- Email: [email protected]
Tagging the Human Immunodeficiency Virus Gag Protein with Green Fluorescent Protein
Virology 255, 20–25 (1999) Article ID viro.1998.9573, available online at http://www.idealibrary.com on
Tagging the Human Immunodeficiency Virus Gag Protein with Green Fluorescent Protein Minimal Evidence for Colocalisation with Actin Claire Perrin-Tricaud,* Jean Davoust,† and Ian M. Jones*,1 *NERC Institute of Virology, Mansfield Road, Oxford OX1 3SR, UK; and †Centre d’Immunologie INSERM/CNRS de Marseille Luminy, Luminy Case 906, 13 288 Marseille Cedex 9, France Received August 26, 1998; returned to author for revision November 4, 1998; accepted December 11, 1998 The assembly and budding of human immunodeficiency virus type 1, encoded solely in the Gag protein precursor Pr55Gag, occur at the plasma membrane of infected cells. However, little is known about the routing of the Gag molecule from its site of synthesis in the cytoplasm to the site of budding, with past studies suggesting that the cytoskeleton, particularly actin, may be involved in the translocation. We have constructed a T7 promoter-driven gag gene fusion with green fluorescent protein (GFP) that expresses Gag-GFP in both cells and supernatant. The distribution of Gag-GFP was the same as Gag only, suggesting that cellular routing was not affected by fusion to GFP, and using colabelling techniques, Gag-GFP was shown to have no particular colocalisation with actin. After detergent extraction of expressing cells, Gag and Gag-GFP remained cell associated, whereas GFP only was wholly released. These data suggest that Gag may associate with other cytoskeletal components or, perhaps more likely, that a partial assembly to a large-molecular-weight intermediate occurs before localisation at the plasma membrane. © 1999 Academic Press
causes intracellular assembly. Thus the ability to assemble is not necessarily linked to the site of assembly. Distinct sites of assembly demand mechanisms for protein localisation to those sites, and for retroviruses with C-type morphology, the involvement of the cytoskeleton has long been invoked to explain the migration of Pr55Gag from the cytoplasm to the plasma membrane (reviewed in Krausslich and Welker, 1996). Purified HIV preparations have been reported to contain actin (Arthur et al., 1992), and the release of HIV-1 from the infected cells is reduced by treatment with the actin inhibitor cytochalasin D (Rey et al., 1996; Sasaki et al., 1995). Direct evidence has also been provided for an interaction of actin and Pr55Gag (Rey et al., 1996). However, in these studies, it is difficult to distinguish between physical entrapment of the Gag precursor by cytoskeletal components and a specific interaction. To reassess any possible Gag–actin interaction, we have constructed a fusion between Gag and green fluorescent protein (GFP) that can be detected directly in cells by fluorescent microscopy. Using this fusion protein, we examined the colocalisation of Gag and actin in situ.
INTRODUCTION Retroviruses were originally classified on the basis of their morphogenesis with a clear distinction between those members of the family that assemble at the plasma membrane and those that assemble cytoplasmically (Nermut and Hockley, 1996). Lentiviruses, such as human immunodeficiency virus (HIV), exhibit a C-type morphology in which assembly and budding co-occur at the plasma membrane. Both the assembly and budding signals are encoded within the Gag Pr55 precursor protein, the only protein required for virus assembly, and plasma membrane localisation requires myristoylation of Gag at the amino-terminus (Jones and Morikawa, 1998). The classic morphological distinction between retroviruses that assemble before or during budding has been moderated in recent years by the finding that mutations in Gag can alter the site of assembly. For example, the type D retrovirus Mason–Pfizer monkey virus (MPMV) normally assembles before plasma membrane localisation yet can be switched to plasma membrane assembly by a mutation in the matrix domain of MPMV Gag (Rhee and Hunter, 1991). Similarly, although HIV normally assembles at the plasma membrane, mutation or deletion of the matrix region (Freed et al., 1994; Lee and Linial, 1994) or prevention of myristoylation (Gheysen et al., 1989)
RESULTS AND DISCUSSION Construction of Gag-GFP fusion protein A gene fusion between the coding regions of Gag and GFP was achieved by ligation of DNA fragments generated from each gene through PCR. The gfp gene (Crameri et al., 1996) was amplified by using the forward
1 To whom reprint requests should be addressed. Fax: 44-1865281696. E-mail: [email protected].
0042-6822/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
20
TAGGING THE HIV GAG PROTEIN
21
FIG. 1. Schematic representation of the Gag-GFP constructs used in this work. The position of the T7 RNA polymerase promoter (T7prm) and of the recognition sites for the restriction enzymes EcoRI and ClaI used in the cloning strategy are indicated. Each of the subdomains of the Gag precursor is indicated. The Gag constructs encode the wild-type amino-terminus and so are myristoylated after expression in HeLa cells.
primer 59-CCGGGAATTCATGGCTAGCAAAGGAGAAGAACTTTTCAC-39 and the reverse primer 59-GGCCATCGATTTTGTAGAGCTCATCCATGCCATGTG-39 to generate a 716-bp fragment flanked by unique restriction site EcoRI at the 59 end and ClaI at 39 end. The PCR product was cloned directly into pGEM-T (Promega) downstream of the T7 RNA polymerase promoter to produce plasmid pGEM-GFP. The HIV-1 BH10gag gene was amplified using the forward primer 59-CCGGGAATTCATGGGTGCGAGAGCGTCAGTATTAAG-39 and the reverse primer 59CCGGGAATTCATTAGCCTGTCTCTCAGTACAATCTTTC-39 to generate a 1416-bp fragment flanked by unique EcoRI restriction sites (underlined) and cloned directly into the EcoRI site present in pGEM-GFP to produce pGEM-GagGFP. The gag fragment amplified encoded a Gag protein (p46) that deleted the p1 and p6 domains because this has been shown previously to express a more stable form of Gag and to retain all assembly information (Jowett et al., 1992; Zhang et al., 1996). A plasmid expressing Gag only under control of the T7 promoter was similarly constructed (Fig. 1). Plasmids carrying inserts generated by PCR were sequenced before use.
reduced compared with Gag only (Fig. 2). Both antigens banded at 45% sucrose in velocity gradient analysis of the supernatant antigen, suggesting particle formation similar to that described for the Gag-V3 loop chimeric particles of HIV-2 (Luo et al., 1992). Expression of GFP only lead to antigen in only the cellular fraction (not shown). To examine the distribution of Gag-GFP in HeLa cells after transfection, cells on coverslips were transfected with pGEM-GFP, pGEM-Gag, and pGEM-Gag-GFP as described except for the inclusion of 10 mM hydroxyurea in the media to minimise the cytopathic effects of vaccinia virus infection (Vos and Stunnenberg, 1988). Direct fluorescence was observed by confocal microscopy (Dutartre et al., 1996) at 15 h posttransfection for the GFP and Gag-GFP constructs and indirectly for Gag only after
Gag-GFP fusion protein is efficiently expressed in human cells To determine whether expression of the Gag-GFP fusion protein was comparable to that of Gag only, the expression of Gag and Gag-GFP from the T7 promoter plasmids was compared in HeLa cells after infection with vaccinia virus expressing T7 polymerase (vvT7–3 at 5–10 PFUs/cell) (Fuerst et al., 1986; Perrin et al., 1998). At 72 h posttransfection, the cells and supernatant were harvested and the presence of each antigen was determined by Western blot using Gag-specific monoclonal antibodies. Efficient expression of immunoreactive bands of the expected molecular weights was observed for each protein in both cell and supernatant fractions, although the level of Gag-GFP in the supernatant was
FIG. 2. Expression of Gag and Gag-GFP in HeLa cells after infection with vaccinia virus TF7–3 and transfection with plasmids described in Fig. 1. Cell lysates and supernatants were analysed by 10% SDS–PAGE and Western blotting using a mixture of Gag monoclonal antibodies ARP308 and ARP315 (Holmes, 1991). The positions of Gag and GagGFP are indicated on the right, and those of molecular weight (MW) markers are indicated on the left. The control lane, V, contained cells infected with vvTF7–3 alone.
22
PERRIN-TRICAUD, DAVOUST, AND JONES
FIG. 3. Distributions of GFP, Gag-GFP, and Gag in HeLa cells. HeLa cells were transfected with either pGEM-GFP (a), pGEM-Gag-GFP (b), or pGEM-Gag (c) and fixed 15 h posttransfection with 4% paraformaldehyde. For Gag staining, slides were processed with a mix of anti-Gag monoclonal antibody followed by an anti-mouse secondary antibody conjugated to Texas red.
fixation and immunostaining. GFP fluorescence was seen throughout the cytoplasm and the nucleus of cells as previously reported (Ogawa et al., 1995), whereas Gag-GFP fluorescence was restricted to the cytoplasm (Figs. 3a and 3b). Indirect immunofluorescence staining of Gag after transfection of pGEM-Gag showed the same distribution as Gag-GFP (Fig. 3c). No fluorescence was observed with mock-transfected cells (not shown). From these data, we conclude that the expression and distribution of Gag-GFP are unaltered compared with that of Gag alone and that Gag-GFP represents a meaningful reporter for the localisation of Gag within the expressing cell. Cellular location of Gag-GFP and actin Cellular fractionation experiments on HIV-1-infected CEM cells or COS-7 cells expressing pr55Gag have shown a variable level of Gag antigen associated with the detergent-insoluble fraction (Rey et al., 1996). Indirect immunofluorescence, however, provided only weak direct evidence for colocalisation of actin and Gag, possibly due to the sample processing involved. To reexamine this association with less harsh procedures, HeLa cells were transfected with pGEM-GFP or pGEM-Gag-GFP as above, and the locations of Gag-GFP and GFP were assessed at 15 h posttransfection using GFP intrinsic fluorescence. Actin was visualised in the same cells using phalloidin conjugated to Texas red. Gag expression was distributed throughout the cell with no particular subcellular location, and there was no apparent association with the actin filaments, which were clearly
visible (Fig. 4, middle column). A series of sections rising through the expressing cell showed that lack of interaction was not associated with any one plane of view (Fig. 4, top to bottom), although a few colabeled filaments were seen at the apex of the cell, probably due to the constriction enforced on each antigen when present in such a small area (Fig. 4). Similar results were obtained with GFP-expressing cells (not shown), suggesting no preferential association between actin and Gag. Effects of detergent extraction on the Gag protein A lack of association between Gag and actin in situ could indicate a failure of Gag-GFP to adopt a meaningful protein conformation capable of Gag assembly despite the apparent Gag-specific cellular distribution of antigen (c.f., Figs. 2 and 3). To ensure that this was not the case, we assessed Gag-expressing cells for the ability to release fluorescence after detergent extraction. Lack of detergent-induced antigen release is a hallmark of the assembly competent precursor Gag antigen but not of Gag fragments that do not assemble (Rey et al., 1996). HeLa cells were transfected as for Fig. 4 and analysed at 15 h posttransfection, but in these experiments, cells were incubated for 4 min at room temperature in extraction buffer (1 mM CaCl2, 1 mM MgCl2, 150 mM NaCl, 20 mM HEPES, 300 mM sucrose, 0.5% Triton X-100, 1 mM PMSF) before confocal microscopy. Under these conditions, HeLa cells remained adhered to the glass coverslips and maintained excellent substructure. GFP was completely extracted by this procedure, yet
FIG. 4. Confocal colocalisation analysis of Gag-GFP and actin filaments. HeLa cells were transfected with pGEM-Gag-GFP and processed for the confocal visualisation of actin and GFP as described. Gag location is shown in the left column, and actin location is shown in the center column. Double labelling is shown in the right column (GFP-Gag in green and actin in red). A colocalisation, if it occurs, causes a bright yellow colour. Four rising sections through the same cell are shown from top to bottom.
TAGGING THE HIV GAG PROTEIN
23
24
PERRIN-TRICAUD, DAVOUST, AND JONES
FIG. 5. Distribution of Gag, Gag-GFP, and GFP only in HeLa cells after extraction with Triton X-100. HeLa cells were transfected with pGEM-Gag-GFP (a), pGEM-Gag (b), or pGEM-GFP (c) and visualised at 15 h posttransfection.
Gag and Gag-GFP remained associated with the cell (Fig. 5). Much of the evidence concerning the role of actin in HIV-1 morphogenesis is circumstantial and has relied on cofractionation of variable quantities of cytoskeletal and Gag proteins (Krausslich and Welker, 1996). Even where purified components have been used to assess an interaction, the stoichiometry of Gag binding to actin has been variable (Rey et al., 1996). This argues against a sequence-specific interaction but in favour of an indirect binding dependent on a property of Gag that itself varies. The pathway of HIV-1 Gag assembly from the ribosome to the complete particle is poorly understood but is thought to consist of two broad stages: aggregation of Gag antigen at the plasma membrane, followed by ordered oligomerisation to form the virion shell. However, the in vitro assembly of Gag antigens indicates that there is no formal need to associate with the plasma membrane to assemble (Gross et al., 1997) and that limited assembly may occur before membrane localisation. Indeed, when mixtures of plasma membrane targeted and nontargeted HIV-1 Gag molecules are produced, by coexpression or by treatment with an antimyristoylation drug, both molecules are found in the released particle (Morikawa et al., 1996; Smith et al., 1993; Wang et al., 1994), suggesting limited assembly occurs before membrane localisation. Direct observation of multimeric forms of Gag in solution has also been reported recently (Morikawa et al., 1998). In this work, we confirmed that Gag as a Gag-GFP fusion protein is present in a detergent-nonextractable form after expression in HeLa cells, but we were unable to show a convincing association with actin in situ despite direct observation using confocal microscopy through many planes of the expressing cell. We suggest that this apparent discrepancy may be explained by partial Gag assembly to large-molecular-weight oligomers as part of the normal assembly pathway before membrane localisation. Such oligomers may become
physically entrapped by actin microfilaments after in vivo extraction, giving rise to the various reports of an interaction. The characterisation of such intermediates may help to identify the individual stages of the assembly process. ACKNOWLEDGMENTS We thank Quentin Sattentau and Nicolas Barois for their help and expert assistance. C.P.-T. was the recipient of an INSERM/MRC exchange fellowship.
REFERENCES Arthur, L. O., Bess, J. W., Sowder, R. C., Benveniste, R. E., Mann, D. L., Chermann, J.-C., and Henderson, L. E. (1992). Cellular proteins bound to immunodeficiency viruses: Implications for pathogenesis and vaccines. Science 258, 1935–1938. Crameri, A., Whitehorn, E. A., Tate, E., and Stemmer, W. P. C. (1996). Improved green fluorescent protein by molecular evolution using DNA shuffling. Nature Bio/Technology 14, 315–319. Dutartre, H., Davoust, J., Gorvel, J. P., and Chavrier, P. (1996). Cytokinesis arrest and redistribution of actin-cytoskeleton regulatory components in cells expressing the Rho GTPase CDC42Hs. J. Cell Sci. 109, 367–377. Freed, E. O., Orenstein, J. M., Buckler-White, A. J., and Martin, M. A. (1994). Single amino acid changes in the human immunodeficiency virus type 1 matrix protein block virus particle production. J. Virol. 68, 5311–5320. Fuerst, T. R., Niles, E. G., Studier, F. W., and Moss, B. (1986). Eukaryotic transient-expression system based on recombinant vaccinia virus that synthesizes bacteriophage T7 RNA polymerase. Proc. Natl. Acad. Sci. USA 83, 8122–8126. Gheysen, D., Jacobs, E., de Foresta, F., Thiriart, C., Francotte, M., Thines, D., and De Wilde, M. (1989). Assembly and release of HIV-1 precursor Pr55gag virus-like particles from recombinant baculovirusinfected insect cells. Cell 59, 103–112. Gross, I., Hohenberg, H., and Krausslich, H. G. (1997). In vitro assembly properties of purified bacterially expressed capsid proteins of human immunodeficiency virus. Eur. J. Biochem. 249, 592–600. Holmes, H. C. (1991). “AIDS Reagent Project: Catalogue of Reagents.” Medical Research Council AIDS reagent project, National Institute for Biological Standards and Control, Potters Bar, UK. Jones, I. M., and Morikawa, Y. (1998). The molecular basis of HIV capsid assembly. Rev. Med. Virol. 8, 87–95.
TAGGING THE HIV GAG PROTEIN Jowett, J. B., Hockley, D. J., Nermut, M. V., and Jones, I. M. (1992). Distinct signals in human immunodeficiency virus type 1 Pr55 necessary for RNA binding and particle. J. Gen. Virol. 73, 3079–3086. Krausslich, H. G., and Welker, R. (1996). Intracellular transport of retroviral capsid components. Curr. Top. Microbiol. Immunol. 214, 25–63. Lee, P. P., and Linial, M. L. (1994). Efficient particle formation can occur if the matrix domain of human immunodeficiency virus type 1 gag is substituted by a myristyolation signal. J. Virol. 68, 6644–6654. Luo, L., Li, Y., Cannon, P. M., Kim, S., and Kang, C. Y. (1992). Chimeric gag-V3 virus-like particles of human immunodeficiency virus induce virus-neutralizing antibodies. Proc. Natl. Acad. Sci. USA 89, 10527– 10531. Morikawa, Y., Hinata, S., Tomoda, H., Goto, T., Nakai, M., Aizawa, C., Tanaka, H., and Omura, S. (1996). Complete inhibition of human immunodeficiency virus gag myristoylation is necessary for inhibition of particle budding. J. Biol. Chem. 271, 2868–2873. Morikawa, Y., Zhang, W.-H., Hockley, D. J., Nermut, M. V., and Jones, I. M. (1998). Detection of a trimeric HIV-1 Gag intermediate is dependent on sequences in the matrix protein p17. J. Virol. 72, 7659–7663 Nermut, M. V., and Hockley, D. J. (1996). Comparative morphology and structural classification of retroviruses. Curr. Top. Microbiol. Immunol. 214, 1–24. Ogawa, H., Inouye, S., Tsuji, F. I., Yasuda, K., and Umesono, K. (1995). Localization, trafficking, and temperature-dependence of the Aequorea® green fluorescent protein in cultured vertebrate cells. Proc. Natl. Acad. Sci. USA 92, 11899–11903.
25
Perrin, C., Fenouillet, E., and Jones, I. (1998). Role of gp41 glycosylation sites in the biological activity of human immunodeficiency virus type 1 envelope glycoprotein. Virology 242, 338–345. Rey, O., Canon, J., and Krogstad, P. (1996). HIV-1 Gag protein associates with F-actin present in microfilaments. Virology 220, 530–534. Rhee, S. S., and Hunter, E. (1991). Amino acid substitutions within the matrix protein of type D retroviruses affect assembly, transport and membrane association of a capsid. EMBO J. 10, 535–546. Sasaki, H., Nakamura, M., Ohno, T., Matsuda, Y., Yuda, Y., and Nonomura, Y. (1995). Myosin-actin interaction plays an important role in human immunodeficiency virus type 1 release from host cells. Proc. Natl. Acad. Sci. USA 92, 2026–2030. Smith, A. J., Srinivasakumar, N., Hammmarskjold, M.-L., and Rekosh, D. (1993). Requirements for incorporation of Pr160gag-pol from human immunodeficiency virus type 1 into virus-like particles. J. Virol. 67, 2266–2275. Vos, J. C., and Stunnenberg, H. G. (1988). Derepression of a novel class of vaccinia virus genes upon DNA replication. EMBO J. 7, 3487–3492. Wang, C. T., Stegeman Olsen, J., Zhang, Y., and Barklis, E. (1994). Assembly of HIV Gag-b-galactosidase fusion proteins into virus particles. Virology 200, 524–534. Zhang, W. H., Hockley, D. J., Nermut, M. V., Morikawa, Y., and Jones, I. M. (1996). Gag-Gag interactions in the C-terminal domain of human immunodeficiency virus type 1 p24 capsid antigen are essential for Gag particle assembly. J. Gen. Virol. 77, 743–751.
Tagging the Human Immunodeficiency Virus Gag Protein with Green Fluorescent Protein Minimal Evidence for Colocalisation with Actin Claire Perrin-Tricaud,* Jean Davoust,† and Ian M. Jones*,1 *NERC Institute of Virology, Mansfield Road, Oxford OX1 3SR, UK; and †Centre d’Immunologie INSERM/CNRS de Marseille Luminy, Luminy Case 906, 13 288 Marseille Cedex 9, France Received August 26, 1998; returned to author for revision November 4, 1998; accepted December 11, 1998 The assembly and budding of human immunodeficiency virus type 1, encoded solely in the Gag protein precursor Pr55Gag, occur at the plasma membrane of infected cells. However, little is known about the routing of the Gag molecule from its site of synthesis in the cytoplasm to the site of budding, with past studies suggesting that the cytoskeleton, particularly actin, may be involved in the translocation. We have constructed a T7 promoter-driven gag gene fusion with green fluorescent protein (GFP) that expresses Gag-GFP in both cells and supernatant. The distribution of Gag-GFP was the same as Gag only, suggesting that cellular routing was not affected by fusion to GFP, and using colabelling techniques, Gag-GFP was shown to have no particular colocalisation with actin. After detergent extraction of expressing cells, Gag and Gag-GFP remained cell associated, whereas GFP only was wholly released. These data suggest that Gag may associate with other cytoskeletal components or, perhaps more likely, that a partial assembly to a large-molecular-weight intermediate occurs before localisation at the plasma membrane. © 1999 Academic Press
causes intracellular assembly. Thus the ability to assemble is not necessarily linked to the site of assembly. Distinct sites of assembly demand mechanisms for protein localisation to those sites, and for retroviruses with C-type morphology, the involvement of the cytoskeleton has long been invoked to explain the migration of Pr55Gag from the cytoplasm to the plasma membrane (reviewed in Krausslich and Welker, 1996). Purified HIV preparations have been reported to contain actin (Arthur et al., 1992), and the release of HIV-1 from the infected cells is reduced by treatment with the actin inhibitor cytochalasin D (Rey et al., 1996; Sasaki et al., 1995). Direct evidence has also been provided for an interaction of actin and Pr55Gag (Rey et al., 1996). However, in these studies, it is difficult to distinguish between physical entrapment of the Gag precursor by cytoskeletal components and a specific interaction. To reassess any possible Gag–actin interaction, we have constructed a fusion between Gag and green fluorescent protein (GFP) that can be detected directly in cells by fluorescent microscopy. Using this fusion protein, we examined the colocalisation of Gag and actin in situ.
INTRODUCTION Retroviruses were originally classified on the basis of their morphogenesis with a clear distinction between those members of the family that assemble at the plasma membrane and those that assemble cytoplasmically (Nermut and Hockley, 1996). Lentiviruses, such as human immunodeficiency virus (HIV), exhibit a C-type morphology in which assembly and budding co-occur at the plasma membrane. Both the assembly and budding signals are encoded within the Gag Pr55 precursor protein, the only protein required for virus assembly, and plasma membrane localisation requires myristoylation of Gag at the amino-terminus (Jones and Morikawa, 1998). The classic morphological distinction between retroviruses that assemble before or during budding has been moderated in recent years by the finding that mutations in Gag can alter the site of assembly. For example, the type D retrovirus Mason–Pfizer monkey virus (MPMV) normally assembles before plasma membrane localisation yet can be switched to plasma membrane assembly by a mutation in the matrix domain of MPMV Gag (Rhee and Hunter, 1991). Similarly, although HIV normally assembles at the plasma membrane, mutation or deletion of the matrix region (Freed et al., 1994; Lee and Linial, 1994) or prevention of myristoylation (Gheysen et al., 1989)
RESULTS AND DISCUSSION Construction of Gag-GFP fusion protein A gene fusion between the coding regions of Gag and GFP was achieved by ligation of DNA fragments generated from each gene through PCR. The gfp gene (Crameri et al., 1996) was amplified by using the forward
1 To whom reprint requests should be addressed. Fax: 44-1865281696. E-mail: [email protected].
0042-6822/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
20
TAGGING THE HIV GAG PROTEIN
21
FIG. 1. Schematic representation of the Gag-GFP constructs used in this work. The position of the T7 RNA polymerase promoter (T7prm) and of the recognition sites for the restriction enzymes EcoRI and ClaI used in the cloning strategy are indicated. Each of the subdomains of the Gag precursor is indicated. The Gag constructs encode the wild-type amino-terminus and so are myristoylated after expression in HeLa cells.
primer 59-CCGGGAATTCATGGCTAGCAAAGGAGAAGAACTTTTCAC-39 and the reverse primer 59-GGCCATCGATTTTGTAGAGCTCATCCATGCCATGTG-39 to generate a 716-bp fragment flanked by unique restriction site EcoRI at the 59 end and ClaI at 39 end. The PCR product was cloned directly into pGEM-T (Promega) downstream of the T7 RNA polymerase promoter to produce plasmid pGEM-GFP. The HIV-1 BH10gag gene was amplified using the forward primer 59-CCGGGAATTCATGGGTGCGAGAGCGTCAGTATTAAG-39 and the reverse primer 59CCGGGAATTCATTAGCCTGTCTCTCAGTACAATCTTTC-39 to generate a 1416-bp fragment flanked by unique EcoRI restriction sites (underlined) and cloned directly into the EcoRI site present in pGEM-GFP to produce pGEM-GagGFP. The gag fragment amplified encoded a Gag protein (p46) that deleted the p1 and p6 domains because this has been shown previously to express a more stable form of Gag and to retain all assembly information (Jowett et al., 1992; Zhang et al., 1996). A plasmid expressing Gag only under control of the T7 promoter was similarly constructed (Fig. 1). Plasmids carrying inserts generated by PCR were sequenced before use.
reduced compared with Gag only (Fig. 2). Both antigens banded at 45% sucrose in velocity gradient analysis of the supernatant antigen, suggesting particle formation similar to that described for the Gag-V3 loop chimeric particles of HIV-2 (Luo et al., 1992). Expression of GFP only lead to antigen in only the cellular fraction (not shown). To examine the distribution of Gag-GFP in HeLa cells after transfection, cells on coverslips were transfected with pGEM-GFP, pGEM-Gag, and pGEM-Gag-GFP as described except for the inclusion of 10 mM hydroxyurea in the media to minimise the cytopathic effects of vaccinia virus infection (Vos and Stunnenberg, 1988). Direct fluorescence was observed by confocal microscopy (Dutartre et al., 1996) at 15 h posttransfection for the GFP and Gag-GFP constructs and indirectly for Gag only after
Gag-GFP fusion protein is efficiently expressed in human cells To determine whether expression of the Gag-GFP fusion protein was comparable to that of Gag only, the expression of Gag and Gag-GFP from the T7 promoter plasmids was compared in HeLa cells after infection with vaccinia virus expressing T7 polymerase (vvT7–3 at 5–10 PFUs/cell) (Fuerst et al., 1986; Perrin et al., 1998). At 72 h posttransfection, the cells and supernatant were harvested and the presence of each antigen was determined by Western blot using Gag-specific monoclonal antibodies. Efficient expression of immunoreactive bands of the expected molecular weights was observed for each protein in both cell and supernatant fractions, although the level of Gag-GFP in the supernatant was
FIG. 2. Expression of Gag and Gag-GFP in HeLa cells after infection with vaccinia virus TF7–3 and transfection with plasmids described in Fig. 1. Cell lysates and supernatants were analysed by 10% SDS–PAGE and Western blotting using a mixture of Gag monoclonal antibodies ARP308 and ARP315 (Holmes, 1991). The positions of Gag and GagGFP are indicated on the right, and those of molecular weight (MW) markers are indicated on the left. The control lane, V, contained cells infected with vvTF7–3 alone.
22
PERRIN-TRICAUD, DAVOUST, AND JONES
FIG. 3. Distributions of GFP, Gag-GFP, and Gag in HeLa cells. HeLa cells were transfected with either pGEM-GFP (a), pGEM-Gag-GFP (b), or pGEM-Gag (c) and fixed 15 h posttransfection with 4% paraformaldehyde. For Gag staining, slides were processed with a mix of anti-Gag monoclonal antibody followed by an anti-mouse secondary antibody conjugated to Texas red.
fixation and immunostaining. GFP fluorescence was seen throughout the cytoplasm and the nucleus of cells as previously reported (Ogawa et al., 1995), whereas Gag-GFP fluorescence was restricted to the cytoplasm (Figs. 3a and 3b). Indirect immunofluorescence staining of Gag after transfection of pGEM-Gag showed the same distribution as Gag-GFP (Fig. 3c). No fluorescence was observed with mock-transfected cells (not shown). From these data, we conclude that the expression and distribution of Gag-GFP are unaltered compared with that of Gag alone and that Gag-GFP represents a meaningful reporter for the localisation of Gag within the expressing cell. Cellular location of Gag-GFP and actin Cellular fractionation experiments on HIV-1-infected CEM cells or COS-7 cells expressing pr55Gag have shown a variable level of Gag antigen associated with the detergent-insoluble fraction (Rey et al., 1996). Indirect immunofluorescence, however, provided only weak direct evidence for colocalisation of actin and Gag, possibly due to the sample processing involved. To reexamine this association with less harsh procedures, HeLa cells were transfected with pGEM-GFP or pGEM-Gag-GFP as above, and the locations of Gag-GFP and GFP were assessed at 15 h posttransfection using GFP intrinsic fluorescence. Actin was visualised in the same cells using phalloidin conjugated to Texas red. Gag expression was distributed throughout the cell with no particular subcellular location, and there was no apparent association with the actin filaments, which were clearly
visible (Fig. 4, middle column). A series of sections rising through the expressing cell showed that lack of interaction was not associated with any one plane of view (Fig. 4, top to bottom), although a few colabeled filaments were seen at the apex of the cell, probably due to the constriction enforced on each antigen when present in such a small area (Fig. 4). Similar results were obtained with GFP-expressing cells (not shown), suggesting no preferential association between actin and Gag. Effects of detergent extraction on the Gag protein A lack of association between Gag and actin in situ could indicate a failure of Gag-GFP to adopt a meaningful protein conformation capable of Gag assembly despite the apparent Gag-specific cellular distribution of antigen (c.f., Figs. 2 and 3). To ensure that this was not the case, we assessed Gag-expressing cells for the ability to release fluorescence after detergent extraction. Lack of detergent-induced antigen release is a hallmark of the assembly competent precursor Gag antigen but not of Gag fragments that do not assemble (Rey et al., 1996). HeLa cells were transfected as for Fig. 4 and analysed at 15 h posttransfection, but in these experiments, cells were incubated for 4 min at room temperature in extraction buffer (1 mM CaCl2, 1 mM MgCl2, 150 mM NaCl, 20 mM HEPES, 300 mM sucrose, 0.5% Triton X-100, 1 mM PMSF) before confocal microscopy. Under these conditions, HeLa cells remained adhered to the glass coverslips and maintained excellent substructure. GFP was completely extracted by this procedure, yet
FIG. 4. Confocal colocalisation analysis of Gag-GFP and actin filaments. HeLa cells were transfected with pGEM-Gag-GFP and processed for the confocal visualisation of actin and GFP as described. Gag location is shown in the left column, and actin location is shown in the center column. Double labelling is shown in the right column (GFP-Gag in green and actin in red). A colocalisation, if it occurs, causes a bright yellow colour. Four rising sections through the same cell are shown from top to bottom.
TAGGING THE HIV GAG PROTEIN
23
24
PERRIN-TRICAUD, DAVOUST, AND JONES
FIG. 5. Distribution of Gag, Gag-GFP, and GFP only in HeLa cells after extraction with Triton X-100. HeLa cells were transfected with pGEM-Gag-GFP (a), pGEM-Gag (b), or pGEM-GFP (c) and visualised at 15 h posttransfection.
Gag and Gag-GFP remained associated with the cell (Fig. 5). Much of the evidence concerning the role of actin in HIV-1 morphogenesis is circumstantial and has relied on cofractionation of variable quantities of cytoskeletal and Gag proteins (Krausslich and Welker, 1996). Even where purified components have been used to assess an interaction, the stoichiometry of Gag binding to actin has been variable (Rey et al., 1996). This argues against a sequence-specific interaction but in favour of an indirect binding dependent on a property of Gag that itself varies. The pathway of HIV-1 Gag assembly from the ribosome to the complete particle is poorly understood but is thought to consist of two broad stages: aggregation of Gag antigen at the plasma membrane, followed by ordered oligomerisation to form the virion shell. However, the in vitro assembly of Gag antigens indicates that there is no formal need to associate with the plasma membrane to assemble (Gross et al., 1997) and that limited assembly may occur before membrane localisation. Indeed, when mixtures of plasma membrane targeted and nontargeted HIV-1 Gag molecules are produced, by coexpression or by treatment with an antimyristoylation drug, both molecules are found in the released particle (Morikawa et al., 1996; Smith et al., 1993; Wang et al., 1994), suggesting limited assembly occurs before membrane localisation. Direct observation of multimeric forms of Gag in solution has also been reported recently (Morikawa et al., 1998). In this work, we confirmed that Gag as a Gag-GFP fusion protein is present in a detergent-nonextractable form after expression in HeLa cells, but we were unable to show a convincing association with actin in situ despite direct observation using confocal microscopy through many planes of the expressing cell. We suggest that this apparent discrepancy may be explained by partial Gag assembly to large-molecular-weight oligomers as part of the normal assembly pathway before membrane localisation. Such oligomers may become
physically entrapped by actin microfilaments after in vivo extraction, giving rise to the various reports of an interaction. The characterisation of such intermediates may help to identify the individual stages of the assembly process. ACKNOWLEDGMENTS We thank Quentin Sattentau and Nicolas Barois for their help and expert assistance. C.P.-T. was the recipient of an INSERM/MRC exchange fellowship.
REFERENCES Arthur, L. O., Bess, J. W., Sowder, R. C., Benveniste, R. E., Mann, D. L., Chermann, J.-C., and Henderson, L. E. (1992). Cellular proteins bound to immunodeficiency viruses: Implications for pathogenesis and vaccines. Science 258, 1935–1938. Crameri, A., Whitehorn, E. A., Tate, E., and Stemmer, W. P. C. (1996). Improved green fluorescent protein by molecular evolution using DNA shuffling. Nature Bio/Technology 14, 315–319. Dutartre, H., Davoust, J., Gorvel, J. P., and Chavrier, P. (1996). Cytokinesis arrest and redistribution of actin-cytoskeleton regulatory components in cells expressing the Rho GTPase CDC42Hs. J. Cell Sci. 109, 367–377. Freed, E. O., Orenstein, J. M., Buckler-White, A. J., and Martin, M. A. (1994). Single amino acid changes in the human immunodeficiency virus type 1 matrix protein block virus particle production. J. Virol. 68, 5311–5320. Fuerst, T. R., Niles, E. G., Studier, F. W., and Moss, B. (1986). Eukaryotic transient-expression system based on recombinant vaccinia virus that synthesizes bacteriophage T7 RNA polymerase. Proc. Natl. Acad. Sci. USA 83, 8122–8126. Gheysen, D., Jacobs, E., de Foresta, F., Thiriart, C., Francotte, M., Thines, D., and De Wilde, M. (1989). Assembly and release of HIV-1 precursor Pr55gag virus-like particles from recombinant baculovirusinfected insect cells. Cell 59, 103–112. Gross, I., Hohenberg, H., and Krausslich, H. G. (1997). In vitro assembly properties of purified bacterially expressed capsid proteins of human immunodeficiency virus. Eur. J. Biochem. 249, 592–600. Holmes, H. C. (1991). “AIDS Reagent Project: Catalogue of Reagents.” Medical Research Council AIDS reagent project, National Institute for Biological Standards and Control, Potters Bar, UK. Jones, I. M., and Morikawa, Y. (1998). The molecular basis of HIV capsid assembly. Rev. Med. Virol. 8, 87–95.
TAGGING THE HIV GAG PROTEIN Jowett, J. B., Hockley, D. J., Nermut, M. V., and Jones, I. M. (1992). Distinct signals in human immunodeficiency virus type 1 Pr55 necessary for RNA binding and particle. J. Gen. Virol. 73, 3079–3086. Krausslich, H. G., and Welker, R. (1996). Intracellular transport of retroviral capsid components. Curr. Top. Microbiol. Immunol. 214, 25–63. Lee, P. P., and Linial, M. L. (1994). Efficient particle formation can occur if the matrix domain of human immunodeficiency virus type 1 gag is substituted by a myristyolation signal. J. Virol. 68, 6644–6654. Luo, L., Li, Y., Cannon, P. M., Kim, S., and Kang, C. Y. (1992). Chimeric gag-V3 virus-like particles of human immunodeficiency virus induce virus-neutralizing antibodies. Proc. Natl. Acad. Sci. USA 89, 10527– 10531. Morikawa, Y., Hinata, S., Tomoda, H., Goto, T., Nakai, M., Aizawa, C., Tanaka, H., and Omura, S. (1996). Complete inhibition of human immunodeficiency virus gag myristoylation is necessary for inhibition of particle budding. J. Biol. Chem. 271, 2868–2873. Morikawa, Y., Zhang, W.-H., Hockley, D. J., Nermut, M. V., and Jones, I. M. (1998). Detection of a trimeric HIV-1 Gag intermediate is dependent on sequences in the matrix protein p17. J. Virol. 72, 7659–7663 Nermut, M. V., and Hockley, D. J. (1996). Comparative morphology and structural classification of retroviruses. Curr. Top. Microbiol. Immunol. 214, 1–24. Ogawa, H., Inouye, S., Tsuji, F. I., Yasuda, K., and Umesono, K. (1995). Localization, trafficking, and temperature-dependence of the Aequorea® green fluorescent protein in cultured vertebrate cells. Proc. Natl. Acad. Sci. USA 92, 11899–11903.
25
Perrin, C., Fenouillet, E., and Jones, I. (1998). Role of gp41 glycosylation sites in the biological activity of human immunodeficiency virus type 1 envelope glycoprotein. Virology 242, 338–345. Rey, O., Canon, J., and Krogstad, P. (1996). HIV-1 Gag protein associates with F-actin present in microfilaments. Virology 220, 530–534. Rhee, S. S., and Hunter, E. (1991). Amino acid substitutions within the matrix protein of type D retroviruses affect assembly, transport and membrane association of a capsid. EMBO J. 10, 535–546. Sasaki, H., Nakamura, M., Ohno, T., Matsuda, Y., Yuda, Y., and Nonomura, Y. (1995). Myosin-actin interaction plays an important role in human immunodeficiency virus type 1 release from host cells. Proc. Natl. Acad. Sci. USA 92, 2026–2030. Smith, A. J., Srinivasakumar, N., Hammmarskjold, M.-L., and Rekosh, D. (1993). Requirements for incorporation of Pr160gag-pol from human immunodeficiency virus type 1 into virus-like particles. J. Virol. 67, 2266–2275. Vos, J. C., and Stunnenberg, H. G. (1988). Derepression of a novel class of vaccinia virus genes upon DNA replication. EMBO J. 7, 3487–3492. Wang, C. T., Stegeman Olsen, J., Zhang, Y., and Barklis, E. (1994). Assembly of HIV Gag-b-galactosidase fusion proteins into virus particles. Virology 200, 524–534. Zhang, W. H., Hockley, D. J., Nermut, M. V., Morikawa, Y., and Jones, I. M. (1996). Gag-Gag interactions in the C-terminal domain of human immunodeficiency virus type 1 p24 capsid antigen are essential for Gag particle assembly. J. Gen. Virol. 77, 743–751.