A requirement of serotonergic p38α mitogen-activated protein kinase for peripheral immune system activation of CNS serotonin uptake and serotonin-linked behaviors

Abstract

Alterations in central serotonin (5-hydroxytryptamine, 5-HT) neurotransmission and peripheral immune activation have been linked to multiple neuropsychiatric disorders, including depression, schizophrenia and autism. The antidepressant-sensitive 5-HT transporter (SERT, SLC6A4), a critical determinant of synaptic 5-HT inactivation, can be regulated by pro-inflammatory cytokine signaling. Systemic innate immune system activation via intraperitoneal lipopolysaccharide (LPS) injection rapidly elevates brain SERT activity and 5-HT clearance. Moreover, the pro-inflammatory cytokine interleukin (IL)-1β rapidly stimulates SERT activity in raphe nerve terminal preparations ex vivo, effects that are attenuated by pharmacological p38 MAPK inhibition. To establish a role of serotonergic p38α MAPK signaling in LPS/IL-1β-induced SERT regulation and attendant behavioral responses, we pursued studies in mice that afford conditional elimination of p38α MAPK in 5-HT neurons (p38α(5HT-)). We found p38α(5HT-) and control (p38α(5HT+)) littermates to be indistinguishable in viability and growth and to express equivalent levels of SERT protein and synaptosomal 5-HT transport activity. Consistent with pharmacological studies, however, IL-1β fails to increase SERT activity in midbrain synaptosomes prepared from p38α(5HT-) animals. Moreover, although LPS elevated plasma corticosterone and central/peripheral pro-inflammatory cytokines in p38α(5HT-) animals, elevations in midbrain SERT activity were absent nor were changes in depressive and anxiety-like behaviors observed. Our studies support an obligate role of p38α MAPK signaling in 5-HT neurons for the translation of immune activation to SERT regulation and 5-HT-modulated behaviors.

Figure 1

Activation and elimination of p38 MAPK in dorsal raphe (DR) 5-hydroxytryptamine (5-HT) neurons. Images of 5-HT and p38 MAPK staining for lipopolysaccharide (LPS) studies were collected from the dorsomedial division of the DR nucleus from adult male C57BL/6J mice. Images of the conditional p38α MAPK elimination experiments were obtained in the same manner. (a) Peripheral administration of LPS (0.2 mg kg⁻¹ 1 h before being killed) elevates levels of phospho-p38 MAPK immunoreactivity in 5-HT-labeled neurons of the adult mouse dorsomedial division of the DR. Scale bar, 10 μm. (b) Conditional elimination of p38α MAPK immunoreactivity. Immunofluorescence for p38α MAPK is presented for floxed p38α MAPK mice without (p38α^5HT−) or with (p38α^5HT+) expression Cre recombinase in 5-HT neurons via ePET::Cre as described in Materials and Methods. Scale bar, 5 μm.

Figure 2

Serotonergic expression of p38α MAPK is required for acute interleukin (IL)-1β-induced SERT activation. (a) High-performance liquid chromatography (HPLC) analysis of 5-hydroxytryptamine (5-HT) levels in the midbrain (MB), forebrain (FB), hippocampus (Hip) and striatum (Str) from male p38α^5HT+ and p38α^5HT− mice. Two-way analysis of variance (ANOVA) indicates significant region (F_3,67=71.75, P<0.0001) and genotype (F_{1, 67},=4.12, P=0.001) effects and a nonsignificant interaction (F_3,67=0.53, P=0.67). N for all regions=10 for p38α^5HT+, N=8 for p38α^5HT−. (b) Basal firing rate of dorsal raphe 5-HT neurons as assessed by cell-attached recordings of serotonergic neurons in acute midbrain slices from p38α^5HT+ and p38α^5HT− mice, N=10 neurons from three p38α^5HT+ mice, N=15 neurons from three p38α^5HT− mice, with data derived from two to three slices per animal. Student's two-tailed t-test reveals no significant genotype effect. (c) SERT protein expression in MB and FB of p38α^5HT+ and p38α^5HT− mice assessed by western blotting. Two-way ANOVA indicates no significant region or genotype effects. N=7 for all groups. (d) Transport of [³H]-5HT (50 nM) in MB synaptosomes of p38α^5HT+ and p38α^5HT− mice following treatment with IL-1β (10 ng ml⁻¹), compared with vehicle. Two-way ANOVA indicates significant drug × genotype interaction (F_{1, 20}=6.68, P=0.02), drug (F_{1, 20}=5.62, P=0.02), and genotype (F_{1, 20}=6.68, P=0.02). Bonferroni post hoc shows a significant effect of IL-1β in lipopolysaccharide (LPS)-treated p38α^5HT+, but not p38α^5HT− mice, compared with vehicle control counterparts. *P<0.05; **P<0.01, N=6 for all groups.

Figure 3

Lipopolysaccharide (LPS) induction of central and peripheral pro-inflammatory cytokines, as well as plasma corticosterone, occurs independently of serotonergic p38α MAPK expression. Quantitative PCR (qPCR) analyses of mRNA expression were assessed 1 h post saline (i.p.) or LPS (0.2 mg kg⁻¹, i.p.) injection. (a, b) Two-way analysis of variance (ANOVA) demonstrates significant LPS-induced elevation of spleen interleukin (IL)-1β (F_{1, 21}=20.43, P<0.001) and tumor necrosis factor (TNF)-α (F_{1, 28}=29.20, P<0.001) independent of genotype (nonsignificant genotype and interaction). N=5–11 per group. (c, d) Two-way ANOVA demonstrates significant LPS-induced elevation of midbrain IL-1β, (F_{1, 24}=8.70, P<0.01) and TNF-α (F_{1, 21}=42.63, P<0.01) independent of genotype. N=5–8 per group. (e) Two-way ANOVA demonstrates significant LPS-induced elevation of serum corticosterone (CORT); drug effect F_{1, 24}=15.56, P<0.01, independent of genotype. N=4–8 per group. *P<0.05; **P<0.01; ***P<0.001, Bonferroni post hoc tests.

Figure 4

Serotonergic expression of p38α MAPK is required for lipopolysaccharide (LPS) stimulation of SERT activity and depressive/anxiety-like behaviors. (a) [³H]-5-HT (5-hydroxytryptamine) uptake (50 nM) in midbrain synaptosomes of p38α^5HT+ and p38α^5HT− males 1 h post intraperitoneal (i.p.) saline or LPS (0.2 mg kg⁻¹) injection. LPS increased SERT activity in midbrain synaptosomes of p38α^5HT+, but not p38α^5HT–, mice. Two-way analysis of variance (ANOVA) significant effect of genotype (F_{1, 23}=4.31, P<0.05) and LPS (F_{1, 23}=7.12, P<0.05), Bonferroni post hoc *P<0.05 in p38α^5HT+ group only. N=6–8 per group. (b) Assessment of SERT activity in mice with adult excision of p38α MAPK. Slc6a4-ER-Cre;p38α MAPK^loxP/loxP mice were treated for 5 days with corn oil or tamoxifen as described in Methods. Four weeks later, mice were administered saline or LPS (0.2 mg kg⁻¹ i.p.) 1 h before being killed. SERT activity was measured in synaptosomes prepared from midbrain (MB), forebrain (FB), hippocampus (Hip) and striatum (Str). The percentage change of 5-HT uptake activity for LPS versus saline controls is plotted for each condition. Two-way ANOVA demonstrates significant LPS effect on SERT activity in p38α^ER5HT+ (N=8 for MB and FB; N=9 for Hip; N=4 for Str)-treated animals (F_{3, 55}=8.56, P=0.005) but not in p38α^ER5HT− (N=12 for MB; N=10 for FB; N=7 for Hip; N=4 for Str) animals (P<0.05). (c) Percent time immobile in the forced swim test (FST) for p38α^5HT+ and p38α^5HT- mice 1 h post saline or LPS (0.2 mg kg⁻¹ i.p). Two-way analysis of variance (ANOVA) shows significant interaction between LPS and genotype (F_{1, 15}=7.01; P=0.02). Immobility time was increased in p38α^5HT+, LPS-treated mice (*P<0.05, Bonferroni post hoc; N=5 per for saline and LPS-treated animals), but not p38α^5HT- mice (N=5 for saline-treated and N=4 for LPS-treated animals). (d) Percent time immobile in the tail suspension test (TST) for p38α^5HT+ and p38α^5HT− mice 1 h post saline or LPS (0.2 mg kg⁻¹ i.p.). Two-way ANOVA shows significant interaction between treatment and genotype (F_{1, 17}=7.83; P=0.01). Immobility time was increased in p38α^5HT+ (N=5 per for saline and LPS-treated animals), LPS-treated mice (*P<0.05, Bonferroni post hoc), but not in p38α^5HT− mice (N=6 for saline-treated and N=5 for LPS-treated animals). (e) Open, closed and total number of arm entries in the EPM in p38α^5HT+ (+) and p38α^5HT− (−) mice treated with saline or LPS (0.2 mg kg⁻¹ i.p.). There was no significant effect of genotype or drug. (f) Percent time spent in the open arms of the EPM (6-min test). Two-way ANOVA shows significant effects of LPS (F_{1, 17}=8.81; P<0.01) and genotype (F_{1, 17}=4.71; P=0.04). LPS reduced time spent in open arms in p38α^5HT+ (N=5 in saline- and LPS-treated animals) but not in p38α^5HT− (N=6 and 4 in saline- and LPS-treated animals, respectively) mice (*P<0.05, Bonferroni post hoc).