Development of Memory-Like Autoregulatory CD8+ T Cells Is CD4+ T Cell Dependent

Afshin Shameli, Xavier Clemente-Casares, Jinguo Wang and Pere Santamaria
J Immunol September 15, 2011, 187 (6) 2859-2866; DOI: https://doi.org/10.4049/jimmunol.1101117

Abstract

Progression of spontaneous autoimmune diabetes is associated with development of a disease-countering negative-feedback regulatory loop that involves differentiation of low-avidity autoreactive CD8+ cells into memory-like autoregulatory T cells. Such T cells blunt diabetes progression by suppressing the presentation of both cognate and noncognate Ags to pathogenic high-avidity autoreactive CD8+ T cells in the pancreas-draining lymph nodes. In this study, we show that development of autoregulatory CD8+ T cell memory is CD4+ T cell dependent. Transgenic (TG) NOD mice expressing a low-affinity autoreactive TCR were completely resistant to autoimmune diabetes, even after systemic treatment of the mice with agonistic anti-CD40 or anti–4-1BB mAbs or autoantigen-pulsed dendritic cells, strategies that dramatically accelerate diabetes development in TG NOD mice expressing a higher affinity TCR for the same autoantigenic specificity. Furthermore, whereas abrogation of RAG-2 expression, hence endogenous CD4+ T cell and B cell development, decelerated disease progression in high-affinity TCR-TG NOD mice, it converted the low-affinity TCR into a pathogenic one. In agreement with these data, polyclonal CD4+ T cells from prediabetic NOD mice promoted disease in high-affinity TCR-TG NOD.Rag2−/− mice, but inhibited it in low-affinity TCR-TG NOD.Rag2−/− mice. Thus, in chronic autoimmune responses, CD4+ Th cells contribute to both promoting and suppressing pathogenic autoimmunity.

Generation of productive CD8+ T cell responses requires the activation of APCs by CD4+ Th cells (1, 2) or pathogen-associated molecular patterns (3), a process referred to as APC licensing. Ligation of CD40 or TLRs on APCs by CD40L (CD154) on CD4+ Th cells or pathogen-associated molecular patterns, respectively, induces the upregulation of CD8+ T cell-costimulatory molecules and the secretion of proinflammatory cytokines (4–6). Most importantly, CD4+ Th cells play a critical role in the generation of functional memory CD8+ cells in response to a variety of foreign and self-Ags, such that unhelped memory CD8+ cells generated and maintained in the absence of CD4+ Th cells have significantly reduced long-term survival and effector function capabilities than their helped counterparts (7–11).

Type 1 diabetes (T1D) in both humans and NOD mice results from a chronic, CD4+ and CD8+ T cell-dependent autoimmune response against pancreatic β cells (12). We and others have shown that CD8+ cells play a critical role in initiation and progression of T1D (reviewed in Ref. 13). A significant fraction of islet-infiltrating CD8+ cells in NOD mice employ highly homologous TCRα rearrangements and recognize an epitope from islet-specific glucose-6-phosphatase catalytic subunit-related protein (IGRP206–214 or its mimotopes NRP-A7 and NRP-V7) in the context of the MHC class I molecule Kd (14, 15). This T cell subset, however, is not homogeneously pathogenic and consists of clones engaging peptide–MHC (pMHC) over a wide range of avidities, which correlate with the clones’ pathogenic activity (16–18). In fact, whereas high-avidity IGRP206–214-reactive CD8+ cells differentiate into diabetogenic CTL, their low-avidity counterparts differentiate into memory-like autoregulatory CD8+ cells with powerful antidiabetogenic properties (19).

We have previously shown that differentiation of high-avidity IGRP206–214-reactive CD8+ cells into diabetogenic CTL is a CD4+ Th-dependent process (14). In addition to upregulating the costimulatory properties of autoantigen-loaded dendritic cells (DCs), ligation of CD40 on DCs by CD154 on CD4+ Th cells releases immature DCs from suppression by Foxp3+CD4+ regulatory T cells (Tregs) (20). Because the autoregulatory CD8+ cells that arise from naive low-avidity precursors in response to chronic autoantigenic stimulation display a memory phenotype, we reasoned that their development might also require the licensing of autoantigen-loaded DCs (and their unlocking from Treg-mediated suppression) by CD4+ Th cells. In this study, we test this hypothesis by following the fate of high- and low-avidity IGRP206–214-reactive CD8+ cells in the presence or absence of CD4+ Th cells or surrogate Th-like DC-activation stimuli. Our data demonstrate that CD4+ Th cells promote the development of autoregulatory T cell memory.

Materials and Methods

Mice

NOD.Ltj mice were from The Jackson Laboratory (Bar Harbor, ME). 17.4α-8.3β-NOD and 17.6α-8.3β-NOD mice have been described (14, 19). Tcrα−/− and Rag2−/− TCR-TG NOD mice were generated by backcrossing the TCR transgenes onto NOD.Tcra−/− and NOD.Rag2−/− mice, respectively. All mice were housed and studied in specific pathogen-free facilities. These studies were approved by the University of Calgary Faculty of Medicine’s Animal Care Committee and followed the guidelines of Canadian Council on Animal Care.

Diabetes

Diabetes was monitored by measuring urine glucose level using Diastix (Bayer, Toronto, Ontario, Canada). Animals were considered diabetic after 2 consecutive d of glucosuria.

Flow cytometry and Abs

Processed thymi, spleens, and lymph nodes were stained at 4°C for 30 min in the dark. Stained samples were thoroughly washed and fixed with 1% paraformaldehyde until analysis. Three-color flow cytometry was performed using following mAbs: anti-CD8α–PerCP (53.6.7), anti-CD4–FITC (GK1.5), anti-Vβ8.1/8.2–PE (MR5-2), anti-CD69–PerCP.Cy5.5 (H1.2F3), anti-CD62L–PE (MEL-14) (all from BD Biosciences, San Jose, CA), anti-CD44–FITC (IM7.8.1) (Caltag Laboratories, Burlingame, CA), and anti-CD122–PE (5H4) (eBioscience, San Diego, CA).

In vitro suppression assays

CD8+ cells were purified from the spleens and lymph nodes of transgenic (TG) mice and FACS-sorted into CD44loCD122− or CD44hiCD122+ populations. Sorted cells (1.7–2.3 × 104 NRP-V7/Kd-tetramer+ cells) were cultured in U-bottom 96-well plates with NRP-A7–pulsed (1 μM) bone marrow-derived DCs (BM-DCs; 104). After 20 h of culture, purified 17.4α-8.3β–CD8+ cells (2 × 104) were labeled with CFSE (2.5 μM) and added to the wells. CFSE dilution was analyzed by FACS after 48 h. Percent suppression was calculated as the ratio of the difference in proliferation between responders without suppressors and responders with suppressors to the percentage of proliferation of responder cells without suppressors.

Adoptive transfer

CD4+ T cells and B cells were purified from the spleens of 7–14 wk old NOD.Ltj mice using BD IMag anti-mouse CD4 particles-DM (BD Biosciences) and anti-B220–coated magnetic beads (Miltenyi Biotec, Bergisch-Gladbach, Germany), respectively, according to the manufacturers’ instructions. Thirty million purified cells were transferred i.v. into 5–10-wk-old Rag2−/− TCR-TG mice. Mice were followed for diabetes development for 150 d. Some 17.6α-8.3β-NOD.Rag2−/− mice were bled on days 0, 8, and 18 to assess the percentage of circulating CD44hiCD122+CD8+ cells.

In vitro-activated CTLs were generated by culturing FACS-sorted CD44loCD122− or CD44hiCD122+ CD8+ cells from 17.6α-8.3β-NOD.Tcra−/− mice in the presence of NRP-A7–pulsed (1 μM) BM-DCs and 50 ng/ml recombinant human IL-2 (Takeda) for 3 d. A total of 2 × 106 cells were transferred into NOD.scid recipients and the host mice followed for diabetes development for 30 d.

mAb and peptide-pulsed DC treatment

BM-DCs were prepared by culturing hind leg bone marrow cells in complete media supplemented with recombinant murine GM-CSF (BD Biosciences), and recombinant murine IL-4 (R&D Systems, Minneapolis, MN) (5 ng/ml each) for 7 d. DCs were stimulated with LPS (1 μg/ml) during the last 18 h of culture, purified with anti-CD11c–coated beads (Miltenyi Biotec), and pulsed with 1–100 μM of peptide. TG mice received one or two i.v. injections of 2 to 3 × 106 peptide-pulsed DCs.

The FGK45 (anti-CD40) hybridoma was from Dr. A. Rolink (Department of Biomedicine, University of Basel, Basel, Switzerland); the 3H3 (anti-4-1BB) hybridoma was from Dr. R. Mittler (Emory University, Atlanta, GA). mAb treatment of TG animals involved three i.p. injections, 3 to 4 d apart, of 100 μg anti–4-1BB mAb or 100 μg anti-CD40 mAb. Mice were followed for at least 40 d after the first injection of mAbs or peptide-pulsed DCs for diabetes development.

Statistical analysis

Data were compared using log-rank, Mann–Whitney U, and two-way ANOVA tests. Statistical significance was assumed at p < 0.05.

Results

Diabetes suppression by memory-like autoregulatory CD8+ cells is resistant to strategies that bypass the need for CD4+ T cell help or that potentiate effector CTL responses

NOD mice expressing a low-affinity IGRP206-214-reactive TCR (17.6α/8.3β-NOD) are almost completely resistant to T1D (19). T1D resistance in these mice is not mediated by Foxp3+ CD4+ Tregs, but rather by a memory-like subset of antidiabetogenic CD8+ cells that arises from naive low-avidity precursors (19). This outcome was remarkably different from that seen in mice expressing a nearly identical TCR targeting the same pMHC complex with ∼10-fold higher affinity (17.4α/8.3β-NOD), which develop accelerated T1D (14, 19). Phenotypic analyses indicated that both types of TCRs foster the development and selection of CD8+ cells as compared with non-TG NOD mice (Fig. 1A). As expected, the percentage of CD8+CD4− thymocytes was significantly higher in mice expressing the high-affinity TCR (17.4α/8.3β) than in mice expressing its low-affinity counterpart (17.6α/8.3β) (p = 0.03; Fig. 1A). However, unlike the former, the latter accumulated a sizeable population of CD44hiCD122+CD8+ cells (Fig. 1B). This population primarily arose in the periphery, because thymic CD8+CD4− cells, unlike splenic CD8+ cells, did not harbor CD122+ cells (Fig. 1C). The increased frequency of CD44hiCD122+CD8+ T cells in the peripheral lymphoid organs of 17.6α/8.3β- versus 17.4α/8.3β-NOD mice could not be explained by selective pairing of the 17.6α-chain with an endogenous TCRβ-chain capable of driving the differentiation of the CD8+ T cells of these mice into memory-like CD8+ T cells because allelic exclusion of the endogenous TCRβ locus was very efficient in both TCR-TG strains. In fact, their spleen and pancreatic lymph nodes (PLNs) harbored very low percentages of CD8+ T cell expressing endogenous Vβ elements (0.63 ± 0 and 0.33 ± 0.01, respectively, for 17.4α/8.3β-NOD mice; 2.39 ± 0.31 and 0.28 ± 0.01, respectively, for 17.6α/8.3β-NOD mice). Furthermore, the percentages of CD8+ T cells expressing endogenous Vβ elements in 17.6α/8.3β-NOD mice were significantly lower than the percentage of memory-like autoregulatory CD8+ T cells (10.10 ± 0.66 and 5.02 ± 0.61 in spleen and PLNs, respectively). These memory-like CD8+ cells had regulatory properties, as they readily suppressed the activation of naive high-avidity autoreactive CD8+ cells by peptide-pulsed DCs in vitro (Fig. 1D).

FIGURE 1.
FIGURE 1.

Low-avidity IGRP206–214-reactive CD8+ cells spontaneously differentiate into memory-like autoregulatory cells. A, Representative FACS profiles of thymic and splenic T cells of non-TG NOD, 17.6α/8.3β-NOD, and 17.4α/8.3β-NOD mice. Data (mean ± SEM) correspond to 5–13 TCR-TG mice (7–14 wk old) (p values: 1 versus 2, 0.0210; 1 versus 3, 0.0159; 2 versus 3, 0.0300; 4 versus 5, 0.0084; 4 versus 6, 0.0035; 5 versus 6, 0.0021). B, Representative CD44 and CD122 staining profiles of CD8+ cells developing in 17.6α/8.3β- and 17.4α/8.3β-NOD mice (mean ± SEM). Sample sizes for 17.6α/8.3β-NOD: spleen, n = 11; PLN, n = 8. Sample sizes for 17.4α/8.3β-NOD: spleen, n = 6; PLN, n = 4. p values: 1 versus 2, 0.0005; 3 versus 4, 0.0081. Mice were 8–14 wk old. C, Representative CD122 staining (left panel) and average percentage of CD122+ cells (right panel) in thymic CD8+CD4− and splenic CD8+ T cell populations from 17.6α/8.3β-NOD mice (n = 4). D, Proliferation of CFSE-labeled 17.4α/8.3β-CD8+ cells in response to NRP-A7–pulsed BM-DCs in the presence of naive CD44loCD122− or memory-like CD44hiCD122+CD8+ cells from 17.6α/8.3β-NOD mice. Right panel represents percent suppression of proliferation averaged over four experiments.

We have previously demonstrated that CD4+CD25+ Treg-mediated suppression of T1D is also associated with suppression of autoantigen presentation by DCs in the PLNs (20, 21). Notably, ligation of CD40 on DCs by CD154 on CD4+ Th cells provides a signal that releases DCs from the suppressive activity of CD4+ Tregs (20). As a result, systemic activation of DCs with an agonistic anti-CD40 mAb, CpG DNA, or LPS was readily able to overcome suppression by CD4+CD25+ T cells, leading to T1D (20). To ascertain whether systemic activation of DCs or systemic delivery of activated Ag-pulsed DCs could also overcome tolerance mediated by memory-like autoregulatory CD8+ cells in 17.6α/8.3β-NOD mice, we treated these mice with an agonistic anti-CD40 mAb or with LPS-activated NRP-V7–pulsed DCs. We also asked if polyclonal activation of the CD8+ cells of these mice with an agonistic Ab against 4-1BB, a costimulatory molecule expressed primarily on effector CD8+ cells, could overcome their diabetes resistance [i.e., by promoting effector CD8+ cell survival and by increasing the IFN-γ secretory and cytolytic activities of these T cells, as described in other settings (22)]. Rat-IgG– and LPS-activated DCs pulsed with a T1D-irrelevant peptide (TUM) were used as negative controls. Additional controls involved the use of age- and sex-matched 17.4α/8.3β-NOD mice.

Administration of agonistic anti-CD40 or anti–4-1BB mAbs (but not rat-IgG) dramatically accelerated diabetes in 17.4α/8.3β-NOD mice (Fig. 2A). This was associated with upregulation of CD44 and CD69 and downregulation of CD62L on a significant fraction of splenic and lymph node CD8+ cells, as expected (Fig. 2B). Anti-CD40 and anti–4-1BB mAb treatment, however, failed to induce diabetes in 17.6α/8.3β-NOD mice (Fig. 2A). mAb treatment in these mice did not induce (in anti-CD40 mAb-treated mice), or did so only weakly (in anti–4-1BB mAb-treated mice), the upregulation of CD44 and CD69 or the downregulation of CD62L on CD8+ cells (Fig. 2C). Likewise, whereas a single dose of NRP-V7–pulsed DCs upregulated CD44 in a significant fraction of CD8+ cells (Fig. 2D) and rapidly induced diabetes in 17.4α/8.3β-NOD mice (Fig. 2A), NRP-V7–pulsed DCs could neither upregulate CD44 on CD8+ cells (Fig. 2E) nor trigger disease in any 17.6α/8.3β-NOD mice (Fig. 2A), even after a second dose.

FIGURE 2.
FIGURE 2.

17.6α/8.3β-NOD mice are resistant to strategies that bypass the need for CD4+ T cell help or potentiate effector CTL responses. A, 17.4α/8.3β- and 17.6α/8.3β-NOD mice were treated with anti-CD40 mAb (n = 7 and 5, respectively), anti–4-1BB mAb (n = 7 and 6, respectively), or LPS-activated NRP-V7–pulsed DCs (n = 4 and 3, respectively). As controls, 17.4α/8.3β- and 17.6α/8.3β-NOD mice received rat-IgG (n = 6) or were left untreated (n = 95), respectively, or were treated with LPS-activated TUM-pulsed DCs (n = 6 and 4, respectively). All mice were followed for at least 40 d for development of diabetes. Representative CD44, CD69, and CD62L stainings of splenic, mesenteric lymph node (MLN), or PLN CD8+ T cells in 17.4α/8.3β-NOD (B, D) and 17.6α/8.3β-NOD (C, E) mice that received anti-CD40 mAb, anti–4-1BB mAb, or NRP-V7–pulsed DCs compared with the same staining in mice that received rat IgG or TUM-pulsed DCs.

Taken together, these data indicate that the suppressive activity of the CD44hiCD122+CD8+ cells of 17.6α/8.3β-NOD mice cannot be overcome by strategies that mimic (hence bypass the need for) CD4+ T cell help, such as agonistic anti-CD40 mAb treatment and NRP-V7–DC transfer. The inability of anti–4-1BB mAb to activate CD8+ cells or trigger T1D in 17.6α/8.3β-NOD mice further indicates that these mice do not harbor preactivated diabetogenic CD8+ cells, suggesting that the memory-like autoregulatory CD8+ cells afford these mice a profound state of autoregulation.

In the absence of CD4+ T cells or B cells, naive low-avidity 17.6α/8.3β-CD8+ cells spontaneously differentiate into diabetogenic effectors

The above observations prompted us to consider the possibility that differentiation of naive 17.6α/8.3β-CD8+ cells into memory-like autoregulatory T cells might be a CD4+ Th cell-dependent/potentiated process and thus enhanced (rather than overcome) by the various Th-like stimuli employed above. This seemed a reasonable interpretation of the data given that CD4+ Th cells have been shown to play important roles both in the primary activation and differentiation of naive T cells into CTL and in the generation of functional CD8+ T cell memory against viral (8, 10), bacterial (9) and self-Ags (11).

We thus sought to investigate if development of the memory-like autoregulatory CD8+ cells of 17.6α/8.3β-NOD mice was a CD4+ T cell-dependent process. To this end, we followed the fate of the 17.6α/8.3β and 17.4α/8.3β TCR transgenes in TCR-TG NOD.Rag2−/− mice, which cannot express endogenous TCR or Ig rearrangements and therefore harbor monospecific CD8+ T cell repertoires devoid of CD4+ T cells and B cells. As expected, TG expression of the low-affinity 17.6α/8.3β TCR led to positive selection of CD8+ cells, although to a lesser extent than its high-affinity 17.4α/8.3β TCR counterpart (Fig. 3A). Nearly all thymic and peripheral CD8+CD4− cells in both types of mice expressed the TG TCRβ-chain and bound NRP-V7/Kd tetramers (Fig. 3A and data not shown). Furthermore, both strains lacked CD4+CD8- thymocytes as well as CD4+ T cells and B cells in their peripheral lymphoid organs (Fig. 3A and data not shown). We compared the development of memory-like autoregulatory CD8+ cells in Rag2- versus Tcra−/− TCR-TG mice because Tcra−/− TCR-TG mice have a more restricted TCRαβ repertoire than their Tcra-competent counterparts, and therefore (as is also the case in Rag2−/− TCR-TG mice), most of their peripheral CD8+ cells bind NRP-V7/Kd tetramers (data not shown). In addition, and despite their inability to express endogenous TCRα-chains, Tcra−/− 17.6α/8.3β- and 17.4α/8.3β-TCR-TG mice support the development of CD4+ cells expressing endogenous TCRs (Fig. 3B).

FIGURE 3.
FIGURE 3.

CD4+ T cell help is required for development and/or maintenance of functional memory-like autoregulatory CD8+ cells. A and B, Representative FACS profiles of splenic T cells of 17.6α/8.3β- and 17.4α/8.3β-NOD.Rag2−/− or NOD.Tcra−/− mice. Data (mean ± SEM) correspond to five to nine mice per strain (7–14 wk old) (p values: 1 versus 3, 0.0022; 5 versus 7, 0.0002; 2 versus 6, 0.0002; 4 versus 8, 0.0022). C and D, Disease survival curves of 17.6α/8.3β-NOD.Rag2−/− (n = 16) versus 17.4α/8.3β-NOD.Rag2−/− (n = 106) and 17.6α/8.3β-NOD.Tcra−/− (n = 14) versus 17.4α/8.3β-NOD.Tcra−/− (n = 28). E, Representative CD44 and CD122 profiles of splenic and PLN CD8+ cells developing in 17.6α/8.3β-NOD.Tcra−/− and 17.6α/8.3β-NOD.Rag2−/− mice. Percentages and absolute numbers of splenic (F) and PLN (G) CD44hi CD122+ CD8+ cells in 17.6α/8.3β-NOD.Tcra−/− and 17.6α/8.3β-NOD.Rag2−/− mice. Sample sizes for left panel in F: n = 13 for NOD.Tcra−/− and n = 8 for NOD.Rag2−/− TCR-TG mice. Sample sizes for right panel in F: n = 7 for NOD.Tcra−/− and n = 6 for NOD.Rag2−/− TCR-TG mice. Sample sizes for left panel in G: n = 9 for NOD.Tcra−/− and n = 7 for NOD.Rag2−/− TCR-TG mice. Sample sizes for right panel in G: n = 3 for NOD.Tcra−/− and n = 6 for NOD.Rag2−/− TCR-TG mice. H, Left panel, Proliferation of CFSE-labeled 17.4α/8.3β-CD8+ cells to NRP-A7–pulsed DCs in the presence of naive CD44loCD122− or memory-like CD44hiCD122+CD8+ cells from 17.6α/8.3β-NOD.Tcra−/− or 17.6α/8.3β-NOD.Rag2−/− mice. Right panel, Percent suppression of proliferation averaged over 24 experiments using 17.6α/8.3β-NOD.Tcra−/− CD8+ cells or five experiments using 17.6α/8.3β-NOD.Rag2−/− CD8+ cells. I, Survival curves of NOD.scid mice transferred with in vitro-activated CD44hiCD122+ (n = 4) and CD44loCD122− (n = 4) CD8+ cells from 17.6α/8.3β-NOD.Tcra−/− mice.

Unlike 17.6α/8.3β-NOD.Tcra−/− mice, 17.6α/8.3β-NOD.Rag2−/− mice developed diabetes spontaneously (Fig. 3C, 3D). In fact, the incidence and average age at onset of disease in 17.6α/8.3β-NOD.Rag2−/− mice were indistinguishable from those seen in 17.4α/8.3β-NOD.Rag2−/− mice (Fig. 3C). Remarkably, the peripheral lymphoid organs of 17.6α/8.3β-NOD.Rag2−/− mice contained significantly reduced percentages and absolute numbers of memory-like autoregulatory CD8+ cells (CD8+CD44hiCD122+) (Fig. 3E–G) as compared with their 17.6α/8.3β-NOD.Tcra−/− counterparts. Furthermore, whereas the CD8+ CD44hi CD122+ cells of 17.6α/8.3β-NOD.Tcra−/− mice could efficiently suppress the proliferation of 17.4α/8.3β-CD8+ cells in response to peptide-pulsed DCs in vitro, the few CD8+ CD44hi CD122+ cells arising in 17.6α/8.3β-NOD.Rag2−/− mice lacked such suppressive activity (Fig. 3H).

To confirm that the naive CD8+ cells of 17.6α/8.3β-NOD.Tcra−/− mice had intrinsic diabetogenic potential (rather than being irreversibly silenced), we compared the ability of CD8+CD122+ and CD8+CD122− cells sorted from 17.6α/8.3β-NOD.Tcra−/− mice to transfer diabetes into NOD.scid recipients after in vitro activation with NRP-A7–pulsed DCs. In agreement with the ability of the CD8+CD122+ cells of 17.6α/8.3β-NOD mice to suppress diabetes transfer by splenocytes from prediabetic NOD donors into NOD.scid hosts (19), all NOD.scid mice that were transfused with in vitro-activated CD8+ CD122+ cells remained diabetes free for at least 30 d after transfer. In contrast, all NOD.scid mice that received in vitro-activated CD8+CD122− cells from the same donors developed diabetes shortly after transfer (Fig. 3I).

Thus, in the presence (but not absence) of B cells and/or endogenous CD4+ T cells, naive 17.6α/8.3β-CD8+ cells differentiate into memory-like antidiabetogenic autoregulatory T cells rather than diabetogenic effectors. It should be noted that the diabetes resistance of 17.6α/8.3β-NOD.Tcra−/− mice (versus 17.6α/8.3β-NOD.Rag2−/− mice) cannot be attributed to the presence of Foxp3+CD4+ T cells because, as shown previously (19), only CD8+ but not CD4+ T cells from 17.6α/8.3β-NOD.Tcra−/− mice could inhibit the transfer of diabetes by NOD T cells into NOD.scid mice.

CD4+ T cell-assisted differentiation of high- versus low-avidity autoreactive CD8+ cells into diabetogenic effectors and antidiabetogenic suppressors, respectively

Collectively, and in the context of our previous observations (14), the above experiments revealed that abrogation of CD4+ T cell (and B cell) development has opposite effects in 17.4α/8.3β-NOD.Rag2−/− versus 17.6α/8.3β-NOD.Rag2−/− mice: it decelerates disease in the former but fosters it in the latter. This prompted us to consider the possibility that CD4+ Th cells can actively promote both the differentiation of high-avidity autoreactive CD8+ cells into diabetogenic effectors and the differentiation of their low-avidity counterparts into memory-like antidiabetogenic suppressors.

To test this hypothesis, we compared the effects of purified polyclonal CD4+ cells from prediabetic NOD mice on the natural history of diabetes in 17.4α/8.3β-NOD.Rag2−/− and 17.6α/8.3β-NOD.Rag2−/− mice. Whereas a single transfusion of CD4+ cells promoted diabetes development in 17.4α/8.3β-NOD.Rag2−/− mice (Fig. 4A), it decelerated its progression in 17.6α/8.3β-NOD.Rag2−/− mice (Fig. 4B). In contrast, adoptive transfer of purified B cells from prediabetic NOD mice into 17.6α/8.3β-NOD.Rag2−/− mice had no significant effects on the natural history of disease in these mice (Fig. 4C). Notably, the CD4+ T cell-induced delay in diabetes development in 17.6α/8.3β-NOD.Rag2−/− mice was associated with increased frequency of CD8+CD44hiCD122+ cells in peripheral blood (Fig. 4D). Thus, CD4+ Th cells can simultaneously support the differentiation of high-avidity autoreactive CD8+ cells into diabetogenic effectors and the differentiation of their low-avidity counterparts into memory autoregulatory T cells.

FIGURE 4.
FIGURE 4.

CD4+ cells promote the differentiation of high- and low-avidity autoreactive CD8+ cells into diabetogenic effectors and anti-diabetogenic suppressors, respectively. A, Disease survival curves of 17.4α/8.3β-NOD.Rag2−/− mice that were either untreated (n = 98) or received CD4+ T cells from prediabetic NOD mice (n = 7). Survival curves of 17.6α/8.3β-NOD.Rag2−/− mice that were either untreated (n = 16) or received CD4+ T cells (B; n = 8) or B cells (C; n = 5) from prediabetic NOD mice. D, Fold increase in memory-like autoregulatory CD8+ cells in the peripheral blood of 17.6α/8.3β-NOD.Rag2−/− mice that were untreated (n = 5) or received CD4+ T cells (n = 7) or B cells (n = 4) from prediabetic NOD mice. The p values were obtained using two-way ANOVA test comparing the differences between the percentages of memory-like autoregulatory CD8+ cells from 17.6α/8.3β-NOD.Rag2−/− mice that were untreated versus the mice that received CD4+ T cells or B cells at both 8- and 18-d time points.

Discussion

During the progression of T1D, low-avidity autoreactive CD8+ T cell clones are progressively replaced by their less prevalent but immunologically fitter high-avidity counterparts (16). Rather than playing a passive role in the disease process, low-avidity autoreactive CD8+ cells function as a source of autoantigen-specific negative-feedback regulatory loops that aim to counter disease progression. They do so by differentiating into subsets of memory-like autoregulatory CD8+ T cells that acquire the ability to kill and suppress autoantigen-loaded APCs in the pancreas-draining lymph nodes, thus blunting the activation of both cognate and noncognate pathogenic effectors simultaneously (19). In this study, we have shown that genesis of these antidiabetogenic memory-like autoregulatory CD8+ cells from naive low-avidity precursors in vivo is a CD4+ Th-dependent process. Because differentiation of naive high-avidity autoreactive CD8+ cells into diabetogenic effectors is also a CD4+ Th cell-dependent process, we conclude that CD4+ Th cells contribute to both promoting and extinguishing diabetogenic autoimmunity.

The general requirement for CD4+ T cell help in the priming of CD8+ T cell responses against foreign Ags and autoantigens is well documented (23). For example, Rag2−/− 17.4α/8.3β-NOD mice, expressing a monoclonal repertoire of high-avidity IGRP206–214-reactive CD8+ T cells completely devoid of CD4+ T cells and B cells, developed a significantly reduced incidence (and a decelerated form) of diabetes as compared with their Rag2-competent counterparts (14). In addition, adoptive transfer of CD4+ T cells from prediabetic NOD mice into these mice restored the incidence and age at onset of diabetes to the levels seen in the Rag2-competent animals (14). We have further shown that interactions between CD154 on CD4+ Th cells and CD40 on DCs play a critical role in this process by releasing immature DCs from the dominant-suppressive effects of Foxp3+CD4+ Tregs (in the steady state) and by promoting their ability to prime diabetogenic CD8+ T cell responses (20). DC stimuli such as agonistic anti-CD40 or TLR ligands like LPS or CpG DNA could readily overcome the need for CD154-dependent CD4+ T cell help, suggesting that Treg-dependent suppression of DC maturation is a reversible process controlled by CD4+ Th signals and innate stimuli (20). In contrast, agonistic anti-CD40 and anti–4-1BBL mAbs and peptide-pulsed DCs could not overcome the diabetes resistance of 17.6α/8.3β-NOD mice, suggesting that suppression of Ag presentation by memory-like autoregulatory CD8+ cells is more profound than that mediated (in the steady state) by Foxp3+ CD4+ Tregs. Furthermore, adoptive transfer of CD4+ T cells from wild-type NOD mice into 17.6α/8.3β-NOD.Rag2−/− hosts, which spontaneously develop T1D, increased the frequency of memory-like autoregulatory T cells in peripheral blood and decelerated disease development. Because these studies involved a single injection of polyclonal CD4+ T cells, and because peripheral blood is the initial place were changes in the frequency of memory-like autoregulatory CD8+ T cells in mice treated with pMHC-coated NPs are noted (19), it is not surprising that polyclonal CD4+ T cell transfer did not also result in increased frequencies of memory-like autoregulatory CD8+ T cells in spleen or PLNs.

We note that these observations are reminiscent of the need for CD4+ T cell help in effector memory CD8+ T cell formation: CD8+ T cell memory against viral (7, 8, 10), bacterial (9), and self-Ags (11) in the absence of CD4+ cells is defective, resulting in reduced survival of memory CD8+ cells and impaired responsiveness to secondary antigenic challenge. The CD4+ Th cell signals responsible for promoting autoregulatory CD8+ T cell memory formation remain unclear, but might be similar to those involved in effector T cell memory formation (24–27). It has also been suggested that unhelped memory CD8+ T cells tend to commit suicide by secreting TRAIL in response to a secondary antigenic challenge (28). Whether the few unhelped memory-like CD8+ cells of Rag2−/− 17.6α/8.3β-NOD mice produce TRAIL in response to antigenic challenge has yet to be explored but is a possibility. Another noteworthy similarity between the unhelped memory-like CD8+ cells of Rag2−/− 17.6α/8.3β-NOD mice and the unhelped effector memory T cells described in other models lies in their impaired functional activities. Because the suppressive activity of helped memory-like autoregulatory CD8+ cells is mediated via IFN-γ–dependent inhibition of Ag presentation and perforin-mediated killing of APCs (19), it is conceivable that the reduced suppressive activity of their unhelped counterparts is due to defects in these pathways, a phenomenon described for the unhelped memory CD8+ T cells generated in response to pathogens (8–10). Interestingly, the defects in IFN-γ production and cytolysis observed in these studies were not seen in the primary effector response. This observation is consistent with our finding that CD4+ T cells are dispensable for activation of naive 17.6α/8.3β-CD8+ cells into effector CTL, both in vivo and ex vivo.

Although a limitation of these studies is that they are done in oligoclonal TCR-TG systems, following the developmental biology and fate of specific autoregulatory CD8+ T cell subsets in non-TG mice is a difficult, if not impossible, task. For example, individual autoantigenic specificities in non-TG mice are very small in size and phenotypically heterogeneous (containing naive, effector, and autoregulatory T cells over a spectrum of avidities). Furthermore, because NOD CD8+ T cells cannot cause β cell loss (hence autoantigen shedding) in the absence of CD4+ T cells, naive low-avidity CD8+ T cells would not be able to differentiate into memory-like autoregulatory cells in a CD4+ T cell-deficient environment, regardless of the role of CD4+ T cell help in autoregulatory CD8+ T cell development. Because of these limitations, studies on the role of CD4+ Th cells in the homeostasis of autoregulatory CD8+ T cells in non–TCR-TG systems are not currently feasible.

In sum, this study establishes a previously unrecognized, ambiguous role for CD4+ T cell help in chronic autoimmune responses: it fosters the diabetogenic potential of high-avidity autoreactive CD8+ cells while boosting the antidiabetogenic properties of their low-avidity counterparts by promoting the generation and/or survival and suppressive function of memory-like autoregulatory CD8+ cells. This discovery challenges the generally held view that CD4+ T cell help invariably promotes effector T cell responses.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank S. Thiessen, M. Khanbabaei, and M. DeCrom for technical assistance and L. Kennedy and L. Robertson for flow cytometry.

Footnotes

  • This work was supported by the Canadian Institutes of Health Research, the Juvenile Diabetes Research Foundation, the Natural Sciences and Engineering Research Council of Canada, and the Canadian Diabetes Association. A.S. is supported by a studentship from Alberta Innovates-Health Solutions. X.C.-C. is supported by an AXA Foundation scholarship. J.W. is supported by a fellowship from the Canadian Diabetes Association. P.S. is a Scientist of Alberta Innovates-Health Solutions and a Juvenile Diabetes Research Foundation Scholar. The Julia McFarlane Diabetes Research Centre is supported by the Diabetes Association (Foothills).

  • Abbreviations used in this article:

    BM-DC
    bone marrow-derived dendritic cell
    DC
    dendritic cell
    IGRP
    islet-specific glucose-6-phosphatase catalytic subunit-related protein
    PLN
    pancreatic lymph node
    pMHC
    peptide–MHC
    T1D
    type 1 diabetes
    TG
    transgenic
    Treg
    regulatory T cell.

  • Received April 18, 2011.
  • Accepted July 9, 2011.

References

  1. ↵
    1. Bennett S. R.,
    2. F. R. Carbone,
    3. F. Karamalis,
    4. J. F. Miller,
    5. W. R. Heath
    . 1997. Induction of a CD8+ cytotoxic T lymphocyte response by cross-priming requires cognate CD4+ T cell help. J. Exp. Med. 186: 65–70.
    OpenUrlAbstract/FREE Full Text
  2. ↵
    1. Ridge J. P.,
    2. F. Di Rosa,
    3. P. Matzinger
    . 1998. A conditioned dendritic cell can be a temporal bridge between a CD4+ T-helper and a T-killer cell. Nature 393: 474–478.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Janeway C. A., Jr..,
    2. R. Medzhitov
    . 2002. Innate immune recognition. Annu. Rev. Immunol. 20: 197–216.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Bennett S. R.,
    2. F. R. Carbone,
    3. F. Karamalis,
    4. R. A. Flavell,
    5. J. F. Miller,
    6. W. R. Heath
    . 1998. Help for cytotoxic-T-cell responses is mediated by CD40 signalling. Nature 393: 478–480.
    OpenUrlCrossRefPubMed
    1. Schoenberger S. P.,
    2. R. E. Toes,
    3. E. I. van der Voort,
    4. R. Offringa,
    5. C. J. Melief
    . 1998. T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40L interactions. Nature 393: 480–483.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Grewal I. S.,
    2. R. A. Flavell
    . 1998. CD40 and CD154 in cell-mediated immunity. Annu. Rev. Immunol. 16: 111–135.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Belz G. T.,
    2. D. Wodarz,
    3. G. Diaz,
    4. M. A. Nowak,
    5. P. C. Doherty
    . 2002. Compromised influenza virus-specific CD8(+)-T-cell memory in CD4(+)-T-cell-deficient mice. J. Virol. 76: 12388–12393.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Janssen E. M.,
    2. E. E. Lemmens,
    3. T. Wolfe,
    4. U. Christen,
    5. M. G. von Herrath,
    6. S. P. Schoenberger
    . 2003. CD4+ T cells are required for secondary expansion and memory in CD8+ T lymphocytes. Nature 421: 852–856.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Sun J. C.,
    2. M. J. Bevan
    . 2003. Defective CD8 T cell memory following acute infection without CD4 T cell help. Science 300: 339–342.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Shedlock D. J.,
    2. H. Shen
    . 2003. Requirement for CD4 T cell help in generating functional CD8 T cell memory. Science 300: 337–339.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Bourgeois C.,
    2. H. Veiga-Fernandes,
    3. A. M. Joret,
    4. B. Rocha,
    5. C. Tanchot
    . 2002. CD8 lethargy in the absence of CD4 help. Eur. J. Immunol. 32: 2199–2207.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Babad J.,
    2. A. Geliebter,
    3. T. P. DiLorenzo
    . 2010. T-cell autoantigens in the non-obese diabetic mouse model of autoimmune diabetes. Immunology 131: 459–465.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Santamaria P.
    2010. The long and winding road to understanding and conquering type 1 diabetes. Immunity 32: 437–445.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Verdaguer J.,
    2. D. Schmidt,
    3. A. Amrani,
    4. B. Anderson,
    5. N. Averill,
    6. P. Santamaria
    . 1997. Spontaneous autoimmune diabetes in monoclonal T cell nonobese diabetic mice. J. Exp. Med. 186: 1663–1676.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Lieberman S. M.,
    2. A. M. Evans,
    3. B. Han,
    4. T. Takaki,
    5. Y. Vinnitskaya,
    6. J. A. Caldwell,
    7. D. V. Serreze,
    8. J. Shabanowitz,
    9. D. F. Hunt,
    10. S. G. Nathenson,
    11. et al
    . 2003. Identification of the beta cell antigen targeted by a prevalent population of pathogenic CD8+ T cells in autoimmune diabetes. Proc. Natl. Acad. Sci. USA 100: 8384–8388.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    1. Amrani A.,
    2. J. Verdaguer,
    3. P. Serra,
    4. S. Tafuro,
    5. R. Tan,
    6. P. Santamaria
    . 2000. Progression of autoimmune diabetes driven by avidity maturation of a T-cell population. Nature 406: 739–742.
    OpenUrlCrossRefPubMed
    1. Han B.,
    2. P. Serra,
    3. J. Yamanouchi,
    4. A. Amrani,
    5. J. F. Elliott,
    6. P. Dickie,
    7. T. P. Dilorenzo,
    8. P. Santamaria
    . 2005. Developmental control of CD8 T cell-avidity maturation in autoimmune diabetes. J. Clin. Invest. 115: 1879–1887.
    OpenUrlCrossRefPubMed
  18. ↵
    1. Han B.,
    2. P. Serra,
    3. A. Amrani,
    4. J. Yamanouchi,
    5. A. F. Marée,
    6. L. Edelstein-Keshet,
    7. P. Santamaria
    . 2005. Prevention of diabetes by manipulation of anti-IGRP autoimmunity: high efficiency of a low-affinity peptide. Nat. Med. 11: 645–652.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Tsai S.,
    2. A. Shameli,
    3. J. Yamanouchi,
    4. X. Clemente-Casares,
    5. J. Wang,
    6. P. Serra,
    7. Y. Yang,
    8. Z. Medarova,
    9. A. Moore,
    10. P. Santamaria
    . 2010. Reversal of autoimmunity by boosting memory-like autoregulatory T cells. Immunity 32: 568–580.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Serra P.,
    2. A. Amrani,
    3. J. Yamanouchi,
    4. B. Han,
    5. S. Thiessen,
    6. T. Utsugi,
    7. J. Verdaguer,
    8. P. Santamaria
    . 2003. CD40 ligation releases immature dendritic cells from the control of regulatory CD4+CD25+ T cells. Immunity 19: 877–889.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Tang Q.,
    2. J. Y. Adams,
    3. A. J. Tooley,
    4. M. Bi,
    5. B. T. Fife,
    6. P. Serra,
    7. P. Santamaria,
    8. R. M. Locksley,
    9. M. F. Krummel,
    10. J. A. Bluestone
    . 2006. Visualizing regulatory T cell control of autoimmune responses in nonobese diabetic mice. Nat. Immunol. 7: 83–92.
    OpenUrlCrossRefPubMed
  22. ↵
    1. Sun Y.,
    2. J. H. Chen,
    3. Y. Fu
    . 2004. Immunotherapy with agonistic anti-CD137: two sides of a coin. Cell. Mol. Immunol. 1: 31–36.
    OpenUrlPubMed
  23. ↵
    1. Kurts C.,
    2. F. R. Carbone,
    3. M. Barnden,
    4. E. Blanas,
    5. J. Allison,
    6. W. R. Heath,
    7. J. F. Miller
    . 1997. CD4+ T cell help impairs CD8+ T cell deletion induced by cross-presentation of self-antigens and favors autoimmunity. J. Exp. Med. 186: 2057–2062.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Bourgeois C.,
    2. B. Rocha,
    3. C. Tanchot
    . 2002. A role for CD40 expression on CD8+ T cells in the generation of CD8+ T cell memory. Science 297: 2060–2063.
    OpenUrlAbstract/FREE Full Text
    1. de Goër de Herve M. G.,
    2. B. Dembele,
    3. M. Vallée,
    4. F. Herr,
    5. A. Cariou,
    6. Y. Taoufik
    . 2010. Direct CD4 help provision following interaction of memory CD4 and CD8 T cells with distinct antigen-presenting dendritic cells. J. Immunol. 185: 1028–1036.
    OpenUrlAbstract/FREE Full Text
    1. Williams M. A.,
    2. A. J. Tyznik,
    3. M. J. Bevan
    . 2006. Interleukin-2 signals during priming are required for secondary expansion of CD8+ memory T cells. Nature 441: 890–893.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Oh S.,
    2. L. P. Perera,
    3. M. Terabe,
    4. L. Ni,
    5. T. A. Waldmann,
    6. J. A. Berzofsky
    . 2008. IL-15 as a mediator of CD4+ help for CD8+ T cell longevity and avoidance of TRAIL-mediated apoptosis. Proc. Natl. Acad. Sci. USA 105: 5201–5206.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Janssen E. M.,
    2. N. M. Droin,
    3. E. E. Lemmens,
    4. M. J. Pinkoski,
    5. S. J. Bensinger,
    6. B. D. Ehst,
    7. T. S. Griffith,
    8. D. R. Green,
    9. S. P. Schoenberger
    . 2005. CD4+ T-cell help controls CD8+ T-cell memory via TRAIL-mediated activation-induced cell death. Nature 434: 88–93.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section