ABSTRACT
Recombinant rabies virus (RV) vaccine strain-based vectors have been successfully developed as vaccines against other viral diseases (J. P. McGettigan et al., J. Virol. 75:4430-4434, 2001; McGettigan et al., J. Virol. 75:8724-8732, 2001; C. A. Siler et al., Virology 292:24-34, 2002), and safety concerns have recently been addressed (McGettigan et al., J. Virol. 77:237-244, 2003). However, size limitations of the vectors may restrict their use for development of vaccine applications that require the expression of large and multiple foreign antigens. Here we describe a new RV-based vaccine vehicle expressing 4.4 kb of the human immunodeficiency virus type 1 (HIV-1) Gag-Pol precursor Pr160. Our results indicate that Pr160 is expressed and processed, as demonstrated by immunostaining and Western blotting. Electron microscopy studies showed both immature and mature HIV-1 virus-like particles (VLPs), indicating that the expressed HIV-1 Gag Pr55 precursor was processed properly by the HIV-1 protease. A functional assay also confirmed the cleavage and functional expression of the HIV-1 reverse transcriptase (RT) from the modified RV genome. In the next step, we constructed and recovered a new RV vaccine strain-based vector expressing a chimeric HIV-189.6P RV envelope protein from an additional RV transcription unit located between the RV nucleoprotein (N) and phosphoprotein (P) in addition to HIV-1 Pr160. The 2.2-kb chimeric HIV-1/RV envelope protein is composed of the HIV-1 Env ectodomain (ED) and transmembrane domain (TD) fused to RV glycoprotein (G) cytoplasmic domain (CD), which is required for efficient incorporation of HIV-1 Env into RV particles. Of note, the expression of both HIV-1 Env and HIV-1 Pr160 resulted in an increase in the rhabdoviral genome of >55%. Both rhabdovirus-expressed HIV-1 precursor proteins were functional, as indicated by RT activity and Env-based fusion assays. These findings demonstrate that both multiple and very large foreign genes can be effectively expressed by RV-based vectors. This research opens up the possibility for the further improvement of rhabdovirus-based HIV-1 vaccines and their use to express large foreign proteins, perhaps from multiple human pathogens.
Viral vectors expressing proteins of human pathogens show great promise as vaccine vehicles against other infectious diseases. As of today, human immunodeficiency virus type 1 (HIV-1) is the primary target for most of these vaccine vectors, and a large body of information regarding their use is available (for review, see references 2, 30, 33, 39, and 46). These results indicate that each viral vector has its own unique characteristics, and some of these features point toward the use of virus vectors as a vaccine vehicle in humans. In general, research indicates that the viral vector should not be blocked by a preexisting immune response in the human population caused by natural infection with the viral vector or vaccination against the virus used. In addition, the vector should be able to induce both humoral and cellular responses against the expressed antigen(s). Other important requirements are low vector-induced pathogenicity, large cloning capacity for foreign antigens, and high stability of the respective foreign gene in the recombinant vaccine vector.
Rabies virus (RV) is a nonsegmented negative-stranded RNA virus within the family Rhabdoviridae (54). The RV genome is about 12 kb in size and encodes five monocistronic RNAs encoding the nucleocapsid protein (N), phosphoprotein (P), matrix protein (M), the single transmembrane protein G, and the viral polymerase (L) (11). The N protein encapsidates the viral RNA, which is the template for replication and transcription by the viral polymerase complex composed of the P and L proteins (52, 53). The RV M protein bridges the RNP with the cytoplasmic domain (CD) of RV G protein in the host cell-derived viral membrane (35). The RV G protein mediates infection of the host cell (14).
The possibility of using rhabdoviruses such as RV or vesicular stomatitis virus (VSV) as expression vectors was shown earlier for the model gene coding for chloramphenical acetyltransferase (CAT), and the results from these studies indicated that foreign genes can be expressed stably (34). In the next step, the possibility for the use of RV as an HIV-1 vaccine was pursued, and several studies have shown that expression of HIV-1 Env or Gag results in potent immune responses directed against HIV-1 in a small animal model (30, 32, 47).
As mentioned above, safety is a major concern for the use of every live viral vector. However, replication-deficient vectors do not induce responses as potent as those seen for replication-competent vectors, and considering the promising results with an attenuated simian immunodeficiency virus (SIV)-based vector (13), a replication-competent vaccine vector may be required. For VSV, one report indicates that the deletion of the VSV G protein CD can reduce pathogenicity in mice, whereas full-length G protein-containing vectors induce a transient disease in mice after intranasal inoculation, as indicated by 10 to 20% weight loss (40). However, intramuscular inoculation with VSV-based HIV-1 vaccine vector did not cause any vector-associated disease in mice or rhesus macaques (40-42). Even though RV-based vectors are very safe after peripheral inoculation, we recently described several new, highly attenuated second-generation RV-based vaccine vehicles expressing HIV-1 Gag. Our results indicated that by exchanging the arginine at position 333 (R333) within the RV glycoprotein (G) with glutamic acid (E333) or by deleting the RV G CD, the vectors expressing HIV-1 Gag were completely safe in mice and apathogenic after intracranial infection. Of note, these highly attenuated RVs expressing HIV-1 Gag were still as potent as the wild-type RV vectors at inducing cellular responses against HIV-1 Gag (31).
RV-based vectors fulfill most of the basic requirements for a successful HIV-1 vector,but size limitations have been a concern, even though genes up to 2.5 kb have been expressed by RV-based vectors (37, 47). In general, DNA viruses have the largest cloning capacities and are able to express two genes from one vector, whereas positive-stranded RNA viruses seem to be very restricted regarding the size of a foreign gene (for review, see reference 46). For rhabdoviruses, the ability to express foreign sequence up to 4.0 kb was shown for VSV expressing HIV-1 Gag and Env protein (1). Our laboratory has previously shown that both of these genes can be individually expressed from RV-based vaccine vectors (32, 47). The recent finding that rhesus macaques inoculated with an experimental HIV-1 vaccine were not protected, as previously thought, but died within 2 to 3 years after SIV challenge due to mutation within the Gag gene indicates that multiple HIV-1 genes may be required and that a humoral response might be desirable (6, 7). The expression of a large number of HIV-1 proteins from a single RV-based vector may therefore be advantageous. Here we report the expression of the 4.4-kb HIV-1 Pr160 protein from an RV vaccine vector. Pr160 was processed in an HIV-1-like fashion, as indicated by immunostaining, Western blotting, and electron microscopy (EM) showing mature and immature HIV-1 virus-like particles. Moreover, we were able to create a recombinant RV expressing functional HIV-1 Pr160 and Env from a single viral genome. These findings demonstrate that RV-based vaccine vectors can encode more than 50% of their genome size in foreign sequences. Moreover, the genes are expressed functionally from two different transcription units, which further emphasizes their potential utility as efficacious antiviral vaccines.
MATERIALS AND METHODS
Plasmid construction.The plasmids encoding the recombinant RV vaccine vector, pSPBN-333, pSPBN-333-Gag, and pBNSP, were described previously (31). To create a new RV vaccine vector expressing HIV-1 Pr160, the gene encoding the Gag-Pol precursor protein was PCR amplified from pNL4-3 (AIDS Research and Reference Reagent Program [ARRRP]) by using Vent polymerase (Biolabs, Inc.) and the primers RP40 (5′-AAACTCGAGCGTACGACAATGGGTGCGAGAGCGTCA-3′) and RP184 (5′-AAAGCTAGCTTAATCCTCATCCTGTCTAC-3′). The PCR product was digested with BsiWI-NheI and ligated to the previously BsiWI-NheI-digested plasmid pSPBN-333. This plasmid was designated pSPBN-Pr160. To construct a recombinant RV vector, BNSP, which expresses HIV-189.6P Env containing the ectodomain (ED) and transmembrane domain (TMD) of HIV-189.6P Env fused to RV G cytoplasmic domain (CD), the ED and TMD of HIV-189.6P Env were PCR amplified from pKB9SHIV(89.6P) (ARRRP) by using Vent polymerase and the primers RP27 (5′-GGGCTGCAGCTCGAGCGTACGAAAATGAGAGTGAAGGAGATCAGG-3′) and RP32 (5′-GCCCCGTTAACTATAGAAAGTACAGCAAAA-3′). The PCR product was digested with BsiWI-HpaI and cloned into pBS2H-NL4-3-G (18). The resulting plasmid was designated pBS289.6P-RVG. To introduce the chimeric HIV-1/RV G Env gene into the BNSP vector, pBS289.6P-RVG was digested with BsiWI-XbaI, and the 2.2-kb fragment was gel eluted and cloned into the previously BsiWI-NheI-digested plasmid pBNSP. This plasmid was named pBNSP-89.6P-RVG. To create an RV expressing both HIV-189.6P and HIV-1NL4-3 Pr160, pBNSP-89.6P-RVG and pSPBN-Pr160 were digested with NotI-PmlI, and the Env-containing fragment was cloned to the Pr160-containing backbone. This plasmid was designated p89.6-P-RVG-Pr160.
To construct p89.6P-RVG-Pr160 with the human cytomegalovirus (CMV) promoter, T7 RNA polymerase promoter, and a hammerhead ribozyme at the 3′ end of the RV antigenome, a PCR strategy was used to create a new RV vector (cSPBN). First, an ∼700-bp DNA fragment containing the CMV T7 promoter was PCR amplified from pcDNA3.1(+) by using Vent polymerase and the primers RP202 (5′-AAACGCTAGCCAGCTTGGG-3′) and RP203 (5′-TTTCTGCAGCGCGTTGACATTGATTATTGAC-3′). An ∼400-bp-long DNA fragment containing the HH ribozyme, the 3′ end of the RV antigenome, and part of the N gene was PCR amplified with pSPBN and the primers RP204 (5′-TTTTCTAGATTAAGCGTCTGATGAGTCCGTGAGGACGAAACCCGGCGTACCGGGTCACGCTTAACAACCAGATCAAAG-3′) and RP76 (5′-CGGGAGCCTTTCCTAGGG-3′). The first PCR product was digested with NheI, and the second PCR fragment was digested with XbaI, the fragments were ligated, and the 1.1-kb fragment was eluted from the gel. This fragment was reamplified with the primers RP203 and RP76, gel purified, and digested with AvrII-PstI and then was ligated to the previously AvrII-PstI-digested pSPBN. This plasmid was designated cSPBN. To introduce the new promoters and HH ribozyme into p89.6-P-RVG-Pr160, cSPBN was digested with NcoI-PmlI, and the 4.9-kb DNA fragment was cloned into the previously NcoI-PmlI-digested plasmid p89.6-P-RVG-Pr160. The resulting plasmid was named c89.6-P-RVG-Pr160.
Generation of recombinant viruses.For recovery experiments with the recombinant RVs, the previously described RV recovery system was used (16, 47). Briefly, BSR-T7 cells (9), which stably express T7 RNA polymerase, were transfected with 5 μg of full-length RV cDNA in addition to plasmids coding for the RV N, P, L, and G proteins by using a Ca2PO4 transfection kit (Stratagene), as instructed by the vendor. Three days posttransfection, supernatants were transferred onto fresh BSR cells, and infectious RV was detected 3 days later by immunostaining with fluorescein isothiocyanate (FITC) against the RV-N protein (Centacor, Inc.).
Sucrose purification of RV virions.BSR cells (107) in a T150 flask were infected for 60 h at a multiplicity of infection (MOI) of 1 with SPBN-333, SPBN-333-Gag, or SPBN-Pr160; the supernatants were preclarified at 14,000 rpm for 3 min in a tabletop centrifuge (Eppendorf, Inc.) and spun over 20% sucrose in an SW28 rotor (Beckman, Inc.) at 25,000 rpm for 1 h. Virion pellets were resuspended in 100 μl of lysis buffer (50 mM Tris [pH 7.4], 150 mM NaCl, 1% NP-40, 0.1% sodium dodecyl sulfate [SDS], 1× protease inhibitor cocktail; Sigma).
Western blotting (i) Virions.Proteins from lysed virions were separated by SDS-polyacrylamide gel electrophoresis (PAGE) (10% polyacrylamide) and transferred to a polyvinylidene difluoride membrane (PVDF-Plus; Osmonics, Minnetonka, Minn.). Blots were blocked for 1 h (5% dry milk powder in phosphate-buffered saline [PBS], pH 7.4), washed three times with Western blot wash solution (WBWS; 0.1% PBS-Tween 20), and incubated with a human anti-HIV-1 p24 antibody (1:500), a rabbit anti-HIV-1 reverse transcriptase (RT) antibody (1:4,000), or a polyclonal rabbit anti-RNP antibody (1:2,000) overnight at 4°C (17). Blots were then washed three times with WBWS. Secondary Alexa Fluor 546 goat anti-human immunoglobulin G (IgG) (1:500) or Alexa Fluor 532 goat anti-rabbit IgG (1:1,000) (Molecular Probes, Inc.) was added, and blots were incubated for 2 h at room temperature. Blots were washed three times with WBWS and washed once with PBS (pH 7.4). Fluorescence analysis was performed with Molecular Imager FX (Bio-Rad) with QuantityOne as instructed by the vendor (Bio-Rad).
(ii) Cell lysates.BSR cells were infected with SPBN-333, SPBN-333-Gag, SPBN-Pr160, or 89.6P-RVG-Pr160 at an MOI of 5 for 48 h and resuspended in 0.5 ml of lysis buffer on ice for 5 min. The suspension was transferred to a microcentrifuge tube and spun for 3 min at 14,000 rpm to remove cell debris. Proteins were separated by SDS-PAGE (10% polyacrylamide) and transferred to a PVDF-Plus membrane (Osmonics). Membranes were blocked with 5% milk powder in PBS (pH 7.4) for 1 h at room temperature and then probed with a polyclonal sheep antibody directed against HIV-1 gp120 (1:1,000) (ARRRP) or a human monoclonal antibody directed against p24 (1:1,000) (ARRRP) overnight at 4°C. After three 10-min washes with WBWS, blots were probed for 1 h with horseradish peroxidase (HRP)-conjugated donkey anti-sheep antibody (1:5,000) or an HRP-conjugated goat anti-human antibody (1:10,000) (Jackson ImmunoResearch Laboratories). After three 10-min washes with WBWS and one wash with PBS (pH 7.4), chemiluminescence was performed as instructed by the manufacturer (NEN).
HIV-1 p24 antigen-capture ELISA.HeLa or BSR cells were infected at an MOI of 5, and 48 h later, supernatants were collected and cells were resuspended in lysing buffer (Triton X-100 in PBS and 2-chloroacetamide). The supernatants and cell suspensions were transferred to microcentrifuge tubes and spun for 2 min at 14,000 rpm to remove cell debris. The quantification of Gag p24 protein in cell supernatants and lysates was performed using the p24 antigen enzyme-linked immunosorbent assay (ELISA), as described by the manufacturer (ZeptoMetrix, Inc.).
Multicycle growth and one-step growth curves.BSR cells (a BHK-21 clone) were plated in 60-mm-diameter dishes and 16 h later were infected at an MOI of 0.01 (multicycle growth) or 5 (one-step growth) with SPBN-333, SPBN-333-Gag, SPBN-Pr160, or 89.6P-RVG-Pr160. After incubation at 37°C for 1 h, inocula were removed, and cells were washed four times with PBS to remove any unabsorbed virus. Three milliliters of complete medium was added back, and 100 μl of tissue culture supernatants was removed at the indicated time points after infection. Titers of virus aliquots were determined in duplicate on BSR cells.
EM.Human HeLa cells were infected with SPBN-Pr160 for 48 h. Cells were fixed at room temperature in neutral buffered 2.5% glutaraldehyde and gelled into warm agar. They were post-fixed in 1% OsO4, dehydrated in graded ethanol and propylene oxide, and embedded in Spurr's epoxy. Thin sections were cut and stained with uranly acetate and lead citrate and examined with an LEO EM10 electron microscope at 60 kV.
RT assay.BSR or HeLa cells were infected at an MOI of 5 with SPBN-333-Gag, SPBN-Pr160, or 89.6P-RVG-Pr160. Supernatants were harvested at the indicated time points and analyzed as indicated by the manufacturer (Innovagen AB, Sweden; info@innovagen.se).
Tetramer analysis of splenic lymphocytes.Groups of 6- to 8-week-old female BALB/c mice were inoculated intramuscularly with 106 FFU of recombinant RV expressing HIV-1 Gag and Pol (SPBN-Pr160) or Gag, Pol, and Env (89.6PRVG-Pr160). Six weeks postimmunization, mice were challenged with vaccinia virus expressing HIV-1 gag (vP1287). Five days later, spleens were removed, single-cell suspensions were prepared, and erythrocytes were lysed with ACK Lysing buffer. A total of 3 × 106 cells were washed twice in fluorescence-activated cell sorter (FACS) buffer (PBS supplemented with 2% fetal bovine serum) and incubated with rat anti-mouse CD16/CD32 (1 μg/106 cells; Fc Block; Pharmingen) and 33-μg/ml unconjugated streptavidin (Molecular Probes) for 1 h on ice. Cells were washed twice with FACS buffer and incubated with 2 μl of tetramer (H-2Kd/AMQMLKETI) from HIV-1 Gag p24 protein and 2-μg/ml FITC-conjugated rat anti-mouse CD8 antibody (Caltag) for 30 min at room temperature with occasional agitation. Cells were washed three times with FACS buffer, resuspended in 300 μl of PBS containing 2% paraformaldehyde, and analyzed by FACScan (Beckman Coulter XL analyzer with a 488-nm argon ion laser. The fluorescence was excited at 488 nm and analyzed through standard fluorescein optics. The FITC and phycoerythrin fluorescence were collected through a 525 and 575 bandpass filters, respectively.).
RESULTS
Construction of recombinant RVs expressing HIV-1 Gag-Pol precursor Pr160.Our previous studies have shown that foreign proteins such as HIV-1 Env; HIV-1 Gag; or hepatitis C virus E1, E2, and core protein are stably expressed by RV-based vaccine vectors and are able to induce vigorous immune responses in vaccinated mice (29, 32, 51). However, the size of the expressed antigens in these previous studies was limited, and larger proteins might have to be expressed to protect from certain infectious diseases. In the case of HIV-1, recent research results from the rhesus macaque model system indicate that expression of SIV Gag is not sufficient to protect animals from a simian-human immunodeficiency virus (SHIV) challenge due to an SIV Gag-directed cytotoxic T lymphocyte (CTL) escape mutant (6, 7). In addition to HIV-1 Gag, a strong CTL response against HIV-1 Pol was found in long-term nonprogressors and is suggested, at least in part, for protection against AIDS. The HIV-1 Pol precursor protein is expressed by frame-shifting of the Gag-Pol gene reading frame, and the HIV Gag-Pol and Gag products are synthesized at a ratio of ∼1:20 (55). To obtain expression of the HIV-1 Pr160, the sequence encoding HIV-1 Pr160 was cloned into the RV vector SPBN-333 between the RV G and L genes (Fig. 1) by utilizing the single BsiWI and NheI restriction sites. The resulting plasmid was designated pSPBN-Pr160. pSPBN-Pr160 was cotransfected with the plasmids encoding the RV N, P, G, and L proteins into BSR cells (a BHK clone) stably expressing T7 RNA polymerase (9) by standard methods (32), and infectious SPBN-Pr160 was detected in 2 out of 24 wells of transfected cells.
Construction of different recombinant vaccine vectors expressing HIV-1 Gag, Gag-Pol, or Env and Gag-Pol. The top of the figure shows the RV vaccine strain-based vector, SPBN-333, which was used for cloning in the HIV-1 GagNL4-3 (SPBN-333-Gag) and the HIV-1 GagPolNL4-3 (SPBN-Pr160) sequences between the G and L proteins. BNSP contains an additional transcription stop/start signal flanked by two unique restriction sites between the N and P genes where the HIV-1 envelope protein from SHIV89.6PG was inserted (BNSP-89.6PG). For incorporation into virion particles, the sequences for the cytoplasmic domain of HIV-1 envelope were replaced with the cytoplasmic domain from RV (BNSP-89.6PG; black box). SPBN-Pr160 was the target to insert SHIV-89.6PG envelope, resulting in a recombinant RV expressing HIV-1 Env, Gag, and Pol proteins (SPBN-89.6PG-Pr160). All constructs contain an amino acid exchange from an arginine (R) to glutamic acid (E) at position 333 of the RV glycoprotein, which has been previously shown to attenuate RV vectors after intracranial inoculation in mice.
Expression of HIV-1 Gag-Pol precursor by recombinant RVs.In the first step to analyze whether the recombinant RVs were expressing HIV-1 Pr160, BSR cells were infected with SPBN-333-Gag or SPBN-Pr160 at an MOI of 0.1 for 48 h, fixed, and immunostained with an antibody directed against HIV-1 Gag p24 or RV N. The result indicated that both RVs were expressing HIV-1 Gag (data not shown). To determine whether the recombinant RV SPBN-Pr160 also expresses HIV-1 Pol, we infected BSR cells with SPBN-333, SPBN-333-Gag, or SPBN-Pr160 (MOI of 5). Sixty hours later, supernatants from the infected cells were harvested, cells and cell fragments were removed, and the virions were purified over 20% sucrose. Virion proteins were separated by SDS-PAGE, transferred to a PVDF-Plus membrane, and analyzed by Western blotting with antibodies directed against HIV-1 Gag p24, HIV-1 RT, or RV RNP. The results showed that RV infected the cells in all three wells and that RV virions were released into the supernatants, as indicated by the detection of the RV N with a rabbit serum directed against RV N (Fig. 2B). In addition, both SPBN-333-Gag- and SPBN-Pr160-infected cells produced immature HIV-1 VLPs, as shown by the detection of full-length HIV-1 Gag p55 with an HIV-1 p24-specific antibody (Fig. 2). In addition, the HIV-1 capsid protein p24 was detected from SPBN-Pr160-infected cells, suggesting the presence of mature HIV-1 VLPs in the supernatants of this cell culture. Because HIV-1 p55 is only cleaved by an HIV-1 protease (55), the expression of a functional HIV-1 protease encoded in the Gag-Pol precursor protein is very likely (Fig. 2A, lane 3). As expected, HIV-1 p55 and/or p24 was not detected in tissue culture supernatants of SPBN-333-infected control cells (Fig. 2A and C). Finally, we screened one of the membranes with an HIV-1 RT-specific antibody, which detected two proteins of the size of HIV-1 RT at 51 kDa and RT containing the RNase H domain migrating at 66 kDa. These two forms of HIV-1 RT can also be found in HIV-1 virions (27, 45) and were only detected in supernatants of cells infected with SPBN-Pr160.
Western blot analysis of purified RV virions and HIV-1 VLPs. BSR cells were infected with SPBN-333, SPBN-333-Gag, or SPBN-Pr160 (MOI of 5) and lysed 60 h later. Virus and VLPs in the culture supernatants were pelleted over 20% sucrose, and viral proteins were separated by SDS-PAGE and subjected to Western blotting with antibodies specific for HIV-1 p24 (α-p24) (A), RV N (α-N) (B), or HIV-1 RT (α-RT) (C). As shown in panel A, expression of HIV-1 Gag was confirmed with a HIV-1 Gag-specific antibody in SPBN-333-Gag and SPBN-Pr160 supernatants. A band at the expected size of the p24 cleavage product was detected in supernatants from cells infected with SPBN-Pr160, but not from SPBN or SPBN-333-Gag infected cells, indicating a functional protease is being expressed from SPBN-Pr160. A protein at the expected size for RV N was detected for all recombinant RVs (B), indicating that all cells were infected with the respective recombinant RV. Two proteins the size of HIV-1 RT at 51 kDa and RT containing the RNase H domain migrating at 66 kDa were detected (C).
Infection of HeLa cells with SPBN-Pr160 results in the release of p55 and p24 as immature and mature HIV-1 VLPs.The detection of p55 and p24 in the sucrose-purified supernatants of SPBN-Pr160-infected cells indicated that most Gag protein was released in the form of HIV-1 VLPs. However, cell or membrane vesicles associated with p24 or p55 might also be detected after sucrose purification, even though an RV infection does not have a cytopathic effect (CPE) on most cell lines, and therefore the occurrence of such “artificial” particles would be limited. To study the generation of the HIV-1 VLPs in greater detail, we infected HeLa cells at an MOI of 1 for 48 h. Cells were fixed at room temperature in neutral buffered 2.5% glutaraldehyde and gelled into warm agar. These cells were post-fixed in 1% OsO4, dehydrated in graded ethanol and propylene oxide, and embedded in Spurr's epoxy. Thin sections were cut and stained with uranyl acetate and lead citrate and examined by EM. The results shown in Fig. 3A indicate a high level of production of RV virions and HIV-1 VLPs. The VLPs are composed of both immature (black arrows) and mature (white arrows) HIV-1 VLPs (Fig. 3B and C).
Evaluation of recombinant RV expressing mature and immature VLPs by electron microscopy. Hela cells were infected with SPBN-Pr160 at an MOI of 1 for 48 h. Cells were fixed at room temperature in neutral buffered 2.5% glutaraldehyde and gelled into warm agar. Cells were postfixed in 1% OsO4, dehydrated in graded ethanol-propylene oxide, and embedded in Spurr's epoxy. Thin sections were cut and stained with uranyl acetate and lead citrate and examined with a LEO EM10 electron microscope at 60 kV. Panel A shows large numbers of bullet-shaped RV particles and HIV-1 VLPs being released from infected cells. Panels B and C show immature (black arrows) and mature (white arrows) HIV-1 VLPs. Original magnifications, ×43,000 (A) and ×131,000 (B and C).
In the next step, we wanted to analyze the Gag production in the supernatants in a more quantitative way. Our previous results with HIV-1 Gag indicated that SPBN-Gag-infected HeLa and BSR cells produce about 4 and 2 ng of Gag, respectively, 48 h after infection, and similar amounts would also be expected from the construct expressing the Pr160. However, it has been shown that high-level expression of the HIV-1 protease can interfere with HIV-1 VLP formation (36). Because RV infection does not have a CPE on most cell lines, such as HeLa, the majority of the HIV-1 Gag detected should be due to HIV-1 VLPs rather than free p55 from lysed cells. The amount of p24 in the supernatants of SPBN-Pr160-infected HeLa cells was 2 ng 48 h after infection, which was very similar to our previous results for SPBN-Gag. Therefore, expression of the active protease did not interfere with VLP formation.
Multicycle and one-step growth curve of recombinant RVs.One concern about a recombinant RV expressing a very large gene such as HIV-1 Pr160 was that this will affect the viral replication cycle, resulting in a recombinant RV that grows slowly or to low titers. To study the replication kinetics of SPBN-333-Gag and SPBN-Pr160 in greater detail, a multicycle growth curve was performed by infecting BSR cells at an MOI of 0.01. As shown in Fig. 4A, both recombinant RVs grew very similarly and reached approximately the same titer at similar time points, indicating there were no differences in viral spread. We previously observed that the recombinant RV expressing HIV-1 Gag did not show any difference in this in vitro setting compared to the parental vaccine vector. The same seems to be true for SPBN-Pr160. We also performed a one-step growth curve to analyze a single cell cycle of viral replication. As shown in Fig. 4B, all three viruses showed approximately the same viral titer at all four time points. Therefore, the ∼35% larger genome of SPBN-Pr160 did not have any major impact on the viral life cycle, unlike the results for rhabdoviruses expressing HIV-1 Env (47) or a foreign gene from a different location within the viral genome (31).
Multicycle-replication and one-step growth curve of recombinant vaccine vectors. BSR cells were infected with SPBN-333, SPBN-333-Gag, or SPBN-Pr160 at an MOI of 0.01 FFU (multicycle growth) (A) or 5 FFU (one-step growth curve) (B). Aliquots of tissue culture supernatants were collected and viral titers were determined in duplicate as indicated.
Processing of RV-expressed Gag-Pol precursor results in functional HIV-1 proteins, including HIV-1 RT.An important feature of a viral expression vector is the ability to express functional proteins, because certain features may be required to mimic the immunogenicity of the parental protein. For example, important classes of neutralizing antibodies against HIV-1 Env are only induced if these epitopes are exposed on intact virions (38). Our results listed above show that the HIV-1 Gag and protease encoded by Pr160 are functional, as indicated by the formation of mature VLPs and the cleavage of p55. However, we were only able to show expression of HIV-1 RT by Western blotting. For these reasons, we used a functional RT assay that determines RT activity. For this approach, BSR or HeLa cells were infected at an MOI of 5 with SPBN-333-Gag and SPBN-Pr160, and samples were analyzed for RT activity 24, 48, and 72 h after infection. As shown in Fig. 5, RT activity was detected in SPBN-Pr160-infected cells, and the RT activity peaked 48 h after infection, with about 5 ng/ml in the supernatants of HeLa cells, dropping to 0.9 ng/ml 72 h after infection. Control infection with SPBN-333-Gag showed only background levels.
Recombinant RV expressing HIV-1 Gag-Pol produces functional reverse transcriptase (RT). BSR or HeLa cells were infected at an MOI of 5 with SPBN-333-Gag or SPBN-Pr160, and samples were analyzed for RT activity 24, 48, and 72 h after infection. RT activity was detected in SPBN-Pr160-infected cells 48 and 72 h after infection in both BSR and HeLa cells. Control infection with SPBN-333-Gag showed only background levels of RT activity.
Construction of a recombinant RV expressing HIV-1 Gag-Pol and HIV-1 Env from a single RV genome by using a highly efficient RV recovery system.The expression of functional HIV-1 Pr160 from the recombinant RVs was encouraging, and this approach is likely to enhance the efficacy of this construct by presenting a larger number of important HIV-1 CTL epitopes to the immune system than SPBN-333-Gag. However, it is questionable whether this approach alone helps to prevent the previously observed escape of a challenge virus from the cellular response (6, 7). It has been speculated that both arms of the immune system might be required to protect against HIV-1 infection or at least to provide long-term protection from disease. Therefore, we decided to include HIV-1 gp160 in our construct. We previously constructed a new RV vaccine vector expressing HIV-189.6P Env in which we replaced the HIV-1 CD with that of the RV G CD (BNSP-333-89.6P-RVG) (Fig. 1). Such a vaccine vector could be used in addition to SPBN-Pr160, but differences in the replication rates of both viruses might interfere with the induction of an immune response against both antigens. This problem could be circumvented if both genes are expressed from a single vector. Using the single NotI and PmlI restriction sites in pBNSP-89.6P-RVG and pSPBN-Pr160 as a target, we constructed the new RV vector p89.6PG-Pr160. Of note, this construct encodes two foreign genes, thereby increasing the RV genome by 55%. Using our standard RV recovery protocol described above, we failed to recover p89.6PG-Pr160 after three independent experiments using four six-well plates each. Using synthetic genomic RNA analogs, Conzelmann and Schnell previously showed that the recovery frequency is dependent on the size of the respective genome (12), and therefore we were concerned about the possibility to recover such a large recombinant virus. In addition, rhabdoviruses containing an HIV-1 Env protein always grew to a lower titer, which raised additional concerns about the recovery of this RV (47). It has been shown previously that the inclusion of a hammerhead ribozyme 5′ of the RV antigenome, in addition to the 3′ hepatitis virus delta ribozyme, greatly improves the generation of an infectious RV from cDNA (23). We therefore included these sequences in our RV vectors. Using the new c89.6PG-Pr160 instead of p89.6PG-Pr160, we were able to recover 89.6PG-Pr160 in 2 out of 24 wells of transfected cells.
Both HIV-1 Pr160 and Env are expressed from 89.6P-RVG-Pr160.To ensure the expression of the HIV-1 proteins by the recombinant RV, cell lysates from SPBN-333-, SPBN-333-Gag-, SPBN-Pr160-, or 89.6P-RVG-Pr160-infected cells were analyzed by Western blotting with an antibody directed against HIV-1 p24 or Env. As expected, p55 was detected for all three Gag-expressing viruses, whereas p24 was only present in cell lysates from SPBN-Pr160- or 89.6P-RVG-Pr160-infected cells (Fig. 6A, lanes 3 and 4). Immunoblotting with a gp120-specific antibody detected one specific-band at about 120 kDa from 89.6P-RVG-Pr160 lysates (Fig. 6B, lane 8). Whereas the presence of functional protease was indicated by p24 in the lysates from 89.6P-RVG-Pr160-infected cells, the expression of functional HIV-1 Env from 89.6PG-Pr160 was analyzed by a fusion assay in a human T-cell-line (Sup-T1). Because wild-type RV infects cells by receptor-mediated endocytosis, the RV glycoprotein (G) can only cause fusion of infected cells at a low pH (56). In contrast to SPBN-Pr160-infected cells, large syncytium formation was detected in Sup-T1 cells 48 h after infection with 89.6PG-Pr160 (see Fig. 8).
Western blot analysis of recombinant RV expressing HIV-1 Gag-Pol and Env. BSR cells were infected with SPBN-333, SPBN-333-Gag, SPBN-Pr160, or 89.6P-RVG-Pr160 (MOI of 2) and lysed 48 h later. Proteins were separated by SDS-PAGE and subjected to Western blotting with antibodies specific for HIV-1 p24 (α-p24) (A) or HIV-1 Env (α-gp120) (B). (A) Expression of HIV-1 p55 from SPBN-333-Gag, SPBN-Pr160, or 89.6P-RVG-Pr160 was confirmed with an HIV-1 Gag-specific antibody, and the cleavage product, capsid p24, was detected in SPBN-Pr160 and 89.6P-RVG-Pr160, indicating that functional protease is being expressed in both viruses, but not the control viruses SPBN-333 and SPBN-333-Gag. (B) A band at the expected size of gp120 was detected in 89.6P-RVG-Pr160-infected cell lysates, but not in the control lysates.
As described above, the HIV-1 RT assay allows for the detection of functional HIV-1 RT and the quantification of the expressed amount of protein. One concern with 89.6P-RVG-Pr160 was that the introduction of Env upstream of Pr160 would reduce the expression of Pr160 located further downstream. It has been shown for VSV that transcription is attenuated by about 20 to 30% at each gene junction (24). Therefore, BSR cells were infected at an MOI of 5 with 89.6P-RVG-Pr160 or SPBN-Pr160, and 24, 48, and 72 h later, we determined the amounts of RT expressed. The results show that both viruses expressed similar amounts of 4 to 6 ng of RT 48 to 72 h postinfection, which indicates that the likely attenuation at the Pr160 mRNA level did not seem to result in lower protein production of active RT (Fig. 7). The functional expression of HIV-1 89.6P Env was confirmed by a fusion assay on Sup-T1 cells (Fig. 8).
Recombinant RV expressing HIV-1 Env, Gag, and Pol produces functional reverse transcriptase (RT). BSR cells were infected at an MOI of 5 with SPBN-333-Gag, SPBN-Pr160, or 89.6P-RVG-Pr160, and samples were analyzed for RT activity 24, 48, and 72 h after infection. Comparable levels of RT activity were detected in SPBN-Pr160- and 89.6P-RVG-Pr60-infected cells 24, 48, and 72 h after infection. This indicates that the insertion of an additional foreign gene (SHIV 89.6P-RVG) upstream of Gag-Pol does not reduce the RT activity from BSR-infected cells. Control infection with SPBN-333-Gag showed only background levels of RT activity.
Expression of functional HIV-1 envelope from recombinant RV expressing HIV-1 Env and Gag-Pol. Sup-T1 cells, a human T-cell line, were infected with SPBN-Pr160 or 89.6P-RVG-Pr160 at an MOI of 0.25 for 48 h and observed for formation of syncytia. Large syncytia (B, solid arrows) were observed in cells infected with 89.6P-RVG-Pr160 (B), but not in those infected with SPBN-Pr160 (A), indicating functional HIV-1 envelope is being expressed by the recombinant RV expressing HIV-1 Env, Gag, and Pol.
In the next step, we analyzed the growth of 89.6P-RVG-Pr160 by both multicycle one-step growth curves. As mentioned previously, all Env-containing rhabdoviruses grow to a lower titer than rhabdoviruses containing other foreign genes. This finding is also confirmed by the data presented in Fig. 9A. However, at least part of the slower spread observed in the multicycle growth curve is probably due to the location of Env within the RV genome, and similar findings were made for the recombinant RV expressing Gag from a location between the N and P proteins (31). However, the observed titer of 89.6-RVG-Pr160 is sufficiently high for use in vaccine approaches (Fig. 9B). Of note, our previous research results with highly attenuated RVs expressing HIV-1 Gag showed that neither a reduced titer nor a slower spread of these vaccine vehicles interferes with their immunogenicity (31).
Multicycle-replication and one-step-growth curve of 89.6P-RVG-Pr160. BSR cells were infected with SPBN-Pr160 or 89.6P-RVG-Pr160 at an MOI of 0.01 (multicycle growth [A]) or 5 (one-step growth curve [B]). Aliquots of tissue culture supernatants were collected, and viral titers were determined in duplicate as indicated.
Finally, we wanted to analyze whether the expression of the large Gag-Pol precurser protein in the case of SPBN-pr160 or the expression of two large proteins from the 89.6P-RVG-Pr160 vaccine vector interferes with the immunogenicity of RV-based vector in a small animal model. For this approach, BALB/c mice were vaccinated once intramuscularly with 106 FFU of SPBN-Pr160 or 89.6P-RVG-Pr160. Six weeks postimmunization, mice were challenged intraperitoneally with 107 PFU of vaccinia virus Gag. Five days after challenge, two mice in each group were sacrificed, and spleens were removed and stained with the Kd-p24 tetramer specific for an HIV-1 Gag p24 epitope. As shown in Fig. 10, about 12 to 16% of the CD8+ cells were tetramer positive after challenge with vaccinia virus expressing Gag, and similar numbers were previously detected for mice immunized by the same route with the RV vector expressing only HIV-1 Gag, indicating that the expression of multiple and large genes does not result in an attenuation of the cellular immune response.
Splenocytes stained with the Kd-AMQMLKETI major histocompatibility complex-peptide tetrameric complex. Mice were immunized with 106 FFU of SPBN-Pr160 or 89.6P-RVG-Pr160, and 6 weeks postimmunization, mice were challenged with recombinant vaccinia virus expressing HIV-1 Gag. Five days later, splenocytes were isolated and cells were stained with FITC-conjugated rat anti-mouse CD8 antibody and the Kd-AMQMLKETI MHC-peptide tetrameric complex.
DISCUSSION
Here we describe the generation and characterization of two RV-based vectors that express the HIV-1 Gag-Pol precursor protein alone or in addition to a chimeric HIV-189.6P/RV G protein. These recombinant viruses express functional HIV-1 Gag, protease, and RT in addition to HIV-1 Env. The expression of the Gag-Pol protein resulted in the formation and release of both immature and mature HIV-1 VLPs. The functional expression of the major structural proteins of the HIV-1 genome makes these recombinant RVs attractive vaccine candidates, because they closely mimic important events in the HIV-1 life cycle, such as release of VLPs and fusion of cells.
Viral vectors are attractive candidates for vaccine vectors and are widely used as vaccine vehicles against HIV-1 (for review, see references 30 and 46). However, cloning limitations and stability are obstacles for the use of a large number of these vectors (for review, see reference 46). So far, DNA-based viruses, such as poxviruses, are able to express the largest foreign genes, and a modified vaccinia virus Ankara (MVA) expressing HIV-1 Gag, Pol, and Env has been used to protect against an AIDS-like disease in rhesus macaques after DNA priming, but the same approach failed when only HIV-1 Gag-Pol antigen was used (3). These results indicate the importance of a HIV-1 Env-specific immune response in this approach. The failure of the MVA expressing only Gag-Pol to protect monkeys might be due to the expression of a large number of vector-encoded proteins, some of which are immunosuppressive (22). However, several approaches were able to induce potent virus-specific CTL responses in the rhesus macaque model and protected control of a highly pathogenic SHIV-189.9P challenge (4, 5, 8, 42, 49, 50). More recently it has been shown that some of these vaccine approaches, which control the virus rather than prevent infection, are failing due to immune escape of the challenge virus (6; for review, see reference 7). These escapes may be prevented or at least delayed by using a broader HIV-1-specific CTL response. Of note, the “Holy Grail” would be the induction of broadly reactive neutralizing antibodies.
As indicated above, viruses with a large cloning capacity are rare, but two rhabdovirus-based vector are promising candidates (33). A recombinant VSV expressing both HIV-1 Env and Gag has been described, but it has not yet been used in an immunization approach (19). However, the immunogenicity of VSV-based vectors expressing HIV-1 Gag or Env has been shown in both mice and rhesus macaques (20, 21, 42; N. F. Rose, P. A. Marx, A. Luckay, D. F. Nixon, W. J. Moretto, S. M. Donahoe, and J. K. Rose, 8th Conf. Retroviruses Opportunistic Infections, abstr. 23, 2001). An RV vaccine strain-based vector expressing HIV-1 Env induces strong humoral Env-specific immune responses after a protein boost (47), and vigorous CTL responses have been detected in mice immunized with RVs expressing HIV-1 Env or Gag (29, 32). More recently, highly attenuated RVs, which were safe even after intracranial inoculation into mice, have been described (31). As mentioned above, a CTL response directed against HIV-1 p55 does not seem to be sufficient to protect long-term infection from progression to a AIDS-like disease in monkeys (6, 7). In contrast to HIV-1 gp160, the internal HIV-1 proteins are more conserved, and a large body of evidence suggests the importance of a CTL response against the other HIV-1 proteins coded for by the genes coding for Pr160 precursor proteins such as protease (26, 44), integrase (15), and RT (10, 43).
Surprisingly, the incorporation of more than 4 kb into SPBN-Pr160 did not interfere with the growth of the recombinant RV, whereas the titers of an HIV-1 Env-expressing RV are reduced about 10-fold (47). These results are similar to the ones seen for other RVs and VSV, and recombinant VSV showed titers reduced 3-fold (19) or 10-fold (25) after incorporation of HIV-189.6 Env. These are important findings, because the use of two viruses with different replication rates could interfere with the immunogenicity of the slower one. Use of a single vaccine vehicle expressing all immunogens from a single genome can circumvent this problem. 89.6P-RVG-Pr160 still grew to a high final titer of about 108, but the multicycle growth curve shows an additive effect of the expression of Env in addition to slower growth due to the expression of a foreign gene between the N and P genes from the vector BNSP (31). Of note, our previous research results with highly attenuated RVs expressing HIV-1 Gag clearly indicate that neither a reduced titer nor a slower spread of these vaccine vehicles interferes with their immunogenicity, but these factors further reduce their pathogenicity (31). However, side-by-side immunogenicity studies will be required to analyze whether there are differences in the immune response generated by a single vector expressing multiple genes from a single cell or a cocktail of vectors expressing single genes from different cells.
Finally, we utilized a new recovery system for the generation of the l8.5-kb genome of 89.6P-RVG-Pr160, a virus that we failed to obtain with our standard recovery system (31). Earlier work indicated that the three extra cytidines introduced at the 5′ end of the RV antigenome, which are needed for efficient transcription from the T7 RNA polymerase promoter, interfere with the efficient recovery of recombinant RVs (12, 48). Our new system is based on the findings of Inoue et al. that the creation of infectious RV is more efficient when a CMV promoter and two ribozymes flanking the antigenome are used (23, 28). The data presented here showed that the improvement in the recovery frequency is probably due to the generation of the precise 5′ end of the antigenome. This conclusion is based on our finding that the recovery frequencies greatly increased on BSR-T7 cells by switching to a new RV vector containing a CMV, T7 promoter, and the HH ribozyme at the 5′. Of note, the introduction of a CMV promoter now allows us to pursue a T7 RNA polymerase-free system and probably use cell lines approved for human vaccine production, as shown by Inoue et al. (23).
Thus, our results demonstrate that very large RV-based vaccine vectors containing multiple foreign genes of more than 50% of the RV genome size can be utilized and recovered. These immunogens were fusogenic and were able to release immature and mature HIV-1 VLPs. Both features might be helpful for an HIV-1 vaccine because they (i) probably expose HIV-1 Env epitopes only available during the fusion process; (ii) produce highly stable and immunogenic HIV-1 VLPs; and (iii) provide expression of CTL epitopes from additional foreign proteins, such as protease, integrase, and RT. Further research will analyze whether these new Gag-Pol or Gag-Pol-Env immunogens are superior to previously described RV vectors expressing HIV-1 Gag or Env, which are currently under investigation in small and large animal models. In addition, the recent effort to vaccinate or revaccinate large portions of the population against smallpox might interfere with the usefulness of vaccinia virus and its modified version as vaccine vectors.
ACKNOWLEDGMENTS
FACS analysis was kindly provided by Paul L. Hallberg from the Kimmel Cancer Center CORE Flow Cytometry Facility at Thomas Jefferson University. Human monoclonal antibody directed against p24 (30), rabbit antibody directed against HIV-1 RT, sheep antibody directed against HIV-1 gp120, the plasmid pNL4-3 encoding an infectious clone of HIV-1NL4-3 (46), and the plasmid encoding the 3′ region of HIV-189.6P were obtained through the AIDS Research and Reference Reagent Program (ARRRP), Division of AIDS, NIAID, NIH.
This study was supported by NIH grant AI49153 to M.J.S.
FOOTNOTES
- Received 12 May 2003.
- Accepted 15 July 2003.
- Copyright © 2003 American Society for Microbiology
