PKR deficiency alters E. coli-induced sickness behaviors but does not exacerbate neuroimmune responses or bacterial load

David Chun-Hei Poon, Yuen-Shan Ho, Ran You, Hei-Long Tse, Kin Chiu, Raymond Chuen-Chung Chang

Systemic inflammation induces neuroimmune activation, ultimately leading to sickness (e.g., fever, anorexia, motor impairments, exploratory deficits, and social withdrawal). In this study, we evaluated the role of protein kinase R (PKR), a serine-threonine kinase that can control systemic inflammation, on neuroimmune responses and sickness. Wild-type (WT) PKR+/+ mice and PKR-/- mice were subcutaneously injected with live Escherichia coli (E. coli) or vehicle. Food consumption, rotarod test performance, burrowing, open field activity, object investigation, and social interaction were monitored. Plasma TNF-α and corticosterone were measured by ELISA. The percentage of neutrophils in blood was deduced from blood smears. Inflammatory gene expression (IL-1β, TNF-α, IL-6, cyclooxygenase (COX)-2, iNOS) in the liver and the brain (hypothalamus and hippocampus) were quantified by real-time PCR. Blood and lavage fluid (injection site) were collected for microbiological plate count and for real-time PCR of bacterial 16S ribosomal DNA (rDNA). Corticotrophin-releasing hormone (CRH) expression in the hypothalamus was also determined by real-time PCR. Deficiency of PKR diminished peripheral inflammatory responses following E. coli challenge. However, while the core components of sickness (anorexia and motor impairments) were similar between both strains of mice, the behavioral components of sickness (reduced burrowing, exploratory activity deficits, and social withdrawal) were only observable in PKR-/- mice but not in WT mice. Such alteration of behavioral components was unlikely to be caused by exaggerated neuroimmune activation, by an impaired host defense to the infection, or due to a dysregulated corticosterone response, because both strains of mice displayed similar neuroimmune responses, bacterial titers, and plasma corticosterone profiles throughout the course of infection. Nevertheless, the induction of hypothalamic corticotrophin-releasing hormone (CRH) by E. coli was delayed in PKR-/- mice relative to WT mice, suggesting that PKR deficiency may postpone the CRH response during systemic inflammation. Taken together, our findings show that (1) loss of PKR could alter E. coli-induced sickness behaviors and (2) this was unlikely to be due to exacerbated neuroimmune activation, (3) elevated bacterial load, or (4) dysregulation in the corticosterone response. Further studies can address the role of PKR in the CRH response together with its consequence on sickness.