Annette Sommershof & Marcus Groettrup
49
Abstract
Chronic stress is not only implicated to increase the susceptibility to infectious diseases but substantial evidence has linked chronic or recurrent exposure to stress with exacerbation of several inflammatory and autoimmune skin disorders. Skin DCs and in particular Langerhans cells (LCs) are known as critical inducers of cutaneous immune responses, however their fate under chronic social stress situations is less documented. We examined the effects of chronic social stress on the migratory capability of skin DCs after epicutaneous skin sensitisation with the contact allergen FITC. Mice subjected to social stress on six consecutive days prior to contact sensitisation showed a significantly reduced migration of FITC-bearing CD11c+ DCs to regional lymph nodes. Pharmacological blockade of β-adrenergic receptors with nadolol did not reverse the stress-associated decline of CD11c+ DCs migration. Furthermore, in vitro migration assays using ear skin explants showed that CD11c+ DCs from SDR mice emigrated less efficiently out of the skin. These results point at a potential mechanism of how stress could negatively influence cutaneous immune responses.
Introduction
Psychological stress has long been suspected to alter skin immunity and to play an important role in the pathophysiology of numerous inflammatory skin disorders. For example, human epidemiological studies have shown that stressful life conditions are associated with exacerbated symptoms of psoriasis and atopic or allergic contact dermatitis (266-268).
Experimental evidence that chronic psychological stress can indeed affect cutaneous immune responses, for example to contact allergens, arouse from mouse models investigating the impact of various stress conditions on the outcome of experimentally induced allergic contact dermatitis (contact hypersensitivity - CHS or delayed-type hypersensitivity - DTH reactions).
Moreover, recent studies analysed the contribution of stress hormones on the outcome of DTH reactions, demonstrating that exposure to high levels of corticosterone or prolonged exposure to moderate levels suppresses skin DTH reactions in response to the contact allergens dinitrofluorobenzene (DNFB) or oxazolone (OXA), suggesting that GC hormones are major mediators of the stress-induced suppression of DTH reactions (221). Other studies suggest that GC-independent mechanisms also participate in the stress-related suppression of contact sensitivity. For instance, it has been demonstrated that local intradermal injection of epinephrine at the time of allergen-induced sensitisation inhibited the induction of a DTH response to epicutenously administered DNFB (269).
Evidence that LCs are major contributors in CHS reactions aroused from the observation that, during sensitisation of the epidermis they are able to recognise, internalise, and process reactive haptens encountered at the skin surfaces, and transport them from the skin via afferent lymphatics to T cell areas of regional lymph nodes (32). Moreover, recent studies demonstrated that the inducible ablation of LCs in adult Langerin-DTR mice results in a diminished CHS response to different haptens (18) and suboptimal priming of CHS-effector T cells due to inefficient transport of the Ag from the epidermis (19). In this context it has been
Chapter 3
50 shown that in vitro pre-treatment of LCs with epinephrine or norepinephrine resulted in an impaired antigen presentation of these cells, leading to a weakened DTH reaction when reinjected into earlier immunised mice. These results may provide a mechanism for the suppressive effects of stress on DTH responses via inhibition of Ag presentation (269).
Studies in mice investigating the effects of GCs on LC function showed that topical corticoisteroid treatment induces apoptosis in LCs and inhibit the expression of co-stimulatory cytokines (270).
Although there is now increasing evidence that both neuroendocrine systems, the HPA axis and the SNS, are able to modulate contact allergen-induced cutaneous immune responses, little is known about their respective contributions under physiological chronic stress conditions (271). Moreover, the underlying mechanisms by which GCs and/or catecholamines may exert their function to influence certain immune cells engaged in contact hypersensitivity reactions are almost lacking. In this regard, skin DCs, which are implicated to play a pivotal role in the initiation of cutaneous immune responses to contact allergens and in skin allergic diseases, have become the focus of recent research. There is novel evidence that a perturbation of the migratory capability of skin DCs and in particular epidermal LCs is involved in pathologies like psoriasis (272) and autoimmune dermatitis (273).
Given the undisputable role of stress in the pathogenesis of skin disorders, it is important to understand whether and how chronic stress alters skin DC function, particularly in context of their migration capacity, and how these alterations contribute to skin disorders. Our primary aim was to investigate a potential stress-induced modulation of skin DC migration and to identify neuroendocrine modulators mediating these alterations.
Results
Diminished in vivo skin DC migration in SDR mice
In order to elucidate the impact of chronic social stress on the migratory capability of skin DCs, we performed a FITC-induced in vivo migration assay. Fluorescein isothiocyanate (FITC) is a fluorescent marker for migratory DCs and has been used in in vivo epidermal LC migration assays since the 1980ies (274-276). Topical FITC application has been shown to preferentially induce LCs in the epidermis, and FITC-bearing cells in the draining LNs were found to be primarily derived from LCs (277-278). Mice were stressed for six days and the FITC-solution was applied 20 hours after the last stress exposure to the abdomenof SDR and control mice, and the presence of FITC-bearing CD11c+ cells in the inguinal LNs was analysed 24 hours after sensitisation.
51 Figure 1: Effects of social stress on the migratory capability of skin DCs to draining LNs
SDR and control mice were painted with 30 µl of 1% FITC in acetone/dibutylphalate (1/1) on the abdomen.
24 h later, inguinal LNs were collected and single-cell suspensions were stained with APC-labelled anti-CD11c antibody and analysed by flow cytometry for the percentage of FITC+ DCs. (A) Dot plot of CD11c staining and histograms for control and SDR mice are shown, representing the percentage of recovered FITC+ positive cells pre-gated on CD11c+ cells. Negative control represents the FITC background fluorescence obtained from naïve mice. (B) Data are presented as the mean percentage of FITC-positive cells ± SEM of the CD11c+ population.
As shown in Figure 1, 24 hours after FITC-treatment, the proportionof FITC+ CD11c+ cells in the inguinal LN of control mice was markedly higher compared to SDR mice (control:
13.3±2.1% vs. SDR: 4.1±0.5%), indicating that exposure to chronic stress prior to hapten stimulation results in a substantial impairment of CD11c+ cell migration from the epidermis into skin draining LNs. Although we did not provide direct evidence that the immigrated FITC+ cells in the LN originate from the epidermis, hence reflecting LCs, it has been demonstrated previously that after epicutaneous FITC sensitisation the majority of FITC-bearing cells are derived from epidermal LCs (277-278).
Phenotype of migrated CD11c+ skin DCs
In order to correlate the migration of CD11c+ cells with their maturation induced by FITC sensitisation, we analysed the expression of CD86 and CCR7 on recovered FITC+ CD11c+ cells in the inguinal LNs.
As shown in Figure 2 A, immigrated FITC-bearing DCs exhibit a strong expression of the costimulatory molecule CD86 and the chemokine receptor CCR7, compared to resident FITC -DCs exhibiting a semi-mature phenotype with intermediate levels of both molecules.
Moreover, migrated FITC+ CD11c+ from stressed mice did not significantly differ in their expression of CD86 (Figure 2 B) and CCR7 (Figure 2 C), compared with those of unstressed control mice. These data demonstrate that by the time they reach the draining LNs, skin derived DCs apparently display a mature phenotype with respect to CD86 and CCR7 expression. These observations are in line with previous reports, demonstrating that full maturation of skin DCs is a fundamental requirement for their migratory competence under inflammatory conditions (15).
Chapter 3
52 Figure 2: Phenotype of migrated FITC+ CD11c+ cells in the LN
The activation state of migrated FITC+ CD11c+ cells in the inguinal LNs was assessed 24h after abdominal FITC-sensitisation by their expression level of CD86 and CCR7. (A) Representative dot plots of CD11c staining and contour plots, representing FITC versus CD86 and CCR7 profiles of the pre-gated CD11c+ population are shown. (B, C) Data in the bar plot are presented as mean fluorescence of CD86 (B) and CCR7 (C) expression ± SEM of migrated FITC+ CD11c+ cells in control and SDR mice.
Role of catecholamines in the modulation of skin DC migration
To identify the endocrine factors mediating the stress-induced decrease of skin DC migration, SDR and control mice were implanted two days prior to the stress procedure with continuous release-pellets containing the β-adrenergic receptor antagonist nadolol.
Figure 3: Skin DC migration in SDR and control mice treated with slow-release pellets of nadolol or placebo
The frequency of FITC+ CD11c+ cells in the inguinal LNs of placebo-treated SDR mice was not different from that in nadolol-treated mice. Data are presented as the mean percentage of FITC-positive cells ± SEM of the CD11c+ population.
Unstressed control mice treated with nadolol exhibited similar percentage of FITC-positive cells in the inguinal LNs compared to mice receiving placebo pellets, demonstrating that nadolol did not influence the skin DC migration by itself. Notably, the frequency of FITC+ CD11c+ cells in nadolol-treated SDR mice was not significantly altered compared to
placebo-53 treated SDR mice (SDR nadolol: 10.3±1.4% vs. SDR placebo: 10.7±1.6%), indicating that β-adrenergic receptor blockade was not effective in preventing the stress-associated changes of CD11c+ migration into the LN (Figure 3).
Impaired ex vivo emigration of skin DCs in SDR mice
Two possibilities could explain the reduced migration and accumulation of CD11c+ in the skin draining LNs of SDR mice. First, inefficient DC migration is caused by an inability of adequate DC numbers toemigrate out of the skin. The other possibility is that skin DCs are mobilised to the same degree, but fail to enter the afferent lymphatics and/or transit to a lesser extent via lymph to the LN. The latter scenario might involve a defective synthesis of chemotactic signals important for LN homing such as the CCR7 ligands ELC/CCL19 and SLC/CCL21. To test whether the impaired migration of skin derived CD11c+ observed in SDR mice reflect a diminished egress of activated DCs from the skin, we evaluated the migratory potential from SDR and control mice ex vivo inaskin ear explant model.
Mouse skin cultures are commonly used to study skin DC migration in ear skin cultures as an
"ex vivo chemotaxis assay" (279) and previous studies have demonstrated that the migration of DCs out of ear skin explants into the culture medium reflect, in part, their in vivo migratory capacity (280). For example in vitro DCs migrate into the dermis to form so called “dermal cords” in the region of dermal lymphatics as they do under in vivo conditions, before continuing to migrate spontaneously into the medium (280). Moreover, it has been demonstrated that the vast majority of skin-emigrated DCs in the medium represent epidermal LCs (32, 280) and that SLC/CCL21 and ELC/CCL19 induce egress of LCs from explanted skin (279).
Stressed and control micewere sensitised with FITC-solution on the dorsal and ventral side of each ear. Twenty-fourhours later, ears were excised and ventral ear sheets were culturedin presence or absence of relevant chemokines to stimulate the release of DCs from the skin-tissues. As illustratedin Figure 5 A, an average of about 2.300 DCs, identified as CD11c+ cells spontaneously emigrated out of one ventral ear half into the culture medium in the absence of exogenous cytokines. In contrast, the number of ex vivo emigrated CD11c+ cells was markedly reduced in skin explants from SDR mice (control: 2236±446 vs. SDR:
1082±95), indicating that DCs from SDR mice exhibit a reduced capability to egress outof the skin.
Chapter 3
54 Figure 4: Effects of social stress on the mobilisation of skin DCs in ear explants
Mice were subjected to six days of SDR and 20h after the last stress exposure mice were contact-sensitised with 1% FITC solution on both ears. 24 hours after contact sensitisation ears were removed and split into dorsal and ventral halves. Ventral ear halves were incubated dermis-side down in the (A) absence (w/o) or presence of (B) SLC and (C) ELC. After 24 hours, cells that migrated ex vivo from the tissue into the culture medium were collected and stained for CD11c expression. The number of migrated CD11c+ cells was enumerated using an Accuri flow cytometer.
Because DCs areattracted to lymphatic vessels by CCR7-SLC and ELC interaction (281-282), we further determined the magnitude of CD11c+ egress in response to both chemoattractants. As shown in Figure 4 B, C, the migratory potential of skin DCs from SDR mice was also significantly lower in the presence of SLC/CCL21 (Figure 4 B; control:
3193±546 vs. SDR: 1558±44) and ELC/CCL19 (Figure 4 C; control: 3020±238 vs. SDR:
1171±228) with respect to the control counterparts, suggesting that the impaired skin DC mobilisation in SDR mice is not due to a lack of these chemokines.
Since both SLC/CCL21 and ELC/CCL19 are known to be potent chemoattractants for LCs in skin explant cultures and the presence of exogenously added chemokines is established to significantly increase emigration (283-284), absolute numbers of emigrated CD11c+ cells in our experiments cannot be directly compared since the experiments were performed independently and hence vary in absolute cell numbers.
Discussion
In this study we demonstrate that chronic social stress suppresses the migratory capability of skin DCs to regional LNs after contact sensitisation with FITC. These findings were associated with an impaired mobilisation of CD11c+ cells from the skin, as revealed by the frequency of CD11c+ cells migrated ex vivo out of ear skin explants. Moreover, we could show that an action of catecholamines at peripheral β-adrenergic receptors is not involved in the suppressive effect of stress, since in vivo blockage of these receptors by the specific antagonist nadolol did not reverse the stress-associated decline of skin DC migration. We are currently investigating whether GCs are the major mediators of the stress-induced altered skin CD11c+ migration.
Although we did not distinguish between the dermal DC populations and epidermal LCs in our experiments, it is likely that the reduced migration of skin CD11c+ cells reflects an
55 inhibition of epidermal LCs. Previous studies demonstrated that topical FITC application preferentially induces LC migration in the epidermis, and FITC-bearing cells in the draining LNs were found to be primarily derived from LCs (277-278). Thus skin DC migration in this model is mainly independent of other dermal DCs and allows conclusion of the involvement of epidermal LCs. Nevertheless, we cannot rule out the possibilitythat FITC passes through the epidermis and penetrates the dermis to bind to dermal DCs, therefore at least some of the FITC-labeled cells recovered in the draining LNs may represent dermal DCs.
The precise mechanism by which neuroendocrine factors act to inhibit the mobilisation of CD11c+ DCs from the skin is not yet clear. Upon skin sensitisation a rapid release of inflammatory cytokines such as TNF-α and IL-1β induces a variety of mechanisms ensuring the mobilisation of LCs from their environment as well as their homing towards regional LNs (16). Consistent with this idea, it was demonstrated that in response to epicutaneous FITC-application, mRNA levels of TNF-α and IL-1β in epidermal cells are rapidly up-regulated (32). Moreover, inhibition of contact allergen-induced LC migration by prior treatment of mice with neutralizing TNF-α mAbs or an IL-1β antagonist is associated with both impaired epidermal egress and LN accumulation (23, 32). On the one hand, both TNF-α and IL-1β are needed for breaking up E-cadherin bonds endowing the detachment of LCs from neighbouring keratinocytes and their subsequent mobilisation within the epidermis (26-27).
On the other hand, the migratory capability of LCs is tightly linked with the concomitant up-regulation of the chemokine receptors CXCR4 and CCR7 (15) that attract and guide migrating LCs from the epidermis into the lymphatic vessels and further on into the lymph nodes (41). In this context, TNF-α has been shown to induce the up-regulation of CCR7 in LCs (283). Thus a reduced TNF-α and/or IL-1β de novo synthesis or secretion will result in decreased LC detachment and accumulation of LCs in the epidermis, and such mechanism could contribute to our observation. Indeed, there is compelling evidence that GCs are generally able to negatively regulate TNF-α and IL-1β expression both at transcriptional and post-transcriptional levels (285-287). Moreover a previous study demonstrated that the synthetic glucocorticoid DEX affects allergen- and IL-1β −mediated but not TNF-α-induced LC migration and accumulation in the LNs, suggesting that the impaired LC migration in this model is predominantly caused by inhibition of TNF-α secretion by keratinocytes (288).
Since the stress procedure occurs prior to the contact sensitisation, an alternate explanation for the observed impairment of skin DC migration could be a stress-induced decrease of absolute CD11c+ DC numbers in the skin. This reduction could in turn be the result of a GC-mediated apoptosis or a redistribution of DCs in response to the stress procedure, which has occurred prior to the contact sensitisation. In this context it is of importance to note that physiological stress responses are generally associated with significant changes in the distribution of leukocytes within the body (234, 289). For instance, it has been shown that chronic social stress in rats results in decreased numbers of TCD8+ and TCD4+ cells in the blood of subordinate animals (289). Moreover, exposure of mice to social disruption stress is associated with a recruitment of CD11b+ leukocytes from the bone marrow to the blood and the spleen as well
Chapter 3
56 as a reduction of total T cell numbers in the spleen (234). Social stress-induced reduction of T cells in the blood and spleen could also be due to a redistribution of T cells to other immune compartments. This view is supported by a study demonstrating that radiolabeled, adoptively transferred blood T cells accumulate in the bone marrow of defeated recipient rats, wherease their recirculation in the spleen and lymph nodes is decreased (254). Such changes in leukocyte distribution may have significant implications for the outcome of chronic stress exposure on cutaneous immune responses after contact sensitisation (290).
Thus the experiments conducted in this study are yet preliminary and factors triggering the impaired skin CD11c+ DC accumulation in the LNs of SDR mice need to be further elucidated. Moreover our experiments clearly contain several potential limitations, as phenotypic characterization of migrated CD11c+ cells needs to be conducted, allowing an ultimate differentiation between epidermal EpCAM+ LC and the different subsets of EpCAM -dermal DCs. It has been demonstrated previously that the majority of FITC-bearing cells derive from epidermal LCs (277-278). Moreover, additional analysis time points after FITC-sensitisation are required to elucidate the kinetics and magnitude of CD11c+ arrival in the LN, and to determine whether the reduced accumulation observed in SDR mice reflect a delayed arrival or a general migration deficiency.
Various studies have examined the effects of psychological stress on Ag-specific skin immunity by analyzing the outcome of delayed type hypersensitivity reactions (DTH) in
Various studies have examined the effects of psychological stress on Ag-specific skin immunity by analyzing the outcome of delayed type hypersensitivity reactions (DTH) in