Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Influenza A antigen exposure selects dominant Vβ17+ TCR in human CD8+ cytotoxic T cell responses

Influenza A antigen exposure selects dominant Vβ17+ TCR in human CD8+ cytotoxic T cell responses Abstract During acute human viral infections, such as influenza A, specific cytotoxic T lymphocytes (CTL) are generated which aid virus clearance. We have observed that in HLA-A*0201+ subjects, CTL expressing Vβ17+ TCR and recognizing a peptide from the influenza A matrix protein (M158–66) dominate this response. In experimental models of infection such dominance can be due to inheritance of a restricted T cell repertoire or acquired consequent on expansion of CTL bearing an optimum TCR conformation against the MHC–peptide complex. To examine how influenza A infection might influence the development of TCR V0β17 expansion, we studied influenza A-specific CTL in a cross-sectional study of 82 HLA-A*0201+ individuals from birth (cord blood) to adulthood. Primary M158–66 -specific CTL were detected in cord blood, but their TCR were diverse and depletion of Vβ17+ cells did not abrogate specific cytotoxicity. In contrast following natural influenza A infection, TCR Vβ17+ CTL dominated to the extent that only one of nine adult CTL lines retained any functional activity after in vitro depletion of Vβ17+ CTL. These results suggest that the dominance of Vβ17+ TCR among adult M158–66-specific CTL results from maturation and focussing of the response driven by exposure to influenza, and have implications for optimum immunization strategies. affinity, cytotoxic T cell, human, influenza A, TCR CTL cytotoxic T lymphocyte, HA haemaglutinin, HI haemagglutinin inhibition, NA neuraminidase, PBMC peripheral blood mononuclear cell, TdT terminal deoxynucleotidyl transferase Introduction Influenza A is a common respiratory pathogen in man causing frequent epidemics and occasional pandemics associated with considerable morbidity and mortality (1,2). The importance of neutralizing antibodies in immunity to influenza A is clear. The main epitopes for these antibodies are on the surface haemaglutinin (HA) and neuraminidase (NA) proteins, which are highly variable and selected for by the humoral response (3). Antigenic drift and antigenic shift result in structural differences in these proteins amongst the different strains of influenza A and neutralizing antibodies are therefore strain-specific (4). However, CD8+ cytotoxic T lymphocytes (CTL), which are important in the clearance of influenza A and other virus infections (5,6), do not discriminate between major virus subtypes (7). This is because most influenza A-specific CTL epitopes are derived from conserved internal viral proteins (8–11). The TCR repertoire used by CD8+ T cells recognizing a defined viral epitope may be conserved or diverse, but little is known about the factors contributing to this selection in man (12–15). One explanation for conserved TCR usage is that endogenously processed self peptides homologous to an antigenic peptide `trim' the TCR repertoire during thymic selection (13,14,16). Conserved TCR usage could also represent maturation and focusing of the CD8+ T cell response following primary or repeated viral infection. Maturation of B cell responses to antigen is well described, and occurs by a combination of somatic hypermutation in the CDR3 region and antigen-driven selection for cells expressing receptors with the highest affinity for antigen (17). There is little evidence for somatic hypermutation in T cells (18). However, the ability of antigen exposure to selectively expand high-avidity T cell populations has been examined in CD4+ and CD8+ T cell responses in murine models (15,19–22). Conflicting results both supporting narrowing of the TCR repertoire during secondary T cell responses (19,21–23) and development of oligoclonality during primary exposure (15,20) have been reported. These studies have been often limited by use of transgenic mice responding to artificial antigens, responses not surprisingly characterized by TCR conservation in the primary response. To date we do not know the relevance of these observations to acute viral infections such as influenza A in man. Such studies are difficult because of the outbred population, controlling antigen exposure and the ethics of studies involving children. To date human studies have been confined to Epstein–Barr virus (24,25), cytomegalovirus (26) and HIV (27,28) infections which establish life-long persistence rather than repeated acute exposure. The HLA-A*0201-restricted influenza A/M158–66 peptide-specific CTL response is dominated by and highly dependent on CD8+ T cells expressing TCR incorporating the Vβ17 and Vα10 gene segments (12,14). An endogenous peptide from the TAP.2 protein with strong sequence homology to the M158–66 peptide has been described and it was suggested that this peptide may have a role in restricting the M158–66-selected TCR repertoire (14,16). However, it is not known whether this peptide is naturally processed and expressed in the thymus. Furthermore, the TCR Vβ17 usage of primary M158–66-specific CTL has not been investigated. We investigated the role of antigen-driven maturation in this CTL response by determining TCR Vβ17 usage of M158–66-specific CTL in a cross-sectional study of 82 individuals from birth to adulthood. We compared the functional properties of M158–66-specific CTL generated from cord and maternal blood, and show that the oligoclonality of Vβ17+ CTL is probably antigen driven as a consequence of natural infection. Methods Subjects Permission to collect cord blood was obtained from women on arrival at the delivery suite at the University Hospital of Wales, Cardiff. Following delivery of the placenta, the clamped umbilical cord was cleaned and a 14 gauge needle inserted into the central umbilical vessels. Between 15 and 150 ml of blood was collected, the final volume depending on the stage at which the cord was clamped; 50ml of maternal blood was collected whenever possible. The children in this study attended the Paediatric Outpatient Department at the University Hospital of Wales (aged 3 months to 15 years), being treated for idiopathic epilepsy or recurrent febrile convulsions with sodium valproate or vigabatrin with blood level monitoring. After written informed consent was given by the parent/guardian, ~2–5 ml of blood additional to that which was being taken for drug levels was collected into a universal container containing preservative-free heparin (10 IU/ml). Blood was also collected from normal laboratory volunteers. The study was approved by the local research ethics committee and written informed consent was obtained from all subjects or guardians. None of the subjects had been immunized against influenza. HLA typing Individuals expressing HLA-A*0201 were identified using the MA2.1 mAb and single-colour immunofluorescence to screen blood samples (14). This was confirmed for several individuals by DNA typing. Influenza A serology Anti-influenza antibody production was determined with the haemagglutinin inhibition (HI) test. All serum samples were stored at –20°C until assayed. Before investigation, they were treated with receptor destroying enzyme from Vibrio cholerae overnight at 37°C and then inactivated at 56°C for 30 min. The HI test was performed with eight agglutinating doses of virus (A/Taiwan/1/86, A/Bayern/7/95, A/JHB/34/94, A/Wuhan/359/95 and A/Sydney/5/97) as appropriate prototype viruses which had circulated during the course of this study. Each test was performed in duplicate with all antigens. Generation of M158–66 -specific CTL lines HLA-A*0201-restricted influenza A/M158–66 -specific CTL lines were generated as previously described (14). In brief, freshly isolated peripheral blood mononuclear cells (PBMC) from cord blood, or blood obtained from children of various ages or adults were infected with 200 HAU of influenza A (AX/31). After 1 h incubation at 37°C, cells were cultured in RPMI-10(AB) [RPMI 1640 (Sigma, St Louis, MO), 10% pooled human AB serum, 2 mM l-glutamine, 1 mM sodium pyruvate and 25 mM HEPES buffer] for 7 days. On day 7 and weekly thereafter cells were harvested, counted and replated in 24 well plates at 2×106 cells/well in RPMI-10(AB) containing 10 IU/ml IL-2 (Boehringer Mannheim, Mannheim, Germany). Cultures were re-stimulated with irradiated (3000 rad for 7 min) M158–66 peptide-pulsed (5 μg/ml) CIRA2 cells at a responder to stimulator ratio of 10:1. On day 21, cultures were tested for influenza A and M158–66 specificity by 51Cr-release assay, and the CD8+Vβ17+ content of M158–66-specific lines determined by two-colour flow cytometry as described (14). Identical re-stimulation conditions were used to generate HLA-A*0201-restricted M158–66-specific CTL lines from PBMC obtained from cord blood, children and adults. Removal of Vβ17+ cells from HLA-A*0201-specific CTL lines Immunomagnetic depletion was performed to remove Vβ17+ cells or as a control Vβ22+ cells (Dynabeads M-450 sheep anti-mouse IgG; Dynal, The Wirral, UK) as previously described (14). Depletion was typically >97%. Staining and analysis of CD8+ T cells Cells were analysed by flow cytometry using a FACScan (Becton Dickinson, La Jolla, CA). Live cells were gated based on FSC/SSC profiles. TCR Vβ gene segment usage was determined by two-colour immunofluorescence using a panel of FITC-conjugated anti-TCR Vβ mAb (TCR Workshop, San Francisco, 1995) and anti-CD8–phycoerythrin (Becton Dickinson). The staining procedure was as previously described (14). Cytotoxicity assays Standard 51Cr-release assays were performed using 51Cr-labelled CIRA2 cells (2000/well) which express a transfected genomic clone of HLA-A*0201 (29). Targets were pulsed with M158–66 peptide (10–6 M) or infected with 200 HAU influenza A as described (14), and control targets included CIRA2 cells in the absence of peptide and CIRA2 cells pulsed with the HLA-A*0201-binding Plasmodium cp36 peptide (30). Percent specific lysis was calculated as 100×experimental release – spontaneous release)/(maximum release – spontaneous release). Spontaneous release was <10% unless otherwise stated. CTL lines were assayed for specific cytotoxicity after 21 days in culture. Results HLA-A*0201-restricted influenza A-specific cord blood T cell responses in cord blood To minimize the effects of antigen exposure on development of the HLA-A*0201-restricted influenza A/M158–66 response, CTL responses were examined in PBMC obtained from umbilical cord blood. Using a re-stimulation protocol previously described for the generation of in vitro primary CTL responses (31), influenza A/M158–66-specific CD8+ T cell responses were generated from four of five HLA-A*0201+ cord blood samples (data not shown). However, this method could not be used for generating CTL from infants because the quantity of blood available was limited to 1–5 ml, thus precluding the generation of autologous dendritic cells. For three HLA-A*0201+ cord blood and three adult blood samples we confirmed that different re-stimulation protocols did not alter the TCR Vβ gene segment usage of the M158–66-specific CTL lines generated (data not shown). However, in order to compare TCR usage between the different individuals in this study it was essential to use identical re-stimulation conditions. We therefore determined whether a re-stimulation protocol conventionally used for the in vitro detection of secondary responses (7) could be applied to cord blood. Surprisingly, M158–66-specific CD8+ T cell responses were generated from 60% (30 of 50) of HLA-A*0201 cord blood samples using this protocol which was thereafter used for generating CTL lines from all subjects included in this study. All M158–66-specific CTL lines killed influenza A virus-infected targets in addition to M158–66 peptide-pulsed targets (Fig. 1). The possibility that these cord blood CD8+ T cell responses resulted from contamination of cord blood with maternal blood was excluded as M158–66-specific lines could be generated from the cord blood of HLA-A*0201-negative mothers, who were unable to generate M158–66-specific CD8+ T cell responses (data not shown). In addition, the MHC types of CTL lines generated from five paired maternal and cord blood samples were determined, and haplotype differences between maternal and cord M158–66-specific CTL lines confirmed (data not shown). To exclude the possibility of intrauterine exposure to influenza A, anti-influenza A antibodies were measured during the first trimester of pregnancy and postnatally from seven mothers, whose infant's cord blood yielded M158–66-specific CTL responses. Two of seven mothers had serological evidence suggestive of influenza A exposure during pregnancy (data not shown). M158–66-specific CTL lines generated from cord blood samples obtained from these two individuals were no different from other cord CTL lines in terms of specific cytotoxicity or TCR Vβ17 usage. These findings together with the absence of any evidence supporting transplacental transfer of influenza A suggest that M158–66-specific CTL responses in cord blood are likely to be primary CTL responses elicited in vitro. TCR Vβ17 usage by M158–66-specific CD8+ T cells The ability to generate probable primary HLA-A*0201-restricted influenza A/ M158–66-specific CTL responses from cord blood using exactly the same re-stimulation protocol as for children and adults, allowed comparison of the TCR Vβ17 usage between primary and secondary effector CD8+ CTL responses. Freshly isolated PBMC and M158–66-specific CTL lines generated in vitro from nine paired cord and maternal samples were examined. Although all CTL lines were influenza A virus specific and M158–66 specific in 51Cr-release assays, cord-derived CTL lines contained significantly fewer CD8+Vβ17+ cells than the corresponding maternal CTL (Table 1). Further M158–66-specific CTL lines were generated from unrelated cord and adult blood samples, and also from PBMC obtained from children of different ages using the same in vitro re-stimulation conditions. Flow cytometric analyses showed that the frequency of CD8+Vβ17+ T cells in M158–66-specific CTL lines increased with age (Fig. 2). An individual example of an in vitro primary response can be seen in child BH. This child was 3 months old and the M158–66-specific CTL line generated contained 14% CD8+Vβ17+cells (Fig. 2). This child had no serological evidence of exposure to influenza A by the HI test (Table 2). The negative serology also implied that the mother had not been exposed to influenza A during pregnancy as maternal antibodies would otherwise have been detected. An individual example of in vivo pathogen exposure influencing Vβ17 expression can be seen in child DH, for whom two samples were obtained 1 year apart. No CTL were generated from the first sample and only 5% of CD8+ T cells in the day 21 culture were Vβ17+. However, 75% of the CD8+ cells in the second sample expressed Vβ17 and M158–66-specific CTL activity was detected (Table 1). The first serum sample from child DH (age 21months) was negative by the HI test, but antibodies were detected in the second sample (age 3 years) suggesting exposure to influenza A between the two samples (Table 2). As initial serology was negative it is likely that this was the first exposure to influenza A. The cord blood M158–66-specific CD8+ T cell response is less dependent on the presence of the CD8+Vβ17+ T cell population than the adult response Previous studies suggested that the M158–66-specific CTL response in adults is highly dependent on the presence of CD8+Vβ17+ CTL (14). Although the number of CD8+ cells expressing Vβ17 in cord blood M158–66-specific CTL lines was significantly reduced compared to adult lines (Fig. 2) it was still possible that M158–66-specific cytotoxicity might be abrogated following removal of this smaller Vβ17+ population. To determine the relative contribution of Vβ17+ cells in M158–66-specific CTL lines generated from cord blood and subjects of different ages, CTL lines were depleted of either Vβ17+ or, as a control, Vβ22+ cells. Comparison of four maternal and cord CTL lines grown in parallel, confirmed that the cord blood M158–66-specific CTL response was not dependent on the presence of the CD8+Vβ17+ population. Depletion of Vβ17+ cells did not decrease the lytic activity of M158–66-specific CTL lines with low frequencies of Vβ17+ cells (cord DR, SG and JC). (Fig. 3). In contrast, the maternal lines were highly dependent on the presence of these cells, the M158–66-specific response being abrogated in three of four cases and markedly reduced in the other (Fig. 3). The difference in the proportion of CTL clones using the Vβ17 gene segment in cord and adult blood was also determined by comparing the M158–66-specific CTL response in cord blood and the corresponding maternal sample by limiting dilution analysis. Using this technique, 20% of cord blood-derived CTL clones were Vβ17+ compared with 80% of maternal CTL clones (data not shown). These experiments were extended to examine M158–66-specific CTL lines from further adults, cord blood samples and also CTL lines from children at different ages in a cross-sectional study. The M158–66-specific CTL line of child BH contained 14% CD8+Vβ17+ cells and the response was not affected by removing these cells (Fig. 4). The M158–66-specific CTL response for child DH following exposure to influenza A was markedly reduced following depletion of Vβ17+ cells, thus being similar to adult CTL lines (Fig. 4). No cord M158–66-specific CTL lines (out of eight) were dependent on the presence of Vβ17+ cells, whilst only one adult line (MJ) was partially independent of Vβ17+ cells (MJ) (Fig. 4). The diversity of the TCR repertoire in the cord blood M158–66-specific CTL response was confirmed using a panel of anti-TCR Vβ mAb. Expansion of M158–66-specific CD8+ CTL expressing other TCR Vβ gene segments was detected in cord-derived CTL, but not in the paired maternal sample (Fig. 5). In the representative example shown, the cord M158–66-specific CTL line had significant expansions of CD8+ cells expressing Vβ5.1, Vβ6 and Vβ17 gene segments (Fig. 5). Depletion experiments confirmed that the cord blood response was independent of Vβ17+ cells, but removal of Vβ17+, Vβ5.1+ and Vβ6+ cells resulted in loss of >95% of cytotoxic activity (Fig 5). The maternal response was dependent on the presence of the Vβ17+ population (Fig. 5). The Vβ gene segment usage in cord M158–66-specific CTL varied between subjects. Vβ17 dependence of the M158–66-specific CD8+ T cell response and natural exposure to influenza A Influenza A HI serology is difficult to interpret in cord blood and in samples obtained during the first year of life, in that positive serology may represent the transplacental transfer of maternal antibodies. Negative serology in such samples is useful, however, and supports the absence of recent exposure to influenza A. The negative HI serology on child BH provided good evidence that this 3-month-old child had not been exposed to influenza A and that the M158–66-specific response represented an in vitro primary CD8+ T cell response (Table 2). HI serology was performed on serum samples from children when available (Table 2). Although the association between dominant Vβ17 usage and positive serology was not absolute, the in vitro M158–66-specific CD8+ T cell response generated from the majority of children with evidence of exposure to influenza A was dominated by CD8+Vβ17+ cells (Table 2). Discussion In this study we have compared Vβ17 usage by M158–66-specific CTL generated from cord blood, and blood obtained from children of various ages and adults. We have demonstrated that the dominant usage of Vβ17-expressing CD8+ T cells in the adult response is acquired, and probably the result of antigen-driven selection and focusing of the response. While a role for thymic selection involving self-peptides homologous to M158–66 cannot be excluded, our results argue against this mechanism. Diverse TCR usage was found among M158–66-specific CTL generated from cord blood, whereas conserved TCR usage might be expected if a thymic selection mechanism was operating. These results confirm that maturation and focusing of CD8+ T cell responses by clonal selection occurs in man in a similar manner to that recently described in a murine model (22). This focusing of a T cell response has potential advantages for the host organism in that the selection of CTL with the highest avidity for a particular epitope–MHC complex may favour the early detection and subsequent lysis of infected cells following re-exposure to the same virus. In addition to highlighting the ability of the immune system to focus a CTL response after infection, these findings demonstrate the important role of repeated antigenic challenge and the likely ineffectiveness of inappropriately selected single peptide immunizations. Many studies have shown that generation of `primary' CTL responses requires special protocols. Studies in neonatal mice have suggested that `dendritic cell' antigen-presenting cells are essential for the generation of primary CTL (32). Furthermore, the in vitro generation of primary CD8+ T cell responses requires cytokines such as IL-7 (30) and/or extended tissue culture with multiple secondary in vitro re-stimulations (33). Therefore the generation of primary M158–66-specific CTL responses from cord blood using a standard protocol (7) was surprising, especially given that previous attempts using a similar protocol were unsuccessful (14). This discrepancy might be explained by a small difference in culture conditions; we were unable to detect M158–66-specific cytotoxicity from cord lines before day 16, whereas Lehner et al. (14) measured responses on day 14. The presence of HLA-A*0201-restricted M158–66-specific CTL in cord blood could also be explained by contamination of cord blood by maternal blood (transplacentally or during delivery) or in utero exposure of the foetus to influenza A. Contamination by maternal blood was excluded. In utero exposure to influenza would require maternal influenza A infection during pregnancy, influenza viraemia and transplacental transfer of infection. Previous studies have estimated that between 5 and 13% of mothers may be infected with influenza A during pregnancy (34). Two out of seven mothers we tested antenatally and postnatally using HI showed evidence of seroconversion to influenza A antigens. This is still far less than the 65% cord blood CTL response rate observed in this study. There is only very limited evidence for viraemia during acute natural influenza A infection (35) and no evidence for transplacental transfer of influenza in humans (34). These facts suggest that in utero exposure to influenza A is an unlikely explanation for the ability to generate HLA-A*0201-restricted M158–66-specific CTL responses from cord blood in vitro. It is likely, therefore, that cord blood M158–66-specific CTL responses are in vitro primary responses and result from an unusual ability of the HLA-A*0201/M158–66 complex to activate naive as well as memory M158–66-specific CTL precursors. The primary nature of this response is supported by the observation that influenza virus-infected stimulator cells were required at the onset of culture to generate M158–66-specific CTL lines from cord PBMC, whereas CTL lines could be generated from adult PBMC using M158–66peptide alone (data not shown). The TCR Vβ repertoire of M158–66-specific CTL in the in vitro primary cord blood CTL response was more heterogeneous and less dependent on the presence of Vβ17+ cells than that of adults. However, expanded populations of Vβ17+ CTL were frequently observed, in some cases comprising up to 60% of the total CD8+ population. This confirmed that the differences observed between adult and cord lines were not the result of an inability of Vβ17+ cell in cord blood to proliferate in vitro. All studies involving in vitro culture are prone to introduction of bias as a result of preferential outgrowth of those antigen-specific cells with the greatest proliferative capacity (36–39). However, using ELISPOT technology to detect M158–66-specific IFN-γ production by PBMC directly ex vivo, it has been confirmed that the dominance of Vβ17+ M158–66-specific CTL in adults is not an in vitro artefact (40). In this study we limited the potential for in vitro bias by using identical in vitro re-stimulation conditions and culturing cells for the same duration prior to assay. Maternal and cord lines were cultured and re-stimulated in parallel. Failure to generate M158–66-specific CTL lines from some HLA-A*0201 children who had serological evidence of exposure to influenza A may have been due to the technical difficulties involved in establishing CTL lines in vitro from 1–5 ml of blood. It seems likely that the increasing frequency of Vβ17+ CTL that occurs in the M158–66-specific HLA-A*0201-restricted CTL response with age is the result of natural exposure to influenza A. Longitudinal follow up of individuals for several years from birth could confirm this and address a possible alternative explanation. This is based on findings in murine studies which demonstrate that mice deficient in terminal deoxynucleotidyl transferase (TdT) have altered TCR repertoires when compared to wild-type mice (41,42). In particular, a focused TCR Vβ8.2 H-2d-restricted response against SWM110–121 peptide is not seen in TdT-deficient mice (41). TdT is not expressed in mice until several days after birth. TdT expression has not been extensively studied in man (43), however it is possible that differences in TdT expression between cord and adult blood could influence the diversity of the TCR repertoire. While this cannot be excluded, the functional superiority of the Vβ17 TCR for influenza M158–66 peptide compared to no Vβ17 TCR (as demonstrated in the accompanying paper) suggests that exposure to influenza is the major factor in shaping the adult TCR repertoire. In this cross-sectional study we determined previous influenza A exposure by HI which, although a frequently used method for determining influenza A serology, is still associated with a number of drawbacks. We were unable to obtain multiple samples from many of the children in this study for ethical reasons. Isolated samples can still be useful in determining whether the child has been exposed to influenza A, but as HI serology can wane with time, a negative result must be interpreted with caution. In addition interpretation of HI serology in children during the first year of life is complicated by the fact that maternal antibodies may still be present during this period. If the HI serology is negative in this age group, it is unlikely that they have been exposed to influenza A although the possibility that they have been exposed to influenza and failed to respond serologically or exposed to a different strain cannot be excluded. However, it is likely that CTL lines generated from these individuals represent in vitro primary responses. As the switch from a non-Vβ17+-dominated cord response (probable primary response) to a Vβ17+-dominated CTL response occurred at a young age, it is possible that a single in vivo exposure to influenza A may be sufficient to alter the TCR repertoire of the M158–66-specific CTL response as in the case of child DH (Table 2). Neonates are likely to be protected from infection by transfer of maternal antibody up to ~9 months and with a single serum sample later infection would be detected. No major antigenic shift or drift in the circulating influenza A strains, nationally or locally, has been observed during the course of this study, therefore repeated exposure is unlikely. Our findings suggest that the primary M158–66-specific CTL response is mediated by M158–66-specific CD8+ T cells with heterogeneous TCR usage and that the TCR repertoire narrows following the first in vivo exposure to influenza A such that the secondary in vitro response is dominated by Vβ17+ CTL. It is possible that the various M158–66-specific clones have different `burst frequencies' during the primary response. The Vβ17+ CTL may have a larger burst frequency because of a higher avidity for HLA-A*0201/M158–66 and thus a functional advantage, resulting in their dominance of the memory M158–66-specific CD8+ T cells. Recently tetramers comprising soluble HLA class I–peptide complexes have been used to quantitate directly the numbers of anti-viral CTL in blood (44), reducing the need to perform in vitro CTL assays. Furthermore such tetramers have been used as a tool to estimate TCR affinity (22,45). We considered using tetramers to compare the TCR repertoires of M158–66-specific CTL isolated directly from cord and adult blood, thus avoiding potential bias introduced by in vitro culture. However, preliminary experiments showed that there was no detectable binding of HLA-A*0201/M158–66 tetramer to cord blood PBMC due to the low frequency of these cells. Furthermore, in the next paper in this issue we show that although all Vβ17+ CTL clones stain with tetramer, two out of three non- Vβ17+ clones fail to stain. Thus a major proportion of the M158–66-specific CTL in cord blood would not be detected using tetramers. The description of conserved TCR usage by CTL recognizing epitopes from several viruses in man such as cytomegalovirus (26), Epstein–Barr virus (24,25) and HIV (27,28) suggests that maturation of these responses may also occur. Clonal selection of HIV-specific CTL has been noted following the primary HIV seroconversion illness (27,28), but confirmation of `maturation' will require comparison of the TCR repertoires and functional properties of primary and secondary CTL specific for these antigens. The demonstration that a natural antigen-specific CD8+ T cell response in man matures following the first encounter with antigen confirms that factors other than thymic selection can affect the repertoire of T cells responding to a particular peptide epitope and adds to our understanding of the interaction between T cells and targets expressing appropriate MHC–peptide complexes. In vaccination programmes seeking to induce specific CTL responses the implication of these observations is that the first encounter with the antigen is important and hence its formulation, depending on the desire to induce dominant oligoclonal or subdominant CTL, will be crucial. Table 1. TCR Vβ17+ usage and cytotoxicity of M158–66-specific CTL lines generated from paired cord and maternal blood samples Subject  Percent lysis ± SD (E:T 20:1)a  Percent CD8+Vβ17+/total CD8+      PBMC  CTL d21b  aCytotoxicity assays were performed after 21 days in culture. Percent lysis was calculated as follows: 100×(experimental release – spontaneous release)/(maximum release – spontaneous release). Spontaneous release was <10% in all cases. The percent lysis using an E:T of 20:1 is shown.  bThe percentage of CD8+ cells expressing Vβ17 in PBMC was determined prior to in vitro re-stimulation and for M158–66-specific CTL, Vβ17 usage was determined on day 21.  Mother HW  60 ± 10  6  85  Cord HW  40 ± 5  7  3  Mother BE  65 ± 4  4  81  Cord BE  50 ± 8  2  20  Mother TU  55 ± 9  4  75  Cord TU  36 ± 6  2  4  Mother CO  60 ± 11  2.5  50  Cord CO  40 ± 6  4  6  Mother DR  55 ± 5  10  58  Cord DR  40 ± 5  6  7  Mother LC  40 ± 4  3  46  Cord LC  60 ± 10  3  25  Mother SG  40 ± 9  2.5  44  Cord SG  52 ± 8  2  9  Mother CB  60 ± 10  2.8  75  Cord CB  35 ± 8  3  9  Mother JC  50 ± 7  4  68  Cord JC  33 ± 7  3.5  6  Subject  Percent lysis ± SD (E:T 20:1)a  Percent CD8+Vβ17+/total CD8+      PBMC  CTL d21b  aCytotoxicity assays were performed after 21 days in culture. Percent lysis was calculated as follows: 100×(experimental release – spontaneous release)/(maximum release – spontaneous release). Spontaneous release was <10% in all cases. The percent lysis using an E:T of 20:1 is shown.  bThe percentage of CD8+ cells expressing Vβ17 in PBMC was determined prior to in vitro re-stimulation and for M158–66-specific CTL, Vβ17 usage was determined on day 21.  Mother HW  60 ± 10  6  85  Cord HW  40 ± 5  7  3  Mother BE  65 ± 4  4  81  Cord BE  50 ± 8  2  20  Mother TU  55 ± 9  4  75  Cord TU  36 ± 6  2  4  Mother CO  60 ± 11  2.5  50  Cord CO  40 ± 6  4  6  Mother DR  55 ± 5  10  58  Cord DR  40 ± 5  6  7  Mother LC  40 ± 4  3  46  Cord LC  60 ± 10  3  25  Mother SG  40 ± 9  2.5  44  Cord SG  52 ± 8  2  9  Mother CB  60 ± 10  2.8  75  Cord CB  35 ± 8  3  9  Mother JC  50 ± 7  4  68  Cord JC  33 ± 7  3.5  6  View Large Table 2. Association between dependence of M158–66-specific CTL response on the presence of Vβ17+ CTL and previous exposure to influenza A in children of different ages Child  Age  Percent CD8+Vβ17+/total CD8+a  CTL lysis(E:T 20:1)b  Percent lytic activity post Vβ17 depletion c  HI Serologyd    PBMCa  CTL (d21)a    A/Taiwan/1/86 (H1N1)  A/Bayern/7/95 (H1N1)  A/JHB/34/94 (H3N2)  A/Wuhan/359/95 (H3N2)  A/Sydney/5/97 (H3N2)  aThe percentage of CD8+ T cells expressing Vβ17 was determined by FACS analysis for fresh PBMC prior to in vitro stimulation and for M158–66-specific CTL after 21 days in culture.  b51Cr-release assays were performed on day 21 using CIRA2 targets. The percentage lysis of M158–66-pulsed targets is shown for an E:T ratio of 20:1. Lysis of targets pulsed with an irrelevant HLA-A*0201-binding Plasmodium peptide and targets without peptide was <10% in each case.  cDay 24 CTL lines with M158–66 specificity were depleted of Vβ17+ cells or as a control Vβ22+ cells. Vβ22 depletion did not alter cytotoxic activity in a single case. Cytotoxicity was expressed in lytic units and the residual cytotoxicity after depletion of Vβ17+ cells was expressed as a percentage of that observed following depletion of Vβ22+ cells.  dAnti-influenza antibody production was determined with the HI test. The HI test was performed with eight agglutinating doses of five appropriate prototype viruses which had circulated during the course of this study.  ND, not done.  TV  3 months  2  ND  0  ND  <10  <10  <10  <10  <10  BH  3 months  5  14  50  100  <10  <10  <10  <10  <10  DS  6 months  3  4  0  ND  <10  <10  <10  <10  <10  JR  2 years 5 months  4.50  5  30  110  <10  <10  160  120  40  DH (#1)  21 months  3.2  5  0  ND  <10  <10  <10  <10  <10  DH (#2)  2 years 9 months  3.5  70  75  20  <10  <10  <10  40  10  EB (#1)  2 years 8 months  3  52  40  10  <10  <10  160  40  80  EB (#2)  3 years 2 months  3.5  76  75  5  <10  <10  160  160  80  EJ  2 years 9 months  ND  5  10  ND  <10  <10  40  80  40  SP  3 years  ND  80  45  0  <10  <10  80  80  80  IS (#1)  3 years  2.8  4.2  0  ND  <10  <10  <10  <10  <10  IS (#2)  3 years 3 months  ND  ND  0  ND  <10  <10  <10  <10  <10  RP  3 years 6 months  ND  31  60  ND  <10  <10  160  160  30  IR (#1)  3 years 9 months  4  38  40  ND  160  80  320  320  160  IR (#2)  4 years 1 month  ND  54  50  10  160  320  320  320  160  DD  3 years 9 months  4  5  0  ND  <10  <10  <10  <10  <10  HK  4 years  1.9  5  40  ND  <10  <10  80  40  <10  AM  4 years 6 months  ND  40  35  0  <10  <10  80  40  20  CE  5 years  ND  25  60  ND  40  320  40  40  <10  JG  5 years 3 months  4.2  91  50  0  <10  <10  160  160  <10  SM  5 years 7 months  2.9  85  50  0  160  320  640  320  80  AG  6 years  ND  22  41  60  <10  <10  80  80  80  MV  6 years  ND  25  40  90  <10  80  80  80  <10  ACT (#1)  6 years 6 months  4  85  55  0  <10  <10  40  20  <10  ACT (#2)  7 years 2 months  ND  97  80  0  <10  <10  640  640  160  KD (#1)  7 years 6 months  3  50  45  ND  <10  <10  <10  <10  <10  KD (#2)  8 years 2 months  ND  72  60  15  <10  <10  40  160  160  RM  8 years  ND  73  60  0  40  160  40  20  <10  LK  9 years  2.5  32  40  ND  <10  20  40  40  10  SC  12 years  3.2  91  55  0  <10  <10  <10  <10  <10  MS  13 years  5  40  20  0  <10  <10  <10  <10  <10  GW  13 years  4.6  96  80  0  <10  <10  80  <10  <10  Child  Age  Percent CD8+Vβ17+/total CD8+a  CTL lysis(E:T 20:1)b  Percent lytic activity post Vβ17 depletion c  HI Serologyd    PBMCa  CTL (d21)a    A/Taiwan/1/86 (H1N1)  A/Bayern/7/95 (H1N1)  A/JHB/34/94 (H3N2)  A/Wuhan/359/95 (H3N2)  A/Sydney/5/97 (H3N2)  aThe percentage of CD8+ T cells expressing Vβ17 was determined by FACS analysis for fresh PBMC prior to in vitro stimulation and for M158–66-specific CTL after 21 days in culture.  b51Cr-release assays were performed on day 21 using CIRA2 targets. The percentage lysis of M158–66-pulsed targets is shown for an E:T ratio of 20:1. Lysis of targets pulsed with an irrelevant HLA-A*0201-binding Plasmodium peptide and targets without peptide was <10% in each case.  cDay 24 CTL lines with M158–66 specificity were depleted of Vβ17+ cells or as a control Vβ22+ cells. Vβ22 depletion did not alter cytotoxic activity in a single case. Cytotoxicity was expressed in lytic units and the residual cytotoxicity after depletion of Vβ17+ cells was expressed as a percentage of that observed following depletion of Vβ22+ cells.  dAnti-influenza antibody production was determined with the HI test. The HI test was performed with eight agglutinating doses of five appropriate prototype viruses which had circulated during the course of this study.  ND, not done.  TV  3 months  2  ND  0  ND  <10  <10  <10  <10  <10  BH  3 months  5  14  50  100  <10  <10  <10  <10  <10  DS  6 months  3  4  0  ND  <10  <10  <10  <10  <10  JR  2 years 5 months  4.50  5  30  110  <10  <10  160  120  40  DH (#1)  21 months  3.2  5  0  ND  <10  <10  <10  <10  <10  DH (#2)  2 years 9 months  3.5  70  75  20  <10  <10  <10  40  10  EB (#1)  2 years 8 months  3  52  40  10  <10  <10  160  40  80  EB (#2)  3 years 2 months  3.5  76  75  5  <10  <10  160  160  80  EJ  2 years 9 months  ND  5  10  ND  <10  <10  40  80  40  SP  3 years  ND  80  45  0  <10  <10  80  80  80  IS (#1)  3 years  2.8  4.2  0  ND  <10  <10  <10  <10  <10  IS (#2)  3 years 3 months  ND  ND  0  ND  <10  <10  <10  <10  <10  RP  3 years 6 months  ND  31  60  ND  <10  <10  160  160  30  IR (#1)  3 years 9 months  4  38  40  ND  160  80  320  320  160  IR (#2)  4 years 1 month  ND  54  50  10  160  320  320  320  160  DD  3 years 9 months  4  5  0  ND  <10  <10  <10  <10  <10  HK  4 years  1.9  5  40  ND  <10  <10  80  40  <10  AM  4 years 6 months  ND  40  35  0  <10  <10  80  40  20  CE  5 years  ND  25  60  ND  40  320  40  40  <10  JG  5 years 3 months  4.2  91  50  0  <10  <10  160  160  <10  SM  5 years 7 months  2.9  85  50  0  160  320  640  320  80  AG  6 years  ND  22  41  60  <10  <10  80  80  80  MV  6 years  ND  25  40  90  <10  80  80  80  <10  ACT (#1)  6 years 6 months  4  85  55  0  <10  <10  40  20  <10  ACT (#2)  7 years 2 months  ND  97  80  0  <10  <10  640  640  160  KD (#1)  7 years 6 months  3  50  45  ND  <10  <10  <10  <10  <10  KD (#2)  8 years 2 months  ND  72  60  15  <10  <10  40  160  160  RM  8 years  ND  73  60  0  40  160  40  20  <10  LK  9 years  2.5  32  40  ND  <10  20  40  40  10  SC  12 years  3.2  91  55  0  <10  <10  <10  <10  <10  MS  13 years  5  40  20  0  <10  <10  <10  <10  <10  GW  13 years  4.6  96  80  0  <10  <10  80  <10  <10  View Large Fig. 1. View largeDownload slide M158–66-specific CTL lines were generated from 6 HLA-A*0201 cord blood PBMC where the mother was HLA-A*0201 negative. Three representative experiments are shown (maternal data not shown). 51Cr-release assays were performed after 21 days in culture using CIRA2 targets (2000/well) pulsed with M158–66 peptide (5 μg/ml), Plasmodium cp36 peptide (5 μg/ml), no peptide or infected with 200 HAU influenza A. Triplicate wells were used and the SD of the mean is shown in each case. Fig. 1. View largeDownload slide M158–66-specific CTL lines were generated from 6 HLA-A*0201 cord blood PBMC where the mother was HLA-A*0201 negative. Three representative experiments are shown (maternal data not shown). 51Cr-release assays were performed after 21 days in culture using CIRA2 targets (2000/well) pulsed with M158–66 peptide (5 μg/ml), Plasmodium cp36 peptide (5 μg/ml), no peptide or infected with 200 HAU influenza A. Triplicate wells were used and the SD of the mean is shown in each case. Fig. 2. View largeDownload slide TCR Vβ17 usage of influenza A/M158–66-specific CTL generated from subjects of different ages. Day 21 influenza A/M158–66 peptide-specific CTL lines generated from cord PBMC, PBMC from children (3 months to 13 years) and adult PBMC were analysed by two-colour immunofluorescent staining for distribution of Vβ17 within the CD8+ population. Data are expressed as percent CD8+Vβ17+ cells within the total CD8+ subset (CD8+Vβ17+/Total CD8+). The shaded area represents the mean percent CD8+Vβ17+/total CD8+ in fresh PBMC. The percent CD8+Vβ17+/total CD8+ cells in cord lines was significantly lower than that of adult lines (P < 0.0001 by unpaired Student's t-test). Fig. 2. View largeDownload slide TCR Vβ17 usage of influenza A/M158–66-specific CTL generated from subjects of different ages. Day 21 influenza A/M158–66 peptide-specific CTL lines generated from cord PBMC, PBMC from children (3 months to 13 years) and adult PBMC were analysed by two-colour immunofluorescent staining for distribution of Vβ17 within the CD8+ population. Data are expressed as percent CD8+Vβ17+ cells within the total CD8+ subset (CD8+Vβ17+/Total CD8+). The shaded area represents the mean percent CD8+Vβ17+/total CD8+ in fresh PBMC. The percent CD8+Vβ17+/total CD8+ cells in cord lines was significantly lower than that of adult lines (P < 0.0001 by unpaired Student's t-test). Fig. 3. View largeDownload slide Dependence of maternal and cord M158–66-specific CTL lines on Vβ17+ cells. Day 24 CTL lines were depleted of either Vβ17+ or Vβ22+ cells as a control. Depletions were confirmed to be >97% complete by FACS. 51Cr-release assays were performed using CIRA2 targets with (closed symbols) or without (open symbols) M158–66 peptide. Lysis of targets pulsed with a control Plasmodium cp36 peptide was not significantly different from that of targets without peptide. The percent CD8+Vβ17+/total CD8+ for M158–66-specific CTL lines before depletion is shown for each subject. Assays were performed in triplicate and the mean percent specific lysis ± SD determined. SD was <7% in each case. Fig. 3. View largeDownload slide Dependence of maternal and cord M158–66-specific CTL lines on Vβ17+ cells. Day 24 CTL lines were depleted of either Vβ17+ or Vβ22+ cells as a control. Depletions were confirmed to be >97% complete by FACS. 51Cr-release assays were performed using CIRA2 targets with (closed symbols) or without (open symbols) M158–66 peptide. Lysis of targets pulsed with a control Plasmodium cp36 peptide was not significantly different from that of targets without peptide. The percent CD8+Vβ17+/total CD8+ for M158–66-specific CTL lines before depletion is shown for each subject. Assays were performed in triplicate and the mean percent specific lysis ± SD determined. SD was <7% in each case. Fig. 4. View largeDownload slide Dependence of the influenza A/M158–66 -specific CTL response on Vβ17+ cells at different ages. Vβ17+ and control Vβ22+ cells were depleted from day 24 M158–66-specific CTL lines using sheep-anti mouse IgG-coated magnetic beads (Dynal). More than 97% depletion was confirmed by two-colour immunofluorescent staining. The Vβ17-depleted and control Vβ22-depleted CTL lines were tested for M158–66-specificity in a 4-h 51Cr-release assay using CIRA2 targets pulsed with M158–66 peptide (5 μg/ml) or without peptide. To compare cytotoxicity of the Vβ17- and Vβ22-depleted CTL lines, results are expressed in lytic units (LU)/107 cells, 1 LU denoting the quantity of effector cells needed to kill 20% of the M158–66-pulsed target cells. The LU activity of the Vβ17-depleted lines is expressed as a percentage of the LU activity of the Vβ22-depleted lines. Certain subjects have been highlighted for further discussion. Fig. 4. View largeDownload slide Dependence of the influenza A/M158–66 -specific CTL response on Vβ17+ cells at different ages. Vβ17+ and control Vβ22+ cells were depleted from day 24 M158–66-specific CTL lines using sheep-anti mouse IgG-coated magnetic beads (Dynal). More than 97% depletion was confirmed by two-colour immunofluorescent staining. The Vβ17-depleted and control Vβ22-depleted CTL lines were tested for M158–66-specificity in a 4-h 51Cr-release assay using CIRA2 targets pulsed with M158–66 peptide (5 μg/ml) or without peptide. To compare cytotoxicity of the Vβ17- and Vβ22-depleted CTL lines, results are expressed in lytic units (LU)/107 cells, 1 LU denoting the quantity of effector cells needed to kill 20% of the M158–66-pulsed target cells. The LU activity of the Vβ17-depleted lines is expressed as a percentage of the LU activity of the Vβ22-depleted lines. Certain subjects have been highlighted for further discussion. Fig. 5. View largeDownload slide The cord blood M158–66-specific CTL response is polyclonal, whereas the adult response is oligoclonal. M158–66-specific CTL lines were generated from PBMC obtained from a maternal and cord pair. PBMC and day 21 M158–66-specific CTL lines were stained for CD8 and Vβ gene segment usage and analysed by two-colour flow cytometry. The dependence on Vβ17+ CTL was determined by depleting Vβ17+ or control Vβ22+ cells from the lines and repeating cytotoxicity assays. The cytotoxicity of the Vβ17-depleted CTL line (in lytic units) was expressed as a percentage of that of the Vβ22-depleted CTL line; 100% suggests no loss of cytotoxicity and 0% reflects complete loss of M158–66-specific cytotoxicity. Fig. 5. View largeDownload slide The cord blood M158–66-specific CTL response is polyclonal, whereas the adult response is oligoclonal. M158–66-specific CTL lines were generated from PBMC obtained from a maternal and cord pair. PBMC and day 21 M158–66-specific CTL lines were stained for CD8 and Vβ gene segment usage and analysed by two-colour flow cytometry. The dependence on Vβ17+ CTL was determined by depleting Vβ17+ or control Vβ22+ cells from the lines and repeating cytotoxicity assays. The cytotoxicity of the Vβ17-depleted CTL line (in lytic units) was expressed as a percentage of that of the Vβ22-depleted CTL line; 100% suggests no loss of cytotoxicity and 0% reflects complete loss of M158–66-specific cytotoxicity. Transmitting editor: E. Simpson This study was supported by an MRC ROPA award. T. M. L. is an ARC Clinical Lecturer in Rheumatology. S. M. is a Royal Society University Research Fellow. References 1 Ruben, F. L. 1986. Who needs influenza vaccines? In Kendal, A. P. and Patriarca, A. R., eds, Options for the Control of Influenza, p. 139. Liss, New York. Google Scholar 2 Palache, A. M. 1992. Influenza subunit vaccine- ten years experience. Eur. J. Clin. Res.  3: 117. Google Scholar 3 Wiley, D. C., Wilson, I. A. and Skehel, J. J. 1981. Structural identification of the antibody-binding sites of Hong Kong influenza haemagglutinin and their involvement in antigenic variation. Nature  289: 373. Google Scholar 4 Potter, C. W. and Oxford, J. S. 1979. Determinants of immunity to influenza infection in man. Br. Med. Bull.  35: 69. Google Scholar 5 Doherty, P. C., Allan, W., Eichelberger, M. and Carding, S. R. 1992. Roles of αβ and γδ T cell subsets in viral immunity. Annu. Rev. Immunol.  10: 123. Google Scholar 6 Topham, D. J. and Doherty, P. C. 1998. Clearance of an influenza A virus by CD4+ T cells is inefficient in the absence of B cells. J. Virol.  72: 882. Google Scholar 7 McMichael, A. J. and Askonas, B. A. 1978. Influenza virus specific cytotoxic T cells in man: induction and properties of the cytotoxic T cell. Eur. J. Immunol.  8: 705. Google Scholar 8 Huet, S., Nixon, D. F., Rothbard, J., Townsend, A. R. M., Ellis, S. A. and McMichael, A. J. 1990. Structural homologies between two HLA-B27 restricted peptides suggest residues important for interaction with HLA-B27. Int. Immunol.  2: 311. Google Scholar 9 Morrison, J., Elvin, J., Latron, F., Gotch, F., Moots, R., Strominger, J. L. and McMichael, A. 1992. Identification of the nonamer peptide from influenza A matrix protein and the role of pockets of HLA-A2 in its recognition by cytotoxic T lymphocytes. Eur. J. Immunol.  22: 903. Google Scholar 10 Sutton, J., Rowland-Jones, S., Rosenberg, W., Nixon, D., Gotch, F., Gao, X. M., Murray, N., Spoonas, A., Driscoll, P., Smith, M., Willis, A. and McMichael, A. 1993. A sequence pattern for peptides presented to cytotoxic T lymphocytes by HLA-B8 revealed by the analysis of epitopes and eluted peptides. Eur. J. Immunol.  23: 447. Google Scholar 11 DiBrino, M., Tsuchida, T., Turner, R. V., Parker, K., Coligan, J. E. and Biddison, W. E. 1993. HLA-A1 and HLA-A3 T cell epitopes derived from influenza virus proteins predicted from peptide binding motifs. J. Immunol.  151: 5930. Google Scholar 12 Moss, P. A., Moots, R. J., Rosenberg, W. M., Rowland-Jones, S. J., Bodmer, H. C., McMichael, A. J. and Bell, J. I. 1991. Extensive conservation of alpha and beta chains of the human T-cell antigen receptor recognizing HLA-A2 and influenza A matrix peptide. Proc. Natl Acad. Sci. USA  88: 8987. Google Scholar 13 Casanova, J. L. and Maryanski, J. L. 1993. Antigen-selected T-cell receptor diversity and self-non-self homology. Immunol. Today  14: 391. Google Scholar 14 Lehner, P. J., Wang, E. C., Moss, P. A., Williams, S., Platt, K., Friedman, S. M., Bell, J. I. and Borysiewicz, L. K. 1995. Human HLA-A0201-restricted cytotoxic T lymphocyte recognition of influenza A is dominated by T cells bearing the V beta 17 gene segment. J. Exp. Med.  181: 79. Google Scholar 15 Sourdive, D. J., Murali-Krishna, K., Altman, J. D., Zajac, A. J., Whitmire, J. K., Pannetier, C., Kourilsky, P., Evavold, B., Sette, A. and Ahmed, R. 1998. Conserved T cell receptor repertoire in primary and memory CD8 T cell responses to an acute viral infection. J. Exp. Med.  188: 71. Google Scholar 16 Hlavac, F., Choppin, J. and Guillet, J. G. 1997. Self-peptide ligands affect T cell recognition of the homologous influenza A matrix virus peptide M.58–66: modification of the HLA- A2. 1/peptide complex structure and T cell antagonism. Hum. Immunol.  54: 48. Google Scholar 17 Batista, F. D. and Neuberger, M. S. 1998. Affinity dependence of the B cell response to antigen: a threshold, a ceiling, and the importance of the off rate. Immunity  8: 751. Google Scholar 18 Cheynier, R., Henrichwark, S. and Wain-Hobson, S. 1998. Somatic hypermutation of the T cell receptor V beta gene in microdissected splenic white pulps from HIV-1 positive patients. Eur. J. Immunol.  28: 1604. Google Scholar 19 McHeyzer-Williams, M. G. and Davis, M. M. 1995. Antigen-specific development of primary and memory T cells in vivo. Science  268: 106. Google Scholar 20 Maryanski, J. L., Jongeneel, C. V., Bucher, P., Casanova, J. L. and Walker, P. R. 1996. Single-cell PCR analysis of TCR repertoires selected by antigen in vivo: a high magnitude CD8 response is comprised of very few clones. Immunity  4: 47. Google Scholar 21 Bachmann, M. F., Speiser, D. E. and Ohashi, P. S. 1997. Functional maturation of an antiviral cytotoxic T-cell response. J. Virol.  71: 5764. Google Scholar 22 Busch, D. and Pamer, E. 1999. T cell affinity maturation by selective expansion during infection. J. Exp. Med.  189: 701. Google Scholar 23 Zheng, B., Han, S., Zhu, Q., Goldsby, R. and Kelsoe, G. 1996. Alternative pathways for the selection of antigen-specific peripheral T cells. Nature  384: 263. Google Scholar 24 Callan, M. F., Steven, N., Krausa, P., Wilson, J. D., Moss, P. A., Gillespie, G. M., Bell, J. I., Rickinson, A. B. and McMichael, A. J. 1996. Large clonal expansions of CD8+ T cells in acute infectious mononucleosis. Nat. Med.  2: 906. Google Scholar 25 Argaet, V., Schmidt, C. W., Burrows, S. R., Silim, S. L., Kurilla, M. G., Doolan, D. L., Suhrbier, A., Moss, D. J., Kieff, E., Suclley, T. B. and Misko, I. S. 1994. Dominant selection of an invariant T cell antigen receptor in response to persistent infection by Epstein–Barr virus. J. Exp. Med.  180: 2335. Google Scholar 26 Weekes, M. P., Wills, M. R., Mynard, K., Carmichael, A. J. and Sissons, J. G. P. 1999. The memory cytotoxic T-lymphocyte (CTL) response to human cytomegalovirus infection contains individual peptide-specific CTL clones that have undergone extensive expansion in vivo. J. Virol.  73: 2099. Google Scholar 27 Kalams, S. A., Johnson, R. P., Trocha, A. K., Dynan, M. J., Ngo, H. S., D'Aquila, R. T., Kurnick, J. T. and Walker, B. D. 1994. Longitudinal analysis of T cell receptor (TCR) gene usage by human immunodeficiency virus 1 envelope-specific cytotoxic T lymphocyte clones reveals a limited TCR repertoire. J. Exp. Med.  179: 1261. Google Scholar 28 Pantaleo, G., Demarest, J. F., Soudeyns, H., Graziosi, C., Denis, F., Adelsberger, J. W., Borrow, P., Saag, M. S., Shaw, G. M., Sekaly, R. P., Soudeyns, H., Corey, L. and Fauci, A. S. 1994. Major expansion of CD8+ T cells with a predominant Vβ usage during the primary immune response to HIV. Nature  370: 463. Google Scholar 29 Storkus, W. J., Howell, D. N., Salter, R. D., Dawson, J. R. and Cresswell, P. 1987. NK susceptibility varies inversely with target cell class I HLA antigen expression. J. Immunol.  138: 1657. Google Scholar 30 Plebanski, M., Allsopp, C. E. M., Aidoo, M., Reyburn, H. and Hill, A. V. S. 1994. Induction of peptide-specific primary cytotoxic T lymphocyte responses from human peripheral blood. Eur. J. Immunol.  25: 1783. Google Scholar 31 Bakker, A. B., Marland, G., Boer, A. J., Huijbens, R. J., Danen, E. H., Adema, G. J. and Figdor, C. G. 1995. Generation of antimelanoma cytotoxic T lymphocytes from healthy donors after presentation of melanoma-associated antigen-derived epitopes by dendritic cells in vitro. Cancer Res  55: 5330. Google Scholar 32 Ridge, J. P., Fuchs, E. J. and Matzinger, P. 1996. Neonatal tolerance revisited: Turning on newborn T cells with dendritic cells. Science  271: 1723. Google Scholar 33 Cerny, A., Fowler, P., Brothers, M. A., Houghton, M., Schlicht, M. H. and Chisari, F. V. 1995. Induction in vitro of a primary human antiviral cytotoxic T cell response. Eur. J. Immunol.  25: 627. Google Scholar 34 Griffiths, P., Ronalds, C. and Heath, R. 1980. A prospective study of influenza infections during pregnancy. J. Epidemiol. Commun. Health  34: 124. Google Scholar 35 Xu, H., Yasui, O., Tsuruoka, H., Kuroda, K., Hayashi, K., Yamada, A., Ishizaki, T., Yamada, Y., Watanabe, T. and Hosaka, Y. 1998. Isolation of type B influenza virus from the blood of children. Clin. Infect. Dis.  27: 654. Google Scholar 36 Gammon, G., Klotz, J., Ando, D. and Sercarz, E. E. 1990. The T cell repertoire to a multideterminant antigen. Clonal heterogeneity of the T cell response, variation between syngeneic individuals, and in vitro selection of T cell specificities. J. Immunol.  144: 1571. Google Scholar 37 Mackensen, A., Carcelain, G., Viel, S. Raynal, M. C., Michalaki, H., Triebel, F., Bosq, J. and Kercend, T. 1994. Direct evidence to support the immunosurveillance concept in a human regressive melanoma. J. Clin. Invest.  93: 1397. Google Scholar 38 Shilyansky, J., Nishimura, M. I, Yanneli, J. R, Kawakami, Y., Jacknin, L. S., Charmley, P. and Rosenberg, S. A. 1994. T-cell receptor usage by melanoma-specific clonal and highly oligoclonal tumour-infiltrating lymphocyte lines. Proc. Natl Acad. Sci. USA  91: 2829. Google Scholar 39 Cibotti, R., Cabaniols, J. P, Pannetier, C., Delarbre, C., Vergnon, I., Kanellopoulos, J. M. and Kourilsky, P. 1994. Public and private Vβ T cell receptor repertoires against hen egg white lysozyme (HEL) in nontransgenic versus HEL transgenic mice. J. Exp. Med.  180: 861. Google Scholar 40 Lalvani, A., Brookes, R., Hambleton, S., Britton, W. J., Hill, A. V. and McMichael, A. J. 1997. Rapid effector function in CD8+ memory T cells. J. Exp. Med.  186, 6: 859. Google Scholar 41 Gilfillan, S., Bachmann, M., Trembleau, S., Adorini, L., Kalinke, U., Zinkernagel, R., Benoist, C. and Mathis, D. 1995. Efficient immune responses in mice lacking N-region diversity. Eur. J. Immunol.  25: 3115. Google Scholar 42 Gavin, M. A. and Bevan, M. J. 1995. Increased peptide promiscuity provides a rationale for the lack of N regions in the neonatal T cell repertoire. Immunity  3: 793. Google Scholar 43 Bodger, M. P., Janossy, G., Bollum, F. J., Burford, G. D. and Hoffbrand, A. V. 1983. The ontogeny of terminal deoxynucleotidyl transferase positive cells in the human fetus. Blood.  61: 1125. Google Scholar 44 McMichael, A. and O'Callaghan, C. 1998. `A new look at T cells'. J. Exp. Med.  187: 1367. Google Scholar 45 Crawford, F., Kozono, H., White, J., Marrack, P. and Kappler, J. 1998. Detection of antigen-specific T cells with multivalent class II MHC covalent peptide complexes. Immunity  8: 675. Google Scholar © 2001 Japanese Society for Immunology http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Immunology Oxford University Press

Influenza A antigen exposure selects dominant Vβ17+ TCR in human CD8+ cytotoxic T cell responses

Download PDF
9 pages

Loading next page...
 
/lp/oxford-university-press/influenza-a-antigen-exposure-selects-dominant-v-17-tcr-in-human-cd8-0gAC3AXk3a

References (43)

Publisher
Oxford University Press
Copyright
© 2001 Japanese Society for Immunology
ISSN
0953-8178
eISSN
1460-2377
DOI
10.1093/intimm/13.11.1373
Publisher site
See Article on Publisher Site

Abstract

Abstract During acute human viral infections, such as influenza A, specific cytotoxic T lymphocytes (CTL) are generated which aid virus clearance. We have observed that in HLA-A*0201+ subjects, CTL expressing Vβ17+ TCR and recognizing a peptide from the influenza A matrix protein (M158–66) dominate this response. In experimental models of infection such dominance can be due to inheritance of a restricted T cell repertoire or acquired consequent on expansion of CTL bearing an optimum TCR conformation against the MHC–peptide complex. To examine how influenza A infection might influence the development of TCR V0β17 expansion, we studied influenza A-specific CTL in a cross-sectional study of 82 HLA-A*0201+ individuals from birth (cord blood) to adulthood. Primary M158–66 -specific CTL were detected in cord blood, but their TCR were diverse and depletion of Vβ17+ cells did not abrogate specific cytotoxicity. In contrast following natural influenza A infection, TCR Vβ17+ CTL dominated to the extent that only one of nine adult CTL lines retained any functional activity after in vitro depletion of Vβ17+ CTL. These results suggest that the dominance of Vβ17+ TCR among adult M158–66-specific CTL results from maturation and focussing of the response driven by exposure to influenza, and have implications for optimum immunization strategies. affinity, cytotoxic T cell, human, influenza A, TCR CTL cytotoxic T lymphocyte, HA haemaglutinin, HI haemagglutinin inhibition, NA neuraminidase, PBMC peripheral blood mononuclear cell, TdT terminal deoxynucleotidyl transferase Introduction Influenza A is a common respiratory pathogen in man causing frequent epidemics and occasional pandemics associated with considerable morbidity and mortality (1,2). The importance of neutralizing antibodies in immunity to influenza A is clear. The main epitopes for these antibodies are on the surface haemaglutinin (HA) and neuraminidase (NA) proteins, which are highly variable and selected for by the humoral response (3). Antigenic drift and antigenic shift result in structural differences in these proteins amongst the different strains of influenza A and neutralizing antibodies are therefore strain-specific (4). However, CD8+ cytotoxic T lymphocytes (CTL), which are important in the clearance of influenza A and other virus infections (5,6), do not discriminate between major virus subtypes (7). This is because most influenza A-specific CTL epitopes are derived from conserved internal viral proteins (8–11). The TCR repertoire used by CD8+ T cells recognizing a defined viral epitope may be conserved or diverse, but little is known about the factors contributing to this selection in man (12–15). One explanation for conserved TCR usage is that endogenously processed self peptides homologous to an antigenic peptide `trim' the TCR repertoire during thymic selection (13,14,16). Conserved TCR usage could also represent maturation and focusing of the CD8+ T cell response following primary or repeated viral infection. Maturation of B cell responses to antigen is well described, and occurs by a combination of somatic hypermutation in the CDR3 region and antigen-driven selection for cells expressing receptors with the highest affinity for antigen (17). There is little evidence for somatic hypermutation in T cells (18). However, the ability of antigen exposure to selectively expand high-avidity T cell populations has been examined in CD4+ and CD8+ T cell responses in murine models (15,19–22). Conflicting results both supporting narrowing of the TCR repertoire during secondary T cell responses (19,21–23) and development of oligoclonality during primary exposure (15,20) have been reported. These studies have been often limited by use of transgenic mice responding to artificial antigens, responses not surprisingly characterized by TCR conservation in the primary response. To date we do not know the relevance of these observations to acute viral infections such as influenza A in man. Such studies are difficult because of the outbred population, controlling antigen exposure and the ethics of studies involving children. To date human studies have been confined to Epstein–Barr virus (24,25), cytomegalovirus (26) and HIV (27,28) infections which establish life-long persistence rather than repeated acute exposure. The HLA-A*0201-restricted influenza A/M158–66 peptide-specific CTL response is dominated by and highly dependent on CD8+ T cells expressing TCR incorporating the Vβ17 and Vα10 gene segments (12,14). An endogenous peptide from the TAP.2 protein with strong sequence homology to the M158–66 peptide has been described and it was suggested that this peptide may have a role in restricting the M158–66-selected TCR repertoire (14,16). However, it is not known whether this peptide is naturally processed and expressed in the thymus. Furthermore, the TCR Vβ17 usage of primary M158–66-specific CTL has not been investigated. We investigated the role of antigen-driven maturation in this CTL response by determining TCR Vβ17 usage of M158–66-specific CTL in a cross-sectional study of 82 individuals from birth to adulthood. We compared the functional properties of M158–66-specific CTL generated from cord and maternal blood, and show that the oligoclonality of Vβ17+ CTL is probably antigen driven as a consequence of natural infection. Methods Subjects Permission to collect cord blood was obtained from women on arrival at the delivery suite at the University Hospital of Wales, Cardiff. Following delivery of the placenta, the clamped umbilical cord was cleaned and a 14 gauge needle inserted into the central umbilical vessels. Between 15 and 150 ml of blood was collected, the final volume depending on the stage at which the cord was clamped; 50ml of maternal blood was collected whenever possible. The children in this study attended the Paediatric Outpatient Department at the University Hospital of Wales (aged 3 months to 15 years), being treated for idiopathic epilepsy or recurrent febrile convulsions with sodium valproate or vigabatrin with blood level monitoring. After written informed consent was given by the parent/guardian, ~2–5 ml of blood additional to that which was being taken for drug levels was collected into a universal container containing preservative-free heparin (10 IU/ml). Blood was also collected from normal laboratory volunteers. The study was approved by the local research ethics committee and written informed consent was obtained from all subjects or guardians. None of the subjects had been immunized against influenza. HLA typing Individuals expressing HLA-A*0201 were identified using the MA2.1 mAb and single-colour immunofluorescence to screen blood samples (14). This was confirmed for several individuals by DNA typing. Influenza A serology Anti-influenza antibody production was determined with the haemagglutinin inhibition (HI) test. All serum samples were stored at –20°C until assayed. Before investigation, they were treated with receptor destroying enzyme from Vibrio cholerae overnight at 37°C and then inactivated at 56°C for 30 min. The HI test was performed with eight agglutinating doses of virus (A/Taiwan/1/86, A/Bayern/7/95, A/JHB/34/94, A/Wuhan/359/95 and A/Sydney/5/97) as appropriate prototype viruses which had circulated during the course of this study. Each test was performed in duplicate with all antigens. Generation of M158–66 -specific CTL lines HLA-A*0201-restricted influenza A/M158–66 -specific CTL lines were generated as previously described (14). In brief, freshly isolated peripheral blood mononuclear cells (PBMC) from cord blood, or blood obtained from children of various ages or adults were infected with 200 HAU of influenza A (AX/31). After 1 h incubation at 37°C, cells were cultured in RPMI-10(AB) [RPMI 1640 (Sigma, St Louis, MO), 10% pooled human AB serum, 2 mM l-glutamine, 1 mM sodium pyruvate and 25 mM HEPES buffer] for 7 days. On day 7 and weekly thereafter cells were harvested, counted and replated in 24 well plates at 2×106 cells/well in RPMI-10(AB) containing 10 IU/ml IL-2 (Boehringer Mannheim, Mannheim, Germany). Cultures were re-stimulated with irradiated (3000 rad for 7 min) M158–66 peptide-pulsed (5 μg/ml) CIRA2 cells at a responder to stimulator ratio of 10:1. On day 21, cultures were tested for influenza A and M158–66 specificity by 51Cr-release assay, and the CD8+Vβ17+ content of M158–66-specific lines determined by two-colour flow cytometry as described (14). Identical re-stimulation conditions were used to generate HLA-A*0201-restricted M158–66-specific CTL lines from PBMC obtained from cord blood, children and adults. Removal of Vβ17+ cells from HLA-A*0201-specific CTL lines Immunomagnetic depletion was performed to remove Vβ17+ cells or as a control Vβ22+ cells (Dynabeads M-450 sheep anti-mouse IgG; Dynal, The Wirral, UK) as previously described (14). Depletion was typically >97%. Staining and analysis of CD8+ T cells Cells were analysed by flow cytometry using a FACScan (Becton Dickinson, La Jolla, CA). Live cells were gated based on FSC/SSC profiles. TCR Vβ gene segment usage was determined by two-colour immunofluorescence using a panel of FITC-conjugated anti-TCR Vβ mAb (TCR Workshop, San Francisco, 1995) and anti-CD8–phycoerythrin (Becton Dickinson). The staining procedure was as previously described (14). Cytotoxicity assays Standard 51Cr-release assays were performed using 51Cr-labelled CIRA2 cells (2000/well) which express a transfected genomic clone of HLA-A*0201 (29). Targets were pulsed with M158–66 peptide (10–6 M) or infected with 200 HAU influenza A as described (14), and control targets included CIRA2 cells in the absence of peptide and CIRA2 cells pulsed with the HLA-A*0201-binding Plasmodium cp36 peptide (30). Percent specific lysis was calculated as 100×experimental release – spontaneous release)/(maximum release – spontaneous release). Spontaneous release was <10% unless otherwise stated. CTL lines were assayed for specific cytotoxicity after 21 days in culture. Results HLA-A*0201-restricted influenza A-specific cord blood T cell responses in cord blood To minimize the effects of antigen exposure on development of the HLA-A*0201-restricted influenza A/M158–66 response, CTL responses were examined in PBMC obtained from umbilical cord blood. Using a re-stimulation protocol previously described for the generation of in vitro primary CTL responses (31), influenza A/M158–66-specific CD8+ T cell responses were generated from four of five HLA-A*0201+ cord blood samples (data not shown). However, this method could not be used for generating CTL from infants because the quantity of blood available was limited to 1–5 ml, thus precluding the generation of autologous dendritic cells. For three HLA-A*0201+ cord blood and three adult blood samples we confirmed that different re-stimulation protocols did not alter the TCR Vβ gene segment usage of the M158–66-specific CTL lines generated (data not shown). However, in order to compare TCR usage between the different individuals in this study it was essential to use identical re-stimulation conditions. We therefore determined whether a re-stimulation protocol conventionally used for the in vitro detection of secondary responses (7) could be applied to cord blood. Surprisingly, M158–66-specific CD8+ T cell responses were generated from 60% (30 of 50) of HLA-A*0201 cord blood samples using this protocol which was thereafter used for generating CTL lines from all subjects included in this study. All M158–66-specific CTL lines killed influenza A virus-infected targets in addition to M158–66 peptide-pulsed targets (Fig. 1). The possibility that these cord blood CD8+ T cell responses resulted from contamination of cord blood with maternal blood was excluded as M158–66-specific lines could be generated from the cord blood of HLA-A*0201-negative mothers, who were unable to generate M158–66-specific CD8+ T cell responses (data not shown). In addition, the MHC types of CTL lines generated from five paired maternal and cord blood samples were determined, and haplotype differences between maternal and cord M158–66-specific CTL lines confirmed (data not shown). To exclude the possibility of intrauterine exposure to influenza A, anti-influenza A antibodies were measured during the first trimester of pregnancy and postnatally from seven mothers, whose infant's cord blood yielded M158–66-specific CTL responses. Two of seven mothers had serological evidence suggestive of influenza A exposure during pregnancy (data not shown). M158–66-specific CTL lines generated from cord blood samples obtained from these two individuals were no different from other cord CTL lines in terms of specific cytotoxicity or TCR Vβ17 usage. These findings together with the absence of any evidence supporting transplacental transfer of influenza A suggest that M158–66-specific CTL responses in cord blood are likely to be primary CTL responses elicited in vitro. TCR Vβ17 usage by M158–66-specific CD8+ T cells The ability to generate probable primary HLA-A*0201-restricted influenza A/ M158–66-specific CTL responses from cord blood using exactly the same re-stimulation protocol as for children and adults, allowed comparison of the TCR Vβ17 usage between primary and secondary effector CD8+ CTL responses. Freshly isolated PBMC and M158–66-specific CTL lines generated in vitro from nine paired cord and maternal samples were examined. Although all CTL lines were influenza A virus specific and M158–66 specific in 51Cr-release assays, cord-derived CTL lines contained significantly fewer CD8+Vβ17+ cells than the corresponding maternal CTL (Table 1). Further M158–66-specific CTL lines were generated from unrelated cord and adult blood samples, and also from PBMC obtained from children of different ages using the same in vitro re-stimulation conditions. Flow cytometric analyses showed that the frequency of CD8+Vβ17+ T cells in M158–66-specific CTL lines increased with age (Fig. 2). An individual example of an in vitro primary response can be seen in child BH. This child was 3 months old and the M158–66-specific CTL line generated contained 14% CD8+Vβ17+cells (Fig. 2). This child had no serological evidence of exposure to influenza A by the HI test (Table 2). The negative serology also implied that the mother had not been exposed to influenza A during pregnancy as maternal antibodies would otherwise have been detected. An individual example of in vivo pathogen exposure influencing Vβ17 expression can be seen in child DH, for whom two samples were obtained 1 year apart. No CTL were generated from the first sample and only 5% of CD8+ T cells in the day 21 culture were Vβ17+. However, 75% of the CD8+ cells in the second sample expressed Vβ17 and M158–66-specific CTL activity was detected (Table 1). The first serum sample from child DH (age 21months) was negative by the HI test, but antibodies were detected in the second sample (age 3 years) suggesting exposure to influenza A between the two samples (Table 2). As initial serology was negative it is likely that this was the first exposure to influenza A. The cord blood M158–66-specific CD8+ T cell response is less dependent on the presence of the CD8+Vβ17+ T cell population than the adult response Previous studies suggested that the M158–66-specific CTL response in adults is highly dependent on the presence of CD8+Vβ17+ CTL (14). Although the number of CD8+ cells expressing Vβ17 in cord blood M158–66-specific CTL lines was significantly reduced compared to adult lines (Fig. 2) it was still possible that M158–66-specific cytotoxicity might be abrogated following removal of this smaller Vβ17+ population. To determine the relative contribution of Vβ17+ cells in M158–66-specific CTL lines generated from cord blood and subjects of different ages, CTL lines were depleted of either Vβ17+ or, as a control, Vβ22+ cells. Comparison of four maternal and cord CTL lines grown in parallel, confirmed that the cord blood M158–66-specific CTL response was not dependent on the presence of the CD8+Vβ17+ population. Depletion of Vβ17+ cells did not decrease the lytic activity of M158–66-specific CTL lines with low frequencies of Vβ17+ cells (cord DR, SG and JC). (Fig. 3). In contrast, the maternal lines were highly dependent on the presence of these cells, the M158–66-specific response being abrogated in three of four cases and markedly reduced in the other (Fig. 3). The difference in the proportion of CTL clones using the Vβ17 gene segment in cord and adult blood was also determined by comparing the M158–66-specific CTL response in cord blood and the corresponding maternal sample by limiting dilution analysis. Using this technique, 20% of cord blood-derived CTL clones were Vβ17+ compared with 80% of maternal CTL clones (data not shown). These experiments were extended to examine M158–66-specific CTL lines from further adults, cord blood samples and also CTL lines from children at different ages in a cross-sectional study. The M158–66-specific CTL line of child BH contained 14% CD8+Vβ17+ cells and the response was not affected by removing these cells (Fig. 4). The M158–66-specific CTL response for child DH following exposure to influenza A was markedly reduced following depletion of Vβ17+ cells, thus being similar to adult CTL lines (Fig. 4). No cord M158–66-specific CTL lines (out of eight) were dependent on the presence of Vβ17+ cells, whilst only one adult line (MJ) was partially independent of Vβ17+ cells (MJ) (Fig. 4). The diversity of the TCR repertoire in the cord blood M158–66-specific CTL response was confirmed using a panel of anti-TCR Vβ mAb. Expansion of M158–66-specific CD8+ CTL expressing other TCR Vβ gene segments was detected in cord-derived CTL, but not in the paired maternal sample (Fig. 5). In the representative example shown, the cord M158–66-specific CTL line had significant expansions of CD8+ cells expressing Vβ5.1, Vβ6 and Vβ17 gene segments (Fig. 5). Depletion experiments confirmed that the cord blood response was independent of Vβ17+ cells, but removal of Vβ17+, Vβ5.1+ and Vβ6+ cells resulted in loss of >95% of cytotoxic activity (Fig 5). The maternal response was dependent on the presence of the Vβ17+ population (Fig. 5). The Vβ gene segment usage in cord M158–66-specific CTL varied between subjects. Vβ17 dependence of the M158–66-specific CD8+ T cell response and natural exposure to influenza A Influenza A HI serology is difficult to interpret in cord blood and in samples obtained during the first year of life, in that positive serology may represent the transplacental transfer of maternal antibodies. Negative serology in such samples is useful, however, and supports the absence of recent exposure to influenza A. The negative HI serology on child BH provided good evidence that this 3-month-old child had not been exposed to influenza A and that the M158–66-specific response represented an in vitro primary CD8+ T cell response (Table 2). HI serology was performed on serum samples from children when available (Table 2). Although the association between dominant Vβ17 usage and positive serology was not absolute, the in vitro M158–66-specific CD8+ T cell response generated from the majority of children with evidence of exposure to influenza A was dominated by CD8+Vβ17+ cells (Table 2). Discussion In this study we have compared Vβ17 usage by M158–66-specific CTL generated from cord blood, and blood obtained from children of various ages and adults. We have demonstrated that the dominant usage of Vβ17-expressing CD8+ T cells in the adult response is acquired, and probably the result of antigen-driven selection and focusing of the response. While a role for thymic selection involving self-peptides homologous to M158–66 cannot be excluded, our results argue against this mechanism. Diverse TCR usage was found among M158–66-specific CTL generated from cord blood, whereas conserved TCR usage might be expected if a thymic selection mechanism was operating. These results confirm that maturation and focusing of CD8+ T cell responses by clonal selection occurs in man in a similar manner to that recently described in a murine model (22). This focusing of a T cell response has potential advantages for the host organism in that the selection of CTL with the highest avidity for a particular epitope–MHC complex may favour the early detection and subsequent lysis of infected cells following re-exposure to the same virus. In addition to highlighting the ability of the immune system to focus a CTL response after infection, these findings demonstrate the important role of repeated antigenic challenge and the likely ineffectiveness of inappropriately selected single peptide immunizations. Many studies have shown that generation of `primary' CTL responses requires special protocols. Studies in neonatal mice have suggested that `dendritic cell' antigen-presenting cells are essential for the generation of primary CTL (32). Furthermore, the in vitro generation of primary CD8+ T cell responses requires cytokines such as IL-7 (30) and/or extended tissue culture with multiple secondary in vitro re-stimulations (33). Therefore the generation of primary M158–66-specific CTL responses from cord blood using a standard protocol (7) was surprising, especially given that previous attempts using a similar protocol were unsuccessful (14). This discrepancy might be explained by a small difference in culture conditions; we were unable to detect M158–66-specific cytotoxicity from cord lines before day 16, whereas Lehner et al. (14) measured responses on day 14. The presence of HLA-A*0201-restricted M158–66-specific CTL in cord blood could also be explained by contamination of cord blood by maternal blood (transplacentally or during delivery) or in utero exposure of the foetus to influenza A. Contamination by maternal blood was excluded. In utero exposure to influenza would require maternal influenza A infection during pregnancy, influenza viraemia and transplacental transfer of infection. Previous studies have estimated that between 5 and 13% of mothers may be infected with influenza A during pregnancy (34). Two out of seven mothers we tested antenatally and postnatally using HI showed evidence of seroconversion to influenza A antigens. This is still far less than the 65% cord blood CTL response rate observed in this study. There is only very limited evidence for viraemia during acute natural influenza A infection (35) and no evidence for transplacental transfer of influenza in humans (34). These facts suggest that in utero exposure to influenza A is an unlikely explanation for the ability to generate HLA-A*0201-restricted M158–66-specific CTL responses from cord blood in vitro. It is likely, therefore, that cord blood M158–66-specific CTL responses are in vitro primary responses and result from an unusual ability of the HLA-A*0201/M158–66 complex to activate naive as well as memory M158–66-specific CTL precursors. The primary nature of this response is supported by the observation that influenza virus-infected stimulator cells were required at the onset of culture to generate M158–66-specific CTL lines from cord PBMC, whereas CTL lines could be generated from adult PBMC using M158–66peptide alone (data not shown). The TCR Vβ repertoire of M158–66-specific CTL in the in vitro primary cord blood CTL response was more heterogeneous and less dependent on the presence of Vβ17+ cells than that of adults. However, expanded populations of Vβ17+ CTL were frequently observed, in some cases comprising up to 60% of the total CD8+ population. This confirmed that the differences observed between adult and cord lines were not the result of an inability of Vβ17+ cell in cord blood to proliferate in vitro. All studies involving in vitro culture are prone to introduction of bias as a result of preferential outgrowth of those antigen-specific cells with the greatest proliferative capacity (36–39). However, using ELISPOT technology to detect M158–66-specific IFN-γ production by PBMC directly ex vivo, it has been confirmed that the dominance of Vβ17+ M158–66-specific CTL in adults is not an in vitro artefact (40). In this study we limited the potential for in vitro bias by using identical in vitro re-stimulation conditions and culturing cells for the same duration prior to assay. Maternal and cord lines were cultured and re-stimulated in parallel. Failure to generate M158–66-specific CTL lines from some HLA-A*0201 children who had serological evidence of exposure to influenza A may have been due to the technical difficulties involved in establishing CTL lines in vitro from 1–5 ml of blood. It seems likely that the increasing frequency of Vβ17+ CTL that occurs in the M158–66-specific HLA-A*0201-restricted CTL response with age is the result of natural exposure to influenza A. Longitudinal follow up of individuals for several years from birth could confirm this and address a possible alternative explanation. This is based on findings in murine studies which demonstrate that mice deficient in terminal deoxynucleotidyl transferase (TdT) have altered TCR repertoires when compared to wild-type mice (41,42). In particular, a focused TCR Vβ8.2 H-2d-restricted response against SWM110–121 peptide is not seen in TdT-deficient mice (41). TdT is not expressed in mice until several days after birth. TdT expression has not been extensively studied in man (43), however it is possible that differences in TdT expression between cord and adult blood could influence the diversity of the TCR repertoire. While this cannot be excluded, the functional superiority of the Vβ17 TCR for influenza M158–66 peptide compared to no Vβ17 TCR (as demonstrated in the accompanying paper) suggests that exposure to influenza is the major factor in shaping the adult TCR repertoire. In this cross-sectional study we determined previous influenza A exposure by HI which, although a frequently used method for determining influenza A serology, is still associated with a number of drawbacks. We were unable to obtain multiple samples from many of the children in this study for ethical reasons. Isolated samples can still be useful in determining whether the child has been exposed to influenza A, but as HI serology can wane with time, a negative result must be interpreted with caution. In addition interpretation of HI serology in children during the first year of life is complicated by the fact that maternal antibodies may still be present during this period. If the HI serology is negative in this age group, it is unlikely that they have been exposed to influenza A although the possibility that they have been exposed to influenza and failed to respond serologically or exposed to a different strain cannot be excluded. However, it is likely that CTL lines generated from these individuals represent in vitro primary responses. As the switch from a non-Vβ17+-dominated cord response (probable primary response) to a Vβ17+-dominated CTL response occurred at a young age, it is possible that a single in vivo exposure to influenza A may be sufficient to alter the TCR repertoire of the M158–66-specific CTL response as in the case of child DH (Table 2). Neonates are likely to be protected from infection by transfer of maternal antibody up to ~9 months and with a single serum sample later infection would be detected. No major antigenic shift or drift in the circulating influenza A strains, nationally or locally, has been observed during the course of this study, therefore repeated exposure is unlikely. Our findings suggest that the primary M158–66-specific CTL response is mediated by M158–66-specific CD8+ T cells with heterogeneous TCR usage and that the TCR repertoire narrows following the first in vivo exposure to influenza A such that the secondary in vitro response is dominated by Vβ17+ CTL. It is possible that the various M158–66-specific clones have different `burst frequencies' during the primary response. The Vβ17+ CTL may have a larger burst frequency because of a higher avidity for HLA-A*0201/M158–66 and thus a functional advantage, resulting in their dominance of the memory M158–66-specific CD8+ T cells. Recently tetramers comprising soluble HLA class I–peptide complexes have been used to quantitate directly the numbers of anti-viral CTL in blood (44), reducing the need to perform in vitro CTL assays. Furthermore such tetramers have been used as a tool to estimate TCR affinity (22,45). We considered using tetramers to compare the TCR repertoires of M158–66-specific CTL isolated directly from cord and adult blood, thus avoiding potential bias introduced by in vitro culture. However, preliminary experiments showed that there was no detectable binding of HLA-A*0201/M158–66 tetramer to cord blood PBMC due to the low frequency of these cells. Furthermore, in the next paper in this issue we show that although all Vβ17+ CTL clones stain with tetramer, two out of three non- Vβ17+ clones fail to stain. Thus a major proportion of the M158–66-specific CTL in cord blood would not be detected using tetramers. The description of conserved TCR usage by CTL recognizing epitopes from several viruses in man such as cytomegalovirus (26), Epstein–Barr virus (24,25) and HIV (27,28) suggests that maturation of these responses may also occur. Clonal selection of HIV-specific CTL has been noted following the primary HIV seroconversion illness (27,28), but confirmation of `maturation' will require comparison of the TCR repertoires and functional properties of primary and secondary CTL specific for these antigens. The demonstration that a natural antigen-specific CD8+ T cell response in man matures following the first encounter with antigen confirms that factors other than thymic selection can affect the repertoire of T cells responding to a particular peptide epitope and adds to our understanding of the interaction between T cells and targets expressing appropriate MHC–peptide complexes. In vaccination programmes seeking to induce specific CTL responses the implication of these observations is that the first encounter with the antigen is important and hence its formulation, depending on the desire to induce dominant oligoclonal or subdominant CTL, will be crucial. Table 1. TCR Vβ17+ usage and cytotoxicity of M158–66-specific CTL lines generated from paired cord and maternal blood samples Subject  Percent lysis ± SD (E:T 20:1)a  Percent CD8+Vβ17+/total CD8+      PBMC  CTL d21b  aCytotoxicity assays were performed after 21 days in culture. Percent lysis was calculated as follows: 100×(experimental release – spontaneous release)/(maximum release – spontaneous release). Spontaneous release was <10% in all cases. The percent lysis using an E:T of 20:1 is shown.  bThe percentage of CD8+ cells expressing Vβ17 in PBMC was determined prior to in vitro re-stimulation and for M158–66-specific CTL, Vβ17 usage was determined on day 21.  Mother HW  60 ± 10  6  85  Cord HW  40 ± 5  7  3  Mother BE  65 ± 4  4  81  Cord BE  50 ± 8  2  20  Mother TU  55 ± 9  4  75  Cord TU  36 ± 6  2  4  Mother CO  60 ± 11  2.5  50  Cord CO  40 ± 6  4  6  Mother DR  55 ± 5  10  58  Cord DR  40 ± 5  6  7  Mother LC  40 ± 4  3  46  Cord LC  60 ± 10  3  25  Mother SG  40 ± 9  2.5  44  Cord SG  52 ± 8  2  9  Mother CB  60 ± 10  2.8  75  Cord CB  35 ± 8  3  9  Mother JC  50 ± 7  4  68  Cord JC  33 ± 7  3.5  6  Subject  Percent lysis ± SD (E:T 20:1)a  Percent CD8+Vβ17+/total CD8+      PBMC  CTL d21b  aCytotoxicity assays were performed after 21 days in culture. Percent lysis was calculated as follows: 100×(experimental release – spontaneous release)/(maximum release – spontaneous release). Spontaneous release was <10% in all cases. The percent lysis using an E:T of 20:1 is shown.  bThe percentage of CD8+ cells expressing Vβ17 in PBMC was determined prior to in vitro re-stimulation and for M158–66-specific CTL, Vβ17 usage was determined on day 21.  Mother HW  60 ± 10  6  85  Cord HW  40 ± 5  7  3  Mother BE  65 ± 4  4  81  Cord BE  50 ± 8  2  20  Mother TU  55 ± 9  4  75  Cord TU  36 ± 6  2  4  Mother CO  60 ± 11  2.5  50  Cord CO  40 ± 6  4  6  Mother DR  55 ± 5  10  58  Cord DR  40 ± 5  6  7  Mother LC  40 ± 4  3  46  Cord LC  60 ± 10  3  25  Mother SG  40 ± 9  2.5  44  Cord SG  52 ± 8  2  9  Mother CB  60 ± 10  2.8  75  Cord CB  35 ± 8  3  9  Mother JC  50 ± 7  4  68  Cord JC  33 ± 7  3.5  6  View Large Table 2. Association between dependence of M158–66-specific CTL response on the presence of Vβ17+ CTL and previous exposure to influenza A in children of different ages Child  Age  Percent CD8+Vβ17+/total CD8+a  CTL lysis(E:T 20:1)b  Percent lytic activity post Vβ17 depletion c  HI Serologyd    PBMCa  CTL (d21)a    A/Taiwan/1/86 (H1N1)  A/Bayern/7/95 (H1N1)  A/JHB/34/94 (H3N2)  A/Wuhan/359/95 (H3N2)  A/Sydney/5/97 (H3N2)  aThe percentage of CD8+ T cells expressing Vβ17 was determined by FACS analysis for fresh PBMC prior to in vitro stimulation and for M158–66-specific CTL after 21 days in culture.  b51Cr-release assays were performed on day 21 using CIRA2 targets. The percentage lysis of M158–66-pulsed targets is shown for an E:T ratio of 20:1. Lysis of targets pulsed with an irrelevant HLA-A*0201-binding Plasmodium peptide and targets without peptide was <10% in each case.  cDay 24 CTL lines with M158–66 specificity were depleted of Vβ17+ cells or as a control Vβ22+ cells. Vβ22 depletion did not alter cytotoxic activity in a single case. Cytotoxicity was expressed in lytic units and the residual cytotoxicity after depletion of Vβ17+ cells was expressed as a percentage of that observed following depletion of Vβ22+ cells.  dAnti-influenza antibody production was determined with the HI test. The HI test was performed with eight agglutinating doses of five appropriate prototype viruses which had circulated during the course of this study.  ND, not done.  TV  3 months  2  ND  0  ND  <10  <10  <10  <10  <10  BH  3 months  5  14  50  100  <10  <10  <10  <10  <10  DS  6 months  3  4  0  ND  <10  <10  <10  <10  <10  JR  2 years 5 months  4.50  5  30  110  <10  <10  160  120  40  DH (#1)  21 months  3.2  5  0  ND  <10  <10  <10  <10  <10  DH (#2)  2 years 9 months  3.5  70  75  20  <10  <10  <10  40  10  EB (#1)  2 years 8 months  3  52  40  10  <10  <10  160  40  80  EB (#2)  3 years 2 months  3.5  76  75  5  <10  <10  160  160  80  EJ  2 years 9 months  ND  5  10  ND  <10  <10  40  80  40  SP  3 years  ND  80  45  0  <10  <10  80  80  80  IS (#1)  3 years  2.8  4.2  0  ND  <10  <10  <10  <10  <10  IS (#2)  3 years 3 months  ND  ND  0  ND  <10  <10  <10  <10  <10  RP  3 years 6 months  ND  31  60  ND  <10  <10  160  160  30  IR (#1)  3 years 9 months  4  38  40  ND  160  80  320  320  160  IR (#2)  4 years 1 month  ND  54  50  10  160  320  320  320  160  DD  3 years 9 months  4  5  0  ND  <10  <10  <10  <10  <10  HK  4 years  1.9  5  40  ND  <10  <10  80  40  <10  AM  4 years 6 months  ND  40  35  0  <10  <10  80  40  20  CE  5 years  ND  25  60  ND  40  320  40  40  <10  JG  5 years 3 months  4.2  91  50  0  <10  <10  160  160  <10  SM  5 years 7 months  2.9  85  50  0  160  320  640  320  80  AG  6 years  ND  22  41  60  <10  <10  80  80  80  MV  6 years  ND  25  40  90  <10  80  80  80  <10  ACT (#1)  6 years 6 months  4  85  55  0  <10  <10  40  20  <10  ACT (#2)  7 years 2 months  ND  97  80  0  <10  <10  640  640  160  KD (#1)  7 years 6 months  3  50  45  ND  <10  <10  <10  <10  <10  KD (#2)  8 years 2 months  ND  72  60  15  <10  <10  40  160  160  RM  8 years  ND  73  60  0  40  160  40  20  <10  LK  9 years  2.5  32  40  ND  <10  20  40  40  10  SC  12 years  3.2  91  55  0  <10  <10  <10  <10  <10  MS  13 years  5  40  20  0  <10  <10  <10  <10  <10  GW  13 years  4.6  96  80  0  <10  <10  80  <10  <10  Child  Age  Percent CD8+Vβ17+/total CD8+a  CTL lysis(E:T 20:1)b  Percent lytic activity post Vβ17 depletion c  HI Serologyd    PBMCa  CTL (d21)a    A/Taiwan/1/86 (H1N1)  A/Bayern/7/95 (H1N1)  A/JHB/34/94 (H3N2)  A/Wuhan/359/95 (H3N2)  A/Sydney/5/97 (H3N2)  aThe percentage of CD8+ T cells expressing Vβ17 was determined by FACS analysis for fresh PBMC prior to in vitro stimulation and for M158–66-specific CTL after 21 days in culture.  b51Cr-release assays were performed on day 21 using CIRA2 targets. The percentage lysis of M158–66-pulsed targets is shown for an E:T ratio of 20:1. Lysis of targets pulsed with an irrelevant HLA-A*0201-binding Plasmodium peptide and targets without peptide was <10% in each case.  cDay 24 CTL lines with M158–66 specificity were depleted of Vβ17+ cells or as a control Vβ22+ cells. Vβ22 depletion did not alter cytotoxic activity in a single case. Cytotoxicity was expressed in lytic units and the residual cytotoxicity after depletion of Vβ17+ cells was expressed as a percentage of that observed following depletion of Vβ22+ cells.  dAnti-influenza antibody production was determined with the HI test. The HI test was performed with eight agglutinating doses of five appropriate prototype viruses which had circulated during the course of this study.  ND, not done.  TV  3 months  2  ND  0  ND  <10  <10  <10  <10  <10  BH  3 months  5  14  50  100  <10  <10  <10  <10  <10  DS  6 months  3  4  0  ND  <10  <10  <10  <10  <10  JR  2 years 5 months  4.50  5  30  110  <10  <10  160  120  40  DH (#1)  21 months  3.2  5  0  ND  <10  <10  <10  <10  <10  DH (#2)  2 years 9 months  3.5  70  75  20  <10  <10  <10  40  10  EB (#1)  2 years 8 months  3  52  40  10  <10  <10  160  40  80  EB (#2)  3 years 2 months  3.5  76  75  5  <10  <10  160  160  80  EJ  2 years 9 months  ND  5  10  ND  <10  <10  40  80  40  SP  3 years  ND  80  45  0  <10  <10  80  80  80  IS (#1)  3 years  2.8  4.2  0  ND  <10  <10  <10  <10  <10  IS (#2)  3 years 3 months  ND  ND  0  ND  <10  <10  <10  <10  <10  RP  3 years 6 months  ND  31  60  ND  <10  <10  160  160  30  IR (#1)  3 years 9 months  4  38  40  ND  160  80  320  320  160  IR (#2)  4 years 1 month  ND  54  50  10  160  320  320  320  160  DD  3 years 9 months  4  5  0  ND  <10  <10  <10  <10  <10  HK  4 years  1.9  5  40  ND  <10  <10  80  40  <10  AM  4 years 6 months  ND  40  35  0  <10  <10  80  40  20  CE  5 years  ND  25  60  ND  40  320  40  40  <10  JG  5 years 3 months  4.2  91  50  0  <10  <10  160  160  <10  SM  5 years 7 months  2.9  85  50  0  160  320  640  320  80  AG  6 years  ND  22  41  60  <10  <10  80  80  80  MV  6 years  ND  25  40  90  <10  80  80  80  <10  ACT (#1)  6 years 6 months  4  85  55  0  <10  <10  40  20  <10  ACT (#2)  7 years 2 months  ND  97  80  0  <10  <10  640  640  160  KD (#1)  7 years 6 months  3  50  45  ND  <10  <10  <10  <10  <10  KD (#2)  8 years 2 months  ND  72  60  15  <10  <10  40  160  160  RM  8 years  ND  73  60  0  40  160  40  20  <10  LK  9 years  2.5  32  40  ND  <10  20  40  40  10  SC  12 years  3.2  91  55  0  <10  <10  <10  <10  <10  MS  13 years  5  40  20  0  <10  <10  <10  <10  <10  GW  13 years  4.6  96  80  0  <10  <10  80  <10  <10  View Large Fig. 1. View largeDownload slide M158–66-specific CTL lines were generated from 6 HLA-A*0201 cord blood PBMC where the mother was HLA-A*0201 negative. Three representative experiments are shown (maternal data not shown). 51Cr-release assays were performed after 21 days in culture using CIRA2 targets (2000/well) pulsed with M158–66 peptide (5 μg/ml), Plasmodium cp36 peptide (5 μg/ml), no peptide or infected with 200 HAU influenza A. Triplicate wells were used and the SD of the mean is shown in each case. Fig. 1. View largeDownload slide M158–66-specific CTL lines were generated from 6 HLA-A*0201 cord blood PBMC where the mother was HLA-A*0201 negative. Three representative experiments are shown (maternal data not shown). 51Cr-release assays were performed after 21 days in culture using CIRA2 targets (2000/well) pulsed with M158–66 peptide (5 μg/ml), Plasmodium cp36 peptide (5 μg/ml), no peptide or infected with 200 HAU influenza A. Triplicate wells were used and the SD of the mean is shown in each case. Fig. 2. View largeDownload slide TCR Vβ17 usage of influenza A/M158–66-specific CTL generated from subjects of different ages. Day 21 influenza A/M158–66 peptide-specific CTL lines generated from cord PBMC, PBMC from children (3 months to 13 years) and adult PBMC were analysed by two-colour immunofluorescent staining for distribution of Vβ17 within the CD8+ population. Data are expressed as percent CD8+Vβ17+ cells within the total CD8+ subset (CD8+Vβ17+/Total CD8+). The shaded area represents the mean percent CD8+Vβ17+/total CD8+ in fresh PBMC. The percent CD8+Vβ17+/total CD8+ cells in cord lines was significantly lower than that of adult lines (P < 0.0001 by unpaired Student's t-test). Fig. 2. View largeDownload slide TCR Vβ17 usage of influenza A/M158–66-specific CTL generated from subjects of different ages. Day 21 influenza A/M158–66 peptide-specific CTL lines generated from cord PBMC, PBMC from children (3 months to 13 years) and adult PBMC were analysed by two-colour immunofluorescent staining for distribution of Vβ17 within the CD8+ population. Data are expressed as percent CD8+Vβ17+ cells within the total CD8+ subset (CD8+Vβ17+/Total CD8+). The shaded area represents the mean percent CD8+Vβ17+/total CD8+ in fresh PBMC. The percent CD8+Vβ17+/total CD8+ cells in cord lines was significantly lower than that of adult lines (P < 0.0001 by unpaired Student's t-test). Fig. 3. View largeDownload slide Dependence of maternal and cord M158–66-specific CTL lines on Vβ17+ cells. Day 24 CTL lines were depleted of either Vβ17+ or Vβ22+ cells as a control. Depletions were confirmed to be >97% complete by FACS. 51Cr-release assays were performed using CIRA2 targets with (closed symbols) or without (open symbols) M158–66 peptide. Lysis of targets pulsed with a control Plasmodium cp36 peptide was not significantly different from that of targets without peptide. The percent CD8+Vβ17+/total CD8+ for M158–66-specific CTL lines before depletion is shown for each subject. Assays were performed in triplicate and the mean percent specific lysis ± SD determined. SD was <7% in each case. Fig. 3. View largeDownload slide Dependence of maternal and cord M158–66-specific CTL lines on Vβ17+ cells. Day 24 CTL lines were depleted of either Vβ17+ or Vβ22+ cells as a control. Depletions were confirmed to be >97% complete by FACS. 51Cr-release assays were performed using CIRA2 targets with (closed symbols) or without (open symbols) M158–66 peptide. Lysis of targets pulsed with a control Plasmodium cp36 peptide was not significantly different from that of targets without peptide. The percent CD8+Vβ17+/total CD8+ for M158–66-specific CTL lines before depletion is shown for each subject. Assays were performed in triplicate and the mean percent specific lysis ± SD determined. SD was <7% in each case. Fig. 4. View largeDownload slide Dependence of the influenza A/M158–66 -specific CTL response on Vβ17+ cells at different ages. Vβ17+ and control Vβ22+ cells were depleted from day 24 M158–66-specific CTL lines using sheep-anti mouse IgG-coated magnetic beads (Dynal). More than 97% depletion was confirmed by two-colour immunofluorescent staining. The Vβ17-depleted and control Vβ22-depleted CTL lines were tested for M158–66-specificity in a 4-h 51Cr-release assay using CIRA2 targets pulsed with M158–66 peptide (5 μg/ml) or without peptide. To compare cytotoxicity of the Vβ17- and Vβ22-depleted CTL lines, results are expressed in lytic units (LU)/107 cells, 1 LU denoting the quantity of effector cells needed to kill 20% of the M158–66-pulsed target cells. The LU activity of the Vβ17-depleted lines is expressed as a percentage of the LU activity of the Vβ22-depleted lines. Certain subjects have been highlighted for further discussion. Fig. 4. View largeDownload slide Dependence of the influenza A/M158–66 -specific CTL response on Vβ17+ cells at different ages. Vβ17+ and control Vβ22+ cells were depleted from day 24 M158–66-specific CTL lines using sheep-anti mouse IgG-coated magnetic beads (Dynal). More than 97% depletion was confirmed by two-colour immunofluorescent staining. The Vβ17-depleted and control Vβ22-depleted CTL lines were tested for M158–66-specificity in a 4-h 51Cr-release assay using CIRA2 targets pulsed with M158–66 peptide (5 μg/ml) or without peptide. To compare cytotoxicity of the Vβ17- and Vβ22-depleted CTL lines, results are expressed in lytic units (LU)/107 cells, 1 LU denoting the quantity of effector cells needed to kill 20% of the M158–66-pulsed target cells. The LU activity of the Vβ17-depleted lines is expressed as a percentage of the LU activity of the Vβ22-depleted lines. Certain subjects have been highlighted for further discussion. Fig. 5. View largeDownload slide The cord blood M158–66-specific CTL response is polyclonal, whereas the adult response is oligoclonal. M158–66-specific CTL lines were generated from PBMC obtained from a maternal and cord pair. PBMC and day 21 M158–66-specific CTL lines were stained for CD8 and Vβ gene segment usage and analysed by two-colour flow cytometry. The dependence on Vβ17+ CTL was determined by depleting Vβ17+ or control Vβ22+ cells from the lines and repeating cytotoxicity assays. The cytotoxicity of the Vβ17-depleted CTL line (in lytic units) was expressed as a percentage of that of the Vβ22-depleted CTL line; 100% suggests no loss of cytotoxicity and 0% reflects complete loss of M158–66-specific cytotoxicity. Fig. 5. View largeDownload slide The cord blood M158–66-specific CTL response is polyclonal, whereas the adult response is oligoclonal. M158–66-specific CTL lines were generated from PBMC obtained from a maternal and cord pair. PBMC and day 21 M158–66-specific CTL lines were stained for CD8 and Vβ gene segment usage and analysed by two-colour flow cytometry. The dependence on Vβ17+ CTL was determined by depleting Vβ17+ or control Vβ22+ cells from the lines and repeating cytotoxicity assays. The cytotoxicity of the Vβ17-depleted CTL line (in lytic units) was expressed as a percentage of that of the Vβ22-depleted CTL line; 100% suggests no loss of cytotoxicity and 0% reflects complete loss of M158–66-specific cytotoxicity. Transmitting editor: E. Simpson This study was supported by an MRC ROPA award. T. M. L. is an ARC Clinical Lecturer in Rheumatology. S. M. is a Royal Society University Research Fellow. References 1 Ruben, F. L. 1986. Who needs influenza vaccines? In Kendal, A. P. and Patriarca, A. R., eds, Options for the Control of Influenza, p. 139. Liss, New York. Google Scholar 2 Palache, A. M. 1992. Influenza subunit vaccine- ten years experience. Eur. J. Clin. Res.  3: 117. Google Scholar 3 Wiley, D. C., Wilson, I. A. and Skehel, J. J. 1981. Structural identification of the antibody-binding sites of Hong Kong influenza haemagglutinin and their involvement in antigenic variation. Nature  289: 373. Google Scholar 4 Potter, C. W. and Oxford, J. S. 1979. Determinants of immunity to influenza infection in man. Br. Med. Bull.  35: 69. Google Scholar 5 Doherty, P. C., Allan, W., Eichelberger, M. and Carding, S. R. 1992. Roles of αβ and γδ T cell subsets in viral immunity. Annu. Rev. Immunol.  10: 123. Google Scholar 6 Topham, D. J. and Doherty, P. C. 1998. Clearance of an influenza A virus by CD4+ T cells is inefficient in the absence of B cells. J. Virol.  72: 882. Google Scholar 7 McMichael, A. J. and Askonas, B. A. 1978. Influenza virus specific cytotoxic T cells in man: induction and properties of the cytotoxic T cell. Eur. J. Immunol.  8: 705. Google Scholar 8 Huet, S., Nixon, D. F., Rothbard, J., Townsend, A. R. M., Ellis, S. A. and McMichael, A. J. 1990. Structural homologies between two HLA-B27 restricted peptides suggest residues important for interaction with HLA-B27. Int. Immunol.  2: 311. Google Scholar 9 Morrison, J., Elvin, J., Latron, F., Gotch, F., Moots, R., Strominger, J. L. and McMichael, A. 1992. Identification of the nonamer peptide from influenza A matrix protein and the role of pockets of HLA-A2 in its recognition by cytotoxic T lymphocytes. Eur. J. Immunol.  22: 903. Google Scholar 10 Sutton, J., Rowland-Jones, S., Rosenberg, W., Nixon, D., Gotch, F., Gao, X. M., Murray, N., Spoonas, A., Driscoll, P., Smith, M., Willis, A. and McMichael, A. 1993. A sequence pattern for peptides presented to cytotoxic T lymphocytes by HLA-B8 revealed by the analysis of epitopes and eluted peptides. Eur. J. Immunol.  23: 447. Google Scholar 11 DiBrino, M., Tsuchida, T., Turner, R. V., Parker, K., Coligan, J. E. and Biddison, W. E. 1993. HLA-A1 and HLA-A3 T cell epitopes derived from influenza virus proteins predicted from peptide binding motifs. J. Immunol.  151: 5930. Google Scholar 12 Moss, P. A., Moots, R. J., Rosenberg, W. M., Rowland-Jones, S. J., Bodmer, H. C., McMichael, A. J. and Bell, J. I. 1991. Extensive conservation of alpha and beta chains of the human T-cell antigen receptor recognizing HLA-A2 and influenza A matrix peptide. Proc. Natl Acad. Sci. USA  88: 8987. Google Scholar 13 Casanova, J. L. and Maryanski, J. L. 1993. Antigen-selected T-cell receptor diversity and self-non-self homology. Immunol. Today  14: 391. Google Scholar 14 Lehner, P. J., Wang, E. C., Moss, P. A., Williams, S., Platt, K., Friedman, S. M., Bell, J. I. and Borysiewicz, L. K. 1995. Human HLA-A0201-restricted cytotoxic T lymphocyte recognition of influenza A is dominated by T cells bearing the V beta 17 gene segment. J. Exp. Med.  181: 79. Google Scholar 15 Sourdive, D. J., Murali-Krishna, K., Altman, J. D., Zajac, A. J., Whitmire, J. K., Pannetier, C., Kourilsky, P., Evavold, B., Sette, A. and Ahmed, R. 1998. Conserved T cell receptor repertoire in primary and memory CD8 T cell responses to an acute viral infection. J. Exp. Med.  188: 71. Google Scholar 16 Hlavac, F., Choppin, J. and Guillet, J. G. 1997. Self-peptide ligands affect T cell recognition of the homologous influenza A matrix virus peptide M.58–66: modification of the HLA- A2. 1/peptide complex structure and T cell antagonism. Hum. Immunol.  54: 48. Google Scholar 17 Batista, F. D. and Neuberger, M. S. 1998. Affinity dependence of the B cell response to antigen: a threshold, a ceiling, and the importance of the off rate. Immunity  8: 751. Google Scholar 18 Cheynier, R., Henrichwark, S. and Wain-Hobson, S. 1998. Somatic hypermutation of the T cell receptor V beta gene in microdissected splenic white pulps from HIV-1 positive patients. Eur. J. Immunol.  28: 1604. Google Scholar 19 McHeyzer-Williams, M. G. and Davis, M. M. 1995. Antigen-specific development of primary and memory T cells in vivo. Science  268: 106. Google Scholar 20 Maryanski, J. L., Jongeneel, C. V., Bucher, P., Casanova, J. L. and Walker, P. R. 1996. Single-cell PCR analysis of TCR repertoires selected by antigen in vivo: a high magnitude CD8 response is comprised of very few clones. Immunity  4: 47. Google Scholar 21 Bachmann, M. F., Speiser, D. E. and Ohashi, P. S. 1997. Functional maturation of an antiviral cytotoxic T-cell response. J. Virol.  71: 5764. Google Scholar 22 Busch, D. and Pamer, E. 1999. T cell affinity maturation by selective expansion during infection. J. Exp. Med.  189: 701. Google Scholar 23 Zheng, B., Han, S., Zhu, Q., Goldsby, R. and Kelsoe, G. 1996. Alternative pathways for the selection of antigen-specific peripheral T cells. Nature  384: 263. Google Scholar 24 Callan, M. F., Steven, N., Krausa, P., Wilson, J. D., Moss, P. A., Gillespie, G. M., Bell, J. I., Rickinson, A. B. and McMichael, A. J. 1996. Large clonal expansions of CD8+ T cells in acute infectious mononucleosis. Nat. Med.  2: 906. Google Scholar 25 Argaet, V., Schmidt, C. W., Burrows, S. R., Silim, S. L., Kurilla, M. G., Doolan, D. L., Suhrbier, A., Moss, D. J., Kieff, E., Suclley, T. B. and Misko, I. S. 1994. Dominant selection of an invariant T cell antigen receptor in response to persistent infection by Epstein–Barr virus. J. Exp. Med.  180: 2335. Google Scholar 26 Weekes, M. P., Wills, M. R., Mynard, K., Carmichael, A. J. and Sissons, J. G. P. 1999. The memory cytotoxic T-lymphocyte (CTL) response to human cytomegalovirus infection contains individual peptide-specific CTL clones that have undergone extensive expansion in vivo. J. Virol.  73: 2099. Google Scholar 27 Kalams, S. A., Johnson, R. P., Trocha, A. K., Dynan, M. J., Ngo, H. S., D'Aquila, R. T., Kurnick, J. T. and Walker, B. D. 1994. Longitudinal analysis of T cell receptor (TCR) gene usage by human immunodeficiency virus 1 envelope-specific cytotoxic T lymphocyte clones reveals a limited TCR repertoire. J. Exp. Med.  179: 1261. Google Scholar 28 Pantaleo, G., Demarest, J. F., Soudeyns, H., Graziosi, C., Denis, F., Adelsberger, J. W., Borrow, P., Saag, M. S., Shaw, G. M., Sekaly, R. P., Soudeyns, H., Corey, L. and Fauci, A. S. 1994. Major expansion of CD8+ T cells with a predominant Vβ usage during the primary immune response to HIV. Nature  370: 463. Google Scholar 29 Storkus, W. J., Howell, D. N., Salter, R. D., Dawson, J. R. and Cresswell, P. 1987. NK susceptibility varies inversely with target cell class I HLA antigen expression. J. Immunol.  138: 1657. Google Scholar 30 Plebanski, M., Allsopp, C. E. M., Aidoo, M., Reyburn, H. and Hill, A. V. S. 1994. Induction of peptide-specific primary cytotoxic T lymphocyte responses from human peripheral blood. Eur. J. Immunol.  25: 1783. Google Scholar 31 Bakker, A. B., Marland, G., Boer, A. J., Huijbens, R. J., Danen, E. H., Adema, G. J. and Figdor, C. G. 1995. Generation of antimelanoma cytotoxic T lymphocytes from healthy donors after presentation of melanoma-associated antigen-derived epitopes by dendritic cells in vitro. Cancer Res  55: 5330. Google Scholar 32 Ridge, J. P., Fuchs, E. J. and Matzinger, P. 1996. Neonatal tolerance revisited: Turning on newborn T cells with dendritic cells. Science  271: 1723. Google Scholar 33 Cerny, A., Fowler, P., Brothers, M. A., Houghton, M., Schlicht, M. H. and Chisari, F. V. 1995. Induction in vitro of a primary human antiviral cytotoxic T cell response. Eur. J. Immunol.  25: 627. Google Scholar 34 Griffiths, P., Ronalds, C. and Heath, R. 1980. A prospective study of influenza infections during pregnancy. J. Epidemiol. Commun. Health  34: 124. Google Scholar 35 Xu, H., Yasui, O., Tsuruoka, H., Kuroda, K., Hayashi, K., Yamada, A., Ishizaki, T., Yamada, Y., Watanabe, T. and Hosaka, Y. 1998. Isolation of type B influenza virus from the blood of children. Clin. Infect. Dis.  27: 654. Google Scholar 36 Gammon, G., Klotz, J., Ando, D. and Sercarz, E. E. 1990. The T cell repertoire to a multideterminant antigen. Clonal heterogeneity of the T cell response, variation between syngeneic individuals, and in vitro selection of T cell specificities. J. Immunol.  144: 1571. Google Scholar 37 Mackensen, A., Carcelain, G., Viel, S. Raynal, M. C., Michalaki, H., Triebel, F., Bosq, J. and Kercend, T. 1994. Direct evidence to support the immunosurveillance concept in a human regressive melanoma. J. Clin. Invest.  93: 1397. Google Scholar 38 Shilyansky, J., Nishimura, M. I, Yanneli, J. R, Kawakami, Y., Jacknin, L. S., Charmley, P. and Rosenberg, S. A. 1994. T-cell receptor usage by melanoma-specific clonal and highly oligoclonal tumour-infiltrating lymphocyte lines. Proc. Natl Acad. Sci. USA  91: 2829. Google Scholar 39 Cibotti, R., Cabaniols, J. P, Pannetier, C., Delarbre, C., Vergnon, I., Kanellopoulos, J. M. and Kourilsky, P. 1994. Public and private Vβ T cell receptor repertoires against hen egg white lysozyme (HEL) in nontransgenic versus HEL transgenic mice. J. Exp. Med.  180: 861. Google Scholar 40 Lalvani, A., Brookes, R., Hambleton, S., Britton, W. J., Hill, A. V. and McMichael, A. J. 1997. Rapid effector function in CD8+ memory T cells. J. Exp. Med.  186, 6: 859. Google Scholar 41 Gilfillan, S., Bachmann, M., Trembleau, S., Adorini, L., Kalinke, U., Zinkernagel, R., Benoist, C. and Mathis, D. 1995. Efficient immune responses in mice lacking N-region diversity. Eur. J. Immunol.  25: 3115. Google Scholar 42 Gavin, M. A. and Bevan, M. J. 1995. Increased peptide promiscuity provides a rationale for the lack of N regions in the neonatal T cell repertoire. Immunity  3: 793. Google Scholar 43 Bodger, M. P., Janossy, G., Bollum, F. J., Burford, G. D. and Hoffbrand, A. V. 1983. The ontogeny of terminal deoxynucleotidyl transferase positive cells in the human fetus. Blood.  61: 1125. Google Scholar 44 McMichael, A. and O'Callaghan, C. 1998. `A new look at T cells'. J. Exp. Med.  187: 1367. Google Scholar 45 Crawford, F., Kozono, H., White, J., Marrack, P. and Kappler, J. 1998. Detection of antigen-specific T cells with multivalent class II MHC covalent peptide complexes. Immunity  8: 675. Google Scholar © 2001 Japanese Society for Immunology

Journal

International Immunology – Oxford University Press

Published: Nov 1, 2001

There are no references for this article.