Fever is a key response to infection, essential for eliminating harmful pathogens. During infection, the immune system communicates with the brain to propagate immune signals and induce sickness-related behaviors such as appetite loss, lethargy and warmth seeking (Quan and Banks, Brain Behav. Immun., 2007). In some children with autism spectrum disorder (ASD) — in which challenges in social communication is a core feature — fever has been reported to increase eye contact, speech and social interaction. These improvements, however, have mainly been described by parents anecdotally and by clinical case reports. How fever influences the brain circuitry underlying social behavior in ASD is not yet clear. SFARI Investigator Catherine Dulac has been working to better understand the neurobiological basis for the effect of fever in ASD, with support from a SFARI Research Award. In a recent study, Dulac and her collaborators identified a population of cells in the hypothalamus that are necessary for generating fever and influencing the activity of several brain areas associated with sickness symptoms (Osterhout et al., Nature, 2022).

To study fever, the researchers administered pro-inflammatory agents that mimic infection and increase body temperature in mice, and then used measures of gene expression to look at which populations of brain cells were activated and where they were distributed. After fever was induced, twelve brain areas showed an increase in activated cells, including areas involved in appetite, metabolism, temperature regulation, sleep, stress and fear responses. In particular, a population of cells in an area of the hypothalamus, called the ventral medial preoptic area (VMPO), was highly activated compared with controls. The hypothalamus is known to control body temperature and other major functions critical for homeostasis (Tan et al., Cell, 2016), but little was previously known about the fever-related role of the VMPO specifically.

In the study, many of the fever-activated neurons within the VMPO were localized near the vascular organ of lamina terminalis (VOLT), a specialized area lacking a blood-brain barrier and allowing access to circulating substances from outside the brain, including fever-related molecules from the immune system (Nakamori et al., Brain Res., 1993). VMPO neurons were also located near activated non-neuronal cells in and around the VMPO, including glial, endothelial and ependymal cells, which secrete immune molecules. The close proximity of VMPO neurons to these other non-neuronal cells led Dulac and her colleagues to hypothesize that neurons in the VMPO might be activated by nearby immune signals. Indeed, when infection was induced, the researchers observed a high level of expression of immune molecules involved in fever and inflammation — such as prostaglandin-endoperoxide synthase 2 (PTGS2), interleukin-1β(IL-1β), and chemokine ligand 2 (CCL2) — in non-neuronal cells near the VMPO, as well as expression of the corresponding immune receptors in neurons of the VMPO. Further investigation using electrophysiological measures showed that these immune molecules directly increase the excitability of, and indirectly increase excitatory input to, neurons in the VMPO. The findings suggest that, during infection and fever, neurons in the VMPO may be activated by immune signals secreted from nearby populations of non-neuronal cells (a type of cellular communication known as a paracrine mechanism).

Whether VMPO neurons are actually necessary and required for fever to occur was still an open question. To answer this question, Dulac and her colleagues selectively destroyed these neurons and then measured body temperature after inducing infection. Mice without a normally functioning VMPO cell population did not have an increase in body temperature when infection was induced (but did not have trouble maintaining their body temperature under normal conditions). The findings showed that VMPO neurons activated by infection were necessary for a fever to develop. These neurons were also needed to generate other sickness-related behavior such as warmth seeking and were involved in inducing a loss of appetite.

To find out exactly how infection-activated VMPO neurons alter body temperature and appetite, Dulac and her colleagues examined the connections that these neurons have with other brain areas. Using neuroanatomical tract-tracing and optogenetic techniques, they found that neuronal populations in the VMPO are synaptically and functionally connected to brain circuits known to control body temperature (including the anteroventral periventricular nucleus and median preoptic nucleus) and appetite (including the arcuate nucleus). This finding suggests that VMPO neurons can influence body temperature and appetite through their projections to these downstream brain regions.

Based on the study’s findings, Dulac and her colleagues proposed that the function of VMPO neurons during infection is to translate immune signals from the periphery into changes in brain activity to elicit sickness symptoms. This study offers new insights into how the brain alters body temperature and triggers sickness behaviors during infection. The characterization of brain circuitry underlying changes in behavior during fever is an important step toward understanding how fever may change social behavior in children with ASD.