Serum IGF-1 is reduced after TBI
Serum IGF-1 is lower in patients with mild TBI than those with severe TBI; The serum IGF-1 content in moderate and severe TBI at each time point was significantly lower than that of the mild TBI group (p < 0.05). Serum Tau progressively reduced after TBI in the first week, while it began to increase on the seventh day, but it was still remaining higher than the control group at six months post injury (p < 0.05, Fig. 1).
Serum IGF-1 protein was significantly higher in poor outcome group (GOS 1–3) that that in good outcome group (GOS 4–5, p < 0.05; Fig. 1). Serum IGF-1 content at 1, 3, 5, 7, and 14 days after injury (acute and subacute stage) had a significant positive correlation with MoCA score even at 6 months after injury (R2 = 0.2484, p < 0.001).Serum IGF-1 in normal cognition group is higher than it in abnormal cognition group at acute stage of TBI patients as well (Fig. 1).
Astrocytic IGF-1 has neuroprotective role in TBI models
Previously we have shown astrocytic IGF-1 protects against excitotoxic neurons both in-vivo and in-vitro. Therefore, we used astrocytic IGF-1 to treat lateral fluid percussion injury (LFPI) mouse model (done at Shanghai Health College). Mice were tested in the water maze to assess cognitive function 12 weeks after injury. During the acquisition session of the water maze, mice given a FPI and treated with astrocytic IGF-1 displayed significantly faster search times than their vehicle-treated counterparts. There were no significant effects on the measure of crossing time and during in the targeted zone during water maze acquisition period (P > 0.05). Furthermore, the astrocytic IGF-1 improved both cognition and motor function in TBI mice by Morris water maze and NSS assessment (Fig. 2). TBI mice had higher NSS; while astrocytic IGF-1 could reduce it.
iTRAQ and bioinformatic analysis in IGF-1 treated TBI mice
Next, we applied iTRAQ proteomics to investigate the neuroprotective effect of astrocytic IGF-1. According to the iTRAQ method published previously, we compared the targeting proteins with a two-fold alteration in TBI vs. Sham group with vectors and TBI + IGF-1 vs. TBI + vector. We identified 189 candidate proteins after IGF-1 treatment. First, we did a David Analysis with an online tool (https://david.ncifcrf.gov) and clustered these candidate proteins based on KEGG analysis (Supp Fig. 1). We listed top 20 cluster and KEGG pathway in Fig. 3. The Top 20 Cluster pathway (David analysis)showed Hydrolase activity (GO:0016787 ~ hydrolase activity) is the first one, including 16 related proteins (Q9CYC6, Q9CR30, Q50L41, Q9R001, Q6NSR8, Q8CGB6, Q9WUZ9, Q9ET22, Q91ZX6, P35821, P52479, Q8CHE4, Q9CXY9, P97470, P07146, Q14BV6) and the top KEGG pathway is membrane protein (Table S2).
DCP2 and related pathway changes after TBI
DCP2(Q9CYC6)is found to increase in TBI group with more than 2-fold change, and IGF-1 treatment could reduce its expression by 1.4 times. Dcp2 codes m7GpppN-mRNA hydrolase in mammalians [23]. M7G is a 5’capping structure in mRNA and also in miRNA. The METTL1 (another m7G decapping enzyme) has been found to promote the maturation of several miRNAs via disrupting the immature structure for miRNA [24]. And METTL1 knockdown could reduce the expression of a serious of miRNAs including let 7e with a most decreased level [24]. Here, we predicted DCP2 is supposed to be decapping the m7G structure of let 7e as well to promote its maturation and checked the expression of DCP2 and IGF-1R in cell models.
To validate this, we did an in-vitro analysis with an excitotoxic model, which can mimic the brain injury in cells. First we observed KA treated neurons had decreased MAP2 and beta-tubulin expression indicating reduced maturation of neurons; while co-culture with astrocytes could promote the neuronal growth and against the excitotoxic effect with increased expression of both MAP2 and beta-tubulin to validate our previous findings and the successful modeling of in-vitro excitotoxic injury (Supp Fig. 2). Next, we found both DCP2 and let 7e expression increased in KA treated neurons; while coculture with astrocytes could reduce both. These findings were consistent with our TBI in vivo results. And this effect was abolished partly by IGF-1R antagonist, PI3K inhibitor and p38 inhibitor (Fig. 3).
Mir let-7e has previously been reported to regulate IGF-1R [25]. In our study, we did a Targetscan analysis for the binding prediction between IGF-1R 3’UTR and let-7e. Further, we confirmed this with luciferase assay. Next, the dual luciferase reporter gene assay with the psicheck2-based IGF-1R‐wt plasmid containing miR‐let-7e binding site was conducted to further verify this prediction. The activity of the luciferase following cotransfection of IGF1R‐wt and miR‐let-7e mimic was lower than the activity following cotransfection of IGF-1R‐wt and NC (p < 0.05), indicating that there was a regulatory relationship between miR‐let-7e and 3′‐UTR of IGF-1R (Fig. 3). These findings helped to verify that miR‐let-7e targeted IGF-1R.
Mitophagy of brain injury in-vitro and in-vivo
It is reported that IGF-1R can regulate the autophagy and IGF-1 can promote mitophagy via activating AMPK [11]. To confirm the IGF-1 role in mitophagy following brain injury, we did both in-vitro and in vivo analysis of mitophagy. This study further evaluated whether the increased neuronal loss in KA-stressed cells was accompanied by of a loss of mitochondrial potential and if astrocytic IGF-1 could affect this phenomenon. The cells were analyzed with the JC-10 dye that forms red J-aggregates in controlled primary cultured cells but stays a green monomer in cells that have lost mitochondrial integrity. The scatter plots show that majority of the cells treated with KA shifted towards green fluorescence when compared to controls (Fig. 4). Remarkably, in KA treated neurons cocultured with astrocytes, a population shift to the red channel was observed indicating preservation of mitochondrial potential. In addition, this effect was partly abolished by IGF-1R antagonist, PI3K inhibitor and p38 inhibitor. Consequently, KA-stressed cells maintained mitochondrial potential when cocultured with astrocytic IGF-1 (Fig. 4).
In addition, we extracted the mitochondria proteins from the cultured cell to assess the mitophagy following KA. We found KA increased the expression of Atg8, Atg 5, Atg 14, Beclin-1and Lamp while decreased the p62SQSTM1 level. This effect is partly abolished by AG1024, PI3K inhibitor and p38 MAPK inhibitor (Fig. 5).
For the in-vivo study, we applied the IHC to analyze the mitophagy marker (PINK1 and NIX) in the rodent TBI brain. As shown in Fig. 6, FPI in mice induced the decreased NIX and PINK1 expression (shown as average optical density of brown area) in the ipsilateral cortex of TBI mice at three months after injury and IGF-1 treatment could reverse these.