| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Microbiology and Immunology, University of Melbourne, and
Victorian Infectious Diseases Service, Royal Melbourne Hospital, Parkville, Victoria, Australia;
Los Alamos National Laboratory, Los Alamos, NM 87545; and
National Center for HIV Epidemiology and Clinical Research, University of New South Wales, and
¶ Center for Immunology, St. Vincent’s Hospital, Sydney, New South Wales, Australia
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The role of the thymus in T cell homeostasis is critical to our understanding of T cell depletion, T cell turnover, and immune reconstitution. A fraction of recent thymic emigrants carries episomal TCR excision circles (TREC)5 that do not replicate with cell division (3, 4). The number of TREC per million cells is dependent on thymic output, T cell proliferation, and the rate of death of both TREC-positive and -negative T cells. Hazenberg et al. (5) demonstrated an inverse relationship between the number of TREC per million cells and cell division in the naive T cell population from individuals with chronic HIV infection (CHI), suggesting that the reduction of TREC per million cells in HIV infection is a consequence of increased T cell proliferation and is not due to a reduction in thymic output. As suggested by Hazenberg et al. (5), interpretation of these results is complex because T cell proliferation, T cell death, and thymic function are all altered by HIV infection (6, 7, 8) and, in addition, each of these processes will also affect the quantification of TREC (4, 9).
To help unravel these competing processes we quantified TREC in two ways. First, we measured TREC per million T cells (CD4+ and CD4-) and then, to account for changes in T cell number due to either HIV infection or therapy, we calculated TREC per milliliter of blood. Proliferation, for example, is thought to decrease TREC by dilution (5). This is clearly true if one measures TREC per million CD4+ T cells, because division will not increase the number of TREC but will increase the number of cells. However, if one measures TREC per milliliter there will be no change because the number of TREC remains unchanged during cell division. If death occurs simultaneously with proliferation to maintain cell numbers, then TREC per milliliter will decrease as TREC+ cells die. Thus, when multiple processes occur, interpretation may be difficult and precise methods of analysis are needed.
Therefore, we measured the total number and the percentage of proliferating T cells and the number of TREC per million cells in sorted CD4+ and CD4- T lymphocytes, and we calculated the number of TREC per milliliter of blood, in the setting of primary HIV infection (PHI) and CHI, before and after effective antiretroviral therapy. We also developed a mathematical model that can account for changes in TREC due to input from a thymic source, cell proliferation, cell death, and lymphocyte redistribution. We then used this model to interpret the effects of antiretroviral therapy.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Patients were defined as having PHI if they presented with signs or symptoms of acute retroviral syndrome, had a negative or evolving Western blot (WB) pattern with detectable HIV RNA and proviral DNA, or had a positive WB pattern and a negative ELISA/WB result within the previous 6 mo. Patients were defined as having CHI if they had known HIV infection for >6 mo, with consistently detectable HIV RNA during this time. No patient had received prior antiretroviral therapy. Blood samples were obtained from both PHI (n = 19) and CHI (n = 14) patients recruited into Institutional Review Board-approved clinical trials of two nucleoside reverse transcriptase inhibitors (zidovudine, didanosine, lamivudine, or stavudine) plus a protease inhibitor (nelfinavir or indinavir).
Cell sorting and real-time PCR for TREC quantification
Concentrations of CD4, CD8, and naive T cells (CD45RA+62L+) were determined by flow cytometry of whole blood, as previously described (10). PBMC were prepared from blood by Ficoll-Paque (Pharmacia Biotech, Uppsala, Sweden) centrifugation and frozen. After thawing, PBMC were incubated with fluorochrome-labeled mAbs, CD3-PE and CD4-FITC, for 30 min on ice (BD Biosciences, San Jose, CA) and sorted into CD3+CD4- and CD3+CD4+ populations using a cell sorter (MoFlo CLS; Cytomation, Fort Collins, CO). A tight lymphocyte gate in the forward scatter-side scatter diagram was used. At least 2 x 105 cells were sorted for each population. DNA was extracted from the purified cell populations using direct cell lysis (3). Sorting into CD3+CD4+ and CD3+CD8+ cells was not possible due to limitations in the number of fluorochromes available. Therefore, for the CD8 T cell studies there is an assumption that CD8 T cells are the major cells in the CD4- T cell fraction. The percentage of CD3+CD4-CD8- T cells in the CD3+CD4- fraction has been reported to be 5% and to be unaltered by HIV infection, CD4 count, or antiretroviral therapy (11).
Quantification of TREC per million PBMC, CD4+ cells, or CD4- T cells was performed using real-time PCR and molecular beacons as previously described (4). Results for TREC are expressed per million cells or are adjusted for the concentration of CD4+ or CD4- T cells and expressed as CD4+ or CD4- TREC per milliliter (i.e., TREC per milliliter = TREC per cell x cells per milliliter = (TREC per million cells) x 10-6 x (cells per microliter) x 103 = (TREC per million cells) x (cells per microliter) x 10-3).
Intracellular Ki67 staining
The fraction of dividing T cells was assessed by the combination of surface staining with CD3-PE and CD4-PerCP and intracellular staining with Ki67-FITC (DAKO, Carpinteria, CA) as previously described (12). All assays were performed on frozen cells. Direct comparison of fresh and frozen PBMC from both uninfected (n = 4) and HIV-infected (n = 4) donors revealed no effect of freezing on the percentage of positive Ki67 cells (data not shown).
Model
To interpret the data obtained in this study, we developed a
mathematical model (Fig. 1
) that is a
generalization of the model presented by Hazenberg et al.
(5). The model includes T lymphocytes in blood
(TB) and in lymphoid and other tissues
(TL). These populations are then
divided into TREC+
(TB+,
TL+) and
TREC-
(TB-,
TL-) subpopulations.
This division is convenient for our study but does not necessarily
reflect other common subsets. For instance, the
TREC- cells will be a mixture of naive and
memory cells. The blood compartment receives an input of T cells from
the thymus (
), of which a fraction (f)
contains TREC. Here we ignore potential extrathymic sources of new
TREC+ T cells. Although these sources may exist
(13), they are likely negligible (3, 13, 14).
|
i, then
we could include this in the model by adding a term decreasing the
number of TREC+ cells at rate
i and similarly increasing the number of
TREC- T cells at the same rate. Here, as in
other models (5), we make i
= 0.
Because measurements of TREC-containing cells are made in blood, we
also kept track in the model of the effects of cell trafficking. We
assumed that TREC- lymphocytes migrate from
blood to tissue at rate vB, and from
tissue back into blood at rate vL,
whereas the corresponding rates for TREC+ T cells
are vBT and
vLT. Finally, we assume that T cells
die at the same rate in tissue and in blood. Using these assumptions,
the model in Fig. 1
can be converted into the following system of
differential equations for either the CD4+ or
CD4- T cell populations:
 |
 |
 |
 | (1) |
We used the model to determine steady state levels of TREC and to
analyze how antiretroviral treatment, which leads to changes in the
parameters, would change TREC per milliliter and TREC per million cells
in blood. Solution of the model indicates that the steady state levels
of TREC per milliliter and TREC per million cells in blood are,
respectively,
 | (2) |
 |
 | (3) |
shows that the proliferation rates, p
and pT, do not affect the steady state
level of TREC per milliliter, a point of some importance. Similarly,
inspection of Eq. 3 demonstrates that the steady state of TREC per
million cells is not affected by the total thymic output
, but only by the fraction of the output that is
TREC+. We have also considered a variant of Eq. 1
in which T cell proliferation is density dependent, i.e., the term
 |
 |
.
To study the effect of antiretroviral therapy, we calculate the
derivatives of the steady state expressions for TREC per milliliter and
TREC per million cells with respect to each relevant parameter. If the
derivative is positive, then increases in the parameter lead to
increases in the steady state TREC value. If the derivative is
negative, then increases in the parameter lead to decreases in the
steady state TREC value. For example, to see the effect on TREC per
milliliter caused by a decrease in the death rate of
TREC+ cells (dT)
during therapy, we differentiate Eq. 2 with respect to
dT. We obtain
 | (4) |
. Thus, we
conclude that an increase in dT will
cause a decrease in TREC per milliliter, or equivalently that any
decreases in the death rate of TREC+ cells lead
to increases in TREC per milliliter in blood. We note that this
technique is general and obviates the need to use specific parameter
values to assess the impact of parameter changes on TREC per milliliter
and TREC per million cells. To analyze the influence of more than one parameter changing simultaneously, as might be expected during treatment, we make two additional assumptions. First, that TREC per million cells is similar in blood and tissue, as reported in rats (17) and humans (3, 18). Second, that the fraction of T cells in blood remains constant. This assumption has been extensively used in the HIV literature, usually by assuming that 2% of lymphocytes are in blood. These assumptions simplify the expression for the steady state level of TREC per million cells (data not shown) and allow the analysis of the effects of simultaneous parameter changes. The simplified expression depends only on the parameters p, pT, d, dT, and f.
For simplicity of analysis, we assume that migration rates are the same
for TREC+ cells and TREC-
cells, i.e., vBT =
vB and
vLT =
vL. There is experimental evidence
that supports this assumption, at least in murine models (17, 19, 20, 21), where it is argued that differential accumulation of
naive and memory cells in different tissues are due, for example, to
different proliferation rates in diverse compartments, as we assume in
our model, rather than different migration rates. In any case, even if
this simplification is not fully tenable in humans, our results (see
Table I) remain unchanged as long as therapy induces similar changes in
the migration patterns of TREC+ and
TREC- cells (e.g., treatment reduces both
vB and
vBT by the same amount).
|
Comparison of acute and chronic infection was made using the Wilcoxon rank sum test. Comparison of the results under therapy vs baseline was done using a paired, one-tailed Wilcoxon signed rank test. Correlations between continuous variables were assessed using the Pearson correlation coefficient. Significance was determined as p < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The model in Fig. 1 (see Materials and Methods) was
used to explore the expected effects of antiretroviral therapy on the
steady state values of TREC per milliliter and TREC per million cells.
We present these results in Table I. For
example, reduction in the death rate of TREC+
cells (dT) leads, as expected, to an
increase in TREC per million cells and TREC per milliliter. However,
therapy reduces the death rate of TREC-
lymphocytes as well (d), with the opposite effect: it
reduces TREC per million cells and has no effect on TREC per
milliliter. So the net effect of reduced death rates varies depending
on how therapy affects d relative to
dT. Examining the effects of
therapy on extrathymic T cells shows that a decrease in the
proliferation rate of mature T cells after effective antiretroviral
therapy (22) leads to different outcomes in the two
measures of TREC content, increasing TREC per million cells but leaving
the number of TREC per milliliter unchanged (Table I).
In contrast, changes in recirculation rates, which lead to release of
lymphocytes trapped in the lymph nodes to the blood, increase TREC per
milliliter but do not change TREC per million cells, assuming that TREC
per million cells is the same in lymph nodes and blood, as reported in
rats (17) and humans (3, 18). Finally, if
HIV-1 infection affects thymic function, then treatment may lead to an
increase in thymic output. Increases in lead to an
increase in TREC per milliliter but will lead to an increase in the
steady state value of TREC per million cells only if T cell
proliferation is density dependent (see Materials and
Methods). Another possible effect of treatment-induced recovery of
the thymus is to increase the fraction of TREC+
cells leaving the thymus (f) due to reduced thymic
proliferation of maturing T cells. If this occurs, then both TREC per
million cells and TREC per milliliter increase (Table I).
TREC in CD45RA+ and CD45RA- T cells in healthy and HIV-1-infected individuals
In a subset of individuals with HIV-1 infection (PHI,
n = 2; CHI, n = 6) and age-matched
HIV-uninfected individuals (n = 4), we quantified TREC
per million cells in separated CD4+ and
CD4- T cells, further sorted into
CD45RA+ and CD45RA-
subsets. The number of TREC per million cells was significantly higher
in CD45RA+ cells than in
CD45RA- T cells in both healthy and infected
individuals, as previously reported for naive and memory T cells
(3, 4, 9). When we compared TREC per million
CD45RA- cells between infected and uninfected
individuals, we found no statistical difference
(p = 0.154 for CD4+ and
p = 0.383 for CD4-; Fig. 2). This may be a consequence of the
small sample size or, more likely, it may be explained by an increase
in the input of TREC-containing cells into the
CD45RA- pool of HIV-infected individuals, due to
lymphocyte activation or transient acquisition by a
CD45RA+ T cell of a memory-like phenotype
(CD45RA-) after homeostasis-driven proliferation
(23, 24). The data shown here suggest that this added
input balances the decrease in TREC due to increased proliferation and
death rates in the CD45RA- lymphocyte population
of HIV-infected individuals.
|
We calculated TREC per milliliter, as described in Materials and
Methods, for
CD4+CD45RA+ and
CD4+CD45RA- T cells
separately. Even though the number of TREC per million cells is
larger in CD45RA+ T cells, there are many fewer
CD45RA+ T cells than
CD45RA- T cells in these HIV-1-infected
individuals; thus, in terms of TREC per milliliter, only 50% are in
CD45RA+ T cells (95% confidence interval,
31.2–68.7%) (Table II). For this reason
we believe that the TREC contribution of the
CD45RA- population should not be neglected in
the setting of HIV-1 infection.
|
Baseline comparison between CHI and PHI individuals
Baseline characteristics of individuals with PHI and CHI before
treatment are shown in Table III. Table III and previous work (3, 4) show that HIV infection leads
to a decrease in TREC-containing cells. Individuals with CHI, in
relation to individuals with PHI, have significantly lower numbers of
TREC, measured as TREC per million cells or as TREC per milliliter, in
both CD4+ and CD4- T cell
compartments (p < 0.001). In PHI, TREC per
million CD4+ T cells is strongly correlated with
TREC per million CD4- T cells
(r2 = 0.74; p <
0.0001), but this is not observed in CHI
(r2 < 0.1; p
= 0.5).
|
). In individuals with PHI there was no correlation between baseline CD4+ T cell count and the percentage of Ki67+CD4+ T cells or the absolute number of Ki67+CD4+ T cells (p = 0.47 and p = 0.10, respectively). However, such a correlation has been seen in CHI (15, 26). In the CD4- fraction of individuals with PHI, we found a strong correlation between baseline CD4- numbers and both the percentage of Ki67+CD4- and absolute numbers of Ki67+CD4- T cells (p = 0.012 and p < 0.001, respectively), again in contrast to previous findings in CHI (26).
Effect of antiretroviral therapy on T cell turnover and TREC
Over the first 8 wk of effective antiretroviral therapy there was an increase in both the total CD4+ and naive CD4+ T cell populations in PHI and CHI. Higher cell counts in relation to baseline were maintained throughout the follow-up period. The percentage of CD4+ naive T cells in relation to the total pool of CD4+ cells remained constant throughout the period of therapy in both study groups. This indicates that the recovery of these populations (total CD4+ and naive CD4+) proceeds at the same rate during HAART. We also observed a sustained increase over baseline of the number of CD8+ naive cells after treatment of PHI.
In general, the percentage of
Ki67+CD4+ and
Ki67+CD4- T cells declined
after antiretroviral therapy of individuals with PHI and CHI, in
agreement with previous studies (5, 15) (Fig. 3). However, when we measured the total
number of cells expressing Ki67 per microliter of blood, the results
were quite different for PHI and CHI (Fig. 3, C and
D, respectively). In treated PHI, there was a sustained and
significant reduction in the total number of
Ki67+ cells in relation to baseline for both
CD4+ and CD4- populations,
for all of the posttherapy follow-up. In contrast, in CHI the absolute
number of Ki67+CD4+ T cells
was not reduced and we did not detect any significant difference
between the number of
Ki67+CD4+ T cells at
baseline and at 8, 24, and 48 wk of antiretroviral therapy. This result
differs from that obtained by Hazenberg et al. (5), where
a significant reduction in total numbers of
Ki67+CD4+ cells was
observed already after 10 wk of HAART, even though a transient increase
was observed again at 24 wk. Most likely, our individuals with CHI are
a more homogeneous group at a more advanced stage of HIV infection, as
measured by CD4+ T cell counts (median
CD4+ cells: 180 cells per microliter in this
study and 280 cells per microliter in Ref. 5).
|
). In the
CD4+ T cell population there was a clear
difference between PHI and CHI. In individuals with PHI there was only
a transient increase in TREC per milliliter after 8 wk of therapy
(p = 0.028, one-sided test). This significant
increase in relation to baseline rapidly disappeared and was not
observed at 24 and 48 wk (Fig. 4C). In contrast, for CHI
there was a significant increase in CD4+ TREC per
milliliter at all time points (p < 0.02,
paired one-sided test) in relation to baseline and there was either a
significant increase or a trend for an increase in the number of TREC
per million CD4+ T cells
(p = 0.091 at 8 wk, p = 0.046
at 24 wk, and p = 0.055 at 48 wk) (Fig. 4D).
|
). After treatment of CHI there was a sustained
increase in TREC per milliliter and TREC per million cells from
baseline (with p < 0.02 for all comparisons except
TREC per million CD4- cells at 8 wk
(p = 0.076)). However, after treatment of PHI
there was only a transient trend for increased TREC per milliliter at 8
wk, which did not reach significance (p =
0.060), whereas the number of TREC per million
CD4- cells increased slowly and the difference
in relation to baseline was significant only at 48 wk
(p = 0.046).
We next looked for correlations between T cell proliferation and TREC
content in both acute and chronic infection. Before treatment for both
CD4+ and CD4- T cells
there was no correlation between TREC per million cells or TREC per
milliliter and the percentage of Ki67+ or total
Ki67+ T cells for either PHI or CHI. After
effective antiretroviral therapy of PHI and CHI, there was a
significant and consistent reduction in the percentage of
Ki67+ cells. However, we could not detect any
significant correlation between reduction in the percentage of
Ki67+ and changes in TREC per million cells, as
suggested before (15). In Fig. 5A, we show some clear
examples (PHI 1-3, CHI 1) of individuals with a profound reduction in
their CD4+ proliferation rates, without an
increase in their TREC per million values, particularly during the
first 8 wk of treatment. In fact, analysis of changes in single
individuals showed considerable variation in the posttherapy evolution
of TREC per million cells and TREC per milliliter (Fig. 5). The
patient-to-patient variability in TREC after treatment of PHI and CHI
is consistent with our previous work (4) and reinforces
the idea that the observed reduction in proliferation rates was not the
driving force behind the measured changes in TREC.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). The preservation of
TREC in CD45RA- T cells in HIV infection may be
a reflection of the higher degree of immune activation and consequent
transition from a CD45RA+ to a
CD45RA- state in both CD4+
and CD4- T cells (27).
Alternatively, as described in mouse models of lymphopenia, naive T
cells may acquire a memory-like phenotype during homeostasis-driven
proliferation (23, 24) without loss of TREC.
Furthermore, our analysis of TREC per milliliter in
CD45RA+ and CD45RA- T
cells shows that, in the setting of HIV-1 infection, only 
50% of
TREC are in "naive"
CD45RA+CD4+ cells. We
calculated TREC per milliliter from TREC per million cells by adjusting
for the absolute numbers of CD45RA+ and
CD45RA- T cells per milliliter of blood. When we
performed this calculation (Table II) we found equivalent amounts of
TREC per milliliter in
CD45RA+CD4+ and
CD45RA-CD4+ T cells.
Therefore, even though the number of TREC per million cells is larger
in CD45RA+ than in CD45RA-
T cells, because there are more CD45RA- than
CD45RA+ T cells in HIV infection, this will
increase TREC per milliliter for CD45RA- T
cells. This would not necessarily be the case in HIV-infected children
or young adults.
Several processes affect TREC in a cell population. TREC measured in
blood can increase due to the input of TREC+
naive cells from the thymus or input of TREC+
cells from an extrathymic source. Two processes lead to a decrease in
TREC: dilution of TREC due to proliferation (without TREC replication)
and loss of TREC due to the death of TREC+ cells.
These processes differently affect TREC per milliliter and TREC per
million cells (Table I), because TREC per milliliter depends only on
TREC+ cells, whereas the number of TREC per
million cells depends on both TREC+ and
TREC- cells. To account for the changes in TREC
per milliliter and TREC per million cells we developed a model that
incorporates T cell proliferation, death, redistribution, and thymic
production.
During antiretroviral therapy of PHI, we observed a transient increase
in TREC per milliliter and unchanged TREC per million cells. According
to our model, these observations can be explained by lymphocyte
redistribution from lymphoid tissue to blood (Table I). Treatment
reduces the amount of virus in lymph nodes (28, 29, 30) and
hence promotes the release of T cells back into the circulation. The
concomitant input of TREC+ cells into the blood
increases TREC per milliliter, but, because TREC per million cells seem
to be identical in blood and lymph nodes (3, 17, 18),
redistribution leaves the number of TREC per million cells in the blood
unchanged. Redistribution has been implicated in the initial rise of T
cell counts posttherapy but not in the continuing rises seen with
months of therapy (31, 32, 33, 34). Thus, redistribution could
also explain the transient nature of the increase in TREC per
milliliter, because redistribution would not constitute a permanent
source of TREC+ cells into the blood. Other
effects do not offer a convincing explanation for the transient
increase in TREC per milliliter. For example, increases in the thymic
output of TREC+ cells lead to continuing
increases in TREC per milliliter and TREC per million cells, which we
do not observe. However, we cannot rule out a small increase in thymic
output due to therapy. During PHI, thymic output only contributes a
small fraction of the TREC already present in blood (5),
and thus a small increase due to therapy may be difficult to detect.
Surprisingly, proliferation seems not to be a major influence on the
changes in TREC in CD4+ and
CD4- T cells in PHI, because, despite a
consistent and pronounced decrease of markers of proliferation, we do
not observe a consistent increase in TREC per million cells (Fig. 5A). This result is consistent with a recent report on early
HIV infection that suggested that the decrease in
CD4+ T cell TREC per million during HIV infection
could not be the result of increased turnover of naive
CD4+ T cells (9). In addition,
decreases in proliferation should not affect TREC per milliliter (see
below).
In CHI, after initiation of treatment, we observe a sustained increase
in both TREC per million cells and TREC per milliliter. In the model,
this can only be a reflection of increased thymic production of
TREC+ cells ( or f) or a
preferential reduction of the death rate of TREC-containing cells
(Table I). The TREC per million cells exiting the thymus is larger than
the TREC per million cells of the blood; thus, new thymic cells
contribute not only to increases in TREC per milliliter but also to
increases in TREC per million cells. Because TREC only disappear when
cells die, reduction in the death rate of TREC+
cells due to therapy results in increases in TREC per milliliter if
there is an unchanged or increased source of TREC from the thymus.
However, reduction of the death rate of TREC+
cells can lead to the observed increase in TREC per million cells only
if there is a smaller reduction in the death rate of
TREC- cells. We do not know of any biological
reason or observation for the preferential reduction of the death rate
of TREC+ cells, and thus reduction of the death
rate per se is unlikely to explain the observed changes in TREC with
therapy of CHI. Reductions in proliferation have been proposed as
responsible for the increase in TREC per million cells (9, 15). However, proliferation does not affect TREC per milliliter
(Eq. 2), and thus reductions in proliferation cannot account for the
observed increase in TREC per milliliter.
In summary, only the increased output of TREC+ T cells from the thymus explains the observed changes in TREC with treatment of CHI. Moreover, we find a very significant correlation between the changes in TREC per milliliter and the changes in CD4+ naive cells per microliter posttherapy (p = 0.001, with a random effects model), suggesting that the increase in naive cells is mainly due to recent thymic emigrants (35) rather than to expansion of existing naive cells.
Our results are partly in contrast with a recent finding of a positive correlation between the increase in TREC per million cells and the fold decrease in proliferating naive T cells 6 mo after effective therapy for HIV infection (15). Another study also describes an inverse correlation between TREC per million naive cells (CD45RA+) and naive cell division (measured by Ki67 expression in CD45RO-CD27+ cells), although these results were cross-sectional and did not include detailed treatment analysis (5). One can speculate that this inverse correlation may have been a consequence of the inclusion of cells with a phenotype between that of true naive and effector/memory cells (so called "transitional cells") into the naive cell gate, which may artificially increase the percentage of Ki67 in the "naive" fraction, as argued before (9). In fact, when transitional cells are excluded from the analysis there is no association between TREC per million cells and Ki67 in the naive cell fraction in individuals with early HIV infection (9). We also did not find a correlation between TREC per million total CD4+ or CD4- T cells and Ki67 expression in total CD4+ or CD4- T cells either before treatment or following 24 wk of treatment. Whether this discrepancy is due to differences in the duration of infection at the time of treatment initiation, differences between naive T cells and total T cells, or other factors is unknown. However, we believe caution is necessary in interpreting associations between decreases in proliferation and increases in TREC.
It is clear that changes in proliferation do not affect TREC per
milliliter, because TREC are not created nor destroyed when cells
divide (3) (see Model, Eq. 2, and Table I). In
contrast, a reduction in proliferation rates leads to an increase in
TREC per million cells, but only in the context of unchanged death
rates. However, such a scenario leads to a decrease in cell counts,
which is clearly not observed with HIV treatment (Fig. 5). In fact, it
is well established that treatment results in reduction of both
proliferation and death rates (22, 36, 37). Our model
shows that if proliferation (p, pT)
and death (d, dT) rates are
reduced by the same proportion, then the equilibrium number of TREC per
million cells will remain unchanged. This is because the total number
of cells depends on p, pT and d,
dT, whereas the number of TREC+
cells depends on dT; if all these
parameters are reduced equally then both the
TREC+ cells and the total cells increase by the
same proportion and TREC per million cells remains unchanged.
Finally, we emphasize that in our patients, even though the the percentage of Ki67+CD4+ cells correlates well with the percentage of Ki67+CD4- cells in both PHI and CHI (r2 = 0.8, p < 0.0001 in PHI; r2 = 0.4, p = 0.037 in CHI), TREC per million CD4+ cells correlates with TREC per million CD4- cells only in PHI (r2 = 0.7, p < 0.0001) and not in CHI (r2 < 0.1, p = 0.5). This indicates that similar high levels of proliferation are maintained in both CD4+ and CD4- T cells during infection, with different impacts on TREC per million cells in those two lymphocyte populations. Thus, discussions in terms of the effect of single processes on TREC should be taken with caution, because several processes—proliferation, death, thymic output, redistribution—occur simultaneously (9). Taken together our results demonstrate that, to accurately interpret TREC data, one needs to use more sophisticated theoretical models, such as the one presented here, with concomitant assessment of disease status, TREC per milliliter, TREC per million cells, T cell proliferation, and T cell death, in sorted naive and memory cells.
We have shown that perturbations in T lymphocyte dynamics differ significantly in PHI and CHI, and we demonstrate that reduction in TREC after HIV infection cannot be explained by proliferation alone. Furthermore, after effective antiretroviral therapy of PHI and CHI, we have identified different patterns of increases in TREC per milliliter and TREC per million cells. Using a mathematical model to study the likely mechanism explaining these changes, we conclude that the changes in TREC after effective therapy for PHI are due to T cell redistribution from lymphoid tissue to blood. However, after treatment for CHI, an increase in thymic output is necessary to explain TREC increases, in particular the observed increase in TREC per milliliter. Thymic contribution is an important component of immune reconstitution after treatment of CHI and should be the target of future development of new therapeutic agents.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 This work was presented in part at the Eighth Conference on Retroviruses and Opportunistic Infections, Chicago, IL, February 4–8, 2001.
3 Address correspondence and reprint requests to Dr. Sharon R. Lewin, Department of Microbiology and Immunology, University of Melbourne, Royal Parade, Parkville, Victoria, 3052, Australia. E-mail address: s.lewin{at}microbiology.unimelb.edu.au
4 Current address: Outpatient Department of Internal Medicine, University Hospital, CH-4031 Basel, Switzerland.
5 Abbreviations used in this paper: TREC, TCR excision circle; PHI, primary HIV infection; CHI, chronic HIV infection; TB, T lymphocytes in blood; TL, T lymphocytes in lymphoid and other tissues; WB, Western blot; HAART, highly active antiretroviral therapy.
Received for publication January 17, 2002. Accepted for publication August 8, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
TCR- T cells is increased HIV disease. AIDS Res. Hum. Retroviruses 17:977.[Medline]
locus V(D)J recombination by-products are abundant in thymocytes and mature T cells. Mol. Cell. Biol. 16:609.[Abstract]
on the circulation pattern. J. Immunol. 152:1744.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  S. Saleh, A. Solomon, F. Wightman, M. Xhilaga, P. U. Cameron, and S. R. Lewin CCR7 ligands CCL19 and CCL21 increase permissiveness of resting memory CD4+ T cells to HIV-1 infection: a novel model of HIV-1 latency Blood, December 15, 2007; 110(13): 4161 - 4164. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-L. Dion, R. Bordi, J. Zeidan, R. Asaad, M.-R. Boulassel, J.-P. Routy, M. M. Lederman, R.-P. Sekaly, and R. Cheynier Slow disease progression and robust therapy-mediated CD4+ T-cell recovery are associated with efficient thymopoiesis during HIV-1 infection Blood, April 1, 2007; 109(7): 2912 - 2920. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Delobel, M.-T. Nugeyre, M. Cazabat, K. Sandres-Saune, C. Pasquier, L. Cuzin, B. Marchou, P. Massip, R. Cheynier, F. Barre-Sinoussi, et al. Naive T-cell depletion related to infection by x4 human immunodeficiency virus type 1 in poor immunological responders to highly active antiretroviral therapy. J. Virol., October 1, 2006; 80(20): 10229 - 10236. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. van den Dool and R. J. de Boer The Effects of Age, Thymectomy, and HIV Infection on {alpha} and beta TCR Excision Circles in Naive T Cells J. Immunol., October 1, 2006; 177(7): 4391 - 4401. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Di Mascio, I. Sereti, L. T. Matthews, V. Natarajan, J. Adelsberger, R. Lempicki, C. Yoder, E. Jones, C. Chow, J. A. Metcalf, et al. Naive T-Cell Dynamics in Human Immunodeficiency Virus Type 1 Infection: Effects of Highly Active Antiretroviral Therapy Provide Insights into the Mechanisms of Naive T-Cell Depletion J. Virol., March 15, 2006; 80(6): 2665 - 2674. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Krenger, H. Schmidlin, G. Cavadini, and G. A. Hollander On the Relevance of TCR Rearrangement Circles as Molecular Markers for Thymic Output during Experimental Graft-versus-Host Disease J. Immunol., June 15, 2004; 172(12): 7359 - 7367. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||
), TREC per
milliliter (•), the percentage of Ki67+ cells (
), and
total cell counts (
), all for CD4+ T cells in single
individuals, during treatment of PHI (A) and CHI
(B).