SEARCH

SEARCH BY CITATION

Keywords:

  • atrophy;
  • cytokines;
  • glucocorticoids;
  • interferon- γ;
  • thymus

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
  10. Supporting Information

Thymic atrophy is known to occur during infections; however, there is limited understanding of its causes and of the cross-talk between different pathways. This study investigates mechanisms involved in thymic atrophy during a model of oral infection by Salmonella enterica serovar Typhimurium (S. typhimurium). Significant death of CD4+ CD8+ thymocytes, but not of single-positive thymocytes or peripheral lymphocytes, is observed at later stages during infection with live, but not heat-killed, bacteria. The death of CD4+ CD8+ thymocytes is Fas-independent as shown by infection studies with lpr mice. However, apoptosis occurs with lowering of mitochondrial potential and higher caspase-3 activity. The amounts of cortisol, a glucocorticoid, and interferon-γ (IFN-γ), an inflammatory cytokine, increase upon infection. To investigate the functional roles of these molecules, studies were performed using Ifnγ−/− mice together with RU486, a glucocorticoid receptor antagonist. Treatment of C57BL/6 mice with RU486 does not affect colony-forming units (CFU), amounts of IFN-γ and mouse survival; however, there is partial rescue in thymocyte death. Upon infection, Ifnγ−/− mice display higher CFU and lower survival but more surviving thymocytes are recovered. However, there is no difference in cortisol amounts in C57BL/6 and Ifnγ−/− mice. Importantly, the number of CD4+ CD8+ thymocytes is significantly higher in Ifnγ−/− mice treated with RU486 along with lower caspase-3 activity and mitochondrial damage. Hence, endogenous glucocorticoid and IFN-γ-mediated pathways are parallel but synergize in an additive manner to induce death of CD4+ CD8+ thymocytes during S. typhimurium infection. The implications of this study for host responses during infection are discussed.


Abbreviations
S. typhimurium

Salmonella enterica serovar Typhimurium

CFU

colony-forming units

IFN-γ

interferon-γ

MLN

mesenteric lymph node

PE

phycoerythrin

PI

propidium iodide

SS

Salmonella-Shigella

TNF-α

tumour necrosis factor-α

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
  10. Supporting Information

The thymus is a critical organ of the immune system and is the primary site for the development and differentiation of T lymphocytes. The thymocyte population consists of cells at different stages of development and are broadly classified into immature (double-negative CD4 CD8 and double-positive CD4+ CD8+) and mature (single-positive CD4+ or CD8+) cells. Thymic atrophy is a recurrent theme during aging, stress and starvation and is characterized by a loss of thymic weight and cellularity. A fine balance between thymopoiesis, thymic output and thymic progenitor depletion is regulated through an intricate network of molecules, e.g. cytokines, chemokines, sex hormones, the ratio of Bcl2 (anti-apoptotic) to Bax (pro-apoptotic) and their role in mitochondrial function, the increase in fat deposits with age, etc.[1-7]

Thymic atrophy is a characteristic feature of many chronic and acute infections and the loss of thymic cellularity affects the subsequent replenishment of the peripheral lymphocyte populations.[4, 8] Therefore, a deeper understanding of the mechanisms involved is essential to design therapeutic strategies to block thymic atrophy and increase thymic output. A variety of mechanisms have been implicated during infection-induced thymic atrophy, including: (i) direct pathogen-induced atrophy due to excessive production of reactive oxygen species;[9] (ii) increase in glucocorticoid levels;[10-16] (iii) Fas-mediated signals;[17] (iv) modulation of Bcl2 family members;[14, 18, 19] (v) extracellular ATP-induced plasma membrane permeabilization in CD4+ CD8+ thymocytes;[20] and (vi) inflammatory cytokines.[13, 16, 19, 21-24]. Although various mechanisms of thymic atrophy during infection have been explored, there is no consensus on a common conclusive pathway. In addition, there is limited understanding of the interactions between the various pathways and the downstream signals involved.

Infections with intracellular bacterial pathogens, such as Francisella tularensis and Listeria monocytogenes, are known to lead to thymic atrophy.[10, 13] Different strains of Mycobacterium, a well-studied intracellular pathogen, have also been reported to cause thymic atrophy.[25, 26] We reported earlier that there is a drastic reduction in the size of the thymus upon intraperitoneal infection with the intracellular Gram-negative bacterial pathogen, Salmonella enterica serovar Typhimurium (S. typhimurium).[27] However, the mechanisms leading to thymic atrophy were not studied. Infection in humans results in a mild form of gastroenteritis and diarrhoea whereas infection of mice with S. typhimurium results in systemic, typhoid-like disease and mice die as a consequence of the colony-forming units (CFU) burden. This is similar to typhoid caused by Salmonella typhi in humans, making it an excellent mouse model for typhoid.[28, 29]

In this report we establish that S. typhimurium infection via the oral, i.e. the physiological, route of infection leads to significant death of CD4+ CD8+ thymocytes but not single-positive thymocytes or peripheral lymphocytes. Using this model of infection-induced thymic atrophy, we have used a small-molecule pharmacological compound (e.g. RU486) together with genetic approaches (lpr and Ifnγ−/− mice) to investigate the pathways involved in the death of CD4+ CD8+ thymocytes during S. typhimurium infection.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
  10. Supporting Information
Bacterial strains and inhibitors

Salmonella typhimurium NCTC 12023 was used for all oral infection experiments.[30] The bacterial culture was revived from glycerol stocks stored at −20° and streaked on SalmonellaShigella (SS) agar plates. A single colony from an SS agar plate was inoculated in Luria broth and cultured overnight. Fresh Luria broth (50 ml) was inoculated with this overnight culture at 0·2% and cultured for 3–4 hr (log-phase culture). Bacteria were centrifuged, washed twice in sterile PBS and resuspended in sterile PBS for further experiments. RU486 (Mifepristone, a progesterone and glucocorticoid receptor antagonist) was purchased from Sigma Aldrich Ltd (St Louis, MO), dissolved in DMSO and injected intraperitoneally at the indicated doses.

Mice and infection protocol

Mice were bred and maintained at the Central Animal Facility of IISc. 6- to 8-week-old C57BL/6, C57BL/6.Ifnγ−/− (Ifnγ−/−) and C57BL/6.lpr (lpr) mice were obtained from the Central Animal Facility of IISc. They were infected with an oral dose of ~ 108 CFU S. typhimurium in a total volume of 0·5 ml PBS. Each experimental group had between three and nine mice and each experiment was repeated at least thrice. All experiments were performed upon approval and in accordance with the guidelines stated by the Institutional Animal Ethics Committee, IISc.

CFU analysis and survival

On indicated days post infection, mice were euthanized and the tissue samples/organs were collected in PBS. The samples were weighed and homogenized in 1 ml PBS; dilutions were plated on SS agar plates. CFU were enumerated after incubation at 37° for 12–16 hr. Survival was monitored at 12-hr intervals for a period of 10–12 days after infection.

Cytokines and cortisol measurement

Blood collected by cardiac puncture immediately upon killing the mice, was allowed to clot, centrifuged and serum was separated and stored at −20°. Serum amounts of tumour necrosis factor-α (TNF-α) and interferon-γ (IFN-γ) were measured using ELISA kits (eBioscience, San Diego, CA) and cortisol was also measured by an AccuBind ELISA kit (Monobind, Inc., Lake Forest, CA) according to the manufacturers' instructions.

Isolation of lymph node cells and thymocytes

On indicated days post-infection, mice were sacrificed and mesenteric lymph nodes (MLN) and thymi were dissected. The tissues were washed in PBS, weighed and disrupted with a pair of forceps and fine wire mesh to prepare single cell suspensions in RPMI + 5% fetal calf serum.[27] Viable cell numbers were determined using Trypan blue and counting with a haemocytometer.

Flow cytometric analysis

Aliquots of cells were used for surface marker and intracellular signalling analysis on a BD Canto II fluorescence-activated cell sorter. Fluorescein isothiocyanite (FITC) -conjugated monoclonal antibodies against CD4 and phycoerythrin (PE) -conjugated antibodies against CD8 were obtained from eBioscience. Cells were incubated with fluorophore-tagged antibodies against surface markers, washed and resuspended in small volumes of 0·5% paraformaldehyde before acquiring FACS data. Live cell populations were gated on the basis of forward and side scatter. All FACS data were analysed using the BD Diva and WinMDI software.

Cell death and mitochondrial assays

Cell death was assayed using Annexin-V and propidium iodide (PI) kits from BD Biosciences (Frankland Lakes, NJ). Briefly, thymocytes and MLN cells were incubated with AnnexinV-FITC in binding buffer for 30 min on ice. The PI was added after washing cells in PBS and data were acquired on a BD Canto II. Caspase 3 activity colorimetric assay was performed according to the manufacturer's instructions using a kit from Sigma Aldrich Ltd. Mitochondrial status was analysed by incubating cells with 100 nm MitoTracker Red CMXROS (Invitrogen, Carlsbad, CA) at 37° for 30 min, washing cells and acquiring data on a BD Canto II. Cells with damaged mitochondria had less accumulation of MitoTracker Red CMXROS and these damaged populations were quantified.

Histological examination

Mesenteric lymph node and thymi were dissected, washed in PBS and fixed in 10% neutral buffer formalin. The tissues were processed and embedded in paraffin wax and sectioned. The sections were mounted on slides and stained with haematoxylin & eosin (H&E). The slides were observed under an inverted light microscope and 100 × magnified images were acquired on a Nikon camera attached to the microscope.

Statistical analysis

Data were analysed using commercially available software (graphpad prism 5; GraphPad Software Inc., La Jolla, CA) and statistically significant differences between various parameters were assessed by means of an unpaired two-tailed Student's t-test with 95% confidence intervals. Significance was represented using P-values as follows *< 0·05, **< 0·01, ***< 0·001.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
  10. Supporting Information

Salmonella typhimurium infection leads to depletion of CD4+ CD8+ thymocytes

To study the extent of thymic atrophy during infection, time–course experiments were performed with C57BL/6 mice orally infected with ~ 108 CFU S. typhimurium. The CFU in Peyer's patches, spleen, liver, thymus and MLN increased from day 2 to day 4 in all organs (Fig. 1a and data not shown). More CFU were recovered on day 4 in MLN than in thymus (Fig. 1a). Also there was a significant drop in the number of viable cells recovered from the thymi of infected mice on days 2 and 4 post-infection compared with uninfected mice. By day 4 there was a ~ 1·5-fold to 2-fold drop in the viable MLN cell numbers whereas there was a ~ 8-fold to 10-fold drop in viable thymocyte numbers (Fig. 1b). Additionally, upon infection of C57BL/6 mice with increasing CFU doses of S. typhimurium, there was a corresponding increase in the CFU recovered on day 4 from the MLNs and thymi (see Supplementary material, Fig. S1a) along with a dose-dependent effect on the survival of mice (Fig. S1b). Also, a CFU dose-dependent depletion in thymocyte numbers, but not MLN cell numbers, was observed upon S. typhimurium infection (Fig. S1c).

image

Figure 1. Salmonella typhimurium infection leads to severe reduction in CD4+ CD8+ thymocytes. (a) S. typhimurium colony-forming units (CFU) in mesenteric lymph nodes (MLN) and thymi were analysed on day 2 and day 4 after oral infection with ~ 108 bacteria. (b) Viable cell numbers from MLN and thymi of uninfected and infected mice on day 2 and day 4 after oral infection were analysed using Trypan Blue exclusion assay. Changes in (c) MLN and (d) thymocyte sub-populations upon infection were assessed by FACS on the basis of surface CD4 and CD8 expression. Each group contained five to eight mice and the experiments were repeated at least thrice. Student's t-test was used for statistical analysis with **P < 0·01, ***< 0·001.

Download figure to PowerPoint

Importantly, when mice were fed with 108 CFU live and 109 and 1010 CFU heat-killed S. typhimurium orally there were no significant differences in the cell numbers recovered from MLN. On the other hand, the thymocyte numbers decreased significantly in the group infected with live S. typhimurium, whereas, administering 10 or 100 times more heat-killed bacteria to the mice did not decrease the number of viable thymocytes (see Supplementary material, Fig. S2b). Also, when thymocytes from uninfected mice were cultured in vitro with 102, 104 or 106 CFU live or 106 CFU heat-killed S. typhimurium, no significant changes in viable cell numbers were observed compared with cells alone, even after 36 hr of culture (Fig. S2c). Therefore, it is unlikely that the direct interactions of S. typhimurium with thymocytes lead to thymic atrophy.

Next, surface phenotype analysis was performed but no significant changes in the percentage or total cell numbers of CD8+, CD4+ and CD4 CD8 cell populations in the MLN were observed upon infection on days 2 and 4 by FACS analysis (Fig. 1c). However, the proportion of CD4+ CD8+ thymocytes decreased significantly, by approximately eightfold, in the thymus by day 4 with a concomitant increase in the percentage of CD4 CD8+ and CD4+ CD8 sub-populations (Fig. 1d). Upon calculating the number of thymocytes in each sub-population on the basis of the percentage data and the total viable cell number obtained per organ, no significant changes in the actual numbers of CD4 CD8+ and CD4+ CD8 cells were observed. The apparent increase in the percentage of CD4 CD8+ and CD4+ CD8 cells in the FACS plots was a consequence of the decrease in the numbers of CD4+ CD8+ thymocytes. These results indicate that immature CD4+ CD8+ thymocytes were the most susceptible population, compared with other subsets, during in vivo oral infection of mice with live S. typhimurium.

Thymocyte depletion during S. typhimurium infection occurs via apoptotic processes

To assess whether the depletion of CD4+ CD8+ cells was the result of enhanced apoptotic processes upon infection, the extent of Annexin-V-FITC staining and PI uptake of the cells were analysed. There were no significant differences on days 2 and 4 in the apoptotic populations in the MLN upon infection (Fig. 2a). However, in the thymus, the Annexin-V-positive population increased approximately twofold while the Annexin-V-positive and PI-positive population increased approximately threefold upon infection, indicating that thymocytes from infected mice have more apoptotic cells (Fig. 2a).

image

Figure 2. Thymocytes, upon infection, show increase in apoptosis markers, mitochondrial damage and caspase 3 activity. (a) Apoptotic single cell populations from thymus and mesenteric lymph nodes (MLN) of control and infected mice were assessed by Annexin V and propidium iodide (PI) staining. (b) Mitochondrial damage in MLN and thymic cells from control and infected mice was quantified using MitoTracker Red. (c) Caspase-3 activity in MLN and thymic cell lysates was measured using a colorimetric assay. Each group had five mice and the experiments were repeated at least thrice. Student's t-test was used for statistical analysis with **< 0·01, ***< 0·001.

Download figure to PowerPoint

A feature of the apoptotic process is the substantial damage to mitochondrial integrity, leading to loss of the mitochondrial transmembrane potential and less accumulation of MitoTracker Red CMXRos.[31, 32] The MLN cell populations did not show any significant differences in the mitochondrial status upon infection; however, thymocytes with low mitochondrial potential were significantly more by day 4 post-infection compared with control or day 2 post-infection (Fig. 2b). This indicates that during infection, the mitochondrial membrane potential of thymocytes is lowered. Also, effector caspase-3 activation was significantly higher in the thymus than in the MLN by 4 days after infection (Fig. 2c). These results together indicate that the depletion of CD4+ CD8+ thymocytes during S. typhimurium infection occurs via apoptotic mechanisms involving membrane depolarization and caspase-3 activation.

Death of CD4+ CD8+ thymocytes during infection is Fas independent

Upon infection of C57BL/6 mice with S. typhimurium, up-regulation of Fas and Fas ligand (FasL) was observed on thymocytes (Fig. 3a). Since lpr mice are deficient in the expression of Fas protein, the role of extrinsic death signals via Fas/Fas L interactions was investigated. As expected, thymocytes from lpr mice did not express Fas, but induction of FasL was observed upon infection (Fig. 3a). Upon investigating the extent of thymic atrophy in the lpr mice upon S. typhimurium infection, no differences in CFU and depletion of thymocytes were observed in these mice compared with C57BL/6 mice (Fig. 3b,c). Also, as the extent of CD4+ CD8+ thymocyte death was comparable in C57BL/6 and lpr mice upon infection, it was concluded that the apoptotic thymocyte death process was not Fas/FasL dependent (Fig. 3d,e).

image

Figure 3. Fas–Fas ligand (FasL) interactions are not involved in Salmonella typhimurium infection-induced thymic atrophy. C57BL/6 and lpr mice were orally infected with ~ 108 colony-forming units (CFU) and euthanized 4 days after infection. (a) Thymocytes from control and infected C57BL/6 and lpr mice were stained for surface expression of Fas and FasL. (b) CFU and (c) cellularity of mesenteric lymph nodes (MLN) and thymi were compared in these mice upon infection. (d) Extent of apoptosis in control and infected groups was assessed via Annexin V and propidium iodide (PI) staining. (e) Changes in thymic sub-populations in both C57BL/6 and lpr animals were monitored on the basis of CD4 and CD8 expression. Each group had four to six mice and the experiments were repeated thrice. Student's t-test was used for statistical analysis with *< 0·1, **< 0·01, ***< 0·001.

Download figure to PowerPoint

Elevated endogenous corticosteroids play a partial role during thymic atrophy

Lymphocytes and thymocytes are known to be susceptible to glucocorticoid-mediated apoptotic signals.[33] To understand whether endogenous corticosteroids contribute to thymic atrophy upon S. typhimurium infection in mice, the amounts of cortisol in sera of mice were quantified. Cortisol amounts in sera were found to be significantly increased from day 2 to day 4 post-infection compared with control mice (Fig. 4a). RU486 (25 mg/kg), which binds to the glucocorticoid receptor and prevents downstream glucocorticoid-mediated signalling was administered intraperitoneally to infected mice 12 hr after infection. Administration of the vehicle alone to infected mice had no effect on infection progression or the reduction of thymocyte numbers (data not shown). No differences in cortisol amounts in sera and CFU in MLN and thymus were observed, but the drop in viable thymocyte numbers in the RU486-treated mice was only about threefold to fourfold compared with the approximately eightfold to tenfold drop in vehicle-treated mice (Fig. 4b–d). The administration of RU486 did not affect the decrease in MLN cell numbers and the survival of mice upon infection (Fig. 4d; see Supplementary material, Fig. S3c). Importantly, the extent of rescue in thymocyte numbers was not enhanced using higher doses of RU486, (Fig. S3d), by pre-treatment or by multiple doses, e.g. pre-treatment and post-treatment (Fig. S3e). These data indicate that corticosteroid signalling is partially responsible for the S. typhimurium-induced thymic atrophy.

image

Figure 4. Inhibition of endogenous corticosteroid signalling results in partial rescue of thymic atrophy. (a) Serum cortisol amounts were estimated by ELISA upon oral infection with Salmonella typhimurium and (b) treatment of infected mice with vehicle or RU486 (25 mg/kg). (c) Colony-forming units (CFU) and (d) viable cell numbers from the mesenteric lymph nodes (MLN) and thymi of control and infected mice were analysed. Each group had five mice and the experiments were repeated at least thrice. Student's t-test was used for statistical analysis with **< 0·01, ***< 0·001.

Download figure to PowerPoint

Elevated endogenous IFN-γ contributes to death of CD4+ CD8+ thymocytes

Infections are known to elicit inflammatory responses from host, including production of various cytokines.[29, 30, 34] The amounts of TNF-α and IFN-γ in sera of infected mice were found to be elevated from day 2 to day 4 compared with control (Fig. 5a). To address the role of an inflammatory cytokine, e.g. IFN-γ, experiments with C57BL/6 and mice lacking IFN-γ, i.e. Ifnγ−/− mice, were performed. Upon infection of these mice with S. typhimurium, higher CFU was recovered from Ifnγ−/− mice on day 4 post-infection (Fig. 5b). Also, the Ifnγ−/− mice died by 4–5 days, whereas C57BL/6 mice died by 7 days (see Supplementary material, Fig. S4). Despite the higher CFU and susceptibility of Ifnγ−/− mice, there was a small and significant rescue in the number of viable thymocytes recovered from Ifnγ−/− infected mice compared with the C57BL/6-infected mice (Fig. 5c).

image

Figure 5. Roles of host encoded interferon-γ (IFN-γ) in thymic atrophy. (a) Tumour necrosis factor-α (TNF-α) and IFN-γ amounts in the sera of uninfected and S. typhimurium-infected C57BL/6 mice were estimated by ELISA on days 2 and 4. (b) Colony-forming units (CFU) and (c) viable cell numbers from the mesenteric lymph nodes (MLN) and the thymi of infected C57BL/6 and Ifnγ−/− mice were analysed in the control and infected groups. Each group had five to eight mice and the experiments were repeated at least thrice. Student's t-test was used for statistical analysis with *< 0·05, **< 0·01, ***< 0·001.

Download figure to PowerPoint

Synergy between IFN-γ and corticosteroids leads to death of CD4+ CD8+ thymocytes

Since the inhibition of glucocorticoid-mediated signals and the lack of Ifnγ lead to a partial rescue in the extent of thymic atrophy upon S. typhimurium infection, it was reasonable to assess the inter-relationship between the two pathways. To this end, RU486 (25 mg/kg) was administered to C57BL/6 and Ifnγ−/− mice post-infection. Serum cortisol amounts were elevated in both strains of mice upon infection; however, administration of RU486 did not modulate the elevated cortisol amounts (Fig. 6a). Also, amounts of TNF-α and IFN-γ were elevated in C57BL/6 mice upon infection but RU486 did not affect the increased cytokine amounts (Fig. 6b,c). Interestingly, in infected Ifnγ−/− mice, with or without RU486 treatment, the amounts of TNF-α in sera were significantly lower compared with the C57BL/6 counterparts (Fig. 6b). Higher CFU was recovered from the Ifnγ−/− mice upon infection compared with C57BL/6 and the administration of RU486 to either group did not affect the CFU burden (Fig. 7a), neither did it affect the survival of either strain of infected mice (Fig. 7c). Also, the number of viable MLN cells was comparable in C57BL/6 and Ifnγ−/− mice and RU486 treatment did not rescue the small drop in numbers (Fig. 7b). Importantly, the decrease in number of total and CD4+ CD8+ thymocytes observed upon RU486 administration to infected Ifnγ−/− mice was ~ 1·5-fold, compared with the three-fold to fourfold drop in C57BL/6 mice given RU486 or vehicle-treated, infected Ifnγ−/− mice (Fig. 7b,d).

image

Figure 6. Interferon-γ (IFN-γ) and glucocorticoid-mediated pathways are independent. C57BL/6 and Ifnγ−/− mice were infected orally with Salmonella typhimurium and treated with RU486 (25 mg/kg). Animals were euthanized on day 4 post-infection. (a) Cortisol, (b) tumour necrosis factor-α (TNF-α) and (c) IFN-γ amounts in sera of control and infected mice were quantified using ELISA. Each group had five to eight mice and the experiments were repeated at least thrice. Student's t-test was used for statistical analysis with *< 0·05, **< 0·01, ***< 0·001.

Download figure to PowerPoint

image

Figure 7. Synergy between interferon-γ (IFN-γ) and glucocorticoids mediates the reduction in CD4+ CD8+ thymocytes during Salmonella typhimurium infection. C57BL/6 and Ifnγ−/− mice were infected orally with S. typhimurium and treated with RU486 (25 mg/kg). Animals were sacrificed on day 4 post-infection. (a) Colony-forming units (CFU) and (b) viable cell numbers from the mesenteric lymph nodes (MLN) and thymi of control, infected and infected plus RU486-treated C57BL/6 and Ifnγ−/− mice were estimated using Trypan blue. (c) Survival of mice in each of these groups was monitored. (d) Phenotypic characterization of subpopulations was performed on the basis of CD4 and CD8 expression. Each group contained five to eight mice and the experiments were repeated at least thrice. Student's t-test was used for statistical analysis with *< 0·05, **< 0·01, ***< 0·001.

Download figure to PowerPoint

This increase in the recovery of viable thymocytes from Ifnγ−/− mice treated with RU486 correlated with the decrease in Annexin-V-positive apoptotic populations (Fig. 8a). Also, the extent of mitochondrial damage upon infection was less in the Ifnγ−/− mice treated with RU486 compared with RU486 administration in C57BL/6 or infection of Ifnγ−/− mice alone (Fig. 8b). Similarly, the amounts of activated caspase-3, upon infection, in thymi of Ifnγ−/− mice treated with RU486 were comparable to those in the thymi of control uninfected animals (Fig. 8c). Hence, pathways mediated by endogenous glucocorticoids and inflammatory cytokines, such as IFN-γ, synergize to enhance death of immature thymocytes during S. typhimurium infection in an additive manner.

image

Figure 8. Upon infection, glucocorticoids and interferon-γ (IFN-γ) enhance apoptosis, mitochondrial damage and caspase-3 activity in the thymus. C57BL/6 and Ifnγ−/− mice were infected with Salmonella typhimurium and treated with vehicle or RU486. Mice from control and infected groups were euthanized on day 4 post-infection. (a) Apoptotic mesenteric lymph node (MLN) and thymic cell populations were analysed using Annexin V by FACS. (b) Mitochondrial status of the MLN and thymic cells was assessed via uptake of MitoTracker Red. (c) Caspase-3 activity in lysates of MLN and thymi was quantified using colorimetric assay. Each group contained five to eight mice and the experiments were repeated at least thrice. Student's t-test was used for statistical analysis with **< 0·01, ***< 0·001.

Download figure to PowerPoint

Histological analysis of thymic tissues upon infection

Thymic architecture is distinctive with a densely populated cortex that stains darker than the less populated medulla. The cortex consists of the immature thymocytes whereas the medulla harbours the more mature thymocytes that are poised to be sent out into the periphery. The H&E-stained thymic sections from different groups of control, infected and RU486-treated C57BL/6 and Ifnγ−/− mice were observed. The thymi of control C57BL/6 and Ifnγ−/− mice showed the characteristic denser medullar and sparse staining cortical distinction (see Supplementary material, Fig. S5a,d). Upon infection, the pattern of staining in the infected C57BL/6 thymus was altered with dramatic loss of cells from the cortex, but not the medulla, as indicated by the change in intensity of the staining (Fig. S5b). Upon RU486 treatment in infected C57BL/6 mice, the architectural integrity was not retained but the loss in cellularity was less when compared with the infected thymi (Fig. S5c). Interestingly, the loss in cellularity in infected Ifnγ−/− mice was less compared with the C57BL/6 infected mice and the rescue in tissue damage was enhanced upon RU486 administration to Ifnγ−/− mice (Fig. S5e,f).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
  10. Supporting Information

Thymic atrophy has been reported during infections by different pathogens and the mechanisms involved vary depending on the pathogen and the infection process.[8] Thymic atrophy occurs transiently during infections with mouse hepatitis virus and simian immunodeficiency virus infection of macaques.[18, 35] During cases of prolonged chronic infection, thymic atrophy is observed in infected individuals that survive for a long period of time. In such cases, thymic atrophy may lead to the generation of lower numbers of mature T cells, which may hamper the host's ability to respond to immune challenges. A classic example is patients with HIV infection, who demonstrate low thymic function with regard to generation of new thymic emigrants.[36] The observation that disturbed thymic architecture and thymocyte death are a consequence of infection with HIV suggests that interventions in this pathway may be important in ensuring better therapeutic efficacy.[37] Importantly, proper CD4+ reconstitution may not occur in AIDS patients undergoing highly active anti-retroviral therapy in the absence of a proper functional thymus. Concurrently, thymic function has been shown to be higher in long-term survivors of HIV infection.[38-42] These studies clearly reinforce the importance of studying infection-induced thymic death as it has clear implications for health and disease.

Acute thymic atrophy occurs during several infections but the underlying mechanisms are not well understood.[10, 11, 13, 27, 43] Active infection by live and virulent microorganisms is important for the development of thymic atrophy.[13, 44] In fact, thymic atrophy is dependent upon a critical number of live pathogens and lowering of infection using anti-bacterial agents lowers thymic atrophy.[27, 45] Anti-retroviral therapy restores the proportion of CD4+ CD8+ thymocyte populations and increases the numbers of ‘TCR excision circles-positive’ lymphocytes in the periphery demonstrating an increase in thymic function.[46] In this study, a model for infection-induced thymic atrophy was developed using oral dosing of mice with a high dose of live S. typhimurium, an intracellular bacterial pathogen. Infection with lower doses of S. typhimurium did not result in the death of immature thymocytes (Fig. S1). Therefore, higher CFU dosing with live bacteria was required in this acute model of S. typhimurium infection of mice and thymic atrophy occurred at a later stage (day 4) of infection by which time the mice were heavily infected and eventually died (day 5–9) (Fig. 1b and Fig. S1b). Heat-killed bacteria were unable to induce thymic atrophy, probably because of their inability to enter the peripheral circulation and elicit a strong host response (Fig. S2b). Additionally, thymocyte depletion was observed only upon in vivo infection of mice but not when thymocytes were exposed in vitro to high numbers of S. typhimurium (Fig. S2c). These results indicate that the infection-induced thymic atrophy was not a result of the direct interactions of S. typhimurium and thymocytes. Most likely, S. typhimurium multiplication occurs within macrophages and dendritic cells within immune organs, including the thymus, and the thymocyte death was a consequence of multiple host factors.

In most cases of infection-induced thymic atrophy, the major population that is depleted consists of CD4+ CD8+ thymocytes;[11, 13, 15, 16, 23] however, infection by Plasmodium berghei causes depletion in CD4 CD8, CD4+ CD8, CD4 CD8+ and CD4+ CD8+ populations.[47] Studies with Trypanosoma cruzi and Plasmodium berghei have reported the presence of higher numbers of CD4+ CD8+ lymphocytes in the periphery, probably as a result of increased migration of thymocytes;[43, 48] however, this was not seen in the present study. Also, the peripheral lymphocyte pool is highly susceptible to death during infections with simian immunodeficiency virus and other pathogens.[9, 11, 18] In this model of S. typhimurium oral infection, there was a marginal decrease in total MLN cell numbers (Fig. 1b). Importantly, the small depletion in MLN cell numbers was not accompanied by phenotypic changes with respect to CD4 and CD8 expression (Fig. 1c) so the possibility that the depletion of CD4+ CD8+ cells from the thymus was the result of enhanced egress of immature or mature thymocytes to the periphery upon infection could be ruled out. Hence, on oral infection of mice with S. typhimurium, the immature CD4+ CD8+ thymocyte population is preferentially depleted eightfold to tenfold within a period of 4–5 days as a consequence of the infection process (Fig. 1d).

An important death pathway through which peripheral lymphocytes and thymocytes receive apoptotic signals is the Fas–FasL pathway. Though the role of Fas-mediated death signals during age-related thymic atrophy is known,[49] its involvement in infection-induced thymic atrophy is unclear. Experiments with infection of gld mice with Trypanosoma cruzi and Francisella tularensis indicate that Fas signalling is not involved in thymic atrophy[13, 50] whereas during HIV infection, both Bcl2 and Fas pathways contribute to lymphocyte and thymocyte death.[17] Though up-regulation of Fas and FasL occurred upon infection, this pathway was not involved in the S. typhimurium-induced thymic atrophy (Fig. 3). In the thymocytes recovered from infected mice, the mitochondria were significantly damaged, as indicated by their loss of membrane potential (Fig. 2b). Together, these data indicate that the intrinsic apoptotic pathway may be activated upon S. typhimurium oral infection of mice leading to death of immature thymocytes.

Glucocorticoids are anti-inflammatory and immunosuppressive steroid hormones that are released from the adrenal glands and are capable of inducing apoptosis in T and B lymphocytes and thymocytes.[33] Classically, the roles of glucocorticoids have been studied by means of adrenalectomy or administration of the synthetic steroid, RU486 (generic name, Mifepristone), a progesterone and glucocorticoid receptor antagonist. RU486 is a synthetic steroid analogue whose effects are mediated through competitive binding to the cytoplasmic glucocorticoid receptor.[51] RU486 is a potent anti-glucocorticoid both in vitro and in vivo, with a binding affinity equal to or exceeding that of the naturally occurring glucocorticoid agonists in rodents and humans. RU486-bound receptors undergo a trans-conformation in the ligand-binding domain and are unable to undergo intracellular trafficking, which has been proposed as a possible reason for the anti-glucocorticoid activities of the molecule.[52, 53] Increase in adrenal corticosteroid secretion leads to thymocyte death in several infection models, such as the early phase of Listeria monocytogenes infection,[10] lipopolysaccharide-treated mice[12] and high-dose intraperitoneal challenge with type A Francisella tularensis.[11, 13] Also, higher glucocorticoid amounts play a partial role in triggering apoptosis of thymocytes during a model of polymicrobial sepsis.[14] On the other hand, high amounts of circulating cortisone are detected in acute Trypanosoma cruzi infection but inhibition by RU486 or adrenalectomy increased the cortisol upon infection and only partially rescued the development of thymic atrophy.[15, 16] However, CD4+ CD8+ thymocyte death caused by concanavalin A[19] or infection by mouse hepatitis virus[35] is not dependent on glucocorticoids. Given these discrepancies in literature, the in vivo role of corticosteroids during S. typhimurium-infection-induced thymic atrophy was assessed. Indeed, the levels of cortisol in the sera of infected animals were higher than in control animals (Fig. 4a). Administration of RU486 did not affect CFU or survival of mice upon infection and a reduction in the drop in thymocyte numbers was observed. Notably, administration of higher saturating doses of RU486 did not augment the rescue, indicating that glucocorticoid-mediated pathways are only partially involved in the observed thymic atrophy (Fig. 4 and Fig. S3).

Most infections elicit a distinct inflammatory response from the host immune system along with the production of various cytokines. Thymocytes are highly sensitive to inflammatory signals and both TNF-α and IFN-γ have been shown to lead to thymocyte apoptosis due to endotoxic shock or sepsis.[21, 22, 24] Intraperitoneal injection of bacteria induces a sepsis-like condition that causes thymic atrophy, which is dependent on TNF-α.[23] Also, in concanavalin A-mediated thymocyte atrophy, extra-thymic IFN-γ and TNF-α together mediate the death of thymocytes.[19] Parasitaemia due to Trypanosoma cruzi, together with serum cortisol and cytokine amounts, is greater in mice lacking both receptors of TNF-α. Consequently, the survival of these mice is reduced but there is no effect on thymic atrophy.[16] On the other hand, thymic atrophy was greatly reduced in mice lacking both the receptors of TNF-α upon infection with Francisella tularensis.[13] Interferon-γ has been shown to be important in restricting S. typhimurium infection during the early phase[54] and Ifnγ−/− mice are hyper-susceptible to infection by less virulent strains of S. typhimurium.[30] In this study, serum amounts of TNF-α and IFN-γ were significantly elevated in C57BL/6 mice upon infection (Fig. 5a). Notably, infected Ifnγ−/− mice had lower serum amounts of TNF-α (Fig. 6b), which is consistent with a previous study.[30] More thymocytes were recovered from Ifnγ−/− mice infected with S. typhimurium despite the fact that Ifnγ−/− mice had higher CFU and died earlier compared with C57BL/6-infected mice (Fig. 5 and Fig. S4). Hence, IFN-γ enhances inflammatory cytokines and perhaps other inflammatory mediators thereby reducing CFU and enhancing survival of mice upon infection with S. typhimurium. However, IFN-γ-mediated inflammatory mediators are likely to enhance thymocyte death. This study clearly demonstrates the pleiotropic effects of IFN-γ with respect to in vivo host responses to S. typhimurium.

As elevated amounts of both endogenous glucocorticoids and IFN-γ contributed partially to the depletion of CD4+ CD8+ thymocytes during S. typhimurium infection of mice, it was of interest to identify the interactions, if any, between these pathways. This aspect is important to understand the relationship between these two pathways because one is an immunosuppressive pathway whereas the other leads to inflammation. Interestingly, glucocorticoids have been shown to reduce the amounts of TNF-α and interleukin-6 and increase the survival of mice during Trypanosoma cruzi infection.[15, 16] Although the role of glucocorticoids in thymic death is well known their effects on host responses during S. typhimurium infection are poorly studied. Despite the fact that IFN-γ has been shown to be important for resistance to intracellular pathogens,[34, 55-57] its role in infection-induced thymic atrophy is not well understood. The cortisol amounts in C57BL/6 and Ifnγ−/− mice were comparable whereas the cytokine amounts were not modulated by RU486 administration (Fig. 6). Also, RU486 treatment did not affect the survival of C57BL/6 or Ifnγ−/− mice (Fig. 7c). Upon treatment of infected C57BL/6 and Ifnγ−/− mice with RU486, the rescue in CD4+ CD8+ thymocyte numbers was significantly higher in Ifnγ−/− mice treated with RU486 (Fig. 7d). Thymocytes from infected Ifnγ−/− mice treated with RU486 had significantly lower apoptotic populations, caspase-3 activity and less mitochondrial damage compared with all other infected groups (Fig. 8). Also, there was lesser tissue damage in this group, indicating that thymic architecture was better preserved (Fig. S5). The results presented here indicate that glucocorticoids do not affect the S. typhimurium infection yet contribute to thymocyte death. On the other hand, IFN-γ has a protective effect during S. typhimurium infection and, importantly, is partially responsible for the thymic atrophy observed. The biological relevance of this study is underscored by the demonstration of the functional roles of the glucocorticoid and IFN-γ-induced pathways, which are parallel but synergize to enhance the death of CD4+ CD8+ thymocytes during S. typhimurium infection (Fig. 9).

image

Figure 9. Schematic representation of the mechanisms involved in Salmonella typhimurium-induced CD4+ CD8+ thymocyte death. The two major pathways that are involved in the death of CD4+ CD8+ thymocytes during S. typhimurium infection are the endogenous glucocorticoids and interferon-γ (IFN-γ) -mediated responses. These are parallel pathways that synergize, in an additive manner, to reduce the survival of CD4+ CD8+ thymocytes during infection. Also, the likely involvement of other unidentified pathways is indicated in this model.

Download figure to PowerPoint

The S. typhimurium infection model used for this study is an acute and lethal mouse model and thymic atrophy was observed to occur as a consequence of high amounts of CFU. Here, the extent of thymic atrophy was de-linked from the survival of mice, as revealed by studies using RU486 and Ifnγ−/− mice. This study has demonstrated the efficacy of this thymic atrophy model to understanding host responses during infections. Overall, the mechanistic details revealed here suggest a broad framework to understand the roles of different pathways and mediators during thymic atrophy. Importantly, some host responses elicited upon infection and leading to thymic atrophy with different pathogens may be conserved.[58] Therefore, studies such as this have important implications and may lead to the development of better strategies to block or reduce thymic atrophy during other infections (e.g. HIV), thus leading to enhanced host cellular defence networks.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
  10. Supporting Information

We thank the Central Animal Facility, IISc for the supply of mice for experiments. The Divisional FACS facility, IISc has been most patient and helpful in acquiring samples. The support and encouragement of all members of the DpN laboratory are greatly appreciated. M.D.L. was supported by the SPM fellowship, CSIR, Government of India. This study was funded by grants from the Department of Biotechnology, Government of India.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
imm12047-sup-0001-FigS1-S5.pdfapplication/PDF655K

Figure S1. Thymic atrophy increases with dose of infection.

Figure S2. Thymocyte depletion occurs during in vivo infection with live Salmonella typhimurium.

Figure S3. Glucocorticoids play a partial role during Salmonella typhimurium infection-induced thymic atrophy.

Figure S4. Ifnγ−/− mice are more susceptible to infection than C57BL/6 mice.

Figure S5. Histological analysis of thymi from C57BL/6 and Ifnγ−/− upon Salmonella typhimurium infection.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.