Organization of Genes in Prototypic Retroviruses

Although the order of the genes common to all retroviruses—gag, pro, pol, and env—is invariant, the relationship of their reading frames is not. The differences in these relationships have important consequences for gene expression. Diagrammatic representations of prototypic genomes from the presently accepted genera of retroviruses (Chapter 1) are shown as proviral DNAs in Figure 5 (left), and properties of the protein products of these genes are listed in Table 1. Figure 5 (right) shows examples of viruses of the ASLV genus and the MLV genus that carry oncogenes.

The most common arrangement of gag, pro, and pol is typified by HIV-1. pro and pol are in the same (–1) reading frame relative to gag. Thus, the viral protease is translated at the same level—about 5% of the level of Gag—as the downstream domains for reverse transcriptase and integrase. In MLV and its close relatives, protease, reverse transcriptase, and integrase are translated together as in HIV-1, but the pro and pol genes are in the same reading frame as gag, reflecting the readthrough of the termination codon at the end of gag. The ASLVs are unusual in that pro is in the same frame as gag, and thus protease forms the carboxy-terminal domain of the Gag protein. As a consequence, about 20 times more protease is translated and incorporated into virions than in HIV-1 or MLV. The fact that the Gag proteins and protease are synthesized in stoichiometric amounts suggests that the ASLV protease may have a structural role in the virion, and indeed there is some evidence for this possibility, protease deletions reducing budding in some cell types (Chapter 7). Yet another arrangement of these genes is exemplified by M-PMV, MMTV, and HTLV. In these cases, the pro gene stands in its own reading frame, requiring two frameshift events for complete synthesis of the Gag-Pro-Pol precursor.

The env gene in all retroviruses is expressed by translation of a subgenomic mRNA. Soon after the discovery of RNA splicing in adenovirus and SV40 in 1977, it became clear that this process also is responsible for generation of subgenomic RNAs in retroviruses (Hayward 1977; Mellon and Duesberg 1977; Weiss et al. 1977). Splicing fuses the leader portion of the primary transcript to the body of the gene. Typically, the splice donor is upstream of gag, but in some viruses, exemplified by ASLV, the splice donor is positioned a few codons into the gag gene, resulting in a primary Env translation product that includes a few amino-terminal amino acid residues in common with Gag. In many cases, env overlaps the carboxy-terminal portion of pol; neither the functional significance nor the evolutionary significance of this arrangement is known. The Env protein is synthesized on membrane-bound polyribosomes and transported by the cell's vesicular traffic to the plasma membrane, where it is incorporated into viral particles (see below).

The products of most retroviral accessory genes, which were discovered first in HTLV (Seiki et al. 1983), regulate transcription of viral DNA and processing of RNA, but in lentiviruses, they include proteins that also serve other functions (Trono 1995). Accessory genes are located in a variety of places downstream from pol; they commonly partially overlap env and U3, as well as each other, perhaps reflecting the evolutionary pressure implied by the limit to the size of an RNA molecule that can be packaged into a virion (Chapters 6 and 9). All accessory genes are expressed from singly or multiply spliced RNAs. In the simplest cases, the genes are located only downstream from env, sometimes extending into the U3 region of the LTR. This situation is exemplified by the bel gene of HFV and the nef gene of HIV-1. In more complicated versions of the genome, found, for example, in HTLV, the accessory genes tax and rex comprise two exons, one of which is upstream or overlapping the 5′region of env and the other downstream or overlapping the 3′region of env. Some lentiviral accessory genes are positioned entirely upstream of env, as, for example, the vif gene of HIV-1. Others, such as tat and rev, comprise exons upstream of as well as within env, but in different reading frames. A summary of the functions of proteins encoded by accessory genes is presented in Table 1.