Effects of exposure of mice to hindlimb unloading on leukocyte subsets and sympathetic nervous system activity

Abstract

The hindlimb unloading (HU) rodent model was developed to simulate some of the aspects of spaceflight conditions. Our previous studies showed that exposure to HU for 48 h (h) followed by bacterial challenge, reduces the ability of mice to resist infection. The purpose of this study was to investigate the physiological changes in mice during the 48 h of exposure to HU to understand the mechanisms involved in the increased susceptibility to infection observed in mice subjected to these conditions. Female Swiss Webster mice were hindlimb-unloaded during 48 h. Blood samples, spleen and peritoneal cells were removed before and after 18 or 48 h of HU-exposure. Leukocyte subset analysis was performed in spleen and peritoneal cells by flow cytometry, and catecholamine levels were measured in plasma and whole spleen by a catecholamine enzyme immunoassay. Catecholamine levels measured in plasma and spleen were significantly greater in mice exposed to HU compared to control. This increase coincided with significant reductions in spleen size in the HU group. Flow cytometric analyses showed a significant reduction of splenic CD19 + B-cells and NK1.1+ cells in mice exposed to HU with a concomitant increase in T-cells. These results suggest that exposure to HU increases the activity of the sympathetic nervous system (SNS) and induces lymphocyte sub-population changes that may contribute to the deregulation of immunity seen in mice exposed to HU and, more importantly may predispose the otherwise healthy host to the subsequent reduced ability to resist infections.

• Catecholamines
• hindlimb unloading
• immune cell subpopulations
• prolonged inactivity
• spaceflight
• stress

Introduction

Exposure to stressors causes the activation of hypothalamic-pituitary-adrenal axis (HPA) and sympathetic nervous system (SNS) with resultant increases in catecholamine production and alterations in the immune response (Eskandari and Sternberg 2002; Elenkov and Chrousos 2006). Interaction between the nervous and immune systems contributes to maintaining homeostasis while protecting the body against invading pathogens. Under stress, norepinephrine (NE) is released from the sympathetic nerve terminals in various immune organs including the spleen. Stress hormones, including catecholamines and glucocorticoids, have been shown to affect key immune functions including antigen presentation, leukocyte proliferation and trafficking, cytokine and antibody production, and the balance of T helper responses (Madden et al. 1995; Elenkov and Chrousos 2006). It is possible that stress hormones act via interaction with cognate receptors present on immune cells. The inability to re-establish homeostasis after exposure to stress may lead to unbalanced and defective immune responses. The clinical importance of such immunological dysregulation is highlighted by the increased risks of disease development including infectious disease (Godbout and Glaser 2006).

Hindlimb unloading (HU) of rodents is a ground-based model that has been successfully used to study the effects of spaceflight conditions on the physiology of the host and resistance to infection (Chapes et al. 1993; Miller and Sonnenfeld 1994; Belay et al. 2002; Sonnenfeld 2003, 2005a,b; Aviles et al. 2003a, 2005). In this model, mice are tethered and suspended by the tail allowing for uninhibited movement about the cage but with no load bearing on the hindlimbs. Exposure to HU results in physiological changes including bone loss, muscle atrophy, and shifts of fluids, which are similar to those that occur during spaceflight or prolonged bed-rest (Morey-Holton and Globus 2002, Morey-Holton et al. 2005). In addition, significant increases in plasma NE and epinephrine levels (Sonnenfeld et al. 2002; Aviles et al. 2005; Morey-Holton et al. 2005) occur with HU-exposure, supporting the idea that HU is an effective stressor and stimulant of the SNS. Notably, one of the most critical aspects attributed to exposure to this spaceflight mimetic is the reported changes in immune parameters (Murdaca et al. 2003; Morey-Holton et al. 2005; Sonnenfeld 2005b), which have adverse consequences on resistance to infection (Belay et al. 2002; Aviles et al. 2003a,b), and therefore, are a considerable health concern for mission crew members. Mice exposed to HU and infected with 50 percent lethal doses (LD₅₀) of K. pneumoniae (Aviles et al. 2003b) or P. aeruginosa (Aviles et al. 2003a) displayed significantly increased mortality and reduced times to death along with a reduced ability to clear bacteria from organs compared with control.

This work extends our previous studies on altered resistance to infection in mice exposed to HU (Aviles et al. 2003a,b, 2005). Although it seems clear that exposure to HU increases the susceptibility to infection, the intrinsic mechanisms involved in this effect have not been well defined. The aim of this study was to gain insight into the mechanisms by which mice exposed to 48 h of HU are more likely to succumb to bacterial challenge compared with control non-exposed mice. We hypothesized that exposure to HU for a short time (48 h) will induce changes in immune parameters and that the SNS could play a role in those changes. The understanding of the mechanisms leading to altered resistance to infection is critical to design countermeasures to alleviate these effects and reduce the risk of potentially developing life theatening infections in spaceflight personnel as plans for long-term missions in space development.

Materials and methods

Animals

Specific pathogen-free female outbred Swiss/Webster mice 8–10 weeks of age, weighing 20–25 g were purchased from Harlan Laboratories (Indianapolis, IN). The mice were housed in a quiet, isolated room with controlled temperature (21–22°C), a 12 h: 12 h dark light cycle (lights on 07.00 h, off 19.00 h), and access to food and water ad libitum. Experimental procedures commenced after one-week acclimation. The mice were randomly assigned to groups: normally caged control (NC) and HU. The mice subjected to HU were housed individually in specially designed cages (Morey-Holton and Globus 2002; Belay et al. 2002; Aviles et al. 2003b, 2004) and suspended, but not restricted, by the tail using Skin Trac adhesive (Zimmer, Warsaw, IN) at approximately 15° head-down tilt with no load bearing on the hindlimbs, and with unlimited access to food and water. A restrained control group used in our previous HU studies was not included because no differences between restrained and normally housed control were observed in those studies (Belay et al. 2002; Aviles et al. 2003a). Four mice per group at each time point were used in these experiments. Some additional mice were used for spleen weight and blood collection to give adequate sample numbers for statistical power. With the exception of the data in Figure 2, experiments were performed at least in duplicate. Data are presented as group means +/ − sem. Mice were killed by cervical dislocation, under anaesthesia (Isofluorane, 2.5% for induction then 3%), between 08.00 and 10.00 h, at 18 or 48 h after commencement of HU. All experiments were approved by the University Institutional Animal Care and Use Committee of the State University of New York at Binghamton and were carried out under the supervision of a veterinarian.

Sample collection

Tissues were collected under anaesthesia, before cervical dislocation. Blood was obtained by cardiac puncture using 3 ml syringes containing sodium citrate. Plasma was obtained after centrifugation at 750g for 10 min and stored at − 20°C until use. Body weights and spleen weights for each mouse were recorded at time of harvest.

Tissue and cell preparation

For splenocytes, the spleen was removed aseptically and single-cell suspensions were prepared using a cell strainer (BD Falcon, Franklin Lakes, NJ). Erythrocytes were removed by incubation with lysis buffer (BD Biosciences, San Jose, CA) following manufacturer's instructions. For thioglycollate-elicited peritoneal cells (TEPC), mice were injected with 2 mL of sterile 3% thioglycollate, an inflammatory agent, into the peritoneal cavity four days prior to harvest. Peritoneal cells were collected by lavage with 8 ml of sterile RPMI-1640 tissue culture media. Cells were washed, resuspended in Dulbecco's PBS, pH 7.4 (Mediatech, Inc), 2% fetal bovine serum (Hyclone), 0.09% sodium azide (Sigma) and counted. Cell viability was determined by trypan blue (Sigma) exclusion.

Flow cytometry

Samples of 10⁶ cells resuspended in stain buffer (Dulbecco's PBS, pH 7.4, Mediatech, Inc), 2% fetal bovine serum (Hyclone), 0.09% sodium azide (Sigma) were Fc-blocked and labelled with antibodies purchased from BD Biosciences (San Jose, CA). Lymphocytes were discriminated by forward versus side scattering and 10,000 cells were gated and analysed on a dual-laser FACSCalibur flow cytometer (BD Biosciences) using CellQuest Pro software for each sample. Subpopulations of cells were identified with antibodies against lineage-specific markers: PerCP-conjugated anti-mouse CD3ϵ chain (clone 145-2C11), FITC-conjugated anti-mouse CD8α (clone 53-6.7), PE-conjugated CD4 (clone GK1.5), APC-conjugated anti-mouse CD19 (clone 1D3), PE-conjugated anti-mouse NK1.1 (clone PK136). 1 μg of each conjugate was used per 10⁶ cells. Matched isotype controls were used to set negative staining criteria.

Catecholamine enzyme immunoassay

Quantitative measurements of epinephrine and NE were determined by enzyme immunoassay using the 2-CAT EIA kit (Labor Diagnostica Nord GmbH & Co. KG, Germany, Rocky Mountain Diagnostics, Colorado Springs, CO) by following the manufacturer's instructions. As suggested by the manufacturer, levels from haemolysed plasma were not used due to false low values. For plasma NE and epinephrine concentrations, 150 μl of plasma were extracted, acylated, and subjected to competitive enzyme immunoassay. Absorbances were read at 450 nm (Beckman Coulter DTX 880 Multimode Detector) and concentrations were interpolated from the standard curve and corrected for dilution. For tissue NE content, whole spleens were weighed, individually homogenized, and normalized with 10 μl 0.1 N HCl per milligram of tissue (10%, weight/volume). Tissue debris was removed by centrifugation at 11,000g and 10 μl of homogenate were handled as above. NE concentrations were directly interpolated from the standard curve. Assay sensitivity for epinephrine was 11 pg/ml (plasma), the intra-assay variation from two samples was 15.0 and 6.9%, and the inter-assay variation from two samples was 13.2 and 15.4%. For NE, assay sensitivity was 44 pg/ml (plasma), the intra-assay variation from two samples was 16.1 and 9.8%, and the inter-assay variation from two samples was 8.5 and 15.0%.

Statistical analysis

Data were analysed using Staview 5.0.1 with alpha set a priori at p ≤ 0.05. ANOVA followed by post hoc Fisher's test was used to test differences between more than two groups and among groups at different time points. Student's t-test was used to test statistical significance between any two the groups.

Results

Exposure to HU and catecholamine concentrations

In this study, concentrations of catecholamines were measured not only in plasma samples but also in the spleens obtained from HU- exposed and NC mice. Consistent with previous studies, peripheral catecholamine levels were greater in mice under unloading conditions. Mice exposed to HU had significantly elevated plasma epinephrine concentrations above control mice at both 18 h (p ≤ 0.0001) and 48 h (p ≤ 0.001), F(2,29) = 8.02, p = 0.0017. Plasma NE concentrations were elevated in the HU group with significant differences seen at 48 h (p ≤ 0.03) of HU exposure compared to NC mice (Table I).

Table I.  Effects of exposure to HU on catecholamine levels measured in plasma and spleen.

Condition	Plasma^*		Spleen^†
	Epinephrine	Norepinephrine	Norepinephrine
NC^‡	480 ± 32 (n = 9)	3245 ± 436 (n = 9)	39 ± 3 (n = 8)
HU-18^¶	1064 ± 108** (n = 12)	4524 ± 965 (n = 12)	49 ± 2* (n = 8)
HU-48^§	1226 ± 185** (n = 11)	6213 ± 1303* (n = 11)	67 ± 7**# (n = 8)

*Significant differences compared to NC group; *p ≤ 0.05; **p ≤ 0.01. #Significant differences between hindlimb unloaded groups

^* Values represent the means ± SE of plasma catecholamine levels measured in pg/ml

^† Values represent the means ± SE of spleen NE levels measured in ng/ml

^‡ NC mice

^¶ Mice hindlimb unloaded for 18 h

^§ Mice hindlimb unloaded for 48 h.

Neural input to the spleen is exclusively sympathetic (Nance and Sanders 2007); therefore, relative levels of NE from the whole spleen homogenates of mice were determined. HU showed a significant effect on spleen NE concentrations, F(2,21) = 9.53, p = 0.0011. Post hoc analyses showed spleen NE concentrations in mice exposed to 18 h of HU were significantly increased (p ≤ 0.04) by 23% above values in normally caged controls. Strikingly, the levels of NE at 48 h of HU were significantly increased (p ≤ 0.003) approximately 70% above levels found in control mice (Table I). Splenic epinephrine concentrations were below detectable limits of the assay (data not shown). The prominent local increase in NE concentration in the spleens of HU-mice strongly corresponds with the decreased spleen cell proliferation and cytokine production seen in HU-mice in our previous studies (Aviles et al. 2005).

Exposure to HU and spleen size and numbers of elicited peritoneal cells

Previously it was determined that spleen size in mice exposed to HU is significantly decreased compared to control mice (Yamauchi et al. 2002; Aviles et al. 2005) at 24 h (Aviles et al. 2005) while 12 h exposure was not sufficient to produce a change. Here we included the 18 h time point to more precisely determine the minimal exposure time to HU that results in reduced spleen size. ANOVA showed a significant effect of HU on spleen size, F(2,49) = 7.916, p = 0.001. Post hoc analyses showed that spleen weights per gram of body weight were significantly reduced in the 18 h (p ≤ 0.02) as well as in the 48 h HU group (p ≤ 0.0006) compared to normally caged controls (Figure 1).

Figure 1 Effects of exposure to HU conditions on spleen size. Mice were normally caged (NC, white bar) or exposed to HU for 18 h (HU-18, grey stippled bar), or 48 h (HU-48, black bar). Spleen size is expressed as the means of spleen weights (gram) divided by body weights (gram) for each mouse ± SE. * indicates significant difference compared to normally caged group (*p ≤ 0.05, **p ≤ 0.01). # indicates significant difference between hindlimb unloaded time point groups (p ≤ 0.05).

TEPC were enumerated and compared between experimental groups; this showed reduced cell numbers with significance established at the 48 h time point. Compared with the controls, TEPC numbers were reduced, but not significantly, with HU-18 h exposure and significantly reduced with HU-48 h (t(6) = 1.928, p ≤ 0.05) exposure (Figure 2). Cytometric analyses of these cells showed no statistically significant differences among groups at any time point (data not shown).

Figure 2 Effects of exposure to HU on numbers of TEPC. Mice were injected with 2 ml of sterile 3% thioglycollate, an inflammatory agent, into the peritoneal cavity four days prior to harvest and normally-caged (NC, white bar) or exposed to HU for 18 h (HU-18, grey stippled bar) or 48 h (HU-48, black bar) prior to harvest. Values are expressed as the means of total numbers of peritoneal cells collected ± SE. Asterisk indicates significant difference compared to normally caged group (p ≤ 0.05).

Exposure to HU and splenic immune cell populations

The observed size reduction of the spleen, the major secondary immune organ, as a result of HU-exposure that occurred in the absence of antigenic challenge was suggestive of cellular regressive and/or apoptotic events, which could play an important role in rendering the host more susceptible to infection. To address this issue, splenocytes from HU-exposed and normally caged mice were stained for cell surface antigen markers and subjected to flow cytometric analyses (Figure 3). ANOVA showed that there were significant effects of HU on percentages of splenic lymphocyte sub-populations with no significant changes in myelocytic cells (B cells, F(2,21) = 3.1, p = 0.034; CD8 T cells, F(2,21) = 4.14, p = 0.031; CD4 T cells, F(2,21) = 3.82, p = 0.038; and NK1.1 cells, F(2,20) = 4.54, p = 0.024. HU also had effect on T/B cell ratios, F(2,21) = 4.64, p = 0.022 but not on CD4/CD8 ratios, F(2,21) = 1.26, p = 0.3. The percentage of CD19 + B cells was significantly reduced in spleens of mice exposed to 18 (p ≤ 0.05) and 48 h (p ≤ 0.05) of HU compared with NC mice (Figure 3A,B,E). Both CD8+ and CD4+ T cell percentages were concomitantly and significantly increased in the spleen with exposure to HU (Figure 3A–C,G,H); however, the CD4+ to CD8+ ratios showed only a trend to be reduced with 18 and 48 h of HU-exposure (Figure 3C,I) indicating that the T cell population might have been altered with HU-exposure. Taken together, T cell to CD19+ B cell ratios were increased with HU at both 18 and 48 h (p ≤ 0.01) when compared to control (Figure 3F). In addition, percentages of NK1.1+ natural killer cells were significantly reduced in the HU groups compared to control (p ≤ 0.05 in the HU-18 h group, and p ≤ 0.01 in the HU-48 h group; Figure 3D,J). Percentages of Gr-1+ and CD11b+ cells of myeloid lineage in the spleen and in the peritoneum (TEPC) were not significantly different between groups in this study (data not shown).

Figure 3 Effects of exposure to HU conditions on splenic immune cell type percentage distribution. Samples of 10⁶ resuspended spleen cells from normally caged mice (NC, white bar) or mice HU-exposed for 18 h (HU-18, grey stippled bar), or 48 h (HU-48, black bar) were stained with CD8α-FITC, CD4-PE, CD19-APC (A–C); CD8α-FITC, NK1.1-PE, CD3-PerCP (D); or matched isotype controls. (A–D) Flow cytometric dot plots; inset represents mean percent ± SE. (E–J) Bar graphs; values are recorded as percent ± SE of cells within the lymphocyte gate stained positive for the indicated surface antigen. In (F), “T” is defined as percent CD4+ plus CD8+ cells. Asterisks indicate significant difference compared to normally caged group (*p ≤ 0.05, **p ≤ 0.01).

Discussion

Results from this study show that exposure to HU conditions produced locally elevated NE level in the spleen, reduced overall spleen size as early as 18 h of HU exposure, decreased the number of elicited peritoneal cells, and induced specific changes in lymphocyte subpopulations in HU mice. The catecholamine level changes observed as a result of HU may contribute to the previously reported impairment of immunity that occurs with subsequent pathogen challenge (Aviles et al. 2003a,b). However, more studies are needed to determine if elevated splenic NE seen with HU-exposure is a consequence of increased biosynthesis, reduced NE release, or reduced degradation. The increased catecholamine levels, especially splenic NE, during the HU period coincided with decreased spleen size (Table I and Figure 1) and reduced numbers of TEPC (Figure 2). We report that percentages of CD19 + B lymphocytes and NK1.1+ natural killer cells are significantly reduced and propose that these reductions are pivotal. The reduction in splenic B cell percentages (Figure 3A,B,E,F) in the absence of pathogen challenge could explain some of our previous results were exposure to HU resulted in decreased spleen cell proliferation and cytokine production after in vitro LPS stimulation (Aviles et al. 2005) and reduction in specific anti P. aeruginosa IgG antibody production (Aviles et al. 2003a).

NK cells function in early phases of infection to contain infection before the adaptive immune response generates antigen specific lymphocytes (Lodoen and Lanier 2006). Studies have shown that NE reduces NK cell cytolytic activity with coinciding reductions in perforin and granzyme B mRNA and protein levels in the spleen (Gan et al. 2002; Dokur et al. 2004). Therefore, the elevated splenic NE levels seen in HU-exposed mice (Table I) may be critical in determining the magnitude or kinetics of ensuing immune responses. The reduced NK1.1+ natural killer cell population in the spleen in HU mice (Figure 3D,J) may also translate into immune impairment if pathogen challenge should subsequently occur.

The SNS acts directly on cells through adrenergic receptors present on most types of immune cells. Interestingly, higher adrenergic receptor densities reportedly occur on B cells and NK cells (Landmann 1992; Sanders and Kavelaars 2007) compared to T cells and here we reported that CD19 + B cells and NK1.1+ cells percentages are the most reduced under HU-exposure. Additional studies are necessary to further test the relationship between the SNS and immune system in our model using catecholamine receptor (specifically the β-adrenergic receptors) antagonists.

An effective host immune response to infection must control pathogen dissemination without inducing immunopathology. Impairment of the immune system imposed by stressors including HU can increase susceptibility to infections (Belay et al. 2002; Aviles et al. 2003a,b). Here, we have extended our previous studies and effort to uncover evidence linking HU-induced stress-related immune change to disease vulnerability by examining the healthy individual. Data from this study indicates that the status of the host even in the absence of pathogen challenge is altered by exposure to HU and therefore plays a critical role in susceptibility to infection. Also, the data suggest that the process of adaptation to unloading conditions may promote immunological changes that could have detrimental effects on the ability of the host to resist infection. It appears that HU acts on biological homeostatic systems in a way that may result in the exacerbation of infectious diseases. Understanding the mechanisms that promote the immune parameter changes induced by HU will provide insights to discern how exposure to stressors influences the outcome of diseases; particularly infection with pathogens that under normal conditions are readily controlled and resolved by a healthy immune system.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Acknowledgements

This work was funded in part by a grant from the Amino Up Chemical Company (Sapporo, Japan). We thank the animal facilities personnel at Binghamton University for their support during these studies.