Nerve-derived Transmitters Including Peptides Influence Cutaneous Immunology

Abstract

Clinical observations suggest that the nervous and immune systems are closely related. For example, inflammatory skin disorders; such as psoriasis, atopic dermatitis, rosacea and acne; are widely believed to be exacerbated by stress. A growing body of research now suggests that neuropeptides and neurotransmitters serve as a link between these two systems. Neuropeptides and neurotransmitters are released by nerves innervating the skin to influence important actors of the immune system, such as Langerhans cells and mast cells, which are located within close anatomic proximity. Catecholamines and other sympathetic transmitters that are released in response to activation of the sympathetic nervous system are also able to reach the skin and affect immune cells. Neuropeptides appear to direct the outcome of Langerhans cell antigen presentation with regard to the subtypes of Th cells generated and neuropeptides induce the degranulation of mast cells, among other effects. Additionally, endothelial cells, which release many inflammatory mediators and express cell surface molecules that allow leukocytes to exit the bloodstream, appear to be regulated by certain neuropeptides and transmitters. This review focuses on the evidence that products of nerves have important regulatory activities on antigen presentation, mast cell function and endothelial cell biology. These activities are highly likely to have clinical and therapeutic relevance.

1. Introduction

Anecdotal evidence has long suggested that the nervous and immune systems are closely related. Many inflammatory diseases, such as atopic dermatitis, psoriasis, acne and rosacea, are believed to be aggravated in response to stress (Fortune et al., 2005; Misery, 2011; Khansari et al, 1990; Sirinek and O’Dorisio, 1991). Furthermore, there is substantial evidence that nerves play a key role in the pathogenesis of psoriasis, discussed below (Dewing, 1971; Raychaudhari and Farber, 1993; Perlman, 1972). These observations are now supported by a growing body of research indicating a key role for neuropeptides and neurotransmitters in influencing cutaneous immunity.

Neuropeptide transmitters such as vasoactive intestinal polypeptide (VIP), pituitary adenylate cyclase-activating peptide (PACAP), calcitonin gene-related peptide (CGRP) and substance P (SP), can be released by sensory nerves, specifically unmyelinated afferent C-fibers (Fernandes et al, 2009; Zhang et al., 1995; Nolano et al., 2012). This type of fiber innervates the skin (Schmelz, 2011). Important actors of the immune system, including Langerhans cells (LCs) (dendritic antigen presenting cells that reside in the epidermis) and mast cells, have been found to be anatomically associated with these nerves, making them likely targets for secreted nerve products (Hosoi et al., 1993; Forsythe and Bienenstock, 2012). Indeed, it has been shown that neuropeptides and adrenergic transmitters modulate LC and mast cell function (Hosoi et al., 1993; Forsythe and Bienenstock, 2012; Seiffert et al., 2002; Ding et al., 2012; Kodali et al., 2004; Kodali et al., 2003). Additionally, dermal blood and, probably, lymphatic vessels are associated with both sensory and sympathetic nerves (Coventry and Walsh, 2003; Dalsgaard et al., 1984; Dalsgaard et al., 1983; Sacchi et al., 1994). Endothelial cells (ECs) may be important in this regard. ECs serve key functions including regulation of hemostasis, vasomotor tone, barrier function, cell and nutrient trafficking and angiogenesis (Aird, 2003), may be important in this regard. ECs can release many cytokines including interleukin (IL)-6, an important differentiation factor for Th17 cells (Swerlick and Lawley, 1993; Mantovani and Dejana, 1989). Furthermore, ECs release a variety of chemoattractant molecules (Swerlick and Lawley, 1993; Mantovani and Dejana, 1989) and express cell surface molecules, including selectins, vascular cell adhesion molecule (VCAM)-1 and intercellular adhesion molecule (ICAM)-1 (Cid, 2002; Springer, 1994) that facilitate leukocyte extravasation. Recent data demonstrate that CGRP and the sympathetic co-transmitter adenosine triphosphate may have important regulatory activities on ECs (see below).

This review will focus on the evidence that products of nerves have important regulatory activities on antigen presentation, mast cell function and endothelial cell biology. The likely clinical and possible therapeutic relevance of these findings will be discussed.

2. Antigen Presentation

Much of the work on effects of neurotransmitters (peptides and non-peptides) has focused on Langerhan cells (LCs). LCs are dendritic antigen-presenting cells (APC) of the epidermis. They capture antigen in the periphery and traffic to regional lymphoid organs to present to lymphocytes. LCs mature in culture and present antigens for many immune responses (Inaba et al., 1986; Grabbe et al., 1991). In the maturation process, LCs upregulate CD80, CD86, CD54, CD40, CD83, DC-LAMP, IL-12p40 and CCR7 while downregulating Langerin (Nakagawa et al., 1999; Berthier-Vargnes et al., 2005)). Additionally, macropinocytosis is downregulated with maturation although some receptor-mediated endocytosis appears to remain operative (Sparber et al., 2010). Thus, classically, LCs were felt to be potent antigen presenting cells (APCs) in vivo, responsible for initiating immune responses. More recent evidence, however, suggests that in the steady-state in vivo, LCs may downregulate immunity (Kaplan et al., 2005) and induce tolerance (Igyártó and Kaplan, 2010), perhaps to prevent reactivity against commensal organisms. In this regard, evidence indicates that TGFβ maintains LCs in an immature state in situ (Kel et al., 2010; Geissman et al., 1999). We speculate that exposure to CGRP in situ may also play a role in maintaining the immaturity of LCs (see below). It further appears as though dermal DCs may play a key role in the inductive phase of immunity (Kaplan et al., 2005; Romani et al., 2012) and that LCs negatively regulate the anti-leishmania response (Kautz-Neu et al., 2011). Additionally, LCs are not responsible for protective Th1 immune responses in the mouse vagina; submucosal dendritic cells (DCs) appear to serve this role (Zhao et al., 2003).Of course, in the presence of danger signals, LCs may play an inductive role as well. Interestingly, LCs express a CD39 ecto-nucleoside triphosphate diphosphohydrolase that serves to hydrolyze ATP and ADP released by keratinocytes in response to chemical irritant exposure, thereby diminishing irritant contact dermatitis (Mizumoto et al. 2002). The picture emerging of the immunogenic role of LCs is certainly not a simple one: new evidence suggests that LCs may be involved in the generation of Th17 cells (Mathers et al., 2009; Igyártó et al., 2011).

2.1. Calcitonin gene-related peptide

CGRP is a 37-amino acid neuropeptide generated by tissue-specific alternative processing of the calcitonin gene (Asahina et al., 1995). It is present in both the central and peripheral nervous systems. CGRP is a potent endogenous vasodilator and is involved in recruitment of inflammatory cells in the setting of inflammation.

CGRP has several important effects on cutaneous immunity. Early experiments demonstrated that LC exposure to CGRP in vitro inhibited their ability to present antigen in several systems including presentation of alloantigens for proliferation of T cells, presentation of an antigen to a T-T hybridoma, and use for eliciting a delayed-type hypersensitivity response in vivo in animals previously sensitized to an antigen (Hosoi et al., 1993). Importantly, epidermal LCs were found to frequently be anatomically associated with unmyelinated nerves that contain CGRP (Hosoi et al., 1993). Thus, an anatomic basis for regulation of LC function by the nervous system through release of CGRP exists. Subsequent studies showed that LC exposure to CGRP inhibited their ability to induce contact immunity in naïve mice after pulsing with an antigen and intradermal administration (Asahina et al., 1995). Effects of CGRP on LC antigen presentation were due, at least partly, to inhibition of NF-κB signaling (Ding et al., 2007). Thus, we originally concluded that CGRP (and certain other neurotransmitters, see below) “inhibited” antigen presentation. This simple interpretation, however, was later shown to be incorrect. It was found that while CGRP inhibited LC antigen presentation for Th1 responses, presentation for Th2 responses was enhanced (as demonstrated by increased IL-4 production and decreased IFN-γ production) (Ding et al., 2008) and CGRP inhibited the stimulated expression of IL-12p40 and IL-1beta by macrophages and a LC-like cell line while augmenting the expression of IL-10 (Torii et al., 1997). It was further shown in the same study that CGRP induces the production of the Th2 chemokines, CCL17 and CCL22 by LCs, while inhibiting the stimulated production of the Th1 chemokines, CXCL9 and CXCL10 suggesting that CGRP biases LCs toward Th2 responses.

Based on the above result, we speculate that chronic and repetitive exposure to CGRP from adjacent nerves in situ participates in preventing LC maturation and, thus, contributes to LC downregulatory function (at least for Th1-type immunity) in the steady-state. We can only speculate as to why exposure to CGRP might shift LC antigen presenting function towards enhanced Th2-mediated responses. It may be that a Th2-mediated response is protective: perhaps it avoids or balances an overactive Th1-mediated response against beneficial commensal organisms or self-tissues while favoring a humoral response against viral agents that enter the epidermis. Indeed, there is recent evidence that Staphylococcus epidermis produces antimicrobial peptides that protect against pathogenic organisms and that other commensal organisms produce bacteriocins that inhibit the growth of other bacterial strains (Gallo and Hooper, 2012). Thus, inhibition of an immune response against these organisms likely has survival value.

2.2. Pituitary adenylate cyclase-activating peptide and vasoactive intestinal polypeptide

VIP and pituitary PACAP are related members of a superfamily of neuropeptides that include secretin, glucagon, and growth hormone-releasing factor (Harmar et al., 1998). These factors have overlapping activities at several receptors. Two of these, VPAC1 and VPAC2, are G protein-coupled receptors that bind PACAP and VIP with similar affinity and activate adenylate cyclase (Harmar et al., 1998). PACAP consists of two forms, a 38 amino acid molecule (PACAP38) and a 27 amino acid form (PACAP27) (Fahrenkrug, 2010). These have identical activities in most biological systems and, although both types can be found in tissues, PACAP38 is the dominant form (Fahrenkrug, 2010). VIP is a 28 amino acid peptide that has 68% homology with PACAP27 (Fahrenkrug, 2010). Both PACAP38- and VIP-immunoreactive fibers are present in human skin (Schulze et al., 1997; Savage et al., 1990; Groneberg et al., 2003).

Initial studies of VIP and PACAP effects on LC function yielded similar findings to those seen with CGRP. As with CGRP, VIP and PACAP were originally found to inhibit antigen presentation to a Th1 clone and a T-T hybridoma and to inhibit LC presentation of antigen to elicit a delayed-type hypersensitivity response in vivo in previously sensitized mice (Kodali et al., 2004; Kodali et al., 2003). Also, like CGRP, effects of PACAP/VIP on LC antigen presentation were found to involve inhibition of NF-κB signaling (Ding et al., 2007). Additional studies showed that VIP and PACAP were found to suppress stimulated production of IL-1β and augment the production of IL-10 (Kodali et al., 2004; Kodali et al., 2003), and VIP was found to inhibit stimulated IL-12p40 production (Kodali et al., 2003).

Thus, as with CGRP, we originally concluded that PACAP/VIP “inhibited” antigen presentation. This too, however, turned out to be incorrect because while PACAP and VIP inhibit LC antigen presenting capacity for generation of Th1 cells, they similarly enhance presentation for development Th2 cells, and, in addition, enhance presentation for differentiation of Th17 cells (Ding et al., 2012). It was demonstrated that PACAP/VIP treatment of LCs prior to in vitro antigen presentation to CD4⁺ T cells enhances IL-17A, IL-6, and IL-4 production, decreases gamma interferon (IFN-γ) and IL-22 release, and increases RORγt and Gata3 mRNA expression (transcription factors associated with Th17 and Th2 cells, respectively) while decreasing T-bet expression (transcription factor associated with Th1 cells). By cytofluorography, the CD4⁺ T-cell population was increased in IL-17A- and IL-4-expressing cells and decreased in IFN-γ-expressing cells (Ding et al., 2012). Thus, exposure of LCs to PACAP or VIP appears to shift Th cell differentiation away from Th1 cells and towards Th2 and Th17 cell responses (although with decreased IL-22 production). VIP has also been shown to enhance Th2 responses in vivo; transgenic mice that overexpress VPAC2 in CD4⁺ T cells show a Th2 preference, whereas mice deficient in VPAC2 show a preference for the Th1 response (Goetzl et al, 2001; Voice et al., 2003; Delgado and Ganea, 2011).

2.3. Catecholamines

Postganglionic sympathetic fibers of the sympathetic nervous system innervate blood vessels, sweat glands and hair follicles of the skin and appear as single nerve fibers in the skin (Metze, 2009). Although norepinephrine is the classic and primary sympathetic neurotransmitter in the periphery, the adrenal medulla produces epinephrine (and some norepinephrine, in a 4:1 ratio), which can reach the skin through the circulation. These hormones are the primary mediators of the sympathetic nervous system, commonly described as responsible for the "fight or flight" response (Khurana, 2006). In addition to norepinephrine in the skin from innervating sympathetic nerve fibers and the circulation, there is also local production of epinephrine by keratinocytes (Schallreuter et al., 1992).

We explored the expression of adrenergic receptors (AR) by murine LCs and assessed their functional role on antigen presentation and modulation of cutaneous immune responses. Both purified LCs and the LC-like cell lines XS52-4D and XS106 (derived from neonatal murine epidermis) were found to express mRNA for the ARs α₁-A and β₂ (Seiffert et al., 2002). XS106 cells and purified LCs were found to express β₁-AR mRNA. Pretreatment of epidermal LCs with epinephrine or norepinephrine in vitro suppressed the ability of these cells to present antigen for elicitation of delayed-type hypersensitivity in previously immunized mice. This effect was blocked by use of the β₂-adrenergic antagonist ICI 118,551 but not by the α-antagonist phentolamine (Seiffert et al., 2002). These results suggest that epinephrine and norepinephrine can suppress the antigen presenting capability of LCs to reduce epidermal immune reactions.

3. Mast cells

Mast cells derive from multipotential hematopoietic stem cells (MHSCs).. They differentiate into two classes, connective tissue and mucosal mast cells, and reside in areas corresponding to their class (Kitamura et al., 2006). Among their many functions, mast cells work to preserve tissue integrity, to warn the immune system of local injury or infection, and to assist in repair of injury (Maurer et al., 2003; Forsythe and Bienenstock, 2012). While mast cells are most commonly discussed in the context of their role in allergic inflammation, they have important functions in both the innate and adaptive immune response and appear to play a key role in immune suppression induced by exposure to mid-range ultraviolet (UVB) radiation (Maurer et al., 2003; Forsythe and Bienenstock, 2012; Halova et al. 2012).

Mast cells can also be modulated by neuropeptides. Although mast cells are distributed throughout the body, they tend to be concentrated near nerve fibers, blood vessels, and lymphatic vessels. This close anatomic proximity facilitates the modulation of mast cells by neuropeptides and secreted products of ECs (Forsythe and Bienenstock, 2012).

3.1. Substance P

Substance P (SP) is an 11-amino acid peptide that belongs to the tachykinin neuropeptide family that includes neurokinin A and B (Pennefather et al., 2004). It is distributed in both the central and peripheral nervous systems and plays an important role in the regulation of the neuronal life cycle, the transmission and perception of pain, autonomic reflexes, and vasodilation (O'Connor et al., 2004). Substance P has also been implicated in the development of cancer; it is overexpressed in many different cancers and can induce the proliferation of tumor cells, angiogenesis, and migration (Rosso et al., 2012).

SP serves to generally enhance cellular immunity. It appears to have several effects on mast cells that contribute to neurogenic inflammation. For example, it induces mast cell production and secretion of TNF-α (Ansel et al., 1993), stimulates mast cell degranulation and release of histamine (Ebertz et al., 1987), and in human cord blood mast cells, induces the mRNA expression and secretion of IL-8 (Castellani et al., 2008). Interestingly, in the setting of exposure to UVB radiation, SP may be released by cutaneous nerves (along with CGRP) to induce mast cell degranulation and release of IL-10 and TNF-α (Streilein et al., 1999). In this context, TNF-α contributes to UVB radiation-induced immune suppression, perhaps through effects on antigen presenting cells, and IL-10 appears to promote immunologic tolerance to an encountered antigen (Streilein et al., 1999).

Many additional roles for SP have been described. It was found that SP improves wound healing when administered exogenously to a laser-induced skin wound (Delgado et al., 2005). Substance P has also been shown to augment contact hypersensitivity (Lotz et al., 1988) and to increase the expression of leukocyte adhesion molecule by ECs (Matis et al., 1990). These proinflammatory effects appear to be exerted via the activation of the transcription factor NF-κB (Marriott et al., 2000), in contrast to the "anti-inflammatory" inhibition exerted by CGRP, VIP, and PACAP.

3.2. Catestatin

Catestatin is a 21-amino acid peptide derived from the neuroendocrine pro-hormone chromogranin A, which is part of the granin family of proteins (Aung et al., 2011; Huttner et al., 1991). The granin family of proteins is found in the secretory vesicles of various endocrine, nerve, and immune cells. Under stress, these cells secrete several endogenous antimicrobial peptides (AMPs) derived from chromogranin A (Aslam et al., 2012). Catestatin serves as an AMP in humanskin and is effective against many skin pathogens, such as gram-positive and negative bacteria, yeast, and fungi (Aung et al., 2011; Radek et al., 2008; Huttner et al., 1991) and is upregulated in response to injury (Radek et al., 2008). Catestatin has also been shown to cause the migration and degranulation of, and production of pro-inflammatory cytokines and chemokines by, human mast cells (Aung et al., 2011). The pro-inflammatory cytokines and chemokines include granulocyte-macrophage colony-stimulating factor, monocyte chemotactic protein-1/CCL2, macrophage inflammatory protein-1α/CCL3, and macrophage inflammatory protein-1β/CCL4 (Aung et al., 2011). The combined findings that catestatin, a neuroendocrine peptide, is upregulated in response to injury, has antimicrobial function, and is capable of modulating mast cell activity, provide yet another link between the nervous and cutaneous immune systems.

4. Endothelial Cells

Endothelial cells serve key functions including regulation of hemostasis, vasomotor tone, barrier function, cell and nutrient trafficking, and angiogenesis (Aird, 2003). They are capable of releasing many cytokines including IL-6, an important differentiation factor for Th17 cells (Swerlick and Lawley, 1993; Mantovani and Dejana, 1989), and expressing important cell surface molecules, such as selectins, VCAM-1 and ICAM-1, that bind blood leukocytes and facilitate their extravasation (Cid, 2002; Springer 1994). Cells can also exit tissue though lymphatic vessels, lined by ECs capable of releasing a variety of chemokines (Swerlick and Lawley, 1993; Mantovani and Dejana, 1989). As leukocytes roll on the endothelium, cell-associated chemokines interact with chemokine receptors on leukocytes that lead to activation of leukocyte integrins, facilitating binding to integrin receptors on the endothelial cells (Huber et al., 1991; Muller, 2012) and diapedesis of leukocytes. A role for IL-8 in transendothelial migration of neutrophils and T cells has been shown directly (Huber et al., 1991; Santamaria Babi et al., 1996). Additionally, if the concentration of a chemokine released by endothelial cells is higher on the ablumenal side of a vessel than in the lumen of the vessel (perhaps because of dilution within the vessel due to blood flow), as hypothesized for monocyte chemoattractant protein-1 (CCL2) by Randolph and Furie (1995), leukocytes may follow the gradient out of the lumen. However, chemokines are primarily found on the cell surface of endothelial cells and a haptotactic gradient of MCP-1 was not found in a model using monolayers of a hybridoma cell line as a model for vascular endothelium (Hardy et al., 2004).

An anatomic relationship exists between dermal vessels and nerves; blood vessels in the dermis are associated with sensory and sympathetic nerves (Coventry and Walsh, 2003; Dalsgaard et al., 1984; Dalsgaard et al., 1983; Sacchi et al., 1994). Lymph nodes and mesenteric lymphatic vessels are innervated (Sacchi et al., 1994; Bellinger et al., 1992) and small nerve bundles are present in the adventitia of lymphatics in the dorsal foot (Boggon and Palfrey, 1973). The large number of nerves in the dermis may even allow lymphatics to be exposed to neuropeptides and neurotransmitters without specific innervation.

4.1. Alpha-melanocyte stimulating hormone

Alpha-melanocyte stimulating hormone (α-MSH) is one of many cleavage products of pro-opiomelanocortin (POMC) (Eves and Haycock, 2010). POMC is produced by keratinocytes response to exposure to UV radiation or endotoxins. It is then cleaved to form adrenocorticotropin hormone, which is further cleaved to form α-MSH (Luger et al., 1993). Other sources of POMC in the epidermis include LCs and melanocytes, which have been found to express mRNA for POMC and to release peptide products of POMC (Luger et al., 2000). ECs also can make α-MSH and its production is stimulated by exposure of these cells to UVB radiation or UVA1 (340–400 nm) radiation (Scholzen et al., 2000).

Alpha-MSH serves a primarily anti-inflammatory role in the skin. It inhibits neutrophil migration and the production of pro-inflammatory cytokines such as IFN-γ, TNF-α, IL-1, and IL-6 by macrophages (Lipton and Catania, 1989). Although α-MSH induces IL-8 and growth-regulated oncogene-α (GRO-α) release from unstimulated ECs in vivo (Scholzen et al., 1999), it inhibits the lipopolysaccharide-stimulated upregulation of ICAM-1, VCAM, and E-selectin by human microvascular ECs and suppresses the adhesion of lymphocytes to monolayers of human microvascular ECs (Luger et al., 2000). It also inhibits the lipopolysaccharide-induced activation of NF-κB in ECs (Luger et al., 2000). In addition to inhibiting pro-inflammatory activity,α-MSH enhances anti-inflammatory factors; it Alpha-MSH also induces the production of IL-10, an anti-inflammatory cytokine, in multiple cells, such as keratinocytes (Redondo et al., 1998) and monocytes (Bhardwaj et al., 1997). The presence of the primary α-MSH receptor, MC-1R, has been demonstrated on a variety of cells, such as ECs, keratinocytes, dendritic cells, mast cells, macrophages and fibroblasts, among others (Luger et al., 2000; Redondo et al., 1998; Bhardwaj et al., 1997; Böhm and Luger, 2004).

Alpha-MSH also has central anti-inflammatory effects; injection of very small amounts of α-MSH into brain ventricles inhibits peripheral inflammation (Lipton and Catania, 1989). Interestingly, this central effect can by blocked by β₂-AR antagonists given systemically (Lipton and Catania, 1989).

4.2. Adenosine 5'-triphosphate

Adenosine 5'-triphosphate (ATP) is incorporated into nucleic acids and is a key chemical energy carrier (Brake and Julius, 1996). It additionally acts as both an intracellular (Kamenetsky et al., 2006) and extracellular (Brake and Julius, 1996) signaling molecule and is well established as a co-transmitter in both sympathetic (Burnstock, 2004) and other nerves. We have shown that the human transformed microvascular endothelial cell line HMEC-1 expresses at least some P2X series (P2X₁, PS2X₄, P2X₅ and P2X₇ with variable expression of P2X₃) ligand-gated receptors and P2Y series (P2Y₂ and P2Y₁₁) G protein-coupled purinergic receptors (Seiffert et al., 2006). When HMEC-1 cells were cultured in the presence of the long-lived, hydrolysis-resistant ATP analog ATPγS, production of the inflammatory cytokine IL-6 and the chemokines IL-8/CXCL8, monocyte chemotactic protein-1/MCP-1, and GRO-α/CXCL1 was stimulated (Seiffert et al., 2006). ATPγS can also induce the production of IL-8 and GRO-α in primary human microvascular ECs and ATP, the parent molecule, can induce these chemokines in both HMEC-1 cells and primary human microvascular ECs (Bender et al., 2008). Interestingly, we have shown that tetracycline inhibits the in vitro ATPγS-induced and TNF-α-induced release of IL-8 and GRO-α by HMEC-1 cells and primary dermal microvascular ECs (Bender et al., 2008). Thus, some of tetracycline's beneficial effects in the treatment of inflammatory disorders may relate to its ability to inhibit stimulated chemokine release by ECs.

4.3. Calcitonin gene-related peptide

A number of studies have demonstrated that systemic or local administration of CGRP to mice inhibits the magnitude of an induced inflammatory stimulus in the skin or elsewhere (Gomes et al., 2005; Raud et al, 1991; O'Kane et al., 2006; Clementi et al, 1995; Clementi et al., 1994). We hypothesized that this might be due to an anti-inflammatory effect of CGRP on ECs. Experiments were performed to examine the ability of CGRP to inhibit the stimulated release of chemokines from dermal microvascular ECs. Initial experiments determined that both HMEC-1 cells and primary human dermal microvascular ECs express mRNA for components of the CGRP and adrenomedullin receptors (Huang et al., 2012). It was found that CGRP inhibits lipopolysaccharide-induced production of the chemokines IL-8, MCP-1 and GRO-α by both HMEC-1 cells and primary human dermal microvascular ECs (Huang et al., 2012) and that this inhibition can be blocked by specific type I CGRP receptors. Importantly, CGRP was found to inhibit NF-κB activation in HMEC-1 cells, as in LCs. Bay 11–7085, an inhibitor of NFκB activation and the phosphorylation of IκBα, was also found to inhibit lipopolysaccharide-induced production of these three chemokines (Huang et al., 2012). Thus, the NF-κB pathway appears to be involved in CGRP-mediated suppression of chemokine production. Notably, CGRP treatment of LPS-stimulated HMEC-1 cells inhibited their ability to chemoattract human neutrophils and mononuclear cells, presumably because of inhibition of chemokine release. These experiments support the likelihood that CGRP has endogenous anti-inflammatory activities separate from its activities on adaptive immunity.

5. In vivo studies

Several studies have been carried out to examine the in vivo relevance of many of the in vitro findings discussed above. Early experiments demonstrated that when CGRP was injected into mice prior to topical immunization at the same skin site with a hapten [contact sensitizers trinitrochlorobenzene (TNCB) or dinitrofluorobenzene (DNFB)], a local, depressed contact hypersensitivity response was observed upon challenge (Asahina et al., 1995). This effect was seen only with immunization at the site of injection and was not observed when immunization was at a distant site. Another group subsequently made a similar observation (Niizeki et al., 1997). Recently, this observation has been extended. Mikami and collaborators (2011) demonstrated that while intradermal administration of CGRP inhibited the induction of contact hypersensitivity to Th1-dominant haptens, such as TCNB and DNFB (discussed above), it led to an enhanced response to immunization with the Th2-dominant hapten fluorescein isothiocyanate. These findings are in accord with the in vitro results discussed above in which CGRP treatment of antigen-presenting cells results in a decreased Th1 response but an increased Th2 response.

Analogous experiments were done with PACAP. Like CGRP, intradermal administration of PACAP prior to application of DNFB resulted in a suppressed contact hypersensitivity response (Kodali et al., 2003). Kodali and colleagues (2003) further showed that PACAP exposure in vitro was able to suppress the ability of epidermal cells enriched for LC content to elicit a delayed-type hypersensitivity response in immunized mice.

Recent experiments have provided a more detailed understanding of the in vivo effects of administration of PACAP or VIP. Naïve mice were injected intradermally with either VIP or PACAP followed by application of DNFB at the injected site. Draining lymph nodes were harvested several days later and CD4⁺ T cells were then isolated and stimulated with anti-CD3 and anti-CD28 monoclonal antibodies. Subsequent cytokine production was quantified by enzyme-linked immunosorbent assays. It was found that IFN- γ production was greatly reduced by intradermal administration of VIP or PACAP, consistent with decreased Th1 immunity, while IL-4 and IL-17A production was enhanced. IL-22 production was also suppressed (Ding et al., 2012). These results are consistent with the in vitro effects already discussed of PACAP and VIP on antigen presentation to CD4⁺ T cells. PACAP and VIP appear to inhibit a Th1 response but to enhance Th2 and Th17 responses.

This type of experiment has also been conducted with epinephrine. Intradermal injection of mice with epinephrine leads to suppressed induction of contact hypersensitivity to a hapten applied topically to the skin at the site of injection or at a distant site (Seiffert et al., 2002). One could speculate that this is due to effects on antigen presenting cells throughout the animal; however, other effects of epinephrine could be responsible. Others have shown that treatment with the β₂-AR antagonist ICI 118,551 during sensitization to fluoroisothiocyanate increases LC migration, the contact hypersensitivity response (CHS), and production of IFN-γ and IL-2 in draining lymph node cells and that norepinephrine reduces IL-12 and stimulates IL-10 production by bone marrow-derived dendritic cells (Maestroni and Mazzola, 2003).

6. Clinical Relevance

Most clinicians accept a close relationship between the nervous system, stress and inflammatory skin disorders. For example, although the association is not universally accepted, there are reports that seborrheic dermatitis is associated with Parkinson’s disease and several other neurologic disorders (Binder and Jonelis, 1983; Plewig and Jansen, 2003). Psoriasis, atopic dermatitis, rosacea and acne are all widely believed to be exacerbated by stress. Although controlled trials are difficult, there are a number of reports indicating that stress can exacerbate these conditions (Dika et al., 2007; Arndt et al., 2008; Sowińska-Glugiewicz et al., 2005; Yosipovitch et al., 2007) and that alleviation of stress can improve some of them, such as atopic dermatitis and psoriasis (Ersser et al., 2007; Zacharieae et al., 1996). In support of this concept, there are animal models demonstrating enhancement of inflammatory skin disorders by experimental stressors (Amano et al., 2008; Pavlovic et al., 2008). There is also literature suggesting that stress increases the risk of developing, or exacerbating existing, non-cutaneous diseases. There are reports that stress may increase the risk for coronary heart disease (Dimsdale, 2008; von Känel et al., 2008; Schulz and Beach, 1999), for infectious diseases (Godbout and Glaser, 2006; Glaser and Kiegolt-Glaser, 1997; Glaser et al., 1999; Cohne et al., 1991; Kiank et al., 2006) and for exacerbation of autoimmune diseases including lupus erythematosus (Wallace, 1987; Stojanovich and Marisavljevich, 2008), rheumatoid arthritis (Wallace, 1987; Cutolo and Straub, 2006) and multiple sclerosis (Mitsonis et al., 2008). In humans there is some evidence that stress can influence susceptibility to infectious skin diseases. There are several reports suggesting that recurrences of cutaneous herpes simplex are associated with psychosocial stress and depression (Pereira et al., 2003; Goldmeier et al., 2008; Chida and Mao, 2009) and that the occurrence of herpes zoster is preceded by more severe life stressors (Schmader et al., 1990; Mehta et al., 2004; Lasserre et al., 2012).

The results of both in vitro and in vivo experiments summarized above provide clues as to how the nervous system and/or stress might affect inflammatory skin disorders. The data summarized herein strongly suggest that some neuropeptides and neurotransmitters direct the outcome of antigen presentation with regard to the subtypes of Th cells generated. It is now well established that Th17 cells are key in the pathophysiology of psoriasis. It has also been known for decades that the full expression of psoriasis depends on innervation. That is, denervation of psoriatic skin leads to improvement or clearing of the psoriasis. One may speculate that the evidence that PACAP and VIP bias LC antigen presentation towards the Th17 pole may play a role in these phenomena. It may be that release of PACAP and/or VIP by cutaneous nerves contributes to psoriasis by biasing immune responses towards generation of Th17 cells. In support of this possibility, there is a well-described murine model of psoriasiform dermatitis involving the actions of Th17 cells where the rash clears upon denervation of affected skin (Ostrowski et al., 2011). We further speculate that LC exposure to CGRP in the epidermis (Fig. 1), may shift antigen presentation towards enhanced Th2-mediated responses in order to protect against an overactive Th1 response: perhaps this shift avoids a Th1-mediated attack against beneficial commensal organisms (as discussed above) or self-tissues. Hypothetically, it may also serve to protect against uncontrolled Th1 responses in a homoeostatic manner. In this model, an immunologic inflammatory stimulus may induce nerve conduction centrally with the impulse propagating past branch points back into the skin (antidromically) to limit antigen presentation for Th1-type immunity.

Fig. 1.

Upper panel: The nervous system regulates immunity in the skin through release of products of nerves that influence Langerhans cell presentation of antigen to naïve T cells. In the epidermis LCs are intimately associated with CGRP-containing nerves. Lower panel: Nerves may also regulate exit of leukocytes from dermal vessels by release of factors that modulate endothelial cell functions. sympathetic nerves associated with dermal blood vessels release ATP, perhaps in response to stress with activation of the sympatho-adrenal response (fight-or-flight response). ATP binds to receptors on endothelial cells inducing expression of ICAM-1 and the chemokines CXCL1, CXCL8 and CCL2. Leukocytes begin to roll on the endothelium through interactions between selectins and develop enhanced expression of integrins through activation by chmokines. They then stick due to interactions between integrins and their receptors such as leukocyte function-associated antigen-1 (LFA-1) and ICAM-1 or very late antigen-4 (VLA-4) and VCAM-1. Cells then transmigrate out of the vessel. However, CGRP counter-regulates this process by inhibiting induction of CXCL1, CXCL8 and CCL2 expression by endothelial cells.

Additional evidence for the immunological influences of the nervous system comes from spinal cord injury (SCI) patients. Following spinal cord injury, these patients tend to present with inadequate immune function reflected by an increased susceptibility to bacterial infection as well as a decrease in the number of natural killer cells, T cells, and cellular adhesion molecules (Cruse et al., 1996; Nash, 2000). Indeed, septicemia and pneumonia are leading causes of death in the SCI patients in the first few years following injury (Campagnolo et al., 2008) and SCI patients have been found to have increased colonization of multi-drug resistant Gram-negative bacilli (Fawcett et al., 1986). The pathophysiology of these changes is not well-understood. In addition to direct neural changes, stressors due to spinal cord injury may have a role (Cruse et al., 1996). These findings appear to be consistent with the findings that psoriasis clears following denervation; the nervous system is necessary for full immune function but abnormal activity may result in inflammatory disease.

An important question is whether neuropeptide levels are associated with immunologic changes in relevant target organs. Stress has been shown to alter measurable levels of some circulating neuropeptides in humans including increased VIP in children undergoing stressful life events (Herberth et al., 2008) and increased PACAP levels in females (but not males) with posttraumatic stress disorder, possibly via estrogen effects that appear to be dependent on a single nucleotide polymorphism in the PAC-1 gene (Ressler et al., 2011). Furthermore, chronic stress increases PACAP mRNA expression in rat brains (Hammack et al., 2009). Knockout models of these neuropeptides appear to confirm their role in emotional or psychological stress; PACAP null mice showed markedly attenuated corticosterone responses to emotional stressors (Tsukiyama et al., 2011; Lehmann et al., 2012).

With regard to the skin, mice stressed for 24 hours demonstrated an increased percentage of SP- or CGRP-positive sensory neurons in skin-innervating dorsal root ganglia (Joachim et al., 2007). Similarly, immobilization stress for several hours reduced contact hypersensitivity responses in concert with increased intensity of CGRP staining of cutaneous nerves (Kawaguchi et al., 1997). In another study, 2 or 8 hours of immobilization stress enhanced the susceptibility of BALB/c mice to cutaneous leishmania infection (with challenge of organisms at the conclusion of the stress period) and the mice stressed for 8 hours and showed greater CGRP and SP immunoreactivity upon immunostaining of skin and this persisted for one week post-infection (Ruiz et al., 2003). Four and 8 weeks after infection a decrease in CGRP innervation was observed (Ruiz et al., 2003). In another study, susceptible (BALB/c) and resistant mice (C57BL/6) mice were injected subcutaneously with Leishmania major parasites. By 1 week post-infection CGRP levels were reduced in the injected skin (Ahmed et al., 1998) of both mouse strains. The CGRP concentration was increased 1week post-infection in the ipsilateral dorsal root ganglia in the resistant C57BL/6 mice while no change was seen in the susceptible BALB/c mice. A reduction in the CGRP level of the ipsilateral dorsal root ganglia in both strains at later timepoints (Ahmed et al., 1998). In nude mice infected with Mycobacterium leprae in the hind footpads, SP- and CGRP-immunoreactive nerves were reduced 6 months after infection and to a greater degree 12 months after infection (Karanth et al., 1990).

These findings suggest a complex interaction between stress, neuropeptide expression, and cutaneous immunity. Stress may enhance expression of neuropeptides such as CGRP in the skin, which, as discussed earlier, appears to shift LC antigen presentation away from Th1-mediated responses and toward Th2-mediated responses. This shift may then increase susceptibility to infection by certain parasites and bacteria. Whether different types of skin infections lead to changes in the neuropeptide milieu that are specific for enhancement (or degradation) of immunity to those infections is a question of great interest that should be examined.

The in vivo immunoregulatory activity of CGRP was demonstrated by the observation that targeted expression of CGRP to pancreatic beta cells in nonobese diabetic (NOD) mice prevented insulin-dependent diabetes in male mice and reduced its incidence in females by 63% (Khachatryan et al., 1997). Also of interest, knock-out mice lacking the VPAC2 receptor (Goetzl et al., 2001) as well as mice deficient in PACAP (Kemény et al., 2010) exhibit enhanced contact hypersensitivity, consistent with the experiments cited above (section 2.2). However, not all results are easily interpreted. While norepinephrine inhibits LC antigen presenting function and inhibits the induction of contact hypersensitivity after intradermal administration (Seiffert et al., 2002), release of norepinephrine from sympathetic nerves due to psychological stress promotes dendritic cell migration and antigen-specific T cell responses (Saint-Mezard et al., 2003). It is likely that the in vivo effects of some of these factors depend on the temporal and spatial loci of release relative to immune challenge.

Experiments, described above, demonstrating that ATP induces ECs to release chemokines and upregulate ICAM-1 (Fig. 1) might explain how stress exacerbates inflammatory skin disorders. In this model, stress activates the sympathetic nervous system leading to the release of ATP by sympathetic nerves associated with dermal blood vessels. The ATP then binds to receptors on ECs causing the release of chemokines and the upregulation of ICAM-1. These events mediate enhanced leukocyte binding to ECs and diapedesis of leukocytes, resulting in an inflammatory cell infiltrate with the appearance or exacerbation of inflammation clinically.

While this model is highly speculative, it is consistent with the existing experimental data. In support of this concept, there are animal models that define activities and elucidate mechanisms by which the nervous system might regulate immune processes. As outlined above, considerable work has been done to define these mechanisms within the skin. It is not surprising that the effects of stress and neurologic derangement on cutaneous inflammatory disorders are more easily appreciated than in other inflammatory disorders, as the skin is an organ in which the gross pathology of almost any disease is visible. The skin is almost uniquely accessible for observation, biopsying and experimentation, and thus it serves well as a substrate for experiments examining the neuroimmune axis.

7. Conclusions

Although studies on the ability of the nervous system to regulate cutaneous immunity remain in their infancy, the area holds great promise for the development of novel techniques to benefit individuals suffering from disorders characterized by unwanted immune reactivity or, conversely, inadequate immune reactivity such as in the setting of infection or response to vaccines. Many of the clinically relevant mechanisms of nervous system-immune system interactions remain unknown. For example, the mechanisms of regulation of release or non-release of immune regulatory factors by nerves in the skin have yet to be elucidated.

Interestingly, some immune cells themselves can produce regulatory neuropeptides; rat thymocytes, B cells, Th2 cells and peritoneal macrophages (Gomariz et al, 1994; Delgado et al., 1996; Delgado et al., 1999; Vassiliou et al., 2001; Pozo and Delgado, 2004) express VIP and PACAP is expressed in rat thymocytes, B cells and in T cells (Abad et al., 2002–2003). CGRP is expressed in rat thymus and lymph node T cells as well as human blood and spleen T cells (Wang et al., 1999; Xing et al., 2000) and human monocytes (Linscheid et al., 2004; Bracci-Laudiero et al., 2005). These observations suggest an important regulatory role for neuropeptides in immune cell interactions that deserves much more investigation. VIP-knock-out mice infected with murine cytomegalovirus showed better survival, evidence of increased antiviral cellular immunity and faster clearance of virus compared to wild-type controls (Li et al., 2011). They further showed in radiation chimeric mice reconstituted with subsets of hematopoietic cells from knock-out and wild-type donors that VIP production by T cells alone is sufficient to attenuate anti-viral cell-mediated immunity (Li et al., 2011). Of potential significance for nerve-immune physiology, nerve growth factor (NGF) reportedly upregulated CGRP expression by B lymphocytes (Bracci-Laudiero et al., 2002) and monocytes, suggesting a novel mechanism by which NGF may regulate immunity (Bracci-Laudiero et al., 2005; Bracci-Laudiero et al., 2002). We are only just beginning to understand the complex roles of neuropeptides in mediating immune function.

The skin is immediately accessible; thus, it is well-suited for study and for therapeutic manipulation. One can envision the development of topical agents that stimulate or inhibit neurotransmitter or neuropeptide signaling pathways to achieve cutaneous benefits with, perhaps, only a negligible systemic concentration, thereby avoiding systemic adverse effects. A more complete understanding of nervous system-immune interactions would likely lead to unexpected novel observations and therapeutic approaches.