26 • JID 2003:187 (1 January) • Barron et al.
M A J O R A R T I C L E
Influence of Plasma Viremia on Defects
in Number and Immunophenotype of Blood
Dendritic Cell Subsets in Human
Immunodeficiency Virus 1–Infected Individuals
Michelle A. Barron,1Naomi Blyveis,2Brent E. Palmer,1Samantha MaWhinney,3and Cara C. Wilson1,2
University of Colorado Health Sciences Center, Denver
1Infectious Diseases and
2Clinical Immunology, Department of Medicine, and
3Department of Preventive Medicine and Biometrics,
Dendritic cells (DCs) are postulated to be involved in transmission of human immunodeficiency virus (HIV)
type 1 to T cells and in stimulation of HIV-1–specific cell-mediated immunity. BloodDCshavebeencategorized
as myeloid (mDC) and plasmacytoid (pDC) subsets, on the basis of differences in phenotype and function.
Blood DC subset numbers and expression of costimulatory molecules and HIV-1 coreceptors on DCs were
measured in the blood of treated and untreated HIV-1–infected subjects and uninfected control subjects.
Absolute numbers of mDCs and pDCs were lower in HIV-1–infected subjects than in control subjects, most
significantly in those with active HIV-1 replication. Increased surface expression of costimulatory molecules
was observed on both DC subsets in subjects with HIV-1 viremia. Highly active antiretroviral therapy sup-
pression of plasma viremia resulted in increases in blood DC numbers and decreases in DC costimulatory
molecule expression. These findings further define the impact of HIV-1 replication on blood DC subsets in
Dendritic cells (DCs) are potent antigen (Ag)–
presenting cells (APCs) that are capable of activating
and expanding naive, Ag-specific CD4?and CD8?T
lymphocytes . There are several lines of evidence
Received 17 April 2002; revised 10 September 2002; electronically published
13 December 2002.
Presented in part: 8th Conference on Retroviruses and Opportunistic Infections,
Chicago, 4–8 February 2001 (abstract B15e); 9th Conference on Retroviruses and
Opportunistic Infections, Seattle, 24–28 February 2002 (abstract B16e).
The study was approved by the Colorado Multiple Institutional Review Board
at the University of Colorado Health Sciences Center (UCHSC; Denver), and
informed consent was obtained from all study participants.Humanexperimentation
guidelines of the US Department of Health and Human Services and those of the
UCHSC were followed in the conduct of clinical research.
Financial support: National Institutes of Health (grants KO8 AI-01459 and PO1
AI-43664 to C.C.W. and AI-07447 5T32 and AmFAR 2-5-20188 to M.A.B.)
Reprints or correspondence: Dr. Cara C. Wilson, University of Colorado Health
Sciences Center, Div. of Clinical Immunology, Dept. of Medicine, 4200 E. Ninth
Ave., Box B164, Denver, CO 80262 (email@example.com).
The Journal of Infectious Diseases
? 2003 by the Infectious Diseases Society of America. All rights reserved.
suggesting that DCs may be involved in the transmis-
sion and immunopathogenesis of HIV-1 . In addi-
tion to the impact that DCs may have on HIV-1 disease
as a result of their ability to transmit HIV-1 to T cells
[3, 4], it has been suggested that DCs themselves might
be adversely effected by replicating HIV-1. Several in-
vestigators have reported decreased numbers of epi-
dermal or blood DCs in HIV-1–infected persons with
advanced disease [4–9]. In addition, there is evidence
that DCs freshly isolated from HIV-1–infected individ-
uals or DCs infected with HIV-1 in vitro are dysfunc-
tional in their ability to induce a primary immune re-
sponse. This dysfunction may result from direct effects
of viral infection, such as down-regulation of major
histocompatibility complex molecules , or indirect
effects, such as exposure to a cytokineenvironmentthat
is less conducive to the generation of cellular immune
Blood DCs have been traditionally characterized by
the absence of leukocyte lineage (Lin)–specific Ags and
by guest on November 5, 2015
Plasma Viremia and Blood Dendritic Cells • JID 2003:187 (1 January) • 27
by surface expression of HLA-DR and CD4 [10, 11]. They have
been subcategorized into myeloid DCs (mDCs), denoted as
Lin?HLA-DR?CD11c?, and plasmacytoid DCs (pDCs), de-
noted as Lin?HLA-DR?CD11c?IL-3RHi[12, 13]. In addition to
phenotypic differences, different functional attributes, includ-
ing the ability to stimulate specific T helper subsets, were
thought to be associated with the individual subsets on the
basis of their lineage [12, 14]. However, it has recently been
suggested that the DC lineage per se may not actually be the
basis of the Th1/Th2 polarization observed.Theactivationstate
of the DC, the type of activation signal it encounters, and the
cytokine environment all have been postulated to contribute
to T helper polarization by DCs [15–18].
The role of both blood DC subtypes in HIV-1 immuno-
pathogenesis remains an area of intense research. A decrease
in the proportion or absolute number of blood DCs expressing
CD11c?(mDCs) has been observed in a number of cohorts of
HIV-1–infected donors [5, 6, 9, 19]. However, a recent article
by Chehimi et al.  that examined levels of blood DCsubsets
in both treated and untreated HIV-1–infected subjectsreported
to that of uninfected control subjects. In that study, a decrease
in mDC number relative to that of control subjects was only
noted in HIV-1–infected subjects with active viral replication.
It was recently reported that pDCs might be the natural
interferon (IFN)–a/b–producing cells (IPCs) of the immune
system [21, 22] and may serve as a link between innate and
adaptive immune systems. Therefore, the role of this blood DC
subset in HIV-1 disease is of great interest. In a study that
examined IPCs in HIV-1–infected patients, IPC numbers were
decreased in patients with AIDS, and a negative correlationwas
found between IPC number and plasmaHIV-1RNAlevels.
Others have shown a similar decrease in the number of pDCs
[5, 6, 20] and decreased IFN-a production with viral stimu-
lation [6, 20] in patients with HIV-1 infection.
Distilling the findings of all these recent reports into a uni-
fying hypothesis about the fate and function of blood DCs in
the setting of HIV-1 infection has been difficult because of the
diversity of subject cohorts studied, including subjects who
differ in both stage of disease and treatment status. In addition,
the role that blood DC defects play in mediating the observed
cellular immune dysfunction characteristic of progressive HIV-
1 infection, and the degree of reversal in DC defects after in-
itiating highly active antiretroviral therapy (HAART), has yet
to be fully elucidated. Blood DCs and their progeny are likely
to play a critical role in priming T cell responses to HIV-1 and
opportunistic pathogens in vivo. An understanding of the im-
pact of HIV-1 replication on the numbers of circulating blood
DCs, their maturation state, and theirabilitytopresentantigens
to naive and memory T cells is likely to be important in the
development of vaccines and immune-based therapiesforHIV-
1 infection. In the present study, we sought to further char-
acterize the defects in blood DC subsets observed during HIV-
1 infection and to assess the effect of HIV-1 plasma viremia
on blood DC number and phenotype. In addition, we sought
to determine whether a relationship existed betweencirculating
blood DCs and attributes of T cell function in the setting of
treated or untreated HIV-1 infection.
MATERIALS AND METHODS
received their care through the University of Colorado Health
Sciences Center (UCHSC) Infectious Diseases Group Practice
were studied. For comparison, 20 HIV-1–seronegative control
subjects were also studied.
Complete blood counts.
Complete blood count data were
obtained from heparinized blood using an ADVIA 120 (Bay-
er Diagnostics) or from EDTA lavender top vacutainer tubes
(Becton-Dickinson [BD]) using a CELL-DYN 4000 (Abbot
Peripheral blood mononuclear cells
(PBMC) were isolated from heparinized blood using standard
Ficoll-Paque (Pharmacia Biotech AB) density-gradient centri-
fugation. Three-color flow cytometric analysis was performed
on freshly-isolated PBMC stained with the following panel of
monoclonal antibodies (MAbs): fluorescein isothiocyanate
(FITC)–labeled anti-Lineage (Lin?) panel (CD3/CD14/CD16/
CD19/ CD20/CD56) MAb (BD), FITC-labeledanti-CD34MAb
(BD), peridinin chlorophyll protein–labeled anti–HLA-DR
(BD), and either phycoerythrin (PE)–labeled anti-CD11c
(PharMingen [Ph]), -CD123 (IL3R) (BD), -CD40 (Ancell),
-CD4 (Ph), -CD86 (Ancell), -CXCR4 (Ph), or -CCR5 (Ph)
MAbs. Four-color analysis wasperformedonfreshPBMCusing
FITC-labeled anti-Lin anti-CD34 MAbs, tricolor-labeled anti–
HLA-DR MAbs (Caltag), allophycocyanin conjugate–labeled
anti-CD11c (BD), or biotinylatedanti-CD123(IL-3R)(Ph)that
was then labeled with streptavidin-conjugated APC (Ph) per
the manufacturer’s instructions, and either PE-labeled anti-
CD40, -CD4, -CD86, -CXCR4, or -CCR5 MAbs. Cells were
also labeled with appropriate isotype control antibodies in each
experiment. After staining for 30 min with MAbs, cells were
washed, resuspended in fixation buffer (PBS plus 1% parafor-
maldehyde), and stored at 4?C in the dark before analysis.
Three- and 4-color FACS analysis was performed using a FAC-
Scan cytofluorometer (BD Immunocytometry Systems). For
each sample, 50,000–200,000 events were acquired and gated
on Lin?CD34?/HLA-DR?expression, and a scatter gate was
designed to include only viable leukocytes. The values for per-
centage of positive cells or the mean fluorescence intensity
(MFI) of all gated cells reported reflect subtraction of the iso-
type control values. DC subsets were defined as follows: mDCs
Forty-two HIV-1–infected subjects who
by guest on November 5, 2015
28 • JID 2003:187 (1 January) • Barron et al.
Both myeloid DCs (mDCs; A) and plasmacytoid DCs (pDCs; B) nos. correlated positively with peripheral CD4?T cell counts in HIV-1–infected subjects.
Nos. of CD11c?mDCs and IL-3RHi?pDCs in highly active antiretroviral therapy (HAART)–suppressed (HS), HAART failure (HF), and HAART-naive (HN)
subjects were decreased (C and D), compared with uninfected control subjects (HIV?). Bars represent median values (results summarized in table 2).
Blood dendritic cell (DC) subsets serve as markers of immune status in human immunodeficiency virus (HIV) type 1–infected individuals.
Table 1. Human immunodeficiency virus (HIV) type 1–infected donor characteristics.
(n p 14)
(n p 10)
(n p 10)
CD4?T cell count, cells/mL
HIV-1 RNA level, copies/mL
WBC count, ?109cells/L
11,600 (3011–1.612 ? 106)
48,450 (1985–1.076 ? 106)
Data are median (range). HAART, highly active antiretroviral therapy; WBC, white blood cell.
were defined as Lin?CD34?HLA-DR?CD11c?, and pDCs were
defined as Lin?CD34?HLA-DR?CD11c?IL-3RHi. Absolute DC
count was derived by using the percentage of cells in relation
to the mononuclear fraction determined by the automatic dif-
ferential blood count and is expressed as number of cells per
milliliter of blood.
Lymphoproliferative assay (LPA).
lated by Ficoll density centrifugation, resuspended at
cells/mL in RPMI 1640 medium with 10% human AB serum
(Sigma), and 100 mL of cells was added to 96-well plates that
contained 100 mL of HIV-1 p24 and p66 baculovirus-expressed
recombinant proteins (NY5 and IIIB strains, respectively; Pro-
tein Sciences; final concentration, 1 mg/mL). Phytohemagglu-
tinin (5 mg/mL; Sigma) and whole candida protein (10 mg/mL;
Greer) were used as positive controls in each assay. Cells were
incubated at 37?C in a humidified 5% CO2atmosphere for 6
days. Plates were pulsed with 1 mCi/well of tritiated thymidine
Fresh PBMC were iso-
for 6 h, cells were harvested, and radioactivity was counted on
a b-counter (Packard).
All statistical analyses assumed a 2-
sided significance level of .05. Because of small sample sizes,
nonparametric statistics were used. The Mann-Whitney U test
was used for primary comparisonsbetweenHIV-1–infectedand
uninfected control subjects, Wilcoxon signed-rank tests were
used for within group comparisons, and Spearman’s r wasused
to describe correlations;
r 1 0.3
be clinically significant. Overall P values for secondary com-
parisons between the control group and HIV-1–infected sub-
groups used a Kruskal-Wallis test (a nonparametric analysis of
variance), and, given an overall
to the control group were conducted using Dunn’s multiple
comparison adjustment. LPA data were analyzed using uni-
variate logistic regression. Data analyses were performed with
orwas considered to
r ! ?0.3
, pairwise comparisonsP ! .05
by guest on November 5, 2015
Plasma Viremia and Blood Dendritic Cells • JID 2003:187 (1 January) • 29
(HIV) type 1–infected subgroups.
Summary of 3-color flow cytometric analysis comparing uninfected control subjects with human immunodeficiency virus
Parameter Control group
HAART suppressed HAART failure
No. of mDCs, cells/mL
mDCs in total PBMC, %
No. of pDCs, cells/mL
pDCs in total PBMC, %
CD4?DCs in the total PBMC
No. of Lin?HLA-DR?cells,
Non-mDC/pDC in the
81.26 (61.88–92.94) 68.89b(16.59–78.36)34.44b(8.88–83.68)45.66b(7.53–81.97).001
20.14 (10.26–37.25) 15.42 (6.63–27.31)11.87b(3.00–15.87) 15.83 (6.33–44.04).009
20.00 (3.70–38.67)31.53 (10.00–70.18) 32.00 (0–57.14)42.99c(13.85–82.61).022
aKruskal-Wallis test (pairwise comparisons to the control group were conducted for
Data are median (range). HAART, highly active antiretroviral therapy; mDC, myeloiddendriticcells;MFI,meanfluorescenceintensity;pDC,plasmacytoid
using Dunn’s multiple comparison adjustment).P !.05
GraphPad Prism (version 3.00), Splus (Insightful), and SAS
(SAS Institute) software.
Blood DC subtypes are decreased in the blood of HIV-1–in-
fected individuals and serve as markers of immune status.
Using 3-color flow cytometry, Lin?HLA-DR?cells from 34
HIV-1–infected and 20 uninfected subjects were analyzed for
surface expression of CD11c (mDCs), CD123 (pDCs), cos-
timulatory molecules (CD40 and CD86), and the HIV-1 co-
receptor CD4. The HIV-1–infected patients evaluated had a
wide range of peripheral CD4?T cell counts and plasma HIV-
1 RNA levels and included those receiving HAART, as well
as untreated subjects. HIV-1–infected subjects as a group dis-
played decreases in blood mDC and pDC subsets relative to
uninfected control subjects, both when DCs were measured
as a percentage of total PBMC (median mDCs: control sub-
jects, 0.41%; HIV-1–infected subjects, 0.23%;
Mann-Whitney U test; median pDCs: control subjects,0.30%;
HIV-1–infected subjects, 0.15%;
test), and when measured as absolute numbers in the blood
(median mDC count: control subjects, 8070 cells/mL; HIV-
1–infected subjects, 5505 cells/mL;
U test; and median pDC count: control subjects, 5970 cells/
mL; HIV-1–infected subjects, 3180 cells/mL;
Whitney U test). When the relationship between blood DC
,P p .002
, Mann-Whitney UP p .001
, Mann-WhitneyP p .001
, Mann-P p .009
numbers and peripheral CD4?T cell count was evaluated in
HIV-1–infected subjects, absolute numbers of both mDCsand
pDCs were found to correlate positively with peripheralCD4?
T cell counts (figure 1A and 1B).
To better define the effect of plasma viremia on DC phe-
notype and function, subsequent analyses were performed with
the 34 HIV-1–infected subjects separated into 3 groups on the
basis of their treatment status and plasma HIV-1 RNA levels.
HAART-suppressed (HS) subjects were defined as those re-
ceiving combination antiretroviral therapy, with plasma HIV-
1 RNA level !400 copies/mL for ?6 consecutive months.
HAART-failure (HF) subjects were receiving combination an-
tiretroviral therapy without suppression of viral replication
(HIV-1 RNA level 12000 copies/mL) for ?6 consecutive
months. HAART-naive (HN) subjects had never received an-
plasma viremia (minimum plasma HIV-1 RNA level, ?2000
copies/mL). The clinical, immunologic, and virological char-
acteristics of the HIV-1–infected subjects are shown in table 1.
The median duration of HAART therapy was 26 months in
the HS group versus 31 months in the HF group. White blood
cells/L) and control subjects (median,
cells/L) were not statistically different(
The ratio of mDCs to pDCs in the blood was not signifi-
cantly different between the uninfected control group and the
P p .09
by guest on November 5, 2015
30 • JID 2003:187 (1 January) • Barron et al.
in uninfected control subjects (human immunodeficiency virus (HIV)–negative [HIV?]), highly active antiretroviral therapy (HAART)–suppressed (HS),
HAART failure (HF), and HAART-naive (HN) subjects. C, Correlation between plasma HIV-1 RNA levels and DC expression of CD86. D, Percentage of
Lin?HLA-DR?blood DCs expressing CD4 for each subject cohort. E, Relationship between DC expression of CD4 and HIV-1 RNA levels. Bars in each
graph represent median values. Results are summarized in table 2.
Blood dendritic cell (DC) expression of costimulatory molecules and CD4. CD86 (A) and CD40 (B) expression on Lin?HLA-DR?blood DCs
HIV-1–infected groups or among the HS, HF, and HN groups
(data not shown), with an average mDC:pDC ratio of 1.39
found across groups. All 3 HIV-1–infected groups had ab-
solute mDC and pDC numbers lower than those of control
subjects, but only those HIV-1–infected subjects with active
viral replication (HF and HN groups) had median DC num-
bers that were statistically significantly lower than those of
control subjects (figure 1C and 1D; table 2). Of note, both
HIV-1–infected groups with active viral replication also had
lower median counts of both mDC and pDC subtypes than
did the HS group (table 2). When relationships between ab-
solute numbers of DC subsets in the blood and plasma HIV-
1 RNA levels were evaluated in HIV-1–infected subjects, sta-
tistically significant direct correlations between mDCs or
pDCs and plasma HIV-1 RNA levels were not found (mDC:
Spearman’s r, ?0.314;P p .071
; data not shown).P p .127
Effect of HIV-1 plasma viremia on blood DC expression of
costimulatory molecules and CD4.
; pDC: Spearman’s r, ?0.267;
As DCs are activated to
mature, their surface expression of the costimulatorymolecules
CD86 and CD40 tends to increase . Expression of these
surface markers on Lin?HLA-DR?blood DCs was found to
differ among the subject groups tested (figure 2A and 2B; table
2). The median surface expression of CD86 was higher on DCs
from HIV-1–infected subjects with plasma viremia (HF and
HN groups), relative to that of DCs from the HS subjects and
uninfected control subjects. DC expression of CD40 was also
higher in viremic subjects than in uninfected control subjects
or HS subjects, but this difference only reached statistical sig-
nificance for the HF group (table 2). In addition, a strong
positive correlation was found between CD86 expression on
DCs and plasma HIV-1 RNA copies/mL (figure 2C), and a
negative correlation was found between peripheral CD4?T cell
counts and CD86 expression on DCs (Spearman’s r, ?0.496;
; data not shown). Significant correlations between
P p .004
blood DC expression of CD40 and either plasma HIV-1 RNA
levels (Spearman’s r, 0.242;
P p .168
) or peripheral CD4?T
by guest on November 5, 2015
Plasma Viremia and Blood Dendritic Cells • JID 2003:187 (1 January) • 31
then the gates were set on Lin?HLA-DR?CD11c?cells (myeloid DCs [mDCs], R3) (B), with subsequent expression of CD86 (C), CD40 (D), CD4 (E),
CXCR4 (F), and CCR5 (G) analyzed on the mDCs in the R3 gate. H–M, A similar analysis was performed on Lin?HLA-DR?IL-3RHi?cells (plasmacytoid
DCs [pDC], R4). The shaded histogram represents the staining with specific monoclonal antibody, and the outlined histogram represents the respective
isotype antigen-presenting cell (APC) or phycoerythrin (PE) control subjects.
Example of blood dendritic cell (DC) subset analysis by 4-color flow cytometry. First, a gate was drawn around Lin?HLA-DR?cells (A);
and costimulatory molecules was examined on myeloid DCs (mDCs) and plasmacytoid DCs (pDCs) using 4-color flow cytometry in uninfected subjects
and HIV-1–infected subjects with HIV-1 RNA levels 12000 copies/mL. A significant difference between CD4 (A) and CXCR4 (B) expression was not
found (see text). However, mDC expression of CCR5 (C) was significantly increased in the viremic patients (
mean fluorescence intensity (MFI) was significantly higher on DCs from HIV-1–infected individuals (mDC,
U test) (D). DC expression of CD86 was also higher in HIV-positive donors, but these differences were not statistically significant (E). *
Four-color flow cytometric analysis of dendritic cell (DC) subtypes. Expression of human immunodeficiency virus (HIV) type 1 coreceptors
, Mann-Whitney U test). CD40
; pDC,P p .0721
P p .028
P p .029 , Mann-Whitney
P ! .05
cell counts (Spearman’s r, ?0.129;
served (data not shown).
Blood DCs have been shown to express the surface marker
CD4, a molecule that also serves as a coreceptor for HIV-1.
The surface expression of CD4 was measured on Lin?HLA-
DR?blood DCs in the different HIV-1–infected cohorts. The
percentage of blood DCs expressing CD4 was significantly de-
) were not ob-P p .467
creased in all the HIV-1–infected groups, compared with un-
infected control subjects (figure 2D; table 2). In addition, the
percentage of Lin?HLA-DR?CD4?DCs in total PBMC was also
significantly decreased in all HIV-1–infected subgroups (table
2). A significant inverse correlation was found between plasma
HIV-1 RNA levels and percentageof bloodDCsexpressingCD4
by guest on November 5, 2015
32 • JID 2003:187 (1 January) • Barron et al.
after institution of highly active antiretroviral therapy (HAART). CD4?T cell counts, HIV-1 RNA levels, blood myeloid DC (mDC) and plasmacytoid DC
(pDC) nos., and Lin?HLA-DR?DC expression of CD86 and CD40 were measured before and 6 months (mo) after starting HAART. For technical reasons,
analysis of CD86 expression on DCs was performed on only 4 of 5 original subjects. MFI, mean fluorescence intensity.
Longitudinal analysis of dendritic cell (DC) subsets and immunophenotype in human immunodeficiency virus (HIV) type 1–infectedsubjects
Blood DC subtypes have differential expression of surface
costimulatory molecules and HIV-1 coreceptors.
amining Lin?HLA-DR?populations using 3-color flow cyto-
metry, it was noted that ∼20% of the cells in this population
in uninfected control subjects were neither CD11c?nor IL3-
RHi?, which suggests that they did not fall into either mDC or
pDC subsets (table 2). The percentage of non-mDC/non-pDC
cells in the Lin?HLA-DR?populations of HIV-1–infected sub-
jects was generally greater than that of uninfected control sub-
jects (table 2). To assess the respective expression of HIV-1
coreceptors and DC costimulatory molecules directlyonmDCs
or pDCs within the Lin-HLA-DR?population, 4-color flow
cytometric analysis was performed on blood obtained from 8
additional HIV-1–infected subjects with HIV-1 plasma viremia
and 8 uninfected control subjects. The 8 HIV-1–infected sub-
jects studied had a median peripheral CD4?T cell count of
183.5 cells/mL (range, 26–690 cells/mL) and a median HIV-1
RNA level of 47,870 copies/mL (range, 17,400–750,000 copies/
mL) and were selected on the basis of plasma HIV-1 RNAlevels
?2000 copies/mL for ?6 consecutive months without regard
to their HAART treatment status.
We first compared the expression of CD40, CD86, and CD4
and the additional HIV-1 coreceptors CXCR4 and CCR5 on
the 2 DC subtypes in the blood of individual seronegativecon-
trol subjects (figure 3). In these donors, pDCs expressed higher
levels of HIV-1 coreceptors (CD4, CXCR4, and CCR5) than
did mDCs, whereas mDCs expressed higher levels of DC cos-
timulatory or maturation markers (CD86 and CD40) (figures
3 and 4). In general, these trends in surface phenotype of mDC
and pDCs were recapitulated in HIV-1–infected subjects, and
these findings confirm those of Donaghy et al. .
We next compared the respective mDC and pDC expression
of CD4, CXCR4, CCR5, CD40, and CD86 between uninfected
control subjects and HIV-1–infected, viremic subjects (figure
4A–E). CD4 has been reported to be highly expressed on blood
DC subsets and to decrease on mDCs with culture [10, 24].
by guest on November 5, 2015
Plasma Viremia and Blood Dendritic Cells • JID 2003:187 (1 January) • 33
Table 3. Summary of logistic univariate regression analysis of lymphoproliferative responses.
OR (CI)P OR (CI)P OR (CI)P OR (CI)P
CD4?T cell count, cells/mL
Log10HIV-1 RNA level, copies/mL
No. of mDCs, cells/mL
No. of pDCs, cells/mL
CD4?T cell counts, or plasma human immunodeficiency virus (HIV) type 1 RNA levels predicted a positive lymphoproliferative response (stimulation index, ?5)
to recall or HIV-1 antigens in 30 HIV-1–infected subjects. Odds ratios (ORs) are expressed as change per 50 cells/mL CD4?T cells, 0.5 log10plasma HIV-1 RNA,
1000 cells/mL mDCs, and 1000 cells/mL pDCs. Statistically significant values are in boldface type. CI, confidence interval; CMV, cytomegalovirus.
A logistic-regression analysis was performed to determine whether myeloid dendritic cell (mDC) orplasmacytoiddendriticcell(pDC)subsets,peripheral
Although the percentage of CD4?mDCs and CD4?pDCs was
slightly lower in HIV-1–infected, viremic subjects relative to
uninfected control subjects, these differences did not reach sta-
tistical significance (figure 4A; percentage of CD4?mDCs,
; percentage of CD4?pDCs,P p .279
U test for both comparisons). These results suggest that the
marked decrease in Lin?HLA-DR?cells expressing CD4 in
HIV-1–infected subjects using 3-color flow cytometry likely
reflects decreases in absolute numbers of pDCs and mDCs in
the blood of those subjects rather than a true decrease in the
surface expression of CD4 on those cells.
DC expression of CXCR4 on mDCs and pDCs was slightly
higher in HIV-1–infected subjects than in control subjects,but,
again, these differences were not statistically significant (figure
4B; percentage of CXCR4?mDCs,
CXCR4?pDCs, , Mann-Whitney U test for both com-P p .645
parisons). Expression of CCR5 on mDCs, expressed at very low
levels in uninfected subjects, was significantlyincreasedinHIV-
1–infected donors (figure 4C). The expression of CCR5 on
pDCs did not significantly differ between the donor groups.
Higher expression of CD40 was observed on both mDCs and
pDCs in HIV-1–infected subjects with plasma viremia, relative
to uninfected control subjects (figure4D).Althoughthemedian
CD86 expression on both DC subsets was slightly higher in
viremic subjects than in control subjects, these differenceswere
not statistically significant, perhaps because of the small sample
size (figure 4E). These results are consistent with the findings
of increased CD40 and CD86 expression on the Lin?HLA-DR?
DC population in HIV-1–infected viremic subjects using 3-
color flow cytometry, so they likely represent true immuno-
phenotypic changes related to HIV-1 replication.
Longitudinal analysis of blood DCs in antiretroviral-naive
HIV-1–infected subjects after starting HAART.
the effect of in vivo virus suppression on blood DC number
and phenotype, blood DCs were evaluated by flow cytometry
in 5 antiretroviral therapy–naive, HIV-1–infected subjects
from the initial analysis over a period of 6 months after in-
stitution of HAART. Of the 5 subjects studied, 2 had baseline
peripheral CD4?T cells counts 1200 cells/mL (range, 251–368
,Mann-WhitneyP p .279
; percentage ofP p .383
cells/mL), and 3 had CD4?T cells counts !200 cells/mL(range,
57–110 cells/mL). All 5 subjects achieved good viral suppres-
sion during HAART, with undetectable plasma HIV-1 RNA
levels (!200 copies/mL) measured in all subjects after 6
months on therapy (figure 5A and 5B). All the subjects had
an increase in their peripheral CD4?T cells counts after the
initiation of therapy.
Blood DC subsets, mDCs and pDCs, were measured before
therapy and after 6 months of receiving HAART (figure 5C
and 5D). The median baseline (pre-HAART) number of mDCs
in the HIV-1–infected subjects was 5390 cells/mL (range,
850–17,940 cells/mL), and the baseline number of pDCs was
2770 cells/mL (range, 850–10,260 cells/mL). Some degree of
recovery of both mDCs and pDCs was observed after 6 months
of virus suppression, with a more marked and consistent in-
crease with therapy noted in the pDC subset than in the mDC
subset (mean change in mDC number, 530 cells/mL; mean
change in pDC number, 4510 cells/mL), althoughthesechanges
relative to pre-HAART values did not reach statistical signifi-
cance (mDC,; pDC,P p .625P p .125
test). Despite the increase in DC numbers from baseline after
starting therapy, posttreatment pDC and mDC numbers re-
mained lower than the median numbers measured in control
Surface expression of CD40 and CD86 on Lin?HLA-DR?
blood DCs was also evaluated before and after initiation of
HAART in this cohort of subjects (figure 5E and 5F). Four
subjects were evaluated with respect to CD86 expression of
DCs, and all 4 displayed some decrease in CD86 expression on
Lin?HLA-DR?blood DCs after virus suppression (figure 5E).
CD86 expression on Lin?HLA-DR?blood DCs decreased from
a median MFI of 35.85 before therapy to 20.96 at 6 months
of HAART, resulting in values after virus suppression similar
to those observed in uninfected control subjects (medianCD86
MFI, 19.45; table 2). A decrease in DC expression of CD40 was
also observed in 3 of 5 subjects tested after 6 months of therapy,
with the median MFI decreasing from 89.53 to 61.40 (figure
5F). In uninfected control subjects, the median CD40 MFI was
52.96 (table 2).
, Wilcoxon signed-rank
by guest on November 5, 2015
34 • JID 2003:187 (1 January) • Barron et al.
Relationship of blood DC numbers to T cell lymphoproli-
Because DCs are potent APCs important in both
priming and maintaining Ag-specific T lymphocyte responses
in vivo, we reasoned that a relationship might exist between
the number of circulating DCs and measured T cell lympho-
were performed using PBMC from 30 of the initial 34 HIV-
1–infected study subjects. To determine whether numbers of
mDC or pDC subsets or other surrogate markers of HIV-1
disease, such as peripheral CD4?T cell counts andplasmaHIV-
1 RNA levels, were predictive of the probability of a positive
proliferative response to recall or HIV-1 antigens in the LPA
(defined here as a stimulation index ?5), univariate logistic-
regression analyses were carried out (table 3). Peripheral CD4?
T cell counts and plasma HIV-1 RNA levels were statistically
predictive of the probability of a positive LPA response to HIV-
1 p24, HIV-1 p66, and cytomegalovirus (CMV) antigens. The
absolute numbers of blood mDCs or pDCs were not predictive
of a response to either HIV-1 p24 or p66 Ags. However, mDC
numbers were predictive of a CMV-specific LPA response, and
the predictive value of pDC numbers for a CMV-specific re-
sponse trended toward significance. None of the parameters
measured was predictive of an LPA response to Candida Ag.
When a direct relationship between blood DC numbers and
magnitude of lymphoproliferative responses was assessed, sig-
nificant positive correlations were observed between blood
mDC numbers and CMV LPA responses (Spearman’s r, 0.449;
) and between blood pDC numbers and LPAresponsesP p .010
directed against HIV-1 p24 (Spearman’s r, 0.407;
data not shown.).
;P p .021
To date, the role of blood DC subtypes in the immunopath-
ogenesis of HIV-1 infection has not been well defined. In our
analysis, we found that absolute mDC and pDC numbers in
in total PBMCs, were significantly decreased in HIV-1–infected
individuals, compared with control subjects. This deficit was
more notable in subjects with detectable plasma HIV-1 RNA
levels. However, a direct correlation between DC numbers or
the percentage of CD11c?or CD11c?IL-3RHI?DCs and plasma
HIV-1 RNA levels was not found in our study, either when
analyzing all HIV-1–infected subjects or only those with de-
tectable plasma viremia.
The HIV-1–infected subjects in the present study were
grouped byclinical treatment statusintoHS,HF,orHNcohorts
for the purpose of analysis. We observed that the number of
mDCs and pDCs was markedly decreased in the 2 cohorts with
active plasma viremia (HF and HN subjects), compared with
uninfected control subjects, whereas the number of mDCs and
pDCs in subjects with effective virus suppression while receiv-
ing HAART (HS) were lower, but not significantly lower, than
those in control subjects. It remains unclear whether the de-
crease in DC numbers we observed in the blood of subjects
with active HIV-1 replication is a result of DC death and de-
letion, migration to lymph nodes where active viral replication
is occurring, or a down-regulation of traditional DC surface
markers. Regardless of the mechanism, our longitudinal data
from HAART-treated subjects suggest that these defects may,
in part, be reversible.
It has been proposed that pDCs play a role in controlling
plasma viremia through their production of antiviral type I
interferons . As such, pDCs are likely important APCs that
bridge the gap between innate and adaptive immunity and also
exert an antiviral effect. Their deletion in the blood of HIV-
1–infected subjects has been associatedwithdiseaseprogression
and development of opportunistic infections . Chehimi et
al.  found a sustained decrease in pDC frequency in un-
treated subjects that was not normalized in HAART-treated,
virus-suppressed subjects. However, the results of our longi-
tudinal analysis suggest that pDC numbers do recover to some
degree after virus suppression with HAART and that they do
so more quickly and consistently than do mDCs. This degree
of recovery is encouraging. Thedifferenceintherateanddegree
of reconstitution of the different DC subtypes during therapy
suggests that the mechanisms of their depletion from the blood
during active viremia may also differ. Understandingthemech-
anism of the observed defects in blood DCnumberswillrequire
In addition to assessing the effect of HIV-1 replication on
blood DC numbers, we evaluated the relationship between pe-
ripheral CD4?T cell counts and numbers of both mDCs and
pDCs in the blood of our HIV-1–infected subjects, finding a
strong positive correlation betweenthese2parametersandcon-
firming the findings of others [5, 9]. A similar relationship
between peripheral CD4?T cell and DC numbers was observed
in the setting of antiretroviral therapy, as evidenced in our
longitudinal analysis of DCs before and after initiation of
HAART. This positive correlation between blood DC and pe-
ripheral CD4?T cell count suggests that the level of blood DCs
may serve as another measure of immune competence during
HIV-1 infection. In addition, the positive correlationmayspeak
to an intricate relationship between blood DCs and CD4?T
cells. In fact, several murine studies have shown that DCs may
influence the survival of CD4?T cells, naive CD4?T cells in
particular, in the periphery [25, 26].
The finding that both mDC and pDC numbers were de-
creased in the setting of HIV-1 plasma viremia also led us to
present on DCs in this setting. CD4, CXCR4, and CCR5 are
surface receptors expressed on cells, including CD4?T lym-
by guest on November 5, 2015
Plasma Viremia and Blood Dendritic Cells • JID 2003:187 (1 January) • 35
phocytes, monocytes, and some DCs, that are susceptible to
infection with HIV-1 [27–30]. We sought to examine the effect
of HIV-1 infection and active HIV-1 plasma viremia on DC
surface expression of these HIV-1 coreceptors. Using 3-color
flow cytometry, expression of CD4 on the total Lin?HLA-DR?
blood DC population was first evaluated. We observed that the
percentage of Lin-HLA-DR?blood DCs expressing CD4 was
significantly decreased in all HIV-1–infectedsubjects,regardless
of treatment status, and that the percentage of CD4?DCs cor-
related inversely with plasma HIV-1 RNA levels. This initially
raised the question as to whether the DCs that expressed CD4
had become actively infected withHIV-1and,asaconsequence,
were deleted from the blood or whether surface expression of
CD4 on blood DCs had been down-regulated. When the re-
spective DC subsets were examined using 4-color flow cyto-
metry, the percentage of each DC subset expressing CD4 did
not significantly differ between the control group and the vi-
remic subjects. These results suggest that, in all likelihood, the
lower numbers of CD4?Lin-HLA-DR?DCs observed initially
in the blood of HIV-1–infected subjects reflects an absolute
decrease in numbers of CD4-expressing DCs in both DC sub-
sets rather than an actual down-regulation of CD4 expression
on DCs. Whether CD4?blood DCs have been deleted from
the peripheral DC pool or have migrated out of the peripheral
blood remains unclear.
CCR5 and CXCR4 serve both as chemokine receptors and
as coreceptors for HIV-1 infection, and surface expression of
these receptors on DCs has been reported to change as a result
of DC maturation [31–33]. As such, their expression on each
DC subset might be involved in mediating susceptibility to
HIV-1 infection [24, 32], be involved in trafficking of blood
DCs to tissues and lymph nodes , or may simply reflect an
altered DC activation state. Our results, as well as those of some
previous studies, showed that the HIV-1 coreceptors CD4,
CCR5, and CXCR4 were more highly expressed on the surface
of pDCs than on mDCs in control subjects, suggesting a pos-
sible mechanism for increased susceptibility of pDCs to HIV-
1 infection and their subsequent deletion in the blood [5, 32].
This hypothesis also finds support in several in vitro studies
that reported an increased susceptibility of pDCs, compared
with mDCs, to infection with certain strains of HIV-1 [24, 32].
However, as noted above, increased expression of CCR5 was
observed on mDCs in viremic, HIV-1–infected subjects, po-
tentially rendering these DCs more susceptible than usual to
infection with R5-tropic strains of HIV-1, as well as making
them more likely to migrate in response to inflammatory stim-
uli . DC subsets from viremic subjects also expressed
slightly higher levels of CXCR4 than did DCs from seronegative
donors. DC surface expression of CXCR4 has been reported
to increase  or to remain unchanged [32, 33] with matu-
ration in some studies, whereas surface expression of CCR5
was observed to decrease with maturation in one study .
Whether or not the changes in blood DC chemokine receptor
expression that we observed in viremic subjects are implicated
in causing the defect in DC numbers, or whether these changes
simply reflect differences in DC activation or maturation, has
yet to be determined.
The DC costimulatory molecules CD40 and CD86 increase
as DCs mature and are important in T cell activation through
interactions with CD40L and CD28, respectively . We ob-
served that DC expression of both CD40 and CD86 was higher
in HIV-1–infected subjects than in control subjects. Amongthe
HIV-1–infected subjects, those with plasma viremia had higher
median expression of these costimulatory molecules,compared
with those who were HAART suppressed, and a significant
positive correlation was observed between DC expression of
CD86 and plasma HIV-1 RNA levels. These changes in costi-
mulatory molecule expression on Lin?HLA-DR?blood DCs
were also observed on pDC and mDC subsets in subjects with
viremia. Of interest, HN subjects treated with effective anti-
retroviral therapy had decreases in the expression of CD40 and
CD86 on DCs over time with virus suppression, which suggests
that these abnormalities may, in part, be reversible with treat-
ment. Increased DC surface expression of CD40 and CD86,
taken together with the correlation observed between CD86
and plasma viremia, suggests that HIV-1 may serve as an ac-
tivation or maturation stimulus to DCs, either directly or in-
directly via production of proinflammatory cytokines such as
tumor necrosis factor (TNF)–a . Proinflammatory cyto-
kines, such as TNF and interleukin-1, are known to be potent
DC maturation factors [35–37], and elevated levels of these
cytokines have been measured in the blood of HIV-1–infected
individuals . It has also been shown that maturation of
blood DCs may be associated with an up-regulation of CCR7
expression, theoretically resulting in migration of blood DCs
to lymph nodes and other tissues [39–41]. Thus, the decrease
in blood DC numbers observed in HIV-1–infectedsubjectsmay
not simply reflect deletion of these subsets but may represent
an adaptive mechanism by which DC precursors deal with a
large pathogen load by migrating from the blood to T cell–rich
areas in lymphoid organs, where they present Ag and stimulate
T cells. Finally, one potential adverse consequence of increased
DC costimulatory molecule expression might be a subsequent
increase in nonspecific activation of T cells by DCs, thereby
implicating DCs in the cycle of chronic immune activation
associated with unchecked HIV-1 infection .
To better define the relationship between blood DC numbers
and T cell function in our HIV-1–infected cohort, weexamined
the relationship between DC numbers and CD4?lymphopro-
liferative ability. CD4?T cell counts and plasma HIV-1 RNA
levels were statistically predictive of the probability of an LPA
response to most Ags, as might be expected. Although, in gen-
by guest on November 5, 2015
36 • JID 2003:187 (1 January) • Barron et al.
eral, blood DC numbers did not predict a significant lympho-
proliferative response to HIV-1 Ags, they did appear to predict
T cell responses to CMV Ag, responses that are generally larger
in magnitude than those directed against HIV-1 Ags. The in-
ability of DC numbers to predict an HIV-1–specific LPA re-
sponse does not negate the likely importance of DCs in initi-
ating or maintaining these Ag-specific T cell responses in vivo.
The relationship between DCs and T cell function in the setting
of HIV-1 infection may not be based simply on numbers of
DCs but perhaps on more complex aspects of DC function. As
mentioned, blood DCs may play a role in inducing primary
immune responses to evolving viral Ags in the setting of active
viremia or may be involved in sustaining naive T cells in the
periphery during immune restoration on HAART. Sorting out
the complex relationship between blood DCs and T cells will
require further investigation.
Defining the defects in blood DC number and function in
the setting of HIV-1 infection should aid in the design of ther-
apies aimed at fully restoring immune function in chronically
HIV-1–infected individuals. The results presented here further
delineate the impact of activeHIV-1replicationonthenumbers
and immunophenotype of blood mDCs and pDCs and suggest
a possible role for blood DC subsets in the immunopathoge-
nesis of HIV-1 disease.
We thank the physicians, staff, and patients in the Infectious
Diseases Group Practice at the University of Colorado Health
Sciences Center, for their assistance and participation in our
study; Karen Helm and Mike Ashton, for help with the flow
cytometry; Ian McNiece and Steven Rosinski, for technical ad-
vice; Robert Schooley, for critical review of the manuscript;and
the National Institutes of Health Division of AIDS, Vaccineand
Prevention Research Program, for providing reagents used in
1. Banchereau J, Steinman RM. Dendritic cells and the control of im-
munity. Nature 1998;392:245–52.
2. Knight SC, Patterson S. Bone marrow–derived dendriticcells,infection
with human immunodeficiency virus, and immunopathology. Annu
Rev Immunol 1997;15:593–615.
3. Pope M, Gezelter S, Gallo N, Hoffman L, Steinman RM. Low levels
of HIV-1 infection in cutaneous dendritic cells promote extensive viral
replication upon binding to memory CD4?T cells. J Exp Med 1995;
4. Cameron P, Pope M, Granelli-Piperno A, SteinmanRM.Dendriticcells
and the replication of HIV-1. J Leukoc Biol 1996;59:158–71.
5. Donaghy H, Pozniak A, GazzardB, et al.LossofbloodCD11c?myeloid
and CD11c?plasmacytoid dendritic cells in patients with HIV-1 in-
fection correlates with HIV-1 RNA virus load. Blood 2001;98:2574–6.
6. Feldman S, Stein D, Amrute S, et al. Decreased interferon-alpha pro-
duction in HIV-infected patients correlates with numerical and func-
tional deficiencies in circulating type 2 dendritic cell precursors. Clin
7. Belsito DV, Sanchez MR, Baer RL, Valentine F, Thorbecke GJ. Reduced
Langerhans’ cell Ia antigen and ATPase activity in patients with the ac-
quired immunodeficiency syndrome. N Engl J Med 1984;310:1279–82.
8. Macatonia SE, Lau R, Patterson S, Pinching AJ, Knight SC. Dendritic
cell infection, depletion and dysfunction in HIV-infected individuals.
9. Pacanowski J, Kahi S, Baillet M, et al. Reduced blood CD123?(lym-
phoid) and CD11c?(myeloid) dendritic cell numbers in primaryHIV-
1 infection. Blood 2001;98:3016–21.
10. O’Doherty U, Steinman RM, Peng M, et al. Dendritic cells freshly
isolated from human blood express CD4 and mature into typical im-
munostimulatory dendritic cells after culture inmonocyte-conditioned
medium. J Exp Med 1993;178:1067–76.
11. Patterson S, Gross J, English N, Stackpoole A, Bedford P, Knight SC.
CD4 expression on dendritic cells and their infection by human im-
munodeficiency virus. J Gen Virol 1995;76:1155–63.
12. Rissoan MC, Soumelis V, Kadowaki N, et al. Reciprocal control of T
helper cell and dendritic cell differentiation. Science 1999;283:1183–6.
13. O’Doherty U, Peng M, Gezelter S, et al. Human blood contains two
subsets of dendritic cells, one immunologically mature and the other
immature. Immunology 1994;82:487–93.
14. Cella M, Facchetti F, Lanzavecchia A, Colonna M. Plasmacytoid den-
dritic cells activated by influenza virus and CD40L drive a potent TH1
polarization. Nat Immunol 2000;1:305–10.
15. Langenkamp A, Messi M, Lanzavecchia A, Sallusto F. Kinetics of den-
dritic cell activation: impact on primingofTH1,TH2andnonpolarized
T cells. Nat Immunol 2000;1:311–6.
16. Patterson S. Flexibility and cooperation among dendritic cells. Nat
17. Vieira PL, de Jong EC, Wierenga EA, Kapsenberg ML, Kalinski P.
Development of Th1-inducing capacity in myeloid dendritic cells re-
quires environmental instruction. J Immunol 2000;164:4507–12.
18. Ito T, Amakawa R, Inaba M, Ikehara S, Inaba K, Fukuhara S. Differ-
ential regulation of human blood dendritic cell subsets by IFNs. J
19. Grassi F, Hosmalin A, McIlroy D, Calvez V, Debre P, Autran B. De-
pletion in blood CD11c-positive dendritic cells from HIV-infected pa-
tients. AIDS 1999;13:759–66.
20. Chehimi J, Campbell DE, Azzoni L, et al. Persistent decreases in blood
plasmacytoid dendritic cell number and function despite effective
highly active antiretroviral therapy and increased blood myeloid den-
dritic cells in HIV-infected individuals. JImmunol2002;168:4796–801.
21. Cella M, Jarrossay D, FacchettiF,etal.Plasmacytoidmonocytesmigrate
to inflamed lymph nodes and produce large amounts of type I inter-
feron. Nat Med 1999;5:919–23.
22. Siegal FP, Kadowaki N, Shodell M, et al. The nature of the principal
type 1 interferon–producing cells in human blood. Science 1999;284:
23. Soumelis V, Scott I, Gheyas F, et al. Depletion of circulating natural
type 1 interferon–producing cells in HIV-infected AIDSpatients.Blood
24. Patterson S, Robinson SP, English NR, Knight SC. Subpopulations of
peripheral blood dendritic cells show differential susceptibility to in-
fection with a lymphotropic strain of HIV-1. Immunol Lett 1999;66:
25. Brocker T. Survival of mature CD4 T lymphocytes is dependent on
major histocompatibility complex class II–expressing dendritic cells. J
Exp Med 1997;186:1223–32.
26. Kirberg J, Berns A, von Boehmer H. Peripheral T cell survival requires
continual ligation of the T cell receptor to major histocompatibility
complex–encoded molecules. J Exp Med 1997;186:1269–75.
27. Deng H, Liu R, Ellmeier W, et al. Identification of a major co-receptor
for primary isolates of HIV-1. Nature 1996;381:661–6.
by guest on November 5, 2015
Plasma Viremia and Blood Dendritic Cells • JID 2003:187 (1 January) • 37
28. Choe H, Farzan M, Sun Y, et al. The beta-chemokine receptors CCR3
and CCR5 facilitate infection by primary HIV-1 isolates. Cell 1996;
29. Dragic T, Litwin V, Allaway GP, et al. HIV-1 entry into CD4?cells is
mediated by the chemokine receptor CC–CKR-5. Nature 1996;381:
30. Feng Y, Broder CC, Kennedy PE, Berger EA. HIV-1 entry cofactor:
functional cDNA cloning of a seven-transmembrane, G protein–
coupled receptor. Science 1996;272:872–7.
31. Lin CL, Suri RM, Rahdon RA, Austyn JM, Roake JA. Dendritic cell
chemotaxis and transendothelial migration are inducedbydistinctche-
mokines and are regulated on maturation. Eur J Immunol 1998;28:
32. Patterson S, Rae A, Hockey N, Gilmour J, Gotch F. Plasmacytoid den-
dritic cells are highly susceptible to human immunodeficiency virus
type 1 infection and release infectious virus. J Virol 2001;75:6710–3.
33. Sallusto F, Schaerli P, Loetscher P, et al. Rapid and coordinated switch
in chemokine receptor expression during dendritic cell maturation.
Eur J Immunol 1998;28:2760–9.
34. Ferbas JJ, Toso JF, Logar AJ, Navratil JS, Rinaldo CR Jr. CD4?blood
dendritic cells are potent producers of IFN-alpha in response to in
vitro HIV-1 infection. J Immunol 1994;152:4649–62.
35. Cella M, Engering A, Pinet V, Pieters J, Lanzavecchia A. Inflammatory
stimuli induce accumulation of MHC class II complexes on dendritic
cells. Nature 1997;388:782–7.
36. Sallusto F, Lanzavecchia A. Efficient presentation of soluble antigen by
cultured human dendritic cells is maintained by granulocyte/macro-
phage colony-stimulating factor plus interleukin 4 and downregulated
by tumor necrosis factor alpha. J Exp Med 1994;179:1109–18.
37. Sallusto F, Cella M, Danieli C, Lanzavecchia A. Dendritic cells use
macropinocytosis and the mannose receptor to concentrate macro-
molecules in the major histocompatibility complex class II compart-
ment: downregulation by cytokines and bacterial products. J Exp Med
38. Roux-Lombard P, Modoux C, Cruchaud A, Dayer JM. Purified blood
monocytes from HIV 1–infected patients produce high levels of TNF
alpha and IL-1. Clin Immunol Immunopathol 1989;50:374–84.
39. Cyster JG. Chemokines and cell migration in secondary lymphoid or-
gans. Science 1999;286:2098–102.
40. Jarrossay D, Napolitani G, Colonna M, Sallusto F, Lanzavecchia A.
Specialization and complementarity in microbial molecule recognition
by human myeloid and plasmacytoid dendritic cells. Eur J Immunol
41. Lanzavecchia A, Sallusto F. The instructive role of dendritic cells on T
cell responses: lineages, plasticity and kinetics. Curr Opin Immunol
42. Poli G, Fauci AS. Cytokine modulation of HIV expression. Semin
by guest on November 5, 2015