Douglas W. Grosenbach and Dennis E. Hruby

Department of Microbiology, Center for Gene Research and Biotechnology, Oregon State University, Corvallis, Oregon, 97331, U.S.A.

Received 2/27/98 Accepted 3/5/98

2. INTRODUCTION

2.1 Overview of the Vaccinia Life Cycle

Vaccinia virus (VV) is the most extensively characterized member of the Orthopoxvirus family of the Poxviridae (1). Its role as a vaccine in the eradication of smallpox represents one of the major achievements in medicine. Although smallpox vaccines are no longer necessary, VV is widely used as a molecular and cellular biology tool as well as in the development of recombinant vaccines. In order to enhance or optimize VV for these purposes, basic research on the virus itself remains a major area of interest. Current research topics include structure/morphogenesis, replication/resolution of the genome, transcriptional regulation, enzymology, immune modulation, and protein processing.

VV is a large double-stranded DNA virus with a broad mammalian host range. Its entire life cycle occurs in the cytoplasm of the host with little or no requirement for the host cell nucleus (figure 1). Upon entry into the cell, the virion core is partially uncoated allowing transcription of the early class of genes. Their products are involved in genome replication and as transcription factors for the intermediate class of genes. The VV genome is composed of a nearly 200 kilobasepair linear double-stranded DNA molecule with potential to encode more than 200 polypeptides (2). Host protein synthesis is rapidly and efficiently inhibited upon infection so that nearly the entire translational capacity of a cell is harnessed by the virus. Following intermediate gene expression and genome replication the late class of genes are expressed. The majority of the late gene products are involved in the late stages of virion development serving as structural components or scaffolding for nascent virions. Virion morphogenesis occurs in perinuclear macromolecular clusters containing numerous copies of the viral genome and virus encoded proteins referred to as virus factories, viroplasm or virosomes. Immature virions form by packaging viroplasm within membrane crescents most likely derived from the membranes of the intermediate compartment (3). The genome is packaged within these membrane crescents in the proteinacious core with the virus-encoded RNA polymerase and numerous cofactors involved in the early stages of the virus life cycle. Concomitant with the proteolytic processing of the three major core proteins (4), the core condenses, producing the first infectious form of the virus which is referred to as intracellular mature virus (IMV). At this stage the core is wrapped with two proteolipid envelopes.

Figure 1. The VV replication cycle. A diagram of the infected cell is shown with an exaggerated view of the endoplasmic reticulum (ER), cis, medial, trans Golgi and the trans-Golgi network (C, M, T, and TGN respectively). The major stages of the virus life cycle are listed. Following late gene expression, previrion forms assemble to form intracellular mature virus (IMV). The IMV is targeted to the TGN and following envelopment, intracellular enveloped virus (IEV) is formed. IEV are propelled to the cell surface by the polymerization of actin filaments. Once released the virus may remain attached to the membrane as cell-associated enveloped virus (CEV) or be released into the medium as extracellular enveloped virus (EEV).

The IMV particle is targeted to the trans-Golgi network (TGN) and by budding through the compartment acquires two additional membranes (5). After enwrapment by TGN membranes, the quadruple membrane-bound particle referred to as intracellular enveloped virus (IEV), is propelled to the cell surface by the formation of thick actin filaments behind it (6). At the cell surface the outermost virion membrane may be lost by fusion with the cell membrane resulting in the release of a triple membrane-bound particle referred to as extracellular enveloped virus (EEV). If the virion remains attached or reattaches to the cell surface it is referred to as cell-associated enveloped virus (CEV). Some poxviruses also package virions in cytoplasmic inclusion bodies referred to as A-type inclusions (ATI) which are primarily composed of a single protein, the ATI protein (7). The ATI protein is truncated and correspondingly nonfunctional in VV so that no inclusion bodies form although the truncated form of the ATI protein is still expressed at high levels (8).

The enveloped forms of the virus (IEV, EEV, and CEV) are antigenically distinct from the IMV particle in that they contain at least six proteins in their outer envelope(s) that are not present on IMV. They are encoded by the A33R (9), A34R (10), A36R (11), A56R (12), B5R (13, 14) and F13L (15) open reading frames (ORFs) of VV. All of these proteins have been demonstrated to play important roles in the formation, release and/or infectivity of EEV with the exception of the A56R gene which encodes the viral hemagglutinin. None of these proteins affects the formation of IMV. Preliminary studies suggest that the IMV particle also contains proteins not found on the multiply enveloped forms of the virus. The ATI protein and the 4c protein are unique to IMV and although their biological relevance is not known they may represent an evolutionary relic. While VV does not occlude virions in ATIs, the closely related cowpox virus does (7). In that system, it has been demonstrated that the 4c protein is required for occlusion of virions in ATIs. It may be that the association of the ATI and 4c proteins with the VV IMV particle is an abortive attempt at ATI formation.

The virus and its life cycle are complex and not completely understood. Throughout its life cycle the virus uses numerous cellular protein processing pathways to include proteolytic processing, phosphorylation, sulfation, glycosylation, and ADP-ribosylation (16). Here we review the acylation of VV proteins and discuss the significance of these proteins and their respective acyl modifications in the biology of VV.

2.2 Overview of Protein Acylation

Two classes of fatty acylated proteins exist in eukaryotic cells (17) and by extension are present in VV-infected cells as well. Labeling of cultured cells with [³H]-myristic or [³H]-palmitic acid identifies distinct subsets of proteins. The distinction between the two classes may be determined in two ways. First, myristylation of proteins is a cotranslational event and is inhibited by the addition of reagents which block translation. Under these conditions, palmitylation of previously translated proteins occurs normally while myristylation does not. Second, the palmitate-protein bond is labile in the presence of mild alkali or neutral hydroxylamine due to the thioester linkage while the amide-linked myristate is stabile under the same conditions. Many palmitylproteins are membrane associated either directly through the palmitate moiety or as transmembrane proteins anchored by the fatty acid although a few are actually secreted from cells. Cell-associated palmitylproteins are distributed throughout the cell with the greatest concentration at the cell surface. Myristylproteins, on the other hand, may be cytoplasmic or membrane-associated. Some proteins are modified by both myristic and palmitic acid with both acyl moieties contributing to protein function and localization.

Myristylation of proteins involves the transfer of myristate from myristyl-Coenzyme A to the amino-terminal motif MGXXX(S/T/A/C/N) (using the single-letter amino acid code) of proteins by the enzyme N-myristoyl transferase (NMT) (18) This motif bestows substrate specificity for the Saccharomyces cerevisiae-encoded enzyme and although the human homologue has similar specificity, it may not be entirely the same. The initiating methionine is removed by methionine amino peptidase during translation and NMT recognizes the newly generated amino-terminal glycine of the nascent peptide after approximately twenty residues are free of the ribosome. NMT transfers myristate to the glycine residue after which the enzyme releases the peptide and translation proceeds normally. Mutations which change any of the conserved residues of the motif inhibit myristylation, with the greatest inhibition achieved by replacing the penultimate glycine with any residue (19). The residues at position six of the motif are less important with low levels of myristylation occurring even if they are changed. It should be noted though that not all proteins containing the amino-terminal motif are myristylated indicating that there are additional requirements. See figure 2 for the "ball and stick" model of a multiply fatty acylated peptide.

Figure 2. Structure of a hypothetical acylated peptide. A peptide consisting of the canonical amino-terminal myristylation motif is shown. The amino-terminal methionine is cleaved and the penultimate glycine residue is modified by the amide-bound 14 carbon fatty acid myristate. Cysteine, in position 3 is modified by the thioester-linked 16 carbon fatty acid palmitate as in Type 4 palmitylproteins (see text). An internal lysine residue is modified by myristate as well. Alternate acceptable amino acids for each position are indicated in parentheses above the peptide backbone.

Recently, myristylation has been demonstrated to occur on internal residues of proteins (20). Although the motif directing this modification has not been discovered, the myristate acceptor residues appear to be arginine or lysine (figure 2). This may be due to the fact that these residues have free amines in their side-chains and is supported by experiments demonstrating their insensitivity to neutral hydroxylamine suggesting amide linkage for this modification as well. The enzyme(s) responsible for internal myristylation is unknown.

Palmitylation of proteins remains an enigma for researchers. Palmitylproteins are more aptly described as ester-type fatty acylated (21) proteins or even S-acylated proteins (22), in reference to the sulfur atom in the side-chain of the acceptor cysteine. Some proteins are preferably modified by stearic or oleic acids over palmitic acid but for the most part palmitate is the predominant fatty acid on these proteins with small percentages of them being modified by stearate and oleate or even arachidonate (23). While cysteine is the most common acceptor residue, serine or threonine may also serve as palmitate acceptors (24), so even the description of these proteins as S-acylated is not totally accurate.

To date, a consensus motif directing palmitylation of proteins as well as the enzyme(s) responsible for the modification are unknown. Palmitylprotein acyltransferase has been described as a membrane-bound component of cells with in vitro activity detected in endoplasmic reticulum, Golgi and plasma membrane fractions of cells. The enzyme has a requirement for the activated form of palmitate, palmityl-CoA, but little else is known about it. An enzyme capable of removing palmitate from proteins, palmityl-protein thioesterase, has been discovered, although it is secreted from cells and most likely is specific for secreted palmityl-proteins.

Some general characteristics of palmitylproteins are known and have allowed us to classify them into four subclasses (25). Type I palmitylproteins are transmembrane proteins that are modified on cysteines at or near the cytoplasmic membrane face. This group is typified by the G-protein coupled receptors and includes the vesicular stomatitis virus G and the influenza virus hemagglutinin proteins. The palmitylation of Type II proteins occurs in the carboxy-terminal region and is dependent on prior prenylation of cysteine in the CAAX motif at the extreme carboxy-terminus. Members of this group include the ras proteins. Types III and IV are dually fatty acylated in the amino-terminal region. Both groups are myristylated on glycine of the motif MGXXX(S/T/A/C/N). Type III palmitylproteins are modified on one or more cysteines within the first 10 to 20 amino acids while Type IV palmitylproteins are modified on cysteine immediately following the myristylated glycine residue (figure 2). Efficient palmitylation of Types III and IV are dependent on prior myristylation. The alpha subunits of the heterotrimeric G-proteins are grouped as Type III or IV.