Monocytes are vitally important cells in the immune system, as they are the precursor cells to professional antigen-presenting cells (APCs), such as macrophages and DCs. These types of immune cells patrol the bloodstream and tissues, replenishing dying APCs or, in an infection, providing enough of these cells for the body to effectively combat an invading pathogen 5. Undifferentiated monocytes live for only a few days in the bloodstream. Upon differentiation or activation, the life-span of monocytes is significantly prolonged for up to several months 6.

There are two major subtypes of monocytes, those that are highly CD14-positive (CD14⁺⁺CD16^-) and those that are CD16-positive (CD14⁺CD16⁺). CD16⁺ cells make up only a small percentage (around 5%) of the total monocyte population, but they are characterised as more pro-inflammatory and having a greater role in infections than the CD14⁺⁺CD16^- cells 7.

Although monocytes express the required HIV-1 receptors and co-receptors for productive infection 89, they are not productively infected by HIV-1 in vitro. This is possibly due to an overall inefficiency in each of the steps required for virus infection, ranging from viral entry to proviral DNA integration 101112, but not due to a viral nucleocapsid uncoating defect 13. Recent studies have suggested a role for naturally occurring anti-HIV micro-RNA (miRNA) in suppressing HIV-1 replication in peripheral blood mononuclear cells or purified monocytes 14151617. This mechanism could allow for further studies utilising miRNAs as inhibitors of HIV-1 15. However, it has also been shown that HIV-1 is capable of suppressing some inhibitory miRNAs 16, which may reflect an evolutional interaction between HIV-1 and host factors. Further studies are required to understand this interaction and develop a therapeutic approach against HIV-1 infection using miRNAs.

Differentiation of monocytes into macrophages or DCs in vitro enables productive HIV-1 replication in the differentiated cells 141819. Based on current understanding, vaginal macrophages are more monocyte-like than intestinal macrophages and show increased HIV-1 susceptibility 20. Hence, some monocyte characteristics might be required for efficient infection, and these traits may be lost in fully differentiated tissue macrophages.

Given the role of monocytes in the immune system and in HIV-1 replication, a number of HIV-1 proteins have been shown to affect the biology of monocytes.

HIV-1 Tat-mediated transactivation of the viral promoter is essential for HIV-1 transcription 21. Exogenous recombinant HIV-1 Tat protein has been shown to increase monocyte survival through increased expression of the anti-apoptotic protein Bcl-2 22. Using an in vitro model of monocyte death mediated by TRAIL (tumour necrosis factor-alpha-related apoptosis inducing ligand), it has been shown that HIV-1 Tat encourages the survival of monocytes in situations where they would normally be cleared 22. Exogenous HIV-1 Tat has been shown to cause production of the cytokine interleukin (IL)-10 from monocytes in vitro 2324. Significantly increased IL-10 levels were also observed in HIV/AIDS patients compared with healthy controls 25. Furthermore, up-regulation of IL-10 production in HIV/AIDS patients has been correlated with increased levels of monocyte-secreted myeloid differentiation-2 and soluble CD14 25; both proteins are key molecules in the immune recognition of gram-negative bacterial lipopolysaccharide (LPS). Given that high levels of secreted CD14 have been associated with impaired responses to LPS 26, it has been proposed that the release of general immunosuppressant IL-10 by monocytes 27 facilitates the progression to AIDS 25.

HIV-1 Nef is a multifunctional accessory protein that plays an important role in viral pathogenesis 28. Retroviral-mediated HIV-1 Nef expression in primary monocytes and a promonocytic cell line inhibits LPS-induced IL-12p40 transcription by inhibiting the JNK mitogen-activated protein kinases 29. As an inducible subunit of biologically active IL-12, IL-12p40 plays a critical role in the development of cellular immunity, and its production is significantly decreased during HIV-1 infection 29. This study implicates the importance of HIV-1 Nef in the loss of immune function and progression to AIDS.

HIV-1 matrix protein (p17) regulates a number of cellular responses and interacts with the p17 receptor (p17R) expressed on the surface of target cells 30. Upon binding to the cell surface receptor p17R, exogenous HIV-1 matrix protein causes secretion of the chemokine monocyte chemotactic protein-1 (MCP-1, also known as CCL2) from monocytes 30. MCP-1 potentially increases monocyte recruitment to the sites of HIV-1 infection, increasing the available monocyte pool for infection by HIV-1; this recruitment may be of critical importance given the relatively low rate of infection of this cell type 101112.

HIV-1 and HIV-1-derived factors have been shown also to induce up-regulation of programmed death ligand-1 on monocytes in vitro 3132. This ligand, in complex with its receptor, programmed death-1, causes apoptosis of all T cell types 33 and a loss of anti-viral function in a manner similar to known immunosuppressive cytokines 34. Together, these studies suggest that HIV-1 can impair virus-specific immunity by modulating immuno-regulatory molecules of monocytes and T cells.

Of the studies discussed above, those involving Tat, matrix protein and HIV-1-derived factors, were performed using recombinant or purified proteins, whereas the Nef study and the reports on the programmed death ligand-1 were performed using infectious viruses and nef-deleted HIV-1 mutants. Although these results shed light on the influence of individual viral proteins on monocytes in vitro, synergistic or antagonistic effects of HIV-1 proteins on cellular responses cannot be ruled out, nor can the roles played by other host factors in vivo be excluded.

Overall, HIV-1 appears to promote the survival of monocytes as a key step for viral persistence. The interactions between the virus and monocytes may contribute key functions in establishing chronic HIV-1 infection and facilitating the progression to AIDS. These outcomes are likely influenced by the altered immunological function of monocytes and their interactions with other types of HIV-1 target cells (Figure 1).

DCs are professional APCs that are differentiated from monocytes in specific cytokine environments. DCs bridge the innate and adaptive immune responses, as they endocytose and break down invading pathogens in the endolysosome or proteasome and present antigen fragments to T cells, usually in the context of major histocompatability complexes 5. There are three major DC subtypes: myeloid DCs, plasmacytoid DCs (pDC), and Langerhans cells. These DC subtypes are characterised based on their locations, surface markers and cytokine secretion profiles 5.

DC life-span and survival are highly dependent on their anatomical locations and the DC subtypes 35. In general, DC half-lives measure up to a few weeks, and they can be replaced through proliferating hematopoietic progenitors, monocytes, or tissue resident cells 35. It has been shown that productive HIV-1 replication occurs in human monocyte-derived DCs for up to 45 days 36. DCs may survive longer within the lymph nodes due to cytokine stimulation in the microenvironment, which may help spread HIV-1 infection and maintain viral reservoirs.

HIV-1 is capable of directly infecting different DC subtypes (known as cis infection), but at a lower efficiency than HIV-1's ability to infect activated CD4⁺ T cells; therefore, only a small percentage of circulating DCs are positive for HIV in infected individuals 19. Productive HIV-1 replication is dependent on fusion-mediated viral entry in monocyte-derived DCs 37, and mature HIV-1 particles are localised to a specialised tetraspanin-enriched sub-compartment within the DC cytoplasm 38.

Langerhans cells are present in the epidermis or mucosal epithelia as immune sentinels 39. It is interesting that Langerhans cells have been shown to be resistant to HIV-1 infection 40. This resistance appears to be due to the expression of Langerin, which causes internalisation and break-down of HIV-1 particles and blocks viral transmission 40. However, in the context of co-infection with other sexually transmitted organisms, such as the bacterium Neisseria gonorrhoeae and/or the fungus Candida albacans 41 or when stressed by skin abrasion 42, Langerhans cells can become more susceptible to HIV-1 infection and are able to transmit HIV-1 to CD4⁺ T cells effectively 42.

Drug abuse can significantly facilitate HIV infection, transmission and AIDS progression through drug-mediated immunomodulation. Recent studies have suggested that the recreational drug, methamphetamine, increases susceptibility of monocyte-derived DCs to HIV-1 infection in vitro 43 and blocks the antigen presentation function of DCs 44. Although its relevance to the in vivo situation is unclear, this finding is potentially a further risk factor (aside from the use of contaminated needles, etc.) associated with drug use and may explain the high levels of HIV-prevalence among drug abusers.

HIV-1 infection of DCs likely contributes to viral pathogenesis. Notably, HIV-2 is much less efficient than HIV-1 at infecting both myeloid DCs and pDCs, whilst retaining its infectivity of CD4⁺ T cells 45. This observation offers an explanation for the decreased pathogenicity of HIV-2, since HIV-2 will need to infect CD4⁺ T cells directly and, perhaps more importantly, resting or memory CD4⁺ T cells to ensure long-term survival of the virus.

Given the important roles DCs play in the immune response, it is reasonable that HIV-1 proteins or the virus itself have been shown to affect the function of DCs in vitro. Both HIV-1 matrix and Nef proteins have been shown to cause only partial maturation of pDCs in vitro 4647. In the presence of these viral proteins, DCs acquire a migratory phenotype, facilitating travel to the lymph nodes. However, these DCs do not express increased levels of activation markers, such as the T cell co-stimulatory molecules CD80 and CD86, or MHC class II, that would lead to a protective immune response 4647. It is possible, therefore, that the DCs are trapped in the lymph nodes and unable to initiate a protective immune response against the virus. The study of Nef protein's effects on DCs 47 was performed using a mouse DC model in vitro and an immortalised cell line; hence the full relevance of this finding to the in vivo situation is unclear.

Conversely, recombinant Nef protein appears to cause DC activation and differentiation by up-regulating the expression of CD80, CD86, MHC class II and other markers, as well as various cytokines and chemokines associated with T cell activation 48. These effects have led to the proposition that Nef protein is capable of causing bystander activation of T cells via DCs 48, although this activity has not been demonstrated experimentally. Of note, the above study was performed using recombinant Nef alone.

DCs could contribute largely to an anti-HIV innate immunity. It has been demonstrated that pDCs are capable of inhibiting HIV-1 replication in T cells when cultured together in vitro 4950, implicating the importance of pDCs for viral clearance. HIV-1 infected individuals are known to have lower levels of circulating pDCs compared with those of uninfected individuals 51. It has been confirmed that HIV-1 is capable of directly killing pDCs 49, illustrating that the virus can remove a potential block to its replication and dissemination in pDCs.

HIV-1 can block CD4⁺ T cell proliferation or induce the differentiation of naive CD4⁺ T cells into T regulatory cells through pDCs 5253. These mechanisms involve HIV-1-induced expression of indoleamine 2,3-deoxygenase in pDCs. Indoleamine 2,3-deoxygenase is a CD4⁺ T cell suppressor and regulatory T cell activator 5253. HIV-1 envelope protein gp120 has also been shown to inhibit activation of T cells by monocyte-derived DCs 54, suggesting that gp120 may also have a role in the suppression of T cell function and progression to AIDS.

In addition, HIV-1 has been shown to suppress the immune function of pDCs in general by suppressing activation of the anti-viral toll-like receptor 7 (TLR7) and TLR8 55, and by blocking the release of the anti-viral interferon alpha 56. A recent study indicated that divergent TLR7 and TLR9 signalling and type I interferon production in pDCs contribute to the pathogenicity of simian immunodeficiency virus (SIV) infection in different species of macaques 57. These results suggest that chronic stimulation of pDCs by SIV or HIV in non-natural hosts may induce immune activation and dysfunction in AIDS progression 57. Overall, HIV-1 inhibits the function of pDCs to allow maintenance of the virus within the host.

The most interesting aspect of HIV-1 infection in DCs is the ability of the cells to act as mediators of trans infection of activated CD4⁺ T cells, which is the most productive cell type for viral replication. DC-mediated HIV-1trans infection of CD4⁺ T cells is functionally distinct from cis infection 5859 and involves the trafficking of whole virus particles from the DCs to the T cells via a 'virological synapse' 5960. Previous reviews have summarised the understanding of HIV-DC interactions 1961; so here we focus on discussing the latest progress in this field.

DC-mediated HIV-1 trans infection of CD4⁺ T cells is dependent on, or enhanced by, a number of other cellular and viral factors. CD4 co-expression with DC-SIGN (DC-specific intercellular adhesion molecule 3-grabbing nonintegrin), a C-type lectin expressed on DCs, inhibits DC-mediated trans infection by causing retention of viral particles within the cytoplasm 62. HIV-1 Nef appears to enhance DC-mediated HIV-1 trans-infection. Nef-enhanced HIV-1 transmission efficiency correlates with significant CD4 down-regulation in HIV-1-infected DCs 62. Furthermore, the maturation state of the DCs appears to be important for trans infection, with mature DCs showing greater HIV trafficking ability than immature DCs 59636465. These results have highlighted the proposed model that immature DCs might endocytose the virus in the periphery and then transfer it to CD4⁺ T cells upon DC maturation in the lymph node 19.

Recent studies have revealed that the precise trafficking of the endocytosed HIV virion, with regard to the sub-cellular vesicle trafficking networks 64 and cytoskeletal rearrangements associated with synapse formation 63, is critical for trans infection in mature DCs. The host cell-derived glycosphingolipid composition of the viral particle also appears to be important for both the capture of virus in mature and immature DCs and the trans infection process 66. Our recent results suggest that intracellular adhesion molecule-1 (ICAM-1), but not ICAM-2 or ICAM-3, is important for DC-mediated HIV-1 transmission to CD4⁺ T cells 67. The interaction between ICAM-1 on DCs and leukocyte function-associated molecule 1 (LFA-1) on T cells plays an important role in DC-mediated HIV-1 transmission 68. This mechanism might be specific for DC-mediated transmission of HIV-1 to CD4⁺ T cell, as in vitro experiments blocking LFA-1 on HIV-infected CD4⁺ T cells have shown no effect on virus transmission to non-infected T cells 69. In addition, purified host surfactant protein A in the mucosa has been shown to enhance DC-mediated HIV-1 transfer by binding to the viral envelope glycoprotein, gp120 70. This study also showed that surfactant protein A inhibited the direct infection of CD4⁺ T cells 70, suggesting a selection pressure for DC-mediated trans infection at mucosal surfaces.

The precise mechanism of virus transfer from DCs to CD4⁺ T cells has yet to be determined 19. Recent studies have demonstrated a role for small lipid vesicles known as exosomes in immature and mature DC-mediated HIV-1 transmission to CD4⁺ T cells 667172. Immature DCs are capable of constitutively releasing infectious virus in association with exosomes in the absence of CD4⁺ T cells 71. HIV-1 and purified exosomes can be endocytosed by mature DCs into the same intracellular compartment and transferred to co-cultured CD4⁺ T cells 72, suggesting that HIV-1 may exploit an intrinsic exosome trafficking pathway in mature DCs to facilitate viral dissemination. Although interesting for infectious dynamics, these observations on exosome-mediated viral transmission do not sufficiently explain the mechanisms of HIV-1 trans infection. How these models relate to the in vivo situation of DC-mediated HIV-1 transmission is unclear, given that DCs can traffic to the lymph node and effectively transfer virus to CD4⁺ T cells 19. If DCs release HIV-1 in association with exosomes in the tissue as DCs migrate to lymph nodes 71, or if DCs require T cell activation for the release of exosome-associated HIV-1 72, the viral transmission process might be very inefficient in vivo.

Recent studies have also offered the intriguing possibility that HIV-1 can be transferred from cell to cell via cell protrusions, with the virus either transmitting via cellular membrane nanotubes 73 or 'surfing' along the extracellular surface of the cytoplasmic membrane 74. HIV-1 intracellular trafficking is dependent on the viral envelope protein on the membrane of an infected cell to form a stable complex with the protrusion from an uninfected cell 73. This mechanism of viral transmission may be an adaptation of a normal cellular cross-talk process that is used in normal cellular communication, for example, by DCs and T cells during immunological synapse formation. Limitations to the above studies are that they were performed in either CD4⁺ T cells alone 73, immortal CD4⁺ T cells 74, or mainly using a mouse retroviral model 74. Indeed, the potential mechanisms of cell-cell-mediated HIV transmission have yet to be investigated in the DC-T cell trans infection model.

Inhibition of cell-cell mediated HIV-1 transmission can be developed into future therapeutic approaches. Because of the importance of DC-mediated trans infection of CD4⁺ T cells, a number of recent studies have identified factors that block this process, such as the C-type lectin, Mermaid, and natural anti-DC-SIGN antibodies in breast milk 75767778. However, the therapeutic efficacy of these factors has yet to be established.

HIV-2 is incapable of being transferred from DCs 45; and, coupled with its overall lack of cis infection of DCs, these data may explain why HIV-2 is less pathogenic than HIV-1.

Monocytes are implicated as a viral reservoir based on the detection of, or the recovery of, infectious virus from monocytes isolated from HIV-positive individuals on antiretroviral therapy 8788899091. It appears that CD16-positive monocytes (5% of monocyte population 7) are both more susceptible to infection and preferentially harbour the virus long-term 9293, perhaps explaining why only small numbers of monocytes are infected by HIV-1 in vitro. CD14⁺⁺ monocytes express high levels of the low molecular weight form of APOBEC3G (apolipoprotein B mRNA-editing enzyme, catalytic polypeptide-like 3G), which is associated with anti-HIV activity, whereas CD16⁺ monocytes express the high molecular weight form of APOBEC3G that has no anti-HIV activity 92.

The mechanism of HIV-1 latency in monocytes is not fully understood. Recent data suggest that the inhibition of viral replication is host mediated, at least in part, through a lack of the expression of key co-factors for the HIV-1 Tat protein. It appears that the transcription of the integrated viral genome, as transactivated by the viral Tat protein, is inhibited 94. Tat binds to the 5' long terminal repeat sequence of the integrated genome in complex with two host proteins, cyclin T1 (CycT1) and cyclin-dependent kinase 9 (CDK9), collectively known as the positive transcription elongation factor b (P-TEFb) 219596. Monocytes, when compared with activated CD4⁺ T cells and macrophages 96, are known to have much lower levels of CycT1 expression 9497, therefore, they lack functional P-TEFb. However, this is not the only factor responsible for the resistance of monocytes to HIV-1 replication, as transient expression of CycT1 is not sufficient to restore HIV-1 Tat-mediated transactivation in monocytes 94. Cell-cell fusion of monocytes and a HIV-1-permissive cell line restores Tat-mediated transactivation 94. Phosphorylation of CDK9 is known to be vital for the formation of a P-TEFb complex and for Tat-mediated transcription of the HIV-1 promoter 98. Despite having the same levels of CDK9, monocytes have low levels of the active, phosphorylated CDK9 form as compared with macrophages, and this phenotype has been directly correlated with the poor ability of monocytes to support HIV-1 replication 94. In addition, the basal transcription from the HIV-1 LTR in undifferentiated primary monocytes was reported to be undetectable using a transient transfection assay 94.

Studying HIV-1 latency in monocytes is challenging due to generally low viral integration and infection of monocytes 94. However, even when a HIV-1 proviral DNA construct is transfected directly into monocytes, there is no infectious virus production 94. When monocytes differentiate into macrophages, they become increasingly susceptible to HIV-1 infection and permissive to viral gene expression and production of infectious viruses 94. Furthermore, the differentiation of monocytes into macrophages stimulates HIV-1 production in the infected monocytes 94, suggesting a role played by monocytes in both viral latency and reactivation.

Because of the ability of DCs to transfer virus to CD4⁺ T cells, it is conceivable that DCs may act as reservoirs for HIV-1 and 'dose' T cells with the virus over extended periods. DCs are capable of transmitting HIV-1 to T cells over a period of several days, and the viral transmission is dependent on viral replication 99100101. It is possible, therefore, that long-term transfer of HIV-1 to T cells is actually through cis infection, while trans infection is only present in the very early stages 58. This HIV-1 transmission process may be 'trans-like', for example HIV-1 may assemble in endosomes or other intracellular membrane domains in a similar manner as described in macrophages 102103, then the virus may be transmitted across a virological synapse. However, the precise mechanism of virus assembly within macrophages remains a source of debate 104105.

The ability of DCs to act as reservoirs of HIV-1 appears to be highly dependent on the DC sub-type. Follicular DCs (FDCs) have been shown to retain infectious viral particles on their surface, and the retained virus is capable of being transferred to CD4⁺ T cells 106107108109110. FDCs in HIV-1 positive individuals harbour genetically diverse viral strains that are not observed elsewhere in the body 111, indicating that these cells may act as focal points for the rapid emergence of mutations observed in HIV-1 infected individuals.

It also appears that peripheral blood myeloid DCs do not harbour the virus in vivo during antiretroviral therapy 112, suggesting that it is the DCs in the lymph nodes that act as the long-term reservoir. This thinking is further supported by other studies that found HIV-1 in association with myeloid DCs that were isolated from lymph node biopsies or necropsies of individuals on antiretroviral therapy 107. Conversely, a recent study has suggested that Langerhans cells isolated from the oral cavity of HIV-1 positive individuals do not act as reservoirs for HIV-1, despite HIV-1 detection within whole tissue samples from the area 113. This result is perhaps not surprising given the effect of Langerin on inhibiting HIV-1 transmission 40. Moreover, pDCs have also not been implicated as reservoirs of HIV-1, which may be due to inhibiting HIV-1 replication through the secretion of IFNα and an unidentified small molecule by pDCs 4950.

HIV-1 is capable of altering the biology of haematopoietic stem cells in vivo, primarily affecting T cell development 114115116. Undifferentiated monocytic precursor cells, such as CD34⁺ stem cells or partially differentiated haematopoietic precursor cells, may act as reservoirs 117118. These cells in the bone marrow will be relatively shielded from antiviral treatments and may act as the ultimate long-term reservoir of HIV-1 (Figure 1). This mechanism allows for transmission of the virus because the progenitor cells containing integrated HIV-1 genomes will proliferate, differentiate and pass on the virus to progeny monocytes. Indeed, the ability to harbour genes and transfer them to progeny cells makes stem cells attractive targets for gene therapy against HIV-1 infection 119120.

There have been a number of studies that have proposed other mechanisms for latency in CD4⁺ memory T cells. It is possible that these mechanisms also have roles in latency in monocytes and/or DCs, but this remains to be investigated.

It has been proposed that the host cell itself can play a role through inhibition of HIV-1 gene transcription. In a CD4⁺ T cell line and a yeast model of HIV-1 transcription, host chromatin structures slowly accumulate (in one study over 30 days 121) on the long terminal repeat of the integrated viral genome and inhibit viral gene transcription 121122. Moreover, recent studies have suggested a much broader role for host transcription factors in HIV-1 latency in CD4⁺ T cells 123124125.

In light of the evidence that suggests miRNAs play a role in the resistance of monocytes to HIV-1 infection 1415, it is of interest that a number of host miRNAs have been implicated in causing latency in resting primary CD4⁺ T cells 126. Inhibitors of these miRNAs are now being touted as a new generation of treatment to be used in concert with current antiretrovirals [reviewed in 127].

In resting CD4⁺ T cells from HIV-1-infected individuals, HIV-1 multiply spliced RNA transcripts are retained in the nucleus and cannot be translated into functional proteins 128. The lack of a host transcription factor, polypyrimidine tract binding protein, appears to account for the underlying mechanism in resting CD4⁺ T cells. Transient expression of this host protein induces productive HIV-1 replication in resting CD4⁺ T cells that are isolated from HIV-1-positive individuals 128.

However, HIV-1 latency is not always restricted to resting CD4⁺ T cells or explained by limiting cellular factors. In some instances, HIV-1 latency is due to replicative selection for specific viral characteristics. It has been shown that a doxycycline-dependent HIV-1 variant is capable of establishing latency within a dividing CD4⁺ T cell type (SupT1 cell line) normally permissive for viral replication 129. This study showed that only a small proportion (0.1%–10%) of an inducible provirus was rescued from the cells after addition of the inducing doxycycline drug, indicating that HIV-1 is capable of establishing latency in the majority of actively dividing cells. Thus, in some settings, HIV-1 proviral latency is not limited to resting T cells, but can be due to intrinsic viral traits 129.