Rev-dependent lentiviral expression vector
© Wu et al. 2007
Received: 20 December 2006
Accepted: 07 February 2007
Published: 07 February 2007
Skip to main content
© Wu et al. 2007
Received: 20 December 2006
Accepted: 07 February 2007
Published: 07 February 2007
HIV-responsive expression vectors are all based on the HIV promoter, the long terminal repeat (LTR). While responsive to an early HIV protein, Tat, the LTR is also responsive to cellular activation states and to the local chromatin activity where the integration has occurred. This can result in high HIV-independent activity, and has restricted the use of LTR-based reporter vectors to cloned cells, where aberrantly high expressing (HIV-negative) cells can be eliminated. Enhancements in specificity would increase opportunities for expression vector use in detection of HIV as well as in experimental gene expression in HIV-infected cells.
We have constructed an expression vector that possesses, in addition to the Tat-responsive LTR, numerous HIV DNA sequences that include the Rev-response element and HIV splicing sites that are efficiently used in human cells. It also contains a reading frame that is removed by cellular splicing activity in the absence of HIV Rev. The vector was incorporated into a lentiviral reporter virus, permitting detection of replicating HIV in living cell populations. The activity of the vector was measured by expression of green fluorescence protein (GFP) reporter and by PCR of reporter transcript following HIV infection. The vector displayed full HIV dependency.
As with the earlier developed Tat-dependent expression vectors, the Rev system described here is an exploitation of an evolved HIV process. The inclusion of Rev-dependency renders the LTR-based expression vector highly dependent on the presence of replicating HIV. The application of this vector as reported here, an HIV-dependent reporter virus, offers a novel alternative approach to existing methods, in situ PCR or HIV antigen staining, to identify HIV-positive cells. The vector permits examination of living cells, can express any gene for basic or clinical experimentation, and as a pseudo-typed lentivirus has access to most cell types and tissues.
All HIV-dependent expression vectors in common use are based on the HIV long terminal repeat promoter (LTR). An early HIV gene product, Tat, increases the level of transcript that is initiated at the LTR. The placement of reporter genes downstream of the LTR results in a responsiveness to the synthesis of Tat, a measure of HIV replication. The earliest indicator lines made use of reporter enzymes, such as luciferase and β-galactosidase [1–4], permitting a direct measurement of reporter gene induction. Tat appears to increase HIV transcriptional activity by two mechanisms. The first identified Tat activity is not directed towards the proviral DNA promoter, but rather through direct association with the growing nascent RNA chain. Tat associates with a 5' RNA loop structure [5–7], the transactivation response element (TAR), to promote completion of the initiated transcript [8–10], an activity also defined as processivity or anti-termination. More recent work has provided evidence that Tat also stimulates assembly of transcription factors to the DNA promoter ; that is, Tat promotes initiation, as well as elongation . However, the LTR as a promoter is inherently leaky. Following integration of the HIV DNA into the host chromatin, the LTR transcribes the early gene products Tat, Rev, and Nef. That is, there is a required basal level of transcription that is Tat-independent. In addition, the site of integration or insertion of the LTR-based expression vector can mediate high levels of transcriptional activity in some cells , leading to expression from the reporter LTR in the absence of HIV. For example, in the generation of HIV indicator cells with LTR-based reporters, it has been necessary to remove 25% or more of the stably transfected cells [4, 14] since they generate reporter transcript in the absence of HIV. While high expression cells can be removed in the generation of reporter clones, this inherent leakiness prevents the use of viral vectors to deliver the LTR-based reporter construct to detect the presence of existing HIV-positive cells in a mixed population.
By eliminating this non-specific activity, that is, non-HIV induction of signal, from an HIV expression vector, a wider use of this convenient and efficient tool would be possible. In addition to Tat, HIV transcriptional activity is also affected by an early gene product, Rev. Rev binds to a 3' loop structure, the Rev response element (RRE), present in unspliced and singly-spliced HIV transcripts, to permit nuclear export and translation of these mRNAs [15–17]. This viral specific activity exploits an essential cellular process. The removal of non-coding regions of transcripts (introns) prior to translation is critical to all cells. Introns are operationally defined by the presence of strong splicing sites, and HIV exploits this cellular activity by including multiple splice sites with varying activity [18, 19] to generate multiple coding regions within the same stretch of HIV RNA. The existence of these sites results in the generation of fully spliced transcripts in the early phase of HIV infection [20, 21]. Once Rev is expressed, RRE-containing HIV transcripts can be delivered to the cytosolic translational machinery.
It is of interest that in HIV infection protein expression from singly or non-spliced HIV transcripts appears absolutely dependent on Rev expression [15, 22, 23]. This dependency is lost with seemingly minor modifications to HIV DNAs [24, 25], and the earliest reported RRE-containing expression constructs [26, 27] can display varied Rev-independent expression. These were designed to elucidate Rev function, and the focus was not to eliminate background signal. In this report, by inclusion of multiple components of the HIV genome, we have constructed a lentiviral expression vector that displays a full dependency on the presence of HIV. As it is silent in HIV-negative cells, it differs dramatically from LTR-based systems.
The Rev-dependent expression vector displayed highly specific HIV-dependent expression. The application tested here, as a lentiviral reporter for HIV detection, permitted us to address two major concerns: non-specific (HIV-independent) noise and inadequate specific (HIV-dependent) signal. The first concern is based on knowledge that expression from an integrated LTR vector can be affected by the local chromatin activity, resulting in HIV- or Tat-independent high expression . Of course, this is what prevents utilization of LTR-based reporter viruses. With the establishment of LTR-based indicator cells, it's just a matter of removing cells with aberrantly high background reporter expression [4, 14]. This problem appeared to be absent in cells transduced with the Rev-dependent vector system, since expression of GFP in both a T cell line and a primary macrophage population occurred only in HIV-positive cells. Furthermore, we found that GFP-encoding transcript was detected by RT-PCR only in HIV-infected populations. Secondly, we wished to know whether this vector, potentially as a single integrant and with the added stringency of Rev-dependence, would generate an adequately robust response. To address this point we examined reporter function under conditions that maximized single reporter integrants. It does assume that infections follow a Poisson function, a central tenet to most virus infection studies. At an input where 80% of the cells do not become infected (approximately one infectious particle per five cells), less than 2% of the cells should possess more than one reporter provirus. Expression of GFP reporter from the HIV-positive cells closely approximated the input reporter viral vector level (Fig. 2 and 3), suggesting that undetected expression of integrated vectors is not prominent. Note, however, that expression of the reporter gene is achieved in cells that are also highly supportive of transcriptional activity from the HIV provirus, that is, LTR activity is supported. Both HIV and the reporter vector use the LTR promoter.
We also examined high doses of reporter viral vector, where a majority of the cells are infected with the Rev-dependent reporter vector. Under these conditions we did not see intact reporter (unspliced) transcript or protein in HIV-negative cells. The nature of the curve that fits the reporter vector detection of HIV-infected cells (Fig. 2D) suggests that the rate-limiting step towards reporter expression follows Michaelis-Menten kinetics. This is characteristic of a single binding-site event, but at this time we are unable to define this rate-limiting process, such as the viral particle binding to the cellular receptor. Additional mechanisms leading to this non-linear function would include the potentially equivalent susceptibility of increasing numbers of infected cells to additional infection, as characterized by a Poisson distribution. Concentration of this viral particle by centrifugation permitted this examination of high vector input, but it also diminished the infectivity of the viral particles. This data also predicts that most of the HIV-infected cells will be susceptible to detection by the reporter virus. Of course, in mixed cell populations or tissues there will be variable susceptibilities to infection by the reporter virus as presently used. The inability to infect all cells by the reporter virus is perhaps the most limiting aspect of this application of the Rev-dependent vector.
In this report we have utilized the term Rev-dependence to distinguish this construct from the existing LTR-based Tat-responsive HIV reporter systems; however, with the intact LTR, we would speculate that this Rev-dependent vector should maintain responsiveness to Tat expression as well. This dual dependency is likely to contribute to the higher stringency seen. In this document, we have characterized this Rev-dependent vector in only one application (viral reporter vector), an application that is challenging and not feasible with the LTR-based reporters. At the same time we view this application as supplementary to existing methods to test cells for the presence of HIV replication. These include the staining or probing of HIV proteins and nucleic acids in fixed cells, and may include amplification with PCR [34–36]. We have also incorporated the vector into continuous cell lines (data not shown). As introduced into a transformed T cell, we find that expression from this Rev-dependent vector is insensitive to cellular activation states, thus maintaining a negative background in the absence of HIV, but has enhanced sensitivity to HIV infection, relative to many of the existing non-T cell indicator cell lines. Although this report has been limited to the expression of reporter genes, the construct's incorporation into a lentivirus and the specificity acquired should also permit its use as an experimental agent to express any gene in HIV-infected cells.
The inclusion of Rev-dependent processing to an LTR-based vector has resulted in a highly specific HIV-dependent expression system.
Viruses and cells were obtained through the NIH AIDS Research and Reference Reagent Program: pNL4-3.HSA.R+E- from Dr. Nathaniel Landau ; CEM-SS from Dr. Peter L. Nara ; J1.1 cells from Dr. Thomas Folks . HIV-1AD8 is a macrophage-tropic molecular clone . Macrophages were derived from peripheral blood monocytes of healthy donors from the Department of Transfusion Medicine at the National Institutes of Health. Cells were cultured in RPMI media supplemented with 10% fetal bovine serum (Hyclone) and recombinant human macrophage colony stimulating factor (rH-MCSF) (R & D Systems) at 10 ng/ml. Media was changed every 48 hours for ten days.
pCMVΔR8.2 and pMD.G have been described previously . pNL-GFP-RRE was constructed by complete deletion of all HIV ORFs of pNL4-3  by replacing the 8.1 kb BssHII-BlpI fragment of the HIV-1 genomes with an insert containing the GFP ORF and the HIV-1 Rev-responsive element (RRE) including the HIV-1 sequence immediately following the BssHII site and the first 336 nucleotides of the gag ORF (the gag reading frame was disrupted by a frame shift mutation at the ClaI site by blunt end ligation), the GFP ORF was derived from pIRES-hrGFP-1a (Stratagene) by PCR amplification (5' CTCGAAATTAACCCTCACTAAAGG 3'; 5'ATCGTGTACGGCCGAATTGGGTACACTTACCTG 3'), and the fragment containing RRE (corresponding to position 7612 to 8469 of the HIV-1NL4-3 genome). pNL-GFP-RRE-(SA) was constructed by insertion of a PCR fragment into the NotI-SmaI site of pNL-GFP-RRE, in front of the GFP ORF. The insert carrying the HIV-1 A5 splicing acceptor and D4 donor was amplified by primers: 5' ATAAGAATGCGGCCGCATCTCCTATGGCAGGAAG 3'; 5' AATCACCCGGGTGCTACTACTAATGCTACTATTGC 3'. The sequence of pNL-GFP-RRE-(SA) has been deposited in GenBank (accession number EF408805).
Stocks of the HIV-1NL4-3 and HIV-1AD8 were prepared by transfection of HeLa cells with cloned proviral DNA. HIV-1 based lentiviral vectors carrying reporter genes were made by cotransfection of DNA constructs as follows: 2 × 106 293T cells were cotransfected with 10 ug of pNL-GFP-RRE-SA, 7.5 ug of pCMVΔR8.2, and 2.5 ug of pMD.G using the calcium phosphate method (Promega). Viral particles were harvested 2 days after cotransfection and filtered through a 0.45 um filter and stored at -80°. One preparation was concentrated by ultracentrifugation . The titer (TCID50) of the lentiviral indicator vNL-GFP-RRE(SA) virus preparation was estimated by serial dilution into activated (TNF-treated) HIV-positive J1.1 Jurkat cells  following the method of Reed and Muench .
Total cellular poly(A+) mRNA was purified from cells by MicroPoly(A)Pure mRNA isolation kit (Ambion) as recommended by the manufacturer. Reverse transcription was accomplished using the RETROscript First-Strand Synthesis Kit (Ambion) with random decamers as the first-strand primers. Following cDNA synthesis, PCR was carried out using primer 5'TAATCGGCCGAACAGGGACTTGAAAGCGAAAG3' and 5'CAGGCACAAGCAGCATTGTTAG 3' to amplify spliced lentiviral transcripts, and primer 5' TAATCGGCCGAACAGGGACTTGAAAGCGAAAG3' and 5'ATCGTGTACGGCCGAATTGGGTACACTTACCTG3' to amplify non-spliced GFP transcripts. The PCR condition was: 1 × PCR buffer, 125 uM dNTPs, 1.5 mM Mg2+, 50 pmol of each primer, 1 U SuperTaq Plus (Ambion) in 50 ul, with 30 cycles of 20 sec at 94°, and 180 sec at 68°. One fifth of the product was analyzed on 2% agarose gel. Cellular β-actin transcripts were amplified using QuantumRNA β-actin Internal Standards (Ambion) with similar conditions as above except using 20 pmol of each actin primer and runs at 20 sec at 94°, 30 sec at 55°, and 40 sec at 68°.
CEM-SS cells, infected with VSV pseudo-typed HIV-1 NL4-3.HSA.R+E- , were enriched by biotin conjugated rat-anti-mouse CD24 antibody and streptavidin conjugated magnetic beads and further cultured for two weeks to remove the beads. The enriched population was then infected with the Rev-dependent lentiviral vector, vNL-GFP-RRE(SA). At 72 hrs post transduction, cells were stained with R-phycoerythrin conjugated rat-anti-mouse CD24 antibody and analyzed by flow cytometer for CD24 and GPF expression. Curve fitting and statistical analysis were achieved with GraphPad Prism software.
Monocyte-derived macrophages in six well plates were infected with HIVAD8 (36 to 360 ng p24) in 500 ul of RPMI for 3 hours, then washed and returned to 2 ml RPMI. The following day, 5 ng p24 of pNL-GFP-RRE(SA) was applied to the macrophages. At 5 days post infection, cells were intracellularly stained on ice. Briefly, cells were rinsed once with Hank's buffer followed by treatment with Cytofix/Cytoperm (Becton Dickinson) for 20 min, then 2 ml of cold methanol for 15 min. After washing with Permeabilization/Wash buffer (Becton Dickinson), cells were stained with anti-p24 HIV antibody 183.H12-5C (1:250) [43–45] for 90 min, followed by washing and staining with Alexa Fluor-568 goat anti-mouse IgG (Invitrogen) (1:250) for 30 minutes.
The GFP-expressing and anti-p24 stained cells were photographed using a Leica DMIRB/E inverted research microscope attached to a Orca DCAM ER camera (Hamamtsu), using an XF102.2 filter (Omega Optical Inc.) for the Alexa Fluor-568 (p24) detection and a GFP filter (Leica) for the GFP expressing cells. P24 expression and GFP expression were captured with similar exposures, typically 1000 milliseconds. Files were saved as Tiffs and color was added on ImageJ software version 1.33u (Rasband, W., NIH-public domain). Dimensions were determined through a stage micrometer (Electron Microscopy Sciences).
Infected or uninfected cells were re-suspended into 100 ul Hank's staining buffer (Hank's buffer plus 0.1% bovine serum albumin, 0.1% sodium azide, and 25 mM HEPES, pH 7.2) for staining with antibodies at concentrations recommended by manufactures. Following staining for 30 min on ice, cells were either washed with Hank's buffer or, in the case of HIV-infected cells, fixed with 500 ul of Cytofix/Cytoperm (Becton Dickenson) for 20 min on ice for flow cytometry analysis on a FACScan analyzer (Becton Dickenson). The murine CD24 antibody was from Southern Biotechnology.
We thank Dongyang Yu for technical expertise with the PCR run, and Drs. Michael Emerman, Tomris Mustafa, and Sundararajan Venkatesan for their comments and criticisms concerning this manuscript. This research was supported by the Intramural Research Program of the NIMH/NIH and by the NIH grant 5R21NS51130-2 awarded to YW.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.