MYD88-TLR4-Dependent Choroid Plexus Activation Precedes Secondary Brain Injury after Intracranial Hemorrhage

The functional outcome of patients with intracranial hemorrhage strongly relates to the degree of secondary brain injury (SBI) evolving within days after the initial bleeding. The spatiotemporal interplay of specific inflammatory processes within different brain compartments has not been sufficiently characterized, limiting potential therapeutic interventions to prevent and treat SBI.

We used a whole-blood injection mouse model to systematically characterize the spatial and temporal dynamics of inflammatory processes after intracranial hemorrhage using 7-Tesla magnetic resonance imaging (MRI), spatial RNA sequencing, functional BBB assessment, and immunofluorescence-average-intensity-mapping.

We identified a pronounced early response of the choroid plexus (CP) peaking at 12-24h, that was characterized by inflammatory cytokine expression, epithelial and endothelial expression of leukocyte adhesion molecules, and the accumulation of leukocytes. In contrast, we observed a delayed secondary reaction pattern at the injection site (striatum) peaking at 96h, defined by gene expression corresponding to perilesional leukocyte infiltration and correlating to the delayed signal alteration seen on MRI. Pathway analysis revealed a dependence of the early inflammatory reaction in the CP on toll-like receptor 4 (TLR4) signaling via myeloid differentiation factor 88 (MyD88). TLR4 and MyD88 knockout mice corroborated this observation, lacking the early upregulation of adhesion molecules and leukocyte infiltration within the CP 24h after whole-blood injection.

We report a biphasic brain reaction pattern after intracranial hemorrhage with a MyD88-TLR4-dependent early inflammatory response of the CP, preceding inflammation, edema and leukocyte infiltration at the lesion site. Pharmacological targeting of the early CP-activation might harbor the potential to modulate the development of SBI.

Fig. 1 Spatiotemporal gene expression after striatal whole-blood injection. (A) UMAP colored by the unsupervised clustering of the merged spatial transcriptome dataset (n = 6). (B) Gene set scores for specific anatomical compartments projected on the UMAP plot. TH, thalamus; CTX, cortex; CP, choroid plexus; STR, striatum; PAL, pallidum; HY, hypothalamus; MB, midbrain; P, Pons. (C) Mean anatomical gene set expression score and fraction of positive features of each cluster. (D) Spatial projection of the clusters on the control sample and corresponding H&E histology. (E, F) Top five differentially expressed genes in the CP (E) and STR (F) stratified by time after injection.

Fig. 2 Perilesional and CP transcription profiles after striatal whole-blood injection. (A) log2-normalized expression of selected genes (n = 6). (B–D) Spatial projection of the log2 normalized expression of Ifi27l2a, Vcam1, and Lyz2.E-G. Temporal dynamics of the log2 normalized expression of Icam1, Ccl20 and Ptprc in the CP.

Fig. 3 Dextran extravasation and inflammatory protein expression after striatal whole-blood injection. (A) Methodological steps of the semi-automated quantification of spatial protein expression (n = 21). (B–G) Temporal dynamics of the FITC-Dextran (70 kDa), CD45 or ICAM intensity. LV, lateral ventricle; 3rdV, third ventricle.

Fig. 4 MyD88-TLR4 dependence of the CP-response. (A) Gene-set enrichment patterns in the CP and the striatum (STR) (n = 6). (B, C) Temporal dynamics (B) and spatial projection (C) of the MyD88-dependent-TLR4-signaling gene set score 24 hours after injection (n = 6).D.-G. ICAM1 and F4/80 signal intensity 24h after striatal NaCl (sham, n = 6) or whole-blood (n = 6) injection in littermate (wt) or respective knockout mice (n = 6).

Publikationsverlauf

Artikel online veröffentlicht:
05. Februar 2024