpha receptor (markers of myelination) in DA rats with EAE (47), and markers of axonal injury and dendritic aborisation are greater in male than female C57BL/6 mice with MOG-induced EAE (86). However, it is not clear at present if these findings merely reflect the strength of the immune response in the animals of different sexes in these models or if they truly reflect sex differences in target organ resistance in EAE. Adoptive transfer studies of male and female cells (not only of peripheral immune cells into naïve recipients, but also of microglia into microglia-depleted recipients) could help to answer this question. This is a topic that will likely be difficult to decipher, but one that may prove very useful in ultimately answering some important questions related to the differential prognosis of males and females with MS.

Other molecules that are expressed in the CNS can also influence sex differences in other pathophysiological mechanisms in EAE. For example, sphingosine-1-phosphate receptor 2 (S1PR2) is a molecule involved in regulation of the integrity of the blood-brain barrier (BBB). Naïve female SJL mice have significantly elevated levels of S1PR2 compared to SJL males and to C57BL/6 mice of both sexes (92). Induction of PLP-induced EAE in the SJL mice further increases expression of S1PR2 and causes increased vascular permeability in CNS tissue from females, but not males. Furthermore, S1PR2-deficient mice showed decreased EAE severity compared with wild-type controls (92). Interestingly, however, adoptive transfer of CD4 ⁺ cells from TCR1640 transgenic mice (MOG _92-106 -specific cells on SJL background) into naïve male or female SJL recipients show no sex differences in BBB integrity (73).

It may even be that the sex of the experimenter can affect outcomes. Whilst this has not been systematically investigated in EAE, as far as we are aware, there are several reports that rodents of both sexes are more stressed when the person doing the experiments on them is male, and that this can lead to differences in behavioral assays of pain and stress (142, 143). This could potentially add another layer of complexity to investigation of the effects of biological sex on the pathophysiology of EAE.

Overall, biological sex can affect many of the elements of immune function and CNS biology that are required for development of EAE. Sex differences could occur during the induction of disease, where there could be sex differences in response to adjuvant and in antigen presentation or in the effector stage of disease when there could be sex differences in the pathological processes in the CNS. To clarify this, appropriate controls (naïve and adjuvant only) of both sexes need to be included in all experiments looking at mechanisms underlying sex differences in EAE. Much more work remains to understand the basic biological differences between the sexes, both in mice and humans.

The sex chromosomes, X and Y, are the only chromosomes in the body that are represented differently in females and males, and they vary greatly (144). Thus, effects of sex could be due differences in the sex chromosomes and/or sex-related differences in gene expression.

The Y chromosome contains only 50 to 60 unique protein-coding genes that encode primarily for molecules required for gonadal formation and fertility, including the Sry gene, which is necessary for the development of male gonads (145) and the Azf (azoospermia factor) locus that has a role in organs outside the testis, including the brain (146), and in inflammation (147, 148). Variable amounts of gene replication occur on the Y chromosome from different species, so that the Y chromosomes from humans and mice are quite different (149, 150). Even within mouse strains differences can be observed: e.g. some strains have a Y chromosome gene known as Yaa (y accelerated autoimmunity) which has resulted from a translocation of the gene encoding Toll-like receptor 7 (Tlr7) from the X chromosome onto the Y chromosome. Yaa can modulate a mouse’s potential to develop autoimmunity via epistatic interactions with other autoimmunity-related genes (151) (152).

In contrast to the Y chromosome, the X chromosome contains over 900 genes, and is highly conserved across species. Genes on the human X chromosome include genes for reproduction and a disproportionately large number that code for brain function, such as the BEX family (brain expressed X-linked genes), NF2 (neurofibromin 2), GDNF (glial cell derived neurotrophic factor) and PLP1 (myelin proteolipid protein, the most abundant protein in CNS myelin, an autoantigen for EAE, and a target of immune attack in MS (153, 154)). The X chromosome also contains genes for the testis that are subject to selection pressure because they confer advantage in males (155). There are also many genes of relevance to immune function and susceptibility for autoimmunity located on the X chromosome, including those encoding IL2-Rγ, IL13-Rα, CXCR3, CD40 ligand, TLR7 and TLR8, TIMP1 and Bruton’s tyrosine kinase (BTK).

In males, the X chromosome is always of maternal origin. In females, one X chromosome is randomly inactivated in each cell of an early-stage embryo, so that the embryo is a mosaic of cells with either an active paternal X chromosome or an active maternal X chromosome. All descendants of that cell will have the same active X chromosome, and most females have an approximately even spread of paternally- and maternally-derived active X chromosomes (156). However, this inactivation is not always complete and can differ in different tissues and cell types from a single individual (157), potentially leading to X dosage effects, with higher expression of X-linked genes in females compared to males. DNA methylation plays an important role in X chromosome inactivation and recent studies have reported that maternally derived X chromosomes have less DNA methylation than paternally-derived X chromosomes, and that therefore there is greater expression of X-linked genes in male cells (158). These factors could have implication for sex differences in autoimmunity.

The development of a mouse model known as the “4 core genotypes” (FCG) model, in which the sex chromosome complement is unrelated to the animal’s gonadal sex (159), has allowed studies of the effects of the sex chromosome complement on susceptibility to EAE. Early studies compared gonadally-intact FCG mice on a SJL background, (SJL-FCG mice) with FCG mice in which the gonads were removed. These studies show that the presence of XY (male) complement of sex chromosomes leads to greater immune responses to MBP than XX chromosomes, whereas the presence of male sex gonads (and therefore sex hormones) inhibits immune responses to MBP (160). In that same model, when gonads are removed, mice carrying the XX genotype are more susceptible to PLP-induced EAE than are mice carrying a XY genotype, thus demonstrating an autoimmune-enhancing effect of the XX genotype independent from hormonal influences (161). Whether this outcome could be due to dosage effects of the target antigen, PLP, which is encoded on the X chromosome, was not investigated. Bone marrow chimeric SJL-FCG mice, in which the mice have either a XX or XY sex complement in the CNS and an immune system with a common sex chromosomal type, demonstrate that the presence of an XY sex complement in the CNS associates with greater clinical and neuropathological severity of EAE, apparently due to higher expression of TLR7 in the CNS of XY vs XX mice (93). FCG mice on a C57BL/6 background and immunized with MOG show parent of origin effects of DNA methylation of a set of five X chromosome genes in autoantigen-stimulated T cells and B cells, with higher expression in XY than XX genotypes. SJL-FCG mice immunized with PLP also show similar effects, but the X-linked genes affected differ from those seen in the C57BL/6-FCG mice, suggesting that parent of origin effects are not strain specific (158).

The Y chromosome also has an effect on immune cell gene transcription (162) and copy number variation in Y chromosome genes plays a role in paternal transmission of autoimmune disease (163). There is an effect of Y chromosome genes on the incidence and severity of EAE, since studies using progeny of C57BL/6 Y-chromosome substitution strains show that the Y chromosome influences the severity of EAE in both male and female mice (164).

Given the differences between the ability of male SJL and C57BL/6 mice to develop EAE, it is of also of interest to note that the Y chromosome in SJL mice has been inherited from the Mus musculus domesticus subspecies, whereas the Y chromosome in C57BL/6 and most other commonly used inbred laboratory mouse strains comes from Mus musculus musculus (165). Y chromosome substitution (Y consomic) SJL male mice carrying a Mus musculus musculus Y chromosome are more susceptible to EAE than are age-matched SJL males, suggesting that a Y chromosome-linked polymorphism controls the sexual dimorphism to development of EAE seen in SJL mice (68).

It is of interest to note that, as well as genes on the sex chromosomes, there are other sex-specific genetic differences. A study of crosses between EAE-susceptible SJL mice and EAE-resistant B10.S mice identified sex-specific loci that governed susceptibility of SJL mice to relapsing EAE (eae12) and monophasic EAE (eae7 and eae13) (69). In a large study of 26 chromosome substitution (consomic) strains, in which mice inherited chromosomes from the EAE-resistant wild-derived PWD mouse strain on a C57BL/6 background, 19 genetic loci on 18 chromosomes were identified that influenced susceptibility to MOG-induced EAE: of these, six were male-specific and four were female specific (166).

Males and females also differ in the levels of circulating sex hormones, which change over the lifetime. In females this is especially true during pregnancy, when there are increased levels of many different hormones and also increased levels of mediators related to pregnancy (see section 6 for more on pregnancy-related effects). The effects of physiological levels of the major sex hormones could explain some of the observed effects of biological sex in EAE, as shown through comparisons of intact and gonadectomized mice (167, 168). The effects of sex hormones on EAE could be due to immune modulation and/or neuroprotection, as summarized in Table 6 . In addition, it appears that there could be bi-directional effects of sex hormones and the pro-inflammatory molecules produced during EAE (178), as SJL mice with PLP-induced EAE have a significantly shorter estrous cycle (normally around 4-6 days (179)) than mice just immunized with adjuvant. The effects of the major sex hormones are summarized below.

There are effects of both the sex chromosomes and the sex hormones on sexual dimorphism in EAE, but these are not always clear-cut and vary from one EAE model to another. The Y chromosome appears to exert an influence on EAE in both female and male mice through paternal transmission of autoimmune susceptibility. The X chromosome also carries many genes relevant to development of EAE, and X dosage effects may occur. The female sex hormones show fairly subtle effects at physiological levels, but are highly immunosuppressive at higher levels in most models tested. In contrast, male sex hormones appear to have stronger effects at physiological levels.

The effects of pregnancy on EAE are summarized in Table 8 . In general, pregnancy reduces the incidence and severity of disease in all models at some time point; however, the specific details of when the effects were most notable differ, depending on the timing of EAE induction vs the beginning of pregnancy.

In guinea pigs, which have a much longer gestation period (63 days) than mice and rats (~21 days), induction of EAE at any timepoint during pregnancy delays the onset of disease and reduces the severity of EAE compared to induction of disease in non-pregnant animals (256). In rabbits (31-32 days gestation), pregnancy suppresses EAE, with the effects most noticeable when EAE is induced during mid to late pregnancy (day 20-25) (259). Similarly, Lewis rats with EAE induced during the third weeks of pregnancy show the greatest level of protection against the development of disease; rats inoculated in the first week or second week of pregnancy are less protected (256, 257). Interestingly, however, when those rats are re-challenged again 3 weeks after weaning of their litters rats that were protected in pregnancy are more susceptible to the second encephalitogenic challenge, suggesting either that the protective effects during pregnancy are not antigen specific, or that they are not very long-lasting (257). In DA rats, EAE is less severe when disease is induced during the second week of pregnancy, although the rats develop more severe relapses post-partum (258).

In various mouse models, active induction of EAE during late pregnancy leads to a reduction in the incidence of EAE and/or a decrease in clinical severity, compared to virgin controls (261, 262). In contrast, adoptive transfer of activated MBP-specific T cells into pregnant mice during mid-pregnancy, but not in late pregnancy, resulted in a delayed onset of EAE and reduced clinical severity in comparison to virgin controls (264).

The effects of becoming pregnant after the onset of EAE have also been investigated. In both SJL mice with PLP-induced EAE and in C57BL/6 mice with MOG-induced EAE, onset of pregnancy leads to clinical improvement of EAE, but the pathological lesions persist (263). In mice followed until after delivery, EAE scores in the SJL mice return to the same level as in virgin controls, whereas C57BL/6 mice continue to have milder disease than the virgin controls (263).

There have been attempts to transfer the protective effects of pregnancy. Offspring from rats that had EAE during pregnancy show transient protection from EAE, and rats suckled by mothers that had EAE in pregnancy also acquire protection (265).

Several potential mechanisms have been put forward to explain the beneficial effects of pregnancy on EAE. There are alterations of cytokine production (increased expression of mRNA for IL-4 and IL-10) in the spinal cords of pregnant compared to non-pregnant Lewis rats with EAE induced one day after mating, and serum from pregnant rats suppresses lymphocyte proliferation (260, 261). A suppressive serum factor found in late pregnancy has also been implicated in the amelioration of EAE in SJL mice when EAE is induced during the second or third week of pregnancy. Mice with EAE induced during the first week of pregnancy are not protected, suggesting that the serum factor acts on the induction and early effector phases of EAE (261).

Exosomes from both pregnant and non-pregnant mice suppress EAE. However, exosomes are more abundant in pregnancy, and it is suggested that they contribute to protection during pregnancy by suppressing T cell activation and promoting maturation of oligodendrocyte precursor cells and their migration into CNS lesions (266).

Other studies have shown that pregnancy does not appear to affect the induction of myelin-specific T cells, and that there is actually an increase in numbers of such cells in the spleen compared to non-pregnant mice; however, these T cells produce less TNF and IL-17 than those from non-pregnant mice, and there is a higher frequency of IL-10-secreting cells within the CD11b⁺, CD11c⁺, CD19⁺, and CD4⁺/CD25⁺ Treg populations (262). Treg cells bearing TCR clonotypes related to the encephalitogenic TCR are increased in pregnant mice in which EAE has been induced on day 7 of pregnancy, whereas such cells are not seen in virgin controls (267). These results suggest that when an antigen is introduced during pregnancy, an immunoregulatory environment towards that antigen predominates.

Although pregnancy is a state when there are circulating growth factors, and although estrogen and progesterone are neuroprotective, there is little information about the effects of pregnancy on neuroprotection and remyelination in EAE.

Several other factors that could contibute to the suppression of EAE by pregnancy have been identified, but only a limited number of studies has been done on each of them. They include:

Overall, pregnancy (particularly the latter part) is a highly immunosuppressive state in almost every EAE model, and studies of pregnancy across different models show more consistent results than do other studies of the effects of sexual dimorphism in EAE. The effects of pregnancy appear to be driven largely, but not solely by estrogen levels, but there are also numerous other pregnancy-related molecules that have been identified that likely contribute to the overall effects of pregnancy. Further study of these molecules is warranted.