Background
Ischemic brain injury elicits an inflammatory response that involves activation of microglia and monocyte-derived macrophages in the central nervous system (CNS). Within the lesion, these cells adopt unique molecular phenotypes with either protective or detrimental effects on neuron survival [
1]. Despite this basic understanding, the mechanisms controlling functional diversity in microglia/macrophages remain elusive but undoubtedly include differential signaling via chemokines, cytokines, and other cues present in the ischemic microenvironment.
The chemokine (C-X3-C motif) ligand 1 (CX3CL1)/ CX3C chemokine receptor 1 (CX3CR1) signaling pathway has been shown to play an important role in the maintenance of neural-immune communication in health and disease [
2,
3]. Fractalkine (CX3CL1), a membrane-bound chemokine, is expressed predominantly by neurons in the healthy CNS [
4,
5] while its receptor (CX3CR1) is highly expressed on resident brain microglia and on peripheral immune cell populations including macrophages [
6,
7]. CX3CR1 selectively modulates microglia activity in response to its ligand CX3CL1 [
6]. CX3CL1/CX3CR1 also regulates recruitment of circulating leukocytes to sites of injury [
8,
9].
Under neuroinflammatory conditions, loss of signaling via CX3CR1 has been shown to have either protective [
10,
11], or detrimental effects [
12], or no effects at all [
13,
14]. These conflicting observations indicate the complexity and disease-specific regulation of neuron-microglia communication via CX3CL1 and CX3CR1. In regards to acute CNS injury models (transient and permanent brain ischemia, spinal cord injury), the collective data suggest that the absence of CX3CR1 significantly reduces ischemic damage and inflammation [
15‐
17].
Although the processes of microglia activation, cytokine production, and phagocytosis have been studied in several animal models of ischemia injury in CX3CR1-deficient (CX3CR1
-/-) mice, it remains unclear as to how CX3CR1 deficiency alters the function of these cells, resulting in reduced damage during ischemic injury (highlighted by several controversial findings). Following induction of focal cerebral ischemia within CX3CR1
-/- mice, no difference was observed in the number of IL-1β-expressing microglia; rather, decreased leukocyte infiltration is involved in the development of smaller infarcts [
15]. Conversely, in an animal model of spinal cord injury, modest but significantly more recruited monocytes (CD45
hi) accumulate in the spinal cord of CX3CR1
-/- mice by 3 days post-injury [
16]; however, a decrease in the number of CD11b
+/Ly6C
lo/iNOS
+ macrophages after 7 days post-injury was associated with reduced neuropathology and enhanced functional recovery in these CX3CR1-deficient mice [
16]. To date, little is known about the CX3CL1/CX3CR1 pathway in the context of microglia activation following brain ischemia.
In the current study, using a murine model of transient middle cerebral artery occlusion (MCAO), we tested the consequences of the absence of CX3CR1 on microglia/macrophage proliferation/recruitment and on their neurotoxicity after brain ischemia. We find that loss of CX3CR1 signaling attenuates the monocyte recruitment, the expansion of resident microglia and newly recruited monocytes, and reduced their inflammatory capacity after MCAO, contributing to neuroprotection in CX3CR1-deficient mice.
Discussion
The work presented here provides important in vivo evidence for the role of the CX3CL1/CX3CR1 signaling pathway in the activation and neurotoxicity of microglia/ macrophage in cerebral ischemia. Using CX3CR1-/- mice, in which the CX3CR1 gene is substituted with the gene for the GFP on both alleles (CX3CR1GFP/GFP), and a murine model of transient brain ischemia, we find that CX3CR1 deficiency resulted in a decrease in the size of ischemic lesion, a decrease in the number of apoptotic cells (predominantly neurons), marked deregulation of post-ischemic brain inflammatory responses (including ROS and pro-inflammatory cytokines), a significant decrease in proliferation of microglia and infiltrating macrophages in the ischemic lesion, and a marked decrease in the levels of IL-1β/IL-6/TNF-α expressed by microglia/macrophage in ischemic brain. Collectively, these effects due to CX3CR1 deficiency correlate with improved neurological function following MCAO and suggest that blockade of CX3CR1/CX3CL1 signaling may provide neuroprotection against ischemic injury.
Absence of fractalkine receptor in CX3CR1
-/- mice has been previously reported to be protective in ischemia. Critically, significantly smaller (56%) infarcts and blood–brain barrier damage have been observed in CX3CR1
-/- animals compared with CX3CR1
+/- and WT mice 72 hours after a 60-minute MCAO [
15]. CX3CR1 deficiency has also been shown to induce protection from a 30-minute MCAO starting as early as 24 hours [
17]. In an experimental model of spinal cord injury, CX3CR1
-/- mice demonstrate neuroprotection and functional recovery 5 days post-injury [
16]. In agreement with these previous studies on the role of the CX3CL1/CX3CR1 pathway in acute CNS injury, our work finds that mice lacking CX3CR1 are protected from ischemic injury 72 hours after a 90-minute MCAO. Interestingly, we find prevention of exacerbation of ischemic lesion was accompanied with reduced cleaved Caspase 3 positive neurons in the peri-infarct area at 72 hours after initial stroke in CX3CR1
-/- mice compared to WT mice, inconsistent with a previous report [
15]. While these observed differences could be attributed to the differences in the experimental designs, the outcomes between our study and others are identical. Further, it may also imply that the effects of CX3CL1/CX3CR1 signaling may be dependent upon the context of the specific disease state.
CX3CR1 deficiency has been suggested to attenuate tissue damage and improve recovery of function by reducing the recruitment and/or the activation of microglia and macrophages [
15,
28‐
30]. We tested this hypothesis by isolating mononuclear cells from the injured brain of WT and CX3CR1
-/- mice, followed by quantifying CD45
+/CD11b
+ cells by flow cytometry. We find a significant decrease in CD45
+/CD11b
+/Ly6G
– microglia/macrophage in the ipsilateral hemisphere of CX3CR1
-/- mice vs WT mice, which was confirmed by immunohistochemical staining with Iba-1 in brain sections. To determine whether this was a result of suppressed microglia proliferation or decreased monocyte recruitment, the relative expression of CD45 has been used to differentiate resident microglia (CD45
low) from recruited monocytes (CD45
hi) [
31]. CD45
low/CD11b
+/Ly6G
– cells (microglia) predominate in the contralateral side of both genotypes. However, by 72 hours post-injury, accumulation of both CD45
low (microglia) and CD45
hi (macrophage/activated microglia) CD11b
+/Ly6G
– cells dramatically decreased in the ischemic hemisphere of CX3CR1
-/- mice. Moreover, BrdU assays show that the proliferation of microglia (CD45
low) is significantly impaired in CX3CR1
-/- MCAO mice. The proliferation of CD45
hi cells (including activated microglia) was also suppressed; however, not to the extent which could fully compensate for the decreased number of CD45
hi cells seen in CX3CR1
-/- MCAO mice compared to WT MCAO mice. We therefore speculate that the migration of monocyte-derived macrophages from the periphery was inhibited. These data together indicate that both microglia/macrophage proliferation and macrophage recruitment are attenuated by CX3CR1 deficiency within the MCAO model. Our results coincide with a previous study that demonstrated decreased leukocytes infiltration as a possible mechanism for the development of smaller infarcts in CX3CR1
-/- mice after a 60-minute MCAO [
15]. Conversely, more CD45
hi cells were found accumulated in the injured spinal cords of CX3CR1
-/- mice. Although the role of this cell population is not clear (rather, this study revealed a distinct monocyte subset), CD11b
+/Ly6C
lo/iNOS
+ macrophages were associated with reduced neuropathology and enhanced functional recovery [
16]. Indeed, it has long been thought that macrophage (microglia) are not a uniform cell population, such that specific environmental signals, including effector molecules, timing of activation, and degree of injury, may induce their different polarization states [
32,
33]. Therefore, phenotypically distinct monocyte populations appear to be associated with tissue pathology and repair in ischemia and, thus, warrant further study [
16].
Previous reports indicated a marked increase in proinflammatory cytokine levels that peak 12 to 24 hours after ischemic injury [
34]. Our results obtained from the WT mice are in agreement with these temporal dynamics of cytokines. As expected, an initial increase of TNF-α, IL-1β, and IL-6 was detected in the ipsilateral hemisphere after MCAO. However, contrary to the levels of proinflammatory cytokines observed in control samples, we show that the ablation of proliferating microglia cells and recruitment of monocytes, associated with CX3CR1 deficiency, resulted in a significant decrease in the levels of IL-1β, TNF-α, and IL-6, suggesting an important temporal deregulation of the proinflammatory cytokine response. Similar patterns of mRNA expression and temporal deregulations were observed in the NF-κB signaling pathway, a general marker of inflammatory response in cerebral ischemia [
16]. Furthermore, using flow cytometry, we confirmed that the trends of proinflammatory cytokines in the ipsilateral hemisphere of the MCAO brain in CX3CR1
-/- mice are consistent with the levels secreted by CX3CR1
-/- microglia/macrophage isolated from MCAO brains. Therefore, it appears that microglia/macrophage may play an important role in the regulation/modulation of the proinflammatory responses in cerebral ischemia, and their functions are attenuated by CX3CR1 deficiency.
Variations of cell morphology or cell displacement are related to the specific activity of microglia/macrophage [
32]. Following ischemia, activated microglia/macrophage can potentially exert either a protective or detrimental effect, suggesting that these cells may acquire different phenotypes belonging to the classical (M1) or to the alternative (M2) active status [
35]. M1 activation is generally referred to as the pro-inflammatory and cytotoxic phenotype, characterized by nitric oxide, ROS, and proinflammatory cytokine production, while the M2 phenotype is an alternative activation state, associated with a fine tuning of inflammation, scavenging of debris, promotion of angiogenesis, tissue remodeling and repair [
17,
36]. Fumagalli and colleagues [
17] reported that, in CX3CR1
-/- mice, protection from ischemia at early time points after injury is associated with a protective inflammatory milieu, characterized by the promotion of M2 polarization markers. We found that activated M1-like Iba-1
+ cells, which have shorter and thicker processes and bigger cell bodies, were visualized in the WT brain section, while ramified M2-like Iba-1
+ cells were predominantly located in the CX3CR1
-/- brain (Figure
3A) with increased expression of M2 markers (Figure
3E), implying that suppressed activation of microglia/macrophage by CX3CR1 deficiency may be responsible for the neuroprotective effects in the CX3CR1
-/- mice.
Competing interests
The authors declare that they have no competing interests.
Authors’ contribution
ZT and YG designed the experiments. ZT, YG, QL, J-XY and QL performed the experiments. ZT and YG analyzed the data and wrote the paper. JS and F-DS supervised the experimental work. F-DS obtained the funding and resources. All authors read and approved the final manuscript.