| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1-Adrenergic Receptors on Immune Cells Impair Innate Defenses against ListeriaLaboratory of Clinical and Experimental Endocrinology and Immunology, Wadsworth Center, New York State Department of Health, Albany, NY 12201
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-adrenoceptors (ARs). We now show that impaired host resistance by CR cannot be accounted for by a decrease in LM-specific (listeriolysin O91–99 tetramer+) effector CD8+ T cells; this result is consistent with previous observations that CR-induced effects are mainly limited to early anti-LM responses. 2-Adrenoceptor (2AR)–/– FVB/NJ and wild-type FVB/NJ mice had equivalent anti-LM defenses, whereas 1-adrenoceptor (1AR)–/– FVB/NJ mice had lower levels of LM even when subjected to CR treatment. Additionally, host-resistance competency of 1AR–/– mice could be transferred to irradiated wild-type mice reconstituted with 1AR–/– bone marrow progenitors and spleen cells, indicating that 1AR signaling on immune cells reduces anti-LM responses. 1AR–/– mice had improved cellular (delayed-type hypersensitivity) responses while 2AR–/– mice had improved humoral responses (IgG1, IgG2, and IgM), a result that further explains the strain differences in LM defenses. CR-induced expression of 1AR and 2AR mRNA was assessed by real-time PCR. CR treatment significantly increased AR mRNAs in Ficoll-purified and F4/80+-enhanced liver but not splenic homogenates, demonstrating an organ-specific effect of stress that alters host defenses. Finally, CR treatment induced early increases in perforin expression that may enhance immune cell apoptosis and interfere with LM clearance. In conclusion, 1AR signaling has immunomodulatory effects on early cell-mediated immune responses; a lack of 1AR signaling improves antilisterial defenses and cell-mediated immunity, in general. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
The experimental model of 1-h cold (4°C) restraint (CR) is known to elicit both physical and psychological stress (9, 10). We have used CR treatment followed by a low-dose bacterial infection in mice in an attempt to simulate a familiar human condition, namely the experience of psychological or physical stressors and exposure to common infectious agents. Psychological and physiological stresses elicit functional changes in many cell types, by modifying the supply of oxygen and metabolites required for a successful "fight or flight" response (11). In addition to controlling cardiovascular functions, energy metabolism, and thermoregulation, stress factors are known to influence immune cell functions. Numerous neuroendocrine factors, such as prostaglandins, glucocorticoids, catecholamines, and neuropeptides, have regulatory influences on host-pathogen interactions (12). It has been proposed that catecholamines provide a physiologic mechanism to prevent an overactive cell-mediated immune response, by shifting the activity of APCs and Th1 cells from a Th1-promoting to a Th2-promoting response via 2-adrenoceptor (2AR) (13). However, the kinetics of stress-factor interactions with immune cells also may be critical in determining whether stress-mediated neuroimmune interactions have beneficial or detrimental consequences. We hypothesize that if immune cells encounter stress before they begin to respond to a pathogen, the immunosuppressive effects of stress-induced neuroendocrine factors can weaken host defenses and increase the pathogenic burden. It is this acute, stress-induced immunosuppression before immune activation that is the focus of our current studies.
Cell surface dopaminergic and adrenergic receptors, which are expressed on immune cells (12), bind the catecholamines dopamine (DA), epinephrine (Epi), and norepinephrine (NE). Pharmacologic blocking studies provided the first evidence for an immunomodulatory role of 1-adrenoceptor (1ARs) in host immunity (7). Specifically, peripheral administration of atenolol (a 1AR-specific antagonist commercially available as Tenormin) blocked the CR-induced delay in bacterial clearance whereas 
AR- or 2AR-specific antagonists (phentolamine and ICI118,551, respectively) did not. It is well documented that 2AR is ubiquitously expressed on Th0, Th1, and B cells; 2AR is down-regulated on Th2 cells by histone deacetylation (14), thereby implicating this receptor in the regulation of humoral and Th1 functions by catecholamines (15). Depending upon the activation state and developmental stage of the Th cell, the 2AR signal can enhance (16, 17) or suppress (18) IFN-
production. Despite a well-established role of sympathetic neuroimmune modulation and despite also the widespread clinical use of " blockers" and synthetic catecholamines, the immunologic relevance of 1ARs has been largely unexplored. The current in vivo studies used wild-type (WT) mice (either BALB/c or FVB/NJ) and FVB/NJ mutant strains with a deficiency of 1AR or 2AR (1AR–/– and 2AR–/–, respectively) to investigate the role of stress-induced -adrenoceptors (ARs) on host immunity, and to evaluate overall in vivo immunity to KLH. Based on previous studies, our a priori hypothesis was that CR stress-induced 1AR signaling impairs host defenses against a low-dose LM infection by altering innate immune mechanisms.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
All in vivo studies were performed using mice housed at the Wadsworth Center Animal Production Unit in accordance with the Institutional Animal Care and Use Committee Guidelines. The listeriolysin O91–99 (LLO91–99) (I-Ad) tetramer study used BALB/c mice from Taconic Farms. The 1AR–/– and 2AR–/– mice (provided by Dr. B. Kobilka, Stanford University, Palo Alto, CA) were bred on the FVB/NJ background. As previously described, both AR-deficient strains were derived by the insertion of a neomycin resistance cassette into the genetic sequence of the fourth transmembrane domain of the receptor, rendering the transcription of the complete receptor impossible (19). Strain verification studies were performed using real-time PCR and RT-PCR to confirm the lack of 1AR and 2AR transcript in both liver and spleen of 1AR–/– and 2AR–/–mice, respectively (data not shown). Control mice were left in their original cages undisturbed, while mice subjected to CR treatment were individually restrained in well-ventilated plastic 60-ml syringes at 4°C for 1 h in the dark. CR represents a physical and a psychological stress. CR was performed between 8 and 11 a.m. on day 0; the mice were infected immediately after CR. LM was originally isolated from a meningitis patient and has been maintained as previously described (20). Mice were i.v. injected with a sublethal dose of LM (2–3 x 103 CFU/i.v. injection for FVB/NJ, and 3.5–10 x 103 CFU/i.v. for BALB/c mice) with or without CR administered before the inoculation. Except for anti-KLH Ig and DTH responses, all other measurements were obtained from different groups of mice depending upon the background strain or timing that was required for each assay.
Determination of viable LM burden in liver and spleen
To determine the bacterial load in mice following LM infection, we performed enumeration of viable LM as described previously (5). Briefly, mice were sacrificed by lethal CO2 anesthesia, and the spleen and liver were removed aseptically and homogenized in sterile 0.9% NaCl. Serial dilutions of organ homogenates were plated on blood-agar plates and cultured overnight for enumeration of viable LM. Bacterial burdens are expressed as number of viable LM CFU per organ.
Flow cytometric analysis of T cells
Spleens were removed aseptically from 6- to 8-wk-old BALB/c male mice as described above, and single-cell suspensions were prepared by grinding the tissue between two frosted microscope slides. RBCs were lysed and cells were washed twice with PBS, and total cell numbers were determined; 1 x 106 cells/tube were used for staining. MHC class I (H2-Kd) tetramers containing an immunodominant LM epitope from the pore-forming toxin listeriolysin (LLO91–99) were used to quantify LM-specific responses; the tetramers were provided by Dr. E. G. Pamer (Sloan-Kettering Institute, New York, NY). All mAbs against cell surface molecules were purchased from BD Biosciences Pharmingen. Cells were treated with FcR-blocking buffer containing 20 µg/ml streptavidin and anti-mouse CD16/CD32 FcR (1 µg/tube) in 50 µl of staining buffer (0.5% BSA and 0.02% NaN3 in PBS (pH 7.5)) for 20 min on ice. Further staining with PE-MHC class I tetramer (LLO91–99), allophycocyanin-anti-mouse-CD62L, and FITC-anti-mouse-CD8 was performed for 60 min on ice. Enumeration of LM-specific CD8+ populations was accomplished using TruCount tubes (BD Biosciences Pharmingen). Regulatory T cell subsets following LM infection with or without CR-treatment were analyzed using a combination of FITC-CD4/PE-GITR/allophycocyanin-CD25 Abs. The acquisition and analysis of all specimens were performed on a BD FACSCalibur using CellQuest software at the Wadsworth Center Flow Cytometry Core.
Perforin expression by Western blot analysis
Protein was isolated from frozen livers and spleens of 2- to 3-mo-old male BALB/c mice and 1AR–/– male and female mice were treated with or without CR and infected with 3.6 x 103 LM by homogenization of a portion of each organ in 0.5 ml of mammalian protein extraction reagent (Pierce) with protease inhibitor mixture (Sigma-Aldrich). Lysed homogenates were centrifuged for 30 min at 15,000 x g. Protein concentrations were measured using the BCA protein assay kit (Pierce), and 100 µg of protein from each sample was diluted with one-half volume of sample buffer containing 30% (v/v) glycerol, 10% (v/v) 2-ME, and 0.25% (w/v) bromophenol blue in 62.5 mM Tris-HCl buffer (pH 6.8). SDS-PAGE was performed in 4–20% gradient separating gels; the gels were then placed upon nitrocellulose paper and subjected to blot transfer. The blotted proteins were blocked with 5% fish gelatin in PBS 0.01% NaN3 for 1 h at room temperature (RT) and washed in TBS-T (25 mM Tris-HCl, 125 mM NaCl, and 1.0% Tween 20 (pH 8.0)). Blots were incubated with mAb to mouse -actin (Sigma-Aldrich) and either rabbit anti-rat perforin (catalog no. CPP100; Cell Sciences) or rabbit anti-mouse granzyme B (catalog no. RB-9015-PO; NeoMarkers) in blocking buffer overnight at 4°C. Blots were washed twice in TBS-T and then incubated with HRP-conjugated secondary Abs (goat anti-rabbit IgG for perforin and granzyme and goat anti-mouse IgG, for actin; Sigma-Aldrich) for 2 h at RT with rocking. Blots were washed three times for 20 min each in TBS-T and then developed with Super Signal/chemiluminescent substrate (Pierce) for 5 min and assayed with a LAS-1000plus (Fuji).
Splenocyte and liver preparation for RNA isolation
Male BALB/c mice were either CR-treated or left undisturbed in their home cage. Immediately after CO2 administration, mice were perfused through the right ventricle with 30 ml of PBS, followed by 10 ml of digestion buffer (0.05 M TES (C6H15NO6S) and 0.36 M CaCl (pH 7.5)). Livers and spleens were placed into digestion buffer containing collagenase IV (200 µg/ml; Sigma-Aldrich) and DNase I (50 µg/ml; Sigma-Aldrich) and were coarsely chopped with a sterile razor blade before 37°C incubation for 20 min. Splenocytes were then homogenized as described above, whereas livers were pipetted through a Teflon mesh. Single-cell suspensions were washed twice in PBS and then layered over mouse Ficoll (Ficoll:metrizoate 12:5, density 1.090 g/ml) (21) and centrifuged for 15 min at RT at 1300 x g. The total Ficoll-purified cells were counted and pelleted for immediate isolation of total RNA.
F4/80+ Kupffer cells were isolated from liver homogenates following the above procedure with slight modifications. After single-cell suspensions had been obtained by passage through a Teflon mesh, liver homogenates were centrifuged (110 x g for 2 min at 4°C) to remove hepatocytes (22). F4/80+ cells were enriched from the single-cell suspension by use of a SpinSep mouse enrichment mixture, SpinSep mouse dense particles, and density gradient centrifugation cell separation procedures per the manufacturer’s protocol (StemCell Technologies).
RNA isolation from immune cells and AR mRNA determination by TaqMan real-time PCR
Cell pellets containing 5 x 105–1 x 106 Ficoll-purified liver or spleen cells or F4/80-enriched cells from liver homogenates were resuspended in 50 µl of PBS and lysed with RLT/2-ME buffer from Qiagen. Cell lysates were homogenized with the Shredder Spin Columns provide by Qiagen, and RNA was isolated using the Qiagen RNA capture minicolumns, according to the manufacturer’s protocol for isolation of RNA from cultured cells. Isolated RNA was resuspended in 60 µl of RNase-free water and quantified using a Beckman DU 640 spectrophotometer. After dilution of each RNA preparation to a working concentration of 0.1 µg/µl, cDNA was prepared from 1 µg of total RNA using a High-Capacity cDNA Archive kit from Applied Biosystems. Following the synthesis reaction, an aliquot of 100 µl of PCR-grade water was added to each tube, and 5 µl was used for amplification of murine 1AR, 2AR, and GAPDH using TaqMan Gene Expression Assay kits from Applied Biosystems. Amplifications were conducted in quadruplicate in a 7500 Real Time PCR instrument (Applied Biosystems). Amplification conditions used were specified by the manufacturer as follows: 2 min at 50°C, then 10 min at 95°C, then 40 cycles of 15 s at 95°C, and 1 min at 60°C. Data for the amplification plot was collected during the 60°C step. Relative quantitation results were measured using the comparative cycle threshold method, whereby the amplification of the gene of interest is normalized to amplification of the gene encoding GAPDH, measured from the same cDNA synthesis sample. Four APC lines, RAW 264.7, PMJ2-PC, N9, and A20, were also used for RNA isolation and subsequent AR amplification procedures.
Corticosterone (CORT) measurement
The concentrations of serum CORT at baseline and immediately following 1-hr CR treatment were determined by enzyme immunoassay with CORT EIA Ab (Assay Designs). The sensitivity of the CORT EIA was 32 pg/ml.
Hemopoietic ablation and lymphocyte reconstitution of recipient mice
FVB/NJ WT- and AR-deficient mice that were used as recipient hosts for lymphocyte reconstitution experiments received lethal 137Cs irradiation of 10 Gy (dose rate 2.5 Gy/min). On the following day, a mixture of 1 x 106 bone marrow progenitor cells and 10 x 106 splenocytes was injected i.v. in 200 µl through the tail vein as described previously (23). Transplant recipients were housed in pathogen-free facilities and given neomycin (1 mg/ml) in drinking water for the first 2 wk. After 6–8 wk total, CR administration, LM infection, and subsequent immunologic analysis of host immune responses were performed as described above.
Splenocyte and bone marrow isolation
Spleens were aseptically harvested from euthanized FVB/NJ WT- and AR-deficient donor mice into a sterile petri dish containing Dulbecco’s PBS (without calcium and magnesium; Sigma-Aldrich) on ice. Inside a biosafety hood, spleens were transferred into another sterile petri dish with 5 ml of Dulbecco’s PBS. Each spleen was homogenized between two frosted microscope slides (Erie Scientific), and the cell suspension was transferred into a 15-ml polypropylene tube with a Pasteur capillary pipette. The cell suspension was allowed to settle for 3 min at RT for separation of cellular debris, and the single-cell suspension in the supernatant was then transferred to a new tube. Following centrifugation at 200 x g for 10 min, RBCs were eliminated by use of lysing buffer (0.15 M NH4Cl, 10 mM KHCO3, and 0.1 mM Na2 EDTA (pH 7.2–7.4)). After lysis (5 min at RT), the cell suspension was centrifuged again, and the cell pellet was resuspended in 10 ml of DPBS; 20 µl of this cell suspension was taken for determination of total cell numbers (Coulter Counter). After centrifugation, the cell pellet was resuspended to the appropriate concentration in sterile PBS. Bone marrow progenitors were isolated as described previously (24).
KLH immunization
Female WT (FVB/NJ), 1AR–/–, and 2AR–/– mice (age 2 mo) were immunized with 100 µg of KLH (Calbiochem) plus TiterMax Gold adjuvant in 200 µl of saline at days 1 and 28.
Serum preparation
Peripheral blood was obtained by retro-orbital phlebotomy, into 1.7-ml Eppendorf tubes. After it had been allowed to clot overnight at 4°C, serum was collected following centrifugation.
ELISA for IgG isotype and IgM
IgG isotype and IgM were measured by a standard ELISA, as described previously (25).
DTH assay
DTH assays were performed 21 days after the last immunization with KLH. After measurement of the thickness of the left and right footpad of each mouse, KLH (100 µg/25 µl saline) was s.c. injected into the right footpad and saline only was injected into the left footpad. The DTH response was measured 24 h later with a Spi dial thickness gauge. The response was defined as the difference between the right and left footpad swellings.
Statistical analysis
Data from two or more independent experiments were combined and differences between two experimental groups (where n 
3) were determined using the t test SigmaStat (Jandel Scientific). Effects of CR stress and 1AR or 2AR deficiency on host immunity were considered significant only if the p value was <0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
The in vivo expansion of CD8+ Ag-specific lymphocytes of control or CR-treated BALB/c mice was measured using the immunodominant LLO91–99 tetramer. Tetramer-positive cells were detected by day 8 after LM-infection (Fig. 1A), but no differences were detected between CR-treated and control mice (Fig. 1B). Stress-related changes in the percentage of either CD4+/CD25+ or CD4+/CD25+/GITR+ splenic subsets were not observed at 24 h after LM infection (data not shown).
|
1AR are more resistant to CR-induced immunosuppression than are WT or 2AR–/– mice
The immunosuppressive effect of CR previously reported for BALB/c mice (5, 7, 8, 26) was also observed in WT FVB/NJ mice by day 3 of a low-dose LM infection. However, because FVB/N mice were more sensitive than BALB/c mice to LM, we had to use a lower LM inoculum (
3 x 103 CFU). Liver colonization was significantly increased in CR-treated WT mice compared with control WT mice (Fig. 2A; p = 0.037). CR did not suppress host resistance of 1AR–/– mice. Relative to WT or 2AR–/– mice, 1AR–/– mice had improved host defenses against LM; CR did not impair their resistance in livers (Fig. 2A) or spleens (Fig. 2B). Splenic and overall body weights did not differ statistically between the LM-infected mice treated with CR and those not subjected to CR (data not shown).
|
AR mRNA in immunocytes of liver but not spleen
Because 1AR and 2AR appear to have substantially different effects on host defenses against LM, it was critical that we assess the expression of these receptors by immune cells. We prepared immune cells from both liver and spleen homogenates of BALB/c mice and analyzed them for mRNA expression of 1AR and 2AR transcripts (Fig. 3). Immediately following CR treatment and in the absence of LM infection, significant increases in 1AR and 2AR expression were measured in livers (Fig. 3A; p = 0.04 and p = 0.004, respectively). We further investigated the expression of 1AR and 2AR on the Kupffer (F4/80+) cells of livers obtained from stressed and control mice (Fig. 3B). Although both receptor subtypes were identified by real-time TaqMan PCR analysis, stress treatment did not significantly alter the expression levels. Splenocytes expressed greater levels of 2AR than 1AR (Fig. 3C; p < 0.005); stress treatment did not significantly alter the expression of either receptor. Although radioisotope-binding assays have provided evidence for ARs on monocyte-derived cells, we used real time PCR to confirm expression of 1AR and 2AR in two cell lines of monocyte origin (RAW 264.7 and PMJ2-PC), a microglial cell line (N9), and a B cell line (A20) (data not shown).
|
Although CORT is known to play a significant role in the stress response, both WT- and AR-deficient strains appear to have equally robust increases in CORT following CR treatment (Table I). Furthermore, our previous research using pharmacological blocking agents had suggested a protective role for CORT in this model of stress and infection rather than an immunosuppressive one (5).
|
The expression of the cytotoxic protein perforin was analyzed in livers and spleens of BALB/c mice (Fig. 4) and 1AR–/– mice (Table II). Early changes in perforin expression were evident in livers of CR-treated BALB/c mice (Fig. 4A). The greatest increases in perforin were observed immediately after stress treatment, and 4 h after LM infection (p < 0.05). No changes in perforin expression were observed in the spleen (Fig. 4B), nor were differences in granzyme B expression detected in either liver or spleen (data not shown). No significant increases were observed in liver homogenates of 1AR–/– mice immediately following CR treatment or 24 h after LM infection (Table II).
|
|
1AR+ immune cells
AR-competent immune cells were reconstituted in AR-deficient, irradiated mice, and AR-deficient immune cells were reconstituted in AR-competent, irradiated mice to aid us in assessing the effects of AR nonimmune and immune cells on immune defenses after CR. WT (FVB/NJ) recipient mice reconstituted with 1AR–/– immune cells had a statistically significant decrease (by 3-fold) in bacterial colonization in the liver, relative to mice reconstituted with WT immune cells (Fig. 5D; p = 0.03). The transfer of 2AR–/– cells into WT mice produced a nonsignificant decrease in liver colonization. The transfer of WT immune cells into 1AR–/– hosts or 2AR–/– immune cells into 2AR–/– hosts gave rise to a greater bacterial burden in liver than did the transfer of 1AR–/– immune cells into 1AR–/– hosts (Fig. 5, E and F; p = 0.05). 1AR–/– mice reconstituted with their own immune cells had significantly lower bacterial colonization of the liver than did control mice reconstituted with their own immune cells (Fig. 5, D and E; p = 0.04). 2AR–/– hosts reconstituted with either WT or 1AR–/– immune cells had lower bacterial loads in their livers and spleens on day 3 of the infection than did the same 2AR–/– hosts reconstituted with 2AR–/– donor cells, but these effects were not statistically significant.
|
1AR enhances cell-mediated immunity, whereas 2AR deficiency improves humoral responses
The production of Abs specific for KLH was measured in 1AR–/–, 2AR–/–, and WT (FVB/NJ) mice (Fig. 6, A–C). IgG subtypes and IgM were highest in serum from 2AR–/– mice and lowest in 1AR–/– mice; the single exception was IgG1 for which WT mice had the lowest production. The differences in production of IgG1 between 2AR–/– and WT mice were statistically significant (Fig. 6A). In contrast to humoral immunity, the DTH response in 1AR–/– mice was more robust than the DTH response in either WT or 2AR–/– mice (Fig. 6D). The ratio of humoral to cell-mediated responses was significantly higher in 2AR–/– mice than in WT mice (Fig. 6E).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
2AR-signaling modulates immunity, this is the first report to describe 1AR inhibition of a DTH response and cell-mediated host immune defenses. Additionally, we have shown that CR treatment inhibits host defenses against LM in mice with functional 1AR signaling (WT FVB/NJ and 2AR–/– mice), whereas 1AR–/– mice are not immunologically impaired by the CR treatment. Previous studies have demonstrated normal immune functions in 2AR–/– mice (27). Our results agree with these findings in that WT and 2AR–/– mice coped similarly with Listeria infection. Moreover, both WT and 2AR–/– mice were immunocompromised by CR treatment. The present data demonstrating improved host defenses against LM in the absence of 1AR corroborate the findings of previous pharmacological studies in which 1AR was identified as a stress-signaling molecule that weakens host resistance (5, 7, 8, 26). The CR-induced increase in bacterial colonization of WT and mutant FVB/NJ mice and mRNA expression of ARs in the liver, but not in the spleen of BALB/c mice, also delineates the existence of stress-induced responses that are organ specific.
Based on the lack of CR-induced inhibition of host immunity in CD4–/– BALB/c mice (8), it had been suggested that CD4+ T cells, possibly regulatory T cells, or downstream CD8+ T cell effectors are involved in the stress-induced immunosuppression. However, CR-induced early changes in the number or phenotype of peripheral lymphocytes have not been detected (8). In this study, we show that CR does not alter the expansion of Ag-specific CD8+ lymphocytes of LM-infected BALB/c mice, indicating that CR-treatment does not interfere with the priming of LM-specific effectors, which occurs within the first 12 h of infection (28). The observed LM-specific CD8 responses further confirm the previous suggestion (5) that CR-treatment inhibits defenses only early (days 2–3) after infection. The investigation of LM-specific CD8+ T cells in AR-deficient mice was not performed because the LLO91–99-specific class I (H2d) tetramers do not match the MHC of the FVB/NJ strain (H2q). Additionally, an increase in the number of regulatory T cells does not appear to be able to account for the inhibition of host defenses against LM.
Because an increase in perforin expression would likely be perceived as beneficial for handling early primary LM infection (29), we were surprised that CR treatment actually increased perforin expression in the liver of BALB/c mice. This stress-induced increase of perforin was still seen at 4 h after LM infection of the CR-treated mice but declined by 24 h. Stress-mediated alteration of NK activities in mice has been previously reported; the duration and type of stress treatment had differential effects on the immunologic outcome in that 1 h of hyperthermic treatment (30) or the stress treatment of electric foot shock for 1 h daily for 3 days (31) both decreased perforin. NK perforin-dependent cytotoxic activities (YAC-1 cytotoxicity) of mononuclear liver preparations were also diminished 24 h after Epi injection (20 µg i.p.); however, NKT cell Fas-dependent killing of syngeneic thymocytes was increased (32). Interestingly, Oya et al. (32) found a significant 50% decrease in T cells (NK1–CD3high) 24 h after a 12-h restraint stress treatment, the decrease was posited to be due to stress-induced apoptotic mechanisms. In another study, stress (H2O2, heat, or high-density growth) increased the susceptibility of activated T cells to NK cell perforin-dependent killing (33). Thus, we suggest that the early CR-induced increase of the cytotoxic perforin molecule plays an important role in our experimental model of neuroimmunomodulation by increasing apoptosis, which consequently increases susceptibility to infection.
LM infection is known to induce lymphocyte apoptosis through the immunodominant LLO protein (34), and type 1 IFNs increase the susceptibility of lymphocytes to apoptosis (35). LM-induced apoptosis is thought to attenuate innate antilisterial responses because the engulfment of apoptotic lymphocytes by macrophages favors IL-10 production (36). It was recently shown that NE also can induce apoptosis of lymphoid cells (37), and our previous studies have demonstrated a stress-induced increase in TNF- production in the liver between days 1 and 3 of an LM infection (26); such an increase could confer increased susceptibility of lymphocytes to apoptosis. Therefore, a possible explanation for the observed stress-induced immunosuppression of host defenses is that stressed mice display premature release of cytotoxic factors, and these factors increase lymphocyte apoptosis, thereby overwhelming the innate mechanisms required for both bacterial clearance and activation of optimal innate and adaptive responses. Perforin-induced death of LM-infected cells does not influence total bacterial burden (38). Therefore, our observation that LM burdens are similar in both stressed and control animals until day 2–3 of the infection is consistent with a possible role of perforin-induced apoptosis early in the infection. Although differences in regulatory T cell subsets were not detected at 24 h after LM infection, it is possible that other CD4+ T cell subsets (i.e., CXCR3+) induce apoptosis in the liver via mechanisms not involving proinflammatory cytokines (39). Additionally, the stress-induced increase in AR receptor expression in the liver, specifically 1AR signaling on F4/80+ cells, implicates catecholamine-induced changes in monocytes. It is known that catecholamines alter macrophage redox state (40, 41, 42) and early inflammatory cytokine profiles (43, 44, 45).
Although our research aim has been to elucidate stress-induced mechanisms of immunosuppression, the improved host resistance that we observe in 1AR–/– infected mice, relative to WT (FVB/NJ)-infected mice, exemplifies the importance of sympathetic nervous system activities on host-pathogen interactions. The role of sympathetic neuronal activity in regulation of host susceptibility to infectious organisms, in particular LM, is well documented (46). Along with our laboratory, others have shown that early host defenses against LM are enhanced by treatment with the neurotoxicant 6-hydroxydopamine (6-OHDA). 6-OHDA selectively and temporarily ablates peripheral dopaminergic nerves, thereby depleting peripheral tissues of catecholamines because DA is the precursor to NE. In these studies, LM colonization of livers and spleens in sympathectomized mice was significantly reduced at days 3–5 after infection, as compared with the colonization of control mice with innervated organs (5, 47, 48, 49). Additionally, peripheral administration of 6-OHDA blocked stress-impaired immunity in spleen but not in liver (5), a pattern that differs from the complete abrogation of stress-induced immunosuppression in both organs of 1AR–/– mice. It is possible that the livers of 6-OHDA-treated mice are still susceptible to stress-induced release of Epi from the adrenal glands, and circulating Epi could influence liver defenses through functional 1AR signaling.
The enhanced host resistance of mice with denervated peripheral organs had been correlated with increased numbers of splenic neutrophils (48) and activated peritoneal macrophages (49) during the first 3 days of an i.p. infection. However, between days 5 and 7 of LM infection, 6-OHDA treatment decreased splenic leukocyte numbers (48). In a similar study of HSV infection, 6-OHDA treatment was shown to suppress the generation of both primary and secondary virus-specific CTL (50). It is important to note that all observed effects of 6-OHDA treatment reported in the cited articles were reduced if mice were pretreated with the catecholamine uptake blocker desipramine, which prevents the destruction of noradrenergic neurons by blockage of 6-OHDA uptake (5, 47, 48, 49, 51). Theses studies suggest that peripheral innervations suppress early innate mechanisms, but not adaptive responses, during an infection; this conclusion is in agreement with our quantification of LM-specific CD8+ cells.
Because the 1AR–/– mice used in the current study can synthesize catecholamines both centrally and peripherally but lack the capacity to signal NE and Epi via 1AR, an assessment of the role of CNS innervation is important in an understanding of the implications of our results. Interestingly, peripheral administration of 6-OHDA does not affect the CNS because this molecule cannot pass the blood-brain barrier (52). However, central ablation of dopaminergic neurons by intrastriatal injection of 6-OHDA (51), bilateral injection of 6-OHDA into the lateral ventricles (53), or genetic removal of DA -hydroxylase (an enzyme required for the production of DA) (54) impaired peripheral host immunity. In other words, the elimination of CNS neurons that release DA (51, 53) or the prevention of DA synthesis by dopaminergic neurons (54) is immunosuppressive, yet the loss of DA neurons in the periphery is immunoenhancing, as noted above. How the elimination of catecholamine signaling in the brain functions to suppress peripheral immune activity is as yet unresolved, but it likely involves inappropriate hypothalamic-pituitary-adrenal axis activation (51) and dysregulation of peripheral catecholamines essential for immune homeostasis (53).
Based on the hemopoietic reconstitution experiments, we infer that the presence of 1AR or 2AR expression in the brain, or lack thereof, does not detectably affect peripheral immunity against LM. Although central release and signaling of DA are critical factors in maintaining immunocompetence, signaling of NE and/or Epi through the 1AR in peripheral organs is immunosuppressive. Improved DTH responses in 1AR–/– mice suggest the existence of cell-mediated specific mechanisms of immunosuppression by 1AR signaling; when these are absent, antilisterial defenses are improved. We observed significant increases in Ag-specific humoral responses of 2AR–/– mice relative to WT mice, following immunization with KLH, whereas previous studies showed equivalent humoral responses produced in vivo by WT FVB/NJ and 2AR–/– mice after immunization with the trinitrophenyl hapten conjugated to KLH (27). The discrepancy between our results and theirs may be a function of differing numbers of naive and/or memory KLH-specific or hapten-specific T and B cells in the two studies. In any case, expression of 1ARs and 2ARs clearly has multiple influences on immunity.
In summary, our study corroborates previously published pharmacological evidence for a 1AR-associated mechanism of immunosuppression. Furthermore, macrophage and/or Kupffer cell expression of both 1AR and 2AR is posited as a key mediator of the stress-induced changes that impair early antilisteria responses. The differential bactericidal activities of immune cells in the liver and spleen after exposure to sympathetic catecholamines may be due to a variation in the expression of 1AR and 2AR by immune cells in these organs.
|
Acknowledgment |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 Address correspondence and reprint requests to Dr. David A. Lawrence, Wadsworth Center, New York State Department of Health, Molecular Medicine, Empire State Plaza, C419, Albany, NY 12201. E-mail address: David.Lawrence{at}Wadsworth.org 
2 Abbreviations used in this paper: LM, Listeria monocytogenes; 1AR, 1-adrenoceptor; 2AR, 2-adrenoceptor; AR, -adrenoceptor; CORT, corticosterone; CR, cold restraint; DA, dopamine; DTH, delayed-type hypersensitivity; Epi, epinephrine; LLO91–99, listeriolysin O91–99; NE, norepinephrine; RT, room temperature; WT, wild type.
Received for publication September 1, 2006. Accepted for publication February 1, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
1-adrenergic receptors. Brain Behav. Immun. 17: 121-133. [Medline]2-adrenergic receptor stimulation in the modulation of Th1, Th2 and B lymphocyte function. Adv. Exp. Med. Biol. 437: 269-278. [Medline]2-adrenergic receptor. J. Immunol. 159: 4857-4867. [Abstract] production by Th1 cells generated from naive CD4+ T cells exposed to norepinephrine. J. Immunol. 166: 232-240. 2-adrenergic receptor by Th1 and Th2 clones. J. Immunol. 158: 4200-4210. [Abstract]2-adrenergic receptor gene. J. Biol. Chem. 274: 16694-16700. 2-adrenoceptors in norepinephrine-induced TNF- production. Biochim. Biophys. Acta 1537: 49-57. [Medline] knockout BALB/c mice. Toxicol. Sci. 89: 444-453. 2-adrenergic receptor. Brain Behav. Immun. 17: 55-67. [Medline]2-Adrenergic receptor stimulation inhibits nitric oxide generation by Mycobacterium avium infected macrophages. J. Neuroimmunol. 101: 68-75. [Medline]1- and 2-adrenergic receptors. Immunology 100: 359-363. [Medline]1 receptors induces heme oxygenase-1 production in RAW264.7 cells. FEBS Lett. 579: 5494-5500. [Medline]2-Adrenergic receptor stimulation inhibits LPS-induced IL-18 and IL-12 production in monocytes. Immunol. Lett. 101: 168-172. [Medline]2-Adrenergic receptor agonist induces IL-18 production without IL-12 production. J. Neuroimmunol. 151: 137-147. [Medline]. J. Neuroimmunol. 160: 87-91. [Medline]-hydroxylase deficiency impairs cellular immunity. Proc. Natl. Acad. Sci. USA 96: 2274-2278.
| ||||||||||||||||||||||||||||||||||||||||||||||||
