Spinal cord injury (SCI) disrupts communication within the nervous system, leading to the loss of the motor/sensory function (Van Middendorp et al., 2011; Mcdonald and Sadowsky, 2002). While rehabilitation can enhance partial spontaneous functional recovery, no effective treatment for chronic SCI patients has yet been developed (Hutson and Di Giovanni, 2019; Kobayakawa et al., 2019a; Shah et al., 2013). In the chronic phase of SCI, the glial scar acts as a physical barrier to regeneration and produces chondroitin sulfate proteoglycans (CSPGs), which function as a chemical barrier to regeneration (Tran et al., 2018; McKeon et al., 1991).

Astrocytes, the main cellular components of the glial scar, undergo a typical change of hypertrophy after SCI, showing process extension and increased glial fibrillary acidic protein (GFAP) expression and presenting a characteristic morphology known as reactive astrocytes (RAs) (Hasel and Liddelow, 2021; Liddelow and Barres, 2017; Shinozaki et al., 2017). These RAs then overlap each other and subsequently transform into scar-forming astrocytes (SAs) that make a glial scar and inhibit the regeneration of axons, leading to limited recovery after SCI (Okada et al., 2006; Silver and Miller, 2004).

In contrast, in experiments that ablate reactive astrocytes or glial border astrocytes using transgenic mice or conditional ablation of astrocytes after SCI, increased leukocyte infiltration, progressive demyelination, and increased neuronal cell death were observed (Voskuhl et al., 2009; Sofroniew, 2005; Faulkner et al., 2004). In addition, experiments using mice specifically conditioned for reactive astrocytes showed no spontaneous axonal regeneration after severe SCI, either by blocking glial scar formation or by attenuation of scar astrocytes, or by the removal of astrocytes in the chronic phase (Anderson et al., 2016), suggesting diverse roles of astrocytes in the SCI pathology from the acute to the chronic phase.

We previously reported that the transformation of RAs into SAs could be stopped by blocking the integrin-N-cadherin pathway and that RAs could revert in retrograde to naïve astrocytes (NAs) by changing the extracellular environment (Hara et al., 2017). These findings suggest the possibility of therapy with an anti-integrin antibody targeting RAs in the acute/sub-acute phase. However, whether or not SAs in the chronic phase can change their morphology and gene expression phenotype by manipulating their extracellular environment is unclear.

In this study, we clarified the time course of changes in gene expression profiles in astrocytes after SCI by performing a selective gene expression analysis of NAs, RAs, SAs, and chronic-phase astrocytes (CAs) with laser microdissection (LMD). By manipulating the extracellular environment of scar-forming astrocytes in vitro, we investigated the change in the expression of SA marker genes and the reversibility of the morphological transformation characteristic of SAs. In particular, we focused on the influence of the integrin-N-cadherin pathway on the plasticity of SAs, as this pathway stops the transformation of RAs into SAs (Hara et al., 2017). Furthermore, we conducted transplantation of SAs into naïve spinal cord and analyzed the morphological/biological characteristics of transplanted astrocytes, indicating the dichotomic character of SAs regarding homeostasis and plasticity.