Detection and initial characterization of protein entities consisting of the HIV glycoprotein cytoplasmic C-terminal domain alone.

Employing antibodies against the cytoplasmic tail of the HIV-1 glycoprotein (Env-CT), in addition to gp160/gp41, we have identified several novel small Env proteins (<25kD) in HIV-1 transfected and infected cells. Mass spectrometric and mutational analyses show that two mechanisms contribute to their generation. Thus the protein, designated Tr-Env-CT (for truncated Env-CT), consists of the C-terminal 139 amino acids (aa) of Env (aa 718-856) with the N-terminal Q718 modified to pyroglutamic acid. It is likely derived from full-length Env protein by proteolytic processing. A further heterogeneous set of slightly larger proteins, termed Env-CT* species, are rather derived from spliced mRNAs containing only those Env C-terminal residues (aa 719-856) which overlap with the second tat and rev coding exons. They are N-terminally extended in the same reading frame. It is conceivable that essential Env-CT functions may be fulfilled by these novel species rather than by the full-length glycoprotein itself.

Mutational analysis of the internal membrane proximal domain of the HIV glycoprotein C-terminus.

This study focuses on the long stretch of highly conserved amino acids in the membrane proximal part of the HIV-1 cytoplasmic tail (Env amino acids (aa) 706-718) upstream of the overlap with the tat and rev second coding exons. Changes in Env aa 713 and 715, although they did not affect Env function, abrogated replicative spread. Other amino acid substitutions, i.e., 706-712, 714 and 716, despite their conservation, did not result in defective replicative phenotypes even in primary peripheral blood lymphocytes. Our results point to their involvement in presently unrecognized essential Env functions pertinent only in in vivo. Interestingly, changes in the codons for residues 717-718 as well as some mutations in residues 714-716 abrogated Gag expression but still allowed expression of functional Env in a rev-independent manner. This could be due to the inactivation of a rev-regulated negative element within the respective nucleotide sequence (8354-8368).

An intronic G run within HIV-1 intron 2 is critical for splicing regulation of vif mRNA.

Within target T lymphocytes, human immunodeficiency virus type I (HIV-1) encounters the retroviral restriction factor APOBEC3G (apolipoprotein B mRNA-editing enzyme, catalytic polypeptide-like 3G; A3G), which is counteracted by the HIV-1 accessory protein Vif. Vif is encoded by intron-containing viral RNAs that are generated by splicing at 3' splice site (3'ss) A1 but lack splicing at 5'ss D2, which results in the retention of a large downstream intron. Hence, the extents of activation of 3'ss A1 and repression of D2, respectively, determine the levels of vif mRNA and thus the ability to evade A3G-mediated antiviral effects. The use of 3'ss A1 can be enhanced or repressed by splicing regulatory elements that control the recognition of downstream 5'ss D2. Here we show that an intronic G run (G(I2)-1) represses the use of a second 5'ss, termed D2b, that is embedded within intron 2 and, as determined by RNA deep-sequencing analysis, is normally inefficiently used. Mutations of G(I2)-1 and activation of D2b led to the generation of transcripts coding for Gp41 and Rev protein isoforms but primarily led to considerable upregulation of vif mRNA expression. We further demonstrate, however, that higher levels of Vif protein are actually detrimental to viral replication in A3G-expressing T cell lines but not in A3G-deficient cells. These observations suggest that an appropriate ratio of Vif-to-A3G protein levels is required for optimal virus replication and that part of Vif level regulation is effected by the novel G run identified here.

Conformation-specific display of 4E10 and 2F5 epitopes on self-assembling protein nanoparticles as a potential HIV vaccine.

The self-assembling protein nanoparticle (SAPN) is an antigen-presenting system that has been shown to be suitable for use as a vaccine platform. The SAPN scaffold is based on the principles of icosahedral symmetry, beginning from a monomeric chain that self-assembles into an ordered oligomeric state. The monomeric chain contains two covalently linked α-helical coiled-coil domains, an N-terminal de novo-designed pentameric tryptophan zipper and a C-terminal de novo-designed trimeric leucine zipper, which assemble along the internal symmetry axes of an icosahedron. In this study, we incorporated the membrane proximal external region (MPER) of HIV-1 gp41 from HXB2 into the N-terminal pentamer, referred to as MPER-SAPN, attempting to reproduce the α-helical state of the 4E10 epitope while maintaining a structurally less-constrained 2F5 epitope. Sprague-Dawley rats were immunized with MPER-SAPNs, and their sera were analyzed for induced humoral anti-HIV-1 responses. We show that high membrane proximal external region-specific titers can be raised via the repetitive antigen display of MPER on the SAPN without the need for adjuvant. However, none of the sera displayed a detectable neutralizing activity against HIV-1. Thus, 4E10- and 2F5-like neutralizing antibodies could not be elicited by MPER conformationally restrained in the SAPN context.

Role of the C-terminal domain of the HIV-1 glycoprotein in cell-to-cell viral transmission between T lymphocytes.

BACKGROUND: Mutant HIV (HIV-Env-Tr712) lacking the cytoplasmic tail of the viral glycoprotein (Env-CT) exhibits a cell-type specific replication phenotype such that replicative spread occurs in some T-cell lines (referred to as permissive cells) but fails to do so in most T-cell lines or in PBMCs (referred to as non-permissive cells). We aim to gain insight on the underlying requirement for the Env-CT for viral spread in non-permissive cells. RESULTS: We established that in comparison to HIV-Wt, both cell-free and cell-to-cell transmission of mutant HIV-Env-Tr712 from non-permissive cells were severely impaired under naturally low infection conditions. This requirement for Env-CT could be largely overcome by using saturating amounts of virus for infection. We further observed that in permissive cells, which supported both routes of mutant virus transmission, viral gene expression levels, Gag processing and particle release were inherently higher than in non-permissive cells, a factor which may be significantly contributing to their permissivity phenotype. Additionally, and correlating with viral transfer efficiencies in these cell types, HIV-Gag accumulation at the virological synapse (VS) was reduced to background levels in the absence of the Env-CT in conjugates of non-permissive cells but not in permissive cells. CONCLUSIONS: During natural infection conditions, the HIV-Env-CT is critically required for viral transmission in cultures of non-permissive cells by both cell-free and cell-to-cell routes and is instrumental for Gag accumulation to the VS. The requirement of the Env-CT for these related processes is abrogated in permissive cells, which exhibit higher HIV gene expression levels.

Identification of the cellular prohibitin 1/prohibitin 2 heterodimer as an interaction partner of the C-terminal cytoplasmic domain of the HIV-1 glycoprotein.

Our studies aim to elucidate the functions carried out by the very long, and in its length highly conserved, C-terminal cytoplasmic domain (Env-CT) of the HIV-1 glycoprotein. Mass spectrometric analysis of cellular proteins bound to a tagged version of the HIV Env-CT led to the identification of the prohibitin 1 and 2 proteins (Phb1 and Phb2). These ubiquitously expressed proteins, which exist as stable heterodimers, have been shown to have multiple functions within cells and to localize to multiple cellular and extracellular compartments. The specificity of binding of the Phb1/Phb2 complex to the Env-CT was confirmed in various manners, including coimmunoprecipitation with authentic provirally encoded, full-length Env. Strong binding was dependent on Env residues 790 to 800 and could be severely inhibited by the double mutation L799R/L800Q but not by mutation of these amino acids individually. Analysis of the respective mutant virions revealed that their different abilities to bind Phb1/Phb2 correlated with their replicative properties. Thus, mutated virions with single mutations [HIV-Env-(L799R) and HIV-Env-(L800Q)] replicated similarly to wild-type HIV, but HIV-Env-(L799R/L800Q) virions, which cannot bind Phb1/Phb2, exhibited a cell-dependent replicative phenotype similar to that of HIV-Env-Tr712, lacking the entire Env-CT domain. Thus, replicative spread was achieved, although somewhat delayed, in "permissive" MT-4 cells but failed to occur in "nonpermissive" H9 T cells. These results point to binding of the Phb1/Phb2 complex to the Env-CT as being of importance for replicative spread in nonpermissive cells, possibly by modulating critical Phb-dependent cellular process(es).

HIV pseudovirion vaccine exposing Env "fusion intermediates"-response to immunisation in human CD4/CCR5-transgenic rats.

Immune responses to a pseudovirion-based HIV vaccine enriched in Env conformations, which have been induced to an authentic intermediate fusion stage by interaction with the cellular HIV receptor complex, have been analysed in human CD4/CCR5-transgenic rats. High titre Env-binding antibodies were elicited. However, these immune sera failed to neutralise HIV-1, but rather led to an enhancement of infection in vitro. This enhancing activity appeared to be directed towards contaminating cellular proteins in the vaccine and was able to mask neutralisation of potent, mixed-in neutralising antibodies. The induced Env-specific antibodies, purified on the basis of binding to monomeric Env, retained high-binding activity, but failed to be neutralising. Thus, it remains unclear whether vaccines based on induced HIV Env fusion intermediates can elicit broadly neutralising responses.

Optimized protocol for the large scale production of HIV pseudovirions by transient transfection of HEK293T cells with linear fully deacylated polyethylenimine.

HIV vaccine strategies which employ pseudovirions (PVs) as the source of antigen require large amounts of particles. These are typically generated by transient transfection of mammalian cells and purification of the released PVs from the culture supernatant. Since efficiency and cost of transfection are key issues, in this report the transfection efficiencies, achieved by employing a panel of high-molecular-weight linear polyethylenimines (PEIs) and small cross-linked PEIs, were analyzed and compared to those obtained by transfections with calcium phosphate or the commercial reagent Polyfect. High efficiencies were obtained using several of the modified PEIs, and the transfections with these inexpensive reagents were very robust. The observed efficiencies (as quantitated by amounts of expressed gene product) were two to four fold superior to calcium phosphate transfection and approximately equal to that achieved using Polyfect which is, however, prohibitively expensive for large scale applications. An optimized and rapid protocol for the large scale production and purification of HIV-PVs from 293T cells growing in so-called cell stacks and transfected with the best reagent identified, PEI87, is described here. The generated PVs, obtained with a yield in the range of 0.4mg virion-associated HIV-CA/liter culture supernatant, exhibited only very minimal contamination with non-viral proteins and were thus suitable for vaccination applications.

Cell-free infectivity of HIV type 1 produced in nonpermissive cells is only moderately impacted by C-terminal Env truncation despite abrogation of viral spread.

Mutant HIV virions, encoding C-terminally truncated Env proteins, exhibit a cell-specific replication defect, i.e., they can replicate in a few T cell lines (termed permissive cells) but not in the majority of T cell lines (termed nonpermissive cells). We have studied the properties of two mutant virions (pNL-Tr712 and pNL-Tr752), encoding Envs with C-terminal truncations of 144 and 104 amino acids, respectively. We show that although unable to give rise to a spreading infection in nonpermissive H9 cells, both cell-free pNL-Tr712 and pNL-Tr752 virions, produced in these cells, still exhibit relatively high levels of infectivity (30-80% of wildtype) when tested in nonpermissive target cells. Compatible with this high remaining infectivity, we observed that the levels of Env incorporation into mutant virions, produced in nonpermissive cells, were not drastically reduced as has been reported by others. The high remaining infectivity of cell-free mutant virions in nonpermissive cells is difficult to reconcile with the complete lack of spreading infection in these cells. We demonstrate that nonpermissive cells are less susceptible to single-round infection with cell-free virus than permissive cells. It is thus conceivable that in these cells other transmission routes, e.g., cell-cell transmission, may be more important for total virus spread and that this route may be more severely impacted by the C-terminal Env truncations.

Incorporation of chimeric HIV-SIV-Env and modified HIV-Env proteins into HIV pseudovirions.

Low level incorporation of the viral glycoprotein (Env) into human immunodeficiency virus (HIV) particles is a major drawback for vaccine strategies against HIV/AIDS in which HIV particles are used as immunogen. Within this study, we have examined two strategies aimed at achieving higher levels of Env incorporation into non-infectious pseudovirions (PVs). First, we have generated chimeric HIV/SIV Env proteins containing the truncated C-terminal tail region of simian immunodeficiency virus (SIV)mac239-Env767(stop), which mediates strongly increased incorporation of SIV-Env into SIV particles. In a second strategy, we have employed a truncated HIV-Env protein (Env-Tr752(N750K)) which we have previously demonstrated to be incorporated into HIV virions, generated in infected T-cells, to a higher level than that of Wt-HIV-Env. Although the chimeric HIV/SIV Env proteins were expressed at the cell surface and induced increased levels of cell-cell fusion in comparison to Wt-HIV-Env, they did not exhibit increased incorporation into either HIV-PVs or SIV-PVs. Only Env-Tr752(N750K) exhibited significantly higher (threefold) levels of incorporation into HIV-PVs, an improvement, which, although not dramatic, is worthwhile for the large-scale preparation of non-infectious PVs for vaccine studies aimed at inducing Env humoral responses.

Effects of signal peptide exchange on HIV-1 glycoprotein expression and viral infectivity in mammalian cells.

In certain cell systems, exchange of the human immunodeficiency virus (HIV) Env signal peptide (SP) sequence with that of heterologous SPs has been shown to increase gp120 transport and secretion. Here we demonstrate that exchange of the HIV-Env-SP with those from erythropoietin or tissue plasminogen activator in the proviral context does not increase wild-type membrane-bound Env expression or incorporation into released virions. In fact, virion infectivity was decreased. These infectivity decreases were largely due to effects on Env transport and/or function and only to a minor extent to cis effects as a result of the sequence exchanges themselves. Thus, in fact, it is not advantageous to employ heterologous SPs to achieve high-level expression of functional cell surface membrane- or virion-associated HIV-Env.

Generation of H9 T-cells stably expressing a membrane-bound form of the cytoplasmic tail of the Env-glycoprotein: lack of transcomplementation of defective HIV-1 virions encoding C-terminally truncated Env.

H9-T-cells do not support the replication of mutant HIV-1 encoding Env protein lacking its long cytoplasmic C-terminal domain (Env-CT). Here we describe the generation of a H9-T-cell population constitutively expressing the HIV-1 Env-CT protein domain anchored in the cellular membrane by it homologous membrane-spanning domain (TMD). We confirmed that the Env-TMD-CT protein was associated with cellular membranes, that its expression did not have any obvious cytotoxic effects on the cells and that it did not affect wild-type HIV-1 replication. However, as measured in both a single-round assay as well as in spreading infections, replication competence of mutant pNL-Tr712, lacking the Env-CT, was not restored in this H9 T-cell population. This means that the Env-CT per se cannot transcomplement the replication block of HIV-1 virions encoding C-terminally truncated Env proteins and suggests that the Env-CT likely exerts its function only in the context of the complete Env protein.

Selection and characterization of a replication-competent human immunodeficiency virus type 1 variant encoding C-terminally truncated env.

A long cytoplasmic C-terminus (Env-CT) on the human immunodeficiency virus type 1 (HIV-1) Env protein is a highly conserved feature in vivo. Mutant HIV lacking the Env-CT cannot replicate in PBMCs and in the majority of T cell lines (nonpermissive cells, e.g., H9 cells) in vitro. We report here that a single amino acid change (N750K) in the context of the mutant virus pNL-Tr752 lacking 104 C-terminal Env amino acids gives rise to a virus variant pNL-Tr752(N750K), which can now replicate in nonpermissive H9 cells and, albeit to a lower extent, in PBMCs. We have analyzed the properties of replication-competent pNL-Tr752(N750K) in comparison to its defective counterpart pNL-Tr752 and to wild-type virus in H9 cells. In all cases, the respective glycoproteins were functional in inducing membrane fusion and were incorporated into particles. In comparison to pNL-Tr752 and pNL-Wt, pNL-Tr752(N750K) glycoprotein exhibited increased fusion induction and 2- to 3-fold increased incorporation into particles, properties that may contribute to the observed replication competence.

Inter-retroviral fusion mediated by human immunodeficiency virus or murine leukemia virus glycoproteins: independence of cellular membranes and membrane vesicles.

We have recently demonstrated for the first time that inter-retroviral membrane fusion, i.e., membrane fusion between individual retroviral particle populations with incorporated HIV-1 Env and cellular receptors, respectively, can occur (Sparacio et al. 2000, Virology 271: 248-252). We have extended these analyses here and confirmed that fusion between particles can occur in the extracellular medium independent of any cellular membranes and that luciferase transduction, mediated by the fused structures, is independent of significant potential contribution by contaminating membrane vesicles. We have additionally analyzed whether membrane fusion between HIV-like particles can be mediated by amphotropic murine leukemia virus (MuLV) glycoprotein and its respective cellular receptor, PiT-2. We demonstrate that PiT-2 can be incorporated into HIV-like particles and can fuse with MuLV-Env-carrying particles. This occurs only in the situation in which the incorporated MuLV-Env protein has been activated to fusion activity by HIV protease-mediated removal of the C-terminal R-peptide and is completely inhibited when the respective particles are generated in the presence of the HIV protease inhibitor, Saquinavir.

Inhibition of release of lentivirus particles with incorporated human influenza virus haemagglutinin by binding to sialic acid-containing cellular receptors.

Mutants of the haemagglutinin (HA) gene of human influenza virus A/Aichi/2/68 (H3N2) encoding HA proteins that are proteolytically cleaved intracellularly, defective in binding to cellular receptors or defective for acylation within the cytoplasmic C terminus have been generated. Here, the properties of these mutated HA molecules are described and their incorporation into the lipid membrane of released human immunodeficiency virus (HIV)-like particles is analysed. It is demonstrated that, when produced from cells coexpressing any of the binding-competent Aichi-HA molecules, release of HIV-like particles into the extracellular medium is reduced and the particles that are released fail to incorporate Aichi-HA. These blocks in release and incorporation, respectively, can both be overcome. The release of normal amounts of particles with incorporated HA can be achieved either by mutation of the receptor-binding site on the Aichi-HA molecule or by removal of sialic acid from surface proteins with neuraminidase. In contrast, as a result of blockage of the sialic acid-binding site by sialidated oligosaccharides on the HA itself, the HA of influenza virus A/FPV/Rostock/34 (H7N1) is efficiently incorporated into HIV-like particles. These results, namely that particle release can be inhibited by interactions between the incorporated glycoprotein and the cell surface and/or that interactions with other cellular components can be inhibitory to incorporation into retrovirus envelopes, probably reflect general principles that may hold for many viral and cellular glycoproteins.