A new cell-based assay for measuring the forward mutation rate of HIV-1.

Over 20 years into the ever-worsening AIDS pandemic, genetic variation remains the greatest obstacle for treating and preventing HIV-1 infection. Mutation rate assays for HIV-1 have been reported; however, none measure directly the forward mutation rate during replication of the virus in cell culture while still retaining the ability to propagate and further study mutant proviruses. Therefore, the objective of the current study was to develop such a phenotypic cell-based assay for measuring the forward mutation rate of HIV-1. Conventional recombinant DNA techniques and polymerase chain reaction were used to create a replication defective HIV-1 vector, pNL4-3Delta+cass, which is based on the NL4-3 strain and contains the thymidine kinase gene from human herpes virus type 1 as the mutational target. A series of transfection and infection steps were used to introduce the vector into 143B cells, which are negative for thymidine kinase function, and produce vector virus for a single cycle of replication. Viral titers were measured by counting the number of drug resistant colonies on the assay plates, and forward mutation rates were calculated from the viral titers. Mutant proviruses were sequenced to determine the types of genetic alterations that occurred. The average forward mutation rate for HIV-1 was 2.2 x 10(-5)mutations/base/cycle. The majority of mutations were base substitutions, including high frequencies of C-->U and G-->A transitions. Single adenosine insertions were also observed frequently. The new assay is economical and provides a direct measurement of the mutation rate during a single cycle of viral replication. Target cells containing mutant proviruses survive the drug selection process and may be propagated for further analysis. The new assay is novel and has many advantages over previous mutation rate assays and thus will be very useful in future studies on genetic variation of HIV-1.

Restricted entry of R5 HIV Type 1 strains into eosinophilic cells.

A cell culture system previously developed by our laboratory demonstrated that T cell-tropic (CXCR4-using) but not macrophage-tropic (CCR5-using) HIV-1 strains productively infected eosinophilic cells. In the current study, an improved model was used to determine the level of this viral restriction by assessing viral entry and coreceptor usage. The model was improved by using AML14.3D10 cells that were engineered to express CCR3 in addition to the major HIV-1 coreceptors, CD4, CXCR4, and CCR5, thus making them more like primary eosinophils. A polymerase chain reaction (PCR) assay was used to detect viral entry. In the PCR assay, primers specific for early reverse transcription products were used to amplify minus strand viral DNA from HIV-1-infected AML14.3D10-CCR3 eosinophilic cells. Coreceptor blocking experiments, using the CXCR4 antagonist AMD3100, were performed to determine coreceptor usage by the CXCR4-using (X4) strain known to productively infect the cells. Virus production was measured by p24 immunoassay. As expected, viral DNA was detected in AML14.3D10-CCR3 cells infected with X4 HIV-1, and cell viability was decreased during maximal viral production. Conversely, viral DNA was not detected in eosinophilic cells exposed to a CCR5-using (R5) HIV-1 strain that is also capable of using CCR3, indicating that R5 HIV-1 is unable to enter eosinophilic cells despite the presence of the appropriate coreceptors. Infection of AML14.3D10-CCR3 cells by HTLV-III(B) was completely inhibited by AMD3100, indicating that X4 HIV-1 enters the AML14.3D10-CCR3 cell line by using the CXCR4 coreceptor exclusively. Since X4 strains predominate during the late stages of HIV-1 infection in many patients, when eosinophil numbers also tend to increase, the ability of these HIV-1 strains to infect eosinophilic cells has important implications for the involvement of eosinophils in the pathogenesis of AIDS.

The envelope glycoprotein of human endogenous retrovirus HERV-W induces cellular resistance to spleen necrosis virus.

Human endogenous retrovirus type W (HERV-W) envelope glycoprotein (Env) has recently been reported to induce fusion in cells expressing the RD-114 and type D retrovirus receptor (RDR) and to serve as a functional retroviral envelope protein. In this report, another biological function for HERV-W was demonstrated by testing its ability to protect cells against retroviral infection. Spleen necrosis virus (SNV), a gammaretrovirus was chosen for testing resistance because it uses RDR to enter cells. An HERV-W Env expression plasmid was transfected into canine osteosarcoma cells (D-17), which are permissive for SNV infection. Cell fusion assays were performed to demonstrate biological function of HERV-W Env in D-17 cells. The presence of HERV-W env sequences was confirmed in stably transfected cell clones by using polymerase chain reaction. Viral infectivity assays were performed with SNV and amphotropic Murine leukemia virus (MLV-A) pseudotyped vector viruses to measure titers in D-17 cells expressing HERV-W Env and in negative control cells. The HERV-W Env caused fusion of D-17 cells in culture and greatly reduced infection by SNV vector virus. A 1000- to 10,000-fold decrease in SNV infectivity was observed for D-17 cells expressing HERV-W Env as compared to D-17 cells that were not expressing HERV-W Env. In contrast, infection by MLV-A pseudotyped vector virus was not significantly reduced. Thus, HERV-W Env confers host cell resistance to infection by SNV. This is the first report of a human endogenous retrovirus gene product blocking infection by any exogenous retrovirus.

High fidelity of homologous retroviral recombination in cell culture.

Genetic variation continues to be a major obstacle in the development of therapies and vaccines against retroviral infections and contributes extensively to viral pathogenesis and persistence. Recombination is one mechanism that increases retroviral variation by shuffling mutations from different genomes. Recent studies suggest that recombination not only shuffles the mutations but also generates them at high rates during reverse transcription. In contrast to these recent studies, this investigation shows that recombination does not generate mutations during recombination. A spleen necrosis virus (SNV)-based homologous recombination system was used to test the hypothesis that retroviral recombination is a high-fidelity process during replication of the virus in cell culture. The system consisted of a pair of SNV vectors expressing two drug resistance genes. The vectors were constructed so that cells containing recombinant proviruses could be selected by a double drug-resistant phenotype. Restriction enzyme digestion and agarose gel electrophoresis were used to map the location of recombination within 182 proviruses. Sequencing and single-strand conformation polymorphism techniques were then used to check for mutations within the recombinant proviruses. Since no mutations were detected among the 182 recombinants that were analyzed, homologous recombination is a high-fidelity process for retroviruses in cell culture.

Strain-dependent productive infection of a unique eosinophilic cell line by human immunodeficiency virus type 1.

Eosinophils are granulocytic leukocytes that function in both protective and pathological immune responses. They can be infected by HIV-1, but characterization of the infection has been hindered by lack of a productive cell culture model. In the present study, the unique eosinophilic cell line AML14.3D10 was used as a model to test the hypothesis that HIV-1 productively infects eosinophilic cells in a strain-dependent fashion. The AML14.3D10 cell line was cultured with one T cell-tropic (T-tropic) strain and two macrophage-tropic (M-tropic) strains of HIV-1 (HTLV-IIIB, HIV-1AdaM, and HIV-1Ba-L strains, respectively). Cytopathic effects were evident in living cultures and in stained slide preparations of AML14.3D10 cells infected with the T-tropic strain of HIV-1. Culture supernatants from infected AML14.3D10 cells contained high levels of HIV-1 p24 protein that peaked at approximately 7-10 days postinfection. A line of AML14.3D10 cells chronically infected with HTLV-IIIB and continuously producing high levels of virus was established. In contrast to the T-tropic strain, the M-tropic strains of HIV-1 did not productively infect the eosinophilic cell line. Thus, the AML14.3D10 eosinophilic cell line was permissive for a T-tropic strain but not for M-tropic strains of HIV-1. Flow cytometry revealed that uninfected AML14.3D10 cells were positive for the HIV-1 receptor CD4 and coreceptors CXCR4 and CCR5; the cell line was negative for CCR3. The lack of productive infection by M-tropic strains despite CCR5 expression indicates that strain-dependent infection may not be determined at the coreceptor level in AML14.3D10 cells.

Retroviral recombination is nonrandom and sequence dependent.

Sequence variation plays a significant role in the pathogenesis and persistence of retroviral infections and is a major obstacle in the development of vaccines as well as therapies against lethal diseases caused by retroviruses. Recombination is one means by which sequence variation is generated. However, the basic molecular mechanisms of recombination are not adequately understood. In the present study, a spleen necrosis virus (SNV) recombination system was used to ask whether a known hot spot for mutation was also a hot spot for retroviral recombination. The system consisted of a pair of SNV vectors expressing two drug-resistance genes, constructed so that recombinants could be selected by a double resistant phenotype. Restriction enzyme site differences engineered into the vectors were used to map the location of recombination sites within relatively small intervals (55 to 420 bp). The vectors were modified to create two pairs that differed only by the presence of runs of identical nucleotides. The runs of identical nucleotides had been shown previously to be hot spots for frameshift mutations during SNV reverse transcription. Each vector pair was introduced into DSDh helper cells by infection. Viruses were harvested from doubly infected DSDh helper cells and used to infect D-17 target cells. Proviral sequences from 228 cell clones were analyzed by polymerase chain reaction and restriction enzyme digestion. Significant differences in the patterns of recombination were found between the two pairs of vectors. In particular, the frequency of recombination was higher than expected in the interval immediately following the runs. For both pairs of vectors, the overall pattern of recombination was nonrandom and one region was refractory toward recombination.

Direct demonstration of retroviral recombination in a rhesus monkey.

Recombination may be an important mechanism for increasing variation in retroviral populations. Retroviral recombination has been demonstrated in tissue culture systems by artificially creating doubly infected cells. Evidence for retroviral recombination in vivo is indirect and is based principally on the identification of apparently mosaic human immunodeficiency virus type 1 genomes from phylogenetic analyses of viral sequences. We infected a rhesus monkey with two different molecularly cloned strains of simian immunodeficiency virus. One strain of virus had a deletion in vpx and vpr, and the other strain had a deletion in nef. Each strain on its own induced low virus loads and was nonpathogenic in rhesus monkeys. When injected simultaneously into separate legs of the same monkey, persistent high virus loads and declines in CD4+ lymphocyte concentrations were observed. Analysis of proviral DNA isolated directly from peripheral blood mononuclear cells showed that full-length, nondeleted SIVmac239 predominated by 2 weeks after infection. These results provide direct experimental evidence for genetic recombination between two different retroviral strains in an infected host. The results illustrate the ease and rapidity with which recombination can occur in an infected animal and the selection that can occur for variants generated by genetic recombination.

Significance of premature stop codons in env of simian immunodeficiency virus.

The location of the translational termination codon for the transmembrane protein (TMP) varies in three infectious molecular clones of simian immunodeficiency virus from macaques (SIVmac). The SIVmac251 and SIVmac142 infectious clones have premature stop signals that differ in location by one codon; transfection of these DNAs into human HUT-78 cells yielded virus with a truncated TMP (28 to 30 kilodaltons [kDa]). The SIVmac239 infectious clone does not have a premature stop codon in its TMP-coding region. Transfection of HUT-78 cells with this clone initially yielded virus with a full-length TMP (41 kDa). At 20 to 30 days posttransfection, SIVmac239 virus with a 41-kDa TMP gradually disappeared coincident with the emergence of a virus with a 28-kDa TMP. Virus production dramatically increased in parallel with the emergence of a virus with a 28-kDa TMP. Sequence analysis of viral DNAs from these cultures showed that premature stop codons arising by point mutation were responsible for the change in size of the TMP with time. A similar selective pressure for truncated forms of TMP was observed when the SIVmac239 clone was transfected into human peripheral blood lymphocytes (PBL). In contrast, no such selective pressure was observed in macaque PBL. When the SIVmac239 clone was transfected into macaque PBL and the resultant virus was serially passaged in macaque PBL, the virus replicated very well and maintained a 41-kDa TMP for 80 days in culture. Macaque monkeys were infected with SIVmac239 having a 28-kDa TMP; virus subsequently recovered from T4-enriched lymphocytes of peripheral blood showed only the 41-kDa form of TMP. These results indicate that the natural form of TMP in SIVmac is the full-length 41-kDa TMP, just as in human immunodeficiency virus type 1. Viruses with truncated forms of TMP appear to result from mutation and selection during propagation in unnatural human cells.