Fig 2: Serum and brain cytokine levels in tumor necrosis factor (TNF)-injected pups at P5 and P16. Serum cytokine levels at P5: n = 16 Vehicle (Veh), n = 16 TNF (20 mg/Kg) (A). Brain cytokine levels at P5: n = 11 Veh, n = 11 TNF (B). Serum cytokine levels at P16: for TNF, n = 30 Veh, n = 31 TNF. For IL-1β, IL-6, IL-10, IFN-γ, IL-5, and CXCL1, n = 20 Veh, n = 21 TNF (C). Brain cytokine levels at P16: n = 11 Veh, n = 12 TNF (20 mg/Kg) (D). For panels (A–D), data are presented as means ± SEM. Mann-Whitney’s U-tests: *p < 0.05, **p < 0.01, and ****p < 0.0001. Only statistically significant differences are presented.

Fig 3: Tumor necrosis factor (TNF) impacts reflex acquisitions in pups but not overall growth and developmental milestones. Timeline of the experiment (A). P1 to P16 body mass (BM) of pups injected with TNF (20 mg/Kg) or vehicle, presented as means ± SEM, note overlapping curves and reduced SEM (B). GEE estimates, 95% CI and p-values associated with the effects of treatment, time and cohort on BM: p(Treatment) = 0.2183, p(Time) < 0.0001, p(Cohort) < 0.0001 (C). Ambulation abilities scored from 0 to 3 in pups aged P6–P12, presented as means ± SEM, note overlapping curves and reduced SEM (D). GEE estimates, 95% CI and p-values associated with the effects of treatment, time and cohort on ambulation abilities: p(Treatment) = 0.5604, p(Time) < 0.0001, and p(Cohort) = 0.0216 (E). Ear development of pups aged P3–P5: a score of 0, 1 or 2 was set according to the number of ears everted/animal, presented as means ± SEM (F). GEE estimates, 95% CI and p-values associated with the effects of treatment, time and cohort on ear development: p(Treatment) = 0.729, p(Time) < 0.0001, p(Cohort) < 0.0001 (G). Eye opening of pups aged P12–P16: score of 0, 1 or 2 was set according to the number of eyes opened/animal, presented as means ± SEM (H). GEE estimates, 95% CI and p-values associated with the effects of treatment, time and cohort on eye opening: p(Treatment) = 0.212, p(Time) < 0.0001, p(Cohort) < 0.0001 (I). Righting reflex latency to turn of pups aged P2–P8, presented as means ± SEM (J). GEE estimates, 95% CI and p-values associated with the effects of treatment, time and cohort on righting reflex acquisition: p(Treatment) = 0.0308, p(Time) < 0.0001, p(Cohort) = 0.840 (K). Acoustic startle reflex of pups aged P10–P14: a score of either 0 or 1, where 1 is given to pups with startle and 0, to pups without startle, presented as means ± SEM (L). GEE estimates, 95% CI and p-values associated with the effects of treatment, time and cohort on the acoustic startle acquisition: p(Treatment) = 0.0332, p(Time) < 0.0001, p(Cohort) = 0.0101 (M). n = 29–30 PBS, n = 29–32 TNF depending on the test. *p < 0.05, ****p < 0.0001. Only statistically significant differences are presented.

Fig 4: IRE1 kinase activity does not affect insulin signaling through Akt. Phosphorylation of Akt (pAkt) (Ser473) relative to total Akt (tAkt) was measured by immunoblotting in 3T3-L1 adipocytes in the absence or presence of 1 μM KirA6. Cells were pretreated with or without KirA6 for 1 h before addition of ligand for 48 h. Cells were treated with or without 100 nM insulin for the final 10 min after cells were exposed for 48 h to: A) PGN (10 μg/ml), B) thapsigargin (1 μM), C) isoproterenol (2 μM), D) LPS (500 ng/ml), or E) TNF (10 ng/ml). Control (C), FK565 (FK), KirA6 (K), thapsigargin (Th or Thap), isoproterenol (Iso), lipopolysaccharide (LPS), tumour necrosis factor (TNF). Values are mean ± SEM (N = 6–8). Statistical significance was measured as p < 0.05 using two-way ANOVA. Post hoc analysis was performed using Tukey’s multiple comparisons test. Conditions with different letters (a, b) denote a statistical difference compared with all other conditions without the same letter.

Fig 5: Inflammatory ligands stimulate lipolysis via IRE1 kinase not RNase activity.A–C, glycerol release rates (μM/h) in 3T3-L1 adipocytes treated with increasing concentrations of the IRE1 inhibitor KirA6 (A; N = 4), the IRE1 RNase inhibitor KirA8 (B; N = 16), or the IRE1 RNase inhibitor 4μ8C (C; N = 12–16). D, expression of spliced Xbp1 in 3T3-L1 adipocytes treated with 1 μM Thapsigargin for 3 h with and without 10 μM KirA8. E, glycerol release rates in differentiated 3T3-L1 adipocytes treated with or without 1 μM KirA6 (IRE1 inhibitor) and one of vehicle (mock treatment), 10 μg/ml FK565 (PGN), 1 μM thapsigargin (ER stress activator), 2 μM isoproterenol (β-adrenergic agonist), 500 ng/ml LPS (TLR4 activator), or 10 ng/ml TNF, as indicated (N = 12–14). F, glycerol release rates in 3T3-L1 adipocytes treated with 1 μM KirA6 and 0.5 mM 8-Br-cAMP (N = 10–12). G, phosphorylation of IRE1 was measured in 3T3-L1 adipocytes after stimulation with vehicle, PGN (10 μg/ml), or thapsigargin (1 μM) for 3 h, with or without 1 μM KirA6 (N = 4–6). Blots were stripped and reprobed for total IRE1. H, quantification of western blots for the ratio between phosphorylated IRE1 and total IRE1, expressed relative to the basal control sample. I–K, RNA isolated from 3T3-L1 adipocytes was treated for 3 h with vehicle, FK565 (PGN, 10 μg/ml), thapsigargin (1 μM), or isoproterenol (2 μM) in the absence or presence of KirA6 (1 μM). Expression of BiP (I), CHOP (J), and spliced Xbp1(K) (N = 12). Values are mean ± SEM. Statistical significance was measured as p < 0.05 using two-way ANOVA. Post hoc analysis was performed using Tukey’s multiple comparisons test. Conditions with different letters (a, b) denote a statistical difference compared with all other conditions without the same letter.