Background
One unique pathological change after primarily spinal cord injury (SCI) is the secondary injury, which is characterized by continuous tissue loss, reactive astrogliosis and chronic inflammation, and usually leads to gradual expansion of the lesion center and formation of a spinal cavity [
1,
2]. Elucidating the mechanisms of tissue loss, particularly nerve cell death is important for preventing the expansion of the lesion area. Previous studies have paid attention to the glutamate-induced apoptosis of neurons and oligodendrocytes within and around the lesion center [
3‐
5]. However, how reactive astrocytes, which are the major component of the glial scar, play diverse roles in SCI [
6] and are particularly important in supporting neuronal survival [
6], are eliminated remains poorly investigated. Understanding the mechanism of astrocytic death post-SCI may yield new insights into understanding the mechanism of secondary SCI and improving functional recovery.
Chronic inflammation plays an essential role in stimulating astrocyte activation and progressive cavitation [
7,
8]. After SCI, inflammation is mainly generated by activated microglia/macrophages, which are constituted by their two phenotypically distinct subpopulations, the pro-inflammatory M1, and the anti-inflammatory M2 microglia/macrophages [
9]. Compared to the immune reaction in peripheral tissue injury, the polarization of microglia/macrophages post-SCI is M1 predominant and lasts longer [
10‐
12]. Previous study has revealed the apoptosis-inducing effects of M1 microglia/macrophages on neurons and oligodendrocytes [
11]. Whether and how the activity of M1 microglia/macrophages affects the survival of reactive astrocytes remains unclear.
In the present study, we analyzed the death of reactive astrocytes in mice and human after spinal contusion, and reported that reactive astrocytes die through necroptosis, a type of programmed necrosis for which the molecular mechanisms have been recently unraveled [
13], and is induced by M1 microglia/macrophages, partially via TLR/MyD88 signaling. Our data indicated that blocking the necroptosis of reactive astrocytes might reduce secondary injury and promote functional recovery after SCI.
Discussion
Necrosis has been traditionally thought to account for the acute cell loss post-SCI and be uncontrollable [
31]. Recent progress in the field of cell death has identified a novel type of programmed necrosis, necroptosis [
23], and unveiled the underlying molecular mechanism, which is mediated by an intracellular RIP1/3/MLKL signaling cascade [
32,
33], thereby offering an opportunity for re-examining necrosis after SCI. Previous studies have reported protective effects of Nec-1 on SCI in rats without knowing the cell types that Nec-1 targets [
34,
35]. Our
in vivo PI-labeling showed that astrocyte was the major type of cells that undergo necrosis after SCI. The ultrastructural localization of RIP3 and MLKL on the cytoplasmic glial fibrils confirmed the astrocytic necroptosis in SCI. Interestingly, PI-positive astrocytes persisted for 2 weeks in injured spinal cord, indicating that chronic necrosis may be an important contributor of cavity formation post-SCI. We recently reported that microglia/macrophages undergo necroptosis after SCI [
36], which was consistent with our observation that RIP3 was primarily expressed by reactive astrocytes, and secondarily expressed by microglia/macrophages (Fig.
2b, c). Considering the fact that inflammation plays critical roles in the cavity formation after SCI, the beneficial effects of necroptosis inhibition on SCI may be results from the protection of astrocytes, microglia/macrophages, as well as other cells.
The innate immune reaction produced by microglia/macrophages has been demonstrated to contribute to the cavity formation and enlargement after SCI [
7]. The destructive effects of activated microglia/macrophages were largely attributed to their M1 sub-group, which was activated quickly after SCI and expressed high levels of pro-inflammatory cytokines, including the well-studied necroptosis inducing factor TNFα [
23,
37]. It is therefore reasonable to speculate a link between M1 micorglia/macrophages and astrocyte death after SCI, which has been poorly investigated. Our data showed that
in vitro, conditioned medium of M1 microglia/macrophages could induce necroptosis of astrocytes.
In vivo, depletion of M1 microglia/macrophages by GdCl
3 or transplantation of M1 macrophages can reduce or enhance necroptosis of astrocytes respectively. These results indicated a critical role of M1 microglia/macrophages in inducing the necroptosis of astrocytes after SCI. Considering that GdCl
3 also affects neutrophils [
38], which are abundant in the injury epicenter after SCI, and that iNOS can also be expressed by neutrophils [
39]. The beneficial effects of GdCl
3 treatment may also be contributed by the inhibition of neutrophils.
Although the identities of death factors released by M1 microglia/macrophages remain unclear, our data showed that TLR4 and MyD88 were up-regulated in necroptotic astrocytes after SCI and M1 CM could increase the expression of TLR2, TLR4 and MyD88 in astrocytes
. Previous researches have reported that TLR4 is involved in the necroptosis of macrophages
in vitro and in the activation of astrocytes after SCI [
40,
41]. These results indicated that M1 microglia/macrophages may induce the necroptosis of astrocytes by activating TLR4/MyD88 signaling. In consistent, inhibiting MyD88 could partially block the necroptosis-inducing effect of M1 microglia/macrophages
in vitro. In addition, the expression of TLR4/MyD88 in human necroptotic astrocytes indicated a common response of this signaling pathway after SCI. Considering that MyD88 mediates the downstream signals of multiple TLRs, and TLZ stimulates both the expression of TLR2 and TLR4
in vitro, the involvement of other TLRs in the M1 microglia/macrophages-induced astrocytic death is not excluded. It has been demonstrated that necroptotic cells release factors that modulate inflammation [
42], whether necroptotic astrocytes could regulate the chronic inflammation after SCI is of interest to be further investigated.
As the major component of glial scar, reactive astrocytes exhibit heterogeneous properties and exert multi-faceted functions in SCI, such as providing nutritive and metabolic support for neurons, inhibiting axonal growth and modulating inflammation [
43,
44]. Our data showed that necroptotic astrocytes were less supportive for neuronal survival, and inhibiting astrocytic necroptosis could rescue the neurotrophic function of reactive astrocytes, thereby reducing cavity area and promoting the survival of neurons surrounding lesion center which otherwise underwent apoptosis during the secondary injury [
45,
46]. It is still unknown how the properties of reactive astrocytes change when necroptotic signaling is activated. Nevertheless, our data, for the first time, have revealed a novel mechanism for the astrocytic death after SCI, implying that astrocytic necroptosis may be manipulated for preventing secondary SCI in the future.
Acknowledgments
This work was supported by grants from the National Natural Science Foundation of China (NSFC, 81571224, 31271583,) to Dr. Ya-Zhou Wang, NSFC (81371364) to Dr. Gong Ju, and NSFC (81272072) to Dr. Lequn Shan. And research fellowship from Fourth Military Medical University (2013D09) to Dr. Hong Fan. The RIP3−/− mice were received from Dr. Dixit at Genentech. We appreciate the technical assistances from Drs. Junjun Kang, Jialei Yang and Haifeng Zhang. The authors thank Drs. Biswas Sangita, Fuzheng Guo and Wenbin Deng (University of California at Davis) for language editing.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
HF performed most experiments, collected and analyzed data, and wrote the paper. KZ contributed to cell culture. KC contributed to morphological quantification. SL and KZ contributed to analysis of human spinal tissues. FK contributed to TLR/MyD88 study. HM contributed to the studies of RIP3−/− mice. GJ and YW designed the experiments, analyzed data, provided financial support and wrote the paper. All authors read and approved the final manuscript.