Published online before print August 21, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Johns Hopkins School of Medicine,
* Department of Neurology and
Center for Flow Cytometry, School of Public Health, Baltimore, Maryland, USA; and
Aaron Diamond AIDS Research Center, The Rockefeller University, New York, New York, USA
1 Correspondence: Johns Hopkins University School of Medicine, Department of Neurology, 600 North Wolfe St., Baltimore, MD 21287. E-mail: abrown76{at}jhmi.edu
ABSTRACT
Macrophages are recognized as a putative reservoir for HIV-1, but whether HIV can establish latent infection in this cell type is not known. An in vitro model using long-term cultured primary human monocyte-derived macrophages (MDM) infected with an M-tropic, enhanced green fluorescent protein (EGFP) tagged reporter virus was developed to test the hypothesis that HIV can establish a latent infection of this cell type. The EGFP-IRES-Nef cassette allowed detection of early gene transcription. The expression of GFP+ MDM was followed with time and the GFP- population was purified and analyzed for evidence of latent infection. Interestingly, in MDM cultures propagated for over two months, distinct subpopulations of infected GFP+ cells were observed and quantitated. In particular, infected MDM that displayed a high level of transcription, characterized as the GFP hi group, yet produced low levels of the late viral gene product, p24, increased with time and represented 10% of the GFP+ population in long-term cultures. The high level production of early genes such as Nef, a protein that can facilitate viral immune escape, but low level of structural proteins such as p24 in the GFP hi population suggests that a subset of infected MDM can exhibit an alternative mode of replication. The GFP- MDM population obtained by a two-step purification protocol using flow cytometry and laser ablation contained integrated provirus as assessed by Alu-LTR real-time PCR analyses. A subset of these, were replication competent as shown by their ability to express GFP and/or p24 antigen after reactivation with IL-4.
Key Words: GFP • latency • laser capture • flow cytometry
INTRODUCTION
Highly active anti-retroviral therapy (HAART) is very effective in suppressing HIV-1 replication, thus allowing restoration of the immune system and overall is a lifesaver for infected individuals. However, discontinuation of therapy leads to a rapid rebound in plasma viral load often to pretreatment levels, in as little as two weeks [1 ]. Studies on the turnover of virus in the plasma of HAART treated patients revealed three phases of decay [2 , 3 ]. The longest phase with a half-life now estimated as 44 months, is attributed to the resting CD4+ T cell pool that contains integrated proviral genome that is transcriptionally silent [4 ]. Upon cellular activation, replication of HIV from the infected resting T-cells is reignited and viruses found in the plasma can be traced back to the latent T cell pool [5 ]. Infected macrophages, the second major target cell type in which HIV can replicate productively, were proposed, based on half-life estimates determined in rodents, to account for the second phase of decay with a half-life of 14 days [3 , 4 ] and therefore, could over time be eradicated. The location of macrophages in tissues makes sampling of viral load from these cells difficult. Yet several features make macrophages, particularly those residing in the brain, attractive candidates for a viral reservoir. It is known that macrophages require higher concentrations of anti-virals to inhibit HIV-1 replication compared with T-cells [6 ]. In addition, the expression of high levels of P-glycoprotein drug transporter at the blood brain barrier as well as the poor penetration of many protease inhibitors create conditions of suboptimal viral suppression. In contrast to T-cells, HIV infection is not cytopathic for macrophages and this combined with the half-life of these cells on the order of months to years depending on the specific type of macrophage [7 ] could facilitate the ability of the virus to persist. Importantly, infected monocytes have been detected in the peripheral blood of patients receiving HAART [8 9 10 ]. The trafficking of such monocytes to the brain and their subsequent differentiation into macrophages may also account for a portion of the residual virus replication seen in treated patients. Additionally, in macrophages virions are assembled and can accumulate to great numbers within multivesicular bodies [11 , 12 ], remaining a source of infectious virions that can be transmitted to susceptible T-cells [13 ].
To begin to study the viral and cellular mechanisms for HIV persistence and perhaps latent infection of macrophages we developed an in vitro model based on the long-term culture of HIV-infected primary human monocyte derived macrophages (MDM) infected with an enhanced green-fluorescent protein (EGFP)-tagged reporter virus. The EGFP and Nef genes are contained on a multiply spliced bicistronic mRNA that is synthesized early in the virus replication cycle. Detection of GFP protein is indicative of Tat-mediated transactivation of the HIV promoter and the expression of early viral gene products such as Nef, Tat and Rev. Despite the direct correlation between the overall level of GFP (early gene) and p24 (late gene) expression, distinct populations of GFP expressing MDM were detected. Interestingly, evidence for host cells exhibiting phenotypes such as persistent viral producer, GFP hi p24 lo, or GFP- HIV DNA+ could all be found within the same long-term culture. In particular among the infected macrophages, the GFP hi p24 lo group increased from 0.7% in day 7 cultures to 9.9% at days 60–78 postinfection. The presence of this latter population in which early gene products such as Nef is produced, but late gene expression is reduced, suggests that the host cell and/or virus possess mechanisms for down-regulating viral particle production. Elucidation of these mechanisms will be important for developing better therapies, particularly those aimed at inhibiting HIV transcription.
MATERIALS AND METHODS
Cell culture and infection.
The pSF162R3 Nef+ viral stocks were generated by transfection of proviral plasmid DNA into HEK293T cells as described previously [14
]. Monocytes were purified from buffy coats of normal donors by ficoll and percoll gradient centrifugation as described previously [14
]. 107 monocytes were seeded onto T-25 flasks and allowed initially to differentiate for 2–3 days in RPMI1640 supplemented with 10% human AB sera (Cambrex, Rockland, ME) and 20% FBS (Invitrogen, Carlsbad, CA) and then subsequently in the same medium without human sera (20% RPMI1640) for an additional 5-4 days before infection. No exogenous cytokines were added. On day 7 post-differentiation the medium was removed from the monocyte-derived macrophages (MDM) and replaced with 1 ml of pSF162R3 Nef+ viral supernatant (1–2 µg p24) in 1 ml of 20% RPMI1640 for 5 h. The inoculum was removed, monolayer washed and replaced with 8 ml of 20% RPMI1640. All medium was exchanged every 3–4 days and clarified supernatants frozen at –70°C for p24 ELISA (Beckman-Coulter, Miami, FL).
Fluorescence and confocal microscopy and quantitative image analysis
MDM were grown and immunostained for fluorescence or confocal microscopy and analyzed as described previously [14
]. In addition, an anti-Nef monoclonal antibody ARP3109 was used at 1:100 (Drs. Jane McKeating and Christine Shotton, and the Centralized Facility for AIDS Reagents, NIBSC, UK). The mean fluorescent intensity (MFI) of GFP was determined by using the freehand tool in the Image J software (NIH, Bethesda, MD) to draw a polygon around the circumference of the macrophage. The average background fluorescence intensity, as defined by GFP– cells in the same image, were subtracted from each GFP+ cell. The same procedure was used for calculating the MFI of p24 staining. A total of 438 cells for day 7 and 375 for day 60–78 postinfection were quantitated.
Flow cytometry and laser ablation
MDM monolayers cultured for at least 30 days were incubated with PBS-5 mM EDTA for 10–15 min and then gently scraped (Fisher Scientific, Pittsburgh, PA) to detach the cells and resuspended in 20% RPMI1640 with 5 mM EDTA. Before sorting, cell suspensions were filtered through a 35 mm nylon mesh (BD Biosciences, Palo Alto, CA) and propidium iodide added to gate out dead cells. Sorting was performed on the MoFlo Sorter using a 100 µm nozzle and the sort single 1 mode. The sorted GFP- macrophages were plated onto a small area of 35 mm laser capture dishes (PALM) in 
500 µl of medium for several hours to allow the cells to adhere and then the volume of the medium was increased to 2 ml. For laser ablation the contaminating GFP+ MDM were marked and catapulted using the following settings on the P.A.L.M. Microlaser (Bernied, Germany): energy-cut of 60; energy-lpc of 86; focus-cut of 75 and focus-lpc of 92. The monolayers were thoroughly washed with PBS before the generation of cell lysates. Depending on the yield of macrophages post sorting, 1x105-4x105 cells were used for cell lysates.
Quantitative real-time PCR
Primers and probes were designed in a conserved region of the pol gene as indicated by the base pair position on the HXB2 reference genome AQZ as given in the primer label; forward primer, RT-F-3698, 5'TGGGTACCAGCACACAAAGG; reverse primer, RT-R-3850 5'ATCACTAGCCATTGCTCTCCAAT; RT probe 3720, 5'ATTGGAGGAAATGAAC. The Taqman MGB probe carried the 5'VIC label (Applied Biosystems, Foster City, CA). The design of the ß-globin primer/probe set was based on Townson et al. [15
]. The ß-globin probe was labeled with 5'6-carboxy fluorescein (FAM), and 3'eclipse quencher (Eurogentec, San Diego, CA). All primers were used at a final concentration of 900 nM (Invitrogen). PCR amplification was performed in triplicate in 25 µl reactions with 2µl of cell lysates (1000/2000 cells/µl) 2X Taq master mix (Eurogentec) and 250 nM of probe. The kinetic PCR reaction conditions were 1X at 50°C, 2 min; 1X at 95°C, 10 min; 50X at 95°C, 15 s, 60°C, 1 min on the ABI Prism 7000 Sequence Detection System (Applied Biosystems). The method of Alu-LTR PCR was performed as reported [15
] except that only the Alu 1 primer (10 pmols) was used in the first reaction with primer SF162-LM667 (10 pmols), 5'ACT-GAT-GCC-ACG-TAA-GCG-AAA-CTC-TGG-CTA-GCT AGG-GAA-CCC-ACTG under the following reaction conditions: 0.6 units of AmpliTaq Gold (Applied Biosystems), 1X, 95°C, 10 min; 15X, 94°C, 20 s; 60°C, 20 s; 72°C, 2.5 min. Amplification of single-strand viral DNA primed by the LM667 oligonucleotide during the first PCR step can give rise to template for the kinetic reaction and lead to an overestimate of the integrated DNA copy number [16
]. To correct for this, first step PCR reactions that contained only the LM667 primer were also set up. One-tenth volume (2.5 µl) of the first reaction was then promptly amplified as given above by kinetic PCR with primers AA55M T-60, 5'GCT-AGA-GAT-TTT-CCA-CAT-TGA-CTA-AAA-GG and Lambda T-61, 5'ACT-GAT-GCC-ACG-TAA-GCG-AAA-CT to generate a product of 142 bp that was detected by the LTR-FL, Vic labeled probe [16
]. Lysates of serially diluted ACH2 cells that harbor a single proviral genome were used to generate standard curves. After quantification by the relative standard curve method using the ABI PRISM 7000 software, the copy number obtained from the LM667 reactions was subtracted from the values determined in the corresponding Alu+LM667 samples. The real-time assay for total and integrated HIV DNA was sensitive to 10 copies per 105 cells.
Statistical analysis
The Prism 4 (GraphPad software, San Diego, CA) statistics program was used to determine the P values for experiments involving two comparisons with an unpaired Student’s two-tailed t-test. Significance was determined by P values less than 0.05.
RESULTS
Long-term HIV-infected human macrophage culture model
To study the mechanisms of HIV-1 persistence in macrophages, we developed an in vitro model based on the long-term culture of primary human monocyte-derived macrophages (MDM) infected with the M-tropic EGFP-tagged reporter virus, pSF162R3 [14
]. MDM could be kept in viable culture for up to 78 days postinfection (Fig. 1
). To assess the expression of early and late viral gene products like Nef and the capsid protein, p24 respectively on a single-cell basis, MDM cultures were immunostained and analyzed by fluorescence and confocal microscopy. At day 10 postinfection, a direct correlation between the expression of GFP and Nef was found (Fig. 1A)
. This was not unexpected as both genes are expressed from the viral promoter via a bicistronic mRNA [14
]. Nef expressed in the context of the pSF162R3 reporter virus is functionally active in that GFP+ MDM have lower levels of the HLA-A2 molecule on their surface, as a result of the receptor down-regulation activity of Nef [14
]. Therefore, GFP can serve as a surrogate marker for Nef expression. In cultures examined at day seven post-infection, 100% of the GFP+ MDM co-expressed p24 antigen (Fig. 1B)
. However, in long-term cultures, a subpopulation of GFP+ MDM exhibited a discordant phenotype between the level of GFP and p24 antigen expression (Fig. 1B
, arrows). In these cultures no significant viral spread occurred after 3–4 weeks as indicated by the lack of increase in GFP+ MDM. However, extracellular particle secretion was detected throughout the culture period, although p24 levels varied with the donor cells and the number of GFP+ infected MDM (Fig. 2
, top panel, 2% GFP+). To quantitate the distinct infected subpopulations in four different donors (two at each time point), the GFP expressing MDM were divided into three statistically different groups (P<.0001), low (1–20), medium (21–40) and hi (41+) representing the GFP mean fluorescent intensity (MFI) (Fig. 2)
. For each GFP+ cell, the MFI of p24 expression was also determined and subdivided into low (1-20), medium (21-40) and hi (41+). In day 7 cultures, 72% of the infected MDM were GFP Lo, 24% GFP Med and only 3% fell into the GFP Hi group (Fig. 2)
. The majority of GFP+ MDM at day 7 (59%) expressed low levels of p24 that could reflect a stage before the peak of virus replication. In this regard, an analysis of the cultures at day 16, near the peak of infection showed only a small decrease in the GFP lo fraction to 55% (data not shown). Interestingly, by D78 post-infection, shifts in the population of GFP expressing MDM were observed. The GFP lo group was reduced to 44% and within this pool the p24 lo fraction decreased from 59% at day 7 to 33% at day 78. The GFP Med fraction rose modestly from 24 to 29%, while the largest increase was observed in the GFP hi group from 3% at day 7 to 27% at day 78 (Fig. 2)
. These results suggest that there was a shift with time to higher levels of GFP expression. Although there was a direct correlation between the level of GFP and p24 expression at both time points (Spearman r=0.99, Day 7 and r=0.98, D78), in long-term cultures there was a large increase in the percentage of GFP hi p24 lo MDM to 9.9% from 0.7% in early cultures. The EGFP gene is under the control of the viral promoter and therefore, the expression of GFP is a downstream indicator of Tat-mediated transactivation function. The presence of the GFP hi p24 lo population suggests that transcriptionally active HIV-infected MDM that produce abundant early proteins such as Nef, but little late gene products for viral particles can persist in long-term cultures and may represent a type of latent infection not described previously for primary cells.
 View larger version (45K): [in a new window] |
Figure 1. Confocal and fluorescence microscopic analyses of MDM infected with pSF162R3 Nef+ in short or long-term cultures. (A) Confocal image of infected MDM at day 10 postinfection. Cells were fixed, permeabilized and immunostained for viral antigens with monoclonal antibodies against p24 (red) and Nef (blue) and labeled with goat anti-mouse Alexa-568 and goat anti-rat Alexa-633 secondary antibodies respectively. (B) Infected MDM monolayers in T-25 flasks were fixed and stained as above at day 7 or 78 postinfection for p24 antigen expression. The purple arrow indicates an infected MDM that has the GFP hi p24 hi phenotype, while the yellow arrows point out macrophages that express variable levels of GFP and p24.
|
 View larger version (15K): [in a new window] |
Figure 2. Quantitation of distinct HIV-infected subpopulations in long-term cultures. In the top panel, a representative donor is shown in which extracellular p24 antigen levels were quantitated by ELISA on supernatants from long-term cultures. Scatter plots relating the mean fluorescence intensity (MFI) of GFP and p24 expression for each individual cell that was quantitated are shown (day 7, n=438 and days 60–78, n=375). The Spearman r, a measure of the degree of correlation between two values was 0.99 for day 7 and 0.98 for day 60–78. The GFP+ population (after background subtraction) was divided based on MFI into lo (1–20), med (21–40) and hi (41+). These three groups were then subdivided based on the MFI of p24 expression into lo (1-20), med (21-40) and hi (41+). The percentages of the individual populations are indicated. The GFP and p24 groups were highly statistically different with P values <.0001.
|
 View larger version (20K): [in a new window] |
Figure 3. Two-step purification of the GFP negative population from pSF162R3 Nef+ infected MDM cultures. (A) The panels show the GFP (left column) and propidium iodide (right column) histograms of the pre-sorted infected MDM population (top), the post-sorted GFP+ fraction (middle), and the GFP- pool (bottom). The purity for the GFP- fraction was typically 99%. However, upon examination of the sorted GFP- cells by microscopy, a few remaining GFP+ MDM were detected. (B) Post-sort GFP- MDM were plated on polyethylene-naphtalate dishes and laser ablation was used to catapult any contaminating GFP+ MDM from the monolayer (yellow arrows). Staining of the cell membranes with wheat-germ agglutinin-Alexa 594 (red) allowed visualization of the GFP- macrophages.
|
 View larger version (28K): [in a new window] |
Figure 4. Quantitation of HIV DNA in sort-ablated GFP negative MDM. Top panels: Agarose gel analysis of amplimers from PCR assays detecting total, integrated and ß-globin gene for determination of cell equivalents. The amplimer sizes for total, integrated and ß-globin DNA were 149, 142 and 100 bp respectively. Lanes 1, 6, 12, 100 bp ladder; lane 2, 7, 13, ACH2 cells (2x105); lane 3, 10, 14, sort-ablate donor 1 (SA-1; 1x105cells); lane 4, 11, 15, sort-ablate donor 2 (SA-2; 1x105 cells); lanes 5, 8, 16, no template control and lane 9, no viral primer control. In the middle panels are shown the real-time amplification curves for the ACH2 diluted standards from 2 x105 to 2 x 102 copies per reaction. The graphs below represent from left to right, the total and integrated DNA copy number per 1000 cells detected in chronically infected and in the sort-ablated GFP- MDM samples. The R2 values for total, integrated and ß-globin were 0.984, 0.996, and 0.997.
|
 View larger version (47K): [in a new window] |
Figure 5. Reactivation of GFP in sort-ablated GFP negative MDM. (A) GFP- MDM obtained post-sorting were plated on laser capture dishes (1x105 cells/dish) and submitted to laser ablation to remove any contaminating GFP+ cells (the frequency was low 5x10–5). The dishes were thoroughly washed and maintained in growth medium for one week and inspected for any remaining GFP+ MDM. IL-4 (10 ng/ml) or no drugs were added and the dishes inspected every 4–7 days. GFP+ cells were detected at day 7 post-drug addition and images of two (no drug) or three different regions (IL-4) of the plate were taken at day 10 post-drug addition (4X magnification). (B) Immunostain of reactivated sort-ablated GFP- MDM for p24 antigen. Top panel, phase contrast, middle panel, GFP epifluorescence (green) and third panel, anti-p24 (red) staining performed as given in Fig. 3
. Arrow designates a cell that is GFP+ p24lo.
|
The ability to detect HIV-infected macrophages using GFP has provided new insights into the relationship between the level of Tat-mediated viral transcription and early and late viral protein expression in macrophages. In this in vitro model using primary MDM, distinct HIV-infected subpopulations coexist within the same culture. Of particular interest were the persistent producers, the GFP hi p24 lo, and the GFP– DNA+ MDM. In the long-term cultures there was little viral spread and hence the observed populations represent the types of infected MDM present at this late time point. MDM that have the persistent producer phenotype continue to release viral particles with time and in this regard, a recent report using a recombinant HIV in which viral transcription is regulated by the presence of doxycycline, demonstrated that only infected cells with transcriptionally active provirus release extracellular p24 antigen [13 ]. In our MDM cultures there was a shift with time from day 7 to 78 to higher GFP expression levels as detected by the increase in mean fluorescent intensity (Fig. 2) that may be indicative of a positive feedback loop stimulating virus production. However, the presence of a significant population of GFP hi p24 lo MDM suggests that high-level Tat-mediated transcription does not predictably translate into abundant particle production. This latter group of cells that appear to accumulate in long-term cultures, in which early, but not late gene synthesis is abundant, may represent a form of latent infection not previously recognized in primary cells. The expression of Nef in the absence of structural proteins would allow HIV to persist and use the functions of Nef, such as MHC-I [17 , 18 ] and CD86 [19 ] receptor down-modulation to mask the infected macrophage from elimination by adaptive immune responses. Further analyses of the GFP hi p24 lo population are required to determine whether the phenotype is the result of a defective genome and if they produce viral particles. Clearly, the development of drugs to inhibit the viral Tat protein would be invaluable in further reducing the damaging effect of HIV on the immune system of infected patients.
Several different pathways could explain the varied levels of GFP expression observed in long-term HIV-infected MDM cultures. Mature blood monocytes are a heterogenous population of cells [7 ] that when differentiated vary in their potential to be infected by HIV. Viral replication may induce the expression of Tat transactivators such as CycT1 [20 ], but the degree of response from the host macrophages in our long-term cultures appears to differ on a cell by cell basis and therefore, this factor alone does not determine the ultimate level of viral transcription. Host cell properties such as the activation state likely have a role in determining the threshold for viral transcription levels. A recent study would suggest that the different GFP expressing MDM in our long-term cultures result from random fluctuations in Tat levels [21 ]. Distinct populations of Jurkat T-cells transduced with an LTR-GFP-IRES-Tat cassette were sorted into GFP bright, mid, dim, and off pools and followed in culture with time. The clonal dim pool derived from a single, proviral integration, gave rise to all four GFP phenotypes [21 ]. In our model that uses non-dividing primary cells and therefore the populations are not clonal, there was a shift with time that resulted in a decrease of the GFP lo MDM group and an increase in the GFP med and hi pools that may provide evidence in support of the idea that cells expressing low levels of Tat are sensitive to feedback loops (random or otherwise) that would alter its activity. The site of proviral integration can influence basal LTR activity [22 ] that in turn can affect the level of Tat-mediated transactivation [23 ] and could also potentially explain the GFP phenotypes observed in our long-term MDM cultures. Lastly, the GFP+ p24 hi MDM that account for 6.6% of the GFP+ cells at day 7 and increase to 20% by day 78, could in addition to representing persistent producers, reflect cells in which intracellular virions have accumulated. In this regard, it has recently been shown that intracellular virions in macrophages are infectious and can be transmitted to susceptible cells [13 ].
Whether macrophages actively sequester virions for cell-to-cell dissemination or particle accumulation is a result of the assembly process is not yet known. However, this fact combined with the observation that a subpopulation of macrophages can produce and release viral particles at a high and persistent level in the absence of significant cell death, reveals the advantages of this host cell as a viral reservoir. To begin to explore whether these cells can also serve as a latent reservoir similar to that defined for T-cells, as a provirus (es) integrated into the host genome in which there is an absence of unspliced HIV-1 RNA and no viral particles, but production can be reactivated by stimuli that activate cells [5
], we developed methodologies to isolate and purify the GFP- MDM population. The frequency of GFP– integrated DNA+ MDM in long-term cultures was 10–2, at least one to two logs higher than that reported for resting CD4+ T cells [24
]. In chronically infected MDM cultures total DNA was very high at 6–7 copies per cell. The total DNA measurement includes circular, integrated and unintegrated forms. It is now recognized however that circular forms of HIV DNA are stable intracellularly and are lost largely through cell division [25
26
27
]. As mature macrophages do not divide, HIV episomes are therefore likely to persist for long periods. Recombinant forms of HIV episomes were shown to persist for four months in the monoblastoid cell line U937 [28
], while a more recent report found that 2-LTR circles persisted in macrophages out to 21 days postinfection [29
]. The high number of integrated copies observed in chronically infected MDM, may be related to the presence of numerous multinucleated giant cells in such cultures. The current paradigm to explain the generation of latent HIV-1 in vivo includes both features of the host cell and properties of HIV-1 replication. HIV-1 does not induce a latent state of infection in vivo, but rather the virus gets trapped by the biological program of cycling T cells that results in the generation of memory cells with long-lived quiescent properties [5
]. Investigations into the mechanisms of viral latency have employed the promonocytic U1 [30
], promyelocytic OM-10.1 [31
] and T cell ACH-2 lines [32
]. Several pathways leading to latency have been identified. The cloned cell lines harbor proviruses with mutations in Tat or TAR suggesting that disruption of the viral transactivator or interactions with it and cellular factors can lead to a latent state of infection [33
34
35
]. Virus production can be reactivated with phorbol esters or TNF-
, factors [28
, 32
] that stimulate NF-
B translocation to the nucleus and thereby act independently of Tat [36
]. In contrast, latently infected T-cells in vivo contain replication competent integrated HIV-1 genomes [4
] indicating that current models do not completely reflect events that occur within patients. A subset of the GFP– DNA+ MDM in our long-term cultures harbor latent replication competent genomes at a frequency of 10–4, which is two to three logs higher than that reported for primary latent CD4+ T-cells (10–6 to 10–7) [4
, 24
]. Interestingly, IL-4 in contrast to phorbol ester or LPS was most effective at reactivating latent replication competent genomes in macrophages. This result supports prior findings that IL-4 enhances HIV replication [37
38
39
] and that NF-B independent pathways contribute to the viral life cycle.
The viral burden contained in tissue macrophages of infected patients remains unknown and consequently, the putative role of these cells in contributing to the latent pool is largely overlooked. If we use the data from this in vitro study for the purpose of discussion, the high frequency of latently infected macrophages observed would present a formidable barrier to treatment. Early studies on brain tissue from AIDS patients with encephalopathy, detected HIV-infected perivascular and parenchymal macrophages [40 , 41 ]. However, a definitive study in one SIV-macaque model, demonstrated that the perivascular and not the parenchymal macrophage is the predominant cell-type infected [42 ]. This insight was critical because there are large differences in the turnover rates of these two populations with perivascular cells having a half-life of months, vs. years for parenchymal macrophages [43 44 45 ]. The ability of anti-retrovirals regimens, particularly those containing a drug that penetrates the central nervous system, to suppress plasma viremia suggests that indeed both infected T-cells and macrophages turnover [3 ]. However, if treatment is electively terminated, virus replication resumes in as little as four to six weeks, often to pre-treatment levels [1 ]. Both latently infected resting CD4+ T-cells and macrophages could contribute to the resurgence of virion production. In this regard, studies in a SIV/macaque model in which CD4+ T cells are rapidly depleted, found that viral production from macrophages could sustain high levels of plasma viremia [46 ]. Furthermore, it will be important to determine if infected macrophages that produce early gene products, but no viral particles are present in infected patients, as this would have major implications for both drug therapy and vaccine development.
The herein described long-term culture model in which distinct infected populations can be isolated, and studied independently will be useful in uncovering the viral and host cell molecular mechanisms that regulate HIV-1 replication and entry into latent states in macrophages and lead to further insights into the dynamics of this host-pathogen interaction.
ACKNOWLEDGEMENTS
This work was supported by NIH Grants R01 AI057007 to S. Gartner, NIH/NIDA 1 K08 DA16160 to C. A. Pardo and the Center for AIDS Research Grant 1P30 AI42855. We thank the NIH AIDS Research and Reference Reagent Program NIAID, NIH, anti-p24, #183-H12-5c, from Bruce Chesebro and Kathy Wehrly, and Jeff Rothstein for use of the Nikon E2000U microscope.
Received February 28, 2006; revised June 6, 2005; accepted June 12, 2006.
REFERENCES
B Nature 339,70-73[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  M. Q. DeGottardi, S. K. Lew, M. Piatak Jr., B. Jia, Y. Feng, S. J. Lee, J. M. Brenchley, D. C. Douek, T. Kodama, J. D. Lifson, et al. Comparison of Plasma Viremia and Antibody Responses in Macaques Inoculated with Envelope Variants of Single-Cycle Simian Immunodeficiency Virus Differing in Infectivity and Cellular Tropism J. Virol., January 1, 2008; 82(1): 321 - 334. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. J. Montaner, S. M. Crowe, S. Aquaro, C.-F. Perno, M. Stevenson, and R. G. Collman Advances in macrophage and dendritic cell biology in HIV-1 infection stress key understudied areas in infection, pathogenesis, and analysis of viral reservoirs J. Leukoc. Biol., November 1, 2006; 80(5): 961 - 964. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
