Virology 270, 251–253 (2000) doi:10.1006/viro.2000.0320, available online at http://www.idealibrary.com on
MINIREVIEW Viruses at the Edge of Adaptation Esteban Domingo 1 Centro de Biologı´a Molecular “Severo Ochoa,” (CSIC-UAM) Universidad Auto´noma de Madrid, Cantoblanco, 28049 Madrid, Spain Received February 25, 2000; accepted March 14, 2000 How vulnerable is the line that separates adaptation from extinction? Viruses, in particular RNA viruses, are well known for their high rates of genetic variation and their potential to adapt to environmental modifications (Drake and Holland, 1999; Domingo et al., 2000). Yet, fitness variations—both increases and decreases—can be spectacularly rapid, and the simple genetic stratagem of forcing virus multiplication to go through repeated genetic bottlenecks can induce fitness losses, at times near viral extinction. New information has been recently obtained on the two sides of the survival line: the edge of adaptation and the edge of extinction. © 2000 Academic Press
present in the population. Furthermore, this particular variant was systematically selected, while, when absent from memory, its selection was rare (Ruiz-Jarabo et al., 2000). In agreement with the concept that memory genomes are part of the mutant spectrum of the quasispecies, memory was erased when the populations were subjected to genetic bottlenecks (Fig. 1). Because of the existence of a molecular memory, a viral quasispecies may be capable of reacting swiftly to a selective constraint which has been already experienced by the same virus population. Immune and other internal physiological responses in infected hosts are hardly uniform and continuous; the same must apply to concentrations of antiviral inhibitors when they reach the target replication sites of viruses. Memory—a feature of biological, complex adaptive systems (such as the immune system)—is likely to represent an advantage for the virus whenever fluctuating selective pressures occur. Are viruses which are endowed with such subtle adaptive strategies immune to extinction? A complementary line of research suggests that, hopefully, they are not.
SUBTLE ADAPTIVE STRATEGIES RNA virus quasispecies are highly effective in the exploration of new genomic sequences (Eigen and Biebricher, 1988). It has long been known that mutant swarms (the mutant spectra that populate viral quasispecies) contain potentially useful variants [those with increased resistance to antiviral agents, altered interferoninducing capacity, or antibody-escape and cytotoxic T lymphocyte (CTL)-escape mutants, among others]. A new feature of quasispecies dynamics which may turn out to be important for viral adaptation has been revealed: the presence of a memory of past evolutionary history, imprinted on minority components of the mutant spectrum (Ruiz-Jarabo et al., 2000). The experiments took advantage of a number of well-characterized mutants of the important animal pathogen foot-and-mouth disease virus (FMDV). One of the memory markers was an unusual internal polyadenylate, known to be associated with fitness decrease of FMDV clones subjected to repeated plaque-to-plaque transfers (Escarmı´s et al., 1996, 1999). When clones were allowed to recover fitness, genomes with additional adenylate residues were no longer detected in the consensus sequence, but were maintained in the mutant spectrum after at least 225 doublings of the amount of infectious particles. The same was true of an antigenic variant of FMDV. Its presence in the quasispecies memory resulted in an increase in the frequency of antibody-escape mutants
THE PRICE OF MISTAKES The maintenance of genetic information of any organism requires a minimal fidelity level during copying of template molecules in the process of genome replication. This applies to viruses (Eigen and Biebricher, 1988; Schuster and Stadler, 1999). Excess mutations—the socalled violation of the error threshold—should abolish virus infectivity (Fig. 2). Experimental support for this concept was obtained initially for poliovirus, vesicular stomatitis virus (Holland et al., 1990), and a retroviral vector (Pathak and Temin, 1992), all of which underwent
1 To whom correspondence should be addressed. Fax: 34-913974799. E-mail:
[email protected].
251
0042-6822/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
252
MINIREVIEW
FIG. 1. Schematic representation of memory in viral quasispecies. Viral quasispecies are represented by genome distributions (horizontal lines) with an average of about four mutations (symbols on lines) per genome. To illustrate memory, the mutant spectrum on the left is divided into three frequency levels: 1, the most frequent genomes; 2, minority components potentially occupied by memory genomes; and 3, basal minority components as a result of mutational pressure and competitive rating of mutants (Eigen and Biebricher, 1988; Schuster and Stadler, 1999; Domingo et al., 2000). Initially, a marker “A” (indicated on the genome) belongs to the basal minority components. When this marker is selected, it becomes dominant (middle distribution, with accompanying random mutations, since marker A belongs to a category of genomes, not to a single genome). When genomes with A show a selective disadvantage relative to competitors lacking A (for example, revertants), genomes with A eventually will not be detected in the consensus sequence after extensive replication. However, they may remain at level 2 as part of memory genomes of the mutant spectrum (upper right distribution). In the case of antigenic variants analyzed by Ruiz-Jarabo et al. (2000), the initial selection for A resulted in a number of alternative variants (with different antigenically relevant amino acid substitutions), each isolated with low frequency. In contrast, once a specific variant was incorporated into memory its selection became deterministic. When bottleneck passages intervene (small, filled arrow), memory is lost and genomes with A return to level 3 (lower right distribution). Different memory levels can be attained for different genetic markers, depending on fitness values and mutation rates (Ruiz-Jarabo et al., 2000), in what appears to be a highly dynamic process which is now under further investigation.
drastic losses of infectivity when subjected to chemical mutagenesis during their replication. Recently, additional evidence has been obtained with HIV-1, in a series of elegant experiments employing mutagenic deoxynucleoside analogs (Loeb et al., 1999). These authors have coined the term “lethal mutagenesis” to describe the mutagen-induced loss of viral infectivity. Could increased mutagenesis be turned into an effective antiviral strategy (Loeb and Mullins, 2000)? At least two conditions should be fulfilled: the mutagenic drug should target viral RNAdependent RNA or DNA polymerases, but not host enzymes. Also, drug concentrations should be effective
enough so as to induce the irreversible transition into error catastrophe (violation of the error threshold) of viral replicons (Eigen and Biebricher, 1988; Schuster and Stadler, 1999; Domingo et al., 2000). These may seem at present harsh demands. However, this new line of research is of utmost importance in view of the continuing difficulties in developing effective vaccines for many viral diseases, the limited success of current antiviral therapies, and a strikingly sustained incidence of emerging and reemerging viral diseases. Both the concept of memory and the strategy of lethal mutagenesis have arisen as consequences of quasispe-
MINIREVIEW
253
ACKNOWLEDGMENTS I am indebted to John J. Holland for many years of collaboration and to C. M. Ruı´z-Jarabo, S. Sierra, A. Arias, E. Baranowski, and C. Escarmı´s for important contributions to the FMDV research reviewed here. This work was supported by Grants PM97-0060-01, FIS 98/0054-01, and FAIR5PL97-3665 and Fundacio´n Ramo´n Areces.
REFERENCES
FIG. 2. Error catastrophe and lethal mutagenesis. One of the implications of the quasispecies nature of RNA viruses is the existence of an error threshold relationship, which links the maximum meaningful genetic information with the superiority of the master (dominant) genomic sequence (over the mutant spectrum) and the copying fidelity during replication (Eigen and Biebricher, 1988; Schuster and Stadler, 1999; Domingo et al., 2000). In this simplified scheme, the error threshold (vertical line) marks a critical copying fidelity value which separates meaningful viral sequences (an organized quasispecies) from absence of information (random sequences or “no sequences”). Violation of the error threshold (arrow) means entry into error catastrophe; this process has been described as “melting” of genetic information. A new trend in antiviral research attempts to exploit violation of the error threshold as a means to force viral populations into extinction (Holland et al., 1990; Loeb et al., 1999; Loeb and Mullins, 2000). This is yet another important and promising outcome of the application of the quasispecies concept to virology.
cies dynamics, which open new avenues for understanding and controlling viruses and pose interesting new questions and challenges.
Domingo, E., Biebricher, C., Holland, J. J., and Eigen, M. (2000). “Quasispecies and RNA Virus Evolution: Principles and Consequences.” Landes Bioscience, Austin. Drake, J. W., and Holland, J. J. (1999). Mutation rates among RNA viruses. Proc. Natl. Acad. Sci. USA 96, 13910–13913. Eigen, M., and Biebricher, C. K. (1988). Sequence space and quasispecies distribution. In “RNA Genetics” (E. Domingo, P. Ahlquist, and J. J. Holland, Eds.), Vol. 3, pp. 211–245. CRC Press, Boca Raton. Escarmı´s, C., Da´vila, M., Charpentier, N., Bracho, A., Moya, A., and Domingo, E. (1996). Genetic lesions associated with Muller’s ratchet in an RNA virus. J. Mol. Biol. 264, 255–267. Escarmı´s, C., Da´vila, M., and Domingo, E. (1999). Multiple molecular pathways for fitness recovery of an RNA virus debilitated by operation of Muller’s ratchet. J. Mol. Biol 285, 495–505. Holland, J. J., Domingo, E., de la Torre, J. C., and Steinhauer, D. A. (1990). Mutation frequencies at defined single codon sites in vesicular stomatitis virus and poliovirus can be increased only slightly by chemical mutagenesis. J. Virol. 64, 3960–3962. Loeb, L. A., Essigmann, J. M., Kazazi, F., Zhang, J., Rose, K. D., and Mullins, J. I. (1999). Lethal mutagenesis of HIV with mutagenic nucleoside analogs. Proc. Natl. Acad. Sci. USA 96, 1492–1497. Loeb, L. A., and Mullins, J. I. (2000). Lethal mutagenesis of HIV by mutagenic ribonucleoside analogs. AIDS Res. Hum. Retroviruses 13, 1–3. Pathak, V. K., and Temin, H. M. (1992). 5-Azacytidine and RNA secondary structure increase the retrovirus mutation rate. J. Virol. 66, 3093– 3100. Ruiz-Jarabo, C. M., Arias, A., Barnowski, E., Escarmı´s, C., and Domingo, E. (2000). Memory in viral quasispecies. J. Virol. 74, 3543–3547. Schuster, P., and Stadler, P. F. (1999). Nature and evolution of early replicons. In “Origin and Evolution of Viruses” (E. Domingo, R. G. Webster, and J. J. Holland, Eds.), pp. 1–24. Academic Press, San Diego.