ScienceWeek

EVOLUTION: ON EPISTASIS IN RNA VIRUSES

Notes by ScienceWeek:

Whether viruses are classified as "living" or "non-living" is arbitrary, but in either case one is dealing with a remarkable entity. As a group, viruses are in the size range 20 to 300 nanometers, compact bundles of genetic information that have apparently existed for billions of years. They are unable to replicate except inside a living host cell, and outside the host cell they are inert. Viruses pass through filters that trap bacteria, and they can be seen only be electron microscopy. Like other biological systems, each virus contains molecular genetic information in the form of nucleic acid, but in viruses this information comes in the form of DNA or RNA, never both. The viral genome is transcribed and replicated only with a host cell, with variations in process dependent on the type of virus and the type of host cell.

When viruses are categorized in terms of their genomes, there are two general types: a) viruses with a DNA genome (DNA viruses), and b) viruses with an RNA genome (RNA viruses). RNA viruses are unique: only in these viruses do we find genomes consisting of RNA; all other biological entities, DNA viruses, bacteria, plant cells, animal cells, etc., contain DNA genomes.

There are more than 2500 groups of different viruses now recognized and at least partially characterized. In each case, for both DNA and RNA viruses, once it enters the host cell, the general challenge for the virus is the same: directly or indirectly, the viral genome must bring about the production of the *messenger RNAs needed by the host *ribosomes to produce the specific proteins necessary for replication of the complete virus.

With DNA viruses, the DNA viral genome acts as the template for the production of messenger RNA. With RNA viruses, however, the process is more complicated.

In general, with some types of RNA viruses, the RNA genome ("plus-sense"; "positive-strand") can itself act as messenger RNA for host ribosomes; while other types of RNA viruses, the RNA genome ("minus-sense"; negative-strand) must first produce a complementary RNA, which then acts as messenger RNA for the host ribosomes. The replication process in minus-sense RNA viruses is complex, since host cells do not carry enzymes that can polymerize complementary RNA from an RNA template, and such viruses therefore must carry their own special enzymes ("RNA-dependent transcriptases") to achieve this synthesis.

A third and special type of RNA virus is the so-called "retrovirus", of which there are many versions. Retroviruses are single-stranded RNA viruses that have an enzyme called reverse transcriptase, and with this enzyme the viral RNA is used as a template to produce viral DNA from host-cellular material. This DNA is then incorporated into the host cell's genome, where it codes for the production of messenger RNA and the ultimate synthesis of viral components. The HIV virus, for example, is a retrovirus.

As a class, retroviruses are usually spherical, 80 to 110 nanometers in diameter, with an RNA genome of approximately 7000 to 10,000 nucleotide bases. An outstanding characteristic of such viruses is that if they kill host cells at all it is usually only after a long latent period (although there are certain important exception). In addition, these viruses are apparently capable of altering, or affecting the expression of, host cell genes involved in cancer (oncogenes).

The three primary genes of the retrovirus genome are called "gag", "pol", and "env". The gag gene encodes the protein of the virus capsid, the protein coat directly encapsulating the viral genome; the pol gene encodes a reverse transcriptase involved in replication of the genome; the env gene encodes the protein of the membrane envelope of the virus when it is outside a host cell (the membrane envelope of the "virion"). (These 3 genes actually produce more than 3 different proteins; what these genes encode for are precursor proteins, each of which is a precursor for several varieties of proteins with different viral functions. In addition, different proteins can be produced by splices of elements from the 3 primary genes.

Concerning the general structure of the retrovirus, the internal nucleic acid genome is encapsulated by a protein coat (the capsid), and the capsid in turn is surrounded by the external lipoprotein envelope. Beyond this general scheme, there is no single morphology for retroviruses.

The protein encoded by the env gene (called the Env protein) is important in recognition of host-cell surface receptors, with which it interacts to secure entry of the virus into the host cell. In addition, after infection the host-cell expresses Env protein on its own surface, and this evidently prevents reinfection of that host cell after new viruses are released.

Perhaps the most important property of retroviruses is their ability to splice into the host-cell genome pieces of their own genome, the result either the introduction of new genes into the host genome or the activation or inactivation of specific nearby genes of the host genome. In principle, either of these scenarios can lead to corruption of the growth regulation process of the host genome, and thus to a line of malignant cells.

Viruses have been implicated in the etiology of several types of human cancers, including cervical cancer and liver cancer. The viruses that have been strongly associated with human cancers include human papillomaviruses, Epstein-Barr virus, hepatitis B virus, and a human retrovirus. Many viruses can cause tumors in animals, either as a result of natural infection or after experimental inoculation. Animal viruses are studied to learn how a limited amount of genetic information (only one or a few viral genes) can profoundly alter the growth behavior of host cells, and ultimately convert a normal cell into a malignant cell. For example, classical studies of RNA tumor viruses first revealed the involvement of cellular *oncogenes in cancer, and similar studies with DNA tumor viruses first implicated a role for *tumor suppressor genes in cancer. In the 1980s, these discoveries revolutionized thinking about the molecular mechanisms of carcinogenesis.

The following points are made by Y. Michalakis and D. Roze (Science 306:1492):

1) If one crosses two black guinea pigs one may get several albino pups. This could happen, for example, if the boar and sow are both heterozygous at the C locus for the c-a allele, which prevents the production of the pigment melanin. Guinea pig fur color is determined by many genes, but the C locus strongly affects the expression of these genes in homozygous c-a individuals. The C locus is said to be "epistatic" because it affects the expression of genes at other loci.

2) Developmental biology is replete with examples of epistatic gene interactions. In contrast, evolutionary biology is desperately searching for them. Epistatic interactions are usually important in evolutionary biology whenever multilocus genetics are of significance. The recent explosion of interest in epistatic interactions is epitomized by publication of a book devoted to the topic [3]. Regarding his work on guinea pig fur color and its determination, Wright has admitted that this study was one of four factors that led him to propose his "shifting balance theory" of adaptive landscapes [4]. Epistatic interactions are also prominent in studies of speciation and reproductive isolation [5]. And, inevitably, they are central to the evolution of genetic recombination.

3) There are very few examples of epistatic gene interactions in evolutionary biology, principally because of the difficulties inherent in performing such studies. One either needs to know all the genes determining a trait, which is typically rare in evolutionary biology, or one needs to conduct relatively laborious breeding experiments. Moreover, the relation between the trait's value and its fitness benefit is crucial and must be at least partly known. It is not surprising, therefore, that the few examples of epistasis in evolution come from microorganisms. New RNA virus studies have provided measures of epistatic gene interactions allowing fresh insights into their effects on evolution [1,2].

4) Sanjuan et al[1] examined vesicular stomatitis virus (VSV), which does not exhibit recombination. They analyzed a set of single-nucleotide mutations generated by directed mutagenesis. The individual fitness effects of these mutations, deleterious or beneficial, were previously established. The authors reported the fitness effects of pairs of mutations that are either deleterious or beneficial when single. Remarkably, singly beneficial mutations when paired exhibit negative epistasis, whereas singly deleterious mutations when paired exhibit positive epistasis. In both cases, the mutation pairs act antagonistically: Their combined effect is less than that expected from their individual effects.

5) In a related study, Bonhoeffer et al[2] presented evidence for positive epistasis in another RNA virus, human immunodeficiency virus 1 (HIV-1). These authors examined the evolution of recombination in HIV-1, which unlike VSV is known to recombine frequently. If recombination in HIV-1 had been selected for because of gene interactions, then epistasis between pairs of genes should be slightly negative. But Bonhoeffer et al[2] reported that gene interactions in HIV-1 exhibit positive epistasis.

6) The two RNA studies[1,2] demonstrate that the pattern of epistasis in RNA viruses is not compatible with current genetic theories of sexual reproduction and recombination, which assume that mutations affecting fitness exhibit negative epistasis.

References (abridged):

1. R. Sanjuan, A. Moya, S. F. Elena, Proc. Natl. Acad. Sci. U.S.A. 101, 15376 (2004)

2. S. Bonhoeffer, C. Chappey, N. T. Parkin, J. M. Whitcomb, C. J. Petropoulos, Science 306, 1547 (2004)

3. J. B. Wolf, E. D. I. Brodie, M. J. Wade, Eds., Epistasis and the Evolutionary Process (Oxford Univ. Press, New York, 2000)

4. S. Wright, J. Anim. Sci. 46, 1192 (1978)

5. S. Gavrilets, Fitness Landscapes and the Origin of Species (Princeton Univ. Press, Princeton, NJ, 2004)

Science http://www.sciencemag.org

--------------------------------

Related Material:

CLONAL INTERFERENCE AND THE EVOLUTION OF RNA VIRUSES

Notes by ScienceWeek:

Studies of the dynamics of evolution indicate that populations adapt through the appearance and subsequent fixation of beneficial mutations. In large populations, beneficial mutations may arise frequently enough so that two or more are simultaneously present in independent lineages. Once beneficial mutations have arisen, there is a certain probability of losing them by "*genetic drift" while their frequency is low, but after an early period dominated by genetic drift, beneficial mutations reach a substantial frequency in the population. In a sexual system, these beneficial mutations will eventually combine, ensuring their fixation together. But if the system is asexual, the lineages created by beneficial mutations will compete, and only the mutation with the largest effect will be fixed. This competition of beneficial mutations in asexual populations is called "*clonal interference", and it is a phenomenon that ensures that beneficial mutations that do achieve fixation are of large effect.

Excluding *prions, whose categorization is still uncertain, viruses are the smallest infectious agents, ranging from approximately 20 to 300 nanometers in diameter, and containing only one kind of nucleic acid (RNA or DNA) as their genome. In general, the nucleic acid is encased in a protein shell, which may be surrounded by a lipid-containing membrane, and the entire infectious unit is termed a "virion". Viruses are inert in the extracellular environment, replicating only in living cells, and they are essentially genetic-level parasites: the viral nucleic acid contains information necessary for programming the infected host cell to synthesize various virus-specific macromolecules required for the production of viral progeny [*Note #1].

RNA viruses are viruses whose genome consists of RNA, and they are involved in a number of serious human diseases, including hemorrhagic fevers and human immune deficiency syndrome (AIDS). One important characteristic of RNA viruses is that as a group they show the highest mutation rates in nature. This, together with their potentially large effective population sizes, and the fact that their reproduction can be asexual, suggests that clonal interference may play an important role in their adaptive evolution.

The following points are made by R. Miralles et al (Science 1999 285:1745):

1) The authors report a study of clonal interference in an RNA virus population (*vesicular stomatitis virus), the study involving two variants differing only in their ability to grow in the presence of a *monoclonal antibody. The authors report their results provide evidence that clonal interference does indeed occur in viral populations, and that this evidence along with models of clonal interference allows certain properties of the adaptive evolution of RNA viruses to be inferred. The authors suggest the following:

2) Adaptive substitutions appear as discrete rare events, regardless of mutation rate or population size. They often do not occur simply as the result of a single mutation but instead represent the best of several competing mutations. The authors suggest this fact has consequences for the dynamics of drug resistance and the search for resistance mutations.

3) In medium to large populations, the rate of fitness increase is hardly affected by changes in either mutation rate or population size.

4) Resident populations are protected from invaders simply because of their numerical advantage. A high-fitness vesicular stomatitis viral clone seeded at low frequency into a resident population of low-fitness variants was displaced by the low-fitness competitors. When its initial frequency was above a certain threshold, however, the high-fitness clone always outcompeted the low-fitness variants in the resident population. The authors suggest that the existence of a frequency threshold for dominance imposes an element of uncertainty in virus sampling during disease outbreaks.

Science http://www.sciencemag.org

--------------------------------

Notes by ScienceWeek:

genetic drift: The term "genetic drift" refers to the random fluctuations of gene frequencies due to sampling errors. Genetic drift occurs in all populations, but its effects are most evident in populations that are small.

clonal interference: In this context, a "clone" is a lineage derived from a single ancestor.

prions: Prions are a class of poorly understood proteins implicated in a number of exotic human neurological diseases and in some common animal diseases such as sheep scrapie and bovine spongiform encephalopathy in cattle ("mad cow disease").

Note #1: Over 4000 animal and plant viruses have been identified (as of 1995), these entities categorized into 71 families, 11 subfamilies, and 164 genera, with hundreds of viruses still unassigned. 24 families contain viruses that infect humans and animals. Classical categorizations of viruses were based on the diseases they produces, but modern categorizations are based on molecular biological parameters.

vesicular stomatitis virus: This virus causes a disease of cattle. The viral entity is bullet-shaped, approximately 75 nanometers in diameter, 180 nanometers in length, the genome single-stranded RNA, 13-16 kilobases in size. The virus replicates in cell cytoplasm, with viral assembly involving budding from the cell membrane.

monoclonal antibody: In general, a monoclonal antibody is an immunoglobulin protein derived from a single clone of plasma cells. Such antibodies are chemically and structurally identical and constitute a pure population with highly specific antigen-binding properties. In general, an "antigen" is any chemical entity that activates an immune response, especially an entity originating outside the body, and an "antibody" is a specific immunoglobulin protein produced by an immune cell, the protein specifically binding a particular antigen. In the context of this report, a typical antigen would be a moiety in the surface coat of the virus.

ScienceWeek http://scienceweek.com