Abstract
The improved efficacy of high avidity CTL for clearance of virus has been well-documented. Thus, elucidation of the mechanisms that confer the increased sensitivity to peptide ligand demonstrated by high avidity CTL is critical. Using CTL lines of high and low avidity generated from a TCR transgenic mouse, we have found that functional avidity can be controlled by the expression of CD8αα vs CD8αβ and the ability of CTLs to colocalize the TCR and CD8 in the membrane. Colocalization of these molecules was mediated by lipid rafts and importantly, raft disruption resulted in the conversion of high avidity CTL into a lower functional avidity phenotype. These novel findings provide insights into the control of functional avidity in response to viral infection.
The amount of peptide Ag that is required to elicit effector function is a reflection of the functional avidity of a CTL. High-avidity CTLs are responsive to relatively low amounts of peptide presented by MHC class I molecules, whereas low-avidity CTLs require substantially higher levels of peptide/MHC. The biological relevance of this increased sensitivity to peptide Ag has been demonstrated in vivo. Our previous work has shown that following adoptive transfer of established CTL lines, only high avidity CTL reduced viral burden (1). In view of the differential in vivo efficacy of high vs low avidity CTL, it is critical that we develop a detailed understanding of the mechanisms that regulate CTL responsiveness to peptide Ag.
The ability of CTLs to respond functionally to Ags displayed on the surface of APCs depends on a number of molecular interactions. Tetramer binding studies have shown that in some cases the affinity of the TCR for peptide/MHC complexes correlates with functional avidity (2, 3, 4). However, it should be noted that this correlation does not hold true in all cases. In addition, the levels of CD8 and LFA-1 have been shown to influence the amount of peptide/MHC required for activation (5, 6, 7, 8, 9).
Using a lymphocytic choriomeningitis virus (LCMV)3 TCR transgenic (P14) mouse crossed onto a recombinase-activating gene-2−/− background, we demonstrated that CTLs of high and low functional avidity can be generated independently of TCR affinity (10). Low avidity CTLs, although expressing a TCR identical with high avidity CTLs, required more TCR engagement events than high avidity CTLs to become activated. This appeared to be the result of TCR engagement events that do not produce a productive signal.
Furthermore, we made two new important observations concerning the role of CD8 and functional avidity. First, although the surface expression of CD8α was similar between high and low avidity CTLs, CD8β expression was significantly increased in high avidity CTLs. These data are consistent with the notion that high avidity CTLs express a greater proportion of CD8αβ heterodimers relative to low avidity CTLs. Next, we demonstrated that low avidity CTLs were less efficient than high avidity CTLs at cointernalizing the CD8 molecule with the TCR. Because the cointernalization of CD8 with TCR following CD3 cross-linking occurred in the absence of engagement of CD8 with MHC class I molecules, we concluded that the TCRs on the surface of high avidity CTLs are more highly associated with the CD8 molecule than on low avidity CTLs. The increased internalization of CD8 with TCR demonstrated by high avidity CTLs and the ability to respond functionally to fewer TCR engagement events suggested that the higher functional avidity apparent in these CTLs may be a direct result of the organization of these molecules within membrane microdomains.
Studies by Zhang et al. (18) have shown that TCR-mediated signaling is enhanced upon oligomerization (11). This aggregation of TCR resulting in increased signaling efficiency may be influenced by the recruitment of these receptors into lipid rafts upon engagement with peptide/MHC molecules on APCs (12). In addition to TCR, a number of signaling molecules have been shown to reside in lipid rafts, including p59fyn, CD8, the linker of activation of T cells, and p56lck (LCK) (12, 13, 14, 15, 16, 17, 18). The kinase p56lck serves to phosphorylate ITAMS of the CD3ζ chains. Once phosphorylated, these tyrosine residues serve as “docking sites” for molecules containing src homology domain 2 binding motifs, such as ZAP-70 and p59fyn (19, 20), further amplifying signal transduction cascades. Additionally, Lck has src homology domain 2 domain, and can bind to phosphorylated ZAP-70 associated with the CD3ζ chain, effectively coupling the coreceptor to the TCR (21). Indeed, the cointernalization of TCR and CD8 results from the association of coreceptor-associated Lck to ZAP-70/CD3ζ of triggered TCR (22). Given our previous findings that high avidity CTLs are more sensitive to TCR engagement and are better able to cointernalize CD8 with the TCR (10), we hypothesized that high avidity CTLs would demonstrate a greater degree of colocalization of TCR with CD8, and that this colocalization would be mediated in part by organization of these molecules into lipid rafts.
The following study uses high and low avidity lines generated from TCR transgenic mice to examine the role of membrane microdomain organization of CD8 and TCR in determining the functional avidity of CTL. We report the novel finding that high and low avidity CTLs differ dramatically in the spatial arrangement of CD8 and TCR on their surface. Specifically, high avidity CTLs colocalize substantially more TCR with CD8 compared with low avidity CTLs, which fail to associate a significant portion of TCR following CD8 capping. The efficiency with which high avidity CTLs cocap TCR with CD8 is abolished following treatment with methyl-β-cyclodextrin (MBCD), indicating that the arrangement is mediated by lipid rafts. Importantly, the ability of high avidity CTLs to respond functionally to fewer TCR engagement events than low avidity CTLs is directly related to the integrity of lipid rafts on their surface, as treatment with MBCD decreases the apparent functional avidity of CTL. These data support a mechanism whereby high functional avidity is determined by the optimal membrane localization of TCR and CD8, which is mediated by compartmentalization of these molecules into lipid rafts.
Materials and Methods
Mice and cell lines
C57BL/6 mice were purchased from Frederick Cancer Research and Development Center (Frederick, MD). The TCR LCMV P14/recombinase-activating gene-2 mice were obtained from Taconic Farms (Germantown, NY). The LCMV P14 peptide (KAVYNATM) encompasses residues 33–41 of the gp33 protein and was synthesized at the Comprehensive Cancer Center Protein Analysis Core Laboratory at Wake Forest University School of Medicine (Winston-Salem, NC).
Generation of CTL lines
For CTL lines generated from transgenic mice, 2 × 106 spleen cells were cocultured with 3.5 × 106 C57BL/6 splenocytes (2000 rad irradiated) previously pulsed with either high (10−5 M) or low (10−10 M) concentrations of LCMV (P14) peptide. Before coculture, stimulators were washed to remove unbound peptide. Cultures were maintained in 24-well plates containing 2 ml of RPMI 1640 medium supplemented with 2 mM l-glutamine, 0.1 mM sodium pyruvate, nonessential amino acids, 100 U/ml penicillin, 100 μg/ml streptomycin, 2-ME (0.05 mM), 10% FBS, and 10% T-stim (Collaborative Biomedical Products, Bedford, MA) as an IL-2 source. CTL lines were established from primary cultures and maintained by weekly restimulation of 3–5 × 105 cells/well in the presence of 5 × 106 irradiated (2000 rad) C57BL/6 spleen cells pulsed with the appropriate concentration of peptide.
Flow cytometry and confocal microscopy
For intracellular staining, cells were fixed in 3% paraformaldehyde solution for 20 min at room temperature and permeabilized with a solution of 1% saponin, 10% FBS, and 0.1% NaN3 in PBS on ice for 30 min. Before adding primary Ab, cells were blocked with 5% normal goat serum for 20 min on ice to prevent nonspecific binding of Abs. Cells were stained with anti-Lck Ab (BD Transduction Laboratories, San Diego, CA) for 20 min on ice. Cells were washed and stained with the appropriate FITC-conjugated secondary Abs (Caltag Laboratories, Burlingame, CA). Lipid raft ganglioside GM1 was labeled using FITC-conjugated cholera toxin B subunit (CT-X; Sigma-Aldrich, St. Louis, MO). Expression of CD8α and CD8β were measure by flow cytometry using an anti-CD8α (clone CT-CD8a) or anti-CD8β (clone CT-CD8b) FITC-conjugated Ab (Caltag Laboratories). For CD8α/CD8β cocapping studies, cells were labeled with FITC-conjugated anti-CD8β Ab (clone CT-CD8b; Caltag Laboratories) for 30 min on ice followed by washing and cross-linking with anti-rat IgG for 30 min at 37°C to induce capping. Cells were washed in ice-cold PBS containing sodium azide stained with Cy5-conjugated anti-CD8α (clone 53-6.7; BD PharMingen, San Diego, CA), and fixed for analysis by confocal microscopy. In Fig. 2⇓, the CD8α is depicted in green and the CD8β in red. These changes in representation were made to maintain consistency with the subsequent figures in which CD8α is depicted in green. For TCR/CD8 cocapping studies, cells were incubated for 30 min on ice with PE-conjugated anti-TCR Ab (clone H57-597; Caltag Laboratories) and FITC-conjugated rat anti-mouse CD8α (clone CT-CD8a; Caltag Laboratories). Cells were washed in cold PBS supplemented with 2% FBS. CD8α cross-linking was done on ice by adding goat anti-rat IgG, followed by washing and further cross-linking with rabbit anti-goat IgG on ice. Cells were then resuspended in 37°C cell culture media and incubated in a 5% CO2 incubator for 20 min to induce capping. Cells were pelleted and resuspended in ice-cold PBS containing 0.2% sodium azide. Cells were fixed to poly(L) lysine-coated coverslips and fluorescence measured by confocal microscopy. The goat anti-rat secondary, used to cap the CD8α Ab, was tested to ensure that it was not capable of binding the anti-TCR Ab. For cocapping studies following raft disruption, CTLs were treated with 5 mM MBCD as described below and subjected to the same capping method as above.
Disruption of lipid rafts
Seven days after routine stimulation, CTLs were washed twice in PBS and resuspended in serum-free OPTI-MEM (Life Technologies, Rockville, MD) containing MBCD (Sigma-Aldrich, St. Louis, MO) at a concentration of 5 mM. Following a 30-min incubation at 37°C, cells were washed three times in PBS and resuspended in OPTI-MEM for use in the IFN-γ ELISA or CD8/TCR cocapping studies.
IFN-γ ELISA
Seven days following routine stimulation, CTLs (5 × 104/well) were either treated with MBCD or left untreated and incubated in 96-well flat-bottom plates that had previously been incubated overnight at 4°C with various concentrations of anti-CD3 Ab (clone 2C11; BD PharMingen) followed by PBS washing. CTLs were cultured for 18 h at 37°C in a 5% CO2 incubator serum-free OPTI-MEM. Supernatant was harvested and assayed for the presence of IFN-γ by ELISA (OptEIA Mouse IFN-γ set; BD PharMingen).
Results
Differences in functional avidity cannot be explained by disparate expression levels of Lck
The activation of a CTL depends on the phosphorylation of ITAMs in the CD3ζ chains by src family kinases. The 56-kDa kinase Lck plays a significant role in these phosphorylation events, and signal transduction resulting in the internalization of the TCR is dependent on p56lck activity (23). Our previous studies have shown that high avidity CTLs down-regulate surface TCR at lower concentrations of anti-CD3 Ab compared with low avidity CTLs (10). Presumably, an increase in the expression of Lck could translate into an increased ability to phosphorylate the ITAMs of the TCR following engagement, which could in turn result in an increased efficiency of TCR down-regulation following engagement with anti-CD3 Ab. Thus, we tested the hypothesis that TCR signal transduction is more efficient in high avidity CTL due to increased Lck expression. Intracellular levels of Lck were measured by flow cytometry, and Fig. 1⇓ shows that the expression of Lck between high and low avidity CTL was nearly identical. These findings were reproduced in an additional pair of high and low avidity CTL lines. Therefore, differential expression of Lck does not explain the increased sensitivity to CD3 engagement in high avidity CTL.
High and low avidity CTLs express similar levels of Lck. On day 7 poststimulation, high (bold histogram) and low (solid histogram) avidity CTLs were permeabilized and stained for intracellular Lck. Negative control histograms for low avidity (dotted histogram) and high avidity (dashed histogram) CTL lines are shown. Data are representative of three independent experiments.
High and low avidity CTLs differ in the ability to colocalize TCR and CD8
The above data demonstrate that the overall level of expression of Lck does not explain the difference in sensitivity to TCR engagement between high and low avidity CTLs, but did not preclude the possibility that the localization of Lck might influence the efficiency of ITAM phosphorylation. One mechanism by which Lck is localized to the TCR complex is by associating with the coreceptor CD8, which binds the α3 domain of the MHC class I molecule. The coordinate binding of both CD8 and TCR with the same MHC class I molecule allows for efficient localization of CD8-associated Lck with the TCR. In addition, TCR/coreceptor colocalization brings p56lck into proximity of TCR-associated kinases such as fyn and Zap-70, which facilitate kinase-dependent signal transduction cascades (20, 24).
In an earlier study, we found that the increased efficiency of anti-CD3-triggered cointernalization of CD8 with TCR was due at least in part to the increased expression of CD8αβ heterodimers vs αα homodimers by high avidity CTL, as CD8 molecules expressing the β-chain were selectively cointernalized with the TCR (10). This increased coreceptor internalization in the absence of ligand binding provides strong functional evidence for an increased association of CD8 and TCR in high avidity CTL relative to low avidity CTL. As previously demonstrated (10), the high and low avidity CTL lines generated from TCR transgenic mice express similar levels of CD8α, but differed dramatically in the expression of CD8β. In Fig. 2⇓A, low avidity CTLs express ∼40% less CD8β than high avidity CTLs, yet express similar levels of CD8α. Although expression levels of CD8α and CD8β by flow cytometry were consistent with the notion that there were more CD8αα homodimers on low avidity CTL relative to high avidity CTL, this technique was insufficient to provide direct evidence of a differential expression of CD8αα and CD8αβ isoforms.
High avidity CTLs express CD8 predominantly in the αβ heterodimeric form while low avidity CTLs express both CD8αβ and CD8αα on their surface. A, On day 7 poststimulation, expression levels of CD8α and CD8β were determined for both high (red histogram) and low (black histogram) avidity CTL lines by FACS analysis. B and C, Confocal fluorescence microscopy. Low avidity (B) and high avidity (C) CTLs were labeled with anti-CD8β Ab on ice. The anti-CD8β Ab was cross-linked with anti-isotype Abs and transferred to 37°C for 30 min to induce capping. Cells were washed with ice-cold PBS containing sodium azide, stained with anti-CD8α Ab, fixed and analyzed by confocal microscopy. CD8α is represented in green and capped CD8β is represented in red. CD8αβ heterodimers are represented in yellow. The presence of CD8α that is not colocalized with CD8β indicates the presence of CD8αα homodimers.
To demonstrate the presence of CD8αα homodimers in low avidity CTL and to determine whether CD8αβ heterodimers vs CD8αα homodimers could be differentiated within the intact membrane, CD8 molecules expressing the β-chain were fluorescently labeled and capped, and the distribution of CD8α and CD8β were visualized by confocal fluorescence microscopy. In Fig. 2⇑, B and C, the staining patterns of CD8α molecules (green) and CD8β molecules (red) are shown for low (Fig. 2⇑B) and high (Fig. 2⇑C) avidity CTLs. Under these conditions, CD8αβ heterodimers are identified by colocalization of CD8α and CD8β (yellow). CD8α is found in low avidity CTL in both capped and uncapped regions. In contrast, all of the CD8α in high avidity CTL is found colocalized with CD8β. These data provide direct evidence that high avidity CTLs express CD8 on their surface predominantly in the CD8αβ heterodimeric form, whereas a substantial portion of surface CD8 molecules on low avidity CTLs are in CD8αα homodimeric form.
Because our previous findings showed that high avidity CTLs were better able to cointernalize CD8 molecules with TCR in the absence of MHC engagement (10), we hypothesized that CD8 was more highly associated with the TCR in high avidity CTLs. The ability of molecules to cocap has been used previously as a measure of their association in the membrane. Studies by Kwan Lim et al. (25) have shown that capping with Abs to the CD8αβ heterodimer resulted in the efficient cocapping of TCR. Thus, we tested the ability of TCR and CD8 to cocap in both high and low avidity CTLs. Importantly, the colocalization of CD8 and TCR was independent of its engagement with peptide/MHC. In Fig. 3⇓, the CD8 molecule is shown in green and the TCR in red for both low (Fig. 3⇓A) and high (Fig. 3⇓B) avidity CTLs. Two important findings are readily apparent. First, the morphology of the cocapped regions differs between high and low avidity CTLs. The low avidity CTLs appear to have incomplete capping (a pattern designated as patching by others), while the high avidity CTLs show a more concentrated or uniform CD8/TCR cocap. These data suggest that the cytoskeletal-dependent organization of membrane receptors may be suboptimal in the low avidity CTL. Second, the ability of the TCR molecule to cocap with CD8 is reduced in low avidity CTL. Although a large portion of the TCR complexes are colocalized with CD8 (yellow), a significant amount of TCR does not colocalize with the capped CD8 molecules, suggesting that engagement of these TCR would not lead to efficient signal transduction. In contrast, the high avidity CTLs demonstrate an increased ability to cocap the TCR with the CD8 molecule, with nearly all of the TCR colocalizing with CD8. One would predict that this close association of TCR and CD8 independent of TCR/CD8 interaction with class I MHC would be very efficient for signal transduction.
Cocapping studies demonstrate that low avidity CTLs fail to colocalize TCR with CD8 as efficiently as high avidity CTLs. On day 7 poststimulation, low (A) and high (B) avidity CTLs were labeled with FITC-conjugated anti-CD8α and PE-conjugated anti-TCR (clone H57-597) on ice. Anti-CD8α Ab was cross-linked with anti-isotype Abs on ice, washed, and transferred to 37°C for 30 min to induce capping. Cells were fixed and analyzed by confocal microscopy. CD8α is represented in green and TCR is represented in red. Colocalized CD8 and TCR are shown in yellow. Expression levels of TCR and CD8α were similar between high and low avidity CTLs as determined by flow cytometry (data not shown).
To quantitate the frequency of cells with colocalized vs noncolocalized TCR and CD8 molecules, a series of confocal images were collected for a representative high and low avidity CTL line. These images were then coded, randomized, and evaluated by two independent investigators in an effort to reduce bias. Individual cells were scored as either colocalized or noncolocalized, and the results were tabulated. The results from these studies are summarized in Table I⇓. Approximately 2% of the confocal images generated from the low avidity CTL line were scored as colocalized and appeared similar to those in Fig. 3⇑A. Conversely, 91% of the images taken of the high avidity CTL line were scored as colocalized, appearing similar to those in Fig. 3⇑B. These data demonstrate a different spatial arrangement of TCR and CD8 within the plasma membrane between high and low avidity CTLs.
Colocalization of TCR and CD8 for high and low avidity CTL
Raft disruption abolishes CD8/TCR cocapping and shifts the phenotype of high avidity CTL to a lower functional avidity
The importance of lipid rafts in facilitating the optimal orientation of signaling molecules is well-documented (16, 26, 27, 28, 29). The reduced cocapping of CD8 and TCR observed in low avidity CTL might be explained by a decrease in the expression of lipid rafts on the surface of these cells. To test this hypothesis, the expression level of the ganglioside GM1 was measured using FITC-labeled CT-X. GM1 is preferentially associated with lipid rafts and is considered to be a raft marker (30). As shown in Fig. 4⇓, high and low avidity CTLs bind CT-X with similar efficiency, suggesting that a difference in the expression of lipid rafts is not responsible for the differences in capping. However, these results did not preclude the possibility that the organization of lipid rafts or the arrangement of molecules within lipid rafts might differ between high and low avidity CTLs, resulting in the differential display of TCR and CD8 on the cell surface.
High and low avidity CTLs bind CT-X equally. On day 7 poststimulation, high (bold histogram) and low (solid histogram) avidity CTLs were incubated with FITC-conjugated CT-X to measure surface expression of lipid rafts. Similar binding of CT-X suggests that high and low avidity CTLs do not differ significantly in the expression of lipid rafts on their surface. These data are representative of three independent experiments. Negative controls for high (dotted histogram) and low (dashed histogram) avidity CTLs are shown.
To test this hypothesis, we evaluated the impact of raft disruption on the colocalization of CD8 and TCR following CD8 cross-linking. Treatment of cells with MBCD has been shown to disrupt lipid rafts by depleting cholesterol from the membrane (31). High and low avidity CTLs were treated with MBCD, and staining and cocapping were performed as in Fig. 4⇑. Following cholesterol depletion with MBCD, both high and low avidity CTLs display a nearly uniform distribution of TCR and CD8 on their cell surface (Fig. 5⇓). These data show that the ability to cocap TCR with CD8 has been completely abrogated, and thus support the hypothesis that the colocalization of TCR following capping of CD8 is mediated by lipid rafts.
Treatment with MBCD abrogates CD8/TCR cocapping in high and low avidity CTLs. On day 7 poststimulation, low avidity (A) or high avidity (B) CTLs were treated with 5 mM MBCD to disrupt lipid rafts by cholesterol depletion. Cells were washed, placed on ice, and cocapping was performed as in Fig. 4⇑. CD8α (green) and TCR (red) are randomly distributed on the cell surface of both high and low avidity CTLs, and no distinct capped regions are evident. Data are representative of two independent experiments.
Given the apparent increased association of CD8 with the TCR in high vs low avidity CTLs shown in Fig. 4⇑, it seemed likely that this increased colocalization could be responsible for the increased sensitivity of high avidity CTL to TCR engagement. Optimal localization of Lck and TCR would allow for improved efficiency of phosphorylation of TCR ITAMs, translating into increased sensitivity to TCR engagement. Therefore, we hypothesized that disruption of lipid rafts (and thus colocalization of CD8 with TCR) in high avidity CTLs would induce a lower avidity functional phenotype. The effect on low avidity CTLs should be less significant, as they already demonstrated a reduced ability to colocalize CD8 with TCR. To test this hypothesis, we measured the effect of MBCD treatment on the ability of high and low avidity CTLs to produce IFN-γ in response to anti-CD3 stimulation. Treatment of high avidity CTLs with MBCD shifted the dose-response curve toward that of the low avidity CTLs (Fig. 6⇓). High avidity CTLs now required 8-fold more anti-CD3 Ab to induce half-maximal IFN-γ production (average of three experiments). Treatment of low avidity CTLs with MBCD had only a minimal effect, shifting the dose-response curve by only 1.5-fold (average of three experiments). The viability of both lines was 95% in the presence of MBCD. Therefore, the differential effect on the dose-response curves was not the result of a difference in the toxicity of the drug in high vs low avidity lines. Similar shifts in dose response were seen following treatment with nystatin, which also disrupts lipid raft integrity (data not shown). These data demonstrate that high avidity CTLs are much more sensitive to raft disruption compared with low avidity CTLs. These results support a mechanism in which high avidity CTLs display TCR and CD8 on their surface in an array that is favorable for optimal signaling following TCR engagement. Disruption of lipid raft integrity abrogates the optimal positioning of TCR and CD8, resulting in a decreased efficiency of signaling through the TCR, which translates into a lower functional avidity phenotype.
Raft disruption results in a lower avidity functional phenotype. On day 7 poststimulation, low (dotted lines) or high (solid lines) avidity CTLs were treated with 5 mM MBCD (○) for 30 min as in Fig. 5⇑, or left untreated (•). Cells were then plated in serum-free media (Opti-MEM; Life Technologies) in flat-bottom 96-well plates that had previously been coated with titrated concentrations of anti-CD3 Ab. Cells were then placed in a 37°C 5% CO2 incubator overnight. Supernatants were harvested at 18–20 h and IFN-γ was measured by ELISA. Data are representative of three independent experiments.
Discussion
Although the relative independence of CD8 engagement for high avidity CTLs has been well-established, the basis for the reduced requirement for CD8 engagement is not fully understood. One mechanism postulated to explain the relative CD8 independence of high avidity CTLs is the expression of a TCR with high affinity, thus reducing the need for the increased adhesion provided by the engagement of CD8 with MHC. The data presented herein using high and low avidity CTL lines generated from TCR transgenic mice provide a novel mechanism that could contribute to the decreased requirement of high avidity CTLs for CD8 engagement. Our data suggest that high avidity CTLs are able to reach their threshold of activation in response to lower levels of TCR engagement due in part to the colocalization of CD8 and the TCR in lipid rafts on their surface. Although there is presently no direct evidence that this mechanism is used in nontransgenic CTL lines, we have identified high and low avidity CTL lines generated from nontransgenic mice that differ in their sensitivity to TCR engagement, similar to what was observed with the TCR transgenic lines. The mechanism responsible for this difference in sensitivity is currently under investigation.
Our previous work demonstrated that high and low avidity CTLs differed in the expression levels of CD8β but not CD8α expression. These data were consistent with the notion that high avidity CTLs express CD8 predominantly as CD8αβ heterodimers, whereas low avidity CTLs express a significant population of CD8 as CD8αα homodimers. In the present study, using confocal microscopy we formally demonstrate that high avidity CTLs display CD8 on their surface predominantly in the CD8αβ heterodimeric form, whereas the low avidity CTLs express a significant portion of CD8 in the CD8αα homodimeric form. The expression of CD8αα homodimers on the surface of low avidity CTLs has a direct effect on the localization of Lck, which binds to the cytoplasmic tail of the CD8α-chain. Whereas both CD8αα and CD8αβ molecules are able to bind Lck, only CD8αβ molecules are localized into lipid rafts by virtue of the palmitoylation site found in the cytoplasmic tail of CD8β (13). Therefore, the form in which CD8 is expressed (CD8αα or CD8αβ) can have a dramatic impact on the localization of CD8-associated Lck relative to raft-associated TCR.
We hypothesize that the preferential expression of CD8 as αβ heterodimers allows for efficient localization of CD8-associated Lck with raft-resident TCR resulting in a high functional avidity phenotype. This hypothesis is supported by our finding that treatment with MBCD, which completely abrogated the cocapping of TCR and CD8 for both high and low avidity CTLs, resulted in an 8- to 10-fold increase in the amount of anti-CD3 Ab required for 50% maximal IFN-γ production in high avidity CTLs, while shifting the dose-response curve for low avidity CTLs only minimally (1- to 2-fold more anti-CD3 Ab). These studies provide functional evidence that the optimal localization of TCR and CD8 can enhance CTL functional avidity by improving the efficiency of signaling. This localization is highly dependent on the organization of these molecules within lipid rafts. These findings are in agreement with a recent study by Arcaro et al. (32), which demonstrated that in T cell hybridomas, CD8β couples the TCR to raft-associated CD8/p56lck complexes. Specifically, the portion of the CD8β molecule containing the palmitoylation site that allows for CD8β to associate with lipid rafts was shown to be critical for efficient TCR/CD8 association. These findings are also consistent with our CD8 cointernalization studies (10), which demonstrated that CD8 molecules expressing the β-chain were selectively internalized with TCRs that had transduced a signal. The present study extends our understanding of the functional consequences of the increased association of CD8αβ and TCR by establishing a correlation between the raft-dependent localization of these two molecules and the sensitivity to TCR engagement (functional avidity).
Importantly, the morphology of cocapped TCR and CD8 was also found to differ between high and low avidity CTLs. High avidity CTLs displayed very discrete, concentrated caps that were highly polarized. The capping of CD8 and TCR on low avidity CTLs appears much less polarized, with patches of colocalization appearing in a comparatively random distribution on the cell surface. The clustering of lipid rafts on the surface of T cells is known to be dependent on actin cytoskeleton reorganization (33, 34). Therefore, the differences in the morphology of capped CD8/TCR indicate that high avidity CTLs demonstrate an enhanced cytoskeletal rearrangement of these raft-resident molecules into more defined polarized regions relative to low avidity CTLs, resulting in more efficient propagation of signals. Another possible outcome of an increased ability to concentrate TCR and coreceptor into highly polarized caps is the increased phosphorylation of the ITAMs on unengaged TCR. Such “bystander” phosphorylation events may serve to amplify signaling through the TCR when TCR ligand is limiting.
Because the preferential expression of CD8 as an αβ heterodimer in high avidity CTLs correlated with an increased ability to associate with the TCR, we predicted that disruption of lipid raft integrity would decrease the ability of TCR to cocap with CD8 and induce a lower avidity phenotype. We found that treatment with MBCD completely abrogated the cocapping of TCR and CD8 for both high and low avidity CTLs. Importantly, these results demonstrated that the ability to efficiently concentrate TCR and CD8 into distinct caps is dependent upon the organization of these molecules into lipid rafts. Furthermore, by disrupting the higher order organization of these molecules through cholesterol depletion, the functional response of high avidity CTLs was dramatically attenuated. High avidity CTLs now display a lower functional avidity, requiring significantly more TCR engagement events following treatment with MBCD to exert effector function (8-fold more anti-CD3 Ab). These studies provide functional evidence that the optimal localization of TCR and CD8 can enhance CTL functional avidity by improving the efficiency of signaling. This localization is highly dependent on the organization of these molecules within lipid rafts.
The ability to regulate functional avidity by the organization of molecules in membrane microdomains or the expression of CD8αα vs CD8αβ opens the possibility that functional avidity may be an inducible phenomenon. Currently, it is unknown whether the functional avidity of a CTL is induced in response to environmental signals during the initial activation of a naive precursor in vivo or whether our ability to generate CTL lines of distinct avidity is a reflection of the preferential expansion of a subpopulation of CTL in vitro. Under conditions in which peptide Ag is abundant, CD8β expression may occur at basal levels. However, when the amount of Ag is very high or when signaling through the TCR is very strong, CD8β expression could be decreased, effectively inducing a lower avidity functional phenotype. Another possibility is that the initial encounter with an APC can influence the organization of molecules within membrane microdomains. Lipid rafts have been shown to play an important role in the efficient presentation of peptide/MHC class II by APCs (26). Thus, it is tempting to speculate that the organization of peptide/MHC on the APC is influencing the initial activation of a CTL such that a defined organization of membrane receptors is established that influences the subsequent response of that CTL.
In summary, the data herein provide a novel mechanism to explain the increased sensitivity to TCR engagement demonstrated by our high avidity CTLs generated from TCR transgenic mice. By expressing CD8 predominantly in the αβ heterodimeric form rather than the αα homodimeric form, high avidity CTLs display CD8 and TCR with greater colocalization relative to low avidity CTLs. This preferential association of CD8 and TCR effectively brings CD8-associated Lck into proximity of the TCR on high avidity CTLs. The enhanced colocalization of CD8 and TCR on high avidity CTLs depends on the integrity of lipid rafts; however, capping studies suggest that the cytoskeletal reorganization of these two molecules differs as well, with high avidity CTLs forming more polarized caps relative to low avidity CTLs. These data extend our understanding of mechanisms that contribute to the control of functional avidity by demonstrating that functional avidity can be influenced by the form of CD8 (αα homodimeric or αβ heterodimeric), the manner in which TCR and CD8 are organized on the cell surface, and the efficiency of cytoskeletal rearrangement of these two molecules. The advances in our understanding of avidity provided by the results presented herein are of significant importance, given the crucial role for functional avidity in determining the outcome of viral infection.
Acknowledgments
We thank Drs. Griffith Parks and Steven Mizel for critical reading of this manuscript.
Footnotes
-
↵1 This work was supported by National Institutes of Health Grant AI43591 (to M.A.A.-M.). A.G.C. was supported by National Research Training Award Grant AI07401.
-
↵2 Address correspondence and reprint requests to Dr. Martha A. Alexander-Miller, Department of Microbiology and Immunology, Wake Forest University School of Medicine, Room 5108, Gray Building, Medical Center Boulevard, Winston-Salem, NC 27157. E-mail address: marthaam{at}wfubmc.edu
-
↵3 Abbreviations used in this paper: LCMV, lymphocytic choriomeningitis virus; MBCD, methyl-β-cyclodextrin; CT-X, cholera toxin B subunit.
- Received May 22, 2002.
- Accepted July 22, 2002.
- Copyright © 2002 by The American Association of Immunologists
References
In this issue
