MΦ/MG have a diverse range of functions during CNS diseases depending on the type of induction caused by unique cytokine stimuli. Although many papers have reported that the MΦ/MG play a role in the induction of inflammation and neural cell death by releasing pro-inflammatory cytokines and producing oxidative stress, recent evidence also suggests the a phenotype of MΦ/MG contributes to the repair and regenerative process after the diseases [3–5]. In the present study, we demonstrated the MΦ/MG activating phenotypes 14 days after SCI. Moreover, to observe the influence of IL-1, we compared the lesion size and MΦ/MG activation using IL-1 KO mice. Our studies clearly showed that IL-1 KO mice have a smaller lesion size and less motor deficit than the wild-type mice. Interestingly, although IL-1 KO mice had a suppressed TNFα level, an inflammatory marker, from the 1st dpo, the animals also had a decreased Ym1 level which is an alternative activating MΦ/MG marker at the 7th and 14th dpo. To confirm the phenomenon, we established adult mouse primary MG cultures, and examined cell responses to the cytokines IFNγ and IL-4 directly with and without IL-1β. These results suggest that IL-1 might participate in the classical and alternative activation of MΦ/MG.
Previous reports have suggested a contribution of IL-1 in acute CNS diseases such as SCI [13, 14, 16, 36], cerebral ischemia [19–21], trauma [37, 38], and subarachnoid hemorrhage . However, no direct evidence from IL-1 KO mice has demonstrated the contribution of IL-1 to SCI. Our results were consistent with previous studies that IL-1 or IL-1 receptor signaling pathway contributes to increase lesion size of the SCI. An increase in IL-1β and a decrease in IL-1ra were observed after SCI, and IL-1β administered into the spinal cord impaired locomotion. Moreover, administering IL-1ra into the spinal cord reduced IL-1β levels and locomotion recovered . IL-1 consists of two molecular subtypes, IL-1α and IL-1β . IL-1α is expressed continuously while IL-1β is inducible in response to injury. We used IL-1α and β KO mice because it has been reported that IL-1α -or β-alone KO mice do not give rise to the neuroprotective phenotype after ischemia . It has also previously been shown that a post-traumatic neuroinflammatory response was involved in the development of injury, and that IL-1 worked as a key inflammatory player that mediated the neuroinflammatory response . This is supported by results showing that IL-1 recruits monocytes to enhance the inflammation mediated by IL-1 receptor I and via a MyD88-dependent pathway . Down-regulation of the IL-1 receptor pathway and IL-1-mediated inflammatory responses becomes a strategy for the suppression of SCI. To this extent, several recent studies have shown that IL-1 receptor antagonists are able to reduce the severity of symptoms after experimental SCI [14, 36, 43, 44]. From these results, we confirmed that our experimental condition did not differ obviously from the previous one and that a deficiency of IL-1 worked as a suppressor of SCI.
We then undertook experiments to determine whether IL-1 deletion affected the inflammatory response. Time-dependent changes in the levels of the pro-inflammatory cytokines IL-1β and TNFα in the spine after SCI were measured with ELISA. It has been reported that pro-inflammatory cytokines including IL-1β, IL-6, and TNFα are induced rapidly following SCI [45–49]. In the present study, IL-1β levels in the spine of wild-type mice after SCI were drastically increased from the 1st dpo and were sustained until the 14th dpo. Immunohistochemical studies suggested that IL-1β was expressed in MΦ and/or MG. The results suggested that IL-1 contributes to the inflammatory responses after SCI. The TNFα level in the wild-type mice was increased in the spine from the 1st dpo and was sustained until the 14th dpo. However, the TNFα level in IL-1 KO mice did not increase after injury and remained at significantly lower levels during the experimental period compared with that seen in wild-type mice. The results suggest that IL-1 participates in the upregulation of expression of TNFα, probably following the induction of a series of inflammation events after injury.
Recently, it has been suggested that IL-1β triggers the proliferation and early differentiation of neural progenitor cells during development of the spinal cord and after hippocampal injury [23, 24]. Moreover, other inflammatory factors such as TNFα and iNOS were implicated in aspects of neural regeneration during wound-repair [26, 27]. In the present study, we found that IL-1β levels remained high up to the 14th dpo despite a decrease in the size of the lesion site. We then postulated that IL-1β might be performing in a different role during part of this period, and, therefore, carried out immunoblotting experiments to examine Ym1 levels in response to SCI. Ym1 has been reported as an excellent marker of alternative activation of MΦ and/or MG , which is one of the activation phenotypes induced by IL-4 and IL-13 [3–5] and plays an important role in the resolution of inflammation and promotion of wound healing [3, 50]. Alternative activating MΦ/MG gene expression increases during the sub-acute stage after SCI . The alternative activation of MΦ promotes axonal growth and overcomes inhibitory substrates . MΦ implanted into the injured spinal cord increase axonal regrowth and/or functional improvement [9, 10, 51, 52]. Immunoblotting for Ym1 revealed higher levels at the 7th and 14th dpo in wild-type mice than in IL-1 KO mice, with immunoreactivity concentrated around the lesion epicenter in injured spinal cord. The Ym1 immunoreactivity coincided with that of immunoreactivity for F4/80 and the growth factor IGF-1, which is known to increase alternative activation of MΦ/MG and plays an important role in neuroprotection [11, 53, 54]. We postulated that IL-1 might contribute to Ym1 expression, and to the induction of alternative activation. Taken together, these results suggest that IL-1 increases the inflammatory response and might also increase tissue repair and anti-inflammatory resolution via the induction of alternative activation of MΦ/MG in response to SCI. Unfortunately, we were unable to differentiate between MG and MΦ because there is no specific immunohistochemical marker available to separate them.
Then, we established adult mouse primary MG cultures and examined cell responses to the cytokines IFNγ and IL-4. Moreover, we added IL-1β to this system to observe its effect because we could not detect endogenous IL-1β in the media after exposing cells to either IFNγ or IL-4 alone. We have previously reported that NOx and TNFα levels in the media of primary cultures of the mouse MG BV-2 line increased in response to exposure to IFNγ alone . Other studies using rodent primary MG obtained from the pups and the BV-2 cell line have also shown an increased expression of inflammatory mediators (TNFα, IL-1β, IL-6, COX-2) and iNOS after MG stimulation by IFNγ and LPS [55–57]. In the present study, while the level of TNFα increased in response to IFNγ treatment, NOx did not. However, NOx was drastically increased by co-treatment with IFNγ and IL-1β; iNOS levels as determined by immunoblotting behaved similarly. Moreover, other alternative activation markers such as arg 1 (activity and protein level), IGF-1, Ym1 and CD206 [3, 7, 54] did not increase upon exposure to IFNγ in the presence or absence of IL-1β. These results indicate that MG polarizes to the classical activating phenotype by IFNγ and/or IL-1β [3, 53]. Some minor differences with other studies exist, with discrepancies perhaps due to differences in the source and type of cells and experimental conditions used.
By contrast, MG exposed to IL-4 showed an increase of arginase activity, as well as increased arg-1, IGF-1, Ym1 and CD206 protein levels, but not NOx, iNOS or TNFα. These characteristics clearly indicated that the MG polarized to the alternative activating phenotype [3, 53]. Surprisingly, co-treatment of MG with IL-4 and IL-1β further increased arg-1 activity, and arg-1 and Ym1 protein levels towards the alternatively activated phenotype. Because treatment of MG with IL-1β alone did not increase these factors, it is suggested that IL-1β has a supportive effect on IL-4-induced responses and supports the induction of the alternative activating phenotype in adult mouse MG. However, another alternative factor, CD206 was not enhanced and IGF-1 tended to decrease following IL-1β co-treatment with IL-4. The co-treatment of MG with IL-4 and IL-1β gave rise to an unexpectedly high TNFα level as well. Because exposure of MG to IL-4 alone did not increase the level of TNFα, the co-treatment is considered to be the result of a synergistic effect between IL-1β and IL-4. To date, no evidence has been reported to show that IL-4 works as an enhancer of the IL-1β response. IL-4/IL-13 has basically been considered to antagonize the IL-1β function  by enhancing the production of IL-1ra and the decoy IL-1β type II receptor [59, 60]. Moreover, IL-4/IL-13 downregulated the pro-IL-1β cleavage enzyme, caspase 1, to convert it to an active mature form [61, 62]. However, a small number of papers have reported that an alternative activating phenotype is classified into sub-phenotypes. A sub-phenotype of MΦ, M2b is influenced by IL-1β. It has been reported that M2b induces TNFα and IL-10 production [54, 63, 64]. However, the main role and phenotype of M2b remain unclear. Moreover, there is no evidence to show that IL-4 participates in the polarization of this phenotype. Different reactions of alternative markers by co-treatment of IL-4 and IL-1β might be due to sub-phenotypes of alternative activating MG. Further studies are needed to clarify the relation between the cytokine network and MG polarization.
Finally, we determined the possible involvement of IL-4 and IL-13 in the adult MG alternative activating response. Many research and review articles have indicated that both IL-4 and IL-13 function similarly as activators of alternatively polarized MΦ [7, 35]; however this has not been studied in detail in adult MG. We applied IL-4, IL-13, or IL-4/IL-13 to primary cultures of adult MG with and without IL-1β to demonstrate a putative signaling mechanism for MG alternative activation, and found that Ym1, arg-1 and CD206 were enhanced by IL-4 and IL-4/IL-13, but not by IL-13 alone. Because the levels of induction between IL-4 and IL-4/IL-13 were very similar, we thought that the effect of induction depended on IL-4. Moreover, even if MG cultures were co-treated with IL-13 and IL-1β, the Ym1 and arg-1 did not further increase in the same manner as for IL-4. Two IL-4 receptors, type I IL-4 receptor (IL-4RI) and type II IL-4 receptor (IL-4RII), mediate IL-4's functions [7, 35]. IL-4RI is exclusive for IL-4, while IL-4RII binds both IL-4 and IL-13. IL-4RI is expressed predominantly in hematopoietic lineage cells and IL-4RII is expressed in hematopoietic and non-hematopoietic cells . Although we did not determine the expression of the IL-4 receptors, the present results suggest that adult MG are polarized to the alternatively activated phenotype by IL-4 but not by IL-13, and that some MG functions might be mediated through IL-4RI. Further analyses are required to determine what IL-4 receptor(s) is involved in the present phenotypes, and what differences exist between MG and MΦ in this respect.
Alternatively activated MΦ are now regarded as a continuum of functionally and phenotypically related cells, with a critical role in the resolution and tissue repair phases [3, 7, 54]. Indeed, it has been reported that immune cells contribute to the maintenance of neurogenesis and spatial learning abilities in adulthood because immune-deficient mice showed impaired hippocampal neurogenesis that could not be enhanced by environmental enrichment . Previously, it was suggested that IL-1β itself contributed to the proliferation and differentiation of neural progenitor cells in the spine and hippocampus, and to nerve regeneration by promoting neurite outgrowth following nerve injury [23, 24, 66]. A mixture of murine recombinant IL-1β, IL-6 and TNFα administered to the lesioned spinal cord four days after the lesion significantly decreased the amount of tissue loss seven days after trauma compared with vehicle-administered controls . Moreover, gene-deficient mice have been used to show that TNFα and iNOS are implicated in neural regeneration during wound-repair stages [26, 27]. This accumulated evidence lends itself to the suggestion that the relationship between IL-1 and IL-4 and the alternative activation of MG might be implicated in neurogenesis. The manner in which the enhancement of alternative activation markers following co-treatment with IL-4 and IL-1β contributes to wound healing, repair and neurogenesis needs to be examined more in detail, as does the way in which immune/inflammatory responses tune the switching to resolution and regeneration following SCI and in other CNS diseases.
In conclusion, we have demonstrated here in in vivo experiments that IL-1 exacerbates the effects of SCI by accentuating the impact of the inflammatory responses. Moreover, the results of in vivo and in vitro experiments suggest that IL-1 participates in the classical and alternative activation of MG. Finally, we suggest that the alternative activation of adult MG is regulated via an IL-4 signaling pathway that could be mediated by IL-4RI.