|
|
||||||||

*
Division of Infectious Diseases and
Center for AIDS Research, University of Pennsylvania, Philadelphia, PA 19104
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
In contrast to mitogen activation, Ag-specific activation is complex. During a response to Ag, both Ag-specific and bystander T cells become activated. Bystander activation represents non-TCR-mediated activation as a result of mediator release or triggering of adhesion or costimulatory molecules (23, 24, 25). Earlier studies of bystander activation suggested that it was common in CD8+ T cells and more limited in CD4+ T cells. With the advent of newer techniques for measuring Ag-specific cells, it appears that bystander activation in CD8+ T cells is, in fact, minimal (26, 27, 28). On the other hand, CD4+ T cell bystander activation is being observed in an increasing number of experimental systems (29, 30). Bystander activation of CD4+ T cells has also been demonstrated to be relevant as part of the immunopathology of multiple viral diseases, including herpes simplex (31), hepatitis B and C (32), diabetes induced by coxsackie virus (33), and dengue (34).
In this report, we identify the populations of
CD4+ T cells that replicate HIV during an
Ag-specific immune response. As a model, we used DC presentation of
superantigen (SA) to and stimulation of CD4+ T
cells. SA are bacteria- or virus-derived proteins that bind to MHC
class II molecules and a constant region of the variable
-chain of
the TCR. The ability of SA to activate a large population of T cells
defined by V
chain expression enables accurate identification of
Ag-specific cells. SA stimulation of CD4+ T cells
closely mimics the recognition of specific Ag (35).
Although there may be subtle differences between Ag- and SA-activated
CD4+ T cells depending on the system used
(36, 37, 38), most studies find that the requirements of Ag
and SA activation are similar. Efficient SA stimulation of T cells uses
adhesion and coactivation molecule interactions shared by peptide Ag
stimulation, including CD4 signaling through
p56lck (39), CD28-B7 (40, 41), ICAM-1-LFA-1 (42), LFA-3-CD2
(43), CD40-CD40 ligand (44), CD49d
subunit-bearing integrins (45), and CD69 (46)
interactions. SA-stimulated cells show a dependence on IL-2
(47) and differential up-regulation of CD2, LFA-1, CD25,
CD28, CD69, and HLA-DR on stimulated cells (48). Thus, a
model system using DC presentation of SA to CD4+
T cells was employed as a physiologic representation of the site of the
linked processes of HIV replication and Ag-specific activation
(10, 12).
| Materials and Methods |
|---|
|
|
|---|
AIM V serum-free medium supplemented with glutamine (2 mM; Life
Technologies, Rockville, MD); GM-CSF, IL-4, and TNF-
(R&D Systems,
Minneapolis, MN); PGE3 (Cayman Chemical, Ann
Arbor, MI); Toxic shock syndrome toxin-1 (TSST-1) and
staphylococcal enterotoxin B SA (Sigma, St. Louis, MO); mAbs, CD16, and
CD56 (Accurate Chemical and Scientific, Westbury, NY); HLA-DR-FITC (41)
and HLA-DR-PE (Becton Dickinson, Franklin Lakes, NJ); CD83-PE
(49) and Ki67-FITC (Immunotech, Fullerton, CA); 2D7-FITC
or -PE, and 12G5-FITC or -PE (PharMingen, San Diego, CA); V
2-biotin
(Immunotech); and streptavidin-PerCP and streptavidin-FITC (Becton
Dickinson) were used. CD16 and CD56 magnetic beads were prepared by
incubating goat anti-mouse IgG beads (Dynal, Lake Success, NY) with
anti-CD16 and anti-CD56 mAb according to the manufacturers
instructions.
Culture system
DC were prepared as previously described (50) with
minor modifications. Briefly, PBMC were isolated from leukapheresis
packs obtained under an institutional review board-approved protocol
from healthy volunteers by Ficoll-Hypaque density gradient
centrifugation. Monocytes were purified from PBMC by discontinuous
Percoll gradient centrifugation. The low density fraction was depleted
of B, T, and, in certain experiments, NK cells using magnetic beads
specific for CD2, CD16, CD19, and CD56 according to the manufacturers
instructions (Dynal). The purified monocytes were cultured at
106 cells/ml in culture medium with GM-CSF (50
ng/ml) and IL-4 (100 ng/ml). TNF-
(1 ng/ml) and
PGE3 (500 nM) were added 1 day before use to
obtain mature, CD83+ (6090%) DC. B cells were
purified from PBMC using positive selection with CD19 beads (Dynal).
Autologous HLA-DR-, CD4+ T
cells were purified by negative selection first by incubating on human
Ig (10 µg/ml; Sigma)-coated plates for 1 h and then with
negative selection using magnetic beads specific for CD8, CD19, CD16,
CD56, and HLA-DR (Dynal). DC and CD4+ T cells
were cocultured, and TSST-1 SA (0.010.02 ng/ml) was added at the same
time.
Analyses
HIV Ba-L, UGO24, SF162, IIIB, TH026 (obtained from
AIDS Reference and Reagent Program), BL2 (51), 89.6
(52), and LL-7 (53) (obtained from Ronald
Collman, University of Pennsylvania, Philadelphia, PA; 520 ng of p24)
were added to 3-day-old DC/CD4+ T cell
(1/10)/SA-TSST-1 (0.01 ng/ml) cocultures (1020 x
106 cells). Sixteen to 48 h later, the
populations of Ag-specific (expressing V
2, which is bound by TSST-1)
and bystander (expressing other V
chains) activated cells were
purified from the bulk culture using flow cytometric sorting
(FACStarPlus; Becton Dickinson) with V
2-biotin
mAb followed by streptavidin-PE and HLA-DR-Fl staining. Ninety-seven to
99% purity of the V
2+ cells was obtained;
9498% purity of the V
2- cells was
achieved. DNA was isolated as described previously (21).
RNA was isolated from cell pellets using Quantum Prep Master Blaster
(Bio-Rad, Hercules, CA) and was reverse transcribed using the
Superscript kit (Life Technologies). HIV infection of
CD4+ T cells was determined by real-time
quantitative PCR (Perkin-Elmer 7700; Perkin-Elmer, Norwalk, CT) for gag
DNA and RNA and GAPDH DNA according to the manufacturers
instructions. The PCR primer sequences for HIV gag were SK38/39 and a
SK19 probe (54) modified with molecular beacon technology.
The PCR primers for GAPDH were: forward, 5'-GGTGGTCTCCTCTGACTTCAACA-3';
reverse, 5'CCAGCCACATACCAGGAAATG-3'; and a molecular beacon-based
probe with the sequence
5'-FAM-GCGAGCCTGGCATTGCCCTCAACGACCACGCTCGC-Dabcyl-3' (Integrated
DNA Technologies, Coralville, IA).
Intracellular p24 gag Ag expression was determined 18 days
postinfection by fixing cells in 2% paraformaldehyde for 30 min,
permeabilizing with 0.1% saponin for 30 min, and including 0.05%
saponin in the staining medium (PBS, 1% FCS, 1% normal mouse
serum, and 0.2% NaN3) in all further
staining steps. Cells were sequentially stained with p24 gag mAb
(183-H12-5C from the AIDS Reference and Reagent Program)
(55), goat anti-mouse IgG-FITC, 5% normal mouse
serum (Sigma) to block further mouse Ig binding, V
2-biotin,
and streptavidin-PerCP plus HLA-DR-PE. Cells were then analyzed on a
FACScan (Becton Dickinson).
Flow cytometry for chemokine receptor expression was performed by
staining cells with V
2-biotin in staining medium for 20 min on ice,
followed by washing twice and incubation with streptavidin-PerCP,
2D7-FITC, and 12G5-PE for 20 min on ice, followed by washing and
analysis on a FACScan.
Cell cycle analysis was performed on CD4+ T cells
with or without DC and SA after 3 days of culture. Cells were stained
with V
2-biotin, streptavidin-FITC, and HLA-DR-PE, fixed in 70%
ethanol for 30 min, and then incubated with actinomycin D (25 µg/ml;
Sigma) for 30 min. Actinomycin D was used to allow three-color
analysis. Cells were analyzed on a FACScan equipped with doublet
discrimination mode to ensure analysis of single cells.
| Results |
|---|
|
|
|---|
Resting CD4+ T cells
(HLA-DR-) were obtained by negative selection
and cultured with autologous DC with or without the SA TSST-1. Since
TSST-1 activates T cells bearing the V
2 TCR chain, cells were
stained for V
2 to distinguish the Ag-specific
(V
2+) and bystander
(V
2-) populations. After 3 days of culture,
HLA-DR-depleted CD4+ T cells demonstrated minimal
activation within both the V
2-positive and -negative populations
(Fig. 1
A). Coculture with
autologous DC (Fig. 1
B) increased activation of both
populations. Based on acquisition of HLA-DR expression, the addition of
TSST-1 resulted in near-complete activation of
V
2+ Ag-specific cells (upper right
quadrant of Fig. 1
C vs 1B) and a
substantial increase in V
2- bystander cell
activation (upper left quadrant of Fig. 1
C vs
1B). Similar bystander activation was observed for DC-T cell
cocultures activated with staphylococcal enterotoxin B (data not
shown). The level of bystander activation correlated in part with the
ratio of DC to CD4+ T cells (Fig. 1
D).
This suggested that some component of cell-to-cell contact was
necessary. If bystander activation of CD4+ T
cells in this coculture system was independent of cell-to-cell contact,
no decrease in bystander activation with low numbers of DC (1/50
CD4+ T cells) should be observed, since the
V
2+, CD4+ T cells
remained almost completely activated (Fig. 1
E).
|
-bearing cells or anergy and T cell depletion depending on the
culture conditions, we chose conditions designed to allow expansion
without depletion of V
2+,
CD4+ T cells (40). To confirm this,
we stained for V
2 expression 2 wk after TSST stimulation and found
no evidence of V
2 depletion. In addition, restimulation with TSST-1
2 wk after initial stimulation again resulted in activation of
V
2+ T cells, demonstrating that anergy had not
been induced. DC activation of CD4+ T cells in
the absence of exogenous Ag can be substantial (315%) especially in
the presence of FCS, which is known as the autologous MLR and can
support HIV replication (10, 12). This phenomenon probably
represents the presentation of Ags obtained by the DC before and during
preparation. The autologous MLR can be substantially, although not
completely, reduced by using serum-free medium for DC generation and T
cell stimulation, which was employed in all experiments in this
report. HIV replication occurs in both Ag-specific and bystander CD4+ T cells
To determine the populations of CD4+ T cells
responsible for HIV-1 replication during SA stimulation,
DC-CD4+ T cell cocultures were infected and
analyzed for infection by intracellular p24 gag expression over the
subsequent 18 days. The HIV-1 strains analyzed included CCR5-using
macrophage-tropic prototype (SF162, BAL) and primary isolates (LL-7,
BL2, TH026); a dual-tropic strain that uses both CCR5 and CXCR4 (89.6);
and CXCR4-using T-tropic prototype (IIIB) and primary (UGO24) isolates.
Both V
2+ Ag-specific and
V
2- bystander cells demonstrated p24 gag
staining (Fig. 2
and data not shown). The
frequency of p24+ expression was typically 2- to
3-fold higher among V
2+ cells than
V
2- bystander cells. However, as only
1020% (11.2% in the experiment shown) of the
CD4+ T cells in cultures with DC plus SA
expressed V
2, approximately two to four times as many bystander
cells were infected as Ag-specific cells. In the absence of SA,
DC-CD4+ T cell cocultures resulted in <1% p24
gag-positive cells, with equal frequencies in the
V
2+ and V
2-
populations.
|
2 expression (Fig. 3
2+ cells in the culture, while within each
experiment, all strains demonstrated a similar distribution of
infection in bystander vs Ag-specific cells. Thus, bystander cells
represented the majority of infected cells, while Ag-specific activated
cells were a more highly productive compartment.
|
2-,
CD4+ T cells. Role of activation in HIV infection of bystander cells
An analysis of bystander cells from DC-SA cocultures demonstrated
a portion of cells expressing HLA-DR, which was dramatically different
from the Ag-specific cells that had high levels of HLA-DR (Fig. 1
).
Therefore, we determined whether infection was restricted to the
HLA-DR+ subset of bystander activated
CD4+ T cells. V
2-
bystander cells in cultures of DC, CD4+ T cells,
and SA were analyzed for HLA-DR and p24 gag expression, which revealed
that cells lacking the activation marker HLA-DR had significant p24 gag
expression (Fig. 4
).
|
|
In addition to the ability of bystander cells to replicate HIV, a
relative inability of Ag-specific cells to support HIV growth was also
observed. In systems of mitogen-activated CD4+ T
cells, high levels of both M- and T-tropic HIV infection and nearly
complete depletion of CD4+ T cells with T-tropic
HIV are typically observed. In contrast, we detected no substantial
loss of V
2+ T cells over time after infection.
Furthermore, the percentage of Ag-specific cells expressing p24 gag was
relatively low even though the Ag-specific cells in our culture system
were highly activated (Fig. 1
, C and E),
proliferated with 10.4% of cells in S phase (Fig. 5
), and contained
the majority of cells incorporating
[3H]thymidine (data not shown). The observation
that highly activated and proliferating Ag-specific
CD4+ T cells were only modestly infected was,
thus, unexpected. To address a potential mechanism for the relative
protection from HIV infection, we analyzed the expression of the
chemokine receptor/HIV coreceptors CXCR4 and CCR5 on the two
populations of CD4+ T cells during an Ag-specific
response. Ag-specific cells demonstrated a significant, but not
complete, down-regulation of CXCR4 expression (Fig. 6
A). Over the next 10 days,
CXCR4 expression on the V
2+ cells never
approached the level of expression on the V
2-
(Fig. 6
B). CCR5 was similarly down-regulated on the
V
2+ cells, and its kinetics of reappearance
were delayed compared with the V
2- cells
(Fig. 6
, C and D). Thus, Ag-specific activation
down-regulated the expression of CCR5 and CXCR4, which probably
explained at least in part the relative resistance to infection.
|
| Discussion |
|---|
|
|
|---|
Model system for studying HIV replication in vitro
The initial identification of HIV as the causative agent of AIDS employed activated T cells (56). Further studies demonstrated that peripheral blood T cells, in the absence of activation, could not complete reverse transcription (1, 2, 3, 4, 5). The study of HIV replication in vitro requires the use of model systems. HIV replication (6, 7) and Ag-specific activation (reviewed in Ref. 57) primarily occur in the paracortical regions of lymphoid organs. Most systems used to study HIV replication in vitro employ PBMC. Unstimulated PBMC typically have <5% of CD4+ T cells expressing HLA-DR or other activation markers; after activation with mitogens or Abs against CD3 nearly all CD4+ T cells become activated. In a lymph node from an HIV-infected individual, up to 75% of the CD4+ T lymphocytes can express varying levels of HLA-DR, while T cells in lymphoid tissue from uninfected subjects have lower, but significant, levels of activation (reviewed in Ref. 58, 59). We and others have used DC activation of CD4+ T cells, with or without the addition of specific Ag, to model both Ag-specific activation and HIV replication based on the above observations (10, 11, 12, 13, 14, 50, 60, 61, 62).
Bystander cells during an Ag-specific response replicate HIV
During an Ag-specific immune response, multiple populations of
cells are activated, only a portion of which recognize the initiating
Ag and become highly activated. In this report we demonstrate that a
population of cells responsible for most of the HIV-replicating cells
were not directly activated by our model Ag. A minority portion of
these bystander cells did become activated, as measured by expression
of the activation markers HLA-DR, CD25, KI67, and CD69 (Fig. 1
and data
not shown), but did not produce higher levels of infection compared
with the activation-negative cells.
It is likely that we are studying multiple populations of bystander CD4+ T cells. One population may be CD4+ T cells that are actively bound via their TCR to an APC that is presenting a peptide Ag bound to MHC class II, but the interaction is not "strong enough" (T cell cannot up-regulate APC function via CD40L, TCR-peptide/MHC interaction too weak, other defective or inefficient interactions between CD28 and B7, LFA-1 and ICAM-3, or CD2 and CD58) to activate the cell (63). When this T cell bound to an APC encounters an efficiently occurring APC-T cell interaction, soluble mediators may be released that allow the poorly interacting T cell to become activated. In addition, transcomplementation of costimulation has been described where another cell that is not presenting Ag to the T cell can supply the necessary B7 interaction and lead to activation (64). A second population of CD4+ T cells that is bound to the APC in an Ag-independent, non-TCR manner has been described that is bound via adhesion or coactivation molecule interactions (11). In the setting of a strong Ag-specific response, they become activated in the absence of TCR signaling. A third population of bystander CD4+ T lymphocytes that occurs without cell to cell contact, through mediator release, also may be present.
The bystander cells in cultures of DC, CD4+ T cells, and SA contain the majority of infected cells as measured by p24 gag Ag expression and gag DNA content. The infected bystander cells express less HIV p24 gag protein and gag RNA per cell, demonstrating a lower production of HIV. The expression of p24 gag on bystander cells is independent of activation marker expression or proliferation, but is associated with some level of activation, because in the absence of DC and SA the cells are not infectable. The source of the activation includes both cytokines released and interactions between the bystander cells and DC, as preliminary studies separating bystander cells from DC-CD4+ T cell-SA cocultures by a membrane demonstrated reduced levels of HIV replication in the separated cells (D. Scales and D. Weissman, unpublished observations). Our data are in agreement with recent reports demonstrating that cytokines are capable of inducing HIV replication in unstimulated CD4+ T cells (65, 66). In addition, an in vivo correlate of our model has recently been reported where in situ examination for viral nucleic acids in tissue from HIV- and SIV-infected subjects has demonstrated low level infected CD4+ T cells lacking activation markers (67). As that study documented the presence of HIV in unactivated cells, but could not determine their activation state at infection, our data extend this observation by demonstrating the infection of nonreplicating CD4+ T cells.
Ag-specific cells are relatively protected from infection HIV
The regulation of the expression of CCR5 and CXCR4 expression on activated CD4+ T cells is complicated by the system used for analysis and the method of detection. Activation of PBMC with PHA was demonstrated to increase the expression of CXCR4. The expression of CXCR4 was found to be a balance between CD3 signaling, which increased CXCR4, and CD28 signaling, which decreased the expression (68). CCR5 was found to be present on a small subset of T cells, (CD26bright, CD45RO memory cells) and was down-regulated soon after PHA stimulation. Expression of CCR5 then slowly increased with the addition of IL-2 (69). Our current and previous studies of CXCR4 expression suggested that it was down-regulated with purification of PBMC and up-regulated after culture in the absence of stimulation (70). In this report Ag-specific activation led to a down-regulation of CCR5 and a significant, but incomplete, down-regulation of CXCR4 compared with bystander cells. Our results using an Ag-specific response in a model of the paracortical region of lymphoid organs are in agreement with other systems designed to mimic APC activation using anti-CD3 and anti-CD28 mAb, where a down-regulation of CCR5 and a protection from infection by R5 viruses was observed (71).
Current approaches to the treatment of HIV infection include the use of anti-viral agents, immune system modulators, and vaccines. With the improvement in virologic control of HIV, new therapies are being developed to target reservoirs of HIV infection. The findings presented in this report suggest that long-lived memory cell reservoirs (reviewed in Ref. 22) may be more dynamic and contain recently infected activation marker-negative CD4+ T cells. The data in this report support the finding of preserved common Ag-specific immune responses, but do not explain the early loss of HIV-specific responses. Further studies directed at defining the controls of bystander replication of HIV will be important as an alternate approach to inhibit viral production as well as for possibly enhancing Ag-specific and HIV-specific responses.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Drew Weissman, University of Pennsylvania, 536 Johnson Pavilion, Philadelphia, PA 19104. ![]()
3 Abbreviations used in this paper: DC, dendritic cell; SA, superantigen; CXCR4, CXC chemokine receptor 4; TSST-1; toxic shock syndrome toxin-1. ![]()
Received for publication October 2, 2000. Accepted for publication March 12, 2001.
| References |
|---|
|
|
|---|
irradiation. J. Immunol. 156:4476.[Abstract]
production. Eur. J. Immunol. 26:1430.[Medline]
-specific mechanism of bystander activation. Cell. Immunol. 161:158.[Medline]
4
7 integrin expression on human activated T lymphocytes. J. Immunol. 156:3668.[Abstract]
. J. Exp. Med. 179:1109.
ccr5 mutation. J. Virol. 71:3219.[Abstract]
This article has been cited by other articles:
![]() |
K. Fernando, H. Hu, H. Ni, J. A. Hoxie, and D. Weissman Vaccine-delivered HIV envelope inhibits CD4+ T-cell activation, a mechanism for poor HIV vaccine responses Blood, March 15, 2007; 109(6): 2538 - 2544. [Abstract] [Full Text] [PDF] |
||||
![]() |
O. T. Keppler, N. Tibroni, S. Venzke, S. Rauch, and O. T. Fackler Modulation of specific surface receptors and activation sensitization in primary resting CD4+ T lymphocytes by the Nef protein of HIV-1 J. Leukoc. Biol., March 1, 2006; 79(3): 616 - 627. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Choi, J. Walker, K. Talbert-Slagle, P. Wright, J. S. Pober, and L. Alexander Endothelial Cells Promote Human Immunodeficiency Virus Replication in Nondividing Memory T Cells via Nef-, Vpr-, and T-Cell Receptor-Dependent Activation of NFAT J. Virol., September 1, 2005; 79(17): 11194 - 11204. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Monie, R. P. Simmons, R. E. Nettles, T. L. Kieffer, Y. Zhou, H. Zhang, S. Karmon, R. Ingersoll, K. Chadwick, H. Zhang, et al. A Novel Assay Allows Genotyping of the Latent Reservoir for Human Immunodeficiency Virus Type 1 in the Resting CD4+ T Cells of Viremic Patients J. Virol., April 15, 2005; 79(8): 5185 - 5202. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Maranon, J.-F. Desoutter, G. Hoeffel, W. Cohen, D. Hanau, and A. Hosmalin Dendritic cells cross-present HIV antigens from live as well as apoptotic infected CD4+ T lymphocytes PNAS, April 20, 2004; 101(16): 6092 - 6097. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. L. Kinter, C. A. Umscheid, J. Arthos, C. Cicala, Y. Lin, R. Jackson, E. Donoghue, L. Ehler, J. Adelsberger, R. L. Rabin, et al. HIV Envelope Induces Virus Expression from Resting CD4+ T Cells Isolated from HIV-Infected Individuals in the Absence of Markers of Cellular Activation or Apoptosis J. Immunol., March 1, 2003; 170(5): 2449 - 2455. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Capodici, K. Kariko, and D. Weissman Inhibition of HIV-1 Infection by Small Interfering RNA-Mediated RNA Interference J. Immunol., November 1, 2002; 169(9): 5196 - 5201. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Alfano, N. Sidenius, B. Panzeri, F. Blasi, and G. Poli Urokinase-urokinase receptor interaction mediates an inhibitory signal for HIV-1 replication PNAS, June 25, 2002; 99(13): 8862 - 8867. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Ni, J. Capodici, G. Cannon, D. Communi, J.-M. Boeynaems, K. Kariko, and D. Weissman Extracellular mRNA Induces Dendritic Cell Activation by Stimulating Tumor Necrosis Factor-alpha Secretion and Signaling through a Nucleotide Receptor J. Biol. Chem., April 5, 2002; 277(15): 12689 - 12696. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |