| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
T Cells in Humans
* Vaccine Research Center, National Institutes of Health, Bethesda, MD 20892; and Departments of
Pediatrics and
Genetics, Stanford University, Stanford, CA 94305
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-chain combined with a 
-chain or a 
-chain combined with a 
-chain. T cells constitute the majority of T cells in human blood throughout life. Flow cytometric analyses presented in this study, which focus on the representation of the developmental (naive and memory) subsets of T cells, show by function and phenotype that this lineage contains both naive and memory cells. In addition, we show that the representation of naive T cells is higher among than T cells in adults and that the low frequency of naive T cells in adults reflects ontological differences between the two major subsets, which are distinguished by expression of V1 vs V2 -chains. V1 cells, which mirror cells with respect to naive representation, predominate during fetal and early life, but represent the minority of cells in healthy adults. In contrast, V2 cells, which constitute the majority of adult cells, show lower frequencies of naive cells than V1 early in life and show vanishingly small naive frequencies in adults. In essence, nearly all naive V2 cells disappear from blood by 1 year of life. Importantly, even in children less than 1 year old, most of the nonnaive V2 cells stain for perforin and produce IFN- after short-term in vitro stimulation. This represents the earliest immunological maturation of any lymphocyte compartment in humans and most likely indicates the importance of these cells in controlling pathology due to common environmental challenges. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-chain combined with a -chain or a -chain combined with a -chain. T cells expressing receptors represent the majority of T cells in human blood. The mechanisms of Ag processing and presentation for the CD4 and CD8 T cell lineages are now well established. In contrast, T cells expressing receptors represent a minority of blood T cells, and many questions remain unanswered with respect to the functions of these cells and the mechanisms through which they recognize Ags (reviewed in Refs. 1 and 2).
Compared with TCRs, there are many fewer and genes available for recombination. Therefore, much of the repertoire is achieved through TCR junctional diversity. In human T cells, only two genes, V1 and V2, are commonly used. mAbs specifically recognize TCRs using either of these genes. Thus, V1- or V2-expressing cells can be identified and studied using flow cytometry.
The representation of V1 and V2 T cells varies with location and with disease status. Although V2 cells typically constitute the majority among T cells in blood, V1 cells predominate at epithelial sites such as in the intestine and skin (3, 4). In certain diseases, the representation of V1 or V2 cells shifts dramatically. For example, in HIV disease V2 cells are lost and V1 cells expand (3, 5, 6). The significance of these changes is not known.
Functional differences have also been demonstrated between V1 and V2 cells. V2 cells respond in vitro to small m.w. nonpeptidic phosphoantigens, originally described for Ags derived from mycobacteria (1, 7, 8, 9). Less is known about the Ag specificities of V1 cells. However, V1 cells in the intestinal epithelium can respond to stress-induced MHC class I-related molecules independent of Ag processing (10, 11).
Although some phenotypic analyses of T cells beyond the subdivision into V1 and V2 subsets have been reported (12, 13, 14), few provide detailed phenotypic, functional, and developmental characterization of T cells assessing multiple markers simultaneously (15). In this study, we apply immunophenotypic classification methods used to distinguish naive and memory T cells (16, 17) to characterize the cells within the T cell compartment early in life and in healthy adults. With these methods, we show that independent of T cell lineage (CD4, CD8), the T cell compartment contains naive and memory subsets that are phenotypically and functionally comparable to those in the compartment. In addition, we show a remarkably consistent assignment of functional attributes to phenotypically defined subsets among CD4, CD8, and lineage T cells, demonstrating that functional and phenotypic maturation is highly parallel across lineage.
We also find distinct differences between the V1 and V2 subpopulations of T cells. Although both V1 and V2 cells display signs of early activation in neonates, V2 cells in particular are activated very early, with the majority of V2 cells showing evidence of prior activation in individuals before 1 year. This represents the earliest immunological maturation of any lymphocyte compartment in humans and most likely reflects the importance of these cells in controlling pathology due to common environmental challenges.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Subjects were recruited from the clinics and inpatient units of the Lucille Packard Children’s Hospital at Stanford University. Blood was drawn at Stanford University and shipped to the National Institutes of Health for analysis. Cord blood was obtained from full-term newborn infants without outside factors for infection or immunosuppression. Other blood samples were drawn from infants and children undergoing elective surgery. All patients were essentially healthy and were presenting for elective surgical procedures. All samples were drawn in the operating room or in the preanesthesia unit. No patients had undercurrent febrile illness or immune deficiency syndrome. Adult blood samples were drawn from healthy volunteers who worked in the hospital clinics or in the laboratory. A complete blood count was obtained for each sample by the clinical laboratory at Stanford University. The study was approved by the Institutional Review Boards both at Stanford and at the National Institutes of Health/National Institute of Allergy and Infectious Diseases.
In vitro stimulation
PBMC were isolated by Ficoll density centrifugation. Cells to be analyzed without stimulation were either prepared immediately for flow cytometric analysis (see below) or placed in culture overnight without stimulation for concurrent analysis with the stimulated cells. For IFN- and IL-2 analysis, cells at 1–2 x 106/ml were stimulated for 6 h with PMA (10 ng/ml; Sigma-Aldrich, St. Louis, MO) and ionomycin (2 µM; Sigma-Aldrich) in 24- or 48-well plates. For phosphoantigen stimulation, cells were stimulated overnight with isopentenyl pyrophosphate (IPP,2 30 µg/ml; Sigma-Aldrich) in 96-well V-bottom plates at a density of 1 million cells per 200 µl well. Monensin (2 µM; BD PharMingen, San Jose, CA) was also included in all stimulations. Depending on the number of cells available, between 3 and 6 million cells were cultured for each stimulation condition in separate wells. After stimulation, the cells from multiple wells for each condition were combined.
Flow cytometric analysis
For analysis of perforin expression, cells were surface stained with all the Ab reagents (18) except for those directed to cytokines, perforin, and CD3. In addition, during this staining, ethidium monoazide bromide (EMA, 1 µg/ml; Molecular Probes, Eugene, OR) was included to label dead cells, and CD163 and CD19, detected in the same channel as EMA, were included to label and exclude monocytes and B cells. An Ab to glycophorin A, CD235, was often also included in this dump channel to exclude red cells. Cells were stained for 10 min at room temperature in the dark, and then placed under an intense fluorescent light source for 10 min to photoactivate the EMA (leading to covalent cross-linking with DNA). After washing, fixation, and permeabilization (Cytofix/Cytoperm kit; BD PharMingen), cells were stained intracellularly for cytokines or perforin, and CD3 (to detect both surface and internalized CD3). The multicolor FACS methods have been described elsewhere (19, 20, 21, 22).
All mAb reagents, either purified or preconjugated, except for Texas Red-PE (TRPE) CD3 and CD45RO, were obtained from BD PharMingen. TRPE CD3 and CD45RO were obtained from Immunotech/Beckman Coulter (Miami, FL). Abs conjugated in our laboratory were done using standard protocols (http:/drmr.com/abcon), and validated by comparison with commercial conjugates after appropriate titration. The fluorescent dyes used in each staining combination include the following: FITC, PE, TRPE, Cy5PE, Cy5.5PE, Cy7PE, Alexa 594, allophycocyanin, Cy5.5 allophycocyanin, Cy7 allophycocyanin, Cascade Blue, and Alexa 430. PE and allophycocyanin were obtained from ProZyme (San Leandro, CA). Cy5, Cy5.5, and Cy7 were obtained from Amersham Life Science (Pittsburgh, PA). The Alexa dyes and Cascade Blue were obtained from Molecular Probes.
A typical 12-color staining combination used for the cytokine analysis was: Cascade Blue CD45RO, Alexa 430 CD57, FITC V1, PE V9, TRPE CD3, EMA/CD163/CD19 (detected in Cy5PE channel), Cy5.5PE CD4, Cy7PE V2, Alexa 594 IFN-, allophycocyanin IL-2, Cy5.5 allophycocyanin CD11a, and Cy7 allophycocyanin CD8. An alternate combination included CD27 and did not include V9. For this panel, the following substitutions were made: Cascade Blue CD27, PE IL-2, Cy7PE CD11a, allophycocyanin V2, Cy5.5 allophycocyanin CD8, and Cy7 allophycocyanin CD45RO. The combination for perforin analysis was similar to this latter panel, substituting: PE V9 and Alexa 594 perforin. Because the perforin analysis used unstimulated PBMC processed by Ficoll, and therefore was not likely to include many dead cells, a panel not including a dump channel was sometimes used: Cascade Blue CD3, Alexa 430 CD57, FITC V1, PE V9, TRPE CD62L, Cy5PE CD27, Cy5.5PE CD4, Cy7PE CD11a, Alexa 594 perforin, allophycocyanin V2, Cy5.5 allophycocyanin CD8, and Cy7 allophycocyanin CD45RO. An alternate panel, not including perforin, was also used for unstimulated cells: Cascade Blue CD62L, Alexa 430 CD57, FITC V1, PE V9, TRPE CD45RO, Cy5PE dump (as above), Cy5.5PE CD4, Cy7PE V2, Alexa 594 CD45RA, allophycocyanin CD3, Cy5.5 allophycocyanin CD11a, and Cy7 allophycocyanin CD8. For Fig. 1A, the following panel was used: Cascade Blue CD57, FITC V1, PE V2, TRPE CD45RO, Cy5PE CD62L, Cy5.5PE CD45RA, Cy7PE CCR7, allophycocyanin CD27, Cy5.5 allophycocyanin CD11a, Cy7 allophycocyanin CD8, and biotin CD3. Biotin was revealed with streptavidin QuantumDot 605, obtained from QuantumDot (Hayward, CA).
|
Statistical analysis
All FACS plots are displayed with five-decade logarithmic axes (except forward scatter, linear) with a data transformation that allows for improved visibility of events that fall at the lower end of the log scale (D. Parks, W. Moore, and M. Roederer, manuscript in preparation). Except for Figs. 3 and 4, FACS plots are displayed as 5% contour graphs with outlying events beyond the last contour shown as dots. In Figs. 3 and 4, the FACS plots are shown as pseudocolor density plots. The subsets in these figures are shown at lower resolution so that the lower frequency events appear as larger dots. For the overlay plots in these figures, one subset is shown either as white contours or white dot plots, depending on the number of events in the subset.
|
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T cell subsets can be identified by surface phenotype
For T cells, there are several surface markers that are capable of differentiating naive cells from cells that are activated or that have previously encountered Ag (i.e., effector/memory cells). Two or more of these markers must be used in combination to achieve the most accurate discrimination of these subsets. Often CD45RA or CD45RO is used in combination with additional markers such as CD62L, CD11a, CD27, and CCR7.
In a small number of adult subjects, we examined the expression of six naive-defining markers simultaneously on and T cells (Fig. 1A). Because either expression of CD45RA, or lack of expression of CD45RO provides similar information regarding naive cell identification (data not shown), we typically use only one of these markers. Therefore, in Fig. 1, we gated on CD45RO- cells and then examined two other markers, CD11a and CD27.
Naive cells express the brightest levels of CD27 and express CD11a at dull levels. Cells expressing low levels of CD27 are CD11abright, and are therefore not naive. Because distinguishing bright from dull levels of CD27 is difficult without an additional marker, we prefer to use CD11a over CD27 when limited to two naive-defining markers. After gating on three markers (CD45RO-, CD11adull, CD27bright), use of an additional marker such as CD62L is generally not necessary, because nearly all of these cells are CD62L+ (Fig. 1A, right panels). CD62L provides clear distinction of naive and memory subsets, but its use is limited by the fact that this marker is usually lost when cells are stimulated or when cells are cryopreserved. In addition, we found that CD62L was lost from a large number of T cells after whole blood was shipped overnight to our laboratory at the National Institutes of Health. We suspect this is likely a consequence of temperature variations during transit. CCR7 is useful as a naive-defining marker for T cells, but not for T cells (Fig. 1A, right panels).
Therefore, we chose to use CD45RO in combination with CD11a as the primary naive-defining markers. We included CD27 as a third naive-defining marker in over one-third of the samples we analyzed. As previously demonstrated for CD4 and CD8 T cells, naive cells do not express CD45RO, express low levels of CD11a, and express high levels of CD27 (16, 23, 24). All other cells are defined as nonnaive.
Fig. 1B shows examples of the expression of CD45RO, CD11a, and CD27 on T cell subsets in two adult blood samples, a cord blood sample, and two pediatric blood samples. Cells are first gated as CD45RO-, and then the CD45RO- cells are further gated as CD11adullCD27bright to identify naive cells. For the CD8 T cells, the naive cells are clearly identified as a separate population.
We used these same markers to determine whether comparable populations can be identified among T cells. Because two major subsets of T cells, V1 and V2, account for the vast majority of T cells in blood and can be identified by surface staining with mAbs, the naive-defining markers have been examined separately on the V1 and V2 subsets. Prior studies have noted a difference in the expression of CD45RO on V1 and V2 cells (12, 13, 14): in the adult, most V2 cells express CD45RO and most V1 cells do not.
As demonstrated by the expression of CD45RO on the subsets in the two adult donors in Fig. 1, we also find that in adults, most V1 cells do not express CD45RO, and most V2 cells express high or intermediate levels of CD45RO. In addition, we find that the lack of expression of CD45RO is not sufficient to phenotypically identify cells as naive. Although there are V1 and V2 cells that are CD45RO-, many are CD11abright (and CD27- or CD27dull), and therefore do not fit the naive phenotype. Therefore, as for T cells, two or more markers must be used or used to phenotypically identify cells as naive.
For both and subsets in adult samples, two markers, CD45RO and CD11a, work very well to identify naive cells because there is a clear demarcation between CD11adull and CD11abright cells. CD11a works less well for cord blood and blood samples from infants. In part, this is due to the fact that there are few CD11abright cells to contrast with the CD11adull cells. In addition, naive cells stain slightly brighter for CD11a than do naive cells in adult and older pediatric samples. However, even with these differences, staining for CD11a is brighter on activated/memory cells. This is demonstrated for the V2 cells from the 14-mo-old child in Fig. 1, in which the lower level of staining for CD27 confirms the transition from the naive phenotype. For 
40% of our samples, we include CD27 as a third naive-defining marker (7 of 14 cord blood, 9 of 37 pediatric, 16 of 27 adult). For these samples, the naive frequency determined by all three markers correlates well with the naive frequency determined by CD45RO and CD11a alone (R2 = 0.98, 0.94, 0.79, and 0.91 for the CD4, CD8, V1, and V2 subsets, respectively). Therefore, the naive frequency is accurately predicted by using only CD45RO and CD11a for the remainder of our samples that were not stained with CD27.
V2 naive cells are lost early in life
Although lack of CD45RO is by itself not sufficient to identify cells as naive, CD45RO expression is sufficient to identify cells as nonnaive. Several previous studies demonstrated that the vast majority of V2 cells in the adult are CD45RO+ and, thus, nonnaive (12, 13, 14). One study has shown that V2 cells in cord blood do not express CD45RO and that CD45RO expression on V2 cells increases with age in children (12). Because there is such a dramatic difference in CD45RO expression between V1 and V2 cells in adults, we examined in greater detail the changes in the naive and memory subpopulations of T cells in early life.
Examples of the staining of a cord blood sample and two pediatric samples are shown in Fig. 1. The majority of CD8, V1, and V2 cells are naive in cord blood and in the infant sample. However, in the sample from a child 14 mo old, most of the V2 cells are already nonnaive. This is in sharp contrast to the other T cell subsets, in which a majority of cells are still naive.
The right panels in Fig. 2 show the kinetics of the disappearance of naive T cells from blood. For both V1 and V2 cells, the major loss of naive cells as a percentage of the subset occurs during the first year of life. This loss is most dramatic for V2 cells in which a very small percentage of the V2 cells remains naive after 1 year. In contrast, a relatively high proportion of naive V1 cells persists through childhood. The proportion of naive CD4 and CD8 T cells also decreases early in life; however, this does not occur as early as for T cells and not to the extent observed for V2 cells. Even before birth, as demonstrated in cord blood, in contrast to T cells, many cells are already nonnaive (Fig. 2, right panels).
|
1 cells, but no change in V2 cells (Fig. 2). This results in an inversion of the V1-V2 ratio, from a median of 3.0 for cord blood to 0.4 for the pediatric group. Naive V1 counts drop between cord blood and the pediatric group, and again between the pediatric and adult groups. For V2 cells, the naive cell drop between the cord blood and pediatric groups is not significant, most likely due to the large number of cells already nonnaive in cord blood (a median of 50%).
Naive and memory T cells have functional similarities to their counterparts
Naive and memory T cells have distinct cytokine profiles in response to short-term polyclonal stimulation: naive cells produce IL-2 and few other cytokines, while memory/effector cells produce multiple cytokines, including the prototypical effector cytokine, IFN-. To determine whether the phenotypic definition of naive and memory cells correlates with function, production of IFN- and IL-2 was examined following in vitro stimulation with PMA and ionomycin. As demonstrated in Fig. 3 for one adult, cells producing only IL-2 are enriched for naive cells, and cells producing only IFN- are enriched for nonnaive cells. This is true for CD8 T cells as well as for the T cells. The tendency for IL-2+ IFN-- cells to be enriched for naive cells is significant when all the adult and pediatric samples examined for cytokine production are included in the analysis (Fig. 3, right panels).
Another difference observed between naive and memory T cells is the expression of perforin. Perforin is expressed in cytotoxic effector cells, but not in naive cells. Fig. 4 shows the expression of perforin as determined by intracellular staining in an adult and an infant. For both the CD8 cells and the cells, the subset of cells not expressing perforin is enriched for naive cells. This figure demonstrates that there are two levels of expression of perforin, high and low, and the data presented for the two subjects shown in this figure suggest that when perforin is expressed by T cells, the V1 cells express high levels and the V2 cells express low levels. When all the perforin data for adult samples are analyzed, there is a significant difference between the percentage of perforin-expressing V1 and V2 cells that express high perforin levels (Fig. 5). This pattern is evident for most subjects, but this figure also shows that several subjects do not fit this profile, e.g., there are three subjects for whom less than 50% of the perforin-expressing V1 cells express high levels of perforin.
|
2 cells show evidence of activation early in life
Consistent with the early loss of most V2 cells of the naive phenotype, by 1 year of age, a large proportion of V2 cells stains for perforin and produces IFN- in response to polyclonal stimulation. Staining examples in two adults are shown in Figs. 3 and 4. In four pediatric samples analyzed, unlike the CD8+ and V1+ cells, the majority of the V2+ cells stain for IFN- (Fig. 6). All four samples are from subjects less than 1.5 years of age, indicating that the V2 cells become responsive to PMA/ionomycin stimulation (in terms of IFN- production) very early in life. This responsiveness persists into adulthood. Interestingly, even in cord blood, a substantial proportion of V2 cells produces IFN- (Fig. 6).
|
production), two of the four show a high proportion of V2 cells staining for perforin (Fig. 6). These two pediatric samples are from subjects 2 and 4 mo old, indicating that it is in this early time that many V2 cells acquire an activated phenotype.
V2 is preferentially paired with V9 in the adult, and these cells respond to phosphoantigen stimulation
The Ag specificities of T cells are not well understood; however, T cells expressing the V2V9 TCR are known to respond to in vitro stimulation with small m.w. phosphoantigens. After overnight stimulation with one such Ag, IPP, responding V2 cells can be detected by intracellular staining with IFN-. Although a variable percentage of V2 cells in adults and children responds to this stimulation, few cord blood cells respond (Fig. 7).
|
2 cells with this specificity are not present in cord blood or that they have not been previously activated and therefore cannot respond within the short time frame of the stimulation. To determine the relative frequency of cells expressing the V9V2 TCR, the receptor responsive to IPP, the percentage of V2 cells coexpressing V9 was determined by costaining with a mAb to V9. As previously shown, the majority of V2 cells in the adult coexpress V9 (Fig. 8). However, the pairing of V2 with V9 is not the dominant TCR in cord blood and young pediatric samples. Among the pediatric samples analyzed in Fig. 8, the proportion of V2 cells coexpressing V9 increases with age. There is no age-dependent preferential pairing of V1 with V9.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T cells in early life. Both the two major types of human T cells, those using either the V1 or V2 TCRs, show signs of prior activation early in life, even before birth. Remarkably, by 1 year of age, almost all V2 T cells in blood have been previously activated and are thus memory T cells. Little is known about the Ag specificities of T cells and about the physiologic roles these cells play. The results of this study provide additional clues as to the nature of these cells, and when interpreted in the context of what has been previously published, help to define the ontogeny of the V1 and V2 T cell subsets in the human.
One difference observed between and T cells is that both V1 and V2 cells include a relatively high proportion of nonnaive cells before birth, as measured in cord blood. Because this activation occurs before environmental exposure, these cells may be responding to self Ags. Consistent with this hypothesis, T cell responses to keratinocytes (25, 26) and stress-induced self Ags have been observed (10, 11), and it has been proposed that V2 cells that recognize phosphoantigens present on foreign pathogens may also recognize homologous self Ags (1).
Shortly after birth, within the first year of life, both the V1 and V2 cells show a rapid decline in the proportion of naive cells. The exposure to environmental Ags that occurs after birth is a likely cause for this early activation. The prevalence of T cells at epithelial surfaces, where they represent the first line of defense, may explain the greater extent of T cell activation compared with T cells. There is also a difference in the kinetics of naive T cell loss between the V1 and V2 cells. Although the V1 naive cells drop to below 40%, the V2 cells drop to well below 10%.
To account for this difference between V1 and V2 cells, we propose that during the ontogeny of these subsets, there are developmental windows when either V1 or V2 cells are generated. This would be similar to the so-called waves of T cell development observed in the mouse and in the chicken (27, 28, 29). These studies have shown that T cells bearing certain TCRs are only generated during fixed periods and that they then home to certain peripheral sites, where they are maintained by self-renewal. Our data suggest that this may be the case for the majority of V2 cells. If they only develop during a fixed period before or near birth, as they become activated, their numbers are not regenerated by the new generation of naive cells, but instead by proliferation of existing cells.
In contrast, because relatively large numbers of naive V1 cells are found in blood into adulthood, these cells may be regenerated continuously by de novo development in the thymus. Consistent with this hypothesis, V1 transcripts are prevalent in the postnatal thymus, while V2 transcripts are prevalent in the fetal thymus (30, 31, 32). Furthermore, the number of V1 naive cells in blood decreases at a rate similar to that for putatively thymus-derived naive CD4 and CD8 T cells, i.e., they persist through childhood, with gradual loss through adulthood thought to be a consequence of thymic atrophy. Note that it is possible that a small subpopulation of V2 cells is also continually produced by the thymus, because small numbers of naive V2 cells are present in adults.
Based solely on phenotypic analyses, we previously hypothesized that V1 T cells in adult humans are extrathymically derived (22). Although the present more-detailed analysis suggests that they are in fact likely to be thymically derived during early development through adulthood, it is possible that extrathymic differentiation of V1 T cells becomes more prevalent later in life, as thymic output dwindles. Nonetheless, the conclusions of our earlier report are open to new interpretation based on this new data.
V2 cells coexpressing V9 are known to respond to small m.w. phosphoantigens present on the cell surface of some pathogenic bacteria, as well as in human cells (8, 9). In the adult, the large majority of V2 cells coexpress V9. Because the proportion of V2 cells coexpressing V9 is low in cord blood and increases with age, it is likely that there is significant positive selection for these cells. There are many different phosphoantigens that are reactive with V2V9 cells, so it may be exposure to a variety of pathogens that results in the selection of these cells. In the macaque, this subset expands in response to mycobacterial infection (33). A study of the life span of T cells in adult thymectomized mice found that the cells had a rapid turnover and displayed an activated/memory phenotype, suggesting a chronic response to environmental Ags (34). However, because this cell type is also potentially autoreactive, it may be exposure to self Ags that arise after birth that leads to their increase.
All of the studies reported in this work study T cells in peripheral blood. This is a limitation especially when studying T cells. Although T cells are present in low proportions in blood, they are present in much higher proportions in other anatomical sites such as epithelial surfaces in the intestine and skin. In fact, it is likely that cells are functionally active at these sites. The changes in the frequencies of these cells that we observe in blood may reflect changes at these peripheral sites and/or altered trafficking between the blood and these sites. Clearly, extrapolation of the results from blood cell studies to other immunological sites should be made cautiously. In general, however, the proportion of naive cells in peripheral sites is much lower than that in blood (because of the selective homing of memory/activated cells to those sites). Thus, the dynamics we report in this work, ontologically early loss of naive T cells, must only be more dramatic for these tissues.
Age-related changes in the repertoire with an expansion of V2 cells and an increase in the expression of CD45RO on V2 cells have been previously reported (12). In this prior report, consistent with our findings, cord blood cells were largely CD45RO-. The authors also noted an age-related increase in CD45RO expression on V2 cells not observed on V1 cells. They reported that early in life, close to 50% of the V2 cells expressed CD45RO, and that this percentage gradually increased in adults. Because they only measured CD45RO expression, they were unable to identify the nonnaive CD45RO- cells that we demonstrate appear very early in life. In addition, they could not identify the loss of the large majority of naive V2 cells within the first year of life. The authors also report that most thymic cells stain for V1, which is consistent with our hypothesis of continual thymic replenishment of V1 cells.
A recent paper reports phenotypic and functional data for naive and memory subsets of V2 cells in blood, lymph nodes, and inflammatory sites (15). The results are consistent with ours, with the exception of reporting a higher proportion of naive V2 cells in adults. This difference is most likely due to the strategy used to identify naive cells. Dieli et al. (15) use only the expression patterns of CD45RA and CD27. As illustrated in Fig. 1A, we found that CD27bright cells are difficult to distinguish from CD27low cells without the use of an additional marker such as CD11a. Therefore, the use of CD27 with CD45RA to identify naive cells can result in nonnaive CD27low cells being classified as naive, leading to a significant overestimation of the frequency of V2 T cells that are naive.
In summary, we describe the developmentally related changes in T cell subsets that occur early in life. These changes are in many ways similar to those observed for other species, suggesting that the human immune system is also capable of waves of T cell development. A significant difference between the and T cell lineages is the much earlier activation and conversion to memory of the T cells, and illustrates the central role that T cells have in addressing Ag challenge from birth onward. This is particularly true for V2 T cells, which show the earliest evidence of developmental maturation of any adaptive immune system.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Abbreviations used in this paper: IPP, isopentenyl pyrophosphate; EMA, ethidium monoazide bromide; TRPE, Texas Red-PE.
Received for publication June 3, 2003. Accepted for publication November 20, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cells: a right time and a right place for a conserved third way of protection. Annu. Rev. Immunol. 18:975.[Medline]
/ T cells. Annu. Rev. Immunol. 14:511.[Medline]
/+ lymphocyte subsets during HIV infection. Clin. Exp. Immunol. 75:206.[Medline]
/ T cell receptor, the CD8 accessory molecule and preferentially uses the V1 gene segment. Eur. J. Immunol. 21:1053.[Medline]
lymphocytes is increased during HIV-1 infection. Clin. Exp. Immunol. 83:187.[Medline]
T lymphocytes in the peripheral blood of patients with human immunodeficiency virus type 1 infection. [Published erratum appears in 1992 J. Infect. Dis. 166:950.]. J. Infect. Dis. 165:917.[Medline]
T cells. Proc. Natl. Acad. Sci. USA 91:8175. T cells by nonpeptidic mycobacterial ligands. Science 264:267. T cells. Nature 375:155.[Medline]
T cells. Science 279:1737./ T cells: implications for innate immunity. J. Exp. Med. 191:937./ repertoire. J. Exp. Med. 171:1597./+ lymphocytes. J. Exp. Med. 171:1833. T lymphocytes and natural killer cells. Int. Immunol. 3:691.2 T cells and migratory routes in lymph nodes or inflammatory sites. J. Exp. Med. 198:391.1 and V2 T cells express distinct surface markers and might be developmentally distinct lineages. J. Leukocyte Biol. 70:518. antigen receptors. Science 252:1430. cells with disrupted primary V gene usage. Science 279:1729./ and / T cells migrate from thymus to the periphery in alternating waves. J. Exp. Med. 186:977./+ clones from thymus and peripheral blood. J. Exp. Med. 170:1521./ transcripts in human thymic ontogeny. EMBO J. 10:83.[Medline]
/ Cells. Annu. Rev. Immunol. 11:637.[Medline]
2V2+ T cells during mycobacterial infections. Science 295:2255./ T cells. J. Exp. Med. 187:357.This article has been cited by other articles:
|
|
 |
|
  L. Li and C.-Y. Wu CD4+CD25+ Treg cells inhibit human memory {gamma}{delta} T cells to produce IFN-{gamma} in response to M tuberculosis antigen ESAT-6 Blood, June 15, 2008; 111(12): 5629 - 5636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Barcy, S. C. De Rosa, J. Vieira, K. Diem, M. Ikoma, C. Casper, and L. Corey {gamma}{delta}+ T Cells Involvement in Viral Immune Control of Chronic Human Herpesvirus 8 Infection |
, n = 11); for other samples, two naive-defining markers were used (CD45RO- CD11adull, squares, n = 11). Data for cord blood samples are not included because the frequency of IFN-
