1 and V2 
T cells express distinct surface markers and might be developmentally distinct lineages
* Vaccine Research Center, National Institutes of Health, Bethesda, Maryland, and
Department of Genetics, Stanford University Medical School, Stanford, California
Correspondence: Stephen C. De Rosa, M.D., Vaccine Research Center, National Institutes of Health, 40 Convent Drive, Room 5612, Bethesda, MD 20814. E-mail: SDeRosa{at}nih.gov
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
T cells found in
human blood, V1 and V2, were found to have markedly different
phenotypes. V2 cells had a phenotype typical of most 
ß T cells
in blood; i.e., they were CD5+, CD28+, and
CD57-. In contrast, V1 cells tended to be
CD5-/dull, CD28-, and CD57+.
Furthermore, although V1 T cells appeared to be "naive" in that
they were CD45RA+, they were CD62L- and on
stimulation uniformly produced interferon-, indicating that they are
in fact memory/effector cells. This phenotype for V1 cells was
similar to that of intestinal intraepithelial lymphocytes, a subset
that can develop in the absence of the thymus. We suggest that the
V1 and V2 T cell subsets represent distinct lineages with
different developmental pathways. The disruption of the supply of
normal, thymus-derived T cells in HIV-infected individuals might be
responsible for the shift in the V2/V1 ratio that occurs in the
blood of individuals with HIV disease.
Key Words: intestinal intraepithelial lymphocytes • CD5 • FACS
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Subsets of T cells also change dramatically in frequency during
HIV disease. Human T cells use a limited set of variable-region
genes (V genes) compared with ß T cells. In fact, only two V
genes, V1 and V2, are commonly found among
T cells in humans. In HIV-uninfected individuals, V2
cells are the more prevalent type found in peripheral blood
[3
], whereas V1 cells are more prevalent among
intestinal intraepithelial lymphocytes (IELs) [4
]. In
HIV disease, there is a dramatic shift in the ratio of these two
T cell types in peripheral blood: V1 cells increase and V2 cells
decline in number and frequency [3
5
6
]. The reason
for this inversion of the V2/V1 ratio is not known. Studies
examining the junctional diversity of V1 and V2 cells in HIV
disease have determined that the inversion in this ratio is caused by
neither an antigen-driven clonal expansion of V1 cells nor a clonal
deletion of V2 cells [7
8
].
To understand the reason for the ratio shift, we used multicolor
fluorescence-activated cell sorting (FACS) to study in detail the
surface markers expressed by V1 and V2 cells. Unexpectedly, most
V1 cells expressed a distinctly different set of surface markers
compared with V2 cells, indicating that most V1 and V2 cells
belong to very different subsets. In fact, the characteristic phenotype
of each subset was not different in HIV-infected adults compared with
uninfected adults; i.e., the marked difference in phenotype between
V1 and V2 cells was preserved during HIV disease progression.
The phenotype of V1 cells differs from most peripheral-blood T
cells, but it is similar to the phenotype of intestinal IELs
[9
10
11
], cells that in the mouse develop in the absence
of a thymus [12
13 ]. In contrast, V2 cells have a
phenotype similar to that of the majority of T cells in blood, and the
loss of V2 cells in HIV disease is correlated with the loss of the
presumably thymus-derived naive CD4 and CD8 T cells. We suggest that
most V1 and V2 cells form functionally and developmentally
distinct subsets, perhaps representing distinct lineages. The changes
in representation of the V1 and V2 subsets might be a consequence
of the HIV-induced impairment of typical thymic T cell development.
Finally, the correlated loss of naive CD4, CD8, and V2 T cells in
HIV disease provides additional evidence for thymic impairment as
an important cause of progressive immunodeficiency during HIV disease.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Flow-cytometric analysis
Peripheral blood mononuclear cells (PBMCs) were
isolated from 7 mL of heparinized blood by Ficoll-Hypaque
centrifugation. Cells were stained with fluorescently labeled
antibodies as described previously [16
]. Most of the
FACS studies reported here used eight-color FACS methods developed in
our laboratory. These methods are described in detail elsewhere
[17
18
19
]. Described briefly here, cells were stained
with combinations of antibodies conjugated to the following eight
fluorescent dyes, which were excited by one of three lasers as
indicated: fluorescein isothiocyanate (FITC), phycoerythrin (PE),
Cy5-PE, and Cy7-PE, excited by a 488-nm argon laser; Texas Red,
allophycocyanin (APC), and Cy7-APC, excited by a dye laser tuned to 600
nm; and cascade blue, excited by a 407-nm krypton laser. Samples were
analyzed on a modified FACStar Plus (Becton-Dickinson, San Jose, CA)
equipped with these three lasers, 10 fluorescence detectors in addition
to forward and side scatter, and MoFlo electronics (Cytomation, Fort
Collins, CO). Data analysis, including compensation after the
data collection, was performed using FlowJo software (TreeStar, Inc,
San Carlos, CA).
For each sample, 2 x 105–3 x 105 events were collected. Absolute counts were determined by multiplying the frequency of each subset relative to lymphocytes (determined by scatter gating) by the absolute lymphocyte count obtained from a complete blood count analyzed by a commercial laboratory on a blood sample drawn simultaneously with the blood sample used for FACS analysis. Examples of some of the eight-color combinations used in this study are shown in Table 1 . The anti-perforin antibody was directly conjugated to Cy5 and collected in the APC channel. Data for Figure 3 (see below) were obtained using the three-color combination FITC CD20, PE CD5, and Cy5-PE CD3.
|
View this table: [in a new window] |
Table 1. Examples of 8-Color Staining Combinations
|
 View larger version (37K): [in a new window] |
Figure 3. CD5- T cells expand in HIV disease and include
and ß T cells. (a) Examples of three-color FACS 5%
probability plots showing the expression of CD5 (PE) and CD3 (Cy5-PE)
on scatter-gated lymphocytes from a representative HIV-uninfected
control and three HIV-infected individuals. The three HIV-infected
individuals showed the three patterns of CD3 expression on the
CD5- T cells: bright CD3 expression (second panel), CD3
expressed at the same levels as most CD5+ T cells (third
panel), or a mixture of these expression levels (fourth panel). The
CD3bright CD5- cells, as compared with the CD5
cells expressing typical levels of CD3, included a higher proportion of
T cells (data not shown). (b) The absolute CD5- T
cell counts are shown for HIV-infected individuals grouped by CD4 count
(n=209) and for the HIV-uninfected controls
(n=42). The boxes with the intersecting lines represent the
interquartile ranges and the medians for each group. The lines above
and below the boxes are the 90th and 10th percentiles, respectively.
The count for the infected group as a whole was significantly elevated
compared with that of controls (P<0.0001). (c) The
percentages of CD5- T cells that were TcR ß+
are shown for HIV-uninfected (n=21) and HIV-infected
(n=69) individuals. As in panel b, the boxes with the
intersecting lines represent the interquartile ranges and the
medians.
|
(IFN-), interleukin (IL)-2, CD3,
CD4, CD5, CD8, CD16, CD28, CD45RA, CD57, CD62L, TcR , and V2
were obtained from BD/PharMingen (San Diego, CA) and conjugated to the
indicated fluorochromes in our laboratory by standard protocols
(http://drmr.com/abcon). FITC-conjugated TCS1 (anti-V1) was
obtained from Endogen, Inc. (Woburn, MA). The antibodies to CD8ß and
perforin (dG9) were kindly provided by E. Reinherz (Dana Farber Cancer
Institute, Boston, MA) and E. Podack (University of Miami, FL),
respectively.
Intracellular cytokine and perforin assays
A FACS assay was used to determine the frequency of cell
staining for IFN- and IL-2 after in vitro stimulation
[20
]. PBMCs (106 cells) were stimulated in
flat-bottom 24-well plates with phorbol myristate acetate [PMA (50
ng/mL)] and ionomycin (1 µM) in the presence of monensin (1 µM)
for 6 h. For staining combinations that included CD62L, a matrix
metalloproteinase inhibitor [KB8301 (10 µM); BD/PharMingen) was
included during the 6-h incubation to prevent stimulation-induced
cleavage of CD62L from the cell surface. PBMCs were stained with the
anti-V1 and V2 reagents (for 15 min at room temperature with one
subsequent wash) before incubation with PMA and ionomycin, because a
nonspecific staining pattern was noted if the PBMCs were stained with
these reagents after the stimulation. After stimulation, cells were
harvested and washed with phosphate-buffered saline (PBS)-bovine serum
albumin (BSA)-azide once and stained for 15 min on ice in the dark to
determine their surface phenotypes. Ethidium monoazide bromide [EMA (5
µg/mL); Molecular Probes] was included with the surface stains to
label the dead cells. Cells were washed 3x in PBS, exposed to
fluorescent light for 10 min to covalently link the EMA, and then fixed
with 2% Formalin in PBS for 20 min at room temperature. After washing
3x in PBS-BSA-azide, cells were resuspended in permeabilization buffer
(0.5% saponin in PBS-BSA-azide) and kept at room temperature for 10
min. Cells were spun down, resuspended in permeabilization buffer
containing the optimal concentration of APC-conjugated anti-IL-2 and
either FITC- or cascade blue-conjugated anti-IFN-, and incubated at
room temperature for 30 min. Finally, cells were washed once with
permeabilization buffer and then 3x with PBS-BSA-azide. Cells were
stored on ice until FACS analysis (within 24–48 h). A similar assay
was used to stain for intracellular perforin, except that fresh PBMCs
were used, without any stimulation.
Statistical analysis
JMP software (version 4 for Macintosh) produced by the SAS
Institute, Cary, NC, was used for all statistical analyses.
Significance values for comparisons between groups were determined by
the nonparametric Wilcoxon’s rank sum analysis. Pearson correlation
coefficients were used to determine significant correlations between
V2 counts and CD4 and CD8 subset counts.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1 and V2 T cells tended to be expressed in distinct
subsets identified by the expression of CD5, CD57, and CD45RA2
cells normally outnumber V1 in the T cell population in
peripheral blood, the majority of the T cells in HIV-infected
people expressed V1 (Fig. 1
). These V1 cells were virtually all CD5- or
CD5dull (Fig. 2
). This increase in the V1 population did not reflect an overall
increase in the frequency of cells. Rather it reflected a change
in the proportion of cells that expressed V1 and belonged to
the CD5- or CD5dull subsets. Thus, although we
found no difference in the overall numbers of T cells between
HIV-infected and -uninfected subjects, there was a marked increase in
the proportion of CD5- (V1) cells in the
HIV-infected subjects (Fig. 1)
. V2 cells, in contrast, were rarely
CD5- and tended to express more CD5 than V1 cells (Fig. 2)
.
 View larger version (15K): [in a new window] |
Figure 1. The T cells that expand in HIV disease often lack expression of
CD5. The absolute counts for overall T cells and the V1 and
V2 T cell subsets in HIV-infected (n=73) and -uninfected
(n=33) individuals are compared. The right panel compares
percentages of the T cells that lack expression of CD5.
Horizontal bars show the medians. The frequencies for the
subsets were determined by FACS from three-color or eight-color
combinations including either an anti- monoclonal antibody
conjugated to FITC or PE, anti-V1 conjugated to FITC, or anti-V2
conjugated to FITC or PE. Absolute counts were obtained by multiplying
this frequency relative to lymphocytes by the absolute lymphocyte count
determined from a complete blood count.
|
 View larger version (48K): [in a new window] |
Figure 2. V1 and V2 cells differ in expression of surface markers, both in
HIV-infected and -uninfected individuals. (a) Example of an eight-color
FACS analysis of PBMCs from an HIV-uninfected individual. Data are
shown either as histograms or as 5% probability plots with or without
outliers. PBMCs (2x105) were scatter gated on
lymphocytes, and then lymphocytes were gated on CD3 to identify T cells
(data not shown); T cells were then gated on V1+ cells
(top panels) and V2+ cells (middle panels). T cells not
expressing V1 or V2 (lower left) and then gated on CD8 (lower
panels) are shown for comparison with the T cells. The levels of
expression of CD57, CD5, CD62L, and CD45RA are shown on these subsets.
The histogram for CD5 expression on CD8+ T cells is
reproduced in gray in the upper panels to show the lower level of CD5
expression of V1 cells. Percentages of cells within the subsets that
were CD57+, CD5-, and
CD45RA+CD62L- are listed. (b) The percentages
of V1 or V2 cells that expressed each listed marker are shown for
HIV-uninfected controls [upper panels (n=19–35)] and for
HIV-infected subjects [lower panels (n=7–62). Various
eight-color FACS staining combinations were used to collect these data.
The boxes with the intersecting lines represent the interquartile
ranges and the medians for each group. The lines above and below the
boxes are the 90th and 10th percentiles, respectively. The
naive phenotype was defined here as coexpression of CD62L and CD45RA.
All the comparisons between V1 and V2 (within HIV-infected or
-uninfected subjects) were significant at P < 0.003,
except for the naive phenotype, which was not significantly different
for the HIV-infected subjects. For comparisons between HIV-infected and
-uninfected cells, those that were significant are indicated as
follows: *, P < 0.001; **, P < 0.02.
(c) The median fluorescence level of expression of CD5 on non-T cells
(CD3-), V1 and V2 T cells, and
CD8+ and CD8- (largely CD4+) T
cells. Data are shown for eight HIV-infected individuals. Horizontal
bars represent the median levels for these eight subjects. P
values indicate the significance for comparisons between the adjoining
groups.
|
1 and V2 also differed markedly in the expression of CD45RA.
Although most V1 cells were CD45RA+, they were not
typical naive T cells in that they generally did not coexpress CD62L
(Fig. 2)
. In studies in which naive T cells have been identified only
by the expression of CD45RA (or the nonexpression of CD45RO), these
cells have been incorrectly counted within the naive population
[21
22 ].
CD28, the T cell-costimulatory molecule, and CD57, a glycoprotein of
unknown function normally expressed on a subset of CD8+ T
cells and some natural killer cells, were two other surface markers
that showed different expression between these subsets; most V1
cells were CD28- and CD57+ (Fig. 2)
. Note that
CD57 appeared to show three levels of expression, with V1 cells
showing the brightest expression level. Some V2 cells expressed CD57
at intermediate levels (examples shown in Fig. 5
below). Because the
demarcation between the CD57- and CD57dull
cells was not always very well resolved, we categorized the
CD57dull cells as "negative" in the figures. Few (if
any) of either the V1 or the V2 cells expressed CD4, and although
a larger proportion of V1 cells expressed CD8 compared with V2
cells, the majority of both V1 and V2 cells were CD4/CD8 double
negative (data not shown).
 View larger version (55K): [in a new window] |
Figure 5. Both V1 and V2 cells stained intracellularly for perforin. Shown
are 5% probability contour plots for the FACS-determined level of
expression of perforin (abscissa) and CD57 (ordinate) for V1, V2,
and CD3+ CD8+ T cells (shown for comparison)
for an uninfected individual (left) and an HIV-infected individual
(right). The staining combination used is listed in the last row of
Table 1
. The percentage of cells in each subset staining for perforin
is indicated. Note that these two samples were not stained on the same
day, so the absolute intensity of staining for CD57 or perforin between
the HIV-infected and the -uninfected samples cannot be compared.
|
1 T
cells. A small population of T cells including both and ß T
cells did not express detectable levels of CD5, and these
CD5- T cells increased in frequency and absolute count in
HIV disease. Figure 3
shows that this subset could be clearly identified as separate in
HIV-infected individuals based on the normal or bright level of
expression of CD3 and the lack of expression of CD5. The absolute count
for this subset became elevated relatively early in HIV disease (CD4 T
cell counts near 500/µL) and remained elevated until late in the
disease process. Although enriched for T cells compared with the
CD5+ population, about half of the CD5- T
cells in HIV-infected subjects and the majority of the
CD5- T cells in HIV-uninfected subjects expressed the
ß T cell receptor (Figure 3)
.
Representation, rather than the phenotype, of V1 and V2 cells
changed in HIV-infected individuals
The V1 and V2 subsets were distinct in both HIV-infected and
-uninfected individuals. For the most part, the characteristic
phenotypic features of these subsets were not changed; however, there
were some differences between infected and uninfected individuals in
the proportions of V1 or V2 cells expressing certain markers (see
Figure 2
). Consistent with the overall expansion of CD5- T
cells in HIV disease, the V1 subset in HIV-infected individuals
included a higher proportion of CD5- cells than did the
V1 subset in uninfected controls. Even the V2 subset included a
small fraction of CD5- cells that were rarely found in
V2 cells from controls. Similar to the loss of naive CD4 and CD8 T
cells in HIV-infected individuals, there were fewer V1 cells in
HIV-infected subjects with a naive phenotype (CD62L+
CD45RA+) compared with V1 cells in uninfected controls.
This difference was largely caused by the decreased proportion of V1
cells expressing CD62L in HIV-infected individuals.
V1 and V2 T cells showed a similar cytokine profile
Using a FACS method of intracellular cytokine staining, we
assessed the cytokine profile of the V1 and V2 T cells. Bulk
PBMCs were stimulated for 6 h with PMA and ionomycin and then
surface stained with a combination of markers that allow identification
of the subset phenotype. Finally, the cells were permeabilized and
stained with fluorescently labeled antibodies to the cytokines IFN-
and IL-2.
The majority of V1 and V2 cells stained for IFN- and very few
of these cells stained for IL-2 (Fig. 4
and Table 2
). This profile was similar to the cytokine profile of typical CD8
memory T cells. Note that in the examples shown in Figure 4V
1
cells showed a lower level of staining for IFN- compared with V2
cells. We observed this staining pattern for most samples analyzed.
This difference in staining suggested that the V1 cells might
produce less IFN- than that produced by V2 cells. However, the
relationship between relative brightness of staining and cytokine
production has not been definitively determined.
 View larger version (31K): [in a new window] |
Figure 4. Intracellular staining for cytokines shows that most V1 and V2
cells stained for IFN- but not IL-2. Shown are 5% probability
contour plots with outliers for the FACS-determined level of expression
of IFN- (abscissa) and IL-2 (ordinate) in V1, V2, and
CD3+ CD8+ T cells (shown for comparison) from
an uninfected individual (left) and an HIV-infected individual (right)
after 6 h of stimulation with PMA and ionomycin. Two eight-color
combinations of reagents were used, including either V1 or V2,
CD3, and CD8, in addition to IL-2 and IFN-. PBMCs
(5x104–2x105 collected) were first
gated on lymphocytes and then gated on V1, V2, or
CD3+ CD8+ cells.
|
|
View this table: [in a new window] |
Table 2. The Majority of Both V1 and V2 Cells Stain for IFN-, but
not IL-2
|
cells from
either HIV-infected or -uninfected individuals. The only difference we
observed that might be associated with HIV disease was that a higher
percentage of V1 cells from uninfected individuals stained for IL-2
(Table 2)
. The V1 cells that stained for IL-2 did not stain for
IFN-, were CD57-, and were naive in phenotype (data not
shown). This was consistent with typical CD5+ ß T
cells, in which naive subsets were IL-2+ and
IFN-- [20
]. Therefore, the loss of these
IL-2-staining cells in the HIV-infected subjects could be accounted for
by the loss of "naive" V1 cells. Because only four uninfected
individuals have been analyzed for cytokine profile, more must be
tested to establish significance.
Both V1 and V2 T cells stained intracellularly for perforin
A similar intracellular staining protocol was used to determine
levels of intracellular perforin, one of the proteins released from
cytotoxic T cells that participate in target cell killing. For this
assay, fresh unstimulated PBMCs were surface stained and then
permeabilized and stained intracellularly for perforin. Figure 5
shows that both V1 and V2 cells stained for perforin. For a
small number of samples analyzed, the majority of V1 and V2 cells
in both HIV-uninfected and -infected individuals stained for perforin
(Table 3 ). There was a higher frequency of V2 cells expressing perforin
in the HIV-infected group (P=0.02), but this difference must
be confirmed with a larger sample size. Figure 5
also shows that there
was some correlation between the expression of perforin and CD57. The
V2 cells, which expressed no or lower levels of CD57 compared with
the V1 and CD8 cells, also expressed lower levels of perforin.
|
View this table: [in a new window] |
Table 3. Most V1 and V2 T Cells Stain for
Perforin
|
2 cell numbers was correlated with the decline
in naive CD4 and CD8 T cells in HIV disease2 cells in the group of
HIV-infected individuals (n=99) that we studied was loosely
correlated with the decline in CD4 T cell counts [r=0.42
(Pearson correlation coefficient); P<0.0001]. This
correlation was largely caused by the correlation of V2 cell counts
with naive CD4 T cell counts [r=0.48; P<0.0001
(Pearson)]. Of the three memory subsets, only the CD45RA-
CD62L+ memory CD4 subset showed a correlation with the
V2 cells [r=0.33; P=0.0009 (Pearson)].
Because CD8 T cell counts showed a pattern of initial expansion with
declines noted only in late-stage AIDS, it was not surprising that the
V2 cell count was not correlated with the overall CD8 T cell
count. However, V2 cells were loosely correlated with the naive CD8
T cell count [r=0.42, P<0.0001 (Pearson)] but
were not correlated with any of the CD8 memory subsets. These
relationships were consistent with the hypothesis that there is a
similar mechanism responsible for the loss of naive T cells and the
loss of V2 cells, perhaps caused by HIV-induced impairment of thymic
function [23
24
25
26
].
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
T cells commonly found in
peripheral blood, V1 and V2, expressed distinct sets of surface
markers that caused them to be characterized as distinct subsets.
Indeed, these markedly different phenotypes suggest that they might be
functionally and developmentally distinct lineages. In HIV disease,
although there is an expansion of the V1 subset and a decline in the
V2 subset [3
5
6
], the phenotypes and likely the
functions of the subsets appear not to change. Therefore, it is the
relative representation of these subsets that changes in HIV disease
and not the nature of the subsets themselves.
CD5 is one surface marker shown to be expressed at different levels on
lineages of T and B cells. For example, CD4 T cells express higher
levels of CD5 than CD8 T cells (Fig. 2)
, and CD5 is expressed at low
levels on a subset of B-1 cells but not on conventional B-2 cells
[27
28
]. V1 and V2 cells differ in the level of
expression of CD5. Many V1 cells lack CD5 expression or express
lower CD5 compared with that of V2 cells. This difference is one
indication that most V1 and V2 cells might belong to different
lineages and that V1 cells might be related to other subsets of T
cells (including ß T cells) that expand in HIV disease and also
show lower levels of CD5 expression. Cells that lack CD5 have been
shown here (and previously) to expand in HIV disease
[29
]. CD5- T cells from HIV-infected adults
were previously reported to express only the ß T cell receptor
[29
]. We found, however, that T cells represent
nearly half of the cells in the enlarged CD5- T cell
subset in HIV-infected subjects.
The use of CD5 expression to distinguish T cell lineages among
intestinal IELs has been questioned in one study [30
].
In this report, murine CD5- CD8ß IELs are shown to
up-regulate CD5 surface expression on stimulation; therefore, the
authors conclude that CD5- and CD5+ CD8ß
IELs cannot be considered as separate lineages. However, these data are
not inconsistent with our proposal. First, that report studies only
ß and not T cells. In addition, when CD5 was up-regulated,
it usually was expressed at intermediate levels, and only infrequently
was it expressed at the levels typical of CD8 T cells found in blood.
Finally, we analyzed resting blood cells and not activated cells. It is
well known that the expression of some surface markers changes on
activation. For, example, CD4 is down-regulated on activation
[31
], and yet these cells remain classified as part of
the CD4+ lineage and not the
CD4-CD8- lineage.
V1 T cells have an overall surface phenotype that is different from
that of most peripheral blood T cells but is similar to that of T cells
found in the intestinal epithelium. Studies of intestinal IELs from
mice show that they include a large proportion of T cells, are
CD5- or CD5dull, and often lack CD28
expression [9
10
11
]. In humans, the majority of
intestinal IELs use ß T cell receptors; however, among the
T cells, almost all use V1 [4
]. Because studies in
mice have shown that intestinal IELs can develop in the absence of a
functioning thymus [12
13
], the phenotypic similarity
between these lymphocytes and V1 cells suggests that these latter
cells might also be capable of extrathymic development, perhaps in the
intestine.
In human peripheral blood, V2 cells usually outnumber V1 cells
[3
]. However, in HIV-infected individuals, V1 cells
become more prevalent due to a decline in absolute counts of V2
cells and an expansion of V1 cells [3
5
6
]. The
reason for this inversion in the ratio of V1 to V2 cells in HIV
disease is not known. An antigen-driven expansion of V1 cells seems
unlikely because the junctional diversity of V1 cells from
HIV-infected individuals does not differ significantly from that of
HIV-uninfected controls [7
8 ].
Based on the similar phenotypes of V1 blood cells and cells in the
intestinal epithelium, we propose that the source of the expanded V1
cells during HIV disease might be intestinal IELs that migrate into the
peripheral blood. This proposed migration of V1 cells from the
intestinal epithelium to the blood might be a consequence of
HIV-associated intestinal inflammation. Consistent with this
hypothesis, expansion of V1 cells in peripheral blood has been
observed in other diseases that are associated with intestinal
inflammation [32
].
The impairment of thymic function that occurs in HIV disease offers
another possible reason why the intestine could be the source of the
increased number of V1 cells in the blood of HIV-infected
individuals. Several studies have indicated that HIV infection is
likely to result in impairment of thymic function. Direct histological
evidence has demonstrated disruption of the thymic microenvironment by
HIV in humans and by simian immunodeficiency virus (SIV) in monkeys
[24
25
26
]. In a model where severe combined
immunodeficiency mice are implanted with human hematolymphoid
tissue (SCID-hu), HIV can infect thymocytes, leading to their
destruction and resulting in a disruption of thymic morphology
[23
]. We have previously reported the progressive
decline in naive CD4 and CD8 T cell counts in HIV disease [1
2
], which is likely a consequence of HIV-induced thymic damage
because the thymus is the major organ responsible for producing naive T
cells.
In HIV disease, there can be increased activity of extrathymic T cell
developmental pathways, as a consequence either of loss of thymic
function or an increased demand in general for T cell development or
both. The intestinal epithelium might be such an extrathymic site where
T cells (enriched for V1 T cells and CD5- ß
T cells) develop and eventually migrate to the periphery.
One of the surface markers that shows differential expression on V1
and V2 cells is CD45RA [21
22
]. This difference has
been shown previously; however, the interpretation then was that Vgd1
cells are naive, based on the high level of expression of CD45RA. We
confirmed that most V1 cells are CD45RA+, but we found
that most are CD62L- and hence not naive in phenotype.
CD45RA alone is not sufficient to distinguish naive from memory
cells; an additional marker such as CD62L or CD11a is required
to make this distinction. In addition, these cells stain for IFN-
but not IL-2, a function consistent with the designation of these cells
as memory/effector cells [20
]. Perforin expression in
both the V1 and V2 cells is also consistent with the designation
of these cells as memory/effector cells and is consistent with previous
reports showing that most T cells in blood express perforin
[33
].
In summary, we have found that most V1 and V2 cells have
characteristic patterns of antigen surface expression in both
HIV-infected and -uninfected individuals, and we suggest that these
phenotypic profiles define developmentally distinct lineages. Although
this conclusion might not be true for all V1 or V2 cells, because
these phenotypic profiles are not exclusive to either subset, it
appears to be valid for most cells in either subset, and therefore
V1 or V2 might be used as single markers to broadly define these
subsets.
Finally, we speculate that the V1/V2 inversion might be a
consequence of thymic impairment by HIV disease. The majority of V2
cells could develop in the thymus. They have a phenotype similar to
most peripheral blood T cells, which are presumably thymus derived. In
addition, the kinetics of the loss of the V2 cells during HIV
disease progression parallels the loss of the naive CD4 and CD8
subsets, whose loss likely reflects the decreased ability of the thymus
to generate T cells. On the other hand, V1 cells might develop
through an extrathymic pathway. They expand in HIV disease and are
similar in phenotype to a subset of cells shown to be capable of
extrathymic development, intestinal IELs. Finally, the changes in
numbers and ratio of V1 to V2 cells during HIV disease might
reflect HIV-induced alterations in the activities of thymic versus
extrathymic T cell developmental pathways.
Received March 7, 2001; revised May 9, 2001; accepted May 11, 2001.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J.-S. Cummings, C. Cairo, C. Armstrong, C. E. Davis, and C. D. Pauza Impacts of HIV infection on V{gamma}2V{delta}2 T cell phenotype and function: a mechanism for reduced tumor immunity in AIDS J. Leukoc. Biol., August 1, 2008; 84(2): 371 - 379. [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 J. Immunol., March 1, 2008; 180(5): 3417 - 3425. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I Tikhonov, C. Deetz, R Paca, S Berg, V Lukyanenko, J. Lim, and C. Pauza Human V{gamma}2V{delta}2 T cells contain cytoplasmic RANTES Int. Immunol., August 1, 2006; 18(8): 1243 - 1251. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. C. De Rosa, J. P. Andrus, S. P. Perfetto, J. J. Mantovani, L. A. Herzenberg, L. A. Herzenberg, and M. Roederer Ontogeny of {gamma}{delta} T Cells in Humans J. Immunol., February 1, 2004; 172(3): 1637 - 1645. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Poles, S. Barsoum, W. Yu, J. Yu, P. Sun, J. Daly, T. He, S. Mehandru, A. Talal, M. Markowitz, et al. Human Immunodeficiency Virus Type 1 Induces Persistent Changes in Mucosal and Blood {gamma}{delta} T Cells despite Suppressive Therapy J. Virol., October 1, 2003; 77(19): 10456 - 10467. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
