Participants
We examined the distribution of blood T lymphocyte subsets in 19 individuals with current PTSD (12 male, 7 female; mean age = 33.6 years, SD = 7.1, range 21-48) according to the DSM-IV (American Psychiatric 296) and 27 non-PTSD control subjects (9 male, 18 female;
mean age = 29.1 years, SD = 8.3, range 19-50). PTSD patients were refugees (4 Africa, 1 Balkan, 14 Middle East and Afghanistan), with chronic (mean symptom duration = 7.2 years, SD = 4.4) and severe (mean sum score in the Clinician Administered PTSD Scale [CAPS]
(343) = 79.6, SD = 18.6) forms of PTSD due to multiple highly stressful war and torture experiences. On average, patients have lived in Germany for 4.9 years (SD = 3.6). All patients were recruited from the Psychotrauma Research and Outpatient Clinic for Refugees, University of Konstanz, located at the Centre for Psychiatry Reichenau, Germany.
The non-PTSD group was recruited through advertisement and was matched to the patient group with regard to age and region of origin (3 Africa, 11 Balkan, 13 Middle East and Afghanistan). Since this control group varied with respect to the number of traumatic event types experienced (range: 0 - 9) some of the analyses were repeated with a three group (PTSD, trauma-exposed and non-exposed controls) design. For this purpose we divided the non-PTSD group by median split into a group with substantial exposure to traumatic stressors (4 - 9 different traumatic event types; n = 14) and a control group with no or few traumatic experiences (0 - 3 traumatic event types; n = 13) based on the number of past traumatic event types assessed with the event checklist of the CAPS (343).
Subjects were excluded if they reported intake of glucocorticoids, had acute or chronic somatic illnesses, or met criteria for additional mental disorders other than stress-related affective or anxiety disorders. Fourteen PTSD patients and 2 trauma-exposed controls met the DSM-IV criteria for a current major depressive episode. Eight PTSD patients and 2 trauma-exposed controls reported current intake of psychotropic medication (PTSD: 2 hypnotics, 3 anxiolytics, 5 antidepressants and 2 neuroleptics; non-PTSD: 1 hypnotic, 1 antidepressant).
Since the pattern of results did not change if we excluded all medicated participants from the statistical analysis, we only report the original analysis here.
Chapter 4
70 Clinical interviews
All participants underwent an extensive standardized clinical interview administered by experienced psychologists and trained translators. PTSD symptoms and the number of traumatic event types experienced were assessed with the CAPS (343). The vivo checklist of war, detention and torture events (344), which assesses common traumatic experiences in conflict regions and during torture, allowed for a detailed evaluation of the number of traumatic event types experienced. The Mini International Neuropsychiatric Interview (M.I.N.I.) (345), was used to screen for potential comorbid mental disorders. In addition, the severity of depressive symptoms was assessed with the Hamilton Depression Rating Scale (HAM-D) (346). After complete description of the study to the subjects, written informed consent was obtained. All procedures were approved by the Ethics Committee of the University of Konstanz.
Blood sampling
Blood was collected between 10 and 11 a.m. in EDTA-treated tubes for T cell phenotyping and in sodium citrate-treated cell preparation tubes for proliferation assays (BD Vacutainer, Franklin Lakes, NY). In order to control for possible HIV and hepatitis A, B and C infections, an additional blood sample was sent to a diagnostic laboratory for standard hepatitis and HIV tests. All samples were negative for HIV or hepatitis C. Subjects classified with acute or chronic hepatitis A or B (n = 3) were excluded from the study. Two patients and one traumatized control showed an infection history for hepatitis B (as indicated by a positive result for hepatitis B core IgG antibody). Since the pattern of results did not change if we excluded them from the statistical analysis, they remained in the sample.
Lymphocyte phenotyping and T cell proliferation
Whole blood was analysed for the percentage of total T cells (CD3+), cytotoxic T cells (CD3+ CD8+) and T helper cells (CD3+ CD4+) as well as B cells (CD45+CD19+), by flow cytometry.
T cell maturation subsets were determined according to their expression profile of the surface molecules CD45RA and CCR7. For quantification of T cell phenotypes, 100 µl whole blood was incubated for 20 min at room temperature with either APC-conjugated anti-CD3 (clone SK7) or a combination of PerCP-conjugated anti-CD3 and APC-conjugated anti-CD8 (clone SK1) or APC-conjugated anti-CD4 (clone RPA-T4), and PE-conjugated anti-CD45RA (clone HI100) and FITC-conjugated anti-CCR7 (clone 150503) monoclonal antibodies (mAbs). For quantification of B-lymphocytes 100 µl blood was stained with PerCP-conjugated anti-CD45 (clone 2D1) and APC-conjugated anti-CD19 (clone HIB19). For quantification of Treg cells, blood samples were stained with PerCP-conjugated anti-CD3, APC-conjugated anti-CD4, FITC-conjugated anti-CD25 (clone M-A251), and intracellular FoxP3 expression was detected using the PE anti-human FoxP3 staining kit (eBioscience, San Diego, CA).
Following antibody staining, standard lyse-wash was performed using BD FACS lysing solution; samples were washed twice, and 1×105 cells were acquired on a FACSCalibur flow cytometer (BD Immunocytometry Systems, San Jose, CA), and analyzed with FlowJo
71 software (Tree Star, San Carlos, CA). All monoclonal antibodies were purchased from BD PharMingen (San Diego, CA), except CCR7 mAb from R&D Systems (Minneapolis, MN).
Absolute lymphocyte numbers (cells/µl) were measured, using an automated hematology analyzer (XT-2000i, Sysmex, Horgen, Switzerland).
For the proliferation assay, 1×105 CFSE-labeled peripheral blood mononuclear cells (PBMCs) were suspended in RPMI medium containing 10% FCS and stimulated for 72h in 96-well flat-bottom microtiter plates coated with anti-human CD3 mAb (2 µg/ml, clone OKT3, eBioscience), and cell proliferation was measured by flow cytometry in triplicates. The investigator who performed the immunological analyses was blind for the group assignment of the probes.
Statistical analyses
Group differences in the immunological parameters were analyzed using ANOVAs. The independent variables were either two (PTSD, non-PTSD) or three groups (PTSD, trauma-exposed and non-trauma-exposed controls). Statistical significance for the immune measures was assessed by nonparametric permutation tests, using 1000 random permutations of group labels (347). Throughout the text and the tables all data are presented as mean ± standard deviation.
In the figures data are displayed as mean + standard errors.
Chapter 5
Discussion
73 Associations between stress and immune functions have been carefully documented in the past, but the specific mechanisms by which chronic stress influences disease susceptibility and outcome remain not fully understood. The aim of the present thesis was to further analyse underlying mechanisms of stress-induced immunosuppression. In chapter 2 we characterised the mechanisms by which chronic social stress affects the outcome of TCD8+ cell-mediated responses during a virus infection in a mouse model. Chapter 3 describes the impact of chronic social stress on the migratory capacity of skin dendritic cells after contact allergen sensitisation in mice. In chapter 4 we describe traumatic stress-associated alterations in peripheral T cell subsets in humans.
Chapter 2 focuses on the impact of chronic social stress on TCD8+ cell-mediated immune responses in a mouse model of LCMV-infection. Unlike most previous studies we directly compared the outcome of different stress protocols in our model. Mice that were subjected to social stress prior to virus infection exhibited a significant reduction of IFN-γ-producing TCD8+
cells specific for the dominant LCMV-derived epitopes GP33 and NP396. Comparison of different readout systems allowed the conclusion that the stress exposure mainly impacts the function of splenic TCD8+ cells by inhibiting antigen-specific IFN-γ secretion. In contrast, the generation of IFN-γ producing TCD8+ cells was not significantly altered in mice receiving the same stress procedure concurrently with the infection. Previous studies in mice have clearly demonstrated that chronic social stress exposure can suppress the generation of TCD8+ cells during the course of a viral infection under some conditions, but it does not do so under all conditions. Our results extend previous work by demonstrating that the timing of the stress exposure is an important factor in determining the direction of the stress-induced immune alterations (246) and add new insight under which circumstances chronic stress results in an immunosuppressive reaction.
We also found that prolonging the stress procedure for three additional days after LCMV-infection decreased the expansion of activated TCD8+splenocytes. Pharmacological blockage of GC receptors in stressed and control mice revealed that these alterations were mediated by GCs. Although GCs are known to induce apoptosis in T cells, we found no differences in the propensity of TCD8+ cells to undergo apoptosis whereas the in vivo proliferation of TCD8+ cells was markedly reduced. Analysis whether the reduced expansion of TCD8+ cells correlates with an inefficient initial activation showed that TCD8+ cells of stressed mice display a significantly lowered expression of the early T cell activation marker CD69. These results suggest that the impairment of TCD8+ cells occurs at the earliest stages of a viral infection.Induction of CD69 on TCD8+ cells is known to induce IL-2, IFN-γ and CD25 expression (62-65), thereby promoting the proliferation of effector TCD8+ cells in an autocrine manner. Analysis of the cytokine secretion capacity of isolated TCD8+ cells proved that TCD8+ cells from stressed mice exhibit an impaired IL-2 and IFN-γ secretion when restimulated in vitro. Previous studies have shown that GCs suppress the activation, proliferation and cytokine production of TCD8+
cells in vitro. These effects have also been confirmed in splenic TCD8+ cells isolated from stressed mice that underwent the social disruption stress procedure. Our findings place these
Discussion
74 previously reported cellular mechanisms of GC-mediated impairment of TCD8+ cell function into the context of a relevant in vivo viral infection model.
It was previously reported that GCs could impair DC maturation and Ag presentation in vitro (209-210). A very recent report has also demonstrated that DCs are the main targets of stress/glucocorticoids in vivo in a model of HSV-1 infection (348). We excluded the possibility of an altered maturation state by analyzing the expression of the costimulatory molecules CD80 and CD86 on splenic DCs. Moreover, we found no differences in their competence to present LCMV-derived epitopes in vitro. We also excluded the possibility that an altered composition of the 20S proteasome contributes to the reduced generation of TCD8+
cells in stressed mice. The impaired TCD8+ responses in stressed mice thus cannot be assigned to an alteration of dendritic cells in our model. Therefore we propose that GCs act directly on TCD8+ cells early during their activation phase. The controversial outcome compared to the previous report (348) could be due to the different stress models used to induce chronic stress in mice (e.g. restraint stress versus social stress; daily recurrent short exposure versus recurrent long exposure). In this context it is important to note that only social disruption stress has been shown to induce an insensitivity of splenic CD11b+ monocytes and CD11c+ DCs to the inhibitory effects of GCs (236). Although the GC insensitivity of CD11c+ DCs is less well documented, it is feasible that this mechanism may contribute to our observation of an unimpaired DC function.
Other observations in our model are less well understood. We cannot yet provide a reasonable explanation why the reduced expansion of TCD8+ cells appears only in the spleen but not in the inguinal lymph node or peripheral tissues. In general, the expression of GC receptors (GRs) is distinct in different tissues and immune cell types and reflects their unique sensitivity to GCs (215). Adoptive transfer experiments showed that splenic TCD8+ cells derived from SDR mice exhibit an impaired in vivo migration to the spleen. An elevated egress of TCD8+ cells might compensate for the reduced generation in the spleen and could provide the peripheral compartments with sufficient effector cells.
Chronic stress is not only implicated to play an important role in the susceptibility to infectious diseases but numerous studies in humans and mice provide accumulative evidence that chronic stress can impact the pathophysiology of several skin inflammatory and autoimmune disorders. In the past, the impact of chronic stress on skin immunity in mice has been focused on antigen-specific T cell-mediated immune reactions in response to contact allergens (DTH reactions). The precise mechanisms how stress hormones (GCs and catecholamines) down-regulate cutaneous immune functions are yet poorly understood.
Although skin DCs and in particular LCs are known as critical inducers of cutaneous immune responses, their fate under chronic stress situations is less documented. In chapter 3 we focused on the effects of social disruption stress on the migratory capability of skin DCs. We show that chronic social stress exposure prior to skin sensitisation with the contact allergen FITC strongly impacts the migration of epidermal FITC-bearing CD11c+ DCs to regional
75 lymph nodes. The impaired migration capacity of skin CD11c+ DCs could not be assigned to an action of catecholamines at peripheral β-adrenergic receptors since in vivo blockage of these receptors did not reverse the effect. This finding is of particular interest since previous reports have suggested a role for the sympathetic neurotransmitter norepinephrine (NE) on the migratory function of LCs (271). However, data on whether NE exerts its effects on LC migration by binding to α- or β-adrenergic receptors are less clear with one report pointing towards an involvement of α-adrenergic receptors (349) while another study suggested signaling through β-adrenergic receptors (350). With regard to stress-induced modulation of leukocyte trafficking it is well established that chronic social stress-induced increase in circulating neutrophils, monocytes and NK cells as well as the decrease in blood T and B cell numbers is at least partly mediated by stimulation of α- and β-adrenergic receptors (229).
Therefore further analyses are needed to elucidate whether catecholamine-mediated α-adrenergic signaling contributes to the altered CD11c+ DCs migration observed in our model.
In vitro assays using ear skin explants demonstrated that CD11c+ DCs from SDR mice emigrated less efficiently out of the skin, even in the presence of the CCR7-relevant chemokines ELC/CCL19 and SLC/CCL21. We conclude that the severe decrease in the accumulation of skin-derived CD11c+ DCs in the lymph nodes of SDR mice may occur through their retention in the skin. These results also suggest that the impaired epidermal DC mobilisation in SDR mice is not due to a lack of ELC/CCL19 and SLC/CCL21 production.
It has been shown that exposure to high as well as prolongedmoderate levels of exogenous corticosterone suppressed skin DTH reactions in response to the contact allergens DNFB or OXA (295). Related to chronic stress previous reports demonstrated a diminished DTH reaction in response to DNFB in mice that were exposed to restraint stress (290). The interaction of allergen-bearing skin CD11c+ DCs with naïve T cellsin the lymph nodes is pivotal for T cell priming and initiation of cutaneous immune responses induced by contact allergens.Vice versa, a deficiency or alteration of skinDC migration is associated with a less effective transport of the Ag to draining lymph nodes and presumably a less effective induction of the DTH response (278, 351-352). The defective migration of allergen-bearing CD11c+ DCs as observed in our model may therefore represent an initial event resulting in the previously described lack of DTH elicitation in stressed mice.
Although these initial findings do not yet deliver sufficient mechanistic insights, further studies along these lines are of high interest, especially with regard to the high prevalence of inflammatory and autoimmune skin disorders associated with stress. Recent reports have highlighted the important role of skin DCs and specifically LCs in the pathogenesis of skin diseases such as psoriasis and autoimmune dermatitis. Different mechanisms have been proposed how defective migration of LCs may contribute to the onset and progression of inflammatory and autoimmune skin disorders. For instance, LCs that retain within the epidermis could present antigens locally to sustain or exacerbate cutaneous inflammatory reactions (272). Another proposed mechanism aroused from the concept that LCs are not only important for protective T cell mediated immunity, but play also an important role in the
Discussion
76 maintenance of peripheral tolerance to self-antigens (353). Thus a reduced migration of LCs to skin draining lymph nodes could also result in a loss of self-tolerance. For instance in a mouse model of lupus dermatitis (autoimmune dermatitis-prone mice) FITC-activated LCs appear to accumulate in the skin, which was shown to precede the onset of dermatitis and correlated with the development and severity of skin inflammation (273). An improved understanding of the mechanism and consequences of impaired skin DC migration in response to chronic stress exposure may help to understand how stress increases the onset and progression of inflammatory and autoimmune skin disorders.
In chapter 4 we describe the phenotypic changes in T lymphocyte subsets in the peripheral blood of severely traumatised human patients. Posttraumatic stress disorder (PTSD) is associated with an enhanced susceptibility to various somatic diseases (300-302). In the past contradictory results were obtained when addressing the question whether an altered composition of the peripheral T cell compartment could be linked to the enhanced susceptibility of PTSD patients to infectious diseases and inflammatory or autoimmune disorders (302). In this thesis, we provide a more differentiated characterization of peripheral T cell subsets with regard to the high diversity of memory cells present among the peripheral T cell pool. Our results demonstrate that PTSD patients exhibit a reduction of naïve TCD3+
lymphocytes that was accompanied by an increased proportion of central (TCM) and effector memory (TEM) cells. Interestingly, the reduction of naïve and the increase of TEM cells were most prominent within the TCD8+ cell population, whereas no significant changes were observed for TCD4+ cells. The same tendency was also observed in trauma-exposed non-PTSD individuals, indicating a cumulative effect of exposure to traumatic stressors on T cell distribution. We could also show that PTSD patients exhibit nearly 50% reduction of regulatory T cells (Tregs) compared to non-PTSD individuals. This new finding is highly important given the fundamental role of regulatory T cells (Tregs) in suppressing immune responses to self-antigens and prevention of autoimmune diseases. These stress-related alterations of the peripheral T cell compartment described in our study might constitute a key factor in the enhanced susceptibility of individuals suffering from PTSD to a range of physical diseases.
Further studies are needed to clarify whether the observed changes in the distribution of T cell maturation subsets are due to alterations in the thymic output of naïve T cells or whether an altered peripheral T cell turnover contributes to the observed changes. A reduced emigration of naive T cells from the thymus may contribute to the reduction of naïve TCD3+ lymphocytes observed in PTSD patients. Such diminished thymic output could be analysed by quantification of peripheral recent thymic emigrants (RTE) and we are currently testing this possiblilty. On the other hand, the maintenance of naïve and memory T cells is distinctively regulated in the periphery. The peripheral naïve TCD8+ and TCD4+ cell pool is maintained through combinatory signals provided by TCR interactions with self-MHCclass II and class I ligands and by cytokines (354). Both IL-4 and IL-7 have been implicated to be important for the survival of naïve T cells in the periphery, while the role of IL-7 in supporting both
77 survival and homeostatic expansion of naïve T cells in vivo is more evident (354). For TCD8+
memory cells, it has been suggested that IL-4 plays a role in the early phases of memory generation, IL-7 in the intermediate stages and IL-15 in long-term maintenance (130). The nature of survival signals for TCD4+ memory cells is less well understood. In conclusion it would be important to determine whether an altered cytokine secretion profile may contribute to the observed reduction of naïve and the increase of TCM and TEM cells in PTSD patients.
memory cells, it has been suggested that IL-4 plays a role in the early phases of memory generation, IL-7 in the intermediate stages and IL-15 in long-term maintenance (130). The nature of survival signals for TCD4+ memory cells is less well understood. In conclusion it would be important to determine whether an altered cytokine secretion profile may contribute to the observed reduction of naïve and the increase of TCM and TEM cells in PTSD patients.