| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print May 9, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Interdisciplinary Graduate Program in Immunology and
Department of Microbiology, University of Iowa, Iowa City
1 Correspondence: Department of Microbiology, University of Iowa, 3-512 Bowen Science Building, 51 Newton Rd., Iowa City, IA 52242. E-mail: john-harty{at}uiowa.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT THE CD8 T CELL...CD8 T CELL DIFFERENTIATION...MEMORY CD8 T CELL...ALTERATIONS OF CD8 T...FORMATION OF MEMORY AFTER...CONCLUSIONSREFERENCES |
|---|
Key Words: vaccination • adaptive immunity • acute infection
|
THE CD8 T CELL RESPONSE TO INFECTION |
|---|
|
TOPABSTRACTTHE CD8 T CELL...CD8 T CELL DIFFERENTIATION...MEMORY CD8 T CELL...ALTERATIONS OF CD8 T...FORMATION OF MEMORY AFTER...CONCLUSIONSREFERENCES |
|---|
and tumor necrosis factor 
(TNF-), as well as the ability to exocytose granules of cytolytic enzymes, including perforin and granzyme B (GrB), upon interaction with an infected cell [17
]. These "effector" CD8 T cells migrate throughout the peripheral tissues of the host and elaborate effector functions upon re-encounter with the antigen on the surface of infected cells. This robust expansion of CD8 T cells is followed by a contraction phase, whereby 90–95% of the effector CD8 T cells present at the peak will undergo apoptosis [18 ]. The factors that control the switch between the expansion phase and the contraction phase are unknown. Several mechanisms have been proposed, however, to mediate the onset of contraction. These include antigen-induced cell death, which is a Fas-FasL-dependent, apoptotic program, initiated as a result of prolonged contact with antigen, and survival cytokine withdrawal as a result of competition between expanding populations of CD8 T cells and other lymphocytes [19 ]. In addition, one potential mechanism supported by data from our laboratory is that the switch between the expansion and contraction phases is programmed early on during the expansion phase [13 ]. As described in more detail later, this hypothesis entails that naïve CD8 T cells that become activated for a sufficient amount of time undergo a program of a set number of cell divisions, independently of further stimulation or contact with antigen. The decision of whether an antigen-specific CD8 T cell will undergo apoptosis or survive to become a memory CD8 T cell is an area of intense research, as is the search for how and when, during the CD8 T cell response, memory cells arise.
After the contraction phase resolves, the remaining population of memory CD8 T cells is believed to be numerically stable for the life of the host [20 ]. The net result of the CD8 T cell response is a population of antigen-specific memory cells, which is increased greatly in number, as well as functionally and phenotypically distinct from their naïve precursors [2 ].
|
CD8 T CELL DIFFERENTIATION PROGRAM |
|---|
|
TOPABSTRACTTHE CD8 T CELL...CD8 T CELL DIFFERENTIATION...MEMORY CD8 T CELL...ALTERATIONS OF CD8 T...FORMATION OF MEMORY AFTER...CONCLUSIONSREFERENCES |
|---|
This programming model further implies that the onset of contraction is also independent of the presence of antigen after the first 24 h [13 ]. Work carried out by our laboratory has demonstrated that the attenuation of the course of LM infection with antibiotics did not substantially affect the timing or the magnitude, with respect to peak numbers, of the contraction phase or the establishment of T cell memory. We were additionally able to show in our system that attenuation of the length of antigen display, found to be linked to the course of infection, did not affect the kinetics of the program beyond the critical first day of T cell stimulation. In total, this body of work indicates that at least with respect to antigen, the expansion, contraction, and memory phases of the CD8 T cell response can be programmed with a relatively brief period of antigen stimulation after infection.
The majority of the studies examining and defining the CD8 T cell response made use of models of infection of mice by lymphocytic choriomeningitis virus or LM. The CD8 T cell response and generation of memory-to-protozoa and protozoan antigens are also fields of intense research. These studies have been reviewed elsewhere [24 25 26 ].
|
MEMORY CD8 T CELL PROPERTIES |
|---|
|
TOPABSTRACTTHE CD8 T CELL...CD8 T CELL DIFFERENTIATION...MEMORY CD8 T CELL...ALTERATIONS OF CD8 T...FORMATION OF MEMORY AFTER...CONCLUSIONSREFERENCES |
|---|
and the exocytosis of cytolytic molecules do not require additional costimulation or further differentiation for memory CD8 T cells in contrast with naïve cells. In addition, memory CD8 T cells require less "lag time" prior to the initiation of cell division and may be able to divide more rapidly than naïve cells [27
].
Memory CD8 T cells can be distinguished from effector and naïve CD8 T cells as a result of the differential expression of several surface markers (Fig. 1
). The identification of memory CD8 T cells or their precursors during the effector stage would allow for the examination of the development of memory and could possibly lead to methods of increasing the representation of this population. Recently, two markers have been proposed to be expressed only on those cells destined to become memory CD8 T cells during the effector phase: the IL-7R chain and the CD8 homodimer.
|
chain is expressed at high levels on naïve cells and upon activation, becomes down-regulated (see Fig. 1
). It was recently reported that a small fraction of CD8 T cells begins to express high levels of the IL-7R chain after lymphocytic choriomeningitis virus (LCMV) infection [29
]. These reports indicated that IL-7Rhi-expressing cells during the effector stage were the precursors to memory CD8 T cells, and the IL-7Rlo-expressing cells were much less effective in the formation of memory. It is interesting that the number of IL-7Rhi-expressing antigen-specific CD8 T cells at the peak correlated with the number of memory CD8 T cells present after the resolution of the contraction phase, and this was true as the contraction phase progressed from the peak of the CD8 T cell response to the memory phase. More recent studies by us and others indicate, however, that this correlation between IL-7R re-expression and the number of memory CD8 T cells is not always the case [30
, 31
]. For example, data generated by our laboratory indicate that DC immunization generated CD8 T cells in which the majority (
65%) of cells expressed high levels of IL-7R at the peak of the response; however, these DC-stimulated CD8 T cells underwent contraction to 10% of the number of CD8 T cells at the peak of the response [31
]. In addition, at the peak of the secondary response, antigen-specific CD8 T cells do not express IL-7R (A. Jabbari and J. T. Harty, unpublished observations), despite undergoing a net contraction phase of a similar magnitude, albeit with relatively delayed kinetics, as a primary CD8 T cell response. These data suggest that expression of the IL-7R is not the sole factor determining whether cells enter the initial memory population.
In addition to its expression in subsets of T cells populating the gut [32
], CD8 homodimers are transiently expressed on effector CD8 T cells, and interactions between the CD8 complex and its ligand, the thymic leukemia (TL) antigen, were recently reported in mouse models to be necessary for the formation of memory CD8 T cells [33
]. TL antigen is a nonclassical MHC class I molecule, not only expressed on murine gut epithelial cells but also on activated DCs [33
], intraepithelial lymphocytes [34
], and activated peripheral T cells [35
]. Initial reports using E8I–/– mutants, which do not express CD8 homodimers but express CD8ß heterodimers, suggested that memory formation was absent when the CD8-TL interaction was disrupted. However, others have produced results in the same systems that suggest that CD8-TL interactions are not essential for the normal formation of CD8 T cell memory [36
], and E8I–/– mice did not manifest deficiencies in CD8 expression in other infection models [37
]. In addition, memory formation was normal in an experimental system, in which only a single MHC class Ia molecule is expressed in the absence of all other ß2m-dependent MHC molecules [38
] (including the TL molecule), further suggesting that the CD8-TL antigen is unnecessary for the formation of memory. It is still undetermined whether CD8 homodimer expression is a reliable marker for the identification of memory CD8 T cell precursors at the effector stage.
One important characteristic of CD8 T cell memory is long-term persistence. Memory CD8 T cells divide at a relatively slow rate in the basal state, presumably to replace dying memory CD8 T cells and maintain their numbers in the host. This mechanism of survival and maintenance is thought to be a result of the common -chain cytokines IL-7 and IL-15 [39
40
41
42
]. The requirements for these factors are demonstrated by knockout (KO) mouse models: IL-15 KO mice are deficient in memory CD8 T cells [43
], and IL-7 KO mice are deficient, not only in memory CD8 T cells [44
] but in all mature, peripheral lymphocytes [45
], indicating a broader role for this cytokine. Furthermore, memory CD8 T cells probably do not require low levels of antigen or antagonist peptide for their survival and maintenance, as naïve cells do [46
]. It is likely that these low-affinity TCR signals are additionally not necessary for the maintenance of memory CD8 T cell function, as memory CD8 T cells maintained in a transporter for antigen processing (TAP)–/– host environment can still elaborate IFN- and robustly expand in vivo in response to antigen-laden DCs [47
].
The role of CD4 T cells in the generation and long-term maintenance of CD8 T cells has been revisited recently by several laboratories [48 , 49 ]. Although the role of CD4 cells on the formation of CD8 T cell responses has been assessed in the past [50 ], modern antigen-specific T cell detection technologies had not been available. It was found that secondary CD8 T cell responses were abrogated in the absence of CD4 T cells. Sun et al. [51 ] went on to demonstrate that CD4 T cells were not necessary in the early stages of the response but, rather, were responsible for the long-term maintenance of the antigen-specific memory CD8 T cell population. It is interesting that the specificity of the CD4 T cells was not an issue; adoptive transfer of memory CD8 T cells from a CD4-bereft environment into an unimmunized mouse with a normal CD4 compartment was sufficient for the memory CD8 population to be maintained. Recent studies by others as well as our own laboratory suggest a role for the molecule TNF-related, apoptosis-inducing ligand in the loss of memory CD8 T cells in the absence of CD4 T cells (ref. [52 ] and Vladimir P. Badovinoc and J. T. Harty, unpublished observations), but this mechanism has yet to be teased out.
In contrast to the numerical loss of memory CD8 T cells in the absence of CD4 T cells, CD8 T cells are numerically maintained but gradually lose function in the face of some chronic infections. In a mouse model of chronic LCMV infection, antigen-specific CD8 T cells go through stages in which they sequentially lose specific effector functions over time [53 ]. Recently, Barber et al. [54 ] determined that blocking programmed death 1 (PD1)-PD ligand interactions allowed for the functional maintenance of memory CD8 T cells in the presence of a chronic LCMV infection. The findings of this study may be useful in the therapeutic rescue of effective CD8 T cell functions necessary to eradicate viruses that establish chronic infections.
Several years ago, a scheme was proposed in which memory CD8 T cells were classified as central memory (TCM) or effector memory (TEM) [55 ]. This classification is based on differential expression of surface homing molecules and distinct trafficking patterns. TCM express relatively high levels of CD62L, a surface protein that mediates homing to lymphoid organs, and TEM express relatively low levels of this molecule. Another observed difference used to distinguish these cell types is the ability to produce IL-2; TEM have a limited capacity to produce IL-2, and TCM can make relatively high levels [16 ]. In general, TCM are thought to be responsible for the proliferative response seen in secondary challenges, and TEM are thought to operate as peripheral first responders to infection. These differences are supported by the increased relative ability of TEM to lyse targets in lytics assays and the increased capacity of TCM to proliferate in the presence or absence of antigenic challenge in mouse models [56 , 57 ].
More recently, however, the properties assigned to these subsets have come under scrutiny. For example, TEM have been shown to undergo robust expansion upon secondary challenge [58 ], and the same group has shown that at least at relatively early time-points after infection, TEM exhibit equal proliferative capacities as TCM [59 ]. In addition, it has previously been shown that TCM may exhibit a higher cytolytic capacity than TEM [60 ]. These data suggest that the TCM and TEM subsets may not be similar in all settings, and memory CD8 T cells are more heterogenous than this classification scheme encompasses.
|
ALTERATIONS OF CD8 T CELL KINETICS: THE ROLE OF INFLAMMATION ON THE GENERATION OF MEMORY CD8 T CELLS |
|---|
|
TOPABSTRACTTHE CD8 T CELL...CD8 T CELL DIFFERENTIATION...MEMORY CD8 T CELL...ALTERATIONS OF CD8 T...FORMATION OF MEMORY AFTER...CONCLUSIONSREFERENCES |
|---|
After an initial LM infection, naïve CD8 T cells undergo a program of differentiation resulting in the formation of antigen-specific memory CD8 T cells. Antigen-specific CD8 T cells undergo expansion and contraction prior to the formation of a stable memory population. In addition, experiments examining the timing of administration of booster infection after LM-priming revealed that 40 days, but no less than 11 days, were required to produce a larger secondary-memory CD8 T cell pool [31 ], indicating that some period of time was required after infection before the population of responding CD8 T cells acquired memory characteristics.
Recently, we had identified experimental conditions in which the CD8 T cell response is made up of a diminished expansion phase without a subsequent contraction phase [61
]. Antibiotic pretreatment of mice prior to LM infection resulted in CD8 T cell responses exhibiting diminished expansion but generated numbers of memory CD8 T cells that matched the peak of the CD8 T cell response. The generation of memory in the absence of contraction correlated with heightened IL-7R expression at Day 7 of the response in antibiotic-pretreated mice, consistent with notions that IL-7R expression marked precursors of CD8 T cell memory [29
]. The resulting memory CD8 T cells generated in antibiotic-pretreated mice were also capable of vigorously expanding and protecting mice against virulent LM infections. Continuing work from our laboratory has also demonstrated that the resulting population of antigen-specific CD8 T cells is phenotypically similar to conventionally generated, memory CD8 T cells at early time-points, corresponding to just after the end of the contraction phase (Vladimir P. Badovinic and J. T. Harty, submitted). From these experiments, we can conclude that the contraction phase is not a necessary step in the formation of protective memory CD8 T cells.
We had additionally identified that DC-priming of mice resulted in the accelerated generation of CD8 T cells with memory characteristics [31
]. The antigen-specific CD8 T cell populations generated by DC-priming were capable of undergoing secondary expansions at much earlier time-points than the analogous, antigen-specific CD8 T cell populations in LM-primed mice. Furthermore, antigen-specific CD8 T cells from DC-primed mice exhibited a rapid transition to a phenotype similar to that of memory CD8 T cells, and this population also was able to elaborate IL-2 after short, in vitro stimulations, similarly to conventional, long-term memory CD8 T cells. As mentioned previously, however, the representation of IL-7R-expressing cells at the peak of the response did not correlate with the numbers of memory CD8 T cells after DC priming, illustrating that IL-7R expression is not always a definitive marker of memory CD8 T cell precursors.
Explorations into the mechanisms underlying early memory CD8 T cell generation suggested a role for the inflammatory milieu and, more specifically, IFN-, in regulating the CD8 T cell response. Our initial clue was that IFN- levels in antibiotic-pretreated mice after LM infection were much reduced when compared with the untreated controls [61
]. Administration of CpG, a Toll-like receptor 9 agonist, which promotes the secretion of proinflammatory cytokines including IL-12, type I IFNs, and IFN-, to the antibiotic-pretreated mice at the time of LM infection, resulted in a CD8 T cell response that exhibited what appeared to be a normal contraction phase; in a sense, CpG coadministration "rescued" the contraction phase. It is interesting that the administration of CpG to antibiotic-pretreated mice deficient in IFN- signaling resulted in no rescue of the contraction phase, implicating IFN- in regulating the CD8 T cell response.
In DC-primed mice, CpG coadminstration prevented the generation of CD8 T cells capable of vigorous secondary expansion at early time-points [31
]. In addition, CpG coadministration prevented the early adoption of a memory phenotype by antigen-specific CD8 T cells generated after DC priming. This again implicates the inflammatory milieu in affecting the CD8 T cell response. Additional experiments in mice deficient in IFN- signaling demonstrated that the effects of CpG were abrogated and support that IFN- can regulate the kinetics of the CD8 T cell response as well as the rate at which CD8 T cells acquire memory characteristics.
Regulation of the CD8 T cell response by IFN- is supported by other studies. Recent experiments by others in which recipients of differentially marked IFN-R–/– CD8 T cells and wild-type (wt) CD8 T cells were infected with LCMV indicated that in the absence of IFN- signals, CD8 T cells fail to expand to the same extent as wt cells in the same mouse [62
]. This experimental model prevented the amount of antigen present in the host from being a confounding variable. Furthermore, this was not a result of different precursor frequencies, as comparisons between TCR transgenic (Tg) IFN-R–/– and wt cells, again in the same mouse, resulted in the same findings. In addition, our laboratory has previously characterized IFN- as playing a role in the contraction phase of the CD8 T cell response to bacterial and viral infections in BALB/c mice [63
]. IFN-–/– mice exhibited a protracted-contraction phase, such that the number of antigen-specific CD8 T cells remained elevated at late time-points as compared with wt animals.
The timing of when IFN- signals affect the CD8 T cell response, however, is not known. Studies conducted in our laboratory indicate that naïve, antigen-specific CD8 T cells are sensitive to IFN-, as assessed by signal transducer and activator of transcription 1 phosphorylation after in vitro coculture with IFN- [64
]. However, these cells lose their sensitivity to IFN- by Day 1 post-infection and regain responsiveness to IFN- only after the contraction phase has started, suggesting that regulation of the expansion/contraction transition by IFN- takes place at early time-points after infection.
In total, it has become clear that the CD8 T cell program can be altered dramatically, depending on the inflammatory context in which the T cell response is initiated. Many of the studies in which perturbations in the CD8 T cell response are demonstrated have the potential for direct, clinical applications. Such applications may be valuable in certain settings, such as immunotherapy or therapeutic vaccination against chronic infections, in which the rapid generation of competent memory CD8 T cells would be of benefit.
|
FORMATION OF MEMORY AFTER SERIAL INFECTIONS |
|---|
|
TOPABSTRACTTHE CD8 T CELL...CD8 T CELL DIFFERENTIATION...MEMORY CD8 T CELL...ALTERATIONS OF CD8 T...FORMATION OF MEMORY AFTER...CONCLUSIONSREFERENCES |
|---|
It has been demonstrated that a secondary infection often leads to a secondary T cell response that is much greater in magnitude than a primary T cell response [13 ]. In addition, the kinetics of the secondary CD8 T cell response are different than the primary response, and this has been documented for the expansion and contraction phases of the CD8 T cell response. For example, secondary CD8 T cell responses exhibit a protracted-contraction phase when compared with primary CD8 T cell responses [13 , 68 , 69 ]. In addition, a population of CD8 T cells undergoing a secondary response exhibit shorter lag time prior to expanding, divide at a faster rate, and may peak earlier than CD8 T cells undergoing a primary response [27 ]. Furthermore, the precursor CD8 T cell population prior to primary and secondary responses exhibits different whole-genome transcriptional profiles, as identified by microarray studies [70 ]. Given these differences in biology, it is not entirely surprising that the resulting memory CD8 T cell populations, after single versus multiple immunizations, are phenotypically and functionally distinct.
To investigate the differences between primary and secondary-memory CD8 T cells, we used a variety of different systems. We initially found that high dose-attenuated LM infection of previously LM-infected mice generated memory CD8 T cells with relatively slow conversion to TCM, as assessed by CD62L expression and the capacity to produce IL-2 after short-term antigen stimulation, compared with their naïve counterparts [71 ]. Next, we examined primary and secondary T cell responses to LM in the same mouse and found again that memory CD8 T cells undergoing a secondary response exhibited a strikingly delayed reacquisition of CD62L. These experiments indicated that differences in the host environment, such as antigen presentation or inflammation, did not account for the disparities seen in CD62L expression by antigen-specific primary and secondary-memory CD8 T cells. Our observations were also not a result of factors unique to LM, as similarly delayed CD62L reacquisition was observed for secondary CD8 T cell responses to LCMV in systems examining primary and secondary LCMV-specific responses in the same mouse. We next used a TCR Tg adoptive transfer system, which allowed for the purification and functional characterization of the populations of CD8 T cell memory. In addition, generation of primary and secondary-memory CD8 T cells with the TCR Tg adoptive transfer system allowed us to examine secondary responses with low numbers of precursor primary-memory cells in the host. This is important, as others have shown previously that the number of precursors prior to a CD8 T cell response to infection impacts the phenotype of the subsequent memory CD8 T cell population [72 ]. Although these experiments compared the responses of naïve CD8 T cells to primary, memory CD8 T cell populations made up of mixed TCM and TEM, other experiments comparing CD62Lhi-purified, primary-memory and naïve CD8 T cell populations indicated that the predominance of CD62Llo cells in the secondary-memory populations was not a result of selective activation of the precursor TEM primary-memory population.
Functional analysis of secondary-memory CD8 T cells further elucidated differences with primary CD8 T cell memory. Secondary-memory CD8 T cells were more protective against virulent LM infection than primary-memory CD8 T cells, despite equal proliferative capacities of the two populations. Consistent with these data, secondary-memory CD8 T cells exhibited increased cytolytic capacity relative to primary-memory CD8 T cells. Elevated GrB expression by secondary-memory CD8 T cells, as compared with primary-memory CD8 T cells, seems to be the likely explanation for the difference in cytolytic capacities.
The molecular mechanism underlying the difference in CD62L re-expression may require rigorous genomic analysis, as will defining the molecular switch that changes a primary-memory CD62Llo cell to a primary-memory CD62Lhi cell. Although others have presented evidence that conversion from a CD62Llo memory cell to a CD62Lhi memory cell may not require cell division [56 ], our laboratory has found that the induction of cell division may speed this transformation. Secondary-memory CD8 T cell populations express lower levels of the IL-15Rß chain than do primary-memory CD8 T cells and do not undergo cell division in response to IL-15 in vitro at concentrations high enough to induce primary-memory CD8 T cell division. However, coculture of secondary-memory CD8 T cells with high concentrations of IL-15 can stimulate cell division of this population, and secondary-memory CD8 T cells that have undergone cell division in response to high concentration IL-15 exhibited an increase in the fraction of CD62Lhi-expressing cells in relation to controls. Furthermore, gating on individual carboxyfluorescein succinimidyl ester peaks or populations of cells that had undergone increasing numbers of cell divisions indicated that populations that had undergone more divisions had higher frequencies of CD62Lhi-expressing cells. These in vitro experiments were corroborated with in vivo experiments in which cell division was induced or not by placement of secondary-memory CD8 T cells into irradiated (lymphopenic) or nonirradiated hosts, respectively. Secondary-memory CD8 T cell populations, which were transferred into lymphopenic hosts proliferated and exhibited increased frequencies of CD62Lhi cells after 2 weeks when compared with secondary-memory CD8 T cells transferred into immune competent hosts, which did not proliferate and did not alter the frequency of CD62Lhi cells. In total, these data indicate that cell division may play a role in inducing re-expression of CD62L. In support of this, secondary-memory CD8 T cells undergo less basal proliferation in vivo than primary-memory CD8 T cells, possibly explaining their reduced reacquisition of CD62L.
The relationship between proliferation and the acquisition of TCM characteristics is of interest and warrants additional experiments. Two models have been hypothesized: CD62Lhi cells accumulate faster than CD62Llo cells in settings of antigen-independent proliferation, leading to an increase in CD62Lhi relative to CD62Llo cells, or some fraction of dividing daughter cells from the CD62Llo population up-regulates CD62L. These models are in addition to the possibility that CD62Llo cells spontaneously become CD62Lhi cells in the absence of cell division. As for the first hypothesis, knowledge of the rates of cell division and the death of the CD62Lhi and CD62Llo populations in the basal state may provide indirect support for this possible explanation if the difference between the rate of cell division and half-life of CD62Lhi cells is higher than that for CD62Llo cells. Further, this model would indicate that the population of a completely CD62Lhi-skewed population of memory cells would need to alter the balance of cell division and death rates to prevent their continued accumulation, potentially at the expense of CD8 T cells of other specificities or other lymphocyte populations. The second model, in which some fraction of daughter cells of CD62Llo memory cells is CD62Lhi, does not require regulation that is as complex. Indeed, our analysis of a purified CD62Llo population may indicate that cell division results in the appearance of more CD62Lhi cells than in the absence of cell division.
What other factors then may account for the differential rate of basal proliferation and CD62L reacquisition after primary and secondary responses? As discussed previously for the generation of primary-memory CD8 T cells, inflammation at the time of initiation of a T cell response may play a role in regulating the phenotype and function of secondary-memory CD8 T cells. It is possible that the inflammatory milieu differs in infected mice depending on the presence of naïve versus primary-memory CD8 T cells, and this difference contributes to the difference in phenotype and function associated with the resulting primary or secondary-memory CD8 T cell population. Experiments examining primary and secondary responses and the resulting primary and secondary-memory CD8 T cell populations in the same mouse indicate that at least with respect to the global host inflammatory milieu, the differences seen in primary and secondary-memory CD8 T cells are not a function of differences in host factors. However, differences in the microenvironments, in which the respective responses are initiated, may still exist. In the case of a primary response, the initiation of a CD8 T cell response requires the interaction of a naïve CD8 T cell with a mature DC. Immunogenic, mature DCs express a litany of surface-costimulatory molecules, including B7.1 and B7.2. In addition, these DCs are producers of cytokines, such as IL-12, which can affect the CD8 T cell response. Indeed, the Mescher laboratory [6 , 7 ] has shown that IL-12 may function as a necessary third signal in the initiation of a naïve CD8 T cell response in an elegant, minimalist, experimental setting. Conversely, secondary CD8 T cell responses or responses by primary-memory CD8 T cells may be stimulated by antigen presented by non-DCs, although this has been called into question by some recent publications [73 , 74 ]. It follows then that the difference between the ensuing memory after primary or secondary CD8 T cell responses may lie in the interaction between the precursor CD8 T cell and the stimulating antigen-presenting cell (APC) type or rather, the inflammatory microenvironment surrounding the precursor T cell at the time of stimulation. For example, lack of DC-derived IL-12 in the APC-primary-memory CD8 T cell interaction, but its presence in the DC-naïve CD8 T cell interaction, may be responsible for observed differences in the resulting primary and secondary-memory CD8 T cell populations, respectively. It is possible that primary and secondary-memory CD8 T cells generated by stimulation of the respective precursors with the same DC population may be phenotypically and functionally equivalent.
Our data indicating that secondary-memory CD8 T cells mostly exhibit a CD62Llo phenotype and a diminished capacity to produce IL-2 in relation to primary-memory CD8 T cells initially indicated to us that these cells were similar to primary TEM. This hypothesis was further supported by sustained expression of GrB by secondary-memory CD8 T cells, a characteristic consistent with memory CD8 T cells generated early after the CD8 T cell response to infection. However, direct comparisons of CD62Llo primary-memory CD8 T cells and secondary-memory CD8 T cells indicated that these two populations are phenotypically and functionally distinct. Secondary-memory CD8 T cells actually express higher levels of GrB in the absence of further stimulation than do CD62Llo primary-memory CD8 T cells. Functional comparisons between these two populations further indicated that secondary-memory CD8 T cells exhibit increased cytolytic activity and are more protective in vivo against LM infection than primary-memory TEM. These data suggest that not all CD62Llo memory CD8 T cells are functionally similar, and the attribution of specific characteristics to CD62Llo and CD62Lhi memory CD8 T cells, at least in the case of secondary-memory CD8 T cells, may be an oversimplification.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTTHE CD8 T CELL...CD8 T CELL DIFFERENTIATION...MEMORY CD8 T CELL...ALTERATIONS OF CD8 T...FORMATION OF MEMORY AFTER...CONCLUSIONSREFERENCES |
|---|
Received February 27, 2006; revised March 26, 2006; accepted March 27, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTTHE CD8 T CELL...CD8 T CELL DIFFERENTIATION...MEMORY CD8 T CELL...ALTERATIONS OF CD8 T...FORMATION OF MEMORY AFTER...CONCLUSIONSREFERENCES |
|---|
detects different frequencies of antigen-specific CD8(+) T cells J. Immunol. Methods 238,107-117[CrossRef][Medline]-mediated survival and differentiation of CD8 memory T cell precursors Science 304,590-593 J. Immunol. 175,5619-5623 homodimer expression and role in CD8 T cell memory generation during influenza virus A infection in mice Eur. J. Immunol. 35,3103-3110[CrossRef][Medline] acts directly on CD8+ T cells to increase their abundance during virus infection J. Exp. Med. 201,1053-1059 Science 290,1354-1358 signaling in antigen-specific CD8+ T cells responding to infection J. Immunol. 174,6791-6802This article has been cited by other articles:
|
|
 |
|
  A. W. M. van der Velden, J. T. Dougherty, and M. N. Starnbach Down-Modulation of TCR Expression by Salmonella enterica serovar Typhimurium J. Immunol., April 15, 2008; 180(8): 5569 - 5574. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Shaw and M. N. Starnbach Antigen Delivered by Anthrax Lethal Toxin Induces the Development of Memory CD8+ T Cells That Can Be Rapidly Boosted and Display Effector Functions Infect. Immun., March 1, 2008; 76(3): 1214 - 1222. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||
