Ty3 Gag3 Protein Forms Ordered Particles in Escherichia Coli

doi:10.1016/j.virol.2007.09.017

Virology. Author manuscript; available in PMC 2009 June 24.

Published in final edited form as:

Virology. 2008 January 20; 370(2): 223–227.

Published online 2007 October 26. doi: 10.1016/j.virol.2007.09.017

PMCID: PMC2701475

NIHMSID: NIHMS36223

Ty3 Gag3 Protein Forms Ordered Particles in Escherichia Coli

Liza S.Z. Larsen,^1,^2,³ Yurii Kuznetsov,⁴ Alex McPherson,⁴ G.Wesley Hatfield,^2,^3,⁵ and Suzanne Sandmeyer^1,^2,³^*

¹Department of Biological Chemistry, University of California, Irvine, CA 92697

²School of Information and Computer Sciences and Institute for Genomics and Bioinformatics, University of California, Irvine, CA 92697

³Department of Microbiology and Molecular Genetics, University of California, Irvine, CA 92697

⁴Department of Molecular Biology and Biochemistry, University of California, Irvine, CA 92697

⁵CODA Genomics, Inc. Laguna Hills, CA 92653

^* Corresponding author: Phone: 949-824-7571, Email: sbsandme/at/uci.edu

The publisher's final edited version of this article is available at Virology

See other articles in PMC that cite the published article.

Publisher's Disclaimer

Abstract

The yeast retrovirus-like element Ty3 GAG3 gene encodes a Gag3 polyprotein analogous to retroviral Gag. Gag3 lacks matrix, but contains capsid, spacer, and nucleocapsid domains. Expression of a Ty3 Gag3 or capsid domain optimized for expression in Escherichia coli, was sufficient for Ty3 particle assembly. Virus-like ordered particles assembled from Gag3 were similar in size to immature particles from yeast and contained nucleic acid. However, particles assembled from the CA domain were variable in size and displayed much less organization than native particles. These results indicate that assembly can be driven through interactions among capsid subunits in the particle, but that the nucleocapsid domain, likely in association with RNA, confers order upon this process.

Keywords: retrotransposon, retrovirus, virus-like particles, capsid, Escherichia coli, yeast Ty3

Introduction

Gag, the major structural component of retrovirus cores, typically contains matrix (MA), capsid (CA), nucleocapsid (NC) and one or more spacer (SP) domains as well as other virus-specific domains. Within the Gag context, MA functions in membrane targeting; CA mediates Gag interaction; and NC packages RNA and promotes Gag interaction (Adamson & Jones, 2004;Scarlata & Carter, 2003). NC is a small domain rich in basic residues that typically contains one or two copies of a CX₂CX₄HX₄C zinc-binding motif. After particle assembly and budding, these domains are separated by proteolytic cleavage and characteristic cores are formed. Expression of Gag in heterologous systems, such as baculovirus and Escherichia coli has shown formation of virus-like particles (VLPs). In the case of HIV, depending on salt and pH conditions CA can form tubes, spheres, and cones in vitro. Inclusion of the NC domain and RNA enhances formation of sheet-like structures or cylindrical forms (Ehrlich, Agresta et al., 1992;Ehrlich, Liu et al., 2001;Campbell & Vogt, 1995;Ganser, Li et al., 1999;Johnson, Scobie et al., 2002;Yu, Joshi et al., 2001). Formation of spheres approximating immature particle forms occurs when regions amino-terminal to CA are included. This effect is attributed to constraint of the amino terminus of CA which rearranges during maturation and is engaged in an internal salt bridge in the mature form (von Schwedler, Stemmler et al., 1998;Gross, Hohenberg et al., 1998;Joshi & Vogt, 2000;Rumlova-Klikova, Hunter et al., 2000;Campbell & Vogt, 1997).

The budding yeast retrotransposon Ty3 has a Gag3 domain comprised of CA, SP3 and NC domains (Fig. 1) (Sandmeyer, Aye et al., 2002). Initial assembly of Ty3 results in a roughly spherical VLP which remains in a similar form even after proteolytic maturation (Kuznetsov, Zhang et al., 2005). The Ty3 precursor differs from the retroviral precursor in that it lacks the amino-terminal extension of CA into MA, as well as the conserved amino-terminal proline of CA. In this work we determined that expression of Ty3 Gag3 was sufficient to form spherical particles in a bacterial expression system. The organization of particles was greatly enhanced by inclusion of the NC domain.

Fig. 1

Ty3 organization

Results

Assembly of recoded GAG3

Because Ty3 Gag3 protein is comprised only of CA, SP, and NC domains, it was of interest to determine whether its expression was sufficient to support spherical particle formation in a heterologous system. A synthetic Ty3 GAG3 gene optimized for bacterial protein expression produced significant levels of Gag3 (Fig. 2, lanes 1-3). Little if any protein expression was observed from the native yeast Ty3 GAG3 gene (Fig. 2, lanes 4-6). About 40% of the recoded Gag3 was soluble (Fig. 3A). The nature of the insoluble protein fraction was not explored, but it could have contained Gag3 inclusion bodies or aggregates.

Fig. 2

Recoded Gag3 is expressed at a higher level in E. coli than native yeast Gag3

Fig. 3

Expression of Gag3 and CA in E. coli

Atomic force microscopy analysis of E. coli cell lysates expressing Gag3

Approximately 500 μg of cleared lysate from cells expressing recoded Gag3 was fractionated over a 20/30/70% sucrose step gradient as previously described (Kuznetsov, Zhang, Menees, McPherson, & Sandmeyer, 2005). Gradient fractions were analyzed by denaturing polyacrylamide gel electrophoresis. A protein of the same size as Gag3 that reacted with anti-CA antibody was the major component of fractions 7 and 8 (Fig. 3B); this is the position of migration of native Ty3 VLPs produced in yeast [(Kuznetsov, Zhang, Menees, McPherson, & Sandmeyer, 2005) and data not shown]. A 5 μl aliquot of the 70/30% interface fraction was used for AFM imaging as described previously. Imaging of the interface fraction showed particles, ranging from 30 to 50 nm in diameter (Fig. 4 A-C). These were of comparable size to protease mutant Ty3 VLPs produced in budding yeast, which had a mean of 49 nm +/- 2.3 nm ((Kuznetsov, Zhang, Menees, McPherson, & Sandmeyer, 2005) and data not shown). A small proportion of VLPs appeared to be joined, mostly in pairs (e.g. Fig. 4C). Similar associations were observed in AFM imaging of wt Ty3 VLPs from yeast (Kuznetsov, Zhang, Menees, McPherson, & Sandmeyer, 2005). Overall, VLPs displayed ordered capsomeres and were generally similar to those previously observed. Tapping disruption of the particles resulted in release of fibrous material from some which resembled RNA (Fig. 4 G-H), suggesting that similar to retroviruses, Ty3 Gag3 can package heterologous RNA in the absence of its native genome.

Fig. 4

AFM analysis of particles resulting from expression of Ty3 Gag3 and CA in E. coli

Expression of CA and AFM analysis of E. coli expression lysates

In order to determine the contribution of SP and NC domains to VLP structure, the CA-coding domain was amplified using recoded GAG3 sequence as the template. Expression of this domain from the IPTG-inducible T7 promoter was efficient, but failed to produce sufficient levels of soluble protein for analysis (data not shown). The construct was recloned into an arabinose-inducible expression vector (pBAD-CODA-CA) where expression was controlled to optimize solubility (data not shown). Under these optimal conditions, much less CA protein was produced than Gag3, but the proportion of soluble protein was comparable to that in the Gag3 expression system (Fig. 3A and C). Approximately 500 μg of cleared lysate from cells expressing pBAD-CODA-CA was fractionated over a 20/30/70% sucrose step gradient and 5 μl of the 30/70% interface fraction was used for AFM imaging as described above. Analysis showed partially structured particles, as well as unstructured amorphous protein aggregates (Fig. 4 D-F and data not shown). Some of these aggregates resembled double particles (Fig. 4F). A random aliquot of particulate material was measured. Particles in this fraction ranged in size from 40 to 157 nm with an overall mean of 94.2 nm.

Discussion

Expression of retroviral gag in heterologous systems is sufficient for particle formation. In the case of retroviruses in vitro assembly using CA with amino-terminal extensions into upstream domains results in spherical particles thought to mimic the immature capsid form, whereas in vitro assembly of CA-NC results in tubes and sheets. This difference is thought to result from constraint in the former situation of the conserved amino-terminal proline of CA, which when released in processing, forms a salt bridge required for the mature particle structure (von Schwedler, Stemmler, Klishko, Li, Albertine, Davis, & Sundquist, 1998;Gross, Hohenberg, Huckhagel, & Krausslich, 1998;Joshi & Vogt, 2000;Rumlova-Klikova, Hunter, Nermut, Pichova, & Ruml, 2000;Campbell & Vogt, 1997). In vitro assembly using MPMV mutant CA-NC lacking the amino-terminal proline residue and unable to form this bridge also resulted in spherical particles (Ulbrich, Haubova et al., 2006). Modeling of the Ty3 CA domain suggests that, similar to retroviral CA, Ty3 CA is composed of amino-terminal and carboxyl-terminal clusters of alpha helices (Larsen, Zhang et al., 2007). However, Ty3 lacks a CA amino-terminal extension and the conserved proline of CA found in retroviruses and is roughly spherical in both immature and mature states. Roughly spherical particles were formed in E. coli upon Ty3 Gag3 expression. This may be similar in some respects to the formation of spherical particles in vitro from MPMV CA-NC lacking the amino-terminal proline. In any case, Ty3 CA, SP and NC domains, likely together with RNA, are sufficient to specify this structure. The nature of interactions which stabilize the mature Ty3 VLPs is not known.

Studies of retroviruses have shown that NC is unnecessary for in vitro retroviral particle formation (Muriaux, Mirro et al., 2001). However, inclusion of the NC domain and RNA in assembly reactions promotes interaction of CA domains and multimer formation (e.g. RSV (Johnson, Scobie, Ma, & Vogt, 2002) reviewed in (Adamson & Jones, 2004)).

Expression of Ty3 CA in E. coli resulted in formation of highly irregular particles, although it is possible that sheets also formed and remained in the extract pellet. These results demonstrated that the SP-NC coupled to an RNA scaffold is not necessary for Ty3 Gag3 multimerization per se. However, the difference in Gag3 and CA particle morphology indicates that the SP-NC domains present in Gag3 and lacking in CA contribute to Ty3 particle organization and symmetry, potentially via an RNA scaffold. Although a role is implied for RNA by the contribution of SP-NC, this role is not likely to be sequence specific. Recoded GAG3 mRNA included none of the 5′ untranslated region previously shown to be important in cis for transposition (Kirchner, Sandmeyer et al., 1992) and by virtue of recoding, was 25% different in sequence within the translated region from the natural GAG3 sequence.

Expression of the Ty1 TYA structural protein from native RNA also resulted in particles in E.coli (Luschnig, Hess et al., 1995). Ty1 is more distantly related to Ty3 than are retroviruses and Ty1 does not contain a zinc-binding motif. However, in this case also, deletion of the nucleic acid binding domain disrupted particle formation. Thus, these yeast retrotransposons both of which assemble spherical, cytoplasmic particles that undergo maturation in the absence of a membrane envelope, showed dependence upon a nucleic acid binding domain for assembly in E. coli.

In summary, Gag3 in the absence of Ty3 native RNA, POL3, and other yeast proteins is sufficient for VLP assembly into structures resembling, but not identical to, native VLPs. CA is sufficient for multimer formation. However, this work shows that the SP-NC domain, possibly together with RNA, contributes to ordering of the particle structure.

Materials and Methods

Strains and culture conditions for E. coli

E. coli strain DH5α (Invitrogen, Carlsbad, CA) was used for DNA cloning and plasmid preparation of CODA-designed genes. E. coli strains BL21 (DE3) ((Novagen, Madison, WI), and Top10 (Invitrogen Inc., Carlsbad, CA) were used for protein expression.

Gene cloning

The Ty3 GAG3 open reading frame was amplified from the pDLC201 plasmid using the primers, 5′-CTATATCTAGCATATGAGCTTTATGGATCA-3′ and 5′-GTATTGGATCCTTATTAAGATCGGTTAGAACTCGCC-3′. The amplified product was digested with NdeI and BamHI and ligated into the complementary sites of the pET-3a expression vector (Novagen, Madison, WI) to give plasmid pLZL2423 (pET3a-yeast-native-GAG3). In order to optimize protein production, the full-length, 876-bp Ty3 GAG3 ORF was recoded to reflect E. coli codon bias and assembled in vitro according to manufacturer's instructions (CODA Genomics Inc., Laguna Hills, CA) (Hatfield GW & Gutman GA, 1992)(Larsen et al., in preparation) (For assembly strategy and intermediate and full-length products, see Supplemental Data Fig. 1; for sequence of oligonucleotides, see Supplemental Data, Table 1.) The yeast CODA Ty3 GAG3 gene was cloned into pET-3a, under an IPTG responsive promoter, to yield clone pLZL2422 (pET3a-CODA-GAG3). The portion of this GAG3 encoding CA (residues 1-207 of Gag3) was amplified using the polymerase chain reaction and primers which annealed to the GAG3 sequence in pET3a-CODA-GAG3 (5′-TACTCCTCGAGATGTCATTCATGGACCAGATTCCGGGCGG-3′ and 5′-GGCATGCAAGCTTTTATTAGCCTATGATGGTATCTCCGTCCGCGTCC-3′). The amplified fragment was restricted and cloned into the XhoI and HindIII sites of the araBAD-based promoter pBAD vector (Invitrogen Inc.) to create clone pLZL2424 (pBAD-CODA-CA). Clones were verified by DNA sequence analysis (Cogenics, Morrisville, NC).

Protein expression

The pET3a-yeast-native-GAG3 and pET3a-CODA-GAG3 were each transformed into E. coli strain BL21(DE3). For optimal protein solubility, a fresh overnight culture was diluted 1:100 into 200 ml LB/amp medium and was grown at 37°C to OD₆₀₀=1.0. The T7 RNA polymerase gene in BL21(DE3) under the control of the lac promoter exhibits leaky expression sufficient to achieve soluble Gag3 by growth at 4°C for 72 h without the use of IPTG. Pelleted cells were resuspended in 20 ml lysis Buffer A (20 mM TrisCl pH 7.5, 20 mM NaCl, 10% glycerol, 5mM DTT, 1mM PMSF, 1mM EDTA) and disrupted by three serial passages through a French pressure cell at 15,000 lb/in² at 4°C. These extracts were clarified by centrifugation at 17K × g for 15 min at 4°C. Pellets were resuspended in 20 ml of buffer A. Laemmli sample buffer (LSB) was added to 10 μl of either resuspended pellet or supernatant to a concentration of 1 × and proteins were analyzed by SDS-10% polyacrylamide gel electrophoresis.

The pBAD-CODA-CA was transformed into E. coli strain Top10 (Invitrogen, Carlsbad, CA). A single E. coli Top10- pBAD-CODA-CA colony was inoculated into 200 ml of LB medium plus 100 μg/ml ampicillin at 30 °C to OD₆₀₀= 0.5. Protein expression was induced with the addition of 0.002% arabinose, and growth was continued for 3 h. Processing of cell pellets was as described above.

VLP preparation from E. coli lysates

Approximately, 500 μg of protein in clarified whole cell lysates produced as described above from E. coli cells expressing Gag3 or CA, was fractionated over a 20/30/70% (542 μl, 253 μl, and 181 μl respectively) sucrose step gradient by centrifugation in a TLS55 rotor in the Optima TL-100 ultracentrifuge (Beckman-Coulter, Fullerton, CA) at 86K × g for 68 min at 4°C. Eight 175 μl fractions were successively removed from the top of the gradient and the interface fraction which contained the majority of Ty3 protein as determined by immunoblotting using anti-Ty3 CA and Coomassie Brilliant Blue R-250 staining (Fig. 3 and data not shown) was used for AFM.

Atomic force microscopy

A 5 μl aliquot of the 70/30% interface fraction collected as described above was used for AFM imaging in fluid using tapping mode (Kuznetsov, Zhang, Menees, McPherson, & Sandmeyer, 2005). The atomic force microscope was a Nanoscope IIIa (VEECO, Santa Barbara, CA). VLPs were scanned in tapping mode with an oscillation frequency of 9.2 KHz and a scan frequency of 1 Hz using oxide sharpened silicon nitride tips in a 75 μl fluid cell (Hansma, Browne et al., 1994;Hansma & Hoh, 1994). Scanning was performed at room temperature. Nucleic acid was extruded from particles which were broken in 1.5M KCl at pH 12 with addition of sodium hydroxide.

Supplementary Material

Click here to view.^{(4.4M, tif)}

Click here to view.^{(24K, doc)}

Click here to view.^{(49K, doc)}

Acknowledgments

This research was supported by Public Health Services grants GM33281 to S.B.S., GM68903 to G. W. H., and GM58868 to AM. We thank M. Zhang for helpful discussions and M. Oakes for assistance with fluorescence microscopy.

Footnotes

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

Adamson CS, Jones IM. The molecular basis of HIV capsid assembly -five years of progress. Rev Med Virol. 2004;14:107–121. [PubMed]
Campbell S, Vogt VM. Self-Assembly In-Vitro of Purified CA-NC Proteins from Rous-Sarcoma Virus and Human-Immunodeficiency-Virus Type-1. J Virol. 1995;69:6487–6497. [PMC free article] [PubMed]
Campbell S, Vogt VM. In vitro assembly of virus-like particles with Rous sarcoma virus Gag deletion mutants: identification of the p10 domain as a morphological determinant in the formation of spherical particles. J Virol. 1997;71:4425–4435. [PMC free article] [PubMed]
Ehrlich LS, Agresta BE, Carter CA. Assembly of recombinant human immunodeficiency virus type 1 capsid protein in vitro. J Virol. 1992;66:4874–4883. [PMC free article] [PubMed]
Ehrlich LS, Liu TB, Scarlata S, Chu B, Carter CA. HIV-1 capsid protein forms spherical (immature-like) and tubular (mature-like) particles in vitro: Structures switching by pH-induced conformational changes (vol 81,pg 586, 2001) Biophys J. 2001;81:586–594. [PubMed]
Ganser BK, Li S, Klishko VY, Finch JT, Sundquist WI. Assembly and analysis of conical models for the HIV-1 core. Science. 1999;283:80–83. [PubMed]
Gross I, Hohenberg H, Huckhagel C, Krausslich HG. N-Terminal extension of human immunodeficiency virus capsid protein converts the in vitro assembly phenotype from tubular to spherical particles. J Virol. 1998;72:4798–4810. [PMC free article] [PubMed]
Hansma HG, Browne KA, Bezanilla M, Bruice TC. Bending and Straightening of Dna Induced by the Same Ligand - Characterization with the Atomic-Force Microscope. Biochem. 1994;33:8436–8441. [PubMed]
Hansma HG, Hoh JH. Biomolecular Imaging with the Atomic-Force Microscope. Annu Rev Biophys Biomol Struct. 1994;23:115–139. [PubMed]
Hatfield GW, Gutman GA. Codon Pair Utilization Bias in Baceria, Yeast and Mammals. In: Hatfield DL, Lee BJ, Pirtle RM, editors. Transfer RNA in Protein Synthesis. CRC PRess; Boca Raton FL: 1992. pp. 157–190.
Johnson MC, Scobie HM, Ma YM, Vogt VM. Nucleic acid-independent retrovirus assembly can be driven by dimerization. J Virol. 2002;76:11177–11185. [PMC free article] [PubMed]
Joshi SM, Vogt VM. Role of the rous sarcoma virus p10 domain in shape determination of Gag virus-like particles assembled in vitro and within Escherichia coli. J Virol. 2000;74:10260–10268. [PMC free article] [PubMed]
Kirchner J, Sandmeyer SB, Forrest DB. Transposition of a Ty3 GAG3-POL3 fusion mutant is limited by availability of capsid protein. J Virol. 1992;66:6081–6092. [PMC free article] [PubMed]
Kuznetsov YG, Zhang M, Menees TM, McPherson A, Sandmeyer S. Investigation by atomic force microscopy of the structure of Ty3 retrotransposon particles. J Virol. 2005;79:8032–8045. [PMC free article] [PubMed]
Larsen LS, Zhang M, Beliakova-Bethell N, Bilanchone V, Lamsa A, Nagashima K, Najdi R, Kosaka K, Kovacevic V, Cheng J, Baldi P, Hatfield GW, Sandmeyer S. Ty3 capsid mutations reveal early and late functions of the amino-terminal domain. J Virol. 2007;81:6957–6972. [PMC free article] [PubMed]
Luschnig C, Hess M, Pusch O, Brookman J, Bachmair A. The gag homolog of retrotransposon Ty1 assembles into spherical particles in Escherichia coli. Eur J Biochem. 1995;228:739–744. [PubMed]
Muriaux D, Mirro J, Harvin D, Rein A. RNA is a structural element in retrovirus particles. Proceedings of the National Academy of Sciences U S A. 2001;98:5246–5251. [PubMed]
Rumlova-Klikova M, Hunter E, Nermut MV, Pichova I, Ruml T. Analysis of Mason-Pfizer monkey virus gag domains required for capsid assembly in bacteria: Role of the N-terminal proline residue of CA in directing particle shape. J Virol. 2000;74:8452–8459. [PMC free article] [PubMed]
Sandmeyer SB, Aye M, Menees TM. Ty3: a position-specific, gypsylike element in Saccharomyces cerevisiae. In: Craig NL, Craigie R, Gellert M, Lambowitz AM, editors. Mobile DNA II. ASM Press; Washington, D.C.: 2002. pp. 663–682.
Scarlata S, Carter C. Role of HIV-1 Gag domains in viral assembly. Biochim Biophys Acta-Biomembranes. 2003;1614:62–72. [PubMed]
Ulbrich P, Haubova S, Nermut MV, Hunter E, Rumlova M, Ruml T. Distinct Roles for Nucleic Acid in In Vitro Assembly of Purified Mason-Pfizer Monkey Virus CANC Proteins. J Virol. 2006;80:7089–7099. [PMC free article] [PubMed]
von Schwedler UK, Stemmler TL, Klishko VY, Li S, Albertine KH, Davis DR, Sundquist WI. Proteolytic refolding of the HIV-1 capsid protein amino-terminus facilitates viral core assembly (vol 17, pg 1555, 1998) EMBO J. 1998;17:1555–1568. [PubMed]
Yu F, Joshi SM, Ma YM, Kingston RL, Simon MN, Vogt VM. Characterization of Rous sarcoma virus Gag particles assembled in vitro. J Virol. 2001;75:2753–2764. [PMC free article] [PubMed]