Microglial Nhe1 cKO mice exhibited accelerated sensorimotor and cognitive function recovery after TBI
We demonstrated in our recent report that our Nhe1 cKO mouse line successfully deleted NHE1 protein expression exclusively in the IBA1+ microglia/macrophages, but remained unchanged in other cell types (21). Survival rate and neurological behavior functions in Ctrl and Nhe1 cKO mice were monitored during 1–30 days post-TBI (Fig. 1a). Neither Sham Ctrl or Sham cKO mice displayed any mortality, while Ctrl and cKO TBI mice exhibited < 10% mortality (Fig. 1b). Ctrl and cKO TBI mice showed similar contusion volume initially at 3 days post-TBI (Fig. 1c). However, the unbiased analysis of NeuN+ neuronal counts by automatic cell counting (using the “count particles” module in ImageJ) revealed significantly higher NeuN+ cell percentages in both CL and IL peri-lesion areas of the cKO brains than the Ctrl brains (Fig. 1d). In neurological behavioral assessments, Ctrl and cKO sham animals similarly showed a brief elevation of sensorimotor deficits at 1 day post-sham, but quickly returned to baseline at 3 days post-procedure (Fig. 1e-f). In comparison, the Ctrl TBI mice exhibited significantly prolonged contact and removal time (~ 20-fold and 10-fold, respectively), as well as significantly more errors in the foot fault test (~ 4-fold) at 1–7 days post-TBI (Fig. 1e-f). However, compared to the Ctrl TBI mice, the Nhe1 cKO mice showed significantly accelerated sensorimotor function recovery during 5–14 days post-TBI, and completely returned to their baseline levels by day 14 (Fig. 1e-f). In testing working memory using the Y-maze test at 30 days post-TBI, the cKO mice exhibited a significantly higher spontaneous alternation rate (~ 73%) than the Ctrl mice (~ 46%, p < 0.0001), indicating a stimulated working memory function in the cKO mice (Fig. 1g). However, the two groups displayed similar locomotor activities reflected by total arm entries (Fig. S1a). Taken together, these neurological function assessment tests demonstrate that the Nhe1 cKO mice exhibited better neurological function (sensorimotor and cognitive) recovery after TBI.
Microglial Nhe1 cKO mice displayed improved white matter resistance against TBI-induced apoptosis and inhibition of oligodendrogenesis
As white matter integrity is important for restoring neurological functions after TBI (2, 8), we tested whether the improved functional outcomes of the cKO mice were in part due to their increased tolerance to TBI-induced damage and/or boosted white matter repair. Myelin basic protein (MBP), a marker for white matter myelination, was used to assess corpus callosum (CC) tract integrity in the Ctrl and cKO brains after sham or TBI procedures. Interestingly, the cKO brains exhibited a thicker CC (midline, same bregma level, Fig. S2) at 24 h after sham procedure (Fig. 2a, p < 0.05), indicating a possible role of microglial NHE1 protein in regulating white matter integrity homeostasis. At 1 d post-TBI, the cKO brains showed a significantly higher CC thickness than the Ctrl brains (Fig. 2a, p < 0.001). Further analysis of Olig2+ oligodendrocyte lineage cells at 3 days post-TBI revealed significantly elevated NG2+Olig2+ OPCs, Ki67+Olig2+ proliferative OLs, and reduced Caspase3+Olig2+ apoptotic OLs in both hemispheres of the cKO brains, compared to Ctrl brains (Fig. 2b, p < 0.05). Importantly, these cKO brains also exhibited increased expression of H3K9me3 in the Olig2+ OLs, a post-translational histone modification marker for OPC differentiation (24) (Fig. 2b, p < 0.01). Moreover, analysis of APC+ mature OLs counts showed that TBI did not affect OL survival in the CL hemispheres of either Ctrl or cKO brains, but induced an immediate decrease of the APC+ mature OLs in the IL hemisphere of the Ctrl mice at 1 d post-TBI (Fig. 2c-e, p < 0.01). In contrast, the cKO TBI brains were resistant to such a loss in the IL hemisphere (Fig. 2c-e). The Ctrl TBI mice continued to lose mature OLs in both CL and IL CC at 3 days post-TBI, while the TBI-induced reduction of mature OLs in the cKO CC was delayed (Fig. 2c-e). Interestingly, by 30 days post-TBI, the mature OLs in Ctrl mice failed to regenerate, while the cKO brains exhibited significantly elevated counts of mature OLs in both hemispheres of the CC tracks (Fig. 2e, p < 0.01). These findings strongly suggest that deletion of microglial NHE1 protein not only provided resistance to the white matter damage induced by TBI, but also promoted oligodendrogenesis by increasing their progenitor cell proliferation and differentiation into mature myelinating OLs.
Selective deletion of microglial Nhe1 increased microglial anti-inflammatory phenotype activation in TBI brains
To understand the underlying mechanisms of the increased oligodendrogenesis in the post-TBI cKO brains, we examined the profiling of microglia and infiltrated myeloid cells in Ctrl or Nhe1 cKO brains at 3 days post-TBI with flow cytometry (Fig. 3a). No difference in the percentage of microglial cells (CD11b+CD45lo) and infiltrated myeloid cells (CD11b+CD45hi) were detected in CL or IL hemispheres of the Ctrl and Nhe1 cKO mice (Fig. 3b, p > 0.05). Further probing of the expression of pro-inflammatory markers CD16/32 and CD86 did not show any differences between Ctrl and cKO microglia and/or myeloid cells (Fig. 3c-d). However, the percentage of anti-inflammatory CD206-positive microglia/myeloid cells and Ym-1hi-positive microglia were significantly increased in the IL hemisphere of the cKO brains, compared to the Ctrl (Fig. 3c-d, p < 0.05). Further characterization of microglial cells or reactive astrocytes in the peri-lesion cortex of Ctrl or cKO brains by immunofluorescent staining revealed that TBI induced significant increases in GFAP+ astrocytes and IBA1+ microglia/macrophage counts in both the Ctrl and cKO brains at 3 days post-TBI (Fig. 3e, p < 0.05), but, the cKO brains showed significantly attenuated GFAP+ astrocyte and IBA+ microglia/macrophage counts were significantly attenuated in the peri-lesion area of the cKO brains, compared to the Ctrl (Fig. 3e, p < 0.001). Taken together, these findings demonstrate that selective deletion of microglial Nhe1 gene promotes the restorative activation of microglia/myeloid cells and reduces astrogliosis in the cKO brains.
Selective deletion of microglial Nhe1 altered inflammation-related transcriptome profile in microglia/macrophages after TBI
To understand how deletion of microglial Nhe1 affects microglial restorative function, we performed bulk RNA sequencing (RNA-seq) of CD11b+ cells isolated from the CL and IL hemispheres of Ctrl and Nhe1 cKO brains at 3 days post-TBI (Fig. 4a). Unsupervised hierarchical clustering analysis demonstrated clear separation between CL and IL hemispheres of cKO and Ctrl brains (Fig. 4b). 123 differentially expressed genes (DEGs) were identified in the IL hemispheres of cKO mice, compared to Ctrl mice (fold change ≥ 1.2 or ≤ -1.2 and FDR q-value ≤ 0.05, Fig. 4c); among those, 56 genes were upregulated and 67 downregulated (Fig. 4d). Ingenuity Pathway Analysis (IPA) showed significantly altered enrichment pathways, including Th1 and Th2 activation pathways (Fig. 4e, p < 0.05), which is known to elicit inflammation in microglia and macrophages (25). Within these pathways, multiple pro-inflammatory genes were triggered by TBI in the Ctrl microglia/macrophages, but significantly decreased in the cKO microglia/macrophages, such as Psen2, Ifi206, Ifi207, and Igsf8 (Fig. 4d, f), all reported to be involved in microglia/macrophage-mediated inflammation (26–29). S1pr1 and Tpm3 genes are also involved in inflammation and expression of these two genes were reduced significantly in cKO IL hemisphere compared to ctrl IL hemisphere (Fig. 4f). Additionally, the cKO microglia/macrophages showed elevated expressions of Fcgr1, Gnb4, and B4galnt1 genes which stimulate anti-inflammatory activation (28, 30). On the other hand, the non-lesion CL hemispheres displayed 178 DEGs between Ctrl and cKO microglia/macrophages (Fig. S3a-b), with IPA analysis showing significantly altered IL-8 signaling pathway with reduced inflammatory genes such as Napepld, Vcam1, Rnd1 (31) in the cKO microglia/macrophages (Fig. S3b-d). Change of selected key pathway genes have been validated by qRT-PCR (Fig. S4). Taken together, our bioinformatic analysis reveals that deletion of microglial Nhe1 attenuates the expression of inflammation-related transcriptomes but stimulates restorative microglial activation transcriptome profiles after TBI.
Post-TBI administration of selective NHE1 inhibitor HOE642 accelerated neurological function recovery
We next explored the efficacy of targeting NHE1 protein with a pharmacological inhibition approach in reducing the TBI-induced functional deficits. Figure 5a illustrated our administration protocol of Veh (DMSO) or a potent NHE1 inhibitor HOE642 (0.3 mg/kg body weight/day, twice per day, i.p.) at 24 h post-TBI. Compared to the Veh-treated TBI mice, HOE642 administration did not affect mortality during 30 days post-TBI (Fig. 5b). However, the HOE642-treated TBI mice exhibited significant improvements in sensorimotor function (adhesive contact/removal test and foot fault test) during the 14 days post-TBI recovery period (Fig. 5c-d, p < 0.05). In assessing the same cohort of mice in cognitive function with Y-maze test at 30 days post-TBI, the HOE642-treated mice exhibited an improved trend of performance in spontaneous alternation rate than the Veh-treated mice (Fig. 5e, p = 0.09), indicating improved working memory function (32). These HOE-treated TBI mice also showed significantly increased locomotor activity reflected by their total arm entries (Fig. 5e, p < 0.05). These outcomes are consistent with the Nhe1 cKO mice shown in Fig. 1.
Characterization of the HOE642-mediated protective effects in TBI mice
Compared to the Veh-treated mice, the HOE-treated mice exhibited significantly smaller contusion volume and increased NeuN+ neurons at 3 days post-TBI (both CL and IL peri-lesion areas) (Fig. 6a-b, p < 0.05), indicating that post-TBI administration of HOE642 has neuroprotective effects. In line with the boosted white matter repair in the Nhe1 cKO mice, the HOE-treated brains showed increased oligodendrogenesis and differentiation properties in the CC white matter tracts compared to the Veh-treated brains at 3 days post-TBI (Fig. 6c, p < 0.05). Moreover, flow cytometry of CD11b+CD45lo microglia and CD11b+CD45hi myeloid cells from the Veh or HOE-treated brains (Fig. S5a-b) showed that the anti-inflammatory phenotype (CD206+) of microglia cells were selectively increased in the IL hemisphere of the HOE-treated brains (Fig. S5c, p < 0.05), while sparing the pro-inflammatory phenotypes of myeloid cells or microglial cells (Fig. S5c-d). Immunostaining further confirmed that IBA1+ microglia/macrophages and GFAP+ reactive astrocytes were significantly reduced in the peri-lesion cortex of HOE-treated brains (Fig. S5e, p < 0.0001). These findings collectively suggest that post-TBI administration of NHE1 protein inhibitor HOE642 is protective of neurons and reduced inflammatory responses and gliosis in the brain, which can concertedly contribute to the improved functional outcome post-TBI.
Long-term effects of NHE1 protein blockade on white matter integrity after TBI
We further assessed whether the Nhe1 cKO brains and HOE-treated brains exhibited long-lasting preservation of white matter integrity by 30 days post-TBI via MRI DTI of the ex vivo brains of the same cohort of mice after completing neurological function testing. The cKO mice exhibited significantly reduced brain lesion volume at 30 days post-TBI, compared to the Ctrl mice (Fig. 7a-b). Interestingly, both the Nhe1 cKO brains and HOE-treated brains displayed increased FA values in the CC and EC white matter tracts (both hemispheres) by 30 days post-TBI, compared to either Ctrl or Veh-treated brains (Fig. 7c, p < 0.05). Additionally, white matter demyelination is associated with higher radial diffusivity (RD) and medial diffusivity (MD), but not necessarily with reduced axial diffusivity (AD) (33, 34). We indeed detected lower RD and MD, but higher AD in the same cohort of either cKO or HOE-treated mice (Fig. S6). Taken together, these data collectively demonstrate that blocking microglial NHE1 protein has sustained long-term protective effects on white matter myelination, and reveals NHE1 protein as a potential therapeutic target for white matter repair after TBI.