| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Published online before print December 6, 2007
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Department of Immunology,*Lerner Research Institute, Cleveland Clinic, Cleveland; and the Department of Pathologyand the Cleveland Clinic Lerner College of Medicine,Case Western Reserve University, Cleveland, Ohio
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
-myosin heavy chain preferentially develop a predominant Th17 lineage response that provides sustained T-cell memory and a high DCM incidence whereas females preferentially develop a predominant Th1 lineage response that becomes anergized to cardiac self resulting in compensatory protection against DCM. The distinct sex-defined disease phenotypes are interchangeable after in vivo manipulation of Th1 (interleukin-2) and Th17 (interleukin-17) cytokines. Our study shows that male and female SWXJ mice differentially respond to cardiac self in ways that lead to distinct autoimmune outcomes and implies that optimized therapy for autoimmunity may require consideration of the qualitatively different ways that males and females engage self.
The basis for female-biased incidence in multiple sclerosis, systemic lupus erythematosus, and rheumatoid arthritis is currently unclear and is particularly confounded by observations that both estrogen and testosterone treatment may be therapeutic.3-5
A common explanation for female-biased autoimmune susceptibility contends that it is attributable to a more aggressive proinflammatory Th1-like immune response generated by females in response to self.1,2
This view is supported by studies correlating increased frequencies of autoreactive interferon (IFN)-
-producing T cells and decreased frequencies of CD4+CD25+FoxP3+ regulatory T cells (Treg) in females with multiple sclerosis and in female mice with experimental autoimmune encephalomyelitis, a widely used animal model for multiple sclerosis.6
However, this view does not apply to myocarditis because male mice develop more severe viral-induced myocarditis7,8
and are often used exclusively in studies involving autoimmune-mediated myocarditis.9,10
In this context, we chose to investigate the immunological features surrounding the male-biased incidence seen in human DCM11
using our recently-developed murine model of experimental autoimmune myocarditis (EAM).12
After adoptive transfer of cardiac -myosin heavy chain (CAM 406-425) peptide-specific CD4+ T cells, SWXJ mice develop EAM with males showing more pronounced complications of myocardial inflammation than females. Male-associated disease is characterized by extensive fibrosis and ventricular dilatation, indicative of cardiac decompensation and DCM.13
Here we show that the increased incidence of DCM in male mice is attributable to persistence of autoimmune T-cell memory and that protection from development of DCM in female mice is attributable to clonal anergy of their autoreactive T cells. We found that the female cardio-specific repertoire is intrinsically myocarditogenic because in vivo potentiation of either interleukin (IL)-17 or IL-2 in females results in the male-like severe DCM phenotype. Likewise, complementary experiments in males involving inhibition of either IL-17 or IL-2 results in anergy and expression of the female-like DCM protective phenotype. We found that programming for anergy in female myocarditogenic T cells occurs during self-priming and is independent of any subsequent sex-specific end-organ modulation of the effector response. Our study shows that males and females differentially engage cardiac self in a manner that leads to qualitatively different autoreactivities and disease outcomes. Our results imply that optimal therapeutic intervention in autoimmune disease may require a sex-based perspective that considers the qualitatively different ways that males and females respond to self.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Female SWXJ.Thy-1b/b (H-2q,s) mice expressing the conventional Thy1.2 isoform of CD90 were purchased from the Jackson Laboratory (Bar Harbor, ME). SWXJ.Thy-1a/b mice expressing both Thy1.1 and Thy1.2 were generated by mating male SJL/J.Thy-1a/a (H-2s) mice (generously provided by Dr. Harley Tse, Wayne State University, Detroit, MI) with female SWR/J.Thy-1b/b (H-2q) mice. As described previously12 for active induction of EAM or for generating primed lymph node cells (LNCs), 6- to 8-week-old SWXJ.Thy-1a/b mice were injected subcutaneously in the abdominal flank on day 0 with 200 µg of CAM 406-425 (KVGNEYVTKGQSVQQVYYSI) and 400 µg of Mycobacteria tuberculosis H37RA (Difco, Detroit, MI) in 200 µl of an emulsion of equal volumes of water and IFA (Difco).
Passive Transfer of EAM
LNCs from SWXJ.Thy-1a/b mice were isolated 10 days after immunization with CAM 406-425 and cultured in 24-well flat-bottomed plates (BD Biosciences, San Jose, CA) at 6 x 106 cells/well in Dulbecco’s modified Eagle’s medium (Mediatech, Herndon, VA) supplemented with 10% fetal bovine serum (Hyclone, Logan, UT), 5% HEPES buffer, 2% L-glutamine, 1% penicillin/streptomycin (Life Technologies, Inc., Grand Island, NY), and 20 µg/ml of CAM 406-425 in a total volume of 2.0 ml. After 4 days, 2 x 107 washed cells were injected intraperitoneally into naive -irradiated (600 rads) SWXJ.Thy-1b/b recipients. Established long-term T-cell lines were derived from peptide-primed Thy1.1-expressing LNCs by alternate cycles of peptide stimulation and the IL-2-supplemented rest on syngeneic -irradiated (2000 rads) splenocyte monolayers. After three rounds of stimulation, T-cell lines were transferred at 2 x 107 cells/mouse into -irradiated (600 rads) Thy1.1– hosts. In experiments involving sex-mismatched passive transfer of EAM, CD8+ T cells were depleted with three 500-µg injections of equal amounts of two different CD8 antibodies (clones YTS169 and T IB105; Ligocyte, Bozeman, MT) every other day starting 3 days before transfer and every 4th day thereafter for 5 weeks. CD8+ T cells represented 16.7 ± 3.3% and 0.76 ± 0.20% of total splenocytes in undepleted and depleted mice, respectively.
Histomorphological Determination of DCM
DCM is traditionally defined by functional criteria involving diminution of cardiac function. In our initial report describing the specific mouse model used in the current study, we showed that cardiac function was severely compromised and that these functional defects involved pathologically altered systolic pressures and ejection fractions as evidenced by pressure recordings from right carotid artery catheterization experiments.14 We also reported that the histomorphological triad of 1) increased heart weight (mg):body weight (g) ratio (HW:BW), 2) cardiac fibrosis, and 3) ventricular dilatation in mice with autoimmune myocarditis invariably accompanied these observed functional abnormalities and as such were pathognomonic for DCM. In our current study, each heart had to meet this rigid triad of criteria for determination of DCM. Moreover, cardiac histological consistency was ensured because all hearts were arrested during diastole as a result of injection of 2 mol/L KCl infusion during euthanasia.
To prevent postmortem clotting, mice were injected intraperitoneally with 100 IU heparin (Elkins-Sinn, Cherry Hill, NJ) 15 minutes before euthanasia. Removed hearts were perfused with phosphate-buffered saline (PBS) and weighed for determining HW:BW. Hearts were fixed in phosphate-buffered formalin overnight and embedded in paraffin. Multiple serial 5-µm cross-sections were obtained from the apex of the heart for cross-sectional analysis of left and right ventricles. Sections were stained with either hematoxylin and eosin or Masson’s Trichrome and examined by light microscopy for the presence of inflammation and fibrosis. The percentage of heart tissue involved in inflammation was determined as described previously.12 DCM was determined by the presence of fibrosis, ventricular dilatation involving an increase in the radius of one or both ventricular chambers without a proportionate increase in mural wall thickness, and an increase in HW:BW. All sections were evaluated in a blinded manner.
T-Cell Assays
Thy1.1+ T cells were purified >95% from recipient spleens by magnetic bead separation on a MidiMACS cell separator using anti-Thy1.1 microbeads (Miltenyi, Auburn, CA) or by FACSVantage (BD Biosciences) cell sorting with anti-Thy1.1-FITC (BD Biosciences). Thy1.1+ T cells (2 x 105)/well were cultured on a monolayer of 5 x 105 -irradiated (2000 rads) sex-matched splenocytes/well in 96-well flat-bottomed microtiter plates (BD Biosciences) in media supplemented as described above in the presence or absence of 100 µg/ml of CAM 406-425.
Enzyme-linked immunosorbent assays (ELISAs) for IFN-, IL-2, IL-4, IL-5, and IL-10 were performed on 48-hour culture supernatants using commercially-available capture/detection antibody pairs as previously described.12
ELISAs were also performed on culture supernatants from T cells removed from heart tissues enzymatically digested by agitation with collagenase type IV (Sigma, St. Louis, MO) in Hanks’ balanced salt solution (5 mg/ml) for 30 minutes at 37°C. The released lymphocytes in suspension were then passed through a cell strainer, washed, and cultured at 2 x 105 cells/well in 96-well flat-bottomed microtiter plates (BD Biosciences). Transforming growth factor (TGF)-β1 was measured according to the manufacturer’s specifications (Promega, Madison, WI). Proliferation assays were performed by 3H-thymidine incorporation at 96 hours of culture as previously described.12
ELISPOT assays were performed by adding 3 x 105 peptide-activated T cells/well in ELISPOT plates (Polyfiltronics, Rockland, MA) containing wells precoated with IFN-, IL-2 (BD Biosciences), or IL-17 (R&D Systems, Minneapolis, MN) capture antibody. At 24 hours, cells were removed, and wells were treated with biotinylated cytokine detection antibodies (BD Biosciences). Spots were visualized using an automated Series-1 immunospot analyzer (Cellular Technology, Cleveland, OH).
In Vivo Manipulation of IL-2 and IL-17
In vivo induction of anergy was performed in male SWXJ mice by injecting 100 µg of CAM 406-425 peptide in PBS three times weekly starting at 2 weeks after adoptive transfer of 2 x 107 Thy1.1+ CAM 406-425-specific LNCs. T cells were released from anergy in vivo by daily intraperitoneal injection of 3 to 6 x 104 IU recombinant human IL-2 (Chiron, Emeryville, CA) throughout a period of 5 weeks in female hosts that had received 2 x 107 Thy1.1+ CAM 406-425-specific LNCs.
IL-17 was depleted in vivo through daily injection of 100 µg of IL-17-neutralizing antibody for 5 weeks starting at 3 weeks after immunization (R&D Systems). In vivo frequencies of IL-17-producing T cells were boosted by culture of CAM 406-425-primed LNCs with 10 ng/ml of IL-23 (R&D Systems) for 4 days before adoptive transfer. The observed increased numbers of IL-17-producing T cells were confirmed by ELISPOT analysis before adoptive transfer.
It is worth noting that our original observations showed sex differences in recall responses to CAM 406-425 8 weeks after adoptive transfer of activated LNCs (Figures 1 and 2)
. Subsequently, we found that female anergy was observed readily by 5 weeks after adoptive transfer and that the anergy did not undergo any spontaneous reversal, particularly between the 5- to 8-week period after transfer. Thus, we often chose the 5-week postadoptive transfer period for analysis of anergy (Figures 3, 4, and 5)
. In the in vivo manipulation experiments (Figure 6)
, we reverted to the 8-week postadoptive transfer time point to maximize any observed impact on DCM.
|
|
|
Cells were analyzed using a FACScan flow cytometer (BD Biosciences) after T-cell culture with or without 100 µg of CAM 406-425 for 72 hours. Fluorescent antibodies against CD3, CD4, CD25, CD40L, CD44, CD62L, CD90.1 (Thy1.1), Annexin V (BD Biosciences), and CTLA-4 (Serotec, Raleigh, NC) were obtained commercially. For cell cycle analysis, 2 x 107 10-day rested T-cell lines taken after two stimulation/rest cycles were co-transferred with 100 µg of lipopolysaccharide (serotype 01111:B4; Sigma), retrieved 5 weeks after passive transfer, and examined by flow cytometry analysis after staining with propidium iodide. For determining Treg frequencies 10-day-primed LNCs were activated in vitro with 25 µg/ml of CAM 406-425 for 72 hours. Cells were then stained with CD4 Per-CP and CD25-PE antibodies (BD Biosciences), permeabilized for 4 hours at room temperature, and stained with FoxP3-FITC antibody (eBioscience, San Diego, CA). Treg frequencies were determined by showing CD4+FoxP3+ double-stained cells in the gated CD25+ population.
Biostatistical Analysis
The two-sided Fisher’s exact test was used to determine differences in DCM incidence between male and female mice. The Student’s t-test was used to analyze differences between all other experimental groups.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
We induced EAM in both male and female SWXJ mice by adoptive transfer of 2 x 107 activated sex-matched CAM 406-425-primed LNCs. Eight weeks after cell transfer, female hearts selectively underwent hypertrophy of their ventricular walls consistent with physiological compensation (Figure 1A
, arrows). In contrast, male hearts preferentially showed a decompensated phenotype with ventricular chamber dilatation and thinning of the mural wall (Figure 1B)
. The difference in DCM incidence between males (13 of 22) and females (2 of 22) was significant (P = 0.015). Moreover, female EAM hearts (Figure 1C
, middle) showed significantly (P = 0.03) less fibrotic replacement (4.0% ± 1.7 SD) than male EAM hearts (13.0% ± 5.4 SD; Figure 1C
, right). Thus, compared to females, males with EAM developed a high incidence of DCM defined as heart size enlargement, parenchymal fibrosis, and ventricular dilatation.
Males but Not Females with EAM Show Sustained Memory to Cardiac Self
Insight into the underlying mechanism for enhanced EAM and more frequent DCM in males emerged after comparison of T-cell recall responses to CAM 406-425 by male- and female-primed LNCs both before (Figure 2
, top) and 8 weeks after (Figure 2
, bottom) adoptive transfer. Memory responses were determined by lymphocyte proliferation (Figure 2
, left), cytokine production (Figure 2
, middle), and ELISPOT frequencies of IFN--producing effector T cells (Figure 2
, right). Despite substantially greater proliferative and cytokine responses before adoptive transfer, female T-cell responses were qualitatively attenuated when compared to analogous male T-cell responses 8 weeks after adoptive transfer. Thus, males showed sustained memory responsiveness to CAM 406-425 whereas females did not. Because females showed a greater number of effector cytokine-secreting T cells and a greater production of Th1-associated cytokines before transfer, we concluded that the inversion of sex-defined response patterns after transfer of autoreactive T cells was not because of a selectively smaller T-cell precursor pool reactive to CAM 406-425 in females nor to a defect in Th1 differentiation occurring pretransfer. Rather, the failure of T-cell memory in females appeared to be attributable to a posttransfer tolerogenic process.
|
We investigated the possibility that failure of T-cell recall responses may be attributable to apoptotic deletion of cardio-specific T cells chronically exposed to high levels of cognate antigen.15,16
Peptide-activated male or female Thy1.1+-primed LNCs or Thy1.1+ peptide-specific long-term T-cell lines were transferred into sex-matched Thy1.1– hosts, and total recovered cells were assessed throughout time. Although we observed a decrease in the recovery of donor Thy1.1+ T cells throughout time, there was no detectable difference between the percentage of recovered male or female Thy1.1+ T cells after transferring 10-day-primed LNCs (Figure 3A
, top) or peptide-specific Thy1.1+ T-cell lines (Figure 3A
, bottom).
The equivalent survival profiles of male and female autoreactive T cells were paralleled both by a gradual decline in female Thy1.1+ cardiac-specific T-cell responsiveness and potentiated autoreactivity of male Thy1.1+ T cells. One week after transfer, female T cells showed greater proliferative responsiveness than male T cells (Figure 3B)
. However, by 4 to 5 weeks, this response pattern became inverted with male T cells showing potentiated autoreactivity both in terms of proliferation (Figure 3B)
and production of IFN- and IL-2 (Figure 3C)
. Because these divergent male and female memory T-cell responses were not associated with a correspondingly discrepant percentage of recovered Thy1.1+ T cells, we concluded that the failure of female T cells to sustain autoimmune memory was not attributable to any female-biased apoptotic deletion.
Thy1.1+ T cells recovered from both male and female hosts consistently expressed the CD4+CD44+CD62L– phenotype of memory T cells (data not shown). Female T cells additionally showed a CD4+CD25+ surface phenotype, prompting investigation as to whether they had acquired regulatory/suppressor activity.17
Female Thy1.1+ T cells were transferred into Thy1.1– recipients, and T cells from donor (Figure 3D
, gate R3) and host (Figure 3D
, gate R2) as well as non-T-cell populations (Figure 3D
, gate R1) were sorted and cultured at varying cell ratios with established CAM 406-425-specific T-cell lines. Donor-derived Thy1.1+CD4+CD25+ T cells, host-derived Thy1.1– T cells, and host-derived non-T cells were unable to suppress responder cell proliferation at any cell ratio tested. However, naturally occurring CD4+CD25+ regulatory T cells purified from normal mouse spleens were able to inhibit proliferative responses of the CAM 406-425-specific T-cell lines. Moreover, when stimulated with anti-CD3, retrieved female Thy1.1+ T cells did not produce substantial amounts of IL-10 or TGF-β (Figure 3E
, top) and did not express CTLA-4 (Figure 3E
, bottom) in a manner comparable to naturally-occurring CD4+CD25+ regulatory T cells.17
To determine whether female-specific suppression of autoreactive CD4+ T cells had occurred in vivo, we transferred female or male CAM 406-425-specific CD4+ Thy1.1+ T-cell lines into sex-matched and sex-mismatched recipients and recovered Thy1.1+ T cells 5 weeks after transfer. To avoid host-directed responsiveness to male-specific antigens, CD8+ T cells were depleted in sex-mismatched recipients by repeated injection with CD8 antibody. Female (Figure 3F
, left) or male (Figure 3F
, middle) hosts that received female-derived Thy1.1+ T-cell lines failed to develop DCM (none of four and none of four, respectively), and the retrieved female Thy1.1+ T cells recovered from both female and male recipients showed hypoproliferative responses to CAM 406-425 (Figure 3F
, right). Hypoproliferative responses to the cognate CAM 406-425 were also elicited by female Thy1.1+ T cells adoptively transferred into neonatally thymectomized Thy1.1– female hosts deficient in production of endogenous Treg populations (data not shown). In contrast, female hosts (Figure 3G
, left) or male hosts (Figure 3G
, middle) receiving male-derived Thy1.1+ T-cell lines developed a relatively high incidence of DCM (three of four and two of four, respectively), and the retrieved male Thy1.1+ T cells from both male and female hosts showed persistent recall responses to CAM 406-425 (Figure 3G
, right). Thus, autoreactive female T cells exhibited no inherent regulatory/suppressor function and showed no indication of being selectively suppressed in the female host milieu.
Autoimmune Memory Failure in Females Is Attributable to T-Cell Anergy
We next determined whether the memory failure in females was attributable to anergy. Five weeks after transfer, CAM 406-425-specific Thy1.1+ donor T cells were retrieved from recipient spleens and cultured in vitro with either 100 µg/ml of peptide alone or with 100 ng/ml of IL-2 for 7 days before addition of peptide. IL-2-mediated recovery from anergy was determined by 3H-thymidine incorporation in response to 100 µg/ml of CAM 406-425. We found that female T cells, whether derived from transfer of primed LNCs or established T-cell lines, were hypoproliferative in response to direct challenge with CAM 406-425 but recovered their responsiveness after in vitro pretreatment with IL-2 (Figure 4A
, left). In contrast, recovered Thy1.1+ male T cells showed recall responsiveness to CAM 406-425 without IL-2 pretreatment and became hypoproliferative only after pretreatment with IL-2 (Figure 4A
, right). Thus, male and female autoreactive T cells show opposite responsiveness to IL-2 with male T cells becoming hypoproliferative and female T cells showing classic IL-2-mediated recovery from anergy.18
|
, top). However, after in vitro activation with 100 µg/ml of CAM 406-425, male Thy1.1+ T cells entered G2 and S phases of the cell cycle at a threefold greater frequency than analogously treated female T cells (27% versus 9%, respectively; Figure 4B
, bottom). G1 arrest of peptide-activated female Thy1.1+ T cells was accompanied by approximately a fourfold lower production of IL-2 when compared to male Thy1.1+ T cells (Figure 4C
, left) and down-modulation of the high-affinity CD25 IL-2 receptor (Figure 4C
, right). Consistent with the features of clonal anergy and not adaptive tolerance,19
peptide-activated female T cells retained the ability to produce IFN- (Figure 4D
, left) and up-regulate CD40L (Figure 4D
, right) in a manner similar to male T cells. The ablation of proliferation in IL-2 pretreated male T cells was found to be attributable to a relative increase in the frequency of apoptotic male T cells as measured by Annexin V staining (Figure 4E)
. Our data indicate that whereas male T cells maintain long-term memory to cardiac self, female T cells become clonally anergized. Anergy and Not Sex-Specific End-Organ Differences Protects Female Hosts from DCM
Several studies have shown uncoupling of effector cytokine production and effector T-cell function. For example, viral-specific CD8+ helpless T cells primed in the absence of CD4+ T cells may undergo several division cycles and up-regulate intracellular stores of IFN- without acquiring the ability to mount protective anti-viral responses.20,21
By analogy, it is possible that female CAM 406-425-specific CD4+ T cells, although producing abundant Th1-like cytokines before adoptive transfer (Figure 2
, top), may not contain myocarditogenic clones and thus may lack the intrinsic ability to induce EAM and DCM. Alternatively, although female and male autoreactive T cells persist in vivo at similar frequencies (Figure 3A)
, it may be possible that transferred Thy1.1+ female T cells undergo selective deletion of high avidity autoreactive T-cell clones.22
Thus, we determined whether female T cells were intrinsically capable of mediating DCM on in vivo recovery from anergy.
CAM 406-425-primed female LNCs were transferred into female hosts that were subsequently given daily injections of IL-2 to ablate anergy in vivo. Sham-treated female controls did not develop DCM (Figure 5A
, left), were relatively unresponsive to CAM 406-425, and recovered autoreactivity after in vitro pretreatment with IL-2 (Figure 5A
, right). In contrast, IL-2-treated females showed a marked increase in the incidence of DCM (7 of 13; Figure 5A
, middle) and showed responsiveness to CAM 406-425 that was ablated after in vitro pretreatment with IL-2 (Figure 5A
, right). Thus, adoptively transferred female T cells contained sufficient numbers of inherently myocarditogenic T cells capable of mediating EAM and DCM.
|
, left) that developed a high incidence of DCM (four of seven) with sustained T-cell memory that was ablated after in vitro pretreatment with IL-2 (Figure 5B
, right), peptide-injected male hosts showed a decreased incidence of DCM (two of nine) and preferentially developed compensatory hypertrophy (Figure 5B
, middle) accompanied by hypoproliferative responsiveness to CAM 406-425 that became enhanced after in vitro pretreatment with IL-2 (Figure 5B
, right). Thus, active in vivo induction of anergy to cardiac self was sufficient to provide DCM protection to males.
Studies have suggested that sex-specific differences related to the end-target organ may differentially modulate the immune effector response.23
To test this possibility, we transferred female or male CAM 406-425-specific Thy1.1+ T-cell lines into sex-matched and sex-mismatched recipients and recovered Thy1.1+ T cells 5 weeks after transfer (Figure 3, F and G)
. Recovered female Thy1.1+ T cells, whether transferred into female hosts (Figure 5C
, top left) or male hosts (Figure 5C
, top right), showed hypoproliferative responses to CAM 406-425 that became enhanced after IL-2 preculture. In contrast, retrieved male Thy1.1+ T cells showed recall responses to CAM 406-425 regardless of transfer into female recipients (Figure 5C
, bottom left) or male recipients (Figure 5C
, bottom right). We concluded that programming for anergy in transferred Thy1.1+ T cells occurred during T-cell priming and persisted independently of end-organ modulation of the effector response. More specifically, female host hearts do not have any selective ability to anergize myocarditogenic T cells.
Sex-Defined EAM Phenotypes Are Attributable to Differential Frequencies of IL-17-Producing T Cells
We next examined the underlying basis for the enhanced ability of male T cells to transfer DCM into either male or female hosts. Since recent studies have shown that severe autoimmune outcomes occur after transfer of a subpopulation of T cells expressing IL-17,24
we compared ELISPOT frequencies of IL-17-producing effector T cells in male and female SWXJ mice 10 days after immunization with CAM 406-425. We observed a consistent differential sex-defined cytokine profile with females showing higher LNC frequencies of IFN--producing T cells (P < 0.003, Figure 6A
) but much lower frequencies of IL-17-producing T cells than males (P < 0.03, Figure 6A
). We hypothesized that the higher frequencies of IL-17-producing T cells in males may account for both their failure to anergize and their severe autoimmune outcome. To test this hypothesis, we boosted frequencies of IL-17-producing T cells in female recipients by treating CAM 406-425-primed LNCs with IL-23 before adoptive transfer.24
Similarly, we neutralized IL-17 in males by treatment with 100 µg of IL-17-neutralizing monoclonal antibody every other day for 5 weeks starting 3 weeks after immunization with CAM 406-425. IL-23 treatment of female LNCs increased IL-17-producing T-cell frequencies (data not shown) resulting in more severe cardiac pathology and a higher incidence of DCM (Figure 6B
, middle) with enhanced production of both IL-17 (P = 0.0000001) and IL-2 (P = 0.002) in splenocyte recall responses to CAM 406-425 (Figure 6B
, right). In a complementary manner, IL-17 neutralization in males resulted in a decreased incidence of DCM (Figure 6B
, middle) as well as abrogated production of both IL-17 (P = 0.03) and IL-2 (P = 0.03) in splenocyte recall responses to CAM 406-425 (Figure 6C
, right). Thus, the severe male nonanergic DCM phenotype and the protected female anergic DCM phenotype may be interchanged through manipulation of frequencies of IL-17-producing T cells.
Analysis of Cardiac-Infiltrating T Cells and Treg Frequencies
Recent reports have implicated enhanced TGF-β production and a Th2 phenotype with the development of cardiac fibrosis and DCM and that IFN- reduces cardiac fibrosis by transcriptional down-regulation of IL-4.25
Although we found that male mice with severe DCM had increased frequencies of IFN--producing T cells 8 weeks after transfer (Figure 2
, lower right), we examined the possibility that the apparent absence of IFN--induced cardio-protection in male mice may have been attributable to decreased local production of IFN- by cardiac-infiltrating T cells. Heart cytokine levels were measured by ELISA 8 weeks after transfer of activated sex-matched CAM 406-425-primed LNCs. We found that when compared to females, T cells that infiltrated male hearts produced substantially higher levels of the Th1 cytokines IFN- and IL-2 (P > 0.05) and that male and female cardiac-infiltrating T cells produced similar low level concentrations of both IL-4 and TGF-β (P > 0.05, Figure 7A
).
|
). Our results are consistent with the view that male and female SWXJ mice have similar frequencies of Tregs and that male mice develop severe DCM in the context of a cardiac milieu involving high levels of IFN- and IL-2 and low levels of IL-4 and TGF-β. |
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The observed sex-defined autoimmune outcomes are directly related to differential CAM 406-425 priming events that predominantly favor a Th1-like response in female SWXJ mice and a Th17-like response in male SWXJ mice. Th1 lineage development is highly dependent on IL-12 whereas the Th17 lineage associated with severe autoimmune outcome has only recently been shown to be induced by IL-6 and TGF-β up-regulation of IL-23 receptors on T cells.26-31 However, the source of enhanced TGF-β in SWXJ males does not appear to be increased numbers of Tregs because males and females showed similar frequencies of CD4+CD25+FoxP3+ T cells in response to CAM 406-425. This latter observation stands in contrast to the increased Treg frequencies implicated in protection of male SJL/J mice from experimental autoimmune encephalomyelitis susceptibility6 and the decreased Treg frequencies implicated in the enhanced myocarditis co-observed in BALB/c males infected with coxsackievirus B3.32 It remains currently unclear whether the selected target antigen or the strain of mice analyzed is more important in determining the role of Tregs in the enhanced severity of autoimmune disease often observed in males.
The dual link to TGF-β in both regulating autoimmunity through Treg function and causing severe autoimmune outcome through Th17 lineage induction implies that a fine balance may exist between these two opposing TGF-β-dependent immune effects and subsequent disease outcomes. This fine balance may be tipped selectively toward Treg-mediated regulation or Th-17-mediated enhanced disease by any of several candidate factors. Bone marrow progenitors, like thymocytes, contain estrogen response elements33 raising the possibility that hormones may govern release from bone marrow or lineage commitment of non-T-cell immune precursors. Sex-disparate terminal differentiation of these precursors may lead to altered ratios of antigen-presenting cell subsets such as the plasmacytoid and myeloid dendritic cells that in humans exert diametrically opposed effects on Th1 and Th2 function.34 In addition, sex hormones may shape sex differential expression of Treg and co-stimulatory molecules3,35 as well as DC maturation, endocytosis, and antigen processing,36 thereby potentially selecting different epitopes and/or altering avidity of the peptide-T cell receptor (TCR) synapse and the subsequent strength of signal that forms the basis for lineage fate selection of several different functional T-cell populations.37-39 Indeed, differential cytokine production may alter intracellular protease function so that differentially-cleaved antigens become alternatively presented during inflammatory responses.40-42 Thus, differentially-processed CAM 406-425 fragments in males and females during T-cell priming may subsequently lead to differential TCR engagement, a process that may fundamentally alter the T-cell response, as has been shown with altered peptide ligands containing point-specific substitutions of TCR contact residues.43
In addition to inducing effects on established T-cell repertoires, sex hormones also have the potential to shape selection of the native repertoire. T-cell repertoire shaping through thymic selection processes occurs prenatally around E17 and E18 in mice coincident with the sex hormone burst that governs formation of the external genitalia and genitourinary tracts. This hormonal burst promotes thymic medullary survival in vivo44,45 and may additionally mold nascent CD4 and CD8 single-positive T-cell ratios postnatally.46,47 Thus, differential autoimmune responses in adult males and females may be attributable in part to sex-defined modulation of thymopoeisis. Prenatal shaping of thymic ontogeny by sex hormones may provide differential sex-based central selection and formation of holistically divergent T-cell repertoires with distinct cellular constituents and/or activation thresholds. It is not difficult to imagine that these different repertoires, in turn, may lead to highly disparate sex-specific responses to a given antigen—whether self or foreign—independent of any postnatal hormonal perturbation of the effector response as we have observed.
It is possible that the observed sex differences in response to cardiac self may represent to some extent the consequence of sex differential responsiveness to the M. tuberculosis adjuvant. However, there are several arguments against this perspective: 1) as a ubiquitous positive control in virtually all of our mouse and human experimentation, we have never observed any differences between males and females with respect to the immune response to M. tuberculosis, including data obtained in the present study (data not shown). 2) We have been unable to find any reputable published reports detailing male and female differences in the immune response to M. tuberculosis. 3) Our model system involves induction of passive disease after adoptive transfer of primed LNCs and/or T-cell lines. Therefore, experimental mice with differential sex-biased development of DCM were never exposed to antigenic challenge with M. tuberculosis.
It is worth noting that our current study and those previously published7,8 support the view that myocarditis, whether induced by virus or autoimmunity, seems to be more severe in males than in females. At the same time, it is less clear that the increased disease severity in males is attributable to a consistently similar qualitative immune response because a variety of immunological conditions have been shown to contribute to the severe DCM outcome in males including a predominant Th2 response,25 a predominant Th17 response independent of Th1/Th2 polarization,48 decreased Treg frequencies,32 decreased T-cell Ig mucin (Tim-3) expression,32 and, as shown in the current study, failure to anergize Th1 and Th17 responses.
In conclusion, our study indicates that male and female mice engage cardiac self in distinctly different manners predisposing females to Th1 lineage selection and T-cell anergy and males to Th17 lineage responses and sustained self-memory. These sex-defined self-recognition patterns ultimately dictate clinical outcome and account directly for the relative protection afforded to females and the enhanced disease severity observed in males. The direct linkage of sex-defined autoimmune responses to differential autoimmune outcomes reveals the potential for optimizing treatment of immune-mediated diseases by developing sex-tailored therapeutic strategies.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the National Institutes of Health (grants AI-051837 and DC-006422 to V.K.T.) and the American Heart Association (fellowship 0215114B to D.J.-W.).
Current address of T.G.F.: Department of Biology, University of Texas at San Antonio, San Antonio, TX.
Accepted for publication September 10, 2007.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
