| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print August 31, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* AIDS Immunopathogenesis Unit and
Laboratory of Clinical Immunology, San Raffaele Scientific Institute, Milan, Italy; and
Vita-Salute San Raffaele University, School of Medicine, Milan, Italy
1 Correspondence: P2-P3 Laboratories, DIBIT, Via Olgettina n. 58, Milano 20132, Italy. E-mail: poli.guido{at}hsr.it
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONPRIMARY MONOCYTES AND MDM:...CELL LINES AS MODELS...ANTIRETROVIRAL AGENTS AND MP...CONCLUSIONSREFERENCES |
|---|
Key Words: mononuclear phagocytes • latency • cytokine • differentiation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONPRIMARY MONOCYTES AND MDM:...CELL LINES AS MODELS...ANTIRETROVIRAL AGENTS AND MP...CONCLUSIONSREFERENCES |
|---|
In addition to CD4+ T cells, productive as well as latent HIV-1 infection of tissue macrophages such as brain macrophages and microglia has been clearly documented in infected individuals [13
14
15
]. Simultaneously, the possibility of infecting productively in vitro monocyte-derived macrophages (MDM) has been demonstrated by several laboratories [16
17
18
]. In more recent years, interest has grown in MP as long-lived cellular reservoirs of viral persistence [19
20
21
22
]. Although the general infectious mechanisms of T lymphocytes and macrophages are similar, peculiarities also exist, and characterize the replicative cycle of the virus in these two cell types that are likely relevant for the pathogenesis of infection in vivo. In this regard, one of the striking differences between MP and T cell infection is the relative resistance of MP to the cytopatic effects of HIV-1, both in vivo and in vitro [20
, 22
]. CD4+T cells die upon acute infection after a few days in culture [9
], a phenomenon reflected in a sharp peak of virus replication rapidly fading away. This may represent at least one of the cytopatic mechanisms of HIV infection in vivo leading to CD4+T cell depletion and immunodeficiency [4
, 23
, 24
]. While no evidence of significant depletion of MP occurs in infected individuals, the half-life of infected macrophages is substantially longer than that of T cells. Infected alveolar macrophages, for example, have an estimated half-life of 
2 months; whereas infected microglia can survive for several years [25
, 26
]. As a result, these cells continue to accumulate and archive replication-competent HIV-1 for prolonged periods of time, even in patients receiving HAART [27
]. As with T-cells, macrophage reservoirs are established soon after primary infection [28
], whereas early (prior to seroconversion) initiation of HAART does not prevent the establishment of both cellular types of viral reservoirs [29
, 30
, 31
].
Studies of the role of MP in HIV infection have been hampered by technical and logistical constraints. The distribution of these cells is widespread in all tissues and organs, including sanctuary tissue sites such as brain and testes [31 , 32 ]. In addition to being difficult to isolate without inducing adherence-mediated activation and differentiation, the recovery of infected tissue macrophages is often very inefficient [33 ]. For these reasons, most investigators have used primary MDM as a reference model system to dissect out the morphological and functional peculiarities of productive and latent HIV infection in this cell type. In addition to MDM, myeloid cell lines at various degree of differentiation along the MP lineage have been heavily studied as complementary or alternative models to MDM or other tissue macrophages.
Over the years, each of these models has provided unique sets of biological properties linked to efficient or inefficient virus entry, replication, and assembly, eventually, favoring viral persistence or productive infection. In this article, we will discuss the essential features of some of the most commonly used in vitro models to investigate HIV latency and replication in MP. Although primary MDM are likely to better reflect HIV-1 infection of macrophages in vivo, cell lines offer the advantage of being readily available, easy to synchronize, of undergoing unlimited expansion, and of not being subject to inter-patient variability. Thus, the combined use of these two models is likely to highlight complementary aspects of virus infection and replication in MP.
|
PRIMARY MONOCYTES AND MDM: A SPECTRUM OF VIRAL RESERVOIRS |
|---|
|
TOPABSTRACTINTRODUCTIONPRIMARY MONOCYTES AND MDM:...CELL LINES AS MODELS...ANTIRETROVIRAL AGENTS AND MP...CONCLUSIONSREFERENCES |
|---|
32-base pair deletion in CCR5 [37
, 38
]. The ability of HIV-1 to replicate in MP is dependent, among other factors, on the stage of cellular differentiation and levels of expression of CD4, as well as on a number of post-entry factors that influence and restrict viral replication [39
40
41
42
]. In contrast to the common belief that monocytes are not infected/infectable by HIV-1, several studies have reported that circulating monocytes are indeed infected in vivo and can serve as important reservoirs of HIV-1, especially in patients with late-stage disease and opportunistic infections [43
44
45
]. In this regard, recent studies suggest that a subset of CD16+ monocytes may be particularly susceptible to HIV-1 and it is dramatically expanded in the peripheral blood of infected individuals, particularly during progression to AIDS [46
, 47
]. CD16+ monocytes share phenotypic and functional characteristics with tissue macrophages and dendritic cells (DC), which suggests that they represent a more advanced stage of MP differentiation than CD16– monocytes [46
, 48
, 49
]. It has been proposed that the CD16+ subset of monocytes may be a preferential target for HIV-1. However, the contribution of the CD16+ monocyte subset to HIV-1 pathogenesis is largely unknown [20 ]. In contrast, monocytes from healthy donors are usually refractory to HIV-1 infection and/or productive replication unless they are induced to differentiate into MDM by in vitro cultivation for 5 to 7 days [41 , 50 , 51 ]. The reason(s) for the apparent discrepancy between these in vivo and in vitro findings is unknown, but it may be related to either activation-induced resistance during monocyte isolation and purification or culture-induced changes in CCR5 expression. Freshly isolated donor monocytes express relatively high levels of CXCR4, which fall to near-undetectable levels after 24 h of cultivation [41 ]. In contrast, CCR5 expression is barely detectable immediately after isolation but increases significantly during cultivation leading to up to >30% of positive cells after 48 h of cultivation [22 , 41 ]. Addition of MP-activating and differentiating cytokines such as GM-CSF or M-CSF to monocyte cultures leads to an even greater increase in CCR5 expression [20 , 52 ].
Patterns of virus replication in MDM
Several studies have shown that the kinetics of HIV-1 replication in MDM are similar to those observed in tissue macrophages, as reviewed in [20
]. After infection with primary or laboratory-adapted viruses that preferentially replicate in MDM rather than in T cells (early defined as "macrophage-tropic" strains of HIV-1, usually R5), the expression of HIV-1 increases linearly over time, reaching a peak level 14 days after viral inoculation with continued high-level virion production for at least 60 days postinfection [53
]. During this period, unspliced and multiply spliced HIV RNAs also increase linearly up until day 10, reaching levels of 1.5 x 108 and 3 x 105 copies/105 cells, respectively, followed by a plateau, with minimal variation detected throughout the remaining culture period [54
]. Interestingly, the replication kinetics of HIV-2 in MDM are distinctly different from those observed for HIV-1. HIV-2 infection of MDM has been reported to occur with lower replication levels and a transient burst of virus production 2 days postinfection, followed by a state of apparent latency; moreover, addition of bacterial LPS to HIV-2 infected MDM induced a switch from latent to productive infection [55
]. When compared with HIV-1, primary HIV-2 isolates have been shown to use a broader range of chemokine coreceptors for entry, while the in vivo infection is characterized by a slower disease progression, low viral load, and reduced transmission rates [56
, 57
].
Differential susceptibility of MDM to R5 and X4 HIV-1 infection
Although peripheral blood monocytes and tissue macrophages express both CD4 and CXCR4, they do not usually support efficient X4 infection. A potential explanation for this phenomenon is that, while CXCR4 is functional in terms of receptor-mediated signaling, it does not support fusion with envelope of X4-tropic variants of HIV-1 [58
]. Alternatively, it has been suggested that a blockade in X4 infection may occur at a later, post-entry stage in MDM, possibly at the level of the pre-integration complex (PIC) and its nuclear import [59
]. When studied in vitro, T cell line adapted HIV-1 isolates, such as the prototypic X4 strain LAI/IIIB, usually do not productively infect MDM in spite of CD4 and CXCR4 expression [58
]. In partial contrast, MDM infection with primary X4 isolates obtained from patients at late-stage HIV-1 disease, although inefficient, has been shown to occur [58
]. Infection with either R5 or dualtropic R5X4 HIV-1 isolates leads to substantially higher levels of viral replication [60
].
Modulation of HIV-1 infection and replication by LPS and other factors controlling HIV-1 transcription in MDM
Macrophages are responsive to a wide variety of positive and negative stimuli with their susceptibility to infection profoundly influenced by the type of stimulus. Exogenous stimuli such as bacterial LPS are powerful inhibitors of acute R5 infection in MDM culture in vitro, due to a rapid down-regulation of CCR5 mediated, at least in part, by the release of endogenous CCR5 binding chemokines [61
, 62
]. This down-regulation is a result of altered recycling of the CCR5 co-receptor rather than a supression of CCR5 mRNA [63
]. Interestingly, even when LPS is added to cultures 3 days after infection, there is a decrease in virion release, which suggests that LPS may also act at a point of the viral life cycle beyond the early events of virus binding and uncoating [61
]. However, LPS can also lead to the induction of viral production by integrated HIV proviruses via activation of NF-kB [65
], and enhancement of the susceptibility of MDM to productive X4 infection has also been reported [66
]. In addition to LPS, engagement of Fc
R receptor by IgG on the surface of MP leads to cell activation and strong inhibition of acute R5 HIV-1 infection and replication in MDM [64
]. Both R5 and X4 viruses are blocked and appear to be restricted after reverse transcription and nuclear translocation of viral DNA [64
].
HIV-1 transcription is controlled by cellular and viral factors that bind the HIV-1 long terminal repeat (LTR). A number of cytokines and chemokines have been reported to modulate HIV-1 replication in MP. TNF-
has been shown to increase viral replication in MDM in vitro through the activation of NF-kB [60
, 67
, 68
], although, like LPS, it may cause down-regulation of CCR5 and prevent acute infection of MDM [63
]. IFN- has been shown to inhibit acute infection of MDM by acting at an early step preceding reverse transcription, but also preventing transcription of integrated virus [61
]. IL-4 and IFN- are bi-functional cytokines that, like TNF-, can either increase or decrease viral replication, depending on their timing of stimulation vs. the infection [69
70
71
].
In addition to NF-kB, C/EBPs are transcription factors that regulate cellular genes involved in a broad range of different processes relating to the synthesis and release of cytokines, as well as immune cell differentiation and inflammation [72 ]. The 5' negative regulatory element (NRE) region of the HIV-1 LTR contains 3 C/EBP sites that bind C/EBPß [73 ]. These are required for proviral DNA transcription in infected MDM but not in CD4+ T-cells [73 ]. Two different proteins can be produced from the same C/EBPß mRNA; the larger 30–37 kD isoform of C/EBPß physically interacts with histone acetyltransferase (HAT) complexes, suggesting that it may affect HIV-1 provirus expression via chromatin remodeling [72 ]. This interaction leads to the recruitment of co-activators and enhanced LTR-mediated stimulation of viral transcription. Interestingly, the smaller isoform (16–20 kD) of C/EBPß has been shown to inhibit HIV-1 replication and LTR-mediated transcription [72 ]. In this regard, stimulation of monocytes with M-CSF during their differentiation into MDM leads to a marked increase in virus production, while GM-CSF leads to a dramatic reduction in both the post-transcriptional and translation activities of HIV-1 [74 ]. Of interest is the fact that, following HIV-1 infection, the smaller isoform is more abundant in GM-CSF but not in M-CSF, stimulated MDM [74 ]. It should be underscored, however, that GM-CSF can simultaneously activate different transcriptional pathways, such as JAK2/STAT5 and ERK-1/-2 and activator protein-1 (AP-1), these latter clearly related to activation of HIV transcription [75 76 77 78 ]. In addition, several protocols of MDM infection include either M- or GM-CSF as stimulants [52 , 75 ].
Thus, as for other stimulants, the ultimate effect on HIV infection and replication may vary according to the timing of the stimulation with respect to the state of infection and of concomitant signals.
Unique features of HIV-1 synthesis and assembly in infected MDM
Unlike CD4+ T lymphocytes, where virion assembly and maturation occurs exclusively at the plasma membrane, viral production in primary MDM also occurs, and sometimes predominates, in cytoplasmic vesicles with features of multi-vesicular bodies (MVB) or late endosomes [27
, 79
80
81
]. This process results in the accumulation and prolonged storage of HIV-1 virions. These virions remain infectious for several months and can rapidly infect CD4+ T-cells [27
]. This feature may play a central role in the dissemination of HIV-1 through a "virological synapse" formed between T cells and macrophages [27
, 82
]. This morphogenetic process in virion synthesis has suggested the model, whereby tissue macrophages act as "Trojan horses" of HIV infection, which hide the viral enemy away from the immune response and, potentially, from some antiviral agents [83
84
85
]. An understanding of the fine mechanisms controlling the preferential synthesis and accumulation of virions in MVB, and the transmission of these stored virions to T cells or other target cells, is crucial for the design of novel strategies aimed at inhibiting HIV-1 accumulation in infected macrophages.
Disadvantages of the MDM as model system for HIV infection
Although primary MDM remain a fundamental in vitro surrogate system for studying HIV infection and replication, this model is endowed with an intrinsic relatively high variability that may lead to conflicting results due to inter-donor factors such as, purity, and activation of the isolated cells, presence and number of contaminant lymphocytes (that are sometimes maintained for several days in culture together with maturing monocytes before removal), degree of plastic adherence, the presence or absence of human serum or of exogenous cytokines and of the type of culture medium. All of these variables can affect the ability of MDM to support productive HIV-1 infection, particularly when exposed to primary rather than laboratory-adapted HIV strains. For these reasons, immortal myeloid cell lines are complementary useful tools to investigate the feature of acute and chronic HIV infection in vitro.
|
CELL LINES AS MODELS OF MP-HIV-1 INTERACTION |
|---|
|
TOPABSTRACTINTRODUCTIONPRIMARY MONOCYTES AND MDM:...CELL LINES AS MODELS...ANTIRETROVIRAL AGENTS AND MP...CONCLUSIONSREFERENCES |
|---|
|
|
, IL-1ß in both cell lines, and of G-CSF and GM-CSF exclusively in Mono Mac 1 cells [87
, 88
, 91
]. LPS stimulation of Mono Mac 1 also results in increased adherence to plastic and development of irregular cell shapes with more extensive projections of the cell membrane [90
]. In addition, Mono Mac 6 cells also express high levels of CD18 and ICAM-1 [91
]. The main difference between Mono Mac cell lines relates to their differential expression of CD4, CCR5, and CXCR4. Approximately 60% of Mono Mac 1 cells display CD4 on the cell surface, while 100% of them expresses CCR5 [90 ]. Consequently, this line is susceptible to both R5 and dualtropic CCR- and CXCR4-dependent (R5X4) viruses but exhibits little, or no, susceptibility to infection by X4 strains of HIV-1 [90 ]. As observed for primary MDM, maturation of Mono Mac 1 leads to increased viral replication and a concomitant down-regulation of CD4 [90 ]. In contrast to Mono Mac 1, Mono Mac 6 cells do not express CCR5, and only <10% of them expresses CD4 [91 ]. As a result, this cell line is unable to support productive infection by R5 viruses. However, Mono Mac 6 cells are positive for CXCR4 expression, and varying levels of X4 HIV-1 replication have been described for up to 300 days postinfection [92 ]. Thus, these paired Mono Mac cell lines provide an excellent model for studying HIV-1 entry and replication in CCR5- vs. CXCR4-positive monocytic cells.
Monocytic THP-1 cell line
The THP-1 cell line was derived from a heterogeneous population of tumor cells cultured from the blood of a patient with acute monocytic leukemia [93
]. THP-1 cells display Fc and C3b receptors, without surface or cytoplasmic expression of Ig [94
]. Therefore, these cells have phagocytic capacity and are characterized by the presence of lysosomes and the ability to produce esterases [95
]. Upon cultivation, THP-1 cells maintain their monocytic characteristics for periods greater than 14 months [94
]. Recently, it has been demonstrated that the THP-1 gene expression profile differs from that of primary monocytes in that their three most abundant gene products (cathepsin G, neutrophil elastase 2 and proteinase 3) are typically expressed by primary neutrophils, rather than monocytes [95
].
When stimulated with phorbol esters such as the tumor-promoting agent PMA, THP-1 cells mimic MDM by becoming adherent to glass or plastic, exhibiting a macrophage-like morphology, and expressing macrophage differentiation markers, namely, the scavenger receptor A (SR-A), apolipoprotein E (apoE), and lipoprotein lipase [95 96 97 98 ]. Unlike monocytes, in which these surface markers are typically down-regulated during their differentiation toward MDM [99 ], this PMA-induced phenotype is associated with an up-regulation of CD14 and IL-1ß [95 ]. This phenotype, however, is not particularly stable and, after prolonged culture, the cells undergo dedifferentiation [94 ].
Relevant to HIV infection, THP-1 cells express low levels of CD4, CCR5, and CXCR4 [100 ]. Treatment with phorbol esters leads to reduced expression of CD4 but has no impact on CCR5 or CXCR4 expression [100 ]. The susceptibility of THP-1 cells to R5 HIV infection is controversial in that both positive [101 ] and negative [102 ] studies have been reported. In this latter study, however, once stimulated with PMA, these cells become highly permissive to HIV-1BaL in spite of a decreased expression of CD4 [102 , 103 ].
Several studies have shown that undifferentiated THP-1 cells can be productively infected with standard laboratory-adapted X4 strains, such as LAI/IIIB, whereas phorbol ester-differentiated cells are resistant to X4 viruses [91 , 104 105 106 107 ]. The reason(s) for these discrepant results are unknown, but they may be related to the inherent instability of THP-1 cells and associated variations in the levels of CD4 expression. In this regard, CD4 surface expression was found to decrease from 80 to 50% after 6 weeks of culture, reaching levels as low as <20% after 3 months [108 ].
Thus, THP-1 cells are particularly useful for studying macrophage-specific factors involved in the differentiation-dependent regulation of CCR5 and CXCR4 expression, and the impact of these processes on differential susceptibility of monocytic cells to R5 vs. X4 HIV-1.
THP-1 derived chronically infected cell lines
Infection of THP-1 cells, as observed with other myeloid and lymphocytic cell lines, leads to the survival of cells carrying integrated proviruses with different profiles of virus expression. Constitutively producing, restricted and completely latent chronically infected THP-1 cell lines have been early described as a multicellular model representing the different forms of viral latency and its inducibility by either demethylating agents or cytokines [109
]. In the case of THP-1 with restricted HIV expression, cellular transcription factors have been proposed to be responsible for the lower transcriptional activity than that observed in the acutely infected cells. In contrast, viral reactivation from latently infected THP-1 cells was induced by 5-azacytidine, suggesting that viral DNA methylation is involved in the maintenance of the viral latent state [109
]. However, it has also been shown that inhibition of DNA methylation by 3-deaza-adenosine analogs inhibits HIV expression in chronically infected THP-1 cells [110
].
Concerning chronically infected THP1 cell lines with a "restricted" phenotype, viral expression was induced by various agents, including LPS, PMA, TNF-, and GM-CSF, and superantigens such as staphylococcal toxic shock syndrome toxin-1 (TSST-1) and staphylococcal enterotoxin A (SEA), which were ineffective in the "truly latent" cell lines [65
, 109
, 111
] (Table 2)
.
Promonocytic U937 cell line and clones
The U937 cell line was established in 1974 from the pleural effusion of a 37-year-old male diagnosed with hystiocytic lymphoma [112
]. This cell line carries markers of the myelomonocytic lineage (CD13, CD15, and CD33), expresses mRNA for the c-myc oncogene, and is less differentiated than monocytic cell lines THP-1 and Mono Mac [94
, 113
] but more differentiated than the myelomonocytic cell line HL-60 (discussed below).
Undifferentiated U937 cells are exclusively susceptible to infection by X4 HIV-1 strains. When cellular clones, obtained by limiting dilution of U937 cells, were tested for susceptibility to HIV infection and replication, two patterns were identified and the cell clones were defined as belonging to either a "Plus" or "Minus" phenotype. Plus cell clones support high levels of X4 replication and relatively fast turnover kinetics, with peak replication between 2 and 3 weeks postinfection, whereas Minus clones support only low-level (sometimes undetectable by reverse transcriptase, RT, activity) replication with slower kinetics (peaking >4 weeks postinfection) [114
, 115
]. This difference in virus expression does not appear to be due to substantial differences in terms of CD4 or CXCR4 expression levels, as assessed by both Northern blot and flow cytometry [116
]. Cellular differentiation triggered by all-trans retinoic acid (ATRA) induced expression of CCR5 in both types of cell clones; however, only Plus clones have been shown to support the replication of R5 viruses in this experimental condition [117
]. Of interest, another differentiating agent such as Vitamin D3 did not induce CCR5 expression in either Plus or Minus U937 clones [117
], but it selectively up-regulated the kinetics and peak of X4 HIV-1 replication exclusively in Minus cells [115
]. Although the original phenotype of Minus clones was linked to the presence of a protease cleaving the p65 subunit of NF-kB [114
], this was proven not to be causatively linked to the inefficient level of replication observed in these clones and it was not repressed by Vitamin D3 [114
]. Thus, a post-entry, NF-kB independent restriction factor(s) overcome by Vitamin D3-induced differentiation appears to play a substantial role for efficient propagation of HIV-1. Plus—but not Minus—U937 cell clones have been shown to constitutively secrete TNF-, whereas its inhibition resulted in decreased efficiency in viral replication [118
]. Finally, Plus clones were shown to lack the IFN-R2 chain and to be unresponsive to the inhibitory effects of IFN- (but not of IFN-) on acute HIV-1 infection [119
].
Taken together, these findings suggest that the Plus and Minus clones of the U937 cell line may provide important tools for assessing the importance of differentiation-dependent factors in determining the relative efficiency or inefficiency of X4 HIV-1 infection. Comparative analyses of Plus and Minus variants stimulated with either Vitamin D3 or retinoic acid may shed new light on the cellular factors controlling the susceptibility to productive R5 HIV-1 infection.
U937-derived chronically infected cell lines (U1 and U33)
Cell lines displaying either constitutive (U33) [120
] or inducible (U1) [120
, 121
] levels of HIV expression have been obtained from the surviving population of U937 cells acutely infected with the X4 HIV-1LAI/IIB strain. These cell lines, particularly U1, have become prototypic models to delineate the regulatory effects of host determinants, including pro- and anti-inflammatory cytokines, class I and II IFNs, and other factors, including ultraviolet light and heat shock [121
122
123
124
125
126
]. Although originally interpreted as consequent to a defective Rev/RRE axis [127
], the relative state of proviral latency characterizing U1 cells has been functionally linked to a defective Tat/TAR interaction [128
]. Transfection of either Tat (but not Rev)-expressing plasmids or incubation with exogenous Tat protein indeed rescued viral expression from U1 cells [128
, 129
].
Among pro-inflammatory cytokines, TNF- [130
]—like phorbol esters—leads to increased HIV transcription and virus expression through activation of NF-kB, as also observed in primary MDM [68
] and PBMC [131
, 132
]. However, other molecules, such as IL-1ß, IFN-, and IL-6 induce HIV expression predominantly or exclusively via activation of MAPK. such as ERK-1/-2 and formation of active AP-1 complexes [77
, 133
, 134
]. Of interest, functional AP-1 DNA binding sites have been described in the LTR but also in an intragenic enhancer present in the coding region of HIV [76
]. While not having a direct effect on viral expression, a number of other factors, including glucocorticoid hormones [135
] and IL-10 [136
] have been shown to potentiate the effects of these pro-inflammatory cytokines.
In addition to responding to exogenous cytokines, PMA stimulation of U1 cells (as well as of OM-10 cells [137
, 138
]) leads to the secretion of TNF- and IL-1ß, acting in an autocrine fashion to up-regulate HIV expression [139
, 140
], as also observed in primary MDM [71
, 141
] and unfractionated PBMC stimulated with IL-2 [132
]. Anti-inflammatory cytokines such as IL-10, IL-4 and glucocorticoid hormones, when used at concentrations high enough to inhibit secretion of these cytokines, also inhibited HIV expression. In the case of IL-4, superinduction of IL-1 receptor antagonist (IL-1ra) was demonstrated to play a major role [140
], whereas TGF-ß as well as ATRA inhibited PMA-induced HIV expression independently from the endogenous release of TNF- [142
, 143
]. Both TGF-ß and ATRA were shown to exert either inhibiting or activating activities depending on the experimental conditions, and particularly the timing of stimulation as a function of that of the acute infection, in primary MDM [142
143
144
].
In addition to cytokines, IFNs have been intensively studied in U1 cells. IFN- has been shown to exert a "post-budding" effect, resulting in the inhibition of the release of virions from the cell surface [145
]. IFN- stimulation results in the up-regulation of HIV expression but also in the inhibition of PMA-stimulated virus production [124
]. This apparent dichotomy is likely explained by the accumulation of virions in MVB under the strong differentiating effect triggered by the combination of the two stimuli, i.e., PMA and IFN-, therefore reproducing the typical feature of HIV morphogenesis described in primary macrophages infected either in vitro [79
, 83
, 84
, 146
] or in vivo [79
]. Of interest, other stimuli, including urokinase-type plasminogen activator [147
, 148
] and CCL2/MCP-1 [149
], have been associated with this phenotype in U1 cells and primary MDM, respectively.
In conclusion, U1 cells (which resemble Minus U937 cell clones, at least in terms of morphology, growth kinetics, and banding pattern of NF-kB) represent an important tool in addition to primary MDM and acutely infected U937 to delineate which factors and signaling pathways are controlling HIV gene expression from latency in a differentiating MP background (Table 2) .
Promyelocytic HL-60 cell line
HL-60 cell line was established in 1977 from a single patient with promyelocytic leukemia [113
]. This cell line encompasses a highly diverse cell population, which, depending on the stimulus, can be induced to differentiate toward either granulocytes or MP [150
]. When stimulated with Vitamin D3, HL-60 cells differentiate along the MP lineage becoming positive for the enzyme -napthyl acetate esterase and acquire the capacity of mediating antibody-dependent cytotoxicity (ADCC) [113
]. Interestingly enough, maturation toward a monocytic phenotype is not associated with a loss in proliferative capacity [113
, 151
]. Phorbol esters induce HL-60 differentiation into a more macrophage-like phenotype, including plastic adherence and formation of prominent pseudopodia, while still retaining some of the properties typical of monocytic cells [151
, 152
].
In its undifferentiated form, HL-60 cells expresses CD4 and CXCR4 but not CCR5 [153
] and, therefore, are not susceptible to infection by R5 HIV-1 [154
]. However, stimulation with PMA induces expression of low levels of CCR5 [155
]. Interestingly, exposure of PMA-treated cells to IL-10 for 24 h leads to an even greater increase in CCR5 expression as previously shown in MDM [155
]. HL-60 cells are, however, susceptible to infection by the laboratory-adapted X4 HIV-1, such as NL4-3 [156
, 157
]. In this regard, it has been reported that initially infection with NL4-3 leads to only low-level virus production, as measured by cell-free RT activity in culture supernatants [156
, 157
]. However, a decrease in the cell viability and a concomitant increase of the percentage of infected cells have been reported after 15 days of cultivation [156
, 157
]. This change in the HIV-1 expression patterns has been correlated to the emergence of viral variant(s) with enhanced cytopathicity for HL-60, other myeloid cell lines, and primary MDM [154
].
HL-60 provides a unique tool for studying the cellular factors regulating the proliferation and differentiation of MP progenitors and for examining the differential expression of HIV-1 co-receptors on cells of the pre-monocytic lineage and its impact on virus susceptibility. As with THP-1, a potential disadvantage of HL-60 cells is their unstable phenotype. The maintenance of HL-60 in culture for periods of 3 to 18 months leads to oncogene amplification, a loss of granulocyte and monocyte lineage markers, and alterations in the cell line growth parameters, which suggests that the cells have undergone a phenotypic drift from a heavily granulated promyelocytic cell to a more undifferentiated agranular blast [150 ]. Interestingly, the actual loss of the promyelocytic phenotype occurs in the first month [150 ]. Whether these changes affect HL-60 susceptibility to HIV-1 infection is unknown.
Chronically infected HL-60 (OM-10) cell lines
As described for U937 cells, clonal derivatives of persistently infected HL-60 cells have been derived from the surviving population after acute infection with HIV-1LAI/IIIB [158
]. These cell lines, named OM-10 and OM-10.1, have served as useful models for investigating the influence of the cell cycle on HIV expression and which signaling pathways could lead to the up-regulation of otherwise latent HIV infection [137
]. Unlike U1 cells, OM-10 cell lines are characterized by oscillating levels of CD4 expression influenced by stimulation with TNF- [158
]. Furthermore, a pulse of TNF- stimulation led to the self-perpetuation of HIV expression by an autocrine TNF--dependent mechanism [159
]; this loop is further enhanced by the up-regulation of cell surface TNF receptors [137
] (Table 2)
.
|
ANTIRETROVIRAL AGENTS AND MP CELL LINES |
|---|
|
TOPABSTRACTINTRODUCTIONPRIMARY MONOCYTES AND MDM:...CELL LINES AS MODELS...ANTIRETROVIRAL AGENTS AND MP...CONCLUSIONSREFERENCES |
|---|
The U937 cell line has been used for studying both anti-viral efficacy and the cellular toxicity of drugs alone or in combination. As a matter of fact, AZT did not inhibit HIV infection in this cell line [163 , 164 ] because of low levels of intracellular phosphorylation [165 , 166 ], a necessary step required by NRTIs in order to become functionally active. U937 cells have proven useful in unraveling multiple complex interactions concerning phosphorylation. While PIs do not appear to influence phosporylation of NRTIs [167 ], ddC and AZT—but not ddI or d4T—caused significant reduction of 3TC phosphorylation [168 ], with AZT favoring and ddC inhibiting the metabolic conversion of ddI in its active form [169 ].
Conflicting results have been reported in U937 cells in terms of mitochondrial toxicity: d4T, but not ddI or AZT, significantly altered the levels and quality of mitochondrial RNA [170 ]; conversely, AZT decreased ATP production by mitochondria, causing dysfunction of cellular redox control and eventually loss of the mitochondrial DNA integrity [171 ].
Previous studies have indicated that the anti-malarial drug chloroquine (CQ) has anti-HIV-1 efficacy and can synergize with some antiretroviral agents [172
]. Indeed, d4T (10 µm) [173
] and ddI (1 µm) in combination with CQ (1 µm) and hydroxyurea (200 µm) [174
] inhibited HIV replication in U937 cells. A marked synergy has been observed between SQV (and other PIs) and mefloquine (MQ), but not CQ [175
]. PIs have been shown to protect U937 cells from apoptosis, possibly by acting on cellular proteases involved in this mechanism of cell death [176
]. The chronically HIV-1 infected U1 cell line has also been useful to dissect out which step of the HIV life cycle is influenced by antiviral agents. While AZT did not inhibit HIV expression, both IFN- [145
] and CQ [174
] inhibited viral expression in PMA- and LPS- or H2O2-stimulated U1 cells, respectively.
Finally, U937 cells have also been used in in vivo studies: SCID mice reconstituted with U937 cells [177
] represent an experimental model to address anti-viral strategies, such as those involving AZT [177
] or IFN- [178
].
Furthermore, it has been shown in THP-1 cells that PIs impair vitamin D3 bioactivation [179 ]. RTV has been shown to either up-regulate [180 ] or decrease CD36 expression in THP-1 cells [181 ], as a potential surrogate marker of the formation of atherosclerotic lesions in vivo [182 ]. HL-60 cells have been useful in documenting the myelotoxicity of AZT and other RTIs [183 ]. In particular, AZT increased the duration of the S phase and decreases that of the G1 phase of the cell cycle [184 ] and mediated the alteration of lipid metabolism in this cell line [185 ]. Finally, it has recently been reported that NFV and SQV, but not RTIs, affected human proteasome function in HL-60 cells [186 ].
Thus, myelomonocytic cell lines are becoming useful tools to compare and delineate the metabolic consequence of antiretroviral agents and of their combination and their overall impact on HIV infection and replication.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONPRIMARY MONOCYTES AND MDM:...CELL LINES AS MODELS...ANTIRETROVIRAL AGENTS AND MP...CONCLUSIONSREFERENCES |
|---|
As a viral reservoir, MP are highly heterogeneous and may show a differential susceptibility to HAART. In addition to an increased repertoire of entry inhibitors, future therapeutics are likely to target cellular factors that can suppress (or eliminate) residual viral replication, not only in memory CD4+ T-cells but also in MVB of infected macrophages. The availability of a broad spectrum of well-characterized myelomonocytic cell lines that could be used for the screening of new antiretroviral agents and determining how their therapeutic efficacy will likely be highly beneficial.
Received March 4, 2006; revised June 5, 2006; accepted June 5, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONPRIMARY MONOCYTES AND MDM:...CELL LINES AS MODELS...ANTIRETROVIRAL AGENTS AND MP...CONCLUSIONSREFERENCES |
|---|
