Fig. 4: Microglial Tbk1 loss dampens the response to systemic LPS, but does not modify motor neuron damage.

a Setup for RNAseq of microglia 24 h after systemic (i.p.) LPS injection, isolated from spinal cords of 4-month-old mice with microglial specific Tbk1 deletion (Tbk1-µG-KO, KO), compared to control littermates (Tbk1-µG-WT, WT). b PCA reveals clustering of samples by treatment. Each point (sample) corresponds to microglia from 2 pooled spinal cords. n = 4 samples per genotype and age, females only. c GSEA showing top regulated immune (Hallmark) pathways between Tbk1-µG-KO and Tbk1-µG-WT microglia, after LPS treatment (adjusted p values < 0.05, NES ≥ ± 1.5, see “Methods”). LPS effect on each pathway in control condition is represented by a red up arrow if activated in WT-LPS vs unstimulated. d Setup to assess the effect of microglial Tbk1 deletion during chronic MN degeneration. Mice with microglial specific Tbk1 deletion were crossed to SOD1G93A ALS mice, to produce SOD1G93A mice with one (G93A/Tbk1-µG-HET) or two (G93A/Tbk1-µG-KO) Tbk1 copies deleted, compared to littermate SOD1G93A mice with control microglial Tbk1 (G93A/Tbk1-µG-WT). e–g Kaplan–Meier plots of ages reached at disease onset (e), early disease (f), and survival age reached at end-stage (g). n = 20-23 mice per genotype (equally sex-mixed). Mean ages in days (d) ± SEM. h Setup to assess the effect of microglial Tbk1 deletion during acute MN injury (unilateral sciatic nerve crush/regeneration) in 4-month-old Tbk1-µG-KO, Tbk1-µG-HET, and littermate control Tbk1-µG-WT mice. Shown are representative toe-spread images (Catwalk) before and after major recovery. i Recovery of the sciatic functional index (SFI) over 4 weeks after crush. n = 6-7 mice per genotype (males only). Adjusted p values are indicated and were determined by mixed effect models. j Recovery was also assessed long-term (over 17 weeks after crush) by measuring the compound muscle action potential (CMAP) using electromyography. n = 5 mice per genotype. Adjusted p values are indicated and were determined by repeated measure two-way ANOVA. a, d, h Created in BioRender. Lenoel, I. (2025) https://BioRender.com/ct8qnd6. i, j Data are shown as means ± SEM. ns not significant.