Brain, Behavior, and Immunity 22 (2008) 1108–1114
Contents lists available at ScienceDirect
Brain, Behavior, and Immunity j o u r n a l h o m e p a g e : w w w . e l s e v i e r. c o m / l o c a t e / y b r b i
Reducing post-traumatic anxiety by immunization Gil M. Lewitus a, Hagit Cohen b, Michal Schwartz a,* a b
Department of Neurobiology, The Weizmann Institute of Science, Rehovot 76100, Israel Ministry of Health, Mental Health Center Anxiety and Stress Research Unit, Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer Sheva, Israel
a r t i c l e
i n f o
Article history: Received 26 March 2008 Received in revised form 2 May 2008 Accepted 3 May 2008 Available online 17 June 2008 Keywords: Psychoneuroimmunology PTSD Lymphocytes Immunization Acute stress ICAM-1
a b s t r a c t Trafficking of T lymphocytes to specific organs, such as the skin and lungs, is part of the body’s defense mechanism following acute psychological stress. Here we demonstrate that T lymphocytes are also traf ficking to the brain in response to psychological stress and are needed to alleviate its negative behavioral consequences. We show that short exposure of mice to a stressor (predator odor) enhanced T-cell infiltra tion to the brain, especially to the choroid plexus, and that this infiltration was associated with increased ICAM-1 expression by choroid plexus cells. Systemic administration of corticosterone could mimic the effects of psychological stress on ICAM-1 expression. Furthermore, we found that the ability to cope with this stress is interrelated with T-cell trafficking and with the brain and hippocampal BDNF levels. Immuni zation with a CNS-related peptide reduced the stress-induced anxiety and the acoustic startle response, and restored levels of BDNF, shown to be important for stress resilience. These results identified T cells as novel players in coping with psychological stress, and offers immunization with a myelin-related peptide as a new therapeutic approach to alleviate chronic consequences of acute psychological trauma, such as those found in posttraumatic stress disorder. © 2008 Elsevier Inc. All rights reserved.
1. Introduction The response to stress is characterized by both emotional and physical manifestations, often leading to activation of various phys iological systems. This evolutionary adaptive response endows the organism with the ability to deal with the stressor by tem porarily adapting the body’s homeostasis to the novel situation. These stress response mechanisms are well regulated, and, in the absence of pathology, enable the return to normal homeostasis when the source of stress is removed. However, when homeostasis is not restored and maintained, long lasting changes can arise; in humans, these changes may lead to posttraumatic stress disorder (PTSD) (Yehuda and McFarlane, 1995). We have recently demon strated in an animal model that the ability to ward off the conse quences of stress is dependent on peripheral immunity (Cohen et al., 2006). Animals with immune deficiency show a reduced ability to deter the consequences of stress. For example, exposure of mice to predator odor results in higher anxiety and startle response in animals suffering from immune deficiency (Cohen et al., 2006). The same conditions were shown to impair cognitive ability, neu rogenesis, and expression of hippocampal brain-derived neurotro phic factor (BDNF) (Kipnis et al., 2004; Ziv et al., 2006b). Different mouse strains have contrasting abilities to cope with mental (Cohen et al., 2008; Shanks and Kusnecov, 1998) and phys ical stress (Kipnis et al., 2001). For example, C57BL/6J and Balb/c
* Corresponding author. Fax: +972 8 9346018. E-mail address: michal.schwartz@weizmann.ac.il (M. Schwartz). 0889-1591/$ - see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.bbi.2008.05.002
mice differ in their behavioral, endocrine and immune response following psychological stress (Shanks et al., 1990; Shanks and Kusnecov, 1998; Cohen et al., 2008). It is interesting to mention that, acute stress enhances delayed-type hypersensitivity in Balb/c but not in C57BL/6J mice (Flint and Tinkle, 2001). Similarly, after CNS injury, T cell infiltration was faster and the levels were higher in the Balb/c mice (Schori et al., 2007). We suggest that stress may enhance T-cell recruitment to the brain as a possible mechanism for brain maintenance. Since stress resilience is manifested by regulation of BDNF expression in the hip pocampus (Kozlovsky et al., 2007), and T cells were shown to affect hippocampal BDNF levels (Ziv et al., 2006b), we hypothesized that T-cell trafficking to the CNS would enable restoration of BDNF lev els and an increased ability to cope with stressful conditions. Here, we show that in mice, acute stress enhances immunetrafficking by up regulating ICAM-1 expression in the brain. Fur thermore, the stress-induced elevation of ICAM-1 could be partly reproduced by exogenous systemic application of corticosterone. We also showed an association between recovery of BDNF levels and increased immune trafficking. Immunization with a CNS spe cific antigen enhanced the adaptation to stress and restored BDNF levels in the stressed animals. 2. Materials and methods 2.1. Animals Adult wild-type mice of the BALB/c/OLA and C57BL/6J strains, all aged 8–12 weeks, were supplied and maintained under germ-free conditions by the Animal Breeding Center of The Weizmann Institute of Science (Rehovot, Israel). The mice
G.M. Lewitus et al. / Brain, Behavior, and Immunity 22 (2008) 1108–1114
were housed in a light- and temperature-controlled room and were matched for age in each experiment. All animals were handled according to the regulations for mulated by the Institutional Animal Care and Use Committee. 2.2. Experimental stress paradigm The mice to be tested (experimental group) were placed for 15 min on thor oughly soaked cat litter (used by a cat for 2 days and sifted for feces) (Cohen et al., 2006).
1109
anti mouse CD54 (ICAM-1) (Chemicon), goat anti mouse CD106 (VCAM-1) (R& D Systems), and rabbit anti BDNF (Alomone Labs). Secondary antibodies used for immunohistochemistry were Cy-3-conjugated donkey anti-rabbit, and Cy-3-conjugated donkey anti-goat. For ICAM-1 staining, biotin conjugated goat anti-hamster was applied for 1 hr, followed by streptavidincy3 for 15 min. All secondary antibodies were purchased from Jackson ImmunoRe search Laboratories Inc. Control sections (not treated with primary antibody) were treated with secondary antibodies to distinguish specific staining from nonspecific staining or autofluorescent components. Sections were then washed with PBS and cover slipped in polyvinyl alcohol with diazabicylo-octane as an anti-fading agent.
2.3. Behavioral testing 2.6. Quantification 2.3.1. Elevated plus-maze The maze we used is a black opaque Perspex platform with four arms in the shape of a plus, elevated 78 cm above the ground, as described by File (File, 1993; Griebel et al., 1995). Each arm was 24 cm long and 7.5 cm wide. One pair of opposite arms was “closed”, and thus the arms were enclosed by 20.5 cm high Perspex walls on both sides and on the outer edges of the platform, while the other pair of arms was “open”, surrounded only by a 3 mm high Perspex lip, which served as a tactile guide for animals in the open areas. The apparatus was illuminated by dim red lighting that provided 40–60 lux in both the open and the closed arms. Mice were placed one at a time in the central platform for 5 min, facing different arms on differ ent days according to a randomized sequence. Between test sessions, the maze was cleaned with an aqueous solution of 5% ethanol and dried thoroughly. Five behavioral parameters were assessed: (1) time spent in the open arms; (2) time spent in the closed arms; (3) number of entries into the open arms; (4) number of entries into closed arms; (5) total number of entries into all arms. Mice were recorded as having entered an open or closed arm only when all four paws crossed the dividing line between the arms and central platform. The number of entries into any arm of the maze (total arm entries) was defined as “exploratory activity”. 2.3.2. Acoustic startle response Pairs of mice were tested in startle chambers. The acoustic startle responses were measured in two ventilated startle chambers (SRLAB System; San Diego Instru ments, San Diego, CA). Each chamber consisted of a Plexiglas cylinder resting on a platform inside a ventilated, sound-attenuated chamber. Movement of the animal inside the tube was detected by a piezoelectric accelerometer located below the frame. The amplitude of the acoustic startle response of the whole body to an acous tic pulse was defined as the average of 100 accelerometer readings, 100 ms each, collected from pulse onset. The readings (signals) were digitized and stored in a computer. To ensure consistent presentation, sound levels within each test cham ber were routinely measured using a sound-level meter (Radio Shack, San Diego Instruments). An SR-LAB calibration unit was routinely used to ensure consistency of the stabilimeter sensitivity between test chambers and over time (Swerdlow and Geyer, 1998). Each startle session started with a 5-min acclimatization period to a background of 68 dB white noise; following habituation, 30 acoustic startle trial stimuli were presented (110 dB white noise of 40 ms duration with 30 or 45 sec inter-trial interval). 2.3.3. Rota-rod treadmills Motor strength and coordination were evaluated on the accelerating Ugo Basile Model 7650 Rota-rod apparatus (Ugo Basile, Camerio, Italy). Each mouse was placed on the cylinder, which increased rotation speed from 5 to 40 rpm over a 300 s period. Mice were first given three trails to become acquainted with the Rota-rod appara tus before the test. For detection, a group of 5 mice were placed on the rotating rod before starting the accelerated program. The time each mouse remained on the rod was registered automatically. If the mouse remained on the rod for 300 s (top speed of the rod) the test was completed and scored as 300 s.
BDNF and ICAM-1 immunoreactivity was quantified blindly with Image Pro Plus 4.5 software (Media Cybernetics) (Ziv et al., 2006b). 2.7. Immunization Adult mice were immunized with 100 lg pMOG35–55(MEVGWYRSPFSRVVHLY RNGK) (Lewitus et al., 2006; Schori et al., 2001), emulsified in an equal volume of CFA (Difco, Franklin Lakes, NJ) containing Mycobacterium tuberculosis (.5 mg/ml, Difco). The emulsion was injected s.c. at a single site in the flank. Control mice were injected with PBS emulsified with CFA. 2.8. Statistical analysis A two-tailed unpaired Student’s t-test was used for analyses of the experiments presented in Table 1. The data from the experiments presented in Figs. 1–4 and Supplementary Table 1 were analyzed by ANOVA, and means were compared using the Tukey–Kramer post hoc analysis test for differences between individual means. Values that differed at P < 0.05 were considered statistically significant. All data are represented as means ± SEM.
3. Results To investigate the recruitment of lymphocyte to the brain follow ing exposure to predator odor we used Balb/c mouse strain, previ ously shown to be less vulnerable to predator odor stress (Cohen et al., 2008). The mice were exposed for 15 min to cat litter (predator odor) and their brains were excised at 3, 24, 48 h as well as 7 days after exposure, and analyzed by immunohistochemistry. We espe cially looked at lymphocyte accumulation in the choroid plexus (Cpx) surrounding the hippocampus, as it forms the main T cells entry point to the CNS (Ransohoff et al., 2003) (Fig. 1a–i). Staining for CD3 (a marker for T cells) revealed that as early as 48 h after the exposure to stress there was a twofold increase in the number of CD3+ cells in the Cpx of the stressed animals (34.2 ± 2.47 average per slice in the stressed mice relative to 18.2 ± 1.35 in unstressed controls) (Fig. 1a), and the numbers of these cells remained high 7 days after the stress, the latest time point that we tested. These results suggest that acute psychological stress induces an increase of brain surveillance by CD3+ cells.
2.4. Corticosterone administration
Table 1 Immunization with pMOG 35¡55 reduces behavioral manifestations induced by acute stress in C57BL/6J mice
Corticosterone (Sigma–Aldrich) was dissolved in polyethylene glycol 400 (PEG) (Sigma–Aldrich), and each mouse received a single s.c. injection of corticosterone (0.6, 6 or 60 mg/kg in 0.01 ml PEG) or PEG alone.
Treatment
2.5. Immunohistochemistry and tissue preparation The animals were deeply anesthetized and perfused transcardially, first with PBS and then with 2.5% paraformaldehyde. Their brains were removed, postfixed overnight, and then equilibrated in phosphate-buffered 30% sucrose. Free-floating 30-lm longitudinal sections were collected on a freezing micro tome (SM2000R; Leica Microsystems) and stored at 4°C prior to immunohisto chemistry. For immunohistochemistry, coronal sections of the brain (30 lm) were treated with a blocking solution containing 20% horse serum (HS), 0.1% Triton X-100 except for sections to be stained for BDNF, for which the blocking solution contained 0.05% saponin (Sigma–Aldrich). Primary antibodies were applied in a humidified cham ber at room temperature. Tissue sections were then labeled overnight with the following primary antibodies (Abs): rabbit anti CD3 (Dako cytomation), hamster
Parameters Time spent in the open arms (min) Number of entries to the open arms Exploratory activity Acoustic startle amplitude Rota rod (sec)
pMOG/CFA
PBS/CFA
Student’s t-test
1.4 ± 0.1
0.8 ± 0.2
3.3 ± 0.3
1.8 ± 1.5
17 ± 1.2
13.9 ± 0.7
344 ± 57.7
571.7 ± 82.1
234 ± 21.8
224 ± 4.5
t17 = 2.23; P = 0.04 t17 = 2.52; P = 0.02 t17 = 2.27; P = 0.03 t17 = 2.27; P = 0.03 n.s.
C57BL/6J mice were immunized with pMOG35¡55 or PBS emulsified with CFA one week before a 15 min exposure to predator odor. pMOG35¡55 immunized mice spent significantly more time exploring the open arms of the elevated plus-maze and showed a reduced acoustic startle response versus PBS-treated mice. Furthermore, there were no motor skill differences between the groups. Values are means ± SEM. n.s., not significant.
1110
G.M. Lewitus et al. / Brain, Behavior, and Immunity 22 (2008) 1108–1114
Fig. 1. Stress enhances immune trafficking in Balb/c mice. (a) Balb/c mice were exposed to predator odor, and were killed at 3, 24, 48 h and 7 days thereafter; sec tions from their brains were analyzed by immunohistochemistry for the presence of CD3+ cells (T cells). Most of the CD3+ cells were found in the Cpx. Bar graph indi cates the average numbers of CD3+ cells per slice. Values represent means ± SEM. (A one-way ANOVA indicated a significant difference between the different time points (F5,24 = 14.55, P = 0.0001); ¤¤¤P < 0.001 (Tukey–Kramer post hoc analysis); n=6 slices from 5 animals). (a-i), a representative image of the Cpx of stressed mice 48 h after stress exposure. CD3+ cells are stained in red and marked by arrows. (b) Analy sis of ICAM-1 expression in the Cpx and hippocampus of stressed Balb/c mice. Graph indicates the density of ICAM-1 in arbitrary units. Values represent means ± SEM. (A one-way ANOVA indicated a significant difference between the different time points (F5,51 = 7.96, P = 0.0001); ¤¤ <0.01, (Tukey–Kramer post hoc analysis); n=5). (bi,ii) Representative images of the Cpx and hippocampus, stained with anti-ICAM-1: from control mice (i), and mice 48 h after exposure to stress (ii). (c) ICAM-1 expres sion in the Cpx after s.c. injection of different concentrations of corticosterone (0.6, 6 and 60 mg/kg). Graph indicates density of ICAM-1 expression relative to ICAM-1 in vehicle-treated control mice. Values are means ± SEM. (A one-way ANOVA indi cated a significant difference between the different time points for each treatment. 0.6 mg/kg (F3,34 = 3.76, P = 0.0194); 6 mg/kg (F3,31=4.01, P = 0.0159) ¤P < 0.05; 60 mg/kg (F3,26 = 3.64, P = 0.0256) ¤P < 0.05 (Tukey–Kramer post hoc analysis); n=5.
To determine which adhesion molecules are involved in T-cell recruitment following stress, we stained brains from stressed mice for VCAM-1 and ICAM-1, the main adhesion molecules that are thought to be involved in CNS immune trafficking (Carrithers et al., 2000; Greenwood et al., 2002). Most of the ICAM-1 expression was observed in the Cpx and on the blood vessels (Fig. 1b-i, ii). There was a significant transient up-regulation of ICAM-1 expression with a twofold increase in expression by 48 h in the Cpx (Fig. 1b).
In contrast, VCAM-1 gave only weak staining, with no observed differences between the various groups; as a positive control for the staining high expression of VCAM-1 was observed in brains in the presence of an inflammatory response (data not shown). These results suggest that stress selectively up regulated ICAM-1 expres sion, and that ICAM-1 might be responsible for the enhanced accu mulation of immune cells in the brain following an acute stress. It was previously shown that corticosterone elevation, in response to acute stress, is one of the hormonal mediators of stressinduced lymphocyte trafficking to peripheral tissues (Dhabhar and McEwen, 1999). Furthermore, cortisol was shown to up-regulate LFA-1 (the ligand for ICAM-1) expression on lymphocytes following acute stress (Tarcic et al., 1995). Therefore, we wished to determine whether administration of exogenous corticosterone could mimic the effect of the stress on ICAM-1 expression in the brain. To this end, we injected mice with corticosterone (0.6, 6 and 60 mg/kg), or with the vehicle, polyethylene glycol (PEG); brains were excised after 3, 24 or 48 h and stained for ICAM-1. The elevation and the timing of ICAM-1 expression in the Cpx were dependent on the cor ticosterone dosage. Three hours after administration of 0.6 mg/kg corticosterone, a slight but not significant elevation of ICAM-1 in the Cpx was seen. When an intermediate dosage of corticosterone (6 mg/kg) was administered, the elevation of ICAM-1 expression at 3 h was statistically significant, and when the highest dosage of corticosterone (60 mg/kg) was administered, the peak of ICAM-1 expression occurred 24 h after the injection (Fig. 1c). These results suggest that the effect of stress on the expression of ICAM-1 might be partly mediated by the elevation of corticosterone. To further understand the functional association between the recruitment of the lymphocytes to the brain and the ability of the mice to adapt to the acute stress, we examined the C57BL/6J mice; this strain has a reduced HPA axis response to stress (Anis man et al., 1998; Shanks et al., 1990), and reduced stress-induced delayed-type hypersensitivity (Flint and Tinkle, 2001). Before look ing at the CD3+ cell recruitment and ICAM-1 expression, we veri fied that the C57BL/6J mice indeed have reduced ability to adapt to the predator odor compared to Balb/c as previously reported (Cohen et al., 2008) (Supplementary Table 1). In the C57BL/6J we found a transient increase of CD3+ cells in the Cpx at 48 h after the exposure to the stressor (63.4 ± 6.03 in the stressed mice relative to 39.73 ± 2.41 in unstressed controls) (Fig. 2a). The expression of ICAM-1 in the Cpx was not significantly affected by the stress (Fig. 2b). It is important to note that the basal level of ICAM-1 was sim ilar between the two strains (data not shown) although the basal levels of CD3+ cells in the Cpx of C57BL/6J animals were higher than in Balb/c. These results further supported a strong association between ICAM-I expression in the Cpx and recruitment of CD3+ cells in response to acute stress. As T cells were shown to affect BDNF levels (Ziv et al., 2006b), we hypothesized that T-cell trafficking to the CNS would enable res toration of BDNF levels and an increased ability to cope with stress ful conditions. To correlate BDNF levels, we examined whether there were strain differences in the effect of stress on BDNF expres sion in the dentate gyrus (DG) of the hippocampus (Fig. 3a). Balb/c and C57BL/6J mice were stained for BDNF at various time points fol lowing exposure to stress. In the Balb/c mice there was a transient reduction of BDNF observed as early as 3 h after stress; however, by day 7, BDNF levels returned to normal (Fig. 3b). In C57BL/6J mice, the reduction of BDNF was seen 3 h after the stress, yet in contrast to Balb/c, BDNF levels remained low even 7 days after stress appli cation (Fig. 3c). Immunization with a CNS-specific antigen was shown to rescue neurons from secondary damage by recruiting autoreactive T cells to the site of injury (Hauben et al., 2000). Therefore, we made an assumption that immunization with such an antigen might reduce the maladaptation to stress in C57BL/6J mice. We immunized the
G.M. Lewitus et al. / Brain, Behavior, and Immunity 22 (2008) 1108–1114
1111
Fig. 2. Stress transiently enhances immune trafficking in C57BL/6J mice. C57BL/6J mice, which have a reduced HPA response to stress, showed a transient infiltration of CD3+ cells to the Cpx after exposure to predator odor. (a) Graph indicates the average numbers of CD3+ cells per slice. Values are means ± SEM. (A one-way ANOVA indicated a sig nificant difference between the different time points (F5,24 = 7.06, P = 0.0003); ¤¤P < 0.01 (Tukey–Kramer post hoc analysis); n=6 slices from 5 animals. (b) ICAM-1 expression in the Cpx of C57BL/6J mice at different time points after stress. Graph indicates ICAM-1 density in arbitrary units. Values are means ± SEM; n=5.
Fig. 3. Expression of BDNF in the hippocampus is associated with adaptation to stress. After exposure to the predator odor, sections of the hippocampus of Balb/c mice and C57BL/6J mice were stained for Brain-derived neurotrophic factor (BDNF). (a) A representative image of BDNF staining in the dentate gyrus (DG) of naïve mice (Balb/c and C57BL/6J) and 7 days after exposure to the predator odor. (b) Analysis of BDNF immunoreactivity in the DG at different time points after stress. Graph indi cates BDNF density in arbitrary units. Values are means ± SEM. (A one-way ANOVA indicated a significant difference between the different time points. For Balb/c: (F5,53 = 9.93, P = 0.0001); ¤P < 0.05, ¤¤¤P < 0.001; for C57BL/6J: (F5,48 = 4.06, P = 0.0037); ¤ P < 0.05, ¤¤P < 0.01, (Tukey-kramer post hoc analysis); n=5.
mice with a MOG-derived peptide, pMOG35–55, emulsified in CFA, 1 week before exposing the mice to predator odor. Though the immu nization increases the potency of the autoimmune cells, it does not lead to autoimmune encephalomyelitis (Lewitus et al., 2006). Mice were tested in the elevated plus maze and for the acoustic startle response 1 week after stress exposure. A significant difference was
Fig. 4. Immunization with pMOG35¡55 enhances BDNF in the hippocampus.C57BL/6J mice were immunized with pMOG35¡55 or PBS emulsified with CFA 1 week before a 15-min exposure to predator odor. (a) Representative image of Brain-derived neurotrophic factor (BDNF) staining in the dentate gyrus (DG) of C57BL/6J mice immunized with pMOG35–55 or PBS emulsified in CFA in naive mice or 7 days after exposure to predator odor. (b) Quantification of BDNF immunoreactivity in the DG of pMOG35–55 or PBS immunized C57BL/6J mice at 24 h and 7 days after exposure to predator odor. Note that the treated mice exhibited a reduction in the BDNF levels 24 h after the stress. However, 7 days after the stress exposure, the levels of BDNF in the mice treated with CFA alone were still low compared to the pMOG35–55 immu nized mice. Values represent means ± SEM. (One-way ANOVA analysis indicated a significant difference between the different time points. (F5,53=20.299, P = 0.0001); ¤¤ P < 0.01, ¤¤¤P < 0.001, n=5).
1112
G.M. Lewitus et al. / Brain, Behavior, and Immunity 22 (2008) 1108–1114
observed between the pMOG35–55 immunized group and the con trol-injected mice. The immunized mice showed lower levels of anxiety, as measured by a weaker acoustic startle response as well as by the larger time spent in the open arms of the elevated plusmaze, and higher exploratory activity (Table 1). To ensure that the reduced exploratory activity of the control mice was not due to reduce motor activity, 5 mice from each group underwent Rotarod test. Both group of mice spent equal time on the accelerating Rota-rod, further suggesting that the observed effect of the immu nization was an outcome of reduced fearfulness (Table 1). These results suggest that manipulation leading to enhanced immunesurveillance of the brain can reduce maladaptation to stress. The fact that we observed association between the BDNF expres sion and recruitment of immune cells to the brain prompted us to examine whether the improved behavior induced by the immu nization with pMOG35–55 also resulted in restoration of BDNF lev els. We therefore repeated the above experiment of immunizing C57BL/6J mice with pMOG35–55 emulsified in CFA, or with CFA alone, and analyzed BDNF levels. The animals were exposed to predator odor 1 week after immunization. The brains were tested for BDNF expression 24 h and 7 days after exposure to the stress. Unstressed animals, immunized with pMOG35–55or treated with CFA, were also analyzed 14 days after immunization. As expected, stress caused a reduction in BDNF levels, which was evident both at 24 h and 7 days. Yet, in the immunized animals, levels of BDNF were significantly restored 7 days following stress (Fig. 4a and b). No significant differences in BDNF levels were observed in the unstressed mice between the pMOG35–55 immunized and the PBStreated mice. 4. Discussion Our results suggest that trafficking of immune cells to the brain is part of the defense mechanism against consequences of psycho logical stress. We also showed that immunization with CNS derived peptide (MOG35–55) reduced the delayed adverse behavioral effects of stress such as anxiety and the acoustic-startle response, by regu lating levels of BDNF. Recently it was shown that adaptation to the acute psycholog ical stressor of predator odor depends on a controlled adaptive immune response recognizing CNS antigens (Cohen et al., 2006). Here we show the enhanced trafficking of T cells to the brain was associated with an enhanced ability to adapt to stress. For example, the Balb/c mice that demonstrated enhanced T-cell recruitment to the brain following the stress had lower levels of anxiety and reduced acoustic startle response relative to the C57BL/6J strain with limited brain lymphocyte recruitment. Similar strain differ ences between these two strains were previously observed in the acute stress-induced enhancement of cutaneous hypersensitivity. In Balb/c mice, acute stress prior to chemical challenge enhances the ear swelling response to antigen; in C57BL/6J mice, acute stress does not affect ear swelling (Flint and Tinkle, 2001). Dhabhar and McEwen were the first to show that acute stress increases immune surveillance and immune responses in organs that can be affected by stress, such as the skin (Dhabhar and McE wen, 1997). They suggested that the ability of stress to enhance the immune response is aimed at protecting the organism from a possible infection or injury caused by the stressful situation (such as possible injury by a predator). Our results suggest that also traf ficking to the brain, induced by the stress, is part of the protective mechanism recruited to fight off the stress. Immune trafficking to the brain has never been proposed as part of the defense mech anism following stress. The enhanced ability of the Balb/c mice to cope with the stress as well as the selective up regulation of ICAM-1 might hint towards a Th2 subset (Biernacki et al., 2001). However, further experiments are needed in order to better char
acterize the different subsets of T cells and to distinguish between the protective and the pathogenic T-cell phenotype in the context of coping with mental stress. Several adhesion molecules, such as ICAM-1 and VCAM-1, were shown to be involved in lymphocyte trafficking to the CNS (Carri thers et al., 2000; Ransohoff et al., 2003) under inflammatory con ditions. However, we show here that psychological stress, induced up regulation of the expression of ICAM-1, but not VCAM-1, pre sumably as a protective immune response. These results are in line with observations showing increased expression of LFA-1 (the ligand for ICAM-1) on peripheral lymphocytes following psycho logical stress in humans and rodents (Bauer et al., 2001; Goebel and Mills, 2000). The enhanced expression of ICAM-1 was mainly observed in the choroids plexus, surrounding the hippocampus and on endothelial blood vessels within the hippocampus. Similar pattern of ICAM-1 expression was observed after cortical contusion trauma in rats, where elevation of ICAM-1 was observed on the epi thelial cells of the choroids plexus and occasional microvessels of the hippocampus (Isaksson et al., 1997). It is important to note that ICAM-1 was previously shown to be solely expressed on epithelial cells of the choroids plexus and not on the choroids plexus endo thelial cells even after the induction of experimental autoimmune encephalomyelitis (EAE) (Steffen et al., 1996; Wolburg et al., 1999). It was suggested that Cpx epithelial cells might have an immuno logical function, as it was shown to express MHC class I and II mol ecules (Nathanson and Chun, 1989). While the exact nature of its immunological function is not known, the presence of the T cells at intra-ventricular space close to the choroid epithelium further emphasizes the importance of the choroid plexus in brain-immune interaction. Possible mediators for the stress-induced immune trafficking to the CNS are the stress hormones, such as Corticosterone. Here we showed that exogenous administration of corticosterone could mimic the elevation of ICAM-1. The dosages of corticosterone influ ence the timing of ICAM-1 expression. Corticosterone was shown to be involved in other aspects of immune trafficking. Dhabhar and McEwen (1996) showed that acute stress induces mobilization of leukocytes from the blood to the periphery, and that acute adminis tration of corticosterone can mimic the change in the distribution of the peripheral leukocytes as seen following acute stress (Dhab har and McEwen, 1996). This data together with ours, suggest that corticosterone may act as the first mediator of the immune response in the brain and in the periphery. The function of the immune system is to be prepared to fight off the consequences of the stressor (such as injury), and also to play a role in maintaining brain homeostasis. There are several lines of evidence suggesting a role for BDNF in the behavioral and cellular response to stress, as well as its path ophysiology (Duman et al., 2000; Vaidya et al., 1997). Acute stress such as immobilization (restraint stress) transiently down-reg ulates BDNF mRNA and protein expression in the hippocampus, especially in the DG (Adlard and Cotman, 2004; Smith et al., 1995). Taken together, the evidence for the involvement of the immune system in the maintenance of BDNF levels in the DG (Ziv et al., 2006b), with the production of BDNF by immune cells (Barouch and Schwartz, 2002; Kerschensteiner et al., 1999), led us to suggest that one of the roles of the enhanced immune-surveillance induced by stress is to facilitate maintaining BDNF levels. In our model of stress, we observed relationships between BDNF expression in the DG, the behavioral response to stress, and the degree of lympho cyte infiltration. The association was observed in the spontane ous response to the stressor and was further substantiated by the beneficial effect of the immunization. The immunization reduced the maladaptative behavior observed a week after stress expo sure, and although the BDNF levels in the DG were reduced 24 h after the stress, by 7 days, the levels of BDNF were restored to the
G.M. Lewitus et al. / Brain, Behavior, and Immunity 22 (2008) 1108–1114
normal, pre-stress levels. These results further emphasized the importance of the immune system in brain homeostasis. In line with our results, it was shown that rats that were exposed to preda tor odor exhibited a reduction in BDNF mRNA, as well as protein lev els in the hippocampus, similar to our findings here. Additionally, a correlation was demonstrated between the behavioral response and BDNF levels (Kozlovsky et al., 2007). It is important to note that the peptide used in this study, under certain protocols, can induce EAE in mice (Mendel et al., 1995). In the present study the protocol included only one immu nization with the peptide (100 lg) without adding Pertussis toxin, used to induce encephalomyelitis (Mendel et al., 1995). We have previously found that there is a delicate balance between beneficial and pathogenic autoimmunity, which is influenced by the type of antigen, dosage as well as the adjuvant used. We have found that en autoimmune response was beneficial only when the EAE symptoms it induced were mild, however, no beneficial effect was observed when EAE was severe (Hauben et al., 2001; Ziv et al., 2006a). Importantly, we used this peptide in the pres ent study as a proof of principle. For therapeutic purposes, weak agonists of self peptides such as altered peptide ligands, as well as additional peptides and carrier, will be tested (Hauben et al., 2001; Ziv et al., 2006a). This study extends the role of ‘protective autoimmunity’ to include protection against mental stress, and further argues in favor of the importance of a distinction between a wellcontrolled immune response that takes place in the acutely stressed brain and a pathological immune response that occurs when immune response looses control (such as Multiple scle rosis (Gold and Heesen, 2006). According to this view, a tran sient and controlled trafficking of T cells in response to an acute stress is a desirable response, and is amenable to boosting. In line with this contention, conditions of PTSD in humans might be a reflection of insufficient or untimely recruitment of the rel evant immune activity. Recognizing that the systemic immune system is a factor in containing mental stress offers new direc tions for the development of a therapy for stress-induced pathol ogies such as PTSD and depression, in the form of T cell-based immunization, which increases the body’s physiological ability to cope with stress. Disclosure/conflict of interest We have no financial interests to disclose. Acknowledgments We thank R. Halper and S. Schwarzbaum for editing the manu script and to S. Ovadia for animal maintenance. M.S. is the incum bent of the Maurice and Ilse Katz Professorial Chair in Neuroimmu nology. This study was supported, in part, by an NARSAD award for Distinguished Investigators awarded to M.S. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbi.2008.05.002. References Adlard, P.A., Cotman, C.W., 2004. Voluntary exercise protects against stress-induced decreases in brain-derived neurotrophic factor protein expression. Neurosci ence 124, 985–992. Anisman, H., Lacosta, S., Kent, P., McIntyre, D.C., Merali, Z., 1998. Stressor-induced corticotropin-releasing hormone, bombesin, ACTH and corticosterone varia tions in strains of mice differentially responsive to stressors. Stress 2, 209– 220.
1113
Barouch, R., Schwartz, M., 2002. Autoreactive T cells induce neurotrophin produc tion by immune and neural cells in injured rat optic nerve: implications for protective autoimmunity. FASEB J. 16, 1304–1306. Bauer, M.E., Perks, P., Lightman, S.L., Shanks, N., 2001. Are adhesion molecules involved in stress-induced changes in lymphocyte distribution? Life Sci. 69, 1167–1179. Biernacki, K., Prat, A., Blain, M., Antel, J.P., 2001. Regulation of Th1 and Th2 lym phocyte migration by human adult brain endothelial cells. J. Neuropathol. Exp. Neurol. 60, 1127–1136. Carrithers, M.D., Visintin, I., Kang, S.J., Janeway Jr., C.A., 2000. Differential adhesion molecule requirements for immune surveillance and inflammatory recruit ment. Brain 123 (Pt 6), 1092–1101. Cohen, H., Geva, A.B., Matar, M.A., Zohar, J., Kaplan, Z., 2008. Post-traumatic stress behavioural responses in inbred mouse strains: can genetic predisposition explain phenotypic vulnerability? Int. J. Neuropsychopharmacol. 11, 331–349. Cohen, H., Ziv, Y., Cardon, M., Kaplan, Z., Matar, M.A., Gidron, Y., Schwartz, M., Kip nis, J., 2006. Maladaptation to mental stress mitigated by the adaptive immune system via depletion of naturally occurring regulatory CD4+CD25+ cells. J. Neu robiol. 66, 552–563. Dhabhar, F.S., McEwen, B.S., 1996. Stress-induced enhancement of antigen-specific cell-mediated immunity. J. Immunol. 156, 2608–2615. Dhabhar, F.S., McEwen, B.S., 1997. Acute stress enhances while chronic stress sup presses cell-mediated immunity in vivo: a potential role for leukocyte traffick ing. Brain. Behav. Immun. 11, 286–306. Dhabhar, F.S., McEwen, B.S., 1999. Enhancing versus suppressive effects of stress hor mones on skin immune function. Proc. Natl. Acad. Sci. USA 96, 1059–1064. Duman, R.S., Malberg, J., Nakagawa, S., D’Sa, C., 2000. Neuronal plasticity and sur vival in mood disorders. Biol. Psychiatry 48, 732–739. File, S.E., 1993. The interplay of learning and anxiety in the elevated plus-maze. Behav. Brain Res. 58, 199–202. Flint, M.S., Tinkle, S.S., 2001. C57BL/6 mice are resistant to acute restraint modula tion of cutaneous hypersensitivity. Toxicol. Sci. 62, 250–256. Goebel, M.U., Mills, P.J., 2000. Acute psychological stress and exercise and changes in peripheral leukocyte adhesion molecule expression and density. Psychosom. Med. 62, 664–670. Gold, S.M., Heesen, C., 2006. Stress and disease progression in multiple sclerosis and its animal models. Neuroimmunomodulation 13, 318–326. Greenwood, J., Etienne-Manneville, S., Adamson, P., Couraud, P.O., 2002. Lympho cyte migration into the central nervous system: implication of ICAM-1 signal ling at the blood–brain barrier. Vascul. Pharmacol. 38, 315–322. Griebel, G., Blanchard, D.C., Jung, A., Blanchard, R.J., 1995. A model of’antipredator’ defense in Swiss-Webster mice: effects of benzodiazepine receptor ligands with different intrinsic activities. Behav. Pharmacol. 6, 732–745. Hauben, E., Agranov, E., Gothilf, A., Nevo, U., Cohen, A., Smirnov, I., Steinman, L., Sch wartz, M., 2001. Posttraumatic therapeutic vaccination with modified myelin self-antigen prevents complete paralysis while avoiding autoimmune disease. J. Clin. Invest. 108, 591–599. Hauben, E., Butovsky, O., Nevo, U., Yoles, E., Moalem, G., Agranov, E., Mor, F., Leibo witz-Amit, R., Pevsner, E., Akselrod, S., Neeman, M., Cohen, I.R., Schwartz, M., 2000. Passive or active immunization with myelin basic protein promotes recov ery from spinal cord contusion. J. Neurosci. 20, 6421–6430. Isaksson, J., Lewen, A., Hillered, L., Olsson, Y., 1997. Up-regulation of intercellular adhesion molecule 1 in cerebral microvessels after cortical contusion trauma in a rat model. Acta Neuropathol. 94, 16–20. Kerschensteiner, M., Gallmeier, E., Behrens, L., Leal, V.V., Misgeld, T., Klinkert, W.E., Kolbeck, R., Hoppe, E., Oropeza-Wekerle, R.L., Bartke, I., Stadelmann, C., Lass mann, H., Wekerle, H., Hohlfeld, R., 1999. Activated human T cells, B cells, and monocytes produce brain-derived neurotrophic factor in vitro and in inflamma tory brain lesions: a neuroprotective role of inflammation? J. Exp. Med. 189, 865–870. Kipnis, J., Cohen, H., Cardon, M., Ziv, Y., Schwartz, M., 2004. T cell deficiency leads to cognitive dysfunction: implications for therapeutic vaccination for schizophre nia and other psychiatric conditions. Proc. Natl. Acad. Sci. USA 101, 8180–8185. Kipnis, J., Yoles, E., Schori, H., Hauben, E., Shaked, I., Schwartz, M., 2001. Neuronal survival after CNS insult is determined by a genetically encoded autoimmune response. J. Neurosci. 21, 4564–4571. Kozlovsky, N., Matar, M.A., Kaplan, Z., Kotler, M., Zohar, J., Cohen, H., 2007. Longterm down-regulation of BDNF mRNA in rat hippocampal CA1 subregion corre lates with PTSD-like behavioural stress response. Int. J. Neuropsychopharmacol. 10, 741–758. Lewitus, G.M., Kipnis, J., Avidan, H., Ben-Nun, A., Schwartz, M., 2006. Neuroprotec tion induced by mucosal tolerance is epitope-dependent: conflicting effects in different strains. J. Neuroimmunol. 175, 31–38. Mendel, I., Kerlero de Rosbo, N., Ben-Nun, A., 1995. A myelin oligodendrocyte glyco protein peptide induces typical chronic experimental autoimmune encephalo myelitis in H-2b mice: fine specificity and T cell receptor V beta expression of encephalitogenic T cells. Eur. J. Immunol. 25, 1951–1959. Nathanson, J.A., Chun, L.L., 1989. Immunological function of the blood-cerebrospi nal fluid barrier. Proc. Natl. Acad. Sci. USA 86, 1684–1688. Ransohoff, R.M., Kivisakk, P., Kidd, G., 2003. Three or more routes for leuko cyte migration into the central nervous system. Nat. Rev. Immunol. 3, 569–581. Schori, H., Kipnis, J., Yoles, E., WoldeMussie, E., Ruiz, G., Wheeler, L.A., Schwartz, M., 2001. Vaccination for protection of retinal ganglion cells against death from glu tamate cytotoxicity and ocular hypertension: implications for glaucoma. Proc. Natl. Acad. Sci. USA 98, 3398–3403.
1114
G.M. Lewitus et al. / Brain, Behavior, and Immunity 22 (2008) 1108–1114
Schori, H., Shechter, R., Shachar, I., Schwartz, M., 2007. Genetic manipulation of CD74 in mouse strains of different backgrounds can result in opposite responses to central nervous system injury. J. Immunol. 178, 163–171. Shanks, N., Griffiths, J., Zalcman, S., Zacharko, R.M., Anisman, H., 1990. Mouse strain differences in plasma corticosterone following uncontrollable footshock. Phar macol. Biochem. Behav. 36, 515–519. Shanks, N., Kusnecov, A.W., 1998. Differential immune reactivity to stress in BALB/ cByJ and C57BL/6J mice: in vivo dependence on macrophages. Physiol. Behav. 65, 95–103. Smith, M.A., Makino, S., Kvetnansky, R., Post, R.M., 1995. Stress and glucocorticoids affect the expression of brain-derived neurotrophic factor and neurotrophin-3 mRNAs in the hippocampus. J. Neurosci. 15, 1768–1777. Steffen, B.J., Breier, G., Butcher, E.C., Schulz, M., Engelhardt, B., 1996. ICAM-1, VCAM-1, and MAdCAM-1 are expressed on choroid plexus epithelium but not endothelium and mediate binding of lymphocytes in vitro. Am. J. Pathol. 148, 1819–1838. Swerdlow, N.R., Geyer, M.A., 1998. Using an animal model of deficient sensorimo tor gating to study the pathophysiology and new treatments of schizophrenia. Schizophr. Bull. 24, 285–301.
Tarcic, N., Levitan, G., Ben-Yosef, D., Prous, D., Ovadia, H., Weiss, D.W., 1995. Restraint stress-induced changes in lymphocyte subsets and the expression of adhesion molecules. Neuroimmunomodulation 2, 249–257. Vaidya, V.A., Marek, G.J., Aghajanian, G.K., Duman, R.S., 1997. 5-HT2A receptor-medi ated regulation of brain-derived neurotrophic factor mRNA in the hippocampus and the neocortex. J. Neurosci. 17, 2785–2795. Wolburg, K., Gerhardt, H., Schulz, M., Wolburg, H., Engelhardt, B., 1999. Ultrastruc tural localization of adhesion molecules in the healthy and inflamed choroid plexus of the mouse. Cell Tissue Res. 296, 259–269. Yehuda, R., McFarlane, A.C., 1995. Conflict between current knowledge about post traumatic stress disorder and its original conceptual basis. Am. J. Psychiatry 152, 1705–1713. Ziv, Y., Avidan, H., Pluchino, S., Martino, G., Schwartz, M., 2006a. Synergy between immune cells and adult neural stem/progenitor cells promotes functional recov ery from spinal cord injury. Proc. Natl. Acad. Sci. USA 103, 13174–13179. Ziv, Y., Ron, N., Butovsky, O., Landa, G., Sudai, E., Greenberg, N., Cohen, H., Kip nis, J., Schwartz, M., 2006b. Immune cells contribute to the maintenance of neurogenesis and spatial learning abilities in adulthood. Nat. Neurosci. 9, 268–275.