Reduced 2,4-dinitro-1-fluorobenzene-induced contact hypersensitivity response in IL-15 receptor α-deficient mice correlates with diminished CCL5/RANTES and CXCL10/IP-10 expression

Authors


Abstract

Using a model of 2,4-dinitro-1-fluorobenzene-induced contact hypersensitivity (CHS) we found that, as compared with wild-type mice, IL-15 receptor α chain (IL-15Rα)-deficient mice showed significantly less ear swelling. This decreased response was associated with diminished expression of CCL5/RANTES and CXCL10/IP-10, chemokines critical for effector cell recruitment, in the inflamed tissue. We determined that both the number of CD8+ T cells infiltrating the affected skin and the production of CCL5/RANTES by antigen-stimulated CD8+ T cells were decreased in IL-15Rα–/– mice. The lower levels of CXCL10/IP-10 suggested that the IL-15Rα–/– mice had reduced production of IFN-γ, the primary inducer of CXCL10/IP-10, which was in fact the case. However, by contrast with CCL5/RANTES, the diminished levels of IFN-γ were likely due to the decreased number of skin-infiltrating CD8+ T cells, since IFN-γ production by antigen-stimulated CD8+ T cells was comparable between wild-type and IL-15Rα–/– mice. Our data suggest a positive, pro-inflammatory feedback loop involving CCL5/RANTES, IFN-γ and CXCL10/IP-10 that underlies the CHS reaction and that is disrupted, likely primarily by a defect in CCL5/RANTES production, in mice lacking IL-15Rα, resulting in impaired leukocyte recruitment and inflammation. Moreover, it is particularly noteworthy that the defect in CCL5/RANTES expression in CD8+ T cells is intrinsic to the absence of IL-15Rα, indicating that IL-15Rα is critical for CCL5/RANTES expression in CD8+ T cells.

Abbreviations:
CCL:

CC chemokine ligand

CXCL:

CXC chemokine ligand

CCR:

CC chemokine receptor

IL-15Rα:

IL-15 receptor α chain

CHS:

Contact hypersensitivity

GAPDH:

Glyceraldehyde 3-phosphate dehydrogenase

GPCR:

G protein-coupled receptor

DNBS:

2,4-Dinitrobenzenesulphonic acid

DNFB:

2,4-Dinitro-1-fluorobenzene

RPA:

Ribonuclease protection assay

Introduction

Although initially identified as a T cell-stimulating cytokine, IL-15 has since been shown to be a pleiotropic cytokine (1 and reviewed in 2) involved in many human inflammatory diseases, suggesting it is an important mediator of the inflammatory response 3. IL-15 targets cells through a heterotrimeric receptor (IL-15R) composed of a private α chain (IL-15Rα), a β chain shared with IL-2R, and a common γ (γc) chain shared with receptors for IL-2, IL-4, IL-7, IL-9 and IL-21 24. The private receptor α chain contributes to high-affinity binding to its specific ligand, while the β and γc chains presumably mediate shared biological activities. Mice deficient in IL-15 or IL-15Rα (IL-15–/– or IL-15Rα–/–) show marked reductions in their CD8+ T cell counts, particularly memory CD8+ T cells 57, as well as reductions in their NK cell, NK T cell and γδ intestinal intraepithelial T cell counts 5, 6. Studies of IL-15–/– and IL-15Rα–/– mice have also confirmed the importance of IL-15 in the maintenance of antigen-specific memory CD8+ T cells in viral infection 8, 9, and a recent study demonstrated the importance of APC-derived IL-15 for the induction of CD8+ T cell-mediated immune responses 10.

Chemokines are members of a family of small, secreted and inducible proteins that target cells through the seven-transmembrane domain G protein-coupled receptors (GPCR) 11. The major physiological function of chemokines is to recruit leukocytes to sites of immune response and inflammation 11, 12; thus, chemokines and their receptors play critical roles in a variety of inflammatory diseases 1315. During an inflammatory response, chemokines attract leukocytes mediating the innate immune response, as well as those mediating an adaptive immune response. Therefore, chemokines facilitate direct communication between the innate and adaptive immune responses 16, 17. Furthermore, because chemokine expression is usually under the control of cytokines, the interplay between cytokines and chemokines is central to the development of such chronic inflammatory diseases as rheumatoid arthritis 1315, 18, active ulcerative colitis 15, 19 and multiple sclerosis 13, 15, 20.

In this study, we employed IL-15Rα–/– mice in a contact hypersensitivity (CHS) model to investigate the effect of the IL-15Rα deficiency on chemokine expression. We chose CHS as a model because cytokines and chemokines are known to act as key mediators during the full development of the inflammatory response in CHS 2123. We found that both CCL5/RANTES and CXCL10/IP-10 expression were significantly decreased in the inflammatory response to 2,4-dinitro-1-fluorobenzene (DNFB) in IL-15Rα–/– mice and that there is a correlation between decreased CCL5/RANTES and CXCL10/IP-10 expression and the reduction of effector cell infiltration, thereby diminishing the CHS response.

Results

CHS responses were significantly weakened in IL-15Rα–/– mice

To examine the effect of IL-15Rα–/– on the CHS response to DNFB, we first sensitized groups of wild-type and IL-15Rα–/– mice by applying DNFB to the skin of their abdomens on days 0 and 1. After 5 days, we challenged these mice, as well as matching groups of untreated mice, by applying DNFB to their ears and to patches of skin on their backs and then measured ear swelling 24 h post challenge. We found that IL-15Rα–/– mice showed significantly less ear swelling than wild-type mice (p<0.0001 for DNFB-sensitized IL-15Rα–/– mice vs. DNFB-sensitized wild-type mice) (Fig. 1).

Figure 1.

IL-15Rα–/– mice show a significant reduction in ear swelling during the CHS response. Wild-type and IL-15Rα–/– mice were left untreated or sensitized with DNFB on their abdominal skin on days 0 and 1. The ears and back were then challenged with DNFB on day 5, and 24 h later, ear swelling was measured by comparing ear thicknesses measured before and after challenge. *p<0.0001 for DNFB-sensitized IL-15Rα–/– mice (n=25) vs. DNFB-sensitized wild-type mice (n=31).

IL-15Rα–/– mice showed reduced expression of CCL5/RANTES and CXCL10/IP-10 in the inflamed tissues during the CHS response

To test whether the weakened CHS response in IL-15Rα–/– mice was the result of diminished chemokine expression in the inflamed skin, which would in turn reduce recruitment of effector cells to the inflamed sites, we used ribonuclease protection assays (RPA) to evaluate chemokine expression in the inflamed ear and back skin of both wild-type and IL-15Rα–/– mice (Fig. 2). We found that expression of multiple chemokine genes (XCL1/lymphotactin, CCL5/RANTES, CCL3/MIP-1α, CCL4/MIP-1β, CXCL1–3/MIP-2, CXCL10/IP-10, and CCL2/MCP-1) was induced in the DNFB-challenged ears (Fig. 2A) and skin (Fig. 2B) of both wild-type and IL-15Rα–/– mice. Notably, the levels of CCL5/RANTES and CXCL10/IP-10 were substantially reduced in IL-15Rα–/– mice compared to wild-type mice, by 2.5- and 1.8-fold, respectively. The level of XCL1/lymphotactin was also reduced in IL-15Rα–/– mice, to a lesser extent. Expression of the other chemokines tested appeared comparable in the two groups. Expression of CCL11/eotaxin, which is associated with type 2 immune responses, was not significantly induced (Fig. 2), confirming CHS being a type 1, but not type 2, immune response.

Figure 2.

Multiple chemokine genes are induced during the CHS response. Total RNA was extracted from the skin of the ears (A) and the challenged sites on the back (B), and 20 µg of total RNA from each sample was tested in an RPA for analysis of chemokine genes. The gels were dried and exposed to Kodak BioMax MR film for 3 days at –70°C. The data are representative of six independent experiments. The data shown are representative of two independent experiments. L32 and GAPDH are housekeeping genes.

The finding of significantly reduced expression of CCL5/RANTES and CXCL10/IP-10 mRNA in the inflamed skin of IL-15Rα–/– mice led us to test whether there were corresponding reductions in the expression of their encoded proteins. Using ELISA, we found that the expression of CCL5/RANTES (Fig. 3A) and CXCL10/IP-10 (Fig. 3B) was significantly up-regulated in the inflamed skin of both wild-type and IL-15Rα–/– mice during the CHS response, but that the degree of up-regulation was significantly reduced in IL-15Rα–/– mice.

Figure 3.

The inflamed ears of IL-15Rα–/– mice contain reduced levels of CCL5/RANTES and CXCL10/IP-10. Ears from DNFB-sensitized and untreated mice were excised and homogenized in lysis buffer. Levels of CCL5/RANTES (A) and CXCL10/IP-10 (B) in the homogenates were assayed using specific ELISA. The data are presented as means ± SEM. *p<0.0001 for CCL5/RANTES (n=15) and p<0.05 for CXCL10/IP-10 (n=13) for DNFB-sensitized IL-15Rα–/– mice vs. DNFB-sensitized wild-type mice.

IL-15Rα–/– mice showed reduced infiltration of CD8+ T cells in the inflamed ear

Given that both CCL5/RANTES and CXCL10/IP-10 are potent chemoattractants for effector CD8+ T cells 2426, that CD8+ T cells are the main effector cells mediating the CHS response 2730 and that IL-15Rα–/– mice showed reduced expression of CCL5/RANTES and CXCL10/IP-10 in inflamed skin (Figs. 2, 3), we would expect that fewer CD8+ T cells would be recruited to sites of inflammation in IL-15Rα–/– mice than in wild-type mice. This prediction was confirmed by the results of immunohistochemical analysis, which showed that CD8+ T cells infiltrated the inflamed ears of DNFB-sensitized mice, but not those of naive mice, and that the ears of DNFB-sensitized IL-15Rα–/– mice had fewer infiltrating CD8+ T cells than wild-type mice (Fig. 4A). Consistent with the presence of fewer CD8+ T cells, real-time PCR showed that IL-15Rα–/– mice expressed lower levels of CD8+ mRNA in the inflamed skin than did wild-type mice (Fig. 4B).

Figure 4.

Fewer CD8+ T cells infiltrate the inflamed skin of IL-15Rα–/– mice. (A) Ear sections from naive and DNFB-sensitized wild-type and IL-15Rα–/– mice were immunostained with PE-conjugated anti-mouse CD8 Ab and counterstained with DAPI; magnification, 200× (A–D) or 400× (E–F). The bar graph shows the quantitative result obtained by counting the number of cells in 10 fields (200×) from three mice. (B) Total RNA was extracted from the challenged skin on the backs of wild-type and IL-15Rα–/– mice, and 5 µg total RNA was used for first-strand cDNA synthesis. The first-strand cDNA and specific primer pairs for CD8 were then subjected to real-time PCR. The relative CD8 mRNA units (2–ΔΔCT) represent the fold induction (with sensitization) over control (without sensitization) from the wild type. The data shown are expressed as the fold induction of relative CD8 mRNA units from DNFB-sensitized wild-type and IL-15Rα–/– mice to unsensitized wild-type mice (the 2–ΔΔCT obtained from unsensitized wild-type mice was set to a value of 1). The data shown are representative of three independent experiments.

CD8+ T cells from DNFB-sensitized IL-15Rα–/– mice showed reduced CCL5/RANTES secretion in response to specific antigen

Given that IL-15Rα–/– mice showed a lower number of CD8+ T cells in the inflamed skin during the CHS response (Fig. 4) and that CD8+ memory T cells highly express CCL5/RANTES 3133, it is likely that the reduced CCL5/RANTES expression found in the inflamed skin reflects, at least in part, the lower number of infiltrating CD8+ T cells in IL-15Rα–/– mice. However, whether CCL5/RANTES production by antigen-stimulated CD8+ T cells would be directly affected by IL-15Rα deficiency was not known. We then examined the CCL5/RANTES expression in both wild-type and IL-15Rα–/– CD8 T cells in response to antigen. When CD8+ T cells isolated from the draining lymph nodes of DNFB-sensitized wild-type and IL-15Rα–/– mice were cultured for 2 days with the antigen 2,4-dinitrobenzenesulphonic acid (DNBS) and irradiated APC obtained from wild-type and IL-15Rα–/– mice, respectively, there was a significant increase in CCL5/RANTES secretion compared to unsensitized mice. Notably, DNFB-sensitized CD8+ T cells from IL-15Rα–/– mice secreted significantly lower levels of CCL5/RANTES than cells from wild-type mice (p<0.01) (Fig. 5A), indicating that production of CCL5/RANTES by antigen-stimulated CD8+ T cells was impaired in IL-15Rα–/– mice. On the other hand, DNFB sensitization did not induce a statistically significant increase in CCL5/RANTES production by CD4+ T cells in either wild-type or IL-15Rα–/– mice (Fig. 5B).

Figure 5.

CCL5/RANTES production is diminished in CD4+ and CD8+ T cells from IL-15Rα–/– mice during the CHS response. CD4+ and CD8+ T cells from the draining lymph nodes of DNFB-sensitized or untreated wild-type or IL-15Rα–/– mice were isolated using a MACS separation column. Isolated CD8+ (A) or CD4+ (B) T cells (1×105) were cultured with irradiated APC (1×105) isolated from spleens in the absence or presence of 50 µg/ml DNBS in a total volume of 200 µl. The 2-day culture supernatants were collected and analyzed for CCL5/RANTES by ELISA. The data presented are the CCL5/RANTES contents measured in culture supernatants from cells cultured with DNBS minus the content of culture supernatants from cells cultured without DNBS. The data are presented as means ± SEM; *p<0.01 for DNFB-sensitized IL-15Rα–/– mice vs. DNFB-sensitized wild-type mice.

Reduced expression of CCL5/RANTES in memory CD8+ T cells in IL-15Rα–/– mice

We then examined whether a reduced CCL5/RANTES expression in IL-15Rα–/– mice was as a result of the deficiency in IL-15Rα. Using RPA to analyze the expression of CCL5/RANTES in T cells from both naive wild-type and IL-15Rα–/– mice, we found that expression of CCL5/RANTES mRNA was indeed diminished in T cells, particularly CD8+ T cells, from IL-15Rα–/– mice (Fig. 6). This suggested that the defect of IL-15Rα leads to a decreased expression of CCL5/RANTES by T cells.

Figure 6.

Expression of CCL5/RANTES mRNA is diminished in T cells isolated from IL-15Rα–/– mice. Total RNA was extracted from purified CD4+ and CD8+ T cells isolated from the spleens and lymph nodes of wild-type and IL-15Rα–/– mice, and 5-µg samples were evaluated in RPA. The intensities of CCL5/RANTES and L32 were obtained by phosphoimage analysis. Shown are relative expression levels normalized to the expression of the housekeeping gene L32. The data are presented as means ± SEM of three independent experiments.

Because CCL5/RANTES is mainly expressed in memory CD8+ T cells 3134 and IL-15Rα–/– mice manifest a severe deficit in that cell type 6, 7, we next examined whether the lower CCL5/RANTES levels in CD8+ T cells from IL-15Rα–/– mice reflected a reduction in IL-15Rα-mediated CCL5/RANTES expression, or merely the smaller number of memory CD8+ T cells. Naive and memory CD8+ T cells from both DNFB-challenged wild-type and IL-15Rα–/– mice were isolated, and their CCL5/RANTES expression was analyzed using real-time PCR. Whereas naive CD8+ T cells from both wild-type and IL-15Rα–/– mice expressed very low levels of CCL5/RANTES, memory CD8+ T cells showed substantial levels of expression (Fig. 7). Notably, CCL5/RANTES expression in memory CD8+ T cells form IL-15Rα–/– mice was significantly lower than in wild-type mice (Fig. 7), supporting the idea that reduced CCL5/RANTES expression in CD8+ T cells from IL-15Rα–/– mice is an intrinsic feature of the deficiency in IL-15Rα.

Figure 7.

Memory CD8+ T cells from IL-15Rα–/– mice show significantly reduced CCL5/RANTES expression. CD8+ T cells were isolated using MACS columns, stained with anti-CD44 Ab and applied to a cell sorter. After sorting the naive (CD44low) and memory (CD44high) cells, total RNA was extracted. Samples of the total RNA equal to that obtained from 3×105 cells were then used for first-strand cDNA synthesis, after which the first-strand cDNA and specific primer pairs for CCL5/RANTES were subjected to real-time PCR. The data shown are expressed as the fold induction in relative CCL5/RANTES mRNA units in DNFB-sensitized wild-type and IL-15Rα–/– mice normalized to the levels in unsensitized wild-type mice (the 2–ΔΔCT obtained from unsensitized wild-type mice was set to a value of 1).

IL-15 stimulated secretion of CCL5/RANTES in CD8+ T cells

The finding of diminished CCL5/RANTES expression in T cells lacking IL-15Rα led us to examine the extent to which IL-15 would stimulate T cells to secrete CCL5/RANTES. We found that IL-15 significantly stimulated CCL5/RANTES secretion from CD8+ T cells, but not from CD4+ T cells, and that CD8+ T cells secreted significantly higher levels of CCL5/RANTES than CD4+ T cells (Fig. 8), indicating that IL-15 significantly regulated CCL5/RANTES secretion in CD8+ T cells, but not in CD4+ T cells. Consistent with our findings, IL-15 was shown to up-regulate CCL5/RANTES expression in human T cells 35; however, no further dissection on T cell subsets responding to IL-15 was reported. Moreover, unlike resting murine T cells highly expressing CCL5/RANTES 31, 32, no significant CCL5/RANTES mRNA was found in resting human T cells 35.

Figure 8.

IL-15 stimulates CCL5/RANTES secretion in CD8+ T cells. CD4+ or CD8+ T cells isolated from wild-type mice were cultured (2×105 cells/200 µl/well) for 24 h with or without 100 ng/ml IL-15, after which the culture supernatants were collected and CCL5/RANTES levels were measured by ELISA. *p<0.0001.

CD8+ T cells from DNFB-sensitized wild-type and IL-15Rα–/– mice secreted comparable amounts of IFN-γ in response to specific antigen

Analyzing IFN-γ secretion by CD8+ T cells cultured with irradiated APC in the presence or absence of DNBS, we found that CD8+ T cells from IL-15Rα–/– mice produced levels of IFN-γ comparable to those seen in wild-type mice (Fig. 9A), indicating that IL-15Rα deficiency did not affect IFN-γ production by DNFB-sensitized CD8+ T cells in the CHS response. We found low levels of IFN-γ mRNA in the inflamed skin of IL-15Rα–/– mice by real-time PCR (Fig. 9B), consistent with the reduced number of CD8+ T cells and with the reduced levels of CXCL10/IP-10.

Figure 9.

Wild-type and IL-15Rα–/– mice produce comparable levels of IFN-γ by antigen-stimulated CD8+ T cells, but IL-15 Rα–/– mice show lower levels of IFN-γ in inflamed tissues during the CHS response. (A) Isolated CD8+ and CD4+ T cells were cultured as described in Fig. 5. After 2 days in culture, the culture supernatants were collected and subjected to IFN-γ analysis by ELISA. The data are presented as means ± SEM. (B) Real-time PCR showing relative levels of IFN-γ mRNA expression in inflamed skins was carried out and quantified as described in Fig. 4B.

Discussion

CCL5/RANTES is a potent chemoattractant for activated and memory T cells, which in vitro show up-regulated surface expression of CCR5, one of the receptors for CCL5/RANTES 25, 26. Indeed, CCL5/RANTES expression and CCR5-bearing T cells are frequently coexistent in inflamed lesions 14. In the present study, we demonstrated that CCL5/RANTES mRNA was significantly lower in CD8+ T cells from IL-15Rα–/– mice than from wild-type mice and that IL-15 significantly stimulated CCL5/RANTES production in CD8+ but not CD4+ T cells, from wild-type mice. Using a CHS model, we also showed that antigen-specific CD8+ T cells in IL-15Rα–/– mice produced significantly lower levels of CCL5/RANTES in response to DNFB and that this reduction in CCL5/RANTES expression is an intrinsic feature of the deficiency in IL-15Rα. Taken together, these findings suggest that, in addition to its role in CD8+ T cell survival and proliferation, IL-15 is involved in regulating CCL5/RANTES expression in CD8+ T cells. It is noteworthy in that regard that although CCL3/MIP-1α and CCL4/MIP-1β are structurally homologous to CCL5/RANTES and share receptors with it 36, their expression in response to DNFB was not affected by IL15Rα deficiency, indicating the specificity of IL-15 in regulating CCL5/RANTES expression.

Murine memory CD8+ T cells contain preexisting CCL5/RANTES mRNA, which results in rapid secretion of CCL5/RANTES after TCR ligation in a transcription-independent manner 31, 32. In human memory/effector CD8+ T cells, the preformed CCL5/RANTES protein is stored in granules and rapidly released following TCR activation 33, 34. Whether IL-15 is involved in regulation of rapid translation of CCL5/RANTES or secretion of CCL5/RANTES from granules in memory CD8+ T cells upon T cell activation is currently under investigation. Notably, we found that at the individual cell level, a reduced CCL5/RANTES mRNA is directly associated with IL-15Rα deficiency in memory CD8+ T cells. It would be of great interest to know how IL-15 receptor signaling mediates transcription of CCL5/RANTES mRNA in memory CD8+ T cells.

Our data indicate that the expression of CCL5/RANTES is up-regulated in CD8+ but not in CD4+ T cells responding to DNBS, consistent with the role of CD8+ T cells as effectors in the CHS response 2730. In addition, we found that IL-15 was able to enhance CCL5/RANTES production in CD8+ T cells and that IL-15 expression was up-regulated in the inflamed skin during the CHS response (data not shown), suggesting that IL-15 plays a key role in the stimulation of CCL5/RANTES production in CD8+ T cells in the CHS response. It is therefore possible that both antigen and IL-15 participate in stimulating effector CD8+ T cells to produce CCL5/RANTES in the inflamed tissue during the CHS response. It is plausible to propose that, upon recognition of the antigen DNFB, the effector CD8+ T cells that are recruited early may rapidly secrete CCL5/RANTES, which in turn promotes the recruitment of additional effector CD8+ T cells to the inflamed site. These recruited effector CD8+ T cells produce CCL5/RANTES in the inflamed site, which recruits more effector CD8+ T cells, as well as other types of leukocytes. Moreover, IL-15, presumably a product of keratinocytes and Langerhans cells 37, further enhances CD8+ T cells producing CCL5/RANTES. Thus, CCL5/RANTES may provide a positive feedback loop mediating the enhanced responsiveness that is characteristic of the elicitation phase of CHS to DNFB. Supporting this hypothesis is our finding that the expression of CCR1 and CCR5, two CCL5/RANTES receptors, is reduced in inflamed tissue from IL-15Rα–/– mice (data not shown).

We found that the chemokine gene most strongly induced in the inflamed skin during the CHS response encoded CXCL10/IP-10, which is consistent with earlier studies 3841. CXCL10/IP-10 was first identified as a gene induced by IFN-γ 42, which is mainly produced by CD8+ T cells during the CHS response 21, 22. The fact that IFN-γ expression in the inflamed skin is associated with the development of a CHS response and with infiltration of hapten-specific CD8+ T cells 28, 30, 39, 43, that CD8+ T cells from DNFB-sensitized IL-15Rα–/– and wild-type mice produced similar levels of IFN-γ , and that fewer CD8+ T cells infiltrated the inflamed skin of IL-15Rα–/– mice, led us to propose that the lower level of IFN-γ in the inflamed skin of IL-15Rα–/– mice is likely due solely to the decreased number of infiltrating CD8+ T cells there. The lower level of IFN-γ would in turn more weakly induce CXCL10/IP-10, leading to a reduction in the recruitment of effector cells. Similar to CXCL10/IP-10, another two IFN-γ-inducible chemokines, CXCL9/Mig and CXCL11/I-TAC, were also significantly induced in the inflamed skin detected by real-time PCR, and their levels of induction were also lower in IL-15Rα–/– mice (data not shown).

In summary, using IL-15Rα–/– mice, we have showed that the role of IL-15 in the CHS response is to regulate CCL5/RANTES expression in CD8+ T cells, which then initiates a pro-inflammatory feedback loop involving CCL5/RANTES and CXCL10/IP-10, which are critical chemokines for the recruitment of effector cells to the site of inflammation in the CHS response. The collective reduction in expression of CCL5/RANTES and CXCL10/IP-10 in the inflamed skin of IL-15Rα–/– mice impairs the recruitment of effector cells to the hapten-challenged sites, thus reducing the CHS response. Our results, together with recent studies using IL-15–/– mice and keratinocyte-targeted IL-15-transgenic mice showing the essential role of IL-15 in initiation of the CHS response 10, 44, suggest that IL-15 plays critical roles in both the induction and elicitation phases of the CHS response.

Materials and methods

Mice

This study was carried out using age- and sex-matched groups of wild-type and IL-15Rα–/– mice. The latter were generated by gene disruption and backcrossed onto the C57BL/6 background as described 7. The IL-15Rα–/– mice used in the present study were 8–12 weeks old and had been backcrossed to C57BL/6 mice for seven to nine generations. The mice were housed under specific pathogen-free conditions at the Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan. Experiments were performed in accordance with the Institutional Guidelines (protocol: RMiIBMLF2001035).

Contact hypersensitivity assay

The mice were sensitized and challenged to elicit a CHS response to DNFB (Sigma-Aldrich, St. Louis, MO). Animals were sensitized by painting their shaved abdomens with 20 µl of 0.5% DNFB in acetone/olive oil (4:1) on days 0 and 1. On day 5, ear thickness was measured, after which each mouse was challenged by applying 10 µl 0.2% DNFB to each side of both ears and to each of three shaved sites (∼1 cm diameter per site) on their backs. Naive mice that were not sensitized but were challenged with DNFB served as negative controls. Ear thickness was measured 24 h after challenge; the extent of ear swelling was determined by subtracting the thickness before challenge.

RPA

Total RNA from tissues or T lymphocytes was isolated using Trizol (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions, after which the levels of various chemokine mRNA were analyzed using multi-probe RPA. Multi-probe templates were purchased from BD PharMingen (San Diego, CA), and the assay was carried out according to the manufacturer's instructions. Briefly, radiolabeled RNA probes were generated from multi-probe templates using T7 RNA polymerase and a mixture of pooled unlabeled nucleotides and [α-32P]UTP. The probes were hybridized overnight with 5–20 µg total RNA and then digested first with an RNase cocktail and then with proteinase K. RNase-resistant duplex RNA were extracted with phenol, precipitated with ammonium acetate, solubilized and resolved on a 5% sequencing gel, which was then dried and subjected to autoradiography and phosphoimage analysis.

Measurement of CCL5/RANTES and CXCL10/IP-10 protein in ears

Ears were excised, weighed and homogenized on ice in protein extraction buffer (BioChain Institute Inc., Hayward, CA) using a tapered tissue grinder (Wheaton Science Products, Millville, NJ). The resultant homogenates were centrifuged at 15,000×g for 20 min at 4°C, after which the supernatants were collected and assayed for CCL5/RANTES and CXCL10/IP-10 using ELISA.

Immunohistochemistry

Ears were soaked in 30% sucrose overnight at 4°C, embedded in OCT (DAKO, Carpentaria, CA), sliced into 6-µm sections on a cryostat microtome, placed on silanized slides and air-dried overnight. The air-dried sections were then fixed for 10 min at RT in 4% paraformaldehyde, washed twice with PBS and blocked for 20 min with 5% normal goat serum in PBS. The sections were then incubated overnight at 4°C with PE-conjugated anti-mouse CD8 Ab (1:200 dilution, clone 53–6.7; BD PharMingen). Thereafter, the sections were washed, permeabilized for 15 min at RT with 0.05% Triton X-100 and counterstained for 5 min at RT with DAPI (0.2 µg/ml in PBS). PE-conjugated rat IgG1 (BD PharMingen) was used as an isotype control. The fluorescent images were observed under a Zeiss Axiovert 200M inverted microscope (Gollingen, Germany) equipped with a digital CCD camera (Cool Snap HQ; Photometrics, Tucson, AZ). MetaMorph imaging software (version 6.106; Imaging Corporation, Downing town, PA) was used for processing the cell image data. The number of infiltrating cells was counted in ten fields under a magnification of 200×.

Real-time PCR analysis

Of total RNA, 5 µg was used for cDNA synthesis using an oligo(dT) primer and SuperScript II reverse transcriptase (Invitrogen) according to the manufacturer's instructions. For real-time PCR analysis, the 2–ΔΔCT method was used to quantify the relative changes in gene expression 45. Real-time PCR for CD8, IFN-γ, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was carried out using Assays-on-Demand Gene Expression products (Applied Biosystems), consisting of a 20× mix of unlabeled PCR primers and TaqMan MGB probe (FAM dye-labeled). The PCR cycling entailed 1 cycle at 50°C for 2 min and then 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 60 s. The resultant PCR products were measured and elaborated using an ABI Prism 7700 Sequence Detection System (Applied Biosystems). All samples were run in duplicate. All quantifications were normalized to the level of GAPDH gene expression. Analysis of the relative quantification required calculations based on the threshold cycle (CT: the fractional cycle number at which the amount of amplified target reaches a fixed threshold) as follows: ΔCT was calculated as the difference between the mean CT values of the samples evaluated with IFN-γ- or CD8-specific primers and mean CT values of the same samples evaluated with GAPDH-specific primers; ΔΔCT was calculated as the difference between the ΔCT values of the samples and the ΔCT value of the calibrator sample; and 2–ΔΔCT was the relative mRNA units representing the fold induction over the control.

Isolation of different populations of T cells from draining lymph nodes

Axillary and inguinal lymph nodes were collected from mice and gently disrupted by pressing the tissues through a nylon mesh cell strainer. The cells were then washed with RPMI 1640 containing 1% FBS, and the CD4+ and CD8+ T cells were isolated by negative selection using a MACS separation column (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's instructions. The purified lymphocytes (92–95% purity) were then subjected to RPA assay or cell culture.

To separate naive and memory CD8+ T cells, the isolated CD8+ T cells were stained with anti-CD44-PE Ab and then sorted using FACSVantage SE (BD Bioscience). Because of the very limited numbers of memory CD8+ cells in IL-15Rα–/– mice, we analyzed CCL5/RANTES expression in naive (CD44low) and memory (CD44high) cells using real-time PCR. Total RNA was isolated using an Absolutely RNA RT-PCR miniprep kit (Stratagene, La Jolla, CA). RNA concentration was quantified using a RiboGreen RNA Quantitation Kit (Molecular Probes, Eugene, OR).

ELISA for CCL5/RANTES, CXCL10/IP-10 and IFN-γ

Isolated CD4+ and CD8+ (1×105) T cells from wild-type or IL-15Rα–/– were cultured with irradiated wild-type or IL-15Rα–/– APC (1×105) purified from spleens in the presence or absence of 50 µg/ml DNBS, a soluble form of DNFB, in a total volume of 200 µl. After 2 days in culture, the culture supernatants were collected and analyzed for CCL5/RANTES and INF-γ using sandwich ELISA. Recombinant murine CCL5/RANTES and CCL5/RANTES ELISA kits were purchased from R&D Systems (Minneapolis, MN), and ELISA were carried out according to the manufacturer's instructions. IFN-γ and CXCL10/IP-10 in the conditioned medium from cultured lymphocytes were determined using a purified unconjugated capture mAb (clone R4–6A2 for IFN-γ; clone 134013 for CXCL10/IP-10) and biotinylated detecting Ab (clone XMG1.2-biotin for IFN-γ; biotin-conjugated goat anti-mouse CRG-2 polyclonal Ab for CXCL10/IP-10). The Ab pairs for IFN-γ and CXCL10/IP-10 were purchased from BD PharMingen and R&D Systems, respectively.

To detect CCL5/RANTES secretion by T cells cultured with IL-15, isolated CD4+ or CD8+ T cells (2×105 cells/well) from wild-type mice were cultured for 24 h in the absence or presence of 100 ng/ml IL-15 (R&D Systems) in a total volume of 200 µl. The conditioned medium was then collected and subjected to analysis of CCL5/RANTES by ELISA.

Statistical analysis

The data are presented as the means ± the standard error of the mean (SEM). The values were compared through analysis of variance (ANOVA) with Tukey-Kramer multiple comparisons test. Values of p<0.05 were considered significant.

Acknowledgements

We thank Ms. Hsuen-Chin Chen for her assistance in immunohistochemistry, Mr. Jen-Chi Hsu for his providing the backcrossed breeders and assistance in genotyping, and Mr. Yihsuan Chang for his help with figures. We also thank Dr. Joshua Farber for the discussion, and Dr. Betty Wu-Hsieh for reading the manuscript. This work was supported by grants from the National Health Research Institutes, National Science Council, and Academia Sinica in Taiwan.

Footnotes

  1. 1

    WILEY-VCH

  2. 2

    WILEY-VCH

  3. 3

    WILEY-VCH

  4. 4

    WILEY-VCH

  5. 5

    WILEY-VCH

  6. 6

    WILEY-VCH

  7. 7

    WILEY-VCH

  8. 8

    WILEY-VCH

  9. 9

    WILEY-VCH

Ancillary