Abstract

INTRODUCTION

Immune cells classically involved in innate responses include natural killer cells, neutrophils, dendritic cells and macrophages that participate in the selective recognition and the clearance of pathogens and toxic cell debris during infection or tissue injury [1, 2]. Nevertheless, these peripheral cells show little immunosurveillance of the brain, and it is now evident that resident cells, glial cells, ependymal cells and even neurons are capable of mounting innate immune responses on their own [3]. The local innate immune response is based on the recognition of non-self and altered self- patterns by molecules and receptors expressed by microglia but also present on astrocytes, oligodendrocytes and neurons [4, 5]. The outcome of these interactions is dependent on the microenvironment, as well as, on the nature of the ligands and on the nature and combination of the ligated receptors. From a therapeutic point of view, it is important to manipulate this innate immune response to promote clearance of pathogens or toxic proteins accumulated during brain injury or in central nervous system (CNS) diseases [5, 6]. However, several innate immune molecules can contribute to cytotoxic and cytolitic activities and must be controlled to avoid neuronal loss and excessive inflammation [7]. There is new emphasis on the role of neuroimmune regulatory proteins (NIRegs) involved in silencing and reshaping an adverse innate immune response such as, CD47 and CD200 expressed by neurons as “don’t eat me” signals that inhibit microglial activity preventing host cell attack, besides to modulate microglial function and phenotype in situations of harmful and excessive inflammation [8]. Here, we present a review about CNS immunity and the modulation of CD200 and its receptor in multiple sclerosis (MS) and in the aging brain and Alzheimer’s disease (AD). We show original data on the effects of an inhibitor of endocannabinoids (eCBs) uptake in the regulation of CD200-CD200 receptor (CD200R) interaction in an experimental model of MS.

CNS INNATE IMMUNE RESPONSE IN THE RESTORATION OF BRAIN HOMEOSTASIS: ROLE OF MICROGLIAL CELLS

Microglia constitute the highly versatile resident macrophages in the CNS, comprising 5% to 20% of the total glial cell population [9]. As reviewed by Yang et al., [10], in 1919, Ramón y Cajal’s disciple, Pío del Río-Hortega recognised microglial cells as a different cell type from the other glial cells. He also characterised microglial transformation from a ramified phenotype into ameboid phagocytic macrophage-like cells in the stab wounds made in animal brains [11]. From these observations, del Río-Hortega concluded that microglia originated from peripheral mononuclear cells [11].

Interestingly, and after a long debate in which even the existence of microglial cells was questioned [12], microglia has finally been established as a distinct glial cell population from a myelomonocity origin [13]. Recently, it has been demonstrated that microglial cells are originated from a distinct myeloid precursor that migrates from the haematopoietic islands in the yolk sac to the developing brain parenchyma at embryonic day 8.5 [14].

Microglia are widely distributed in the adult CNS, but differences in their cellular density among different brain areas have been reported in mice [9, 15] and humans [16]. Thus, microglial cells are more abundant in the telencephalon than in the diencephalon or the mesencephalon, with the rhombencephalon containing the lowest amount of these cells [9, 17]. Not only there is an uneven distribution of microglia across the cerebral anatomical regions but between the grey and white matter (the latter containing higher microglial cell density than the former). Moreover, cumulative data show that microglial expression of tissue macrophage markers varies within the different brain regions, as well as, their response to different stimuli [18, 19, 20].

In the adult healthy brain, microglia has been described to be in “resting” or “quiescent” state with a ramified morphology. However, this term has limitations and can lead to the misleading idea that microglia are dormant cells awaiting a signal that will wake them up. On the contrary, recent elegant in vivo imaging experiments show “resting” microglia as highly dynamic cells, continuously branching thin processes that survey and sample their microenvironment [21, 22] in search for potential threats and danger signals. Thanks to this routine immunosurveillance of the CNS, microglia will remove apoptotic bodies and other potentially toxic cellular debris (myelin debris, amyloid deposits, protein aggregates, etc) [23, 24, 25]. In order to accomplish this task, it is essential that microglial cells are capable of distinguishing between “self” and “nonself” signals (Fig. ​11). Clearence of pathogens and toxic cell debris during infection or tissue damage is based on the recognition of “nonself” and “altered-self” patterns by microglia, but also astrocytes, oligodendrocytes and neurons have been shown to be able to recognise those patterns [1, 4]. According to Medzhitov and Janeway [1], there is a plethora of the so-called “eat-me” signals expressed by pathogens and apoptotic or necrotic cells. Some of these signals are a heterogeneous group of molecules known as pathogen-associated molecular patterns (PAMPs) and are characterised by being highly conserved through evolution with little antigenic variability [22]. These PAMPs are constituents of the microbial structure which induce in the host a strong innate immune reaction directed towards the removal of the pathogen by phagocytosis [26]. The classical example of a PAMP is the lippopolysaccharide (LPS), component of the Gram negative external membrane. Analogous to PAMPS, it has been proposed that cells undergoing programmed cell death express de novo apoptotic cell-associated molecular patterns (ACAMPs) [22, 27, 28]. These ACAMPs would play a key role in the embryonic process in which whole cell populations need to be cleared out without mounting an inflammatory response and minimising the presence of cellular debris [29]. Some ACAMPs include oxidized low density proteins, alteration of the membrane electrical charges, nucleic acids and phosphatydilserine [4, 30]. Similarly, damaged or stressed tissues release/express the so-called danger-associated molecular patterns (DAMPs). Some DAMPs including heat shock proteins (HSP), adenosine, ATP, high motility group box chromosomal protein 1 (HMGB-1), galectins and thioredoxin present adjuvant and pro-inflammatory activity [31]. Phagocytic cells recognise these PAMPs, ACAMPs and/or DAMPs, which can be either membrane-bound or soluble, through their pattern recognition receptors (PRRs) [22, 28, 32]. Some of the PPRs include toll-like receptors (TLR), scavenger and manose receptors, CD14, CD36, complement receptors, phosphatydilserine receptor (PSR) and milk fat globulin (MFG-EGF8) [8]. Therefore, the activation of microglia, rather than an unspecific process, is highly dependent on the stimulus that originated it.

Fig. (1)

Roles of “eat me” and “don’t eat me” signals in the CNS innate immune responses. Scheme showing how microglial cells differenciate “self” from “non-self” cues.

In contrast to this plethora of signals that might elicit an innate immune activation directed to the elimination of the pathogen, the apoptotic cell or the tissue debris; there is a complex set of interactions that silence and reshape microglial response [1]. For instance, electrically active neurons inhibit the interferon-γ (IFNγ)-induced increase in major histocompatibility complex (MHC) class II expression in microglia [33]. Some neurotransmitters have modulatory effects on microglial response whereas others, like substance P enhance the activation of microglia [34]. Neurotrophins secreted by healthy neurons such as nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF) and to a lower extent neurotrophin-3 (NT-3) were able to reverse the induction of MHC class II molecules in microglia [35]. Interestingly, neurons express membrane-bound molecules and/or secrete soluble mediators that function as “don’t eat me” signals, reshaping microglial response and inhibiting their phagocytic activity on the site of injury [4, 36]. These signals were originally grouped in a family of heterogeneous molecules called self-associated molecular patterns (SAMPs). It was suggested that these SAMPs could interact with novel inhibitory PPRs, negatively modulating the innate immune response and promoting tissue repair [4]. These SAMPs signals are present in almost all host “normal-self” cells but not in pathogens and they are downregulated in apoptotic/necrotic cells (altered self). Thus, this “normal-self” identity can be lost. An example of a molecular marker of “normal-self” is the presence of sialic acid in the terminal of glycoproteins and glycolipids, which can interact with different microglial receptors [37]. One of these inhibitory receptors belongs to a family known as siglecs [38]. Siglecs contain immunoreceptor tyrosine-based inhibitory motifs (ITIMs) in their cytoplasmic tail which downregulate phagocytosis in macrophages and microglia [39, 40]. Most microorganisms lack sialic acid in their glycolipids and/or glycoproteins and in some cases, virally infected or transformed cells can show a deficient pattern of sialization, which is a strong signal of “altered-self” to promote phagocytosis. Also in apoptotic cells, a reduced expression of sialic acid has been shown [22].

Of particular interest is the physical interaction (by means of their membrane-bound proteins) between neurons and microglia. Thus, a group of SAMPs were renamed NIRegs to highlight their role in modulating, reshaping an adverse immune response and skewing microglia towards a protective phenotype. Interestingly, many of these NIRegs contain an ITIM domain in their cytoplasmic tails which negatively regulated microglial response to insults.

As mentioned before, siglecs are members of a subgroup of the immunoglobulins (Ig) superfamily that recognize the sialic acid residues on the periphery of cell surface glycolipids and glycoprotein. Siglecs contain an ITIM domain in their cytoplasmic tails which signal via the recruitment of tyrosine phosphatases such as SHP1 [40]. In this sense, when microglial siglecs interact with the neuronal glycocalix, neurotoxicity is alleviated [41].

Another interesting NIReg protein is the pair CD200- CD200R. CD200 is a 41-47 KDa protein belonging to the Ig supergene family characterised by two immunoglobulins superfamily (IgSF) domains [42, 43]. This surface protein is highly conserved and within the CNS is mainly expressed by neurons and vascular endothelium [42, 44, 45, 46], but CD200 is present in almost peripheral cell types, thymocytes [47], T and B cells [43], dendritic cells [48] and even in brain astrocytes and oligodendrocytes [49, 50]. In contrast, CD200R, which also contains two IgSF domains and a longer cytoplasmic tail with an ITIM domain as showed in (Fig. ​22) [42], is chiefly expressed by myeloid cells and microglia [46, 51]. However, CD200R has been described to be present also in thymocytes [52] as well as in T and B cells [43, 53]. Functional studies about the role of CD200-CD200R interaction within the brain indicate that this pair of molecules plays a critical role in neuronal protection in the setting of inflammatory-mediated neurodegeneration (Fig. ​33). In particular, this immune-regulatory system, when deficient, may contribute to chronic inflammation in MS, AD or in the aging brain as we will show in this review. Deletion of CD200 results in myeloid cells dysregulation and enhanced susceptibility to autoimmune response [51, 54].

Fig. (2)

Binding of CD200-CD200R in the CNS. Scheme showing the binding of CD200 expressed in neurons, oligodendrocytes astrocytes and brain endothelial cells with its receptor CD200R expressed mainly in microglial cells. Both molecules are type-1 membrane …

Fig. (3)

Neuron-microglia crosstalk. A) In control conditions the crosstalk between neuron-microglia via the CD200-CD200R interaction is working correctly. The inhibitory signal is contributing to the maintenance of brain homeostasis and immune privilege. …

Other neuron-microglia ligand-receptor interaction that can modulate microglial responsiveness is the neuronal CD47 (also expressed by endothelium, astrocytes and dendritic cells) [55] and the microglial signal regulatory protein SIRPα (CD172 ) which contains three extracellular Ig domains [56]. CD47-CD172 interaction recruits the tyrosine phosphatases SHP-1 and SHP-2 with inhibition of phagocytosis and synthesis of the anti-inflammatory cytokine TGF-β [22]. Another NIReg with a potential critical role in neuroprotection include the pair neuronal CX3CL1 and microglial CX3CR1 [57].

Interestingly, some membrane receptors contain immunoreceptor tyrosine-based activation motifs (ITAM). ITAM-containing signalling adaptor proteins are associated with receptor subunits. After ligand-receptor interaction, tyrosine residues of the ITAMs become phosphorylated by members of the Src kinase family and subsequently serve as docking sites for Src homology 2 (SH2) domains of Syk protein kinases which then mediate cellular activation via a number of downstream cascades [40]. Triggering receptor expressed on myeloid cells 2 (TREM-2) is an innate immune glycoprotein heterogeneously expressed in brain microglia [58]. It belongs to the immunoglobulin and lectin-like superfamily with a short cytoplasmic tail [59]. Interaction of TREM-2 with its uncharacterized ligand induces its association with the ITAM-containing adaptor molecule DAP12 followed by the recruitment of ZAP70, SYK, PI3K and phospholipase Cγ [60]. Activation of TREM-2/DAP12 leads to augmented phagocytosis with reduced expression of tumor necrosis factor-α (TNFα), interleukin (IL)-6 and inducible nitric oxide synthase (iNOS) [61]. Mutation of either TREM2 or DAP12 proteins lead to the rare Nasu-Hakola disease (NHD), also known as polycystic lipomembranous osteodysplasia with sclerosing leukoencephalopathy (PLOSL), an autosomal recessive inherited disease [62]. While NHD patients carrying TREM-2 mutation present an early onset presenile dementia followed by delayed bone symptoms [63], patients with mutations in DAP12 display an early onset combination of presenile dementia and systemic bone cysts [64]. Interestingly, Piccio et al., [65, 66] showed that a soluble form of TREM-2 was present in the cerebrospinal fluid (CSF) of patients with MS and that blockade of TREM-2 with a blocking antibody exacerbated experimental autoimmune encephalomyelitis (EAE) in rodents. These observations suggest a crucial role of TREM-2 in preventing neurodegenerative processes. So far the nature of the TREM-2 ligand remains unknown. It has been suggested that TREM-2 bind anionic ligands of the surface of Gram positive and Gram negative bacteria, including Neisseria gonorrheae, as well as an unidentified ligand in the astrocytoma cell line HTB12 [61]. Importantly, Hsieh et al. [67], using a TREM2/Fc chimera, showed that neurons undergoing apoptosis increased the expression of a TREM-2 ligand (TREM-2L), inducing their phagocytosis by BV2 cells (a murine microglia cell line). This effect could be reversed with the use of a blocking antibody against microglial TREM-2 [67].

This review will examine how the CNS innate immune response maintains a critical balance between protective and potentially harmful effects by the interaction of CD200 and its receptor CD200R in neurological diseases which involve neuroinflammation with special emphasis to MS, AD and aging brain. The balance between the destructive and protective effects of the innate immune response must be precisely regulated in order to promote conditions that support brain repair and re-established tissue homeostasis.

MULTIPLE SCLEROSIS

MS belongs to a larger group of inflammatory demyelinating diseases of the CNS which include besides the different manifestations of MS, acute disseminated leukoencephalitis, Devic’s neuromyelitis optica and Balo’s concentric sclerosis. Although these diseases differ in clinical course, imaging pathology and immunopathogenesis, they share some essential structural features of their lesions, as they all occur on a background of inflammatory reaction composed of lymphocytes, activated macrophages and microglia and show demyelination. MS is a chronic inflammatory and neurodegenerative disease affecting over 2.5 million individuals worldwide. The pathological hallmark of MS is white matter demyelination but there are other features such as axonal and neuronal damage, grey matter demyelination, composition of vascular cuffs (monocytes, T cells, B cells, plasma cells), loss of oligodendrocytes and anatomical lesions that vary between patients [68]. The pathogenesis of MS is still not fully understood, although significant progress has been made in the last decades [69]. Several immunological abnormalities have been identified in patients with MS, such as the presence of oligoclonal bands in the CSF, higher levels of T cell activation and decrease of regulatory T cell function. The presence of inflammation driven by the adaptive immune system in absence of infectious pathogens, the response to immunomodulatory therapy, in addition to the association with human leukocyte antigen (HLA) and the predominance for young women strongly suggest an autoimmune origin. Regarding the susceptibility for suffering MS, familial and twin studies have revealed a concordance in monozygous twin of 30% compared with dyzogous twin and sibling of 2-5%. The identification of genomic factors predisposing for MS have revealed the association with HLA class II and specifically with the HLA-DRB1*1501 allele. In respect to environmental factors, several factors have been validated by different studies and meta-analysis including the association with common infections (Epstein-Barr virus, Herpesvirus type 6 and MS retrovirus), levels of vitamin D and sun exposition [70,71]. Regarding association with infections, the picture is complex because no single pathogen has been linked with MS in a clear way. Therapies that have been developed so far are aimed at multiple mechanisms such as broad immune suppression or inhibition of immune cell migration. Some of these therapies reduce the clinical disease activity and the progression of lesion load as determined by magnetic resonance imaging (MRI), but so far no therapy is available that can cure MS or even halt its progression. The search for more effective approaches is therefore necessary.

IMMUNOREGULATORY MOLECULES IN NEUROINFLAMMATION

As commented before, in the CNS exists several intrinsic mechanisms that tightly regulate the activities of microglia [94]. This is critical, as an uncontrolled inflammatory reaction could be detrimental to a tissue that is known for its poor regenerating capacity. Quiescent microglia inspect and guard their environment [95], but when a certain threshold is exceeded they become activated and show retraction of ramifications acquiring round phenotype and upregulate a number a markers like CD45 and MHC class II. These cells will become motile, phagocytic and might display local proliferation [96]. A proper balance in immune activating and immune inhibitory signals thus regulates their activation. It has been postulated that excessive microglia/macrophage activation in a pathological setting is due to imbalanced control, reflected by impaired immune activation, immune inhibition or both, and could lead to disease progression as seen in MS [97]. This view subsequently implies that correcting the equilibrium by supplying extra inhibitory signals specific for these cells may dampen the pathological inflammatory response and restore the immune suppressed environment of the CNS [8]. A specific therapeutic candidate target to diminish inflammation in MS via the immune inhibitory molecules is CD200 and its receptor CD200R.

CD200-CD200R in MS

CD200 is a membrane glycoprotein belonging to the IgSF, with CNS expression extremely high. Among other cells, neurons express the highest level of CD200 [46]. CD200 has a short cytoplasmic tail with no known signaling motifs, then CD200 itself exerts functions in the cell on which it is expressed its receptor. Precisely, CD200R, homologous to CD200, is expressed mainly on myeloid cells such as macrophages, microglia and dendritic cells [98, 99] and on a subset of B and T cells [53, 100]. CD200R has a cytoplasmic tail which contains three tyrosine residues, one of which forms part of an NPXY motif. Upon binding to its ligand, the tyrosine residues become phosphorylated, and adaptor proteins downstream of tyrosine kinase (DOK) 1 and 2 are recruited. This ultimately leads to inhibition of the MAPK p38, extracellular kinase (ERK) and c-Jun terminal kinase (JNK), the common pathways involved in classical activation of macrophages and microglia. This in turn results in inhibition of release of pro-inflammatory cytokines like IL-1β, IL-6 and TNFα [101, 102], but promote the secretion of IL-10 [103]. This is an important point as CD200R engagement not only decreases the production of pro-inflammatory cytokines, but enhances the production of an anti-inflammatory cytokine. Therefore, through CD200-CD200R interaction, the activity of macrophages and microglia can be down-regulated. Interestingly, using neuron-microglial cell co-cultures model, as described by Hernángomez et al. [103], IL-10 protected neurons from microglia-induced cytotoxicity. In this context, it has been suggested that CD200-CD200R interaction may be one of the mechanisms by which IL-10 protects neurons from inflammatory damage by increasing the expression of CD200 in neurons [103]. In fact, in time-lapse experiments using co-cultures of neurons-microglial BV2 cells activated with LPS/IFNγ, the addition of the anti-inflammatory cytokines IL-4 and IL-10 rescue neurons from inflammation induce death while the blockade of the interaction of CD200 with its receptor enhanced neuronal death induced by the inflammatory stimuli (Fig. ​44).

Fig. (4)

Microglia-neurons co-cultures: The anti-inflammatory cytokines, IL-4 and IL-10, reduce neuronal death- induced by inflammatory stimuli while the impairment of CD200-CD200R interaction enhanced neuronal death. A) Representative time-lapse experiment…

The high expression of CD200 in the CNS is thought to be a mechanism of constitutive immune suppression and is developmentally regulated in the mouse brain (104). In fact, the blockade of CD200-CD200R interaction, for instance, in CD200-/- deficient mice, leads to spontaneous activation and increased proliferation of microglia and macrophages, demonstrated by upregulation of CD45 which is a hallmark of microglial activation and expression of iNOS [51]. In CD200 deficient mice induction of EAE and experimental uveoretinitis showed enhanced macrophage infiltration and a more rapid and severe disease course compared to wild type mice [51, 105].

A blocking antibody against CD200R aggravated EAE in rats by enhancing infiltration of T cells and activated macrophages [106]. In addition, the interruption of CD200-CD200R interaction increased neuronal death in macrophage/microglia co-cultures [103, 106]. These findings implicate that CD200 is a potent immune inhibitory molecule and that reduced inhibitory input from CD200 causes a disturbed equilibrium which subsequently results in cellular activation and tissue damage. This has been confirmed in MS, since the decreased expression of CD200 and other immune inhibitory molecules like CD47 in chronic active lesions was accompanied of macrophage/microglia activation expanding the lesions [97]. A reduction of CD200 and CD200R gene expression in the spinal cord was also observed at later chronic phases in the viral model of MS, TMEV-IDD [103]. Immune inhibition might be hampered in MS and this could facilitate the activated state of macrophages/microglia with demyelinating activity as a consequence. Importantly, findings in animal studies show that increased CD200R signaling can be beneficial. For example, mice that have inherently elevated levels of CD200 show ameliorated disease signs in the EAE model accompanied by enhanced neuroprotection and elevated levels of IL-10 [49]. As expected, these animals displayed fewer activated macrophages/microglia and less demyelination and axonal damage. CD200R is likely to mediate several actions of alternatively activated macrophages (M2) such as the reduced secretion of pro-inflammatory cytokines. Moreover, CD200R expression on T cells seemed to be restricted to Th2 cells [107]. In summary, impaired immune inhibition by reduced CD200/CD200R interaction might result in a pro-inflammatory environment that may contribute to demyelination and axonal injury in MS.

THE CANNABINOID SYSTEM: CANNABINOID RECEPTORS AND ENDOCANNABINOIDS PRODUCTION AND INACTIVATION

Cannabinoids (CBs) activate mainly the cloned CB1 and CB2 receptors, but also can interact with the TRPV1 vanilloid receptor, GPR55 and other yet uncloned receptors such as abn-CBD receptors [108, 109]. In addition, CBs can interact with peroxisome proliferator-activated receptors, PPAR γ and PPARα [110,111]. CB1 receptors are expressed abundantly throughout the brain by many neurons, as well as outside the CNS, but at much lower levels [112, 113, 114]. CB1 receptors are Gi/o protein-coupled receptors that modulate the activity of several plasma membrane proteins and intracellular signaling pathways. Several studies suggest that short-term versus long-term activation of CB1 receptors affects different cellular functions. For example, activation of neuronal CB1 receptors for seconds inhibits presynaptic N-type calcium channels and activates inwardly rectifying potassium channels, reducing neurotransmission [115,116]. In contrast, CB1 receptor activation for minutes to hours changes gene expression, for example, inducing the expression of neuroprotective proteins, such as BDNF, known to counteract cell damage [117]. Under neuropathological conditions whereby the blood-brain barrier (BBB) is disrupted, large quantities of leukocytes enter the CNS and participate in neuroinflammatory responses. Leukocytes express CB1 receptors, although at lower levels compared with neurons, and their activation modulates related immune responses [118]. Thus, CBs may affect neuroinflammation through their action at CB1 receptors expressed by invading leukocytes. CB2 receptors are also Gi/o protein-coupled receptors [119]. They share 44% protein identity with CB1 receptors and display a distinct pharmacological profile and expression pattern. Many laboratories reported that CB2 receptors are not expressed in healthy brain [118,119, 120] suggesting that they are not expressed by healthy neurons, astrocytes, resting microglial cells, and oligodendrocytes. However, more recent studies show that microglia, astrocytes, oligodendrocytes, progenitor neural cells and even neurons can express CB2 receptors [121, 122, 123]. However, CB2 receptors are expressed primarily by immune cells [118]. Δ9–Tetrahydrocannabinol (THC) binds to CB2 receptors as a partial agonist [119, 124], an effect that leads to reduced inflammation by inducing apoptosis of immune cells [125] and inhibiting the ability of macrophages to process antigens and prime helper T cells [126, 127]. The expression and pharmacological dichotomy between CB2 and CB1 receptors have tremendous pharmaceutical potential, as compounds that interact selectively with CB2 receptors may provide anti-inflammatory drugs devoid of CB1-induced psychoactive adverse effects.

In 1992, Mechoulam and coworkers identified the first endogenous compound that binds CB1 receptors with high affinity [128]. This compound, arachidonoylethanolamide (AEA), which was named anandamide, is produced in an activity-dependent manner, activates CB receptors, and is inactivated enzymatically. The activity-dependent increase in AEA production was initially demonstrated with neurons in primary culture and then, with microdialysates from freely moving rats [129, 130, 131]. This response requires a rise in intracellular calcium, which increases the activity of an acyltransferase specific for the precursor of AEA, N-arachidonoylphosphatidylethanolamide (NAPE), and a NAPE-specific phospholipase D, generating AEA (Fig. ​55). In 1995, two laboratories identified a second endogenous ligand, 2-arachidonoylglycerol (2-AG) which is approximately 200 times more abundant than AEA in brain and activates CB1 and CB2 receptors [120, 132]. Neurons and glial cells produce and inactivate 2-AG [129,131, 133, 134]. Its production also requires sustained increase in intracellular calcium levels and the activation of the phospholipase C-diacylglycerol lipase pathway [135, 136]. Several lipids structurally related to AEA and 2-AG can also interact with CB receptors, among them, palmitoylethanolamide (PEA) which contains a 16:0 fatty acid moiety may be a subtype of eCB. PEA is produced and inactivated by cells and induced biological effects that are blocked by a CB2 antagonist although it does not bind to any known CB receptors [131, 137, 138, 139]. The bioactivity of eCBs can be terminated by desensitization and internalization of CB receptors, clearance of eCBs from the extracellular milieu and by enzymatic hydrolysis of eCBs, namely AEA by fatty acid amide hydrolase (FAAH) and 2-AG by monoacylglycerol lipase (MGL), ABH serine hydrolase and cyclooxigenase 2 [140, 141, 142, 143]. The characteristics of AEA and 2-AG uptake into cells are similar and suggest a common mechanism [135]. However, whether a protein mediates this uptake is still controversial. In summary, the eCB family consists of several structurally related lipids that interact with CB receptors. Several studies have identified eCBs as retrograde signaling molecules, being generated in a calcium-dependent manner from postsynaptic terminals, diffusing toward presynaptic terminals and inhibiting neurotransmitter release [144, 145, 146]. Figure 6 shows the current signaling relationship between glia and neurons and the role played by eCBs.

Fig. (5)

Models showing AEA and 2-AG production and inactivation. AEA is produced by phospholipase D hydrolysing NAPE which is produced by N-acyltransferase. AEA is released toward the extracellular space and activates CB1, CB2 and TRPV1 receptors; AEA is …

eCBs Promote CD-200-CD200R Interaction

eCS has arisen as a promising new therapeutic target for the treatment of MS [153, 158, 169]. In animal models of MS, exogenous CB agonists alleviate disease symptomatology [150–152], a benefit also produced by pharmacological interventions aimed at increasing the levels of eCBs [159, 161, 162, 170]. Uncontrolled innate immune responses within the CNS are widely recognized as playing a major role in the development of autoimmune disorders and neurodegeneration, with MS and AD being primary examples. However, the role of the eCS in the regulation of inhibitory immunoregulatory proteins and their impact in microglial function has not been completely elucidated. During the past years, several in vitroand in vivo studies have suggested that eCS participates in the control of brain immune responses as well as in the protection of the CNS against injury [171]. During immune-mediated attack of the brain, it has been hypothesized that the activation of eCBs represents a protective mechanism, aimed at reducing both neurodegenerative and inflammatory damage through various and partially converging mechanisms that involve neuronal and immune cells. The major eCBs, AEA and 2-AG, have been shown to exert neuroprotective activity largely thought to be mediated through two specific CB receptors, the CB1 and CB2 receptors. Regarding innate immunity and in particular, CD200-CD200R, recently, our lab showed that the treatment with AEA induced the recovery of CD200 and CD200R gene expression that was reduced in the TMEV model of MS [103]. This was accompanied by decreased inflammatory mediators and reduced microglial reactivity. Here, in an attempt to further support the ability of AEA to modulate the interaction of CD200 with its receptor, we show that the administration of an inhibitor of AEA uptake (UCM-707) to TMEV-infected mice also induced a significant recovery in the expression levels of CD200 (Fig. ​7A7A; p<0.05) that was decreased in TMEV-diseased mice. Treatment with UCM-707 did not significantly modify the expression of CD200R mRNA (Fig. ​7B7B), although there was a trend toward increase. The effects of UCM-707 on CD200 mRNA were accompanied by a diminished expression of IL-1β mRNA (Fig. ​7D7D; p<0.05) and an increased of IL-10 mRNA expression (Fig. ​7C7C; p<0.05). Therefore, increased AEA tone [162] prevented gene expression of a pro-inflammatory cytokine such as IL-1b and increased the expression of the anti-inflammatory cytokine IL-10, accordingly with a reduction in microglial reactivity, likely by increasing CD200 expression in the spinal cord. Importantly, the axonal damage observed in TMEV-infected mice in comparison to Sham animals (p<0.01) was reduced after treatment with UCM-707 (p<0.05) as evaluated by SMI32 staining (Fig. ​7E7E). Analyses of motor function in TMEV-infected mice treated with UCM-707 confirmed previous data from our group [162], as vertical (p<0.05) and horizontal activity (p<0.001) deficits were ameliorated (Fig. ​7F7F). These data indicate that CD200 negatively correlates with the degree of spinal cord axonal damage induced by TMEV and suggest that increased CD200 expression may be one of the mechanisms contributing to the anti-inflammatory effect observed in the TMEV-IDD model of MS by enhancing eCB tone [161, 162].

Fig. (7)

The inhibitor of AEA uptake, UCM-707, induces a recovery in the expression of CD200 mRNA in the spinal cord of TMEV-IDD mice. TMEV-infected mice were subjected at 60 days pi (during established disease) to the administration of the selective inhibitor …

The mechanisms that lead to reduction of damage and to promote repair involve switching macrophage functional activation to include decreased pro-inflammatory cytokines [79]. In studies from our lab we have described that macrophages infected with TMEV increased their production of AEA and that both, AEA and inhibitors of AEA cellular re-uptake inhibited IL-12p40 release [161]. The identification of microglia as one of the major sources of IL-23 in brain sections from MS patients points out the importance of microglial cells as cellular targets to down-regulate this family of heterodimeric cytokines. In the TMEV-IDD model of MS, it is unknown whether TMEV-infected microglia exhibit the autocrine regulatory mechanism between IL-10 and the cytokines of IL-12 family [172] but data from our lab have established the molecular mechanisms by which AEA might regulate the IL-12p70/IL-23 axis and IL-10 in TMEV-infected microglial cells and in the TMEV-IDD model of MS [173, 174].

Taken together, our findings support the hypothesis that enhancing endogenous anti-inflammatory regulatory systems like the eCS, basic mechanisms of innate immunity like CD200-CD200R interaction that are disrupted in chronic inflammatory diseases such as, MS can be restored representing a new therapeutic strategy that prevents harmful inflammation and axonal damage [175, 176].

REGULATION OF MICROGLIAL ACTIVATION BY CD200-CD200R INTERACTION IN OTHER NEUROINFLAMMATORY CONTEXT: ALZHEIMER’S DISEASE AND THE AGING BRAIN

The role of inflammation in AD has been intensively studied [177, 178]. Pathological studies of AD indicated the presence of activated microglia, reactive astrocytes and complement activation on association with amyloid β (Aβ) containing plaques, suggesting that a kind of chronic inflammation was ongoing [179,180]. A range of in vitro studies using cultured microglia from humans and rodents indicated that aggregated Aβ peptide activates microglia to a pro-inflammatory state capable to produce a wide range of neurotoxic mediators [181, 182, 183]. In this line, anti-inflammatory therapies have been considered for treating this disease [184]. Because clinical trials with anti-inflammatory compounds did not generally show effectiveness at slowing the progression of AD, new inflammatory therapeutic targets are needed. These would include the enhancing of the function of endogenous immune regulatory molecules such as CD200 and its CD200R.

CD200 and CD200R in Human Elderly Brain and in AD

Recent studies have focused in the characterization of CD200 and CD200R expression in elderly human brains, as well as in AD. One of these studies showed a decrease in CD200 mRNA expression in the hippocampus of rats with increasing age [50], in agreement with those observed in MS lesions [97]. The main findings obtained from brain tissues affected by AD showed that there is a deficit of CD200 but also of CD200R [181]. As confirmed in other studies CD200 is abundant in the human brain while CD200R is expressed at lower levels [46]. In fact, low expression of CD200R by microglia was confirmed at the mRNA and protein levels using cultured human microglia compared to blood-derived macrophages [181]. Neuronal expression of CD200 was decreased in brain regions affected by AD pathology evaluated by quantitative measurements of protein and mRNA in brain extracts as well as in antibody stained tissue sections. Interestingly, in the normal human brain differences in CD200 immunoreactivity can be observed depending on the brain area studied, while in AD brain the main decreases of CD200 labeling correspond to the hippocampus and inferior temporal gyrus. In agreement with previous data obtained from C57/BL6 mice [49] not only neurons but also astrocytes were positive for CD200 labeling in the human brains [181]. CD200R mRNA expression was also reduced in AD brains, specifically in regions affected by the pathology in comparison with matched no disease brains. In LPS models of neuroinflammation a decreased expression of CD200R has been described within the brain [185]. In the case of MS, no changes in CD200R were observed in brain lesions from MS patients [97], but in the experimental model of MS, TMEV-IDD a decreased expression of CD200 mRNA was observed at chronic phases of the disease [103]. It is suggested that a deficiency of CD200 in AD brain could contribute to maintaining chronic inflammation as has been hypothesized in MS, but also a deficiency in CD200R provokes a reduced efficiency of the CD200-CD200R complex to limit or control inflammation and therefore, the CD200-CD200R interaction in elderly brain might not be functioning in a correct way. Previous studies showed that the inhibitory response to CD200 is dependent on the level of expression of CD200R in responding cells [101]. Increasing the expression of CD200 and CD200R might provide a way of enhancing the efficiency of the system relevant from a therapeutic point of view [186]. Thus, the treatment with the CD200 fusion protein (CD200 Fc) was able to decrease microglial activation in the hippocampus of aged rats [187] and CD200 immunoadhesin diminished in a murine model of rheumatoid arthritis [188]. In this way, anti-inflammatory cytokines play a relevant role in regulating CD200-CD200R axis as IL-4 has been described to increase the expression of CD200 protein in neuronal cultures from rodents and IL-4-/- mice had lower levels of CD200 in the brain [189]. Similarly, IL-10 has been recently shown that augmented the expression of CD200 in neurons [103]. Additional studies revealed that the expression of CD200R was also strongly increased by IL-4 in microglia and macrophages and IL-13, that share common receptors with IL-4, exert a positive effect in increasing the expression of CD200R [181, 190]. In agreement with the above findings a significant age-related decline of IL-4 has been observed in the hippocampus of rats between 4 months and 22 months which correlated with increased levels of IL-1β [191]. In summary, the above reports suggest that the impairment of CD200-CD200R interaction may initiate a disturbed equilibrium in microglia–neuron interaction, in the way of sensitizing the pro-inflammatory reactivity of microglia and resulting in CNS injures or in the maintenance of chronic inflammatory states such as observed in AD and in the aging brain.

CONCLUSION AND FUTURE RESEARCH

The CNS innate immune response maintains a critical balance between the protective and potentially harmful effects of activating the innate immune system in CNS pathologies like MS and AD, as well as, in the aging brain that loss its homeostatic capacity with the consequent augmentation of morbidity and mortality. The balance between the protective and the deleterious effects of the innate immune response must be tightly regulated to favor conditions that support brain repair and the return to tissue homeostasis. The CD200/CD200R inhibitory immune ligand-receptors system constitutes one of the most suitable endogenous immunoregulatory molecule candidate to restore the immune suppressive status of the CNS altered in chronic neuroinflammatory situations. The elucidation of the immunoregulatory pathways shared between the CNS innate immune system and microglial cells and their crosstalk with neurons, oligodendrocytes and cells of the immune compartment represents an important challenge, but one that is of great therapeutic potential in preventing damage caused by macrophages/microglia in neuroinflammatory and neurodegenerative disorders.

ACKNOWLEDGEMENTS

REFERENCES