Systemic LPS induces spinal inflammatory gene expression and impairs phrenic long-term facilitation following acute intermittent hypoxia

Abstract

Although systemic inflammation occurs in most pathological conditions that challenge the neural control of breathing, little is known concerning the impact of inflammation on respiratory motor plasticity. Here, we tested the hypothesis that low-grade systemic inflammation induced by lipopolysaccharide (LPS, 100 μg/kg ip; 3 and 24 h postinjection) elicits spinal inflammatory gene expression and attenuates a form of spinal, respiratory motor plasticity: phrenic long-term facilitation (pLTF) induced by acute intermittent hypoxia (AIH; 3, 5 min hypoxic episodes, 5 min intervals). pLTF was abolished 3 h (vehicle control: 67.1 ± 27.9% baseline; LPS: 3.7 ± 4.2%) and 24 h post-LPS injection (vehicle: 58.3 ± 17.1% baseline; LPS: 3.5 ± 4.3%). Pretreatment with the nonsteroidal anti-inflammatory drug ketoprofen (12.5 mg/kg ip) restored pLTF 24 h post-LPS (55.1 ± 12.3%). LPS increased inflammatory gene expression in the spleen and cervical spinal cord (homogenates and isolated microglia) 3 h postinjection; however, all molecules assessed had returned to baseline by 24 h postinjection. At 3 h post-LPS, cervical spinal iNOS and COX-2 mRNA were differentially increased in microglia and homogenates, suggesting differential contributions from spinal cells. Thus LPS-induced systemic inflammation impairs AIH-induced pLTF, even after measured inflammatory genes returned to normal. Since ketoprofen restores pLTF even without detectable inflammatory gene expression, "downstream" inflammatory molecules most likely impair pLTF. These findings have important implications for many disease states where acute systemic inflammation may undermine the capacity for compensatory respiratory plasticity.

Fig. 1.

Systemic inflammation (3 h) induced by LPS (100 μg/kg ip) significantly reduced acute intermittent hypoxia (AIH)-induced phrenic long-term facilitation (pLTF). A: representative integrated phrenic neurograms from anesthetized rats during the AIH (3 × 5 min hypoxia: Hx1, Hx2, Hx3) protocol for a vehicle-injected (saline, top trace), or LPS-injected (middle trace), or time control (no AIH, bottom trace) rats. Black dashed line indicates baseline phrenic amplitude in each trace. Development of pLTF is evident as a progressive increase in phrenic nerve amplitude over 60 min in the vehicle-injected animal. B: no change in the short-term hypoxia response was evident. C: group data for vehicle-injected AIH (n = 5), LPS-injected AIH (n = 5), and time control (n = 5) demonstrating a significant reduction in the magnitude of pLTF 60 min post-AIH in LPS treated and time control rats (***P < 0.001 repeated-measures two-way ANOVA, Tukey post hoc test).

Fig. 2.

Systemic inflammation (24 h) induced by LPS (100 μg/kg ip) did not alter the short-term hypoxia response, significantly reduced AIH-induced pLTF, but pLTF was restored with the anti-inflammatory drug ketoprofen (12.5 mg/kg ip, 3 h). A: representative integrated phrenic neurograms from anesthetized rats during the AIH (3 × 5 min hypoxia) protocol for vehicle-injected (saline, top trace), LPS-injected (second trace), LPS + ketoprofen-injected (third trace), and time control (no AIH, bottom trace) rats. Black dashed line indicates baseline phrenic amplitude in each trace. Development of pLTF was evident as a progressive increase in phrenic nerve amplitude over 60 min. B: no change in the short-term hypoxia response was evident. C: group data showing pLTF for vehicle-injected (n = 5) and ketoprofen-injected (n = 6) rats with AIH, and a reduction in pLTF in rats injected with LPS (n = 9). The appearance of pLTF was restored in rats injected with LPS and after treatment with ketoprofen (n = 6). There was no increase in phrenic nerve amplitude in time control rats (n = 15). (***P < 0.001, **P < 0.01 indicates significant difference from vehicle, ketoprofen, and 24 h LPS + ketoprofen).

Fig. 3.

Systemic inflammation induced by LPS (100 μg/kg ip) caused a transient increase in inflammatory gene expression in the spleen. LPS (3 h) caused a significant increase in all inflammatory genes (n = 3) examined compared with the respective vehicle control (n = 4) but returned to baseline levels by 24 h (n = 3) and was not altered with ketoprofen (n = 2) (*P ≤ 0.001 different from all other treatment groups).

Fig. 4.

Systemic inflammation evoked by LPS (100 μg/kg ip) caused transient and differential changes in inflammatory gene expression in isolated microglia (black bars) and homogenates (gray bars) from the cervical spinal cord. A: treatment with LPS (3 h) increased mRNA for iNOS compared with vehicle (microglia n = 15, homogenates n = 14) in both microglia (n = 8) and homogenate (n = 8) samples. Expression of iNOS was reduced 24 h post-LPS (microglia n = 7, homogenates n = 6) compared with 3 h post-LPS in both sample types but was not changed relative to vehicle. After ketoprofen (12.5 mg/kg ip, 3 h), microglia had greater iNOS gene expression (n = 8) compared with homogenates (n = 7). B: treatment with LPS (3 h) increased COX-2 mRNA in both microglia (n = 7) and homogenate (n = 8) compared with vehicle (microglia n = 15, homogenates n = 15), but was reduced 24 h post-LPS. LPS (24 h) alone (microglia n = 7, homogenates n = 6) or with ketoprofen (microglia n = 8, homogenates n = 7) did not alter COX-2 mRNA in either microglia or homogenates compared with vehicle. After ketoprofen treatment, microglia had less COX-2 mRNA compared with homogenates. LPS treatment (3 or 24 h) had no effect on gene expression for TNFα (C) or IL-1β (D). **P < 0.01, ***P < 0.001, significant difference from vehicle; @@P < 0.01, @@@P < 0.001, significant difference from 3 h LPS; #P < 0.05, significant difference between microglia and homogenate samples.