●
Cutting Edge: CTLs Rapidly Capture Membrane Fragments from Target Cells in a TCR Signaling-Dependent Manner1 Denis Hudrisier,2* Joelle Riond,† Honore´ Mazarguil,† Jean Edouard Gairin,3† and Etienne Joly3‡
Upon encounter of a CTL with a target cell carrying foreign Ags, the TCR internalizes with its ligand, the peptide-MHC class I complex. However, it is unclear how this can happen mechanistically because MHC molecules are anchored to the target cell’s surface via a transmembrane domain. By using antigenic peptides and lipids that were fluorescently labeled, we found that CTLs promptly capture target cell membranes together with the antigenic peptide as well as various other surface proteins. This efficient and specific capture process requires sustained TCR signaling. Our observations indicate that this process allows efficient acquisition of the Ag by CTL, which may in turn regulate lymphocyte activation or elimination. The Journal of Immunology, 2001, 166: 3645–3649.
U
pon physiological stimulation, receptors with tyrosine kinase activity (RTK)4 rapidly internalize together with their soluble ligands. The TCR clearly belongs to this family of receptors, and it too is internalized after stimulation (1). Intuitively, it seems unlikely that the TCR ligand, the peptideMHC complex which is membrane-bound, could be efficiently transferred onto CTL. However, recently the absorption and internalization by T cells of MHC and costimulatory molecules derived
*Institut National de la Sante´ et de la Recherche Me´dicale, Unite´ 395, Centre Hospitalier Universitaire Purpan, BP3028; †Centre National de la Recherche Scientifique, Unite´ Mixte de Recherche 5089; and ‡Centre National de la Recherche Scientifique, Unite´ Propre de Recherche 2163, Centre Hospitalier Universitaire Purpan, BP3028, Toulouse, France Received for publication December 8, 2000. Accepted for publication January 18, 2001. 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. 1
This work was supported by grants from Centre National de la Recherche Scientifique (to J.E.G. and E.J.), Institut National de la Sante´ et de la Recherche Me´dicale (to E.J.), Fondation pour la Recherche Me´dicale (to E.J.), and Association pour la Recherche sur le Cancer (contract number 5485 to J.E.G). 2 Address correspondence and reprint requests to Dr. Denis Hudrisier, Institut National de la Sante´ et de la Recherche Me´dicale, Centre Hospitalier Universitaire Purpan, BP3028, 31024 Toulouse Cedex 03, France. E-mail address:
[email protected] 3
J.E.G. and E.J. contributed equally to this work.
4
Abbreviations used in this paper: RTK, receptors with tyrosine kinase activity; PTK, protein tyrosine kinases; B6, C57BL/6; APL, altered peptide ligands; PKO, perforindeficient; FSC, forward light scatter. Copyright © 2001 by The American Association of Immunologists
●
from APC has been shown to occur during the recognition process (2, 3). The observation that CTL that recognized antigenic peptides subsequently became sensitive to fratricide killing (i.e., killing by T cells specific for the same Ag) (2, 3) suggested that the transfer probably involves the peptide in combination with the presenting MHC molecule. Here we examine the molecular processes involved in the capture of target cell molecules by CTL during the recognition process.
Materials and Methods Cell lines and mice The murine cell lines RMA and BALB/c 3T3 transfected with H-2Db (3T3Db) were used as target cells. CTL lines specific for lymphocytic choriomeningitis virus gp33/H-2Db were derived from the P14 TCR-transgenic mouse (4) or produced by standard peptide immunization procedure of C57BL/6 (B6) perforin-deficient (PKO) mice as described previously (5). CTL lines specific for the model peptide FITC-K9L presented by H-2Db were also generated by peptide immunization.
Peptides Synthesis of peptides has been described elsewhere (6). All peptides were HPLC-purified (⬎98%), and their identity was confirmed by mass spectrometry.
Functional assays Cytotoxicity was measured in a classical 4- to 5-h (or 24-h when indicated) 51 Cr release assay. For IFN-␥ production, 100 l were removed after 24 h from the same wells used for cytotoxicity assays and assayed for IFN-␥ content by ELISA, as previously described, using R4-6A2 and biotinylated XMG1.2 (BD PharMingen, San Diego, CA) (7).
Flow cytometric analysis Target cells were stained with the lipophylic dye PKH-26 (Sigma, St. Louis, MO) according to the manufacturer’s instruction, placed in U-bottom 96-well plates (100 l) with the indicated concentrations of peptides (0.2 ⫻ 106cells/well in final volume of 200 l), incubated for 1 h at 37°C, and then washed three to five times. After the last wash, cell pellets were resuspended with 100 l of T cell suspension (0.1 ⫻ 106/ml), centrifuged for 30 s at 1000 rpm to promote conjugate formation, and then left at 37°C for 1 h (or the indicated times). Conjugates were dissociated by washing cells twice in PBS containing 0.5 mM EDTA and were then stained with anti-CD8-Tricolor. In some experiments, T cells were treated with PP2 (10 M; Sigma) or with blocking Abs either before or after conjugate formation. 0022-1767/01/$02.00
3646
Results and Discussion To investigate how T cells acquire the antigenic peptides, we used two different fluorescently labeled antigenic peptides and measured their transfer onto specific CTL by flow cytometry. FITCgp33 corresponded to the immunodominant epitope of the lymphocytic choriomeningitis virus (gp33, 33KAVYNFATC41) (8) to which a FITC moiety was attached on the ⑀ amino group of Lys33, and Cys41 was replaced by Ile to avoid chemical reduction. The second peptide was the fluorescent probe FITC-K9L (FITCKAIENAEAL) (6). Both peptides bind specifically and with a high affinity (Kd ⫽ 10 –50 nM) to cells that express the H-2Db MHC class I molecule (data not shown). FITC-gp33 is recognized in cytotoxicity assays by P14 CTL specific for gp33 derived from transgenic mice expressing the P14 (V␣2, V8.1) TCR (4), albeit more weakly than the original peptide (half-maximal lysis at 0.3 nM vs 1 pM, respectively; data not shown). FITC-K9L is recognized by CTL generated by standard peptide immunization. RMA cells coated with increasing concentrations of either FITC-gp33 or FITC-K9L were thoroughly washed to remove any free peptide and were then incubated with either P14- or K9Lspecific CTL. After dissociation of the conjugates, the amount of FITC-peptide acquired by the CTL was assessed by flow cytometry. As shown in Fig. 1A, CD8⫹ P14 CTL were significantly and reproducibly stained with FITC after incubation with FITC-gp33pulsed, but not FITC-K9L-pulsed, RMA cells. Conversely, antiK9L CTL became positively stained for FITC after incubation with FITC-K9L-pulsed, but not FITC-gp33-pulsed, RMA cells. As shown in Fig. 1B, P14 and K9L-specific CTL selectively acquired FITC staining in a dose-dependent manner when RMA cells were respectively coated with 10 nM-1 M FITC-gp33 or FITC-K9L. These results demonstrate a specific and dose-dependent capture of
CUTTING EDGE the antigenic peptide by CTL and thus provide a molecular basis for the previously reported process of fratricide killing (2). In addition, we observed that FITC-gp33, but not FITC-K9L, induced a down-modulation of the P14 TCR that correlated closely with the acquisition of FITC-gp33 in a dose-dependent manner (data not shown). Other molecules such as B7.1, B7.2, and ICAM-1 are acquired by T cells during the recognition process in parallel with MHC molecules (3, 9, 10). To evaluate the diversity of molecules acquired by CTL, the surface of living 3T3-Db cells was biotinylated. After pulsing half of them with gp33, these cells were incubated with P14 CTL. After 30 min, the CTLs were harvested, whole-cell protein extracts were prepared from them, and they were transferred to nitro-cellulose membrane after SDS-PAGE separation and probed using streptavidin-HRP. As shown in Fig. 1C, when p14 CTLs were mixed with gp33-pulsed 3T3-Db, the amount of detectable biotinylated molecules recovered on P14 was markedly increased (lane 2) compared with a control performed using unpulsed 3T3-Db cells (lane 1). The pattern of these biotinylated proteins differs qualitatively and quantitatively from those of 3T3-Db and even more from those of P14 (lanes 3 and 4, respectively), demonstrating that these biotinylated proteins did not come from a contamination of P14 T cells by biotinylated 3T3-Db cells or from unblocked biotinylation reagent. Therefore, CTL recognition of the peptide-MHC complex at the surface of the target cell induced the preferential transfer of a selected set of target cellderived surface proteins onto CTL. Although it is likely that the major band detected around 45 kDa corresponds to MHC class I molecules, the precise identity of this and other proteins remains to be determined.
FIGURE 1. Specific capture of target cell-derived antigenic peptides by T cells. A, RMA cells pulsed with 1 M of FITC-gp33 (top) or FITC-K9L (bottom) were mixed with P14 CTLs (left panels) or K9L CTLs (right panels). After incubation at 37°C for 1 h, conjugates were dissociated and stained with anti-CD8-Tricolor and then analyzed by flow cytometry. The analysis was performed on gated CD8⫹ live T cells as identified by FSC parameter. The dark line shows green fluorescence intensity on CTLs exposed to peptide-coated RMA, as compared with noncoated ones (thin line). B, Dose-response curves of FITC-gp33 (F) or FITC-K9L (E) transfer from RMA cells to P14 CTLs (left) or K9L CTLs (right). C, 3T3-Db cells biotinylated on their cell surface were pulsed (lane 2) or not (lane 1) with 1 M gp33 and were then incubated with P14 for 30 min at 37°C. P14 CTLs were separated from 3T3-Db, lysed in N-octylglucoside, and the lysates were analyzed by Western blot analysis revealed by streptavidin-HRP and chemiluminescence. Lanes 3 and 4 correspond to small amounts of lysates from biotinylated 3T3-Db and P14, respectively.
The Journal of Immunology How might the passage onto T cells of so many different molecules from the plasma membrane of the target cell be explained? We investigated the possibility that membrane fragments could convey these proteins by analyzing the transfer to T cells of a fluorescent lipid (PKH26) initially incorporated into the target cell membranes. As shown in Fig. 2A, P14 CTL acquired PKH26 when incubated with RMA cells pulsed with 1 M gp33, but not with 1 M of the control peptide K9L or RMA cells left unpulsed. Reciprocally, significant PKH26 transfer onto K9L-specific CTL occurred only when K9L, but not gp33, was used to coat the target cells. The amount of lipid transferred depended on the dose of peptide used in both cases (Fig. 2B). The transfer observed specifically concerns membranes because when RMA cells were stained with CFSE, a stable internal protein dye, no transfer of fluorescence from RMA cells to P14 CTL was detected (data not shown). When we incubated RMA cells that were both pulsed with FITC-gp33 and labeled with PKH26 with P14 CTL, we observed a parallel capture of antigenic peptide and lipid by T cells (data not shown). To a rough approximation based on levels of fluorescence intensity per cell, we estimate that after 1 h CTLs acquire 5–10% of the amount of lipid and peptide initially present on a target cell. Capture of both peptide and lipid was partially inhibited by antiP14 TCR Abs (anti-V␣2 and anti-V8.1), but not by anti-V6 or anti-TCR C control mAb (data not shown). After P14 CTLs were exposed to gp33-pulsed RMA cells, we also observed fratricide killing by fresh P14 CTLs (data not shown). These observations suggest that specific MHC-TCR interactions are directly involved in the capture process. Interestingly, mAb 53.6.7.2 (or its Fab⬘), which binds to CD8␣, markedly inhibited the transfer of both PKH26 (Fig. 3A) and FITC-gp33 (data not shown). Because this mAb stabilizes the TCR-ligand interaction but inhibits TCR-mediated signaling (11, 12), we wondered whether the capture process would require TCR signaling. To investigate this possibility, we made use of PP2, a selective inhibitor of Src family protein tyrosine kinases (PTK),
FIGURE 2. Specific capture of target cell membranes by T cells. A, RMA cells were labeled with PKH26, pulsed with 1 M of gp33 (top) or K9L (bottom), and then incubated with P14 (left) or K9L CTLs (right) for 1 h at 37°C. P14 CTLs were then analyzed by flow cytometry. RMA cells were gated out based on FSChigh and PKH26bright characteristics. The PKH26 fluorescence intensity was measured on T cells exposed to peptidepulsed (line) as compared with unpulsed (gray fill) RMA cells. B, Doseresponse curves for gp33 (F) or K9L (E) used to pulse RMA cells before incubating them with P14 (left) or K9L CTLs (right).
3647
FIGURE 3. Capture of target cell membranes by T cells is rapid and requires sustained TCR signaling. A, PKH26-labeled RMA cells pulsed or not with 1 M gp33 were incubated with P14 for 1 h at 37°C (upper two panels). P14 CTLs were treated with 10 M PP2 or 10 g/ml 53.6.7.2 anti-CD8␣ mAb for 15 min at 37°C before addition to gp33-pulsed RMA cells (lower two panels). B, PKH26-labeled RMA cells pulsed with 1 M gp33 were incubated with P14 at 37°C (F). Cell conjugates were disrupted by washing in PBS-EDTA at the indicated time points and analyzed immediately by flow cytometry. PP2 (10 M) was added to P14 CTLs 15 min before conjugate formation (f) to provide a negative control. C, PP2 was added to gp33/RMA-p14 cell conjugates after incubations for the indicated times. ⫺15, PP2 was added to P14 CTLs 15 min before conjugate formation. PKH26 capture by P14 was assessed after 1 h at 37°C. The first two bars correspond to positive and negative controls in the absence of PP2.
which act very early in the TCR signaling cascade (13). At 10 M, a nontoxic concentration that blocks T cell activation completely (data not shown), PP2 fully inhibited the acquisition of PKH26 by P14 CTL induced by gp33 (Fig. 3A). Even though we cannot exclude an effect of PP2 on endocytosis, this result, together with the Ab-blocking experiments presented above, strongly suggests a role for TCR signaling in the capture process. Interestingly, the transfer of RMA cell membranes to P14 CTL was already detectable after a 5-min incubation and reached its maximum as early as after 30 min, with 50% of the transfer being achieved at 12 min (Fig. 3B). The fact that the increase of PKH26 staining on P14 with time was progressive and followed a monomodal profile (data not shown) suggests that multiple transfers of small amounts of target cell plasma membrane occurred during CTL-target cell recognition rather than a single event involving the capture of a large piece of membrane. As shown in Fig. 3C, PP2 inhibited the transfer of PKH26 to T cells even when added after conjugation, with the extent of inhibition diminishing progressively with time and indicating that a sustained TCR signaling was required to maintain the capture process. A similar time-dependence was observed with anti-CD8 mAb (data not shown). These kinetics are remarkably comparable to those for inhibition of T cell function when similar approaches are used (13, 14). Finally, these data strongly support our view that small fragments of membrane are progressively captured by CTL in a manner that can be interrupted with treatments affecting proximal events of TCR signaling. Because PKH26 transfer required sustained TCR signaling and this signaling can be differentially affected by altered peptide ligands (APL) (15, 16), we next compared the ability of 16 APL to trigger cytotoxicity, cytokine production, and capture of PKH26. The 16 peptides all bound to H-2Db with affinities comparable to the natural peptide (i.e., around 1 nM) (data not
3648
FIGURE 4. Induction of target cell membrane capture by altered peptide ligands. RMA cells pulsed with the indicated concentrations of peptides were tested for their ability to activate P14 cytotoxicity in a 51Cr release assay (upper row), target cell membrane capture by flow cytometry (middle row), and IFN-␥ production by ELISA. The names of the APL used to obtain the data shown here are indicated on the top row, and those of additional APL with similar behavior are indicated at the bottom of each column.
shown). However, their effects on CTL activity were widely divergent and could be classified into six different categories according to their capacity to stimulate P14 CTL functions. As shown on Fig. 4, membrane capture by CTL correlates tightly with their lytic function but not with cytokine production. Considering the tight correlation between this transfer phenomenon and cytolytic activity, we compared the ability for membrane capture by gp33-specific CTLs derived from PKO mice with that of P14 CTLs. As expected, PKO CTLs were incapable of killing their targets in a 4-h assay but could do so in a 24-h assay (Fig. 5A). Interestingly, anti-gp33 PKO CTLs could capture target cell membranes as efficiently as P14 CTLs (Fig. 5B) or CTLs from B6 mice immunized with gp33 (data not shown). The 100-fold difference in the concentration of peptide required
FIGURE 5. Target cell membrane capture by perforin-deficient CTL cytotoxicity. A, RMA cells pulsed with the indicated concentrations of gp33 were challenged with P14 CTLs (F) or gp33-specific CTLs generated in PKO B6 mice (E) for 1 h at 37°C in a 4-h (left panel) or 24-h (right panel) 51Cr release assay. In control samples using K9L-pulsed RMA cells, no killing was observed after short or long incubation (data not shown). B, Capture of target cell membranes was analyzed by flow cytometry after incubating the two CTL populations for 60 min with PKH26-labeled gp33pulsed RMA cells.
CUTTING EDGE for the half-maximal capture to be reached in PKO vs P14 CTLs relates to the avidity of both CTL lines for the Ag (Fig. 5A). Finally, the kinetics of membrane capture was also the same when PKO or P14 CTLs were used (data not shown). Therefore, these observations rule out that this capture process requires the involvement of perforin’s cytolytic activity and indicate that the capture process is not a consequence of perforin-mediated damages caused to target cells. The possibility that CTLs scavenge membrane fragments nonspecifically was ruled out by that fact that K9L-specific CTLs did not capture any PKH26 when coincubated with P14 CTLs in the presence of PKH26-labeled RMA cells pulsed with gp33 (data not shown). The tight correlation between short-term (perforin-dependent) cytotoxicity and membrane capture indicate that either these two processes are independent but have similar requirements or that they are intimately linked. In light of our data, membrane capture is independent of perforin release and could correspond to an initial “love bite” preceding CTL’s kiss of death. Further studies are required to explore the nature of the correlative link we observed between these two events. One point to consider is the role(s) that this capture phenomenon might play in regulating CTL activity. By providing prolonged engagement of the TCR, membrane fragments captured from Agbearing target cells could in the first place favor T cell activation. However, whether soluble or membrane-bound ligands of RTK could participate in signaling once inside the cell is still a matter of debate (17, 18). In turn, in the case of important antigenic load, those activated CTL will inherit enough antigenic material to become sensitized to killing by other CTLs carrying TCRs of the same specificity as their own. This fratricide killing was proposed previously to explain the phenomenon of lymphocyte “exhaustion” (2, 3, 19). Our data not only support this notion, but also provide a mechanistic explanation for this phenomenon.
Acknowledgments We thank S. Julien for blot analysis, J. P. M. van Meerwijk, P. Romagnoli, C. Featherstone, and S. Valitutti for helpful discussion and critical reading of the manuscript, and J. C. Guery for PKO mice and mAb to IFN-␥.
References 1. Valitutti, S., S. Muller, M. Cella, E. Padovan, and A. Lanzavecchia. 1995. Serial triggering of many T-cell receptors by a few peptide-MHC complexes. Nature 375:148. 2. Huang, J. F., Y. Yang, H. Sepulveda, W. Shi, I. Hwang, P. A. Peterson, M. R. Jackson, J. Sprent, and Z. Cai. 1999. TCR-mediated internalization of peptide-MHC complexes acquired by T cells. Science 286:952. 3. Hwang, I., J. F. Huang, H. Kishimoto, A. Brunmark, P. A. Peterson, M. R. Jackson, C. D. Surh, Z. Cai, and J. Sprent. 2000. T cells can use either T cell receptor or CD28 receptors to absorb and internalize cell surface molecules derived from antigen-presenting cells. J. Exp. Med. 191:1137. 4. Pircher, H., E. E. Michalopoulos, A. Iwamoto, P. S. Ohashi, J. Baenziger, H. Hengartner, R. M. Zinkernagel, and T. W. Mak. 1987. Molecular analysis of the antigen receptor of virus-specific cytotoxic T cells and identification of a new V␣ family. Eur. J. Immunol. 17:1843. 5. Hudrisier, D., J. Riond, H. Mazarguil, M. B. A. Oldstone, and J. E. Gairin. 1999. Genetically encoded and post-translationally modified forms of a major histocompatibility complex class I-restricted antigen bearing a glycosylation motif are independently processed and co-presented to cytotoxic T lymphocytes. J. Biol. Chem. 274:36274. 6. Hudrisier, D., H. Mazarguil, M. B. A. Oldstone, and J. E. Gairin. 1995. Relative implication of peptide residues in binding to major histocompatibility complex class I H-2Db: application to the design of high-affinity, allele-specific peptides. Mol. Immunol. 32:895. 7. Hudrisier, D., B. Kessler, S. Valitutti, C. Horvath, J.-C. Cerottini, and I. F. Luescher. 1998. The efficiency of antigen recognition by CD8⫹ CTL clones is determined by the frequency of serial TCR engagement. J. Immunol. 161:553. 8. Oldstone, M. B., J. L. Whitton, H. Lewicki, and A. Tishon. 1988. Fine dissection of a nine amino acid glycoprotein epitope, a major determinant recognized by lymphocytic choriomeningitis virus-specific class I-restricted H-2Db cytotoxic T lymphocytes. J. Exp. Med. 168:559. 9. Patel, D. M., P. Y. Arnold, G. A. White, J. P. Nardella, and M. D. Mannie. 1999. Class II MHC/peptide complexes are released from APC and are acquired by T cell responders during specific antigen recognition. J. Immunol. 163:5201.
The Journal of Immunology 10. Arnold, P. Y., and M. D. Mannie. 1999. Vesicles bearing MHC class II molecules mediate transfer of antigen from antigen-presenting cells to CD4⫹ T cells. Eur. J. Immunol. 29:1363. 11. Luescher, I. F., E. Vivier, A. Layer, J. Mahiou, F. Godeau, B. Malissen, and P. Romero. 1995. CD8 modulation of T-cell antigen receptor-ligand interactions on living cytotoxic T lymphocytes. Nature 373:353. 12. Daniels, M. A., and S. C. Jameson. 2000. Critical role for CD8 in T cell receptor binding and activation by peptide/major histocompatibility complex multimers. J. Exp. Med. 191:335. 13. Muller, S., S. Demotz, C. Bulliard, and S. Valitutti. 1999. Kinetics and extent of protein tyrosine kinase activation in individual T cells upon antigenic stimulation. Immunology 97:287. 14. Valitutti, S., M. Dessing, K. Aktories, H. Gallati, and A. Lanzavecchia. 1995. Sustained signaling leading to T cell activation results from prolonged T cell receptor occupancy. J. Exp. Med. 181:577.
3649 15. Germain, R. N., and I. Stefanova. 1999. The dynamics of T cell receptor signaling: complex orchestration and the key roles of tempo and cooperation. Annu. Rev. Immunol. 17:467. 16. Hudrisier, D., and I. F. Luescher. 2000. Antigen recognition by CD8⫹ CTLs. In Cytotoxic Cells: Basic Mechanisms and Medical Applications. M. V. Sitkouski and P. A. Heukazt, eds. Lippincott Williams & Wilkins, Philadelphia, p. 25. 17. Cagan, R. L., H. Kramer, A. C. Hart, and S. L. Zipursky. 1992. The bride of sevenless and sevenless interaction: internalization of a transmembrane ligand. Cell 69:393. 18. Kramer, H. 1993. Patrilocal cell-cell interactions: sevenless captures its bride. Trends Cell Biol. 3:103. 19. Hanon, E., J. Stinchcombe, M. Saito, B. Asquith, G. Taylor, Y. Tanaka, J. Weber, G. Griffiths, and C. Bangham. 2000. Fratricide among CD8⫹ T lymphocytes naturally infected with human T cell lymphotropic virus type I. Immunity 16:657.
