Antidepressant treatment during social challenge prior to 1 year of age affects immune and endocrine responses in adult macaques

Antidepressants are widely used in treating depression and other behavioral problems in children and adolescents. Little is known about the long-term effects of these agents, particularly on physiological systems. The effects of previous antidepressant treatment during a social challenge in 9-month-old rhesus monkeys (Macaca mulatta) on their adult immune and endocrine responses were studied. Prior to the social challenge, the monkeys were reared either by their mother or in a peer group. Monkeys were treated with either a serotonergic agonist (fluoxetine), a noradrenergic agonist (desipramine), or saline during social separation. Non-separated, saline-treated monkeys served as control monkeys. In order to evaluate immune effects of early antidepressant treatment, adult monkeys were immunized with a novel antigen, tetanus toxoid. Blood samples were collected prior to and at 4-5-day intervals for 28 days after immunization. Plasma total immunoglobulins (IgG and IgM), complement levels (C3 and C4), tetanus antibody titers, and cortisol were assessed. Antibody levels were lowest in monkeys treated with antidepressants regardless of specific drug treatment or early rearing condition. Drug-treated subjects had elevated plasma immunoglobulins and complement protein levels. Cortisol was also highest in drug-treated subjects. These results should be considered when prescribing commonly used antidepressants for treatment of childhood disorders.

A preliminary description of responses of free-ranging rhesus monkeys to brief capture experiences: behavior, endocrine, immune, and health relationships

A cohort of free-ranging rhesus monkeys has been followed since birth in 1994 on the island of Cayo Santiago, Puerto Rico. At 3 years of age, subjects were trapped and blood samples were collected after capture and prior to release the following day. Blood samples were processed for natural cytotoxicity toward xenogeneic tumors, phenotyping, and plasma hormones. Intestinal parasites were determined from fresh stool samples collected during trapping. Data were also available from the previous year for antibody titers to latent viruses prevalent in this population. Behavioral traits of each monkey were characterized using a previously developed trait scale for rhesus monkeys. Natural cytotoxicity toward both K562 and Raji targets declined from capture until release the following day. Plasma cortisol rose and plasma prolactin and growth hormone fell during the period of captivity; a rise in insulin was significant. It was expected that individual differences in behavioral traits might predict immune and hormone levels at the time of capture or changes in these parameters during the capture period. Although behavioral adjectives tended to cluster along three orthogonal dimensions (Insecurity, Irritability, and Sociability), they bore no relationship to the physiological parameters collected acutely (in vitro immune and endocrine parameters). The individual difference markers of gender and maternal rank were not related to the magnitude of the observed changes in these in vitro parameters, either. However, an in vivo measure (CMV titer) was related to individual differences in Irritability. It was concluded that the magnitude of the stress associated with capture overwhelmed the individual difference effects.

Elevated cytotoxicity in combat veterans with long-term post-traumatic stress disorder: preliminary observations

Resting immune [WBC and differential cell counts lymphocyte phenotyping (CD2, CD4, CD8, CD16, CD20, and CD56), and NK activity] and endocrine (cortisol, prolactin, growth hormone, and DHEA-SO4) parameters were measured in 10 male, Vietnam combat veterans diagnosed with long-term post-traumatic stress disorder (PTSD) and 9 control Vietnam combat veterans without a PTSD diagnosis but with a comparable history of alcohol abuse. Subjects completed a battery of psychological questionnaires. We report on preliminary observations of the relationship between PTSD and physiological and psychological parameters. With some important exceptions, PTSD patients did not differ from the age-matched control group with regard to hormone levels or lymphocyte phenotypes. However, NK activity was higher in the PTSD population than in the controls. Beck, Mississippi, and Combat Exposure scores were significantly elevated in the PTSD population. In contrast to previous observations in depressed populations, depression (indicated by elevated Beck scores), comorbid with PTSD, was associated with increased natural cytotoxicity.

Juvenile friends, behavior, and immune responses to separation in bonnet macaque infants

Individual differences in the response to maternal separation in nonhuman primate infants have been attributed to (among other variables) presence or absence of processes that may model social support in humans. Alternative attachments to other members of the social group buffer the infant against a depressive response to maternal separation. This hypothesis was tested in a group of bonnet macaques by manipulating the presence or absence of alternative juvenile attachment figures (friends) during separation. Infants who retained such attachments showed fewer behavioral evidences of depression when separated from their mothers. These infants without friends also showed changes in lymphocyte activation by mitogens or natural cytotoxicity that were not evident in the infants with juvenile friends. Across all separated infants, natural cytotoxicity was positively correlated with juvenile affiliative behavior directed toward the infants during the separation. These results support the hypothesis that social support, available from alternative attachments, can modulate the response to loss, and can account for some of the individual differences seen in these responses.

Natural cytotoxicity toward K562 cells by macaque lymphocytes from infancy through puberty: effects of early social challenge

The consequences of a single maternal separation experience followed by reunion at 6-7 months of age was studied in socially housed pigtail (Macaca nemestrina) and bonnet (Macaca radiata) macaques. At 15 months, these subjects were removed from their natal group and placed in same species social groups, consisting of other separated and matched control subjects. Some subjects were followed until they reached an average age of 4 years while remaining in this group. Blood samples were drawn to permit assessment of the ability of peripheral blood lymphocytes to lyse, in vitro, K562 cells. Maternal separation failed to affect lysis of K562 cells acutely, although lysis by matched control subjects appeared to be transiently reduced 2 h after removal of the adult female. A longer-term influence was noted such that lysis of targets in previously separated subjects was greater than that in matched controls. Lysis rose over time regardless of species or experimental condition. A striking internal consistency in the lysis was also noted. Lysis was highly intercorrelated (r's > .60, p < or = .005) within subjects across time during baseline, separation, and reunion phases of the acute social challenge. In addition, there was a tendency for this correlation to hold over the longitudinal phases. Lysis of K562 targets by macaque lymphocytes would appear to possess trait-like stability; however, the range of lysis may be modified by early experiences.

Morphine-induced alterations in antibody levels: receptor and immune mechanisms

We have previously shown that an acute administration of morphine (10 mg/kg, i.v.) decreases IgG, but not IgM, antibody levels to antigen administered before morphine. Further, decreases in IgG were blocked by previous administration of naltrexone, indicating that receptor binding is critical to the decreased antibody levels. These studies investigated potential receptor and immune mechanisms for these effects. To investigate potential receptor mechanisms, the stereoselectivity and location of receptor binding was determined. The results of these experiments suggest morphine must bind stereoselectively to central sites to decrease antibody levels after antigen administration. To investigate potential immune mechanisms for these changes, antibody secreting cells (ASC) for keyhole limpet hemocyanin-specific IgG and IgM were enumerated. Morphine decreased ASC for IgG but increased ASC for IgM. Two pathways for the genetic switch from IgM to IgG production were investigated. One pathway requires interferon-gamma to stimulate IgM-secreting cells to switch to IgG2a-secreting cells. Another pathway requires interleukin-4 to stimulate IgM-secreting cells to switch to IgG1- secreting cells. IgG1 and IgG2a levels were measured to determine if these pathways were differentially affected and only IgG2a levels were decreased. Further, these decreases were accompanied by decreased IFN-gamma levels but not by altered numbers of splenocytes. These data indicate that morphine may alter the ability of ASC to switch from IgM to IgG2a production, possibly by reducing the availability of IFN-gamma.

A long-term increase in basal levels of corticosterone and a decrease in corticosteroid-binding globulin after acute stressor exposure

Adrenal glucocorticoids play an important role in mediating many of the behavioral and physiological effects of exposure to stressors. Focus has been primarily on the acute stress-induced rise in glucocorticoids [corticosterone (CORT) in the rat]. There are reports, however, that exposure to chronic stressors can produce an increase in basal CORT and a decrease in corticotropin-binding globulin (CBG). These changes occur subsequent to the stress-induced rise in CORT. The following experiments examined whether exposure to an acute stressor (100 5-sec inescapable tail shocks; IS) could also produce long term changes in basal CORT and CBG. We report that a single session of IS results in an increase in basal total serum CORT that persists 48-96 h after IS termination. The increase is present only at the diurnal trough (morning). CBG levels ae also decreased for 24-48 h. The decrease is present at both the diurnal peak (evening) as well as the trough (morning). These changes result in an increase in the percent and amount of biologically active CORT (unbound or free). Thus, glucocorticoid-sensitive targets are exposed to high levels of free CORT for several days after IS termination. The long term increase in free CORT reported here may play an important role in mediating some of the effects produced by IS as well as those produced by other acute stressors.

Individual differences in macaques' responses to stressors based on social and physiological factors: implications for primate welfare and research outcomes

Primates are used extensively in a variety of research settings. Federal regulations in the US mandate that caretakers provide for the 'psychological well-being of laboratory primates'. One of the difficulties in implementing this law has been both in the definition of psychological well-being and in the need to deal with each primate species and, in some cases, age or sex class, uniquely. Non-human primates exhibit distinct individual differences in their behavioural and physiological responses to experimental challenges and caretaking procedures. We have been investigating what factors can predict some of these individual differences, and have found that factors both intrinsic and extrinsic are significant. Extrinsic factors found to predict individual differences in response to stressors include the nature and prior experience with the challenge, the presence of familiar peers and availability of social support. Intrinsic factors include cognitive interpretations of the challenge and temperamental differences in reactivity. These studies highlight the importance of understanding the context and individual psychology of macaques in order to provide laboratory environments conducive to their welfare, and in order to understand the impact experimental and caretaking procedures are likely to have on the health and welfare of our subjects.

Stressed rats fail to expand the CD45RC+CD4+ (Th1-like) T cell subset in response to KLH: possible involvement of IFN-gamma

Exposure to stressors effects various aspects of immune function, including the in vivo antibody response. We have previously reported that rats exposed to an acute session of inescapable tail shock (IS) show long-term reductions in anti-KLH (keyhole limpet hemocyanin) IgM and IgG. The mechanisms responsible for this suppression are currently unknown. Previous work has suggested changes in CD4+ T cells could be important. We report here that exposure to IS results in a reduction in Con A-stimulated IFN-gamma levels in mesenteric lymphocytes and splenocytes taken immediately after IS termination. In addition, IS exposure prevents the KLH-induced increase in the number of CD45RC+CD4+ T cells (Th1-like) in both the mesenteric lymph nodes and the spleen 4 days after immunization. The failure of KLH to expand the CD45RC+CD4+ subset could be due to the stress-induced reduction in IFN-gamma levels reported in cells taken at the time of immunization. Implications of these findings as a mechanism for the decrease in the in vivo antibody response previously reported is discussed.

Total cortisol, free cortisol, and growth hormone associated with brief social separation experiences in young macaques

Many behavioral, immunological, and physiological consequences or brief maternal separation in bonnet (Macaca radiata) and pigtail monkeys (Macaca nemistrina) have been documented. However, the impact of social separation on plasma cortisol and growth hormone is unknown for these particular species. In the present study, the behavioral and endocrinological consequences of a 2-week maternal separation in socially housed infant bonnet and pigtail monkeys were followed. In seven pairs (separated and matched control) of bonnet and six pairs of pigtail infants, plasma was obtained under baseline, separated, and reunion conditions twice weekly for the duration of the study. Blood samples were obtained from both infants of the pair in approximately 10 min. Plasma total cortisol, free cortisol, and growth hormone were measured in these samples. Focal animal behavioral observations were made on all subjects twice daily throughout the study period. In both species, total cortisol and free cortisol rose immediately following maternal separation in comparison to the matched nonseparated controls and returned to basal levels (e.g., that of matched nonseparated controls) following reunion with the mother. In contrast, plasma growth hormone rose only in the pigtail infants over a time course that peaked around the time of reunion. Multiple regression techniques indicated for the first week of separation, in the separated but not control subjects, that mean plasma free and total cortisol was positively related to distress behaviors (vocalization and postural slouch) observed during this week and negatively related to social behaviors (play and proximity to others) noted during the same period. In contrast, plasma growth hormone was related to both species and sex of the subjects but unrelated to behavioral variables.

Stress-induced reduction in the rat mixed lymphocyte reaction is due to macrophages and not to changes in T cell phenotypes

Exposure to aversive events or stressors modulates various aspects of immune function. We have previously reported that exposure to an acute stressor, inescapable tail shock (IS), resulted in a shift in T cell subpopulations in rat mesenteric lymph nodes but not in cervical lymph nodes (Fleshner et al. (1992) J. Neuroimmunol. 41, 131-142). The mesenteric CD4+/CD8+ ratio was increased immediately after exposure to IS and was due primarily to an increase in the percent of CD4+ cells. The present experiments were designed to determine the relationship between the IS-associated phenotypic shift and its significance in the function of CD4+ T cells. The function assessed was the in vitro proliferative response to alloantigens coded for by the Major Histocompatibility Complex (MHC). Using the mixed lymphocyte reaction (MLR), we report that exposure to IS resulted in a decrease in the MLR response of cells from both cervical and mesenteric lymph nodes. Depletion of macrophages (nylon wool adherent cells) eliminated the IS-induced reduction and co-culture of macrophages (irradiation-insensitive cells) from shocked rats produced the suppression. One interpretation of these data is that exposure to IS resulted in the activation of macrophages and the release of a suppressive factor which reduced the MLR response of peripheral lymph node lymphocytes.

Natural killer cell activity is reduced in association with oral contraceptive use

Oral contraceptive (OC) use has been associated with increased incidence of a number of infections, but the mechanisms behind these changes is unclear. The present study compared OC users and nonusers in natural killer (NK) cell activity, NK phenotype, and illness frequency. Subjects were 55 female medical students (19 OC users, 36 nonusers). Three blood samples were obtained, 1 mo apart. Natural cytotoxicity was tested with a 51Cr assay. Self reports of illness symptoms during the previous week were collected at each blood sampling. NK phenotype number was assessed by flow cytometry. Oral contraceptive users had lower natural cytotoxicity and increased frequency of sneezing, gastrointestinal distress, runny nose, sore throat, coughing, and total illness symptoms, relative to nonusers. No differences were found between OC users and nonusers in NK phenotype number. These findings support the hypothesis that differences between users and nonusers in infection rates might be due to alterations in NK activity.

Morphine-induced decreases in in vivo antibody responses

Endogenous opioids have been shown to be released during acute stress and could play a role in immune modulation and activation of the hypothalamo-pituitary-adrenal axis. We investigated the ability of morphine sulfate to mimic stressor effects on decreases in in vivo antibody responses. Sprague-Dawley and Fischer 344 rats were given an intraperitoneal injection of an antigen, Keyhole limpet hemocyanin (KLH), followed by a single intravenous injection of either saline or varying doses of morphine sulfate. The corticosterone and anti-KLH IgG antibody responses to morphine were measured. A dose-dependent increase in corticosterone was observed. Significantly lower levels of anti-KLH IgG antibodies were observed in morphine-treated animals but these effects were strain and dose dependent. In Sprague-Dawley rats, 3 and 10 mg/kg doses of morphine decreased antibody levels while 1.5, 5, and 15 mg/kg did not change antibody responses. In Fischer 344 rats a dose of 5 mg/kg of morphine decreased antibody levels while 10 and 15 mg/kg did not change antibody responses. These results indicate that morphine can decrease antibody levels and that these decreases are not correlated with elevated levels of corticosterone. To determine if opioid binding is critical to these changes, animals received naltrexone prior to the administration of morphine. Naltrexone partially attenuated corticosterone levels, but completely blocked morphine-induced changes in immune function.

Intrinsic and extrinsic factors affect infant responses to maternal separation

Responses of individuals to the loss of a primary attachment object may be quite variable. In humans, it has been suggested that only about 25% of bereavements result in substantial psychological or medical morbidity (Hamburg et al. 1975). In nonhuman primates, which are used to model responses to separation and loss, a similar estimate of about 25% has also been obtained (McKinney 1985). In addition, there are wide-ranging species differences in vulnerability with regard to the nature and severity of the response to maternal separation and/or loss. All of these findings suggest that there are important processes, intrinsic and/or extrinsic to the individual, that contribute to the probability that a loss will produce a major behavioral or physiological response. We have been systematically examining some of the factors that may account for a portion of this variability in two species of macaques (bonnet monkeys Macaca radiata; and pigtail monkeys, M. nemestrina).

Modulation of the in vivo antibody response by a benzodiazepine inverse agonist (DMCM) administered centrally or peripherally

Exposure to stressors can result in changes in immune function. Although there is increasing information concerning the peripheral hormonal and neural mediators of stress-induced changes in immune function, there is little information concerning the central nervous system mechanisms that lead to the peripheral changes. The following experiments examined the possible involvement of the benzodiazepine-GABAA-chloride complex in modulation of the in vivo antibody response. Rats were given either peripheral or intracerebroventricular injections of methyl-6,7-dimethoxy-4-ethyl-beta-carboline-3-carboxylate (DMCM), a drug that has been shown to act at the benzodiazepine-GABAA complex and produces a behavioral state similar to anxiety. Rats were then immunized with keyhole limpet hemocyanin (KLH) and serum levels of KLH-specific antibody were measured for 2 weeks after immunization. Both peripheral and central administration of DMCM modulated the in vivo antibody response. The dose-response relationship of DMCM and changes in antibody levels was nonmonotonic, with high doses resulting in an increase in serum antibody levels and moderate doses resulting in a decrease in serum antibody levels. A possible role of the benzodiazepine-GABAA system in stress-induced immunomodulation is discussed.

Specific antibody levels in free-ranging rhesus monkeys: relationships to plasma hormones, cardiac parameters, and early behavior

Levels of tetanus-specific antibodies were assessed in free-ranging, yearling rhesus monkeys following prophylactic immunization with tetanus toxoid. Each subject's behavior had been observed between 11 and 25 weeks of age and approximately 2 months later during its mother's first concentrated mating period as a part of another study. Prior to immunization, at approximately 1 year of age, cardiovascular parameters, and several plasma parameters [cortisol, adrenocorticotrophic hormone (ACTH), growth hormone (GH), interleukin-2 (IL-2), and total plasma immunoglobulin G (IgG)] were measured during a brief period of captivity. Antibody titers noted approximately 2 weeks after immunization were related to cardiovascular parameters. Thus, yearlings with high heart rates and low heart rate variability during captivity had the highest tetanus-specific serum IgG. Levels of plasma cortisol, ACTH, total IgG, and IL-2 noted at the time of capture were unrelated to subsequent antibody levels. Antibody titers were, however, positively correlated with GH noted immediately following capture on the day prior to immunization. Antibody titers were also related to the infants' behavior observed during their mother's first concentrated mating period. Infants who were most distressed (high levels of distress vocalization) when their mother resumed mating (a time particularly stressful for free-ranging rhesus infants) showed lower antibody titers to tetanus immunization as a yearling. The present observations add support to the existence of a relationship between behavior during exposure to an early stressor and later immune regulation and that certain cardiovascular parameters may be related to certain indicators of immunoregulation.

Anesthesia-induced modulation of in vivo antibody levels: a study of pentobarbital, chloral hydrate, methoxyflurane, halothane, and ketamine/xylazine

The influence of anesthesia on long-term changes in in vivo antibody levels after antigen challenge was examined. Rats experienced a surgical plane of various anesthetics alone (anesthesia/intact) or in combination with laparatomy (anesthesia/laparatomy) and were given 1 or 3 wk to recover. Antigen, keyhole limpet hemocyanin, was then administered, and antibody levels specific to the antigen were measured during the next 14 days. Comparisons were made between anesthesia-treated animals and home cage controls. Pentobarbital and chloral hydrate produced decreases in in vivo antibody levels even 3 wk after exposure, whereas halothane, methoxyflurane, and ketamine/xylazine did not. Ketamine/xylazine produced moderate but not significant decreases in antibody levels when 1 wk intervened between exposure and antigen administration, but not when 3 wk intervened. Surgery did not produce larger changes in antibody levels than did anesthesia itself. These data suggest the possibility that some anesthetics, per se, may contribute to infection that may occur postoperatively.

Blockade of the hypothalamic-pituitary-adrenal response to stress by intraventricular injection of dexamethasone: a method for studying the stress-induced peripheral effects of glucocorticoids

Interest in the mechanisms whereby stressors can influence behavior and physiological functioning has involved the use of a variety of methods to prevent the stress-induced release of glucocorticoids, an important and commonly studied stress hormone. We examined the effect of intracerebral ventricular dexamethasone (ICV DEX) on the stress-induced release of adrenocorticotropic hormone (ACTH), corticosterone, plasma epinephrine (E), and plasma norepinephrine (NE). Male Sprague-Dawley rats were stereotaxically implanted with third ventricle ICV cannulae, administered DEX or vehicle, and exposed to 100 1.6-mA tail shocks. Stress hormones were assessed from blood taken during and after the cessation the shock. We report an ICV DEX injection protocol (10 microgram given four times) that results in blocking the stress-induced release of ACTH and corticosterone, and attenuating the stress-induced release of plasma E and NE. We hypothesize that ICV DEX reduces hypothalamic corticotropin releasing hormone (CRH) synthesis and/or release. This method would be especially useful for those studying the effect of pituitary-adrenal hormones on steroid sensitive peripheral targets, such as the immune system.

Specific changes in lymphocyte subpopulations: a potential mechanism for stress-induced immunomodulation

The mechanisms by which stressors alter immune function are not well understood. One hypothesis for stress-induced immunomodulation is that since immune responses require cooperation of different cell types, stress-induced shifts in cell populations might affect an organism's ability to mount an immune response. We sought to determine if inescapable shock (IS) could alter lymphocyte subpopulations and if so, whether this could be a mechanism for shock-induced immunomodulation. Our results suggest that IS produces changes in lymphocyte subpopulations and that these shifts could be responsible for modulation of in vivo antibody production. Exposure to IS resulted in an increase in the percent of CD4+ mesenteric lymphocytes and a decrease in the percent of CD8+ mesenteric lymphocytes when examined immediately after the cessation of IS. The stressor reduced antibody production to antigen processed at the altered mesenteric nodes, but did not alter antibody production to antigen processed at other sites. No measurable shifts were found in other compartments examined. The changes in CD4+ and CD8+ mesenteric lymphocytes resulted in an increased CD4+/CD8+ ratio that persisted for 1-24 h after stressor termination, becoming absent 48 h after IS termination. The stress-induced reduction in antibody production occurred only when antigen was given immediately prior to but not when antigen was given 48 h post stress. These findings suggest that the effects of a stressor could be specific to the manner in which the antigen enters the body, and that the stress-induced decrease in antibody production could be due to altered lymphocyte subpopulations as reflected by an increased CD4+/CD8+ ratio.

Immune responses following competitive water tests in two species of macaques

A panel of immune parameters (lymphocyte activation by mitogens, natural cytotoxicity, and differential cell counts) was assessed in socially housed pigtail and bonnet macaques 1 and 2 weeks before, 48 h after, and 1 and 2 weeks after a competitive water test. Species differences were found in both baseline measures and the responses to the test: Immune measures observed during baseline periods were lower in pigtail macaques. Furthermore, only the pigtail macaques showed changes in mitogen activation and cytotoxicity at 48 h post-test. Dominance-related behaviors affected these responses both within and across social groups. The species differences may be accounted for by the differences in the behavioral responses of the two species to the test: Pigtail macaques consistently contested access to the water during the test, whereas bonnet macaques did not. These results suggest that the immune system can be modulated by psychosocial behavioral systems, particularly during times of stress.

Exercise does not modify spatial memory, brain autoimmunity, or antibody response in aged F-344 rats

Old F-344 rats were given endurance training over a 10-week period on a motorized treadmill. This treatment resulted in substantial heart-to-body weight ratio increases, indicative of effective training. To determine whether endurance training might alter some of the known immune system and cognitive changes observed during aging, exercised old rats were compared to nonexercised old and young controls on three variables: in vivo antigen-specific immune activity, brain-reactive antibody formation, and spatial memory. The exercise training did not influence any of these measures in the old rats. Both groups of old rats showed poorer antibody response to a specific antigen, more brain-reactive antibody formation, and poorer spatial memory than the young controls. There was, however, a significant relationship between brain-reactive antibody formation and spatial memory performance, regardless of training condition.