Macrophage CD40 signaling drives experimental autoimmune encephalomyelitis

Abstract The costimulatory CD40L–CD40 dyad plays a major role in multiple sclerosis (MS). CD40 is highly expressed on MHCII+ B cells, dendritic cells and macrophages in human MS lesions. Here we investigated the role of the CD40 downstream signaling intermediates TNF receptor‐associated factor 2 (TRAF2) and TRAF6 in MHCII+ cells in experimental autoimmune encephalomyelitis (EAE). Both MHCII–CD40–Traf2 −/− and MHCII–CD40–Traf6 −/− mice showed a reduction in clinical signs of EAE and prevented demyelination. However, only MHCII–CD40–Traf6 −/− mice displayed a decrease in myeloid and lymphoid cell infiltration into the CNS that was accompanied by reduced levels of TNF‐α, IL‐6 and IFN‐γ. As CD40–TRAF6 interactions predominantly occur in macrophages, we subjected CD40 flfl LysM cre mice to EAE. This myeloid‐specific deletion of CD40 resulted in a significant reduction in EAE severity, reduced CNS inflammation and demyelination. In conclusion, the CD40–TRAF6 signaling pathway in MHCII+ cells plays a key role in neuroinflammation and demyelination during EAE. Concomitant with the fact that CD40–TRAF6 interactions are predominant in macrophages, depletion of myeloid CD40 also reduces neuroinflammation. CD40–TRAF6 interactions thus represent a promising therapeutic target for MS. © 2018 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of Pathological Society of Great Britain and Ireland.


Introduction
Multiple sclerosis (MS) is a chronic inflammatory disease of the CNS, characterized by a breakdown of the blood-brain barrier, multifocal inflammatory lesions, demyelination, reactive gliosis, axonal degeneration and neurodegeneration [1,2]. MS affects more than two million people worldwide and is one of the most common causes of neurological disability in young adults [2,3]. The exact etiology of MS is unknown, although it is considered to be an autoimmune disease in which lymphocytes programmed to recognize myelin antigen trigger a cascade of events resulting in the formation of acute inflammatory, demyelinating lesions in myelin sheaths in the white matter areas of the CNS [4].
The costimulatory CD40-CD40L dyad has a critical role in driving immune responses and plays a key role in several chronic inflammatory diseases, such as atherosclerosis, obesity, rheumatoid arthritis and MS [5]. Both CD40l −/− and CD40 −/− mice are protected against experimental autoimmune encephalomyelitis (EAE) [6,7]. CD40 −/− mice have reduced T cell priming against myelin oligodendrocyte glycoprotein (MOG) and these T cells produce less IFN-γ [8]. Furthermore, CD40 deficiency inhibits leukocyte infiltration into the CNS [8]. CD40 is present in MS on different cell types, including B cells, monocytes, macrophages, endothelial cells, T cells and CNS-resident cells [9,10]. Activation of these cells by CD40L results in the secretion of proinflammatory cytokines and chemokines [11,12], and T cell expansion [13], which promotes the ongoing inflammation in the CNS. Antibody-mediated blockade of CD40L blocks CNS inflammation and disease progression in mice with relapsing EAE [14,15]. Accordingly, CD40 antagonists repressed EAE onset and disease severity in marmoset monkeys [16]. Fueled by the success of these experimental studies, clinical trials were initiated. The treatment of MS patients with the anti-CD40L mAb IDEC-131 was successful in a pilot study, although the subsequent phase II trial was halted due to the occurrence of a case of severe thromboembolism in a parallel Crohn's disease trial that evaluated IDEC-131 [17]. Thus, blockade of CD40L is not a feasible therapeutic strategy, as it can result in thromboembolic side-effects due to the disruption of αIIbβ3-CD40L interactions in arterial thrombi [18,19]. Moreover, complete inhibition of the CD40-CD40L dyad may result in immunosuppression [20]. However, modulation of CD40 downstream signaling pathways, preferably in a cell type-specific manner, may offer a strategy to reduce the risk for these side-effects [19,20]. Upon binding of CD40L, CD40 recruits TNF receptor-associated factors (TRAFs) to exert signaling to downstream pathways [20,21]. The intracellular domain of CD40 contains a distal binding domain for TRAF2/3/5 and a proximal domain for TRAF6 [20,21]. Which signaling pathway is activated depends on the cell type and the cellular environment. Here we unravel the role of the CD40-TRAF2 and the CD40-TRAF6 signaling pathways in neuroinflammation by subjecting mice with site-directed mutagenesis for the TRAF6 or TRAF2/3/5 binding site on the CD40 intracellular tail in MHCII + cells to EAE. We found that both signaling pathways play an important, but distinct, role in EAE disease development, with a prominent role for CD40-TRAF6 signaling in neuroinflammation. As CD40-TRAF6 signaling is especially important in the macrophage [22,23], we investigated the role of myeloid-specific CD40 signaling in EAE, using CD40 flfl LysM cre mice and found that macrophage CD40 is indeed an important driver of EAE.

Materials and methods
Mice CD40 flfl mice were generated by custom design at Ozgene Pty Ltd (Bentley, WA, Australia). A conditional allele was created by inserting loxP sites upstream of exon 2 and downstream of exon 3. Mice carrying the floxed CD40 gene were crossed with LysM cre mice [24] to obtain myeloid-specific CD40-deficient mice (CD40 flfl LysM cre mice). Littermates not containing the LysMcre driver were used as controls (CD40 flfl or WT mice). CD40 flfl LysM cre , CD40 flfl , MHCII-CD40-T2/3/5 −/− , MHCII-CD40-T6 −/− and CD40 −/− mice were bred and maintained at the animal facility of the Academic Medical Centre (Amsterdam, the Netherlands). C57Bl6 mice were used as a WT control in the TRAF signaling EAE experiment and purchased from Charles River (Lyon, France). All the experimental procedures were approved by the Ethical Committee for Animal Experiments of the Academic Medical Centre, Amsterdam, the Netherlands (AMC).

EAE induction
To investigate the effects of macrophage CD40 and the CD40-TRAF pathways in EAE, 12-week-old female A control group was included that received only CFA and pertussis toxin (n = 6). Body weight and neurological symptoms were monitored daily. The symptoms were graded using the following scale: 0 = no neurological abnormalities; 0.5 = partial loss of tail tonus; 1 = complete loss of tail tonus; 2 = hind limb paresis; 3 = partial hind limb paralysis; 4 = complete hind limb paralysis; 4.5 = paralysis up to the diaphragm, 5 = death. Scoring of clinical symptoms was performed by an observer who was blinded for the experimental conditions.

Western blot
CD40 flfl LysM cre and CD40 flfl BMDMs were cultured as described above, treated for 10 min or 24 h with FGK45 (30 μg/ml), IFN-γ (5 ng/ml) and LPS (100 ng/ml) and subsequently lysed in RIPA lysis buffer. Protein concentrations were determined using the Pierce BCA Protein Assay Kit (Thermo Scientific, Waltham, MA, USA). Equal amounts of protein samples were loaded on to a SDS polyacrylamide gel and transferred to a nitrocellulose membrane (Bio-Rad). After blocking with 5% BSA in PBS containing 0.05% Tween-20, blots were incubated overnight with anti-CD40 (Abcam, Cambridge, UK) and anti-α-tubulin (Cedarlane, Burlington, ON, Canada). Blots were then washed and incubated with the appropriate HRP-conjugated secondary antibody (1:800; Thermo Scientific) and visualized using the SuperSignal West Pico Plus Chemiluminescent Substrate kit (Thermo Scientific).

RNA isolation and gene expression analysis
Total RNA from the CNS was extracted using TRIzol (Invitrogen, Carlsbad, CA, USA) and RNA from BMDMs was isolated using High Pure RNA Isolation Kits (Roche, Basel, Switzerland). RNA was reverse transcribed using an iScript cDNA synthesis kit (Bio-Rad) and quantitative (q)PCR was performed using a SYBR Green PCR Kit (Applied Biosystems, Leusden, the Netherlands) on a ViiA7 real-time PCR system (Applied Biosystems). Expression levels of transcripts obtained using RT-qPCR were normalized to the mean expression levels of the reference transcripts Gapdh and Rplp0. Primer sequences can be found in supplementary material, Table S1.

Statistical analysis
Results are presented as mean ± SEM. Calculations were performed using GraphPad Prism 7.0 software (GraphPad Software Inc., San Diego, CA, USA). Statistical significance was evaluated using unpaired Student's t-tests. Clinical parameters were analyzed by nonparametric (Mann-Whitney) test. P values < 0.05 were considered statistically significant. n represents the number of biological replicates.

CD40-and CD40-TRAF6-deficient mice are protected against neuroinflammation
Analysis of the CNS (cerebellum and spinal cord) shows that the number of total leukocytes (CD45 +

cells), macrophages (Mac3 + cells) and T cells (CD3 + cells) that infiltrated the CNS was significantly reduced
in MHCII-CD40-T6 −/− and CD40 −/− mice compared with WT mice subjected to EAE ( Figure 3A,B). Although MHCII-CD40-T2/3/5 −/− mice exhibited significantly milder EAE disease symptoms than WT mice, the amount of immune cell infiltration in the CNS did not differ ( Figure 3A,B). Moreover, although not significant, the number of CD45 + cells in the spinal cord of WT versus MHCII-CD40-TRAF2/3/5 −/− mice even seemed to be increased ( Figure 3A,B), whereas clinical symptoms were ameliorated. One of the possible reasons why this increase in CD45 + cells did not result in an aggravated EAE response is that the number of immune-regulatory FoxP3 + T cells was also increased (see supplementary material, Figure S1B), resulting in a 'balanced' immune response [27]. In a previous study, in which MHCII-CD40-T2/3/5 −/− mice were subjected to atherosclerosis, we also observed an increase in CD3 + T cells during disease and found that the increase was due to an increase in regulatory T cells [22]. The balancing of the immune response is also reflected by the fact that gene expression analysis also showed no differences in the expression of inflammatory genes in the CNS of these mice ( Figure 3C). The reduction in EAE severity in the MHCII-CD40-T2/3/5 −/− mice might be explained by the reduced MOG IgG production by the B cells (Figure 2).
The CD40-TRAF6 pathway is especially known to be important for CD40 signaling in the myeloid lineage. CD40-TRAF6 interactions induce classical monocyte activation, monocyte recruitment and macrophage activation via canonical NF-κB signaling [23,[27][28][29]. Although the role of CD40-CD40L signaling in MS is mostly attributed to T cells [30] and B cells [31], the main cell types expressing CD40 in human MS lesions are monocytes, macrophages and activated microglia [7]. Infiltrating macrophages and activated microglia present in active human MS lesions contribute to CNS injury and have mainly M1 characteristics, including high CD40 expression [32][33][34]. Although macrophage CD40 is abundantly present in MS lesions, its role in the pathogenesis of MS is currently unknown.

Myeloid CD40-deficient mice develop less severe EAE and exhibit decreased CNS inflammation
To study the role of myeloid CD40 on neuroinflammation in vivo, CD40 flfl LysM cre mice were generated, which exhibited an 84-89% reduction in CD40 expression in macrophages ( Figure 4A,B). We immunized CD40 flfl LysM cre and CD40 flfl (WT) mice for induction of EAE. Deficiency of myeloid CD40 resulted in significant lower disease severity (maximum disease score 3.0 ± 0.2) and cumulative disease score (AUC 14.4 ± 2.0) compared with WT controls (maximum score 4.0 ± 0.3, p = 0.010, AUC 24.7 ± 1.9, p = 0.0013; Figure 4C and Table 2). Moreover, survival of CD40 flfl LysM cre mice was increased compared with WT mice (100% versus 75%, Table 2). In WT mice, the maximum body weight loss was 13.9 ± 2.0% and resulted in a 4.3 ± 3.4% weight loss at the end of the experiment. CD40 flf LysM cre mice had significantly less severe weight loss, with a maximum of only 7.9 ± 1.7% loss and complete recovery of body weight at the end of the experiment (p < 0.05, Figure 4D and Table 2). As expected from a macrophage-specific CD40 knockout, IgM, total IgG and anti-MOG IgG levels did not differ between CD40 flfl LysM cre and WT mice (see supplementary material, Figure S3). The CNS of CD40 flfl LysM cre mice exhibited a significant reduction in CD45 + leukocyte infiltration, which was due to a reduction in Mac3 + macrophages/active microglia and CD3 + T cells ( Figure 4E,F), revealing a reduction in CNS inflammation in the absence of macrophage CD40. No differences in peripheral immune responses were observed (see supplementary material, Figure S4C). In addition to the previously described role of CNS CD40 on microglial cell activation during EAE [35], we here show that loss of peripheral macrophage CD40 signaling ameliorates EAE and prevents inflammation, revealing a key role for macrophage CD40 signaling in neuroinflammation.
In EAE, myeloid cells cause CNS damage by presenting antigen, secretion of cytokines and phagocytosis of cell debris, resulting in the production   of free radicals [36]. In EAE lesions, macrophages phagocytose myelin and allow the removal of cell debris, critical for effective repair [33]. CNS LFB staining displayed that CD40 flfl LysM cre mice were protected against EAE-induced myelin loss ( Figure 5A). Flow cytometry analysis of myelin-laden BMDMs from CD40 flfl LysM cre and WT mice show that CD40 flfl LysM cre BMDMs ingested significantly less myelin than WT BMDMs (54.8 ± 8.6% reduction, n = 4) ( Figure 5B). CD40 flfl LysM cre BMDMs also expressed less CD36, suggesting that macrophage CD40 mediates clearance of myelin debris ( Figure 5C). Earlier research using bone marrow chimeric mice showed that CD40 expression on hematopoietic cells is essential for the priming of encephalitogenic T cells [8]. FACS analysis revealed that CD40 flfl LysM cre BMDMs exhibited a reduced expression of the classically activated macrophage markers CD80 and CD86, an unchanged expression of MHCII and a decrease in the scavenger receptor and, alternatively, the activated macrophage marker CD204 ( Figure 5D-G). The reduced expression of costimulatory molecules supports the results on the reduced EAE development in these mice as costimulation is an essential signal for T cell activation [37,38]. Gene expression analysis of BMDMs incubated with myelin showed no significant differences (see supplementary material, Figure S4A,B). mRNA levels of Il12b were decreased in CD40 flfl LysM cre macrophages treated with LPS/IFN-γ, whereas Il6, Tnf and Nos2 mRNA levels were increased (see supplementary material, Figure S4A). However, protein levels of these cytokines were not affected (see supplementary material, Figure S4C). CD40 flfl LysM cre macrophages treated with IL-4 showed increased Cdh1 gene expression compared with WT BMDMs, but no differences were observed in mRNA levels of other immunomodulatory genes, including Cd204, Il10 and Tgfb (see supplementary material, Figure S4B). In MS lesions, human foamy macrophages and microglia express proinflammatory (CD40 and iNOS) markers as well as anti-inflammatory markers (MR) [33].
Macrophages are diverse and plastic, their phenotype is dependent on cues from the microenvironment, including inflammatory signals [39]. This might explain why we do not observe a reduced inflammatory phenotype in vitro, whereas the CD40 flfl LysM cre mice display reduced neuroinflammation and are protected against clinical symptoms of EAE in vivo.

Conclusions
We have shown that macrophage CD40 plays a key role during neuroinflammation, probably via CD40-TRAF6 signaling. However, the reduction in neuroinflammation was stronger in mice lacking CD40-TRAF6 signaling in all MHCII + cells than in mice lacking CD40 signaling in macrophages, suggesting that both CD40-TRAF6 signaling in macrophages and other antigen-presenting cells including B cells and dendritic cells drive neuroinflammation. Recently, we have developed small molecule inhibitors of the CD40-TRAF6 interaction, TRAF-STOPs, and we were able to show that these small molecule inhibitors can reduce atherosclerosis, peritonitis, sepsis and metabolic complications of diet-induced obesity in mice [23,28,29]. Interestingly, TRAF-STOP treatment was also able to reduce CNS leukocyte infiltration during EAE in both mice and rats but unfortunately failed to ameliorate disease severity [40]. Further optimization of TRAF-STOPs for more optimal delivery in the CNS may be required to achieve better therapeutic results. Our data reveal that blocking CD40 in a cell type-specific manner, or inhibition of parts of the CD40 signaling, has great therapeutic potential for the treatment of MS, as it will cause limited immune-suppressive side-effects compared with untargeted inhibition of the entire CD40 pathway.