| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
*
Immunobiology Section, Laboratory of Parasitic Diseases, and
Laboratory of Molecular Microbiology, National Institute of Allergy and Infectious Diseases, and
Experimental Immunology Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892;
Department of Obstetrics/Gynecology and Medicine, Division of Infectious Diseases, Northwestern University Medical School, Chicago IL 60611; and
¶
Molecular Histology, Inc., Gaithersburg, MD 20879
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The above-mentioned intracellular pathogens are activators of
CD4+ T lymphocyte function and, as a result,
could promote HIV-1 expression by enhancing viral infectivity, reverse
transcription, integration, and spread within newly infected cells
(1, 8). Perhaps more importantly, these microbial agents
are strong inducers of many of the proinflammatory cytokines (e.g.,
IL-1
, IL-6, and TNF-
) known to stimulate transcription, assembly,
and HIV-1 release within already infected cells (9). In
the case of TNF-, HIV-1 transcription is enhanced through the
induction of NF-
B, which interacts with promoter sequences within
the long terminal repeat
(LTR)4 of the HIV-1
genome (10). Mycobacteria and other intracellular
pathogens are potent inducers of NF-B activity (11, 12), and, thus, coinfection with these microbes could drive
HIV-1 expansion in chronic AIDS patients.
In considering the effects of microbial coinfection on HIV-1
expression, a critical question concerns the cellular source of the
virus induced. Persistently infected memory T cells carrying integrated
provirus represent one potential source susceptible to
NF-B-dependent immune activation (9). Cells of the
monocyte/macrophage lineage may also provide an important reservoir for
viral induction, particularly in the latter stages of the disease when
CD4 lymphocyte counts are greatly reduced (13). HIV-1
infection of macrophages is typically noncytopathic, and cells of this
lineage have a long 14-day half-life (14). In addition, in
vitro virus production by HIV-1-infected macrophages has been shown to
be markedly up-regulated as a result of exposure to microbial stimuli,
with the NF-B pathway of transcriptional regulation again playing a
major role (1, 9, 11, 13). That macrophages can serve as a
major source of latent virus expression has been confirmed recently by
in situ localization of virus in tissues of AIDS patients exposed to a
bacterial opportunistic infection (15, 16).
We have been developing an in vivo experimental model, which employs
transgenic mice that carry complete DNA copies of the HIV-1 genome, to
study the mechanism(s) whereby coinfections induce HIV-1 expression in
latently infected hosts. The tissues of these animals produce low
levels of viral mRNA, as well as infectious virus recoverable by
coculture with human T cells (17). Since no viral spread
occurs in these mice because of the lack of the appropriate receptors
and coreceptors, they are of limited value for the study of HIV
infectivity or AIDS pathogenesis. Nevertheless, because of the presence
of integrated provirus, including an unaltered LTR, the animals can be
used to focus on the factors that regulate latent viral expression. Our
initial work using this model (12) showed that infection
of line 166 transgenic mice with T. gondii, an opportunistic
intracellular protozoan parasite that stimulates a strong
proinflammatory and Th1-biased immune response (18),
results in elevated HIV-1 gene expression in tissues invaded by the
pathogen. The latter response was not seen in an equivalent transgenic
mouse line in which the NF-B-binding regions of the proviral LTR
were genetically inactivated.
In the present study, we have extended our analysis of coinfection-induced HIV-1 expression in the transgenic mouse model by demonstrating that the increased proviral mRNA observed is reflected in the enhanced synthesis of viral protein (p24), as well as the assembly and release of infectious virions. Moreover, we show that a second intracellular opportunistic pathogen, Mycobacterium avium, can also serve as potent microbial stimulus of both localized and systemic HIV-1 expression and demonstrate that virus levels change during the course of infection of line 166 mice with this bacterial agent. Finally, by means of in situ as well as in vitro analyses, we establish that cells of the monocyte/macrophage lineage are the major and selective source of virus production in animals exposed to the above microbial stimuli. Taken together, our findings support the concept that macrophages containing integrated proviral DNA can serve as an important reservoir for microbially induced immune activation and establish a useful and relevant in vivo model for studying this process in the context of mycobacterial infection.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The transgenic mouse line 166 was derived as previously
described (12, 19) by pronuclear injection of FVB/N mouse
embryos with proviral DNA encoding the entire genome of the NL4-3
molecular clone, a T cell-tropic strain of HIV-1. The resulting animals
contained 
20–60 copies of the proviral transgenes present at single
integration sites and transmitted them in a stable Mendelian fashion.
Mice were maintained by homozygous breeding under specific
pathogen-free conditions in an escape-proof facility within the animal
care facilities of the National Institute of Allergy and Infectious
Diseases (NIAID, Bethesda, MD). Nontransgenic FVB/N mice were bred in
the same facility as controls. Age- and sex-matched transgenic and
control animals between 6 and 12 wk old were used in all
experiments.
Infectious agents
The M. avium strain 2-151 SmT was kindly provided by
Dr. A. Cooper, Colorado State University. Stocks were prepared
containing 108 CFU/ml in saline, as previously
described (20), and stored frozen at -70°C. The
avirulent ME-49 strain of T. gondii was maintained by
passage in mice, as previously described (21), and
infectious tissue cysts were derived from the brains of animals
infected at least 1 mo previously. Tachyzoites of the
temperature-sensitive Ts-4 strain of T. gondii were
propagated by tissue culture in human fibroblasts utilizing published
procedures (22).
Experimental infections
For infection with M. avium, mice were each injected i.v. using a dose of either 107 or 108 CFU (suspended in PBS) in a total of 300 µl. To assess bacterial load, spleens were harvested at various time points postinfection, and single cell suspensions were prepared and serially diluted from 106 to 102 cells/ml in buffered saline and cultured on agar plates as previously described (20). CFU per 106 cells were translated to CFU per spleen by multiplying by the total splenocyte count. For infection with T. gondii, mice were inoculated i.p. with 30 brain-derived cysts.
Measurement of HIV-1 gene expression by RT-PCR
Relative levels of HIV-1 mRNA were determined by RT-PCR using an adaptation of a previously described protocol (23). Briefly, single cell suspensions prepared from spleen and other tissues were thoroughly washed, and total RNA was prepared by lysis of 106 cells in RNA-STAT-60 (Tel-Test, Friendswood, TX), followed by precipitation from the aqueous phase as recommended by the manufacturer. Recovered RNA was resuspended in diethyl pyrocarbonate-treated, distilled, deionized water, and cDNA was synthesized. PCR reactions were performed on serial dilutions of cDNA (from 10 µl) in a final volume of 50 µl, and a sample (10 µl) of each PCR reaction was electrophoresed through a 1.0% agarose gel and visualized with ethidium bromide. The number of cycles of PCR amplification used was first determined by amplifying cDNA through 24 to 36 cycles and comparing the product obtained to a standard curve from LPS-stimulated spleen cells. The number of cycles of amplification was chosen to give a PCR product that was easily detected in a gel, while remaining on the linear part of the amplification curve. Ethidium bromide-stained gels were photographed with an Eagle Eye II Still Video System (Stratagene, La Jolla, CA), and the intensity of fluorescence was determined using the associated Eaglesight software. To ensure that equivalent amounts of cDNA were used in each reaction, PCR was performed for hypoxanthine phosphoribosyltransferase (HPRT) from each sample, and the cDNA was adjusted to equivalent levels. Both pairs of primers (which were synthesized at the NIAID) spanned at least one intron, allowing mRNA to be distinguished from any contaminating genomic DNA. Cycle number and primer sequences used were as follows. HPRT (29 cycles): HPRT sense, GTT GGA TAC AGG CCA GAC TTT GTT G; HPRT antisense, GAG GGT AGG CTG GCC TAT AGG CT; gag (30 cycles): gag sense, ATA ATC CAC CTA TCC CAG TAG GAG AAA T; gag antisense, TTT GGT CCT TGT CTT ATG TCC AGA ATG C.
Detection of infectious virus by in vitro culture in MT4 cells
The presence of infectious virions in transgenic mouse tissue
was assessed by coculture with a human T lymphocytic line. Spleen cells
were harvested from uninfected or T. gondii- or M.
avium-infected line 166 mice, or, as controls, FVB/N mice infected
with the same agents. The splenocytes were then distributed in 24-well
Costar chambers (Corning Costar, Cambridge, MA) in 500-µl aliquots at
a concentration of 2 x 106, 2 x
105, or 2 x 104
cells/ml in a culture medium consisting of RPMI 1640 (Life
Technologies, Grand Island, NY), 10% FCS (HyClone, Logan, UT), 50 U/ml
penicillin (National Institutes of Health (NIH) stock), 50 µg/ml
streptomycin (NIH stock), and 2 mM glutamine (NIH stock). Log phase
cells of the human T cell line MT-4 (24) were added to
each well in 500 µl of culture medium at 2 x
106 cells/ml, and culture supernatants were
assayed every 2 days for the presence of reverse transcriptase as
described previously (25). Briefly, 10 µl of culture
supernatant was pipetted into 50 µl of reaction buffer consisting of
50 mM Tris (pH 7.6), 75 mM KCl, 5 mM MgCl2, 0.96
mM EDTA, 0.05% Nonidet P-40 (Sigma, St. Louis, MO), 5 µg/ml poly
(rA) (Pharmacia Biotech, Piscataway, NJ), 1.6 µg/ml
oligo(dT)12–18 (Pharmacia Biotech), 2 mM DTT, and 10 µCi/ml
[-32P]dTTP (Amersham Life Science Products, Arlington
Heights, IL) added immediately before use. This reaction was incubated
for 3 h at 37°C, then terminated by the addition of SDS to
0.1%. A sample of reaction mix (5 µl) was spotted onto DEAE filter
mats (Wallac, Turku, Finland), air dried for 60 min, and washed 3 times
for 10 min in 2x SSC, then twice in 95% ethanol. The filter was dried
for 10 min in a vacuum oven at 80°C, before being added to a bag
containing 4 ml of beta plate scintillation fluid (Wallac) and counted
in a Wallac 1450 beta plate counter. Results were expressed as the mean
cpm incorporated in duplicate reactions.
In an additional series of experiments, spleen cells from uninfected 166 transgenic mice were cultured as above at a 1:1 or 1:10 ratio, with a fixed number of MT-4 cells/ml in the presence of either M. avium (106 CFU/ml), T. gondii Ts-4 strain tachyzoites (106/ml), or with LPS (100 ng/ml; Sigma) as a positive control. Negative controls consisted of MT-4 cells exposed to the same stimuli in the absence of spleen cells. Aliquots of culture supernatants were harvested at 2-day intervals and assayed for RT activity as described above.
Localization of viral expression by in situ hybridization
Portions of spleen, lungs, and liver from M. avium-infected mice were fixed in 1.3 M aqueous formaldehyde for 24 h. In situ hybridization (ISH) was performed as previously described (26). Briefly, two sets of mounted 6-µm paraffin sections were dewaxed and treated with protease to expose viral nucleic acid. They were hybridized with a 33P-labeled antisense probe (HIV-1, IIIB) that represents 9 kb of the HIV-1 genome. A sense probe hybridization was also performed as a control. After hybridization, one set of slides was dipped in NTB 2 Kodak emulsion (undiluted), and the other was used for phosphor storage imaging (27) on a Fuji BAS 5000 instrument.
The autoradiograms were exposed in the dark over desiccant for 4 days at 4°C and then developed in Kodak D-19 (Eastman Kodak, Rochester, NY) for 4 min at 15°C. The phosphor storage images were analyzed with MacBas v2.5 software to determine the number of cells expressing HIV RNA per unit area. Alternatively, autoradiograms were stained with hematoxylin and eosin for morphological and pathological assessment. Some slides were stained with Kinyoun’s carbol fuchsin after development and carefully decolorized to preserve the silver grains. This procedure allowed simultaneous detection of bacteria and HIV-expressing cells (15). In other instances, slides were stained with a rat mAb (clone M3/84; PharMingen, San Diego, CA) to Mac-3 to detect cells of the monocyte/macrophage lineage. The Ab was applied using commercial reagents for immunoperoxidase detection (Dako, Carpinteria, CA).
Detection of p24 Ag levels in plasma and ex vivo culture supernatants
At various time points pre- and postinfection with M. avium, transgenic mice were bled from the tail vein into EDTA-treated vacutainer tubes (Becton Dickinson, Franklin Lakes, NJ). Plasmas were separated and kept frozen at -20°C before assay. p24 antigenemia was determined by ELISA using a commercial kit (Coulter, Miami, FL; detection limit, 4 pg/ml) according to the manufacturer’s instructions. To control for individual variations in basal p24 levels between animals, antigenemia was calculated as the fold increase in p24 vs the preinfection level for each individual mouse. The means and SE were then calculated on the pooled values from all animals assayed at each time point.
In additional experiments, ex vivo production of p24 by freshly isolated spleens from infected and control mice was assessed by incubating single cell suspensions (5 x 106/ml) for 48 h at 37°C in a culture medium consisting of RPMI 1640 (Life Technologies), 10% FCS (HyClone) MEM nonessential amino acid solution (0.1 mM final concentration) (Life Technologies), 5.5 x 10-5 M 2-ME (Life Technologies), 10 mM HEPES (Life Technologies), 100 U/ml penicillin (NIH stock), 100 µg/ml streptomycin (NIH stock), and 2 mM glutamine (NIH stock). p24 levels were then determined by ELISA as described above.
In vitro stimulation of p24 production in splenocyte or peritoneal macrophage cultures from noninfected transgenic mice
To assess the effects of T cell vs macrophage stimuli on HIV-1
expression, spleen cells from uninfected line 166 mice were cultured
for 72 h at 5 x 106/ml in RPMI 1640
culture media (see above) and stimulated in vitro with immobilized
anti-CD3 (10 µg/ml), Con A (Sigma; 2.5 µg/ml), LPS (Sigma; 100
ng/ml), Staphylococcus aureus Cowan (SAC;
Calbiochem-Behring, Pansorbin, La Jolla, CA; 0.01%), or M.
avium (107/ml). Production of p24 in 48-h
and IFN-
in 72-h culture supernatants was measured by two-site
ELISA. The IFN- assay used immobilized HB 170 mAb, a polyclonal
monospecific rabbit anti-mouse IFN- Ab, and peroxidase
conjugated donkey anti-rabbit IgG (Jackson Immunoresearch
Laboratories, West Grove, PA) as previously described
(28). In one set of experiments, peritoneal macrophages
obtained from mice 5 days after elicitation with sterile 3%
thioglycollate, as described previously (12), were
stimulated with M. avium, and the p24 response was measured
using the above culture and ELISA protocols.
Phenotypic analysis of HIV-1-expressing cells by fluorescence in situ hybridization and flow cytometry
To assess the cell surface phenotype of virus-producing cells, 2 x 106 spleen cells from M. avium-infected line 166 mice were incubated in FACS buffer (balanced salt solution, 0.1% BSA, 0.01% sodium azide) with FcR block (PharMingen; anti-CD16/32, 10 µg/ml) for 10 min at +4°C and then stained with PE-labeled Ab (PharMingen; anti-CD4, -CD8, -CD19, and -Mac1) for 10 additional min, at +4°C. The cells were next washed twice with FACS buffer and fixed in 100 µl of PermeaFix (Ortho Diagnostics, Raritan, NJ) for at least 1 h at room temperature. Following further washes first in sterile PBS and then in 2x SSC, the cells were pelleted and resuspended in hybridization solution (2x SSC, 30% formamide, sonicated salmon sperm, and yeast transfer RNA) containing 500 ng of 5-carboxy-fluorescein double-end-labeled, gag-pol-specific oligonucleotides probes or gag-pol sense oligonucleotides as a negative control probe mixture (29). The intracellular hybridization was performed at 42°C for 1 h, followed by successive washes at 42°C in 2x SSC, 0.1% Triton X-100, and in 0.1x SSC, 0.1% Triton X-100. Finally, the cells were resuspended in PBS (pH 8.3), and two color analysis was performed using an Epic XL flow cytometer (Coulter).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
M. avium is a Gram-positive, acid-fast bacterium that
is a major cause of opportunistic infection in AIDS patients
(30). In common with T. gondii, the microbial
stimulus previously studied in this system (12), M.
avium is an intracellular pathogen that induces strong
proinflammatory and Th1-type cytokine production but differs from that
protozoan in selectively infecting macrophage/monocytes (20, 23, 30). To assess the capacity of M. avium to alter HIV-
expression in vivo, line 166 transgenic mice were infected i.v. with
107 CFU/animal. As a positive control, a second
group of transgenic mice was infected with 20 cysts of the ME-49 strain
of T. gondii. At 7 and 14 days postinfection, spleens,
livers, and lungs were removed from the animals, and cell suspensions
were prepared. HIV-1 mRNA expression was then assayed on a fixed number
of cells by semiquantitative RT-PCR using primers for the
gag gene. As shown in Fig. 1
,
M. avium induced 2.5- to 5-fold increases in gag
mRNA in spleen relative to comparable tissue from noninfected mice.
These responses were detected at 7 days and increased further by the
second week of infection. Comparable changes in mRNA levels were
observed in liver and lungs of the same M. avium-infected
animals whereas no viral RNA was detected in infected control
(nontransgenic) FVB/N mice at any time, although the recovery of
bacteria from these animals was indistinguishable from that measured in
transgenic mice (data not shown). Confirming our previous work
utilizing Northern analyses of viral mRNA production (12),
increased gag expression was also detected by RT-PCR in
spleens of mice infected with T. gondii. (Fig. 1). The above
data establish that mycobacterial as well as protozoan infections can
stimulate HIV-1 mRNA expression in tissues of line 166 transgenic
mice.
|
Since transgenic mice carrying full-length copies of proviral DNA
possess the biosynthetic machinery necessary for the assembly of HIV-1
virions (17), we asked whether the elevations in viral
mRNAs seen in tissues of infected line 166 mice would be reflected in
the induction of infectious HIV-1. To assess this possibility, spleen
cells from M. avium-, T. gondii-, or sham-infected
transgenic mice were cocultured with the human T cell line MT-4 at a
ratio of 1:1, 1:10, and 1:100, and supernatants were assayed at 2-day
intervals for the presence of RT. Because RT activity is nearly
exclusively virion associated, its amplification in T cell cultures is
a useful indicator of spreading infection by viral particles.
Supernatants from cocultures containing spleen cells from uninfected
line 166 mice displayed measurable although low levels of RT,
confirming that the proviral transgenes in these animals can indeed be
expressed and that their products yield infectious HIV (Fig. 2A). This RT activity was
detected only at the highest concentration of splenocytes employed,
suggesting that the number of cells releasing virus in uninfected
animals is quite small. In contrast, higher levels of RT were produced
in cultures containing the same number of spleen cells from mice
infected with either M. avium or T. gondii, and
this response was retained at a 1:10 dilution of the splenocytes
relative to the MT-4 indicator cells whereas all activity was lost upon
an additional 10-fold dilution (Fig. 2, B and C).
These results suggest that the initial spleen cell populations from the
M. avium- and T. gondii-infected mice produced
infectious virions at a level at least 10 times that generated by
splenocytes from uninfected mice.
|
Nonetheless, there remained the possibility that the greater viral
yields observed in the spleens of line 166 mice infected with M.
avium or T. gondii reflect the recruitment of
preexisting HIV-1-producing cells into that organ rather than the de
novo induction of virus from normally silent splenocytes. To confirm
that these microbial agents are capable of directly stimulating virus
production, spleen cells from uninfected 166 transgenic mice were
cultured at a 1:1 or 1:10 ratio with a fixed number of MT-4 cells per
milliliter in the presence of either M. avium
(106 CFU/ml), T. gondii tachyzoites of
the avirulent Ts-4 strain (106/ml), or with LPS
(10 µg/ml) as a positive control. As shown in Fig. 3, in vitro exposure of transgenic spleen
cells to each of these stimuli (Fig. 3, B–D, respectively)
resulted in a 10-fold or better increase in the yield of RT activity
relative to that displayed by nonstimulated control cultures based on
cell dilution (Fig. 3A). Addition of any of these microbial
agents to MT-4 cells in the absence of transgenic spleen cells, or to
transgenic T cells in the absence of MT-4 cells, failed to induce
detectable RT activity (data not shown). Together, the above in vitro
findings argue that up-regulated HIV-1 expression observed in vivo upon
T. gondii or M. avium infection can result at
least in part from the de novo induction of viral synthesis in mouse
cells.
|
To further confirm that the increased viral loads seen in tissues
of pathogen-infected transgenic animals are due to the induction of
HIV-1 expression rather than solely to cell migration, we measured
levels of the p24 core protein in plasma as a parameter of systemic
viral production. M. avium was used as the infecting agent
in this and all subsequent experiments because of the ease of
quantitating the infection and for the purposes of standardizing a
defined model for future studies. Low and variable amounts (100–400
pg/ml) of p24 protein were detectable by ELISA in plasma of uninfected
166 mice. However, in animals infected with M. avium
(108 CFU/mouse, administered i.v.), plasma p24
levels rapidly increased to 3-fold the preinfection values at wk 1 and
then gradually declined to baseline by wk 5–6 postinfection. In
contrast, in plasma from age-matched uninfected transgenic mice, p24
levels remained stable during the same period with only minor
fluctuations (Fig. 4A).
|
B). This change roughly
paralleled the increase in bacterial loads (from 0.53 x
108 at wk 1 to 22.3 x
108 CFU/spleen at wk 6) measured in the same
cells during the same period (Fig. 4B). Nevertheless, the
time course of p24 expression in spleen was almost the converse of that
in the plasma (Fig. 4A), suggesting that these two read-outs
reflect distinct events associated with virus production in
vivo. Localization of M. avium-induced viral expression in tissues
The preceding experiments demonstrated that pathogens such as
T. gondii or M. avium can induce the expression
of both HIV-1 mRNA and infectious virus from transgenic murine cells.
Since, in a transgenic animal presumably every cell contains proviral
DNA and is therefore in principle capable of viral synthesis and
secretion, an important question concerns whether the HIV-1 induced by
microbial immune activation originates from specific cell types. As a
first step in examining the cellular origins of the induced virus,
tissues from M. avium-infected line 166 mice were embedded
in paraffin, sectioned, and protease-treated to expose nucleic acids,
and ISH was performed with labeled antisense riboprobe derived from the
HIV genome. No detectable hybridization was observed with tissues of
nontransgenic littermates (data not shown) or with tissues of M.
avium-infected transgenic mice with a control sense riboprobe
(e.g., Fig. 5A). Very low
levels of hybridization (antisense) were detected in all tissues
sampled from uninfected line 166 mice, the signals appearing
sporadically and usually small in size (Figs. 5B and
6A). However, when examined at
day 14 postinfection, the hybridization signals had clearly increased
in both frequency and intensity in spleen (Fig. 5C), lung
(Fig. 6B), and liver (not shown). Densitometric quantitation
revealed an approximate 4- to 5-fold increase in hybridization signal
in tissues of infected vs noninfected animals, consistent with the
results of RT-PCR performed on similar tissues (Fig. 1).
|
|
C). Similarly, in the lungs
of infected mice, hybridization foci were most abundant within the
macrophage-enriched granulomatous lesions resulting from the immune
response to M. avium bacilli (Fig. 6, B and
D). Colocalization of HIV-1 by ISH and of M.
avium by acid fast staining revealed that both bacteria-infected
and uninfected macrophages express virus. Indeed, many of the
ISH-positive cells, although concentrated largely within the
inflammatory lesions, lacked visible bacteria (Fig. 6B).
Immunocytochemistry performed on nearby sections confirmed that the
majority of cells in the areas of inflammation where elevated viral
expression was observed were mononuclear phagocytes, as judged by their
staining with Ab specific for the Mac-3 surface Ag marker (Fig. 6D). Ex vivo analysis of HIV-1-expressing cell population from M. avium-infected mice
To confirm that macrophages constitute the major population
producing virus in response to M. avium infection, spleen
cells from 7-day-infected line 166 mice were stained with PE-labeled
mAb against different lymphocytic and monocytic cell surface markers,
and FISH was performed with fluoresceinated oligonucleotides specific
for HIV gag-pol RNA. Dual color FACS analysis was then performed on the
stained populations. As shown in Table I
and Fig. 5, in spleens of mice infected with M. avium, the
percentage of Mac-1+ cells expressing HIV gag-pol
RNA was dramatically increased, as compared with splenocytes from
uninfected animals. This approximate 12-fold augmentation in HIV
mRNA-expressing cells (from 0.4 to 4.9%) could not be explained by the
increased percentage of Mac-1+ cells in the
spleen (2.4% in uninfected mice vs 5.4% in M.
avium-infected mice). In contrast, no substantial elevations were
observed in the percentage of HIV gag-pol RNA-expressing cells within
the populations stained with T or B cell markers
(CD4+, CD8+, or
CD19+ cells) (Table I). These experiments argue
that macrophages rather than lymphocytes are the major source of HIV-1
expression in spleens of infected transgenic
mice.
|
One possible explanation of the finding that M. avium
infection preferentially induces virus production from macrophages in
transgenic mice is that the former pathogen preferentially invades and
stimulates these cells. To ascertain whether the observed macrophage
responses are not pathogen specific, a series of in vitro experiments
were performed in which splenocytes from uninfected line 166 mice were
exposed in vitro to known monocyte and/or T cell stimuli. As shown in
Table II, immobilized anti-CD3 or Con
A failed to induce significant increases in p24 production compared
with control cultures. In contrast, LPS and SAC stimulated greater than
20-fold increases in p24 levels. The failure of the T cell stimuli to
induce viral expression is not due to a lack of activation of the
lymphocytes since anti-CD3 and Con A triggered strong IFN-
responses in the same cultures. In agreement with both the in vivo
studies described above (Figs. 1, 2, 4, and 5; Table I) as well as the
results of the in vitro experiments measuring expression of HIV-1 by
MT-4 culture assay (Fig. 3), M. avium was found to directly
stimulate p24 production by spleen cells from uninfected transgenic
mice (Table II) and induced a highly significant p24 response in
enriched thioglycollate-elicited macrophages from the same animals
(345 ± 15 pg/ml in M. avium-stimulated vs 46 ±
15 pg/ml in unstimulated cultures). Together these results suggest that
macrophages provide a more readily induced reservoir for HIV-1
expression than T lymphocytes and that M. avium, in common
with other microbial agents such as T. gondii, SAC, and LPS,
can provide an excellent stimulus for virus production from that
cellular source.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A key issue we have addressed concerns the basis of the increased viral
expression observed in tissues of M. avium- or T.
gondii-infected mice. These elevations, which ranged between
approximately 4- and 10 -fold depending on the assay, could reflect
either de novo HIV-1 induction and/or the recruitment and localized
enrichment of virus-producing cells already present at low levels in
transgenic animals before microbial infection. The former explanation
is supported by several observations and, most importantly, the
systemic increases in p24 in plasma detected during the first few weeks
following M. avium infection (Fig. 4). Moreover, focal in
situ hybridization signals appeared to be more intense in tissues of
infected transgenic mice (Figs. 5 and 6), suggesting augmented
transcription from individual cells. Finally, significant increases in
infectious virus, as well as released p24, could be stimulated in
spleen cells in vitro (Fig. 3), a finding that is most consistent with
the direct induction of viral expression. Nevertheless, a partial
contribution of tissue-specific cell recruitment, and in particular of
macrophages (Table I), cannot be ruled out and in fact is predictable,
given the mononuclear inflammatory responses observed in the tissues
studied. Such a recruitment process, if occurring in
HIV+ humans, could play a role in promoting viral
spread.
Since M. avium, in common with both T. gondii and
M. tuberculosis, stimulates strong proinflammatory and
Th1-type host cytokine responses, its ability to trigger viral
expression in transgenic mice was not unexpected and confirms previous
reports describing an association of this bacterial infection with
HIV-1 expression in tissues of AIDS patients (15, 16).
Importantly, the transgenic mouse model has enabled us to follow for
the first time the kinetics of viral expression following M.
avium infection. This analysis revealed a rapid increase in p24
levels in both spleen cell supernatants and plasma by the first week
postinfection. Interestingly, while splenic p24 production continued to
rise during the next 5 wk consistent with the growing bacterial load in
that organ, plasma p24 levels declined in an approximate reciprocal
fashion (Fig. 4). The explanation for this differential response
pattern in plasma and spleen is presently unclear. One possibilty is
that it reflects changes in cytokine expression occurring during the
same period in M. avium-infected mice. For example, while
IFN- synthesis is induced early and reaches a plateau at 2 wk
(20, 23), IL-10 production is delayed and does not occur
at high levels until 3–4 wk postinfection (D. Jankovic and A. Sher,
unpublished observations). The decline in plasma p24 levels may
represent a response to this burst in IL-10 synthesis. Preliminary data
(not shown) indicate that p24 synthesis by spleen cells also becomes
diminished at later time points (i.e., >6 wk), perhaps as a
consequence of the same IL-10 regulatory response. Experiments in which
IL-10 has been experimentally depleted or administered during the
course of M. avium infection in transgenic mice are in
progress to assess the possible role of this down-regulatory cytokine
in the control of immune activation-induced viral expression.
Although the HIV-1 proviral transgene should be present in all somatic
cells in line 166 mice, viral expression in response to M.
avium appears to arise from a restricted cellular source. This
conclusion is evident, from the anatomical localization of HIV-1
expression in situ, from the preferential expression of viral mRNA ex
vivo by Mac-1+ cells from infected mice, and from
the induction of virus expression in vitro by LPS or SAC but not Con A
or anti-CD3. Taken together with our previous observations in
T. gondii-infected transgenic mice indicating that viral
mRNA is produced preferentially by adherent cells in spleen and can be
stimulated in thioglycollate-elicited peritoneal macrophages
(12), these findings strongly implicate cells of the
macrophage/monocyte lineage as the principal source of HIV-1 expression
in line 166 mice infected with either pathogen and argue against a T or
B lymphocyte origin. Although we cannot rule out a role for dendritic
cells in virus production, the paucity of HIV-1+
cells in the white pulp of spleens from M. avium-infected
mice (Fig. 5A) suggests that interdigitating dendritic cells
are not involved.
Despite strong microbial stimulation in vivo or polyclonal stimulation in vitro, T lymphocytes from transgenic mice failed to exhibit elevated viral expression although integrated proviral DNA was clearly present in FACS-sorted cells of this lineage (data not shown). Since the HIV-1 molecular clone used to construct the mouse was T tropic, the unresponsiveness of the transgenic T cells appears to be unrelated to the cellular tropism of that virus. Instead, the data suggest that virus-infected cells of the macrophage/monocyte lineage may be intrinsically more susceptible than T lymphocytes to immune activation. This distinction may be less apparent in persistently infected HIV-1+ human cells because of differences in initial infection levels between T cells and macrophages and because of ongoing receptor-mediated viral entry.
The above considerations suggest that the M. avium-infected transgenic mouse model we have described offers a powerful experimental tool for studying the immune activation of integrated HIV-1 from its in vivo macrophage reservoirs. Although present in lower numbers than HIV+ CD4+ T lymphocytes (31, 32), macrophages containing HIV-1-integrated DNA can constitute a major source of persistent infection since they are relatively resistant to viral cytopathic effects. Macrophages with HIV DNA have been isolated from lymph nodes and spleens throughout the course of infection, while macrophages expressing HIV mRNA are more difficult to detect, suggesting that they may serve as a protected site for replication-competent provirus (33). In addition, in certain anatomical sites, such as the central nervous system, monocyte/macrophages are the only infected cells detected in situ (34, 35). Little information exists on the survival of latently infected macrophages after potent antiretroviral therapy (36), although HIV within macrophages has been reported to be relatively resistant to protease inhibitors in vitro (37). Given that integrated HIV-1 DNA within resting CD4+ T cells can persist in patients receiving prolonged highly active antiretroviral therapy (31, 38, 39), it is likely that virus-infected macrophages are also refractory.
Monocytes/macrophages may be of particular importance as reservoirs for microbially stimulated virus production (1, 11, 15, 16), and thus the ability to infect mice with bacterial and parasitic pathogens commonly found in HIV-1+ individuals offers a unique advantage for studying this interaction in vivo. Because of the rapid (1–2 wk) response of line 166 to microbial (M. avium or T. gondii) infection and the ease of measuring this response by p24 antigenemia in plasma or cell culture supernatants, the model allows the rapid testing of interventions that inhibit or augment immune activation and could be utilized to assess the susceptibility of the macrophage reservoir to antiretroviral or immune-based therapies. We are currently pursuing these approaches with the aim of both better understanding the requirements for microbially induced viral expression and of developing strategies for limiting the impact of that process on HIV-1 progression.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Department of TB Immunology, Statens Serum Institut, 5 Artillerivej, 2300 Copenhagen S, Denmark.
3 Address correspondence and reprint requests to Dr. Alan Sher, Building 4, Room 126, National Institutes of Health, 4 Center Drive, Bethesda, MD 20892-0425. E-mail address:
4 Abbreviations used in this paper: LTR, long terminal repeat; HPRT, hypoxanthine phosphoribosyltransferase; ISH, in situ hybridization; FISH, fluorescence in situ hybridization; SAC, Staphylococcus aureus Cowan.
Received for publication March 10, 1999. Accepted for publication May 24, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
activates human immunodeficiency virus type 1 through induction of nuclear factor binding to the NF-B sites in the long terminal repeat. Proc. Natl. Acad. Sci. USA 86:5974.B and its inhibitor IB-/MAD-3 in monocytes by Mycobacterium tuberculosis and during human tuberculosis. J. Immunol. 159:4109.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  J. Sun, T. Soos, V. N. KewalRamani, K. Osiecki, J. H. Zheng, L. Falkin, L. Santambrogio, D. R. Littman, and H. Goldstein CD4-Specific Transgenic Expression of Human Cyclin T1 Markedly Increases Human Immunodeficiency Virus Type 1 (HIV-1) Production by CD4+ T Lymphocytes and Myeloid Cells in Mice Transgenic for a Provirus Encoding a Monocyte-Tropic HIV-1 Isolate J. Virol., February 15, 2006; 80(4): 1850 - 1862. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Zhang, G. Li, A. Bafica, M. Pantelic, P. Zhang, H. Broxmeyer, Y. Liu, L. Wetzler, J. J. He, and T. Chen Neisseria gonorrhoeae Enhances Infection of Dendritic Cells by HIV Type 1 J. Immunol., June 15, 2005; 174(12): 7995 - 8002. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Bafica, C. A. Scanga, M. Schito, D. Chaussabel, and A. Sher Influence of Coinfecting Pathogens on HIV Expression: Evidence for a Role of Toll-Like Receptors J. Immunol., June 15, 2004; 172(12): 7229 - 7234. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Bafica, C. A. Scanga, O. Equils, and A. Sher The induction of Toll-like receptor tolerance enhances rather than suppresses HIV-1 gene expression in transgenic mice J. Leukoc. Biol., March 1, 2004; 75(3): 460 - 466. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Chougnet Role of CD40 Ligand dysregulation in HIV-associated dysfunction of antigen-presenting cells J. Leukoc. Biol., November 1, 2003; 74(5): 702 - 709. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Bafica, C. A. Scanga, M. L. Schito, S. Hieny, and A. Sher Cutting Edge: In Vivo Induction of Integrated HIV-1 Expression by Mycobacteria Is Critically Dependent on Toll-Like Receptor 2 J. Immunol., August 1, 2003; 171(3): 1123 - 1127. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Equils, M. L. Schito, H. Karahashi, Z. Madak, A. Yarali, K. S. Michelsen, A. Sher, and M. Arditi Toll-Like Receptor 2 (TLR2) and TLR9 Signaling Results in HIV-Long Terminal Repeat Trans-Activation and HIV Replication in HIV-1 Transgenic Mouse Spleen Cells: Implications of Simultaneous Activation of TLRs on HIV Replication J. Immunol., May 15, 2003; 170(10): 5159 - 5164. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Bouchonnet, N. Boechat, M. Bonay, and A. J. Hance Alpha/Beta Interferon Impairs the Ability of Human Macrophages To Control Growth of Mycobacterium bovis BCG Infect. Immun., June 1, 2002; 70(6): 3020 - 3025. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. T. Keppler, F. J. Welte, T. A. Ngo, P. S. Chin, K. S. Patton, C.-L. Tsou, N. W. Abbey, M. E. Sharkey, R. M. Grant, Y. You, et al. Progress Toward a Human CD4/CCR5 Transgenic Rat Model for De Novo Infection by Human Immunodeficiency Virus Type 1 J. Exp. Med., March 18, 2002; 195(6): 719 - 736. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Chougnet, C. Freitag, M. Schito, E. K. Thomas, A. Sher, and G. M. Shearer In Vivo CD40-CD154 (CD40 Ligand) Interaction Induces Integrated HIV Expression by APC in an HIV-1-Transgenic Mouse Model J. Immunol., March 1, 2001; 166(5): 3210 - 3217. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
4
pg/ml. To determine bacterial loads in the same spleens, 1 x
106 cells were serially diluted in PBS and plated in
duplicate onto culture plates containing Middlebrook media. The number
of CFU/spleen (filled squares) was then calculated. Results shown are
the means and SEs of assays performed on individual mice. Asterisks
indicate values significantly different (p < 0.05,
Student’s t test) from those obtained from uninfected
controls.
