Illuminating Neuroimmunity: A Humoral Brain

Abstract

The hypothalamic-pituitary-adrenal axis modulates immunity in response to stress. In a recent report in the May 14, 2020 issue of Nature, Zhang et al. use optogenetic tools to investigate whether the splenic immune response is directly controlled by descending neuronal circuits activated in response to stress.

Stress influences the immune system. Although the immune response is generally enhanced by acute short-term stress, chronic stress has been linked with immune dysfunction (Silberman et al., 2003; Dhabhar, 2009) and even susceptibility to some cancers (Blanc-Lapierre et al., 2017). Stress-dependent activation of the hypothalamus-pituitary-adrenal (HPA) axis results in the systemic release of neuroendocrine systemic mediators (CRH/ACTH/corticosterone) along with sympathetic nerve-derived norepinephrine (NE), the latter of which is thought to play a role in locoregional immunomodulation. However, the relative contribution of either axis has long been difficult to resolve because of a lack of molecular handles with differential and bona fide tissue specificity.

In a recent study published in Nature, Zhang et al. (2020) investigate whether a direct splenic neural circuit is activated in response to stress, in addition to the established neuroendocrine-mediated pathway by which stress modulates immune responses via the systemic release of hormones. Using ethanol-induced ablation of the splenic neurovascular plexus, the authors show a resulting reduction of tyrosine hydroxylase+ (TH⁺) sympathetic neurons normally present within the T cell zones and B cell follicles of the spleen. Despite the loss of sympathetic neurons, the non-tissue-specific nature of the splenic denervation—coupled with existing controversy over parasympathetic or sensory splenic innervation (Kooijman et al., 2015) — does not exclude off-target effects, despite the grossly normal architecture. To evaluate the immunological effects of splenic denervation, mice were immunized 6 weeks following surgery with the T-dependent immunogen NP-KLH (4-hydroxy-3-nitrophenylacetyl conjugated to keyhole limpet haemocyanin). Thirteen days following immunization, the authors observe a reduction in the number of splenic plasma cells (PCs) in denervated mice despite no change in the number of germinal center (GC) B cells. CD138^hi PCs arise both from extra-follicular and GC B cell responses (McCarron et al., 2017). IgM⁺ extrafollicular B PCs are typically short lived, whereas the GC produces higher-affinity, class-switched, long-term PCs. Although the majority of PCs are GC-derived, it is unclear whether this defect in PC generation is a result of defects in the extrafollicular response or B cell dynamics within the GC. Notably, this reduction of PCs could not be recapitulated following immunization with a T-independent immunogen, indicating a degree of T cell dependence in PC development within control mice.

Because of the implied T cell dependence of the PC defect, Zhang et al. (2020) explored whether NE-responsive, acetylcholine (ACh)-producing choline acetyl transferase⁺ (ChAT⁺) T cells play a role in PC formation (Rosas-Ballina et al., 2011). Following confirmation of nicotinic ACh receptor (nAChR) a9, β1, and β4 subunit expression via rtPCR of sorted total, PC, and GC B cells, Zhang et al. (2020) generated mice lacking a9 subunit (Chrna9^−/−), which were used in bone marrow transplant experiments to further refine the mechanism by which ACh enhances PC formation. Defects in Chrna9^−/− B PC formation were observed following the co-transplantation of bone-marrow-derived hematopoietic cells from B cell-deficient μMT and Chrna9^−/− (or Chrna9^−/− control) mice into irradiated wild types (WTs) after immunization with NP-KLH. Interestingly, there was no significant difference between transplanted WT and Chrna9^−/− PC formation in denervated mice following immunization.

After assessing Ach enhancement of PC formation, Zhang et al. (2020) went on to determine whether ACh-producing ChAT⁺ CD4 T cells have a role in this B cell response. The adoptive transfer of ChAT-cre::Rosa26-LSL-TdTomato fluorescent reporter CD4⁺ T cells to T cell deficient (Tcrb^−/− Tcrd^−/−) mice confirmed the presence of ChAT⁺ CD44^hi memory CD4 T cells within the spleen only 8–10 days after immunization with NPKLH. The authors also suggest the colocalization of ChAT⁺ T cells with TH⁺ sympathetic neurons and B PCs, based on 2D fluorescence microscopy. In a similar model, ChAT-expressing cells were selectively ablated via diphtheria toxin administration following immunization, resulting in impaired PC development. This further implicates ChAT⁺ T cells in the B plasma response; nevertheless, a role for parasympathetic-derived cholinergic signals cannot be excluded.

The authors next confirm the presence of descending neural circuits onto the spleen by showing that a retrograde trans-synaptic fluorescently labeled pseudo rabies virus can be detected within the paraventricular nucleus (PVN) and central nucleus of the amygdala (CeA) of the brain 96 h following injection into the spleen. The CeA and PVN consist of corticotropin-releasing hormone (CRH)-producing neurons that drive activation of the HPA axis in response to stress, with concomitant corticosterone release from the adrenal glands. To probe a neural link between the CeA/PVN and spleen, Zhang et al. (2020) stereotactically inject into both the CeA and PVN an adeno-associated virus (AAV) expressing conditional light-activated cation channel channelrhodpsin2 (ChR2). Light activation of CeA and PVN CRH⁺ neurons resulted in increased nerve activity within the neurovascular plexus in vivo demonstrating a neural link—although neurotransmitter identity cannot be resolved using this extracellular electrophysiological technique. In a similar set of experiments, AAV vectors containing either apoptosisinducing caspase 3, clozapine-N-oxide (CNO)-inducible inhibitory or CNO-inducible excitatory designer receptor exclusively activated by designer drugs (DREADD)—expressed in a CRH-credependent manner—were stereotactically injected into the CeA and PVN. Whereas PC formation was impaired following the ablation or inhibition of CRH-producing neurons, DREADD activation of the PVN and CeA enhanced the PC response (Figure 1). Interestingly, this enhancement was not observed in mice with ethanol-denervated spleens. Although the authors do not assess this, the remote stimulation of CRH-producing neurons is likely accompanied by activation of the HPA axis with concomitant corticosterone release. This begs the question of whether systemic changes in CRH or corticosterone underlie the effects they measure on B PC formation, given that glucocorticoids and CRH have been implicated in B cell homeostasis at resting state in mice (Gruver-Yates et al., 2014; Harle et al., 2018) with additional reports of altered activation-induced deaminase (AID) expression following in vitro stimulation of human peripheral blood mononuclear cell-derived B cells (Benko et al., 2014). Thus it is not to be excluded that the lack of difference in PC response when combining ethanol denervation with pharmacological activation of the PVN/ CeA (and therefore HPA activation) is not due to damage to the vasculature of the neuro-vascular plexus, which could affect perfusion of splenic corticosterone.

Figure 1. Activation of the CeA and PNS Enhances Splenic B Plasma Cell Formation during a T-Dependent Immune Response.

Optogenetic and chemogenetic excitation of CRH neurons within the CeA and PVN results in increased splenic nerve activity and drives the HPA axis, resulting in an enhanced B plasma cell response. MZ, marginal zone; LZ, germinal center light zone; DZ, germinal center dark zone; SMH, somatic hypermutation; PC, B plasma cell; BMem, Memory B cell.

To recapitulate enhancement of B PC formation in a physiologically relevant way, Zhang et al. (2020) induce mild stress—which is well known to drive the HPA axis—using an elevated platform standing (EPS) behavioral regimen, in which mice were placed on a small, transparent platform 1.5 m above the ground twice daily for 5 days prior to immunophenotyping. As measured by increases in calcium movement through an AAV CRH-dependent GCaMP6m calcium indicator during EPS, this mild acrophobic stress was sufficient to increase CRH neuronal activity in the PVN and the CeA. It is clear that the PC response to immunization is enhanced with EPS—an effect that is abrogated following the ethanol denervation of the neuro-vascular plexus. Moreover, an EPS-induced increase in immunogen (NP)-specific serum IgG was only observed in mice with functional CRH⁺ CeA/PVN neurons, an intact splenic neuro-vascular plexus, and Chrna9^+/+ B cells, implicating each of these factors in the enhancement of PC formation following exposure to mild stress. The effects of an additional, stronger behavioral induction of stress on B cell response was also assessed using prolonged physical restraint (PPR) for 90 min twice daily from day 8 to day 12 post-immunization; this stress regimen differentially activated the PVN and CeA as measured by calcium flux in the same way as with EPS. Although PC formation was unaltered following this behavioral regimen, the proportion of GC B cells was notably decreased. This contrasting result may be reflective of the significantly increased serum corticosterone levels observed following PPR, which are more than twice as high as controls—in comparison with EPS-treated mice whose serum levels were around 1.5× controls. Taken together, the additional increase of both mutations in the variable heavy chain of the B cell receptor and proportion of NP-specific, long-term, class-switched PCs in the bone marrow 27 days following immunization indicate an enhanced immune response with EPS. It is clear that both regimens increase activity of the HPA axis, making it difficult to decipher an exclusive role for splenic sympathetic nerves on the immune response.

Zhang et al. (2020) demonstrate an ACh-enhancement of PC formation following a T-dependent immune response, which is also induced by behaviorally induced mild stress. However, a number of questions remain given that a direct link between the stress-induced activation of the PVN and CeA and sympathetically stimulated ACh production by ChAT⁺ T cells cannot yet be solidly confirmed. In addition, the inevitable changes in activation of the HPA axis that accompany pharmacological and/or stress induced activation of CeA and PVN CRH⁺ neurons cannot be excluded from playing a role in the B cell response. There are vast implications for the influence of stress on the immune system; however, evidence for the direct neuronal control of the immune system by the brain still remains in the shadow. Illuminating the neuroimmune circuity will require a concerted effort among scientists to develop a tool set of molecular handles with bona fide tissue specificities for the autonomic nervous system.