The life cycle of a retrovirus is that of an obligatory intracellular parasite, and thus HIV cannot replicate outside of human cells.
HIV binding and entry
Infection of the host cell commences when HIV binds to specific receptors on the cell membrane. In general, the interaction requires the recognition of two host-cell surface-receptor proteins by the viral gp120 envelope protein. The presence or absence of these cellular proteins restricts the range of host-cell types that are susceptible to infection by a strain of HIV.
The first HIV receptor described was the CD4 protein  which is present predominantly on cells of the T lymphocyte and myeloid lineages. The distribution of CD4 receptors has been thought to restrict HIV susceptibility to cells of the lymphocyte, monocyte/macrophage and other CD4-expressing lineages, although there is precedent for HIV-1, HIV-2 and the related virus SIV to be able to fuse with cells not expressing CD4 [34, 85, 86].
Subsequently, the requirement of a second coreceptor for viral entry was recognized [87-92]. This function may be performed by a range of proteins within the class of seven-transmembrane receptors, although the most important are CCR5 (CC chemokine receptor 5) and CXCR4 (CXC chemokine receptor 4). The seven-transmembrane class of receptors is very large, containing over 100 related proteins. Several of these proteins have been shown to facilitate the binding of HIV in vitro (Reviewed in  ). The in vivo significance of entry via these minor co-receptors, however, remains unclear.
After HIV gp120 binds to CD4 receptor and the co-receptor, a conformational change in gp41 causes insertion of the N-terminal hydrophobic fusion-peptide region into the target-cell membrane . This insertion results in membrane fusion and the entry of the viral particle contents into the cytoplasm, a process critically dependent upon interactions between the N-and C-terminal regions of the gp41 ectodomain. This intra-protein interaction led to the discovery of a novel class of antivirals called fusion inhibitors, e.g. T-20 (enfuvirtide) which is a short peptide that mimics the structure of the conserved C-terminal region of gp41 . Unlike other viruses, HIV does not require to be endocytosed and in CD4+ T cells, primarily fuses at the plasma membrane .
Although all HIV strains will recognize and bind to CD4, affinity for either CCR5 or CXC4 varies. These differences account for the observed distinct tropisms between HIV strains. Binding ability and tropism of the virus is dependent on the protein structure of gp120. Particular patterns of sequence of the V3 and V4 variable regions, and other regions of gp120, relate to CD4 binding and differential coreceptor affinity [93, 97, 98].
In general, viral strains that bind to CCR5 (R5 strains) infect macrophages, activated CD4+ T cells and subsets of resting CD4+ T cells where CCR5 is expressed. The majority of strains present in vivo are R5 strains and variants of R5 strains known as founder viral strains (see below) are selected during the process of sexual HIV transmission. Other HIV strains that recognise CXCR4 (X4 strains) can exist, but by contrast, infect only T cells and T cell lines and are only present in approximately 40 to 50% of patients during the latter stages of HIV disease . The reason for the lack of CXCR4 variants in early stages of HIV disease, is presently hypothesized to be related to their susceptibility to the adaptive immune response in addition to other factors that limit their growth . The adaptive immune hypothesis does fit well with the time that these isolates emerge, as it is typically at the latter stages of HIV disease when the immune system is being steadily degraded. In theory, CXCR4 isolates can emerge via two mechanisms. Firstly, by CXCR4 viruses being transmitted alongside CCR5 using strains or secondly by progressive genetic evolution of a CCR5 using strain to a CXCR4 using strain. To date the majority of scientific observations support the latter hypothesis . Indeed, investigators would argue that all HIV positive patients would acquire a X4 utilising HIV strain, given enough time (ie. patients often succumb to the effects of R5 HIV prior to the emergence of X4 strains). The emergence of X4 strains is often associated with subsequent accelerated loss of CD4+ T cells and as a consequence accelerated disease progression. This latter phenomenon is a consequence of increased targeting of cells by X4 HIV strains. For instance, whilst CCR5 expression is restricted to a small memory/activated CD4+ T cell subset, CXCR4 is present on upwards of 90% of all CD4 T cells [101-103]. Historically, the growth of X4 strains in vitro has been characterised by the presence of syncytial cells, which are formed by the fusion of multiple infected cells and can be observed by light microscopy. Syncitia formation, however, is largely based on the relative expression of CXCR4 on T cell lines used in these assays. The expression of CCR5 expression to similar levels as CXCR4, can readily result in the formation of syncytia when using R5 strains that replicate at similar levels. Indeed, the mechanistic basis of syncytia formation is the level of HIV envelope expressed on the infected donor cells and the HIV receptor expression (CD4 and CCR5 or CXCR4) on the neighbouring recipient cell. This scenario is largely based on the replication of the virus within the donor cell and indicative of overall viral production. Indeed investigators have demonstrated the spontaneous formation of syncytia in cell lines that are exposed to high titer virus .
Importantly, knowledge of the role of CD4 and associated chemokine receptors in HIV binding to cells led to a new area in HIV therapeutics. A number of small molecules that block binding of HIV to either CCR5 or CXCR4 have been described. Early clinical trials of the CCR5 inhibitor Maraviroc, given in conjunction with an optimised background of other antiretroviral drugs, demonstrated greater virologic and immunologic efficacy profiles compared to placebo in participants having only detectable R5 HIV [105, 106]. X4 specific chemokine receptor inhibitors also exist; the first generation AMD3100 and the second generation AMD11070. Curiously, AMD3100 was discovered in anti-HIV drug screens prior to the knowledge that chemokine receptors work in concert with CD4 to allow HIV entry [107, 108]. Concept trials in humans using the bioavailable AMD11070 CXCR4 inhibitor observed positive outcomes, but further development of this drug class has been delayed due to observations of liver damage in animal models .
The pre-integration complex
The genetic information of HIV is contained within an RNA genome. Following infection of a new host cell, the RNA genome is first reverse transcribed into single-stranded DNA that is then further transcribed to double-stranded DNA. These two polymerase steps are performed by viral reverse transcriptase, which is co-packaged in the viral particle. Self-priming of the single-stranded RNA and DNA and removal of the transcribed RNA strand occur by a complex series of steps dependent upon interactions between the viral LTR and host-cell enzymes. The double-stranded DNA genome forms a complex with host-cell and viral proteins (including matrix, integrase and Vpr) that is actively transported to the nucleus [110, 111].
During the early steps of the HIV-1 replication cycle, the virus counteracts specific host proteins that have evolved to limit retroviral replication. Two key host cell factors that influence early viral events are i) a protein called human tripartite motif 5 alpha (TRIM5alpha), and ii) sterile alpha motif (SAM) domain and Histidine-Aspartic (HD) domain-containing protein 1 (SAMHD1). TRIM5alpha acts in the early steps of the HIV-1 replication cycle, soon after the entry process and before reverse transcription . TRIM5alpha restricts retroviral infection by specifically recognising HIV-1 capsid and promoting its rapid, premature disassembly [113, 114]. TRIM5alpha from rhesus macaques and African green monkeys inhibit HIV-1 replication, whereas the human homologue is inactive against SIV and HIV-1, leading to the susceptibility of human cells to both viruses. The primary mode of action SAMDH1was initially hypothesised to be by cleaving nucleotides needed for efficient HIV reverse transcription and subsequent integration. This is particularly apparent in resting cells like macrophages, dendritic cells and resting CD4+ T cells, where there is already a limited nucleotide pool. However recent studies have observed SAMHD1 to also act on HIV RNA and restrict HIV infection through RNase activity  and it is presently unclear whether it is the RNase activity combined with dNTPase activity that combines to restrict HIV infection.
Integration and transcription
The double-stranded HIV genome is then either integrated into the host-cell genome by means of DNA splicing, performed by the viral Integrase or forms DNA circles . In the context of chromosomal integration, this process is not random, as recent studies have observed preferential HIV DNA integration in active transcriptional sites within infected cells (reviewed by ). The integrated form of HIV is known as the provirus, with identical LTR copies flanking the coding regions. Proviral DNA is replicated as part of the normal cell genome and may persist in this form for long periods and through many rounds of mitotic cell division. Furthermore, integration into long-lived resting CD4+ T cells provides challenges towards attempts at HIV cure as the virus in these cell types in typically dormant/latent.
Transcription of the proviral HIV genome
Transcription is mediated by the promoter in the 5' LTR. Transcription generates a 9.2-kb primary HIV RNA genome that can be directly packaged in the virion through HIV Gag binding to the 5’ LTR and psi packaging elements. In addition this transcript can also be subject to splicing and as a consequence, can form up to 30 mRNA transcripts.
Regulation of HIV RNA transcripts
The primary HIV-1 RNA genome contains several splice donors and splice acceptors to generate the pool of 30 HIV spliced RNA transcripts that are designated as incompletely spliced or fully spliced RNA. Incompletely spliced RNA uses the splice donor site located nearest the 5' end of the HIV RNA genome in addition to one of many splice acceptor sites located in the central genomic region of the virus. Incompletely spliced RNA is approximately 4 to 5 kilobases in length, can express Env, Vif, Vpu, Vpr, and Tat and require the expression of HIV Rev for RNA export from the nucleus (ie. like the full unspliced genome of HIV, are HIV Rev dependent).
Fully spliced HIV RNA has spliced out both introns of HIV and can express Rev, Nef, and Tat. These heterogeneous mRNAs do not require the expression of the Rev protein. As they do not require the expression of HIV Rev for RNA nuclear export, they are often referred to as early viral genes.
HIV assembly and release
For full HIV assembly, transcription and subsequent translation must reach a critical threshold for viral assembly to occur . One common misconception with respect to HIV latency is that lack of HIV transcription and translation exists in all latently infected cells. In fact, latency in resting CD4+ T cells may exist where low levels of detectable viral RNA and proteins are present but at a level that does not lead to assembly and production of fully assembled and infectious virions. Once a threshold of viral protein production is reached, proteins that are destined for virions are recruited and assembled at the plasma membrane . Not all virally produced proteins are incorporated with assembling virions, with the dominant virion associated proteins expressed as either a fusion protein with the structural protein HIV Gag (ie. Gag-Pol) or are able to non-covalently associate with HIV Gag (eg. Vpr).
For post translational events, Gag exists as a monomer and oligomerisation of Gag triggers the exposure of a myristoylation motif that binds to the membrane enriched lipid called phosphatidylinositol 4,5-bisphosphate. Whilst gag oligomerisation can proceed when expressed in isolation, it can also be triggered upon viral RNA binding largely through the nucleocapsid domain of Gag , thus ensuring viral inocula have a greater propensity to bud from the cell membrane with its genetic cargo.
Oligomerisation of Gag at the plasma membrane leads to the formation of what is known as the immature viral shell, primarily consisting of repeating Gag polyprotein hexamers . A partially intact spherical shell is hypothesised to leave the membrane, with the aid of scission from the membrane through the Gag-specific recruitment of cellular proteins involved in the ESCRT complex (Endosomal Sorting Complexes Required for Transport machinery), that effectively act as molecular scissors to liberate virions from the membrane.
Shortly after leaving the membrane, the HIV protease is activated and culminates in the sequential cleavage of Gag and Gag-pol proteins. This process, commonly referred to as viral maturation, is marked by condensation of a large pool of HIV capsid, that leads to encapsulation of many viral proteins including HIV RNA-bound nucleocapsid within the conical shaped viral core.
Whilst the above outlines the specific role of the major structural HIV protein Gag, other viral proteins perform a variety of roles to subvert normal cellular function and facilitate viral replication (summarised in detail above in Table 1). Much about these processes remains poorly understood. Vpr acts to alter host-cell transcription and arrest infected cells at the G2/M phase of cell division . Nef induces downregulation of the CD4 receptor and MHC class I molecules [66, 68, 122]. Vpu promotes degradation of CD4 in the endoplasmic reticulum and counteracts an interferon-induced restriction factor known as Tetherin, which acts by tethering and preventing the release of newly formed HIV particles from the plasma membrane . Vif is necessary for subsequent efficient infectivity of the newly produced viral particles [82, 123] and counteracts cytidine deaminases (enzymes present especially in macrophages and T cells) that are naturally occurring host defence mechanisms against retroviruses. These proteins include APOBEC3G and APOBEC3F and are degraded by HIV .