A Lentiviral cDNA Library Employing Lambda Recombination Used To Clone an Inhibitor of Human Immunodeficiency Virus Type 1-Induced Cell Death

Abstract

Expression cloning technology of cDNAs is a suitable tool for identifying novel functional properties of genes. Here, we generated a lentiviral cDNA library-expressing system for human T cells based on a site-specific recombination system of phage lambda for transferring cDNA libraries with a minimum loss of its complexity. The library-transduced CD4⁺ T cells were challenged with wild-type human immunodeficiency virus type 1 (HIV-1), and the cells that acquired resistance to HIV-1-induced cytopathic effect (CPE) were selected. From these cells, CD14 was isolated and proved to inhibit the entry of HIV-1 and the HIV-1-induced CPE. This cloning system allows rapid identification of genes encoding novel properties in human T cells and probably other mammalian cells.

A number of screening systems from genetic libraries have been developed to identify novel functional properties of the genes. A successful screening with mammalian cells is dependent on the efficiency of the transduction system into the appropriate target cells. A plasmid-based expression system has been generally used (1, 2, 17). However, this system has a limit due to the inefficient transfection into particular cells, such as nonadherent cells. In addition, the introduced genes are expressed only transiently. Therefore, it is desirable to develop a new technology that can efficiently achieve long-lasting expression of genetic information in the nonadherent cells, especially human lymphocytes. Retrovirus vectors appear to overcome these limits (16). Retrovirus infects a wide range of mammalian cell types, including lymphocytes, with a high efficiency. The library-inserted retrovirus vector can integrate into the host's chromosome and is expressed permanently. These properties have been utilized for a gene delivery system for lymphocytes (16). However, the prototype murine leukemia virus-based retrovirus vector infects only dividing cells (12) and, less efficiently, human T cells. Thus, the target cells for screening are limited. Recently, a human immunodeficiency virus (HIV)-based lentivirus vector was developed (12), and such vectors are beginning to be used in many applications (4, 7, 10, 11).

In this report, we describe the development of a lentiviral cDNA library expression system applicable for human T cells. The results showed significant utility of the system to clone genes through a high-throughput screening procedure. This system allowed us to identify genes that render cells resistant to HIV-induced cell death. Our lentivirus system is promising, as it can be applied to many library screening systems, and should accelerate the discovery of novel properties of the genes in many other cells including neurons and hematopoietic stem cells.

MATERIALS AND METHODS

Cells.

Human 293T were maintained in Dulbecco's modified Eagle medium containing 10% fetal calf serum, and MT-4 cells were maintained in RPMI 1640 containing 10% fetal calf serum.

Gateway-compatible lentiviral cDNA library system and HIV-1 challenge.

A Gateway-compatible lentivirus vector DNA (pYK005C) was constructed through the insertion of a Gateway cloning system reading frame cassette (Invitrogen, Carlsbad, Calif.) into the EcoRI site of the multiple cloning sites (MCS) in the HIV-1-based vector DNA, pCSII-elongation factor 1α promoter (EF)-MCS-internal ribosome entry site (IRES)-humanized Renilla green fluorescent protein (hrGFP) (9). For the generation of the entry cDNA library, 10 ng (≈1.5 × 10⁹ copies) of the original cDNA library generated from human peripheral blood leukocytes (Invitrogen) was amplified by PCR with the following primers: 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCT-3′ and 5′-GGGGACCACTTTGTACAAGAAAGCTGGGT-3′ (underlined nucleotides are the attB [B1 and B2] sequences in the forward and reverse primers, respectively). The cycling conditions were 94°C for 2 min, 94°C for 15 s, 55°C for 30 s, and 68°C for 5 min for 15 cycles and 68°C for 10 min. PCR products and pDONR201 DNA (Invitrogen) were incubated with BP Clonase enzyme mix (Invitrogen) for 16 h at 25°C by using the procedure recommended by the manufacturer, and the resulting recombinant molecules were transformed in DH5α. The transformants were selected with kanamycin (50 μg/ml), and the resultant entry cDNA library was prepared from pools of transformants. For the generation of the vector cDNA library, 300 ng of the entry cDNA library and 360 ng of pYK005C vector DNA, which is linearized by digestion with EcoRI, were incubated with LR Clonase enzyme mix (Invitrogen) for 19 h at 25°C. All resulting recombinant molecules were transformed in DH5α and selected on plates containing ampicillin (50 μg/ml). The resultant vector cDNA library was prepared from pools of transformants. For preparation of cDNA-expressing lentivirus vector, a vesicular stomatitis virus (VSV)-pseudotyped lentivirus vector was generated via calcium phosphate-mediated transfection of 293T cells as described before (9). Briefly, 1.2 × 10⁷ cells were divided onto six TC dishes (100 × 20; Nunc, Roskilde, Denmark) 24 h before transfection. Seventeen micrograms of Vector cDNA library DNA, 12 μg of HIV Gag-Pol-expressing vector (pMDLg/pRRE), 5 μg of VSV-G protein-expressing vector (pMD-G), and 5 μg of HIV Rev-expressing vector (pRSV-Rev) per dish were cotransfected, then 48 h later, the culture supernatants were collected, and virus particles were concentrated 30-fold by centrifugation at 6,000 × g for 16 h. The concentrated viruses were titrated with MT-4 cells. For transduction of the cDNA library into T cells and HIV type 1 (HIV-1) challenge, 1.2 × 10⁷ MT-4 cells were infected with 8 × 10⁶ infectious doses of the viral cDNA library. Three days later, the cells were challenged with HIV-1_NL4-3 at a multiplicity of infection (MOI) of 0.05. For recovery of the cDNA sublibrary from surviving cells, MT-4 cells that survived HIV-1 challenge were collected and genomic DNA was extracted. The cDNAs from the surviving cells were amplified by PCR with primers that were used to amplify the original cDNA library as described above. This cDNA sublibrary was transferred to the pDONR201 vector by a BP reaction, and the resultant entry cDNA sublibrary was transferred to pYK005C lentivirus vector DNA by an LR reaction as described above. The viral cDNA sublibrary was prepared via transfection of 293T cells and used for the second round of screening.

Flow cytometric analysis.

Two-color flow cytometric analysis was performed. Briefly, cells were stained with the optimal concentration of antibody for 30 min at 4°C and then washed. Phycoerythrin-conjugated anti-human CD4 and CD14 (eBioscience, San Diego, Calif.) and anti-mouse H-2K^k (Cedarlane, Ontario, Canada) were used. HIV-1 expression was examined with an anti-HIV-1 human serum followed by staining with biotin-conjugated anti-human IgG (Vector Laboratories, Burlingame, Calif.) and streptavidin-conjugated peridinin chlorophyll protein (BD Biosciences, San Jose, Calif.). The data were collected by FACScan (BD PharMingen, San Diego, Calif.) and analyzed with WinMDI software.

Sequence analysis.

cDNA cloned into the pDONR201 vector was analyzed with the 5′-TCGCGTTAACGCTAGCATGGATCTC-3′ primer. The data were collected with the ABI 377 autosequencer. The sequence data were compared with the DNA database at the National Center for Biotechnology Information by using BLAST search.

Determination of individual cDNA length.

The original cDNA library, the entry cDNA library, and the vector cDNA library were applied to Escherichia coli competent cells, and the cells were spread onto Luria-Bertani plates to develop bacterial colonies. cDNA fragments were amplified by PCR from these bacterial colonies containing each cDNA fragment. The PCR products were subjected to agarose gel electrophoresis and visualized with ethidium bromide. The migration distance of each cDNA fragment was compared with a DNA size marker. MT-4 cells transduced with the viral cDNA library were cloned by the limiting dilution method. cDNA fragments were amplified by PCR from the cloned cells. The length of each cDNA was determined as described above.

CD14 cDNA transduction and HIV-1 infection.

A CD14 cDNA-expressing construct was made through the insertion of the Gateway cloning system reading frame cassette (Invitrogen) into the EcoRI site of the pIRES-hrGFP vector (Stratagene, San Diego, Calif.), and then a CD14 cDNA fragment was isolated from the library by an LR reaction. CD4⁺CCR5⁺ HeLa cells (6) were transfected by Lipofectamine 2000 (Invitrogen) with the CD14-expressing construct or empty vector (pIRES-hrGFP) as a control, and then 48 h later, the cells were infected with HIV-1_NL4-3 at an MOI of 2. Cells were harvested 2, 12, 24, and 48 h after HIV-1 infection, and DNA was extracted as described before (20). For CD14 stable transduction, an H-2K^k-expressing lentivirus vector, which was constructed by replacing the mutant Renilla reniformis hrGFP sequence in the Gateway-compatible lentivirus vector DNA (pYK005C) with the H-2K^k sequence, was used. MT-4 or CD4⁺ CCR5⁺ HeLa cells were infected with either the CD14-expressing or control lentivirus vector at an MOI of 1, and then 2 days later, the cells were challenged with HIV-1_NL4-3 at an MOI of 0.05. Cell killing activity was measured by trypan blue staining, and virus production in the culture supernatant was monitored by enzyme-linked immunosorbent assay (ZeptoMetrix Corp., Buffalo, N.Y.) for the HIV-1 p24^gag antigen.

Real-time PCR assay.

For the detection and quantification of individual forms of HIV-1 DNA, strong-stop (early reverse transcript), full-length/1-LTR circle (late reverse transcript), 2-LTR circle, and integrated forms, a real-time PCR assay was used as described previously (20). PCR was performed with an ABI PRISM 7700 sequence detection system (PE Applied Biosystems, Foster City, Calif.) and TaqMan universal PCR master mix (PE Applied Biosystems).

Statistical analysis.

The Mann-Whitney U test was used to determine statistical significance, and P values of <0.05 were considered significant.

RESULTS

Transfer of a cDNA library from a cloning expression vector into a donor vector.

Since some leukocytes would produce antiviral proteins, we started to isolate anti-HIV genes from a cDNA library generated from human peripheral blood leukocytes. Since a plasmid-based expression vector via transfection cannot be used for efficient and stable transduction into T cells, a lentiviral cDNA library-expressing system was used to introduce genes into human T cells. In this system, we used a recombination-cloning system referred to as Gateway (22). The Gateway system, which has been used to transfer individual genes (22), is based on the recombination system of the phage lambda that mediates integration and excision of the phage DNA into and from the E. coli genome, respectively. The integration involves recombination of the attP sites (P1 and P2) of the phage DNA within the attB sites (B1 and B2) located in the bacterial genome (BP reaction) and generates an integrated phage genome flanked by attL (L1 and L2) and attR (R1 and R2) sites. The next excision results in these attL and attR sites back to the attP and attB sites (LR reaction). First of all, the cDNA library fragments inserted between bacterial genome-derived B1 and B2 sites of the pCMV-SPORT6 cloning expression vector, referred to as the original cDNA library, were amplified by PCR. The PCR product was purified and incubated with a donor vector containing an insertion of the P1 and P2 sites, pDONR201, in the presence of BP clonase enzyme mix, which consists of a mixture of the phage protein integrase (Int) and the bacterial protein integration host factor. The BP clonase recombines the B1 and P1 sites as well as the B2 and P2 sites (BP reaction), and as a result, the cDNA library fragments were placed between derivatives of the L1 and L2 sites. The cDNA library-inserted pDONR201 was referred to as the entry cDNA library. Although a similar number of independent cDNA-carrying clones was obtained after this transfer, the mean size of the cDNA was clearly reduced from 1.75 ± 0.82 kb to 1.34 ± 0.66 kb (± standard deviations [SD]; n = 60; P < 0.05, Mann-Whitney U test), as shown in Table 1. Generally, the short DNA fragment tends to be more efficiently amplified during PCR. This property may account for the reduction of the cDNA size. However, omission of the PCR resulted in an obvious reduction of the number of independent cDNA-carrying clones by about 1/25. Thus, the PCR amplification before the BP reaction was indispensable.

TABLE 1.

Quality of cDNA libraries

cDNA library	No. of primary clones	Mean insert size ± SD (kb)a
Original	1 × 107	1.75 ± 0.82
Entry	1.5 × 107	1.34 ± 0.66
Vector	8 × 107	1.26 ± 0.65
Viral	NDb	0.71 ± 0.54

^aMean insert size was determined by electrophoresis of PCR fragments from 60 bacterial colonies (original, entry, and vector libraries) or from 190 cDNA clones in viral cDNA library-infected cells.

^bND, not done.

Transfer of the entry cDNA library into a lentivirus vector.

Next, the entry cDNA library was incubated with a lentivirus vector DNA that had derivatives of the R1 and R2 sites in the presence of the LR clonase enzyme mix that consists of a mixture of the phage protein excisionase (Xis), Int, and integration host factor. The LR clonase recombines the L1 and R1 sites as well as the L2 and R2 sites (LR reaction), and DNA fragments between L1 and L2 were placed between the B1 and B2 sites, respectively. Although cis-acting sequences derived from the lentivirus vector may interfere with wild-type HIV-1 replication, a part of the vector with a deletion of the U3 region (self-inactivating vectors) was not responsive to the interference (3). Therefore, a self-inactivating vector was used in this study. The cDNA library transferred into the lentivirus vector DNA was referred to as the vector cDNA library. Because this lentivirus vector expresses the cDNA library, under the control of elongation factor α promoter, along with GFP expression from a single bicistronic transcript, cDNA library-transduced cells are easily identified by flow cytometry or fluorescent microscopy. After this transfer, no reductions in the number or cDNA size of independent cDNA-carrying clones were observed (Table 1; Fig. 1), suggesting that the library inserted between the L1 and L2 sites would be transferred into another vector without significant loss of library complexity.

FIG. 1.

Histogram analysis of lengths of individual cDNA fragments in each library. The lengths of cDNA fragments were determined as described in Materials and Methods and are plotted in 250-bp increments on the x axes. Percentages of individual clones are indicated on the y axes.

Generation of a cDNA library-expressing lentivirus vector.

Next we prepared a cDNA library-expressing lentivirus vector, referred to as the viral cDNA library, via cotransfection of 293T cells with vector cDNA library DNA, a VSV-G protein expression DNA, an HIV Gag-Pol expression DNA, and an HIV Rev expression DNA. The infectious titer was approximately 4 × 10⁶/ml, measured by using a human CD4⁺-T-cell line, MT-4 cells. On the other hand, the infectious titer of the parental lentivirus vector with no cDNA inserted, CSII-EF-MCS-IRES-hrGFP (9), was 10 to 100 times higher than that of the viral cDNA library (data not shown). The average size of cDNA fragments in the transduced cells was 0.7 kb, which was shorter than that of cDNA fragments in the vector cDNA library, suggesting that the smaller cDNAs were enriched during lentivirus preparation and its infection into cells (Table 1; Fig. 1). To overcome this problem, size fractionation to enrich long cDNA fragments should be performed in future experiments. Nevertheless, some transduced cDNAs were more than 2,000 bp (Fig. 1), suggesting that this vector system can transduce more than 2,000-bp cDNA fragments.

Cloning of genes that prevent cells from HIV-1-induced cell death.

Figure 2 shows an outline of the selection system used to isolate anti-HIV genes from the library used in this study. Twelve million MT-4 cells were infected with the viral cDNA library at an MOI of approximately 0.68. The total number of cDNA-transduced cells was estimated to be around 8 × 10⁶, which was slightly smaller than the number of independent clones of the original cDNA library. Three days after cDNA transduction, the cells were challenged with HIV-1_NL4-3 at an MOI of 0.05. About 30 days after HIV-1 challenge, when nontransduced culture cells had been completely killed, surviving cells, all of which were continuously growing, and GFP⁺ cells were collected and cellular DNA was extracted. The cDNA fragments were recovered by PCR with B1 and B2 primers and transferred into pDONR201 vector DNA through the BP reaction. Then the cDNA sublibrary-expressing lentivirus was generated. After subsequent screening through transduction of the cDNA sublibrary in MT-4 cells and subsequent HIV-1 challenge, more than 25 independent cDNA clones were isolated, which were confirmed in further experiments to confer the cytopathic effect (CPE)-free phenotype in the transduced cells after HIV-1 challenge. Sequence analysis revealed that these clones contained full-length CD14 cDNA, and their sequence was identical to that of BC010507 in the GenBank database. Flow cytometric analysis showed that the anti-CD14 antibody reacted only with the cDNA clone-transduced CD4⁺ cell population (Fig. 3A) identified by GFP expression (Fig. 3B and C).

FIG. 2.

Scheme for strategy used to select genes that arm cells with resistance to HIV-induced CPE. MT-4 cells were infected with the viral cDNA library and then challenged with HIV-1_NL4-3, which is highly cytopathic to MT-4 cells. If the introduced gene has anti-CPE, the cell will survive in the presence of HIV-1.

FIG. 3.

Characterization of a cDNA clone that confers T-cell resistance to HIV-1-induced CPE. Cells transduced with the CD14-carrying vector isolated from this viral cDNA library were stained with anti-CD4 antibody (A), anti-CD14 antibody (B), or isotype-matched control antibody (C) and analyzed by flow cytometry. The results shown are data from one flow cytometry experiment, which is representative of three independent experiments. PE, phycoerythrin.

To verify the effect of CD14 on HIV-1 infection, we independently prepared three cell lines: MT-4 cells transduced with a lentivirus vector that express CD14 along with GFP from a single bicistronic transcript, MT-4 cells transduced with an H-2K^k-expressing lentivirus vector, and nontransduced MT-4 cells. Flow cytometric analysis confirmed that all GFP-expressing cells simultaneously and persistently expressed CD14 on the cell surface (Fig. 4A). The three cell lines were mixed, and the cultures were challenged with wild-type HIV-1. Before HIV-1 infection, the mixed culture consisted of three cell populations: GFP⁺ H-2K^k−, GFP⁻ H-2K^k+, and GFP⁻ H-2K^k− cells (Fig. 4B). Only the GFP⁺ H-2K^k− population survived after HIV-1 infection (Fig. 4C). In contrast, the proportion of the three cell types was consistently maintained in HIV-1-uninfected cultures (Fig. 4D). Trypan blue staining confirmed that all dead cells were GFP⁻ and all GFP⁺ cells were alive (Fig. 4E, F, and G). A subsequent flow cytometric analysis of cells stained by anti-HIV-1 human sera indicated that the CD14-transduced MT-4 cells also expressed HIV-1 antigen (Fig. 4H and I). When the CD14 gene was transduced into human CD4⁺ CCR5⁺ HeLa cells, these cells were also susceptible to HIV-1 infection but resistant to HIV-1-induced cell death (data not shown). To reveal the mechanism of how CD14 is blocking the HIV-1-induced cell death, the effect of CD14 for HIV-1 replication was examined. The surface expression of neither CD4 nor CXCR4 was altered in CD14-transduced MT-4 cells (data not shown). On the other hand, the HIV-1 replication in CD14-transduced cells determined by production of p24^gag antigen in culture supernatant was significantly lower than that in control vector-transduced cells (Fig. 5A and B). To determine the level at which HIV-1 replication is inhibited by CD14, we used a real-time PCR assay to detect individual forms of viral cDNA at various times after HIV-1 infection. Since the preceding transduction with an HIV-1-based lentiviral vector will hamper the real-time PCR assay to measure the level of newly synthesized HIV-1 cDNA only originated from subsequent wild-type HIV-1 infection, a CD14-expressing plasmid DNA or control vector DNA was transfected into CD4⁺ CCR5⁺ HeLa cells. More than 70% of cells were confirmed to express the CD14 molecule on the cell surface determined by flow cytometry 2 days after transfection, and then the culture was infected with HIV-1_NL4-3 (X4 virus). The levels of early reverse transcripts, late reverse transcripts, 2-LTR circle, and the integrated form were significantly lower in the CD14-expressing culture than in control culture (Fig. 5C, D, E, and F). However, when CD4⁺ CCR5⁺ HeLa cells were infected with HIV-1_NL4-3 and then 2 days later transfected with the CD14-expressing DNA, the HIV-1 release, determined by production of the p24^gag antigen in the culture supernatant, was similar in both the CD14-transfected and control cultures (data not shown). These data indicate that CD14 appears to partially inhibit the entry step in HIV-1 replication and provide the HIV CPE-free phenotype.

FIG.4.

Resistance to HIV-1-induced cell death in CD14-transduced T cells. (A) Flow cytometric analysis of MT-4 cells infected with a CD14- and GFP-expressing lentivirus vector was performed. These cells (GFP⁺ H-2K^k−) were mixed with cells infected with a lentivirus vector expressing H-2K^k alone (GFP⁻ H-2K^k+) and uninfected cells (GFP⁻ H-2K^k−) and then challenged with HIV-1_NL4-3. Flow cytometric analysis of the mixed culture is shown before HIV-1 challenge (B), 8 days after HIV-1 challenge (C), and 8 days after mock infection (D). Trypan blue staining (E) and fluorescent microscopic examination (F) were performed 3 days after HIV-1 challenge. A merged image of panels E and F is shown in panel G. Magnification, ×200. Flow cytometric analysis of the HIV-1-challenged culture 10 days after infection (H) or of an uninfected culture (I) was performed by staining with anti-HIV human serum. The results shown are data from one experiment, which is representative of three independent experiments.

FIG. 5.

Inhibition of HIV-1 replication in CD14-transduced cells. HIV-1 replication was evaluated by production of p24^gag antigen in the culture supernatant of CD14- or empty vector-transduced MT-4 (A) or CD4⁺ CCR5⁺ HeLa cells (B) with a lentivirus vector expressing CD14 and H-2K^k or H-2K^k alone, respectively. To determine the level of HIV-1 entry efficiency, CD14- or empty vector-transfected CD4⁺ CCR5⁺ HeLa cells were challenged with DNase-treated HIV-1_NL4-3. Target cell DNA was isolated at the indicated time and used to detect early reverse transcripts (C), late reverse transcripts (D), 2-LTR circle (E), and the integrated form (F). Data are the means ± SD from duplicate experiments. The levels of the p24^gag antigen or HIV-1 DNA in the CD14-transduced cultures were significantly lower than those of the empty vector-transduced cultures (P < 0.05, Mann-Whitney U test). Lines with open circles, CD14 vector; lines with filled circles, empty vector.

DISCUSSION

In this study, we established a lentivirus vector system to transduce a cDNA library into human T cells and successfully isolated an anti-CPE gene against HIV-1 infection. A wide variety of expression systems in mammalian cells have been developed, for example, plasmid-based, virus-based, and transposon-based systems. Among them, retrovirus-based and lentivirus-based expression systems can stably transduce genes into human T cells and are, furthermore, efficient enough to screen 10⁷ genes. Recently, van Maanen et al. (21) reported the potential use of a lentivirus vector in an expression cloning system. However, the vast majority of cDNA libraries have not been yet constructed on lentivirus vectors. If a cDNA library is efficiently constructed on a lentivirus vector, this vector system will be strongly powerful to isolate genes and accelerate functional genomics. Hence, we used a site-specific recombination system, Gateway, for transferring a cDNA library from a transient expression vector to a lentivirus vector. The major advantage of this transfer system is that we can apply the system to many already established libraries. For instance, a cDNA library amplified in phage can be transferred to a mammalian expression vector including the lentivirus vector without significant loss of its complexity. Many reports have suggested that this technology allows for easy transfer of individual cDNA fragments, and this technology has become a powerful tool for a high-throughput screening system in functional genomics (8, 14, 18, 19). In this study, we used a premade cDNA library in a transient expression vector that cannot be efficiently introduced into T cells. Here, we successfully transferred the cDNA library to a lentivirus vector and found that CD14 can confer resistance to HIV-induced cell death in the transduced cells. This observation suggests that the transferred libraries can still be applied for a functional screening system. Because the vector has also been used to transduce genes into nondividing cells such as neurons (11, 12), muscles (7), and hematopoietic stem cells (4), our lentivirus-based system can be applied to expression cloning systems that use such cells.

However, there is a need to improve our library transfer system. Some degree of loss in library complexity was noted. In our experiment, some of the cDNAs, especially long cDNA fragments, appeared to be lost during two steps: PCR amplification and lentivirus vector production and/or infection. This problem would be solved when the long cDNA fragments are enriched before the BP reaction and a genetic library is directly generated on a donor vector for the LR reaction. We think the latter strategy is useful because, if a library is constructed on a donor vector, the library can be transferred to various expression vector systems, which include not only the lentivirus vector but also other traditional vector systems, by only the LR reaction. As shown in Table 1, we did not observe the loss of library complexity during the LR reaction in our experiment. To overcome the loss of parts of the long cDNA fragments during production of lentivirus vector and/or infection, shorter parent lentiviral vector DNAs should be used in future experiments. In the present study, we used a GFP-expressing lentivirus vector DNA (CSII-EF-MCS-IRES-hrGFP). To obtain long cDNA inserts in the lentiviral vector, there is probably a need to delete some parts of the fragment within the vector DNA, such as IRES-hrGFP.

It was reported that the lentivirus (derived from HIV-1) vector-transduced T cells are less susceptible to wild-type HIV-1 infection than nontransduced T cells (3). The transcripts transduced by the vector appears to compete efficiently for encapsidation, resulting in inhibition of its infectivity, probably because cis-acting sequences in the lentivirus vector are responsive to the regulatory protein of wild-type HIV-1. However, the inhibitory effect was completely eliminated in a self-inactivating vector (3). Thus, we used a self-inactivating vector, and we could not actually find any differences in its HIV-1 replication ability between the transduced and nontransduced cells (data not shown).

In the present study, we used a cDNA library as a functional genetic element. In the future, we will be able to choose different genetic libraries, such as ribozyme (8) and peptide libraries (23, 25). The ribozyme library can be efficiently expressed under the control of an RNA polymerase III-dependent promoter (8). When constructing a lentivirus vector containing such a library, the Gateway-based transfer system will be useful. Moreover, since the length of such a library is more homogeneous and shorter than a conventional cDNA library, the lentivirus vector system will be able to more potently deliver a ribozyme library than a cDNA library.

CD14 is known as a coreceptor molecule for lipopolysaccharide (LPS) (24) and is expressed on the surface of myeloid cells via a glycosylphosphatidyl inositol tail. LPS binds to a serum protein, LPS-binding protein (15), and associates with CD14. Subsequently, LPS stimulates Toll-like receptor 4 (13) and activates signaling pathways, mainly the nuclear factor-κB (NF-κB) pathway. HIV-1 also preferentially infects macrophages that express CD14. It is known that macrophages are one of the major target cells for HIV infection, and they behave as cellular reservoirs of virions in HIV-infected patients, probably because the cells are relatively resistant to HIV-induced CPE (5). Although the mechanisms of the low susceptibility of macrophages to HIV-1-induced cell death are poorly understood at present, some explanations may be brought up from the resistance of CD14-transduced cells to HIV-1-induced cell death. One explanation is that overexpression of CD14 can trigger cell survival signals such as NF-κB or induce antiapoptotic genes. Another explanation is that CD14 can reduce the cytotoxicity of HIV-1 infection in T cells through a partial inhibition of HIV-1 replication. In fact, CD14 overexpression resulted in an inhibition of the entry step on HIV-1 replication, as shown in Fig. 5. A determination of the exact mechanisms of CD14 function in HIV-infected cells should enhance our understanding of the cellular events during HIV-induced cell death, which results in immune destruction in HIV-infected individuals.

In conclusion, application of the Gateway system to a genetic library transfer system will allow the use of the lentivirus vector system as a powerful tool for the study of functional genomics of mammalian cells.