| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Immunology |
Departments of 1 Gynecologic Oncology, 2 Breast Medical Oncology, and 3 Immunology, University of Texas M.D. Anderson Cancer Center, Houston, Texas; 4 Department of Surgery, Uniformed Services University of the Health Sciences, Bethesda, Maryland; and 5 Infectious Disease Research Institute, Seattle, Washington
Requests for reprints: Constantin G. Ioannides, Department of Gynecologic Oncology, University of Texas M.D. Anderson Cancer Center, Unit 440, 1515 Holcombe Boulevard, Houston, TX 77030. Phone: 713-792-2849; Fax: 713-792-3575; E-mail: cioannid{at}mdanderson.org.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
+ Perf– and to TCRhi IFN-+ Perf+ cells. Blocking caspase-9 activation at antigen stimulation also enhanced the generation of TCRhi Perfhi cells, demonstrating that TCR density dictated the pathway of death activated by stimulation with the same agonist. Expansion and differentiation of TCRhi Perf+ CTL required an agonist of optimal CH2 side chain length, which in this study was equal to two CH2 groups appended to E75 at the Gly4 position. Side chains one CH2 shorter or longer than optimal were either less stimulatory or induced death of TCRhi Perf+ cells. Differentiation of TCRhi CD8+ cells can be finely tuned by synthetic amino acids in the peptide, whose side chains induce small increments in the affinity of the antigen for TCR below the affinity which induce apoptosis.
Key Words: CTL • perforin • TCR • synthetic amino acids • HER-2
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To elucidate the significance of TCR density in the differentiation of TCRhi cells to cytolytic effectors, we modified the affinity of the ligand (antigen) for the receptor (TCR) and simultaneously analyzed the responses of two polyclonal CD8+ populations from the same individual that differed in the levels of TCR by one order of magnitude. Because the transition between mitosis and apoptosis in TCRhi cells responding to tumor antigen takes place within a narrow range of antigen concentrations (5 x 10–6 to 10–7 mol/L; ref. 3), we modified the affinity of the antigen for TCR in the smallest possible changes (increments/decrements), using only atomic van der Waals forces from methylene (CH2) groups appended to Gly. At only 2 kJ/mol (0.5 kcal/mol), van der Waals forces are the weakest forces between atoms; by comparison, the force involved in the formation of one hydrogen bond after the introduction of one hydroxyl group is 20 kJ/mol (4). Because longer and branched side chains produce steric hindrance and steric hindrance may offset gains from increments in van der Waals forces, we used only short linear CH2 extensions.
We used CTLs isolated from tumor-associated lymphocytes (TAL) that recognize an epitope from the HER-2/neu proto-oncogene, which is present on normal epithelial cells but is also overexpressed in many epithelial cancers of the breast, lung, prostate, and ovary (5). Because HER-2 is a self-antigen, most T cells of high affinity for HER-2 epitopes are deleted (6), and the remainder cells recognize HER-2 peptides with low avidity (at about 10–6 mol/L). Nevertheless, because the same HER-2 CTL epitopes are presented by substantial proportions of tumors, the CTL epitopes from HER-2 become significant for cancer therapy (7).
How undifferentiated TCRhi cells respond to human tumor antigen and what is needed to induce their proliferation and differentiation to functional effector cells remain unknown. To identify the optimal agonist for inducing differentiation of TCRhi cells, we constructed four variants of E75, the HER-2 (369-377) epitope for CTLs, by appending one, two, three, or four methylene groups to the glycine molecule at position 4 (Gly4) to form a linear C side chain. We then selected cells expressing high concentrations of TCRs for E75 (E75-TCRhi) to evaluate the role of TCR density in CTL differentiation upon stimulation with the same ligands. The TCRhi population which usually includes cells staining with antigen-tetramers/dimers with a mean fluorescence intensity (MFI) higher than 102 was divided in two populations, one staining with antigen-pulsed HLA-A2/IgG dimers (dimers) with a MFI (TCR) between 102 and 103, and another which stained with antigen-pulsed dimers with a MFI (TCR) between 103 and 104. These populations were designated as TCRmed and TCRhi, respectively.
The ability of CTLs to synthesize perforin is critical to their ability to lyse target cells (8, 9). Presumably, cells with high-density TCR would need correspondingly high amounts of effector molecules such as perforin for maximum functionality. Perforin is undetectable in naive CTLs but is up-regulated in response to signals from TCRs (8–10). Perforin also controls CD8+ homeostasis independently of its role as an effector molecule (11–13). Large expansion of antigen-specific CD8+ cells that produce IFN- in perforin-deficient mice resulted in lethal disease during viral infections (11, 14). Most T cells with high affinity for self-antigen are deleted upon encountering that self-antigen. Deletion constitutes a mechanism for protection against autoimmunity (6, 15). We sought to determine how to induce differentiation of such cells and to identify the factors controlling their differentiation and survival. The difficulty in addressing these questions is compounded in polyclonal human systems, not only because surviving TALs recognize antigen with low functional avidity because their perforin expression is impaired (16, 17), but also because a large number of TCRs with distinct affinities for antigen are present.
The variant G4.2 (i.e., E75 with two CH2 groups appended at Gly4) was more effective in inducing perforin of E75-TCRhi cells than were variants with longer and shorter (G4.1) CH2 side chains. Moreover, differentiation of TCRhi IFN-– Perf– cells to TCRhi IFN-+ Perf+ cells by G4.2 involved an intermediate step to TCRhi IFN-+ Perf– cells. TCRhi IFN-+ Perf– cells proliferated and differentiated in response to the variant G4.2 to Perf+ cells, which mediated tumor lysis. In contrast, E75-TCRmed IFN-– Perf– cells primed with variants differentiated directly to TCRmed IFN-+ Perf+ cells. Higher TCR density was associated with greater resistance to differentiation to Perf+ cells. Our results indicate that the density of TCR in conjunction with the optimal affinity of the ligand, can avoid induction of death. The ability to finely tune the affinity of antigen for TCR in relation to receptor TCR density should be important for novel strategies to eliminate tumor cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
at levels increasing in tandem with the length of the CH2 chain, indicating that the E75-TCR+ cells in the TAL-1 population were functional and not tolerized (data not shown). Monoclonal antibody (mAb) used for detecting surface antigen and cytokines used for culturing T cells from TAL-1 are described elsewhere (18). APC-conjugated antibody to IFN- (IgG2a), PE-conjugated mouse (IgG2b) antibody to perforin (
G9), empty recombinant soluble dimeric human HLA-A2/IgG1 designated as "dimers," and all specific isotype immunoglobulin controls were obtained from BD PharMingen (San Diego, CA). Proteasome and caspase inhibitors were obtained from R&D Systems (Minneapolis, MN).
Synthetic peptides. The peptides used were E75 (HER-2 [369-377], KIFGSLAFL; ref. 19) and its methylene-appended C side chain variants. Four CH2 variants were generated by substituting the glycine at position 4 (Gly4) with alanine and one of following synthetic amino acids: -aminobutyric acid, norvaline, or norleucine (Advanced Chemtech, Louisville, KY). The abbreviations for the CH2-extended E75 variants (G4.1, G4.2, G4.3, and G4.4) reflect the position (Gly4) and number of the CH2 group extensions (19). Another variant in which the alanine at position 7 (Ala7) was replaced with norleucine was designated A7.3. All peptides were prepared by Dr. Martin Campbell (Peptide Synthesis Core Facility of the University of Texas M.D. Anderson Cancer Center). Amino acids were coupled in sequential format from the COOH terminus using standard N-(9-fluorenyl)methoxycarbonyl peptide chemistry on a Rainin Symphony Automated Peptide Synthesizer and purified by high-performance liquid chromatography. The purity of the peptides ranged from 95% to 97%. Peptides were dissolved in PBS and stored at –20°C as aliquots of 2 mg/mL until use. Molecular modeling of peptide/HLA-A2 complexes was done using as model the crystal structure of Tax peptide bound to HLA-A2 (18).
T cell/peptide-HLA-A2 dimer association and dissociation assays. Expression of TCRs specific for HLA-A2 bound to the E75 peptide (E75-TCR+ cells) was determined by using E75 dimers (dE75). dE75 were prepared as previously described (18). Staining of cells was done as described previously (19–21). Specific geometric (y2) MFI (TCR) for each peptide was calculated by substracting the MFI (TCR) of cells stained with "empty dimers" (dNP) from the MFI (TCR) of cells stained with dE75, dG4.1, dG4.2, dG4.3, and dG4.4. To identify the changes in the affinity of variants for TCR, the y2 (MFI) of TAL-1 cells stained with each peptide dimer, were determined immediately after staining (t0) or 2 hours later (20–23). Empty dimers were prepared in the same conditions as dE75 dimers except that no peptide was added to the HLA-A2-IgG dimers. The increase or decrease in the MFI (TCR) induced by each CH2 group extending the CH2 chain relative to E75 was calculated by subtracting the MFI (E75-TCR) from MFI (TCR) of cells stained with each variant peptide (G4.1, G4.2, G4.3, and G4.4) and dividing the result by the number of CH2 groups appended to each variant (1–4). These values were designated as MFI (TCR) per appended CH2. Among cells staining positive with peptide bound to HLA-A2/IgG dimers, populations were considered to express TCR at low density (TCRlo) if the geometric (y2) MFI for cells staining with that dimer was between 101 and 102, at medium density (TCRmed) if the MFI was between 102 and 103, and at high density (TCRhi) if the MFI was between 103 and 104.
T-cell stimulation by the CH2 variants. Apoptosis was induced in TAL-1 by two successive stimulations with 10,000 nmol/L of each peptide pulsed on T2 cells. Surviving TAL-1 were then primed with 5,000 nmol/L of each peptide pulsed on irradiated T2 cells as described elsewhere (24, 25). Control cultures were stimulated with T2 cells that had been pulsed with peptide A7.3. The death-resistant cells were restimulated in vitro with irradiated T2 cells pulsed with each peptide at a final concentration of 5 µmol/L, expanded in IL-2 (150 IU/mL; C1/A1) and maintained over feeders. Expression of IFN- and perforin, markers of differentiation, was determined by using IFN--APC-conjugated or perforin-PE-conjugated antibodies and matched PE/FITC/APC-conjugated isotype controls on dE75-stained and permeabilized cells (18). Fold expansion by each variant was calculated by dividing the number of E75-TCRhi/TCRmed cells detected in each sample after stimulation by the number of E75-TCRhi/TCRmed cells present before stimulation.
CTL assays. E75-, A7.3-, and G4.1- to G4.4-stimulated TAL-1 were used as effectors in CTL assays. Antigen recognition by the E75 variant–induced CTLs was determined as described elsewhere (26, 27). Recognition of E75 was considered specific when the mean specific lysis of T2 cells pulsed with E75 minus the SD was at least 10% and was at least twice as high as the percentage of specific lysis of T2 cells that had not been pulsed with peptide, plus the SD (27) E75-specific tumor lysis was determined by subtracting the levels of SKOV3.A2 tumor lysis observed in the presence of T2-E75 cells from the levels of SKOV3.A2 tumor lysis observed in the presence of T2-NP cells (27). The tumor cells were then incubated with 10 µmol/L MG132 (28, 29) for 30 minutes, before and during labeling, and used as targets in CTL assays. High- and medium-avidity effector CTLs were distinguished in two ways: first, by their ability to recognize E75 at concentrations at least two times lower than the E75-primed CTL (e.g., at 500 nmol/L instead of at 1,000 nmol/L) and second by the ability of high-avidity CTLs to mediate an effector response (e.g., % specific lysis) that was (a) at least twice as high as the effector response at the same or lower antigen concentration and/or (b) at half the effector-to-target ratio of the medium-avidity CTLs (30).
Caspase inhibitors. The caspase inhibitors Z-IETD-fluoromethyl ketone-(fmk) (specific for caspase-8), Z-LEHD-fmk (specific for caspase-9), and EDVE-fmk (specific for caspase-3), have been reported to participate in perforin-mediated apoptosis (31). For these experiments, 2 x 106 G4.2-induced CTL were incubated with each caspase inhibitor at 37°C for 90 minutes, washed twice with PBS, and stimulated with T2 cells pulsed with 5 µmol/L of G4.2.
Statistical analysis. Differences in the levels of IFN-, perforin, and cytolysis between groups were compared by using unpaired Student's t tests from triplicate determinations. Differences were considered significant at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
– Perf– TAL-1 with E75 or its Gly4 variants induced their differentiation to IFN-+ Perf– cells. TAL- 1 contained significant proportions of ex vivo activated and differentiated T cells. TCRhi and TCRmed cells were of similar size (mean forward scatter, 630), indicating that they were activated, not resting, cells and that the higher TCR density was not related to cell size. Representative results for TCRhi and med expression, cell size, and staining for perforin-positive (Perf+) and IFN-+ cells before and after stimulation are shown in the appendices. To recapitulate the process of differentiation, we first deleted differentiated cells (i.e., TCRhi and TCRmed Perf+ IFN-+ cells) by stimulating TAL-1 twice with 10,000 nmol/L E75, G4.1, G4.2, G4.3, or G4.4 pulsed on T2 cells to ensure that most (if not all) TCRhi Perf+ IFN-+ cells were eliminated. E75 and all of the Gly4 variants with higher affinity for TCR than E75 deleted most of the TCRhi and med Perf+ cells (Fig. 2A) and most of the IFN-+ cells (also appendices). A7.3, the CH2 position control peptide for G4.4, also induced deletion of TCRhi Perf+ cells (Fig. 2A). Differences between the small numbers of surviving cells were not significant. Compared with unstimulated IL-2-cultured TAL-1, the numbers of TCRhi Perf+ cells decreased insignificantly after stimulation with control T2 cells, which present only few endogenous self-peptides.
|
Priming with variants resulted in a modest increase in E75-TCRhi cells. Exception made priming with G4.3 and wild-type E75. However, the resulting numbers of E75-TCRhi cells stimulated by E75 or G4.3 increased by <2-fold compared with cultures which were not stimulated with peptide (group NP). All E75-TCRhi cells were IFN-+; there were no IFN-+ Perf+ cells (Fig. 2B). The few cells (<5% of total) in E75-, A7.3-, and G4.1-primed cultures which expressed perforin were IFN-–.
Most E75-TCRmed Perf+ cells (95-98%) were also IFN-+ indicating that the same variants induced differentiation of a significant part of TCRmed cells (Fig. 2C). IFN-+ Perf+ cells were 36%, 68%, 50%, and 57% in cultures stimulated with G4.1, G4.2, G4.3, and G4.4, respectively, indicating that G4 variants induced complete differentiation of a part of E75-TCRmed cells.
It is unclear whether the E75-TCRhi Perf+ cells derived from the few Perf+ cells which survived apoptosis or from the Perf– cells which differentiated in response to antigen stimulation. It is evident that stimulation with variants had different effects than stimulation with wild-type E75, in perforin and IFN- induction, in E75-TCRhi cells and E75-TCRmed cells. The results in Fig. 2B and C indicate that CH2 appendage was effective in inducing differentiation of both TCRhi and TCRmed cells, but the effects differed, depending on the position of the appendage (G4 or A7) and the density of the TCR. We quantitated the effects of stimulation with variants of polyclonal populations only on the cells expressing E75-TCR. The effects of the variants on cells expressing specific TCR for the variants, or reacting with the variants with higher affinity than with E75 have not been determined. Therefore, it may not be excluded that cells (e.g., E75-TCRmed) are also G4.2-TCRhi cells. The patterns of specific MFI (TCR) for variants-dimer complexes show parallels (Fig. 1F) with the E75-TCRmed Perf+ cells (Fig. 2C).
To identify the effects of restimulation in E75-TCR+ cell differentiation, all cultures were restimulated with T2 cells pulsed with the same amounts of priming peptide, as in vaccination studies. Results in Fig. 2D show a significantly higher increase in the numbers of E75-TCRhi cells in cultures stimulated with G4.1, G4.2, and G4.4 compared with E75-TCRmed cells in the same cultures. E75-TCRhi cells increased by 6- and 7-fold, respectively, in cultures restimulated by G4.1 and G4.4 and by >20-fold in cultures stimulated by G4.2. E75-TCRhi cells stimulated by E75 and G4.3 increased less in numbers and became apoptotic. E75-TCRmed cells expanded less at restimulation with E75, G4.1, and G4.2 (3-, 4-, and 5-fold, respectively) but did not expand at restimulation with G4.3 and G4.4. Because these populations are polyclonal, we cannot distinguish whether expansion of TCRhi cells was due to a higher rate of division of only TCRhi cells or to a 3- to 4-fold increase in the numbers of E75-TCR molecules per cell in populations of E75-TCRmed cells, or both. The increase in E75-TCRhi cells in cells stimulated with G4.1, of lower affinity for TCR than E75, is surprising and suggests a process similar with homeostatic proliferation and differentiation induced by low affinity ligands.
Perforin expression has been associated with terminal differentiation of CTL and their effector function (32–34). To address whether E75-TCR+ cell expansion affected their state of differentiation, we determined the levels of expression of perforin in these cells. Results in Fig. 3A show that E75-TCRhi cells which lacked perforin after priming, expressed high levels of perforin after restimulation. Representative histograms (G4.2 restimulated) are shown in Supplemental Material. In contrast, E75-TCRmed cells which expressed higher levels of perforin at priming, expressed reduced levels of perforin at restimulation. Because both E75-TCRhi and E75-TCRmed cells were present in the same culture, were stimulated in the same conditions, and were stained in the same tube, the results indicate that E75-TCRmed cells expressing higher levels of perforin were deleted at restimulation. Contraction of the Perf+ population may account for weaker expansion of TCRmed cells restimulated with G4.1 and G4.2 and for lack of expansion of TCRmed cells restimulated with G4.3 and G4.4.
|
CTLs induced with G4.2 lyse tumor cells. To address whether variant activated CTL had higher functional avidity for E75 than E75-activated CTL, we assessed the recognition of E75 by variant-induced CTLs. In the first experiment (Fig. 3B), effectors were variant-primed cells, and only the TCRmed populations expressed perforin. In the brief (4-hour assay), only G4.4-CTLs significantly recognized E75 (>10% lysis) at concentrations of 100 and 500 nmol/L, whereas lysis by G4.2 CTL was below the 10% cutoff level to be considered significant. The levels of perforin were slightly higher in E75-TCRmed cells stimulated with G4.2 and G4.4 than in E75-TCRmed cells stimulated with G4.1 and G4.3 (Fig. 3A). The levels of IFN- (G4.2: MFI (IFN-) = 25.9, G4.4 MFI (IFN-) = 18.23), and E75-TCR (G4.2: MFI (y2) = 212, G4.4 (MFI (y2) = 158) were similar in G4.2 and G4.4 primed cells. Thus, the results suggest a better "fit" between E75 presented by the target and the TCRmed of CTL primed by G4.4 than the TCRmed primed by G4.2, or G4.3. E75-CTL did not recognize E75 at this concentration, but rather required 2,500 nmol/L of the peptide for lysis to be detected (data not shown). In the extended (22 hours) CTL assay, variant-induced CTLs mediated similar levels of lysis of T2-E75 (additional data in the appendix). This suggested that the affinity of TCR for the ligand and the TCR-peptide-MHC conformational fit and not the small differences in perforin expression determined the effectiveness of these CTL.
To determine whether expression of perforin in E75-TCRhi cells increased the functional avidity of the effectors, we repeated the assays with variant-restimulated cells at a lower E/T ratio (1:2). (Fig. 3C) We found that G4.2-CTL recognized E75 at a concentration of 50 to 100 nmol/L. The functional avidity of G4.2-CTL for E75 was increased by a factor of at least four at restimulation (from 7% lysis at 1:1 ratio to 19.5% lysis at 1:2 ratio); the functional avidity of G4.4-CTL increased by a factor of two (from 15% lysis at 1:1 ratio to 20% lysis at 1:2 ratio), whereas the functional avidity of G4.1 cells increased by a factor of 10 from (3% lysis at 1:1 ratio to 16% lysis at 1:2 ratio). Therefore, G4.2-expanded TCRhi Perf+ cells of higher functional avidity for E75 than G4.4 and G4.1. The higher functional avidity was a factor of both perforin levels and of the better "fit" TCR-peptide MHC interaction.
G4.3-reactivated and G4.1 reactivated cells died in large numbers and this resulted in elimination of CTL which recognized E75 with high affinity. Specific lysis by G4.1 stimulated CTL for 500 nmol/L E75 decreased to 6% in a second experiment (data not shown). The levels of perforin were similar in G4.1 and G4.4 reactivated cells and higher in G4.2-reactivated cells than in G4.3-reactivated cells. G4.2-expanded more CTL of higher functional avidity for E75 which survived longer than CTL stimulated with E75, G4.1, and G4.3.
To determine whether the higher functional avidity for E75 reflected a high functional avidity for tumor, we assessed E75-specific tumor lysis in E75-blocking experiments. In these experiments, SKOV3.A2 cells were treated or not treated with IFN- to activate antigen presentation. IFN- treatment increased the levels of HLA-A2 by a factor of three, as indicated by an increase in MFI for HLA-A2 from 114 to 342 (data not shown). To verify that E75 was being processed endogenously, we treated targets with the proteasome inhibitor MG132 before adding the effectors. The IFN--treated SKOV3.A2 cells were more sensitive to G4.2-CTL in the 4-hour CTL assay than were the untreated SKOV3.A2 cells (Fig. 3D). Lysis of IFN--treated tumor cells by G4.2-CTL had continued to increase at 20 hours, demonstrating that G4.2-CTL had high and stable functional avidity for E75. MG132 inhibited SKOV3.A2 lysis by 60% in the 4-hour CTL assay (Fig. 3D), indicating that most of the E75 was being processed by proteasomes.
Activation by G4.2 in the presence of caspase-9 inhibitor increases the numbers of TCRhi Perfhi cells. Induction of TCRhi Perf+ cells raised the question of whether these cells were sensitive to antigen-induced apoptosis and, if so, how to avoid antigen-induced cell death. To identify the pathway of preferential deletion by antigen of Perf+ cells, expressing high levels of perforin, the G4.2 cells used in the previous experiment were "rested" by culturing them in the absence of IL-2, treated with an inhibitor of caspase-8 or caspase-9, or remained untreated, and then stimulated with G4.2. IL-2 was added 24 hours later to avoid interference with TCR stimulation. Perforin expression was measured 48 hours later in E75-TCRhi Perf+ cells, and as an internal control in E75-TCRmed Perf+ cells. Although separation of these cells at MFI (103) is arbitrary, from the levels of perforin, two populations were clearly distinguished in both TCRhi and TCRmed cells: one population expressing MFI (Perf) <100 and a second population expressing MFI (Perf) >300.
Pretreatment with the caspase-8 inhibitor increased the number of E75-TCRhi cells by only 15% and decreased their perforin level by 10% compared with cells stimulated with G4.2 in the absence of caspase-8 inhibitor. (Fig. 4A versus B). By contrast, treating the G4.2 cells with the caspase-9 inhibitor doubled the number of TCRhi cells and produced a 50% increase in MFI (perforin) relative to stimulation with only G4.2 (Fig. 4C versus A). These findings indicate that G4.2 induced death in TCRhi Perf hi cells by activating caspase-9. In contrast, in E75-TCRmed cells, the caspase-8 inhibitor was more protective than the caspase-9 inhibitor, increasing the numbers of E75-TCRmed cells by 62% (Fig. 4E) as opposed to only 23% for the caspase-9 inhibitor (Fig. 4F) relative to the cells, which were not pretreated with caspase-inhibitor (Fig. 4D). However, the levels of perforin, in the Perf hi population, increased by 21% in the caspase-9 inhibitor-treated population, suggesting that caspase-9 is activated by TCR to delete cells expressing high levels of perforin.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
These findings were obtained by simultaneous analysis of two human T-cell populations expressing a specific TCR for the same antigen that differed in the amount of TCR expressed per cell by one order of magnitude. The "classic" TCRhi population (of MFI >102; refs. 1–3) was separated in TCRhi [MFI (TCR) >103] and TCRmed [MFI (TCR) <103]. Although this separation was arbitrary, clear differences were observed among TCRhi and TCRmed populations with regard to their responses to CH2 appended E75.
Expansion in TCRhi cells was not associated with affinity of the antigen for TCR or with the stability of peptide-MHC/TCR complexes, as would be expected in classic models of TCR signaling. Responses at priming were weak and they increased at restimulation with G4.2, the agonist of intermediate affinity for TCR inducing stronger expansion than G4.3 and G4.4. Expansion, perforin and IFN- induction in the TCRmed cells paralleled the changes in affinity of the ligand for TCR for G4.2, G4.3, and G4.4. Based on functional assays, G4.2 seemed the most effective among the CH2 variants. G4.2-reactivated cells recognized antigen at lower concentrations than G4.4- and G4.1-reactivated cells.
The differences in ligand-receptor (TCR) affinity reflected differences in atomic forces induced in the CH2 chain by the addition of single CH2 groups. The average affinity for TCR per CH2 group appended was lower in G4.2 than in G4.4 at t0 and was higher in the G4.3 variant than in the G4.2 or G4.4 variants. The fact that this difference corresponds to van der Waals forces of only 0.5 to 1 kcal/mol indicates the extraordinary sensitivity of TCRhi and TCRmed cells to forces that are a full order of magnitude weaker than the forces generated by one hydroxyl group (5 kcal/mol). Unexpectedly, optimal affinity in these experiments corresponded to the linear extension of the side chain of E75 with two CH2 groups, a variant not present in natural amino acids.
Our results indicate that expansion of TCRhi Perf hi cells of high functional avidity for tumor antigen require fine-tuning of the antigen affinity for the TCR to the best "fit" of TCR which activate the survival and lytic programs. In support of these conclusions, mutations in the side chains of some amino acids of HLA-A2 inhibited recognition of epitopes by specific CTL. This seemed the result of conformational changes in the peptide HLA-A2 (35). Conformational changes at the TCR-peptide-HLA-A2 interface after initial binding seem to be essential for recognition of antigen by CTL (36). We cannot address at this time, whether small differences in perforin levels in effectors (e.g., G4.1-CTL versus G4.3-CTL) resulted in differences in the observed tumor lysis. Because expression of high levels of perforin in CTLs is followed by apoptosis when the TCR is restimulated by the ligand, approaches to protect, partially differentiated and differentiated CTL, such as the ones induced by G4.1 and G4.3, using caspase inhibitors, will ultimately result in higher numbers of effector TCRhi Perf+ cells than reactivation with wild-type agonists.
Our results suggest that changes in ligand affinity by substitutions with natural amino acids will be unable to modulate the weak forces executing differential control of effector gene expression in TCRhi cells. Differences of only two methylene groups (e.g., Gly versus -aminobutyric acid) are absent in natural amino acids; in addition, differences of one CH2 group in linear chains of three to four CH2 groups are not present in natural amino acids. The side chains of Val and Leu/Ile differ in one amino acid length and are branched. Differences in van der Waals forces from one CH2 group in Ser and Thr should be masked by the 10-fold stronger forces from –OH groups. Synthetic amino acids are not genetically encoded; although some are generated in humans by enzymatic reactions during metabolism (e.g. -aminobutyric acid, a neurotransmitter), they are not known to be incorporated into proteins. Expression of synthetic amino acids in proteins suggests that posttranslational modifications, such as methylation/demethylation of a CTL epitope, may be taking place if the corresponding bacterial enzymes are present (37). The presence of norvaline and norleucine in bacterial proteins, in recombinant proteins, and in antibiotics, raises the possibility that TCRhi cells specific for self-antigen are periodically activated and inactivated after interactions with bacterial or fungal pathogens.
In summary, our results provide a novel basis for possible control of proliferation and terminal differentiation of human antitumor CTLs that recognize self-antigen. The sensitivity to small changes in atomic force shown by different responses to a one to two CH2 difference in the antigen may be useful in the induction of antitumor responses.
|
Acknowledgments |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Dr. Martin Campbell for peptide synthesis, Dr. Moshe Talpaz for support and encouragement, and Dr. Christinne Wogan for the superb editing of this article.
|
Footnotes |
|---|
Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
Received 6/24/04. Revised 1/ 6/05. Accepted 1/28/05.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Science 2000;290:1354–8. secretion and perforin expression are impaired in CD8+ T lymphocytes from patients with undifferentiated carcinoma of nasopharyngeal type. Cancer Immunol Immunother 2003;52:28–32.[Medline]
This article has been cited by other articles:
|
|
 |
|
  N. Tsuda, D. Z. Chang, T. Mine, C. Efferson, A. Garcia-Sastre, X. Wang, S. Ferrone, and C. G. Ioannides Taxol Increases the Amount and T Cell Activating Ability of Self-Immune Stimulatory Multimolecular Complexes Found in Ovarian Cancer Cells Cancer Res., September 1, 2007; 67(17): 8378 - 8387. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
, 
), TCR med cells (
, 
). F, MFI (TCR) of cells at t0 (i.e., immediately after staining with peptide dimers; 
, 
) cells do not express perforin. Unstimulated TCRhi and TCRlo cells express perforin (Pre). Histograms of the presence of Perf+ and IFN-
), Perf+ IFN-
), total E75+ cells (C5:A5) and (
). Per 106 gated TAL-1 cells. D, restimulation of variant-primed cells with G4.2 results in higher expansion of E75-TCRhi cells than of TCRmed cells, E75-TCRhi cells (
), G4.2 (
), G4.3 (
), and G4.4 (