Our immune and nervous systems influence each other, both locally and at a distance (1–4). Locally, immune responses in the central nervous system (CNS) include activation of resident glia cells and macrophages (3–6), and infiltration of circulating immune cells (5). Many responses rely on cytokines secreted by immune cells for communicating not only to immune cells but also to neurons and glia to control synaptic pruning (7), neuroplasticity (8), and neuroprotection (3). Molecules important in the immune system, such as major histocompatibility complex (MHC) proteins, are also expressed by neurons and glia and likely contribute to neural function. Can the nervous system communicate with the immune system through neurotransmitters—chemicals that relay signals from neurons to target cells—to regulate inflammation and immunity, or even to feed back and regulate the nervous system itself? Two papers in this issue, by Rosas-Ballina et al. on page 98 (9) and Wong et al. on page 101 (10), demonstrate how neurotransmitters directly modulate specific cells and cellular responses in the immune system at a distance.
Previous work pointed to pathways of direct, long-distance neuroimmune cross-talk. For example, stimulation of the vagus nerve, which originates in the brainstem and innervates visceral organs, inhibits cytokine release and attenuates inflammatory damage in endotoxemia and sepsis. The vagus nerve stimulates celiac ganglion adrenergic neurons that innervate the spleen, leading to release of the neurotransmitter acetylcholine (ACh) and activation of the nicotinic ACh receptor (nAChR) on splenic macrophages. This blocks production of the proinflammatory cytokine tumor necrosis factor–α (TNF-α). How does stimulation of adrenergic neurons induce ACh release in the spleen? Rosas-Ballina et al. show that a splenic T cell capable of synthesizing and secreting ACh fulfills this role (see the figure). The authors identified a subpopulation of CD4+ T cells that secrete ACh, express β-adrenergic receptors, and are located adjacent to adrenergic nerve endings in the spleen. Transplanting these T cells into mutant mice devoid of T cells and with endotoxemia rescued the attenuation of TNF-α by vagus nerve stimulation. Furthermore, reducing expression by small interfering RNA of choline acetyltransferase, an enzyme required for ACh biosynthesis, in these T cells before transplantation blocked rescue of TNF-α attenuation after vagus nerve stimulation. Thus, ACh secretion by these T cells is required in this inflammatory reflex. Considering the roles of TNF-α in CNS inflammation (11), this pathway should also be explored with CNS injury and disease models.
Noradrenaline-secreted from peripheral nerve terminals in the spleen and liver regulates pro- and anti-inflammatory cytokine secretion by stimulating specialized T cells. These immune cells and cytokines, in turn, could feed back into the nervous system and regulate neuroinflammation and neural function. APC, antigen-presenting cell.
CREDIT: B. STRAUCH/SCIENCE
Long-distance neuroimmune communication also occurs in poststroke immunosuppression, which protects the brain from inflammatory damage (12) but leaves the body prone to infection, a major cause of stroke-related death (13). Although the fundamental mechanism of this immunosuppression is not known, Wong et al. focused on stroke-activated hepatic invariant natural killer T (iNKT) cells to address how stroke modulates immunosuppression. Using a mouse stroke model, they observed that mice deficient in iNKT cells developed peripheral infection earlier and had higher mortality, suggesting that iNKT cells normally attenuate stroke-induced immunosuppression. They then found that noradrenergic innervation in the liver, rather than a circulating molecule, signals hepatic iNKT cells after stroke to promote systemic immunosuppression. Either depletion of noradrenergic nerve terminals in the liver or blockade of β-adrenergic receptors with propranolol altered the cytokines secreted by iNKT cells, thereby attenuating immunosuppression, bacterial infection, and mortality in wild-type mice. Conversely, noradrenaline injected directly into the liver activated iNKT cells and increased immunosuppression and infection. Unexpectedly, Wong et al. found that another iNKT cell activator, the glycolipid antigen α-galactoceramide (α-GalCer), which acts through MHC proteins, also reduced bacterial infection after stroke. Together, these data suggest that the determining factor in iNKT cell-mediated immunosuppression after stroke may not simply be their activation, but adrenergic regulation of a shift from proinflammatory T helper cell 1 (TH1)–type cytokines [such as interferon-γ (IFN-γ)] to antiinflammatory TH2-type cytokines [such as interleukin-10 (IL-10)]. Pretreatment of mice with IL-10 led to increased lung infections, whereas administration of propranolol and α-GalCer reversed cytokine production from IL-10 back to IFN-γ after stroke. Demonstrating stroke-induced increase in noradrenergic signaling in the liver would bolster this hypothesis.
The findings of Rosas-Ballina et al. and Wong et al. raise the possibility of leveraging specific pathways to extinguish damaging neuroinflammation without compromising the ability to fight infection—for example, by activating nAChR in the spleen and blocking β-adrenergic receptors and activating MHC receptors in the liver. Furthermore, these findings raise several questions: Are T cells and noradrenaline specialized for mediating long-distance cross-talk between the immune and nervous systems, or do other subpopulations of immune cells participate? Various brain insults and neurodegenerative diseases may range from subclinical involvement to primary or secondary damage from inflammation—are these regulated by long-distance feedback loops between the CNS and peripheral immune organs, in addition to local interactions within the nervous system? For example, although noradrenaline in the spleen increased damaging neuroinflammation, and noradrenaline as well as IL-10 in the liver increased infection, both are neuroprotective in the CNS (5, 6). Neuroactivated immune cells and their cytokine signals may enter the CNS and modulate neuronal and/or glial function, prompting investigation of “immunoneurobiology” pathways in which the peripheral immune system may regulate neural plasticity and behavior.
We acknowledge support from the American Heart Association (grant 11PRE7310069 to E.F.T.), NIH (grants EY020913, EY020297, NS061348, and NS074490 to J.L.G.; grant P30-EY014081 from the University of Miami), and an unrestricted grant from Research to Prevent Blindness (University of Miami).