Monitoring Sequence Space as a Test for the Target of Selection in Viruses

doi:10.1016/j.jmb.2004.10.066

J. Mol. Biol. (2005) 345, 451–459

Monitoring Sequence Space as a Test for the Target of Selection in Viruses Celia Perales†, Vero´nica Martı´n†, Carmen M. Ruiz-Jarabo and Esteban Domingo* Centro de Biologı´a Molecular “Severo Ochoa” (CSIC-UAM) Universidad Auto´noma de Madrid, Cantoblanco, 28049 Madrid, Spain

An essential feature of viral quasispecies, predicted from quasispecies theory, is that the target of selection is the mutant distribution as a whole. To test molecularly the mutant composition selected from a viral quasispecies we reconstructed a mutant distribution using 19 antigenic variants of foot-and-mouth disease virus (FMDV). Each variant was marked by a specific amino acid replacement at a major antigenic site of the virus that conferred resistance to a monoclonal antibody (mAb). The variants were introduced in the mutant spectrum of a biological FMDV clone, at a frequency commonly found in FMDV quasispecies. The reconstructed quasispecies (and a number of control populations) were allowed to replicate in the presence or absence of the mAb. The mutant distribution that became dominant as a result of antibody selection included at least ten of the 19 mutants initially used to reconstruct the quasispecies. No such biased mutant repertoire was found in control populations. The results show that a mutant distribution was selected, and are incompatible with selection of an individual genome, which then generated multiple mutants upon further replication. An ample representation of variants immediately following a selection event should contribute to subsequent adaptability of the virus. q 2004 Elsevier Ltd. All rights reserved.

*Corresponding author

Keywords: viral quasispecies; target of selection; monoclonal antibodyescape; foot-and-mouth disease virus

Introduction RNA viruses replicate as complex mutant distributions termed quasispecies. 1–3 Quasispecies theory is a formulation of Darwinian evolution, with emphasis in mutation, that was developed to explain self-organisation and adaptability of RNA (or RNA-like) replicons, in the RNA world stage of the origin of life.1,4,5 In its initial formulation quasispecies theory involved steady-state mutant distributions of infinite size, in equilibrium. Later extensions of the theory to changing environments6,7 have reinforced quasispecies as an adequate † C.P. and V.M. have contributed equally to this work. Present address: C. M. Ruiz-Jarabo, Gladstone Institute of Virology and Immunology, 1001 Potrero Av., Bldg. 3, San Francisco, CA 9410, USA. Abbreviations used: FMDV, foot-and-mouth disease virus; mAb, monoclonal antibody. E-mail address of the corresponding author: [email protected]

theoretical framework to explain the behaviour of RNA viruses at the population level. Quasispecies dynamics is characterised by continuous generation of variant viral genomes, competition among them, and selection of the fittest mutant distributions in any given environment.2,3,8 Such a dynamics captures several features of the biology of RNA viruses such as: (i) rapid adaptability (via mutants with altered pathogenesis profiles or mutants resistant to inhibitors or to components of an immune response);9–11 (ii) suppressive and modulating effects of mutant spectra on replication and expression of subsets of genomes contained in them;12–16 and (iii) a history-dependent component in the form of memory genomes in the mutant spectra.17–22 In addition to these critical features (that stem from the fact that RNA viruses consist of mutant distributions rather than molecularly and functionally defined entities), it is noteworthy that quasispecies evolving in infected hosts constitute an initial stage in the process of genetic diversification of a virus. Mutant distributions are then

0022-2836/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.

452 subjected to positive and negative selection, and random drift (particularly in association with transmission events).3 A central issue in quasispecies theory that has not been tested experimentally in a direct fashion with RNA viruses, is that the target of selection is not an individual but the quasispecies as a whole.1,2,5–7 A difficulty for testing this essential feature experimentally is that if an individual genome is selected, a spectrum of mutants will be generated upon replication of this individual mutant, due to high mutation rates during RNA genome replication (average of 10K3 to 10K5 misincorporations per nucleotide copied or 0.1–10 mutations introduced into any RNA molecule synthesised from a template RNA23,24). To investigate the amplitude of the mutant distribution that is selected when a selective agent is applied, we have reconstructed a viral quasispecies with a collection of antigenic variants of the important animal virus foot-and-mouth disease virus (FMDV) (for review of this virus and the disease it causes, see Rowlands25 and Sobrino & Domingo26). Each of the antigenic variants used includes a single amino acid substitution within a major antigenic site of FMDV located at the G-H

Target Selection in Viruses

loop of capsid protein VP1.27,28 The mutants were generated in the course of serial passages of biological clone FMDV C-S8c129 in cytolytic infections of BHK-21 cells, as a result of a co-evolution of antigenicity and host cell tropism, and were isolated as monoclonal antibody (mAb) SD6-escape mutants, as described.11,30–33 The molecular basis of this co-evolution is an extensive overlap of the set of amino acid residues of the G-H loop of VP1 that belong to the antigenic site and the set of amino acid residues that are involved in the recognition of integrins, cell adhesion proteins that act as receptors for FMDV27,34,35 (reviewed by Jackson et al.36). In particular, a key element is an Arg-Gly-Asp (RGD) triplet and its neighbour Leu (L) (positions 141–144 of VP1 (Figure 1)) that are part of the epitope recognised by mAb SD6,37,38 and are essential for integrin recognition.36,39 Upon serial passage of FMDV C-S8c1 in BHK-21 cells, the residues involved in integrin recognition (in particular RGDL at positions 141–144) became dispensable as a result of acquisition by the virus of the capacity to use alternative receptors for cell entry.11,30–33,40 A consequence of this expansion of host cell tropism is a marked difference in the repertoire of

Figure 1. FMDV genome and repertoire of mAb SD6-escape mutants. Top: scheme of the FMDV C-S8c1 genome (8115 nucleotides, excluding homopolymeric tracts54); boxes indicate encoded proteins and lines indicate regulatory regions (not drawn to scale); the filled dot represents protein VPg covalently linked to the 5 0 end of the RNA and AAAA represents the 3 0 -terminal polyadenylate tract. Below the genome, the amino acid sequence of the G-H loop of VP1 is given; the epitope defined by mAb SD6 is underlined.37,38 The sequence RGDL is boxed. Below the mAb SD6 epitope sequence, repertoires of mAb SD6-escape mutants (each mutant arising from an independent mutational event) are shown.22,37 The parental FMDV populations used to derive the escape mutants are shown on the right. C-S8c1 is our reference FMDV clone29 and C-S8c1p100 is clone C-S8c1 passaged 100 times in BHK-21 cells as described.55 Memory populations refers to the repertoire of mAb SD6-escape mutants analysed to monitor decay of a memory genome encoding REDL instead of RGDL at VP1 positions 141–144.22 Subindices indicate the number of times that each amino acid substitution has been found in mAb SD6-escape mutants. Note the absence of replacements within the RGDL motif in the repertoire of FMDV C-S8c1 mAb SD6-escape mutants, due to the strict requirement of integrins as receptors for C-S8c1, but not for C-S8c1p100 or memory populations.11,30,31,40 The mutants used for the reconstruction of FMDV C-S8c1 (C19m) quasispecies are encircled in blue; (*) in the last row indicates a double mutant. The origin of viruses and experimental procedures are further detailed in the text.

453

Target Selection in Viruses

FMDV C-S8c1 mutants resistant to neutralisation by mAb SD6 found for the initial clone C-S8c1, and the progeny population passaged 100 times in BHK-21 cells (C-S8c1p100) (Figure 1). Indeed, none of a total of 83 mAb SD6-escape mutants of C-S8c1 included amino acid substitutions within the VP1 RGDL sequence whereas 26 out of 46 mAb SD6-escape mutants of C-S8c1p100 included an amino acid replacement within the RGDL (P!0.001; c2 test). Additional mutants within the G-H loop of VP1 were obtained among biological clones analysed in a study of quasispecies memory loss, monitored by the repertoire of mAb SD6-escape mutants20,22 (Figure 1). This ample collection of mutants has permitted the reconstruction of a quasispecies consisting of 4!106 plaque-forming-units (pfu) of FMDV C-S8c1 and about 10 pfu of each of 19 mAb SD6-escape mutants derived from memory populations (of the FMDV RED lineage at passages 15, 25, 45, 50 and 6022 as detailed in Materials and Methods); these populations are related to FMDV C-S8c1p100 (Figure 1). Therefore, the reconstructed quasispecies, termed C-S8c1 (C19m), contained the majority genomes and the corresponding mutant spectrum provided by biological clone C-S8c1, and additional components of the mutant spectrum, selectable by mAb SD6, present at a frequency of 2.5!10K6 each. Then this reconstructed population (or C-S8c1 alone or the 19 mutants alone, termed FMDV 19m) were allowed to replicate in BHK-21 cells in the presence or absence of mAb SD6. Comparison of the mutant spectra of the six progeny populations indicated that an ensemble of mutants was selected. This result underlines the behaviour of quasispecies as a whole during a selection process, shows that selection need not constitute a severe population bottleneck, and implies a higher adaptive potential when the

ensemble is confronted with subsequent selection events.

Results Mutant selection by mAb SD6 BHK-21 cells were infected either with FMDV C-S8c1, FMDV C-S8c1 (C19m), or with FMDV 19m, in the absence or presence of mAb SD6. Viral production and consensus nucleotide sequences of the FMDV genomic region encoding amino acid residues 138–147 of the G-H loop of VP1 were determined at different times post-infection (Figure 2 and Table 1). As expected, the wild-type (C-S8c1) consensus sequence was obtained at all times tested, following infection by FMDV C-S8c1 or FMDV C-S8c1 (C19m). In the infections by the 19 mAb SD6-escape mutants alone, a more complex evolution of the consensus sequence was observed. In this case, the dominance of virus encoding REDL at early times after infection in the absence of mAb SD6 may relate to a slower entry of this mutant in BHK-21 cells. In the presence of mAb SD6 the co-dominance of genomes encoding the consensus RGDL and those encoding RRDL may indicate some selective advantage of the latter mutant. Both observations have not been further investigated (see Discussion on the possible effects of relative fitness of different mAb SD6-escape mutants on the outcome of selection). Clonal analysis of progeny virus To sample the mutant composition of mAb SD6selected populations (and of their unselected counterparts), 12–79 biological clones from each

Figure 2. Kinetics of progeny production upon infection of BHK-21 cells with the indicated FMDV populations in the absence or the presence of mAb SD6. Conditions for cell culture, infections and titration of FMDV are detailed in Materials and Methods. Infections were carried out in parallel taking aliquots of the cell culture supernatant at the indicated time points, for titration of virus infectivity using several plates. Progeny of the infections by 19m at 0, 3 and 6 hours post-infection was not titrated because no RT-PCR amplifiable material was obtained at those times. Titrations were carried out in triplicate and standard deviations are given.

454

Target Selection in Viruses

Table 1. Consensus sequences of progeny FMDV FMDVa

MAb SD6b

C-S8c1

K C K C K

C-S8c1 (C19 m) 19m

C

a b c d e

Nucleotide sequence encoding SD6 epitoped (nucleotides 3619–3648)

Time post-infection (hours)c 0,3,6,9,12,24 0,3,6,9,12,24 0,3,6,9,12,24 0,3,6,9,12,24 0,3,6 9 12,24 0,3,6 9 12 24

GCCAGTGCACGCGGGGATTTGGCTCACCTA GCCAGTGCACGCGGGGATTTGGCTCACCTA GCCAGTGCACGCGGGGATTTGGCTCACCTA GCCAGTGCACGCGGGGATTTGGCTCACCTA GCCAGTGCACGCGAGGATTTGGCTCACCTA GCCAGTGCACGCGRGGATTTGGCTCACCTA GCCAGTGCACGCGGGGATTTGGCTCACCTA Not determined GCCAGTGCACGCRGGGATTTGGCTCACCTA GCCAGTGCACGCGGGGATTTGGCTCACCTA GCCAGTGCACGCRGGGATTTGGCTCACCTA

VP1 amino acids (141–144)e RGDL RGDL RGDL RGDL REDL RGDL/REDL RGDL – RGDL/RRDL RGDL RGDL/RRDL

The origin of the FMDV populations used is detailed in Introduction and Materials and Methods. Presence or absence of mAb SD6 during infection. Hours post-infection at which the nucleotide sequence encoding VP1 amino acids 138–147 was determined. Nucleotides underlined indicate a nucleotide change relative to the C-S8c1 sequence. RZACG. A bar indicates a mixed population.

virus population present at 12 and 24 hours postinfection were analysed (Table 2). Viral RNA extracted directly from each individual clone (viral plaque), without any biological passage, was subjected to RT-PCR amplification, and the cDNA sequenced, as described in Materials and Methods. In the viruses that were the progeny of either C-S8c1 or C-S8c1 (C19m) in the absence of mAb SD6, only clones with the wild-type mAb SD6 epitope sequence were obtained (Table 2). In contrast, in each population produced under selection by mAb SD6, a repertoire of mutants was obtained (Table 2). A subset of the initial mutant distribution was also isolated from the progeny of FMDV 19m. No clones with substitutions within the RGDL sequence of VP1 were isolated among the progeny of FMDV C-S8c1, in agreement with the fact that integrin recognition is essential for infection of BHK-21 cells by this virus, and no mutants with substitutions at positions 141–144 of VP1 can become dominant11,30,31,33 (Figure 1). Significantly, the repertoire of mutants selected by mAb SD6 during infection by C-S8c1 (C19m) closely matched the mutant spectrum of the initial, reconstructed quasispecies (Table 2). Out of 19 mutants included in population C-S8c1 (C19m), ten were represented in the clones analysed. To further ascertain that these selected mutants were derived from the initial 19 mutants used to reconstruct the quasispecies, rather than being selected as progeny mutants of C-S8c1 (that contributed the majority genomes of the reconstructed C-S8c1 (C19m) quasispecies), the nucleotide sequence around genomic position 3797 was determined. In C-S8c1 this position is A while in C-S8c1 p100 and related populations, this position is G. Out of 116 mutants selected by mAb SD6 from population C-S8c1 (C19m), 110 included a G at genomic position 3797, and this is a marker for FMDV population C-S8c1 p100, and all derived populations from which the 19 mutants used for quasispecies reconstruction were obtained (Figure 1); six mutants, I3 and R*3, affecting position 139, originated from C-S8c1, according to marker

position 3797 (Table 2). In the selected distribution, 95% of the genomes analysed were derived from the 19 mutants initially introduced, and 5% from the mutant spectrum contributed by C-S8c1 (Table 2). This analysis shows that minority components of the mutant spectrum participated actively in the selection process, and that selection involved an ample representation of such a spectrum.

Discussion A key feature of quasispecies dynamics is that the quasispecies as a whole, rather than an individual, is the target of selection.1–3,5,6,8 In agreement with this concept, the repertoire of mutants selected by mAb SD6 acting on replicating population C-S8c1 (C19m) could originate only from the selection of an ensemble of mutants present initially as minority components of the mutant spectrum (Table 2). The main arguments in support of this conclusion are: (i) Mutants of FMDV C-S8c1 with substitutions within the RGDL sequence of VP1 are not allowed to become dominant, as documented previously.11,30,31,33 This was directly confirmed here by analysing the mutant repertoire selected by mAb SD6 from population C-S8c1 in which no clones with amino acid replacements within the RGDL sequence were obtained (P!0.001; c2 test; data included in Table 2). (ii) The dominance of genomes with G-3797, a marker for memory populations (Figure 1), among the 116 mutants analysed from the progeny of C-S8c1 (C19m) in the presence of mAb SD6. (iii) An alternative interpretation to the selection of an ensemble of mutants would be that all different biological clones selected by mAb SD6 in replicating C-S8c1 (C19m) originated actually from a single mutant out of the 19 initially introduced, that this mutant reverted to the wild-type sequence, and that then this wild-type originated the observed repertoire. This is extremely unlikely since reversion of antigenic variants of FMDV either has not been observed, or

Table 2. Amino acid substitutions at VP1 residues 138–147 found in biological clones of progeny FMDV FMDV

C-S8c1

MAb SD6 Time post-infection (hours) Total clones analysed (wt)a Amino acidb A138 S139

K 12 70(70)

K 24 62(62)

C-S8c1 (C19m) C 12 12(0)

C 24 36(0)

D1 I1

D1 I1

R3 R*3 N1

R7 R*17 N6

K 12 52(50)

K 24 53(50)

Ic3 R6 R*3c

G142 D143

G1

L144

C 12 44(0)

S1 V1

V1

E8 R10 G3 N1 V3

19 m C 24 72(0)

K 12 42(2)

K 24 41(2)

I3 G20 R2

I3 G12 R2

N10

N6 T5

R1

R1

D2 R23

E3 R19 G9 N3 V10

V6

N1 V1 V3 Q1

H146

V2

C 12 33(0)

C 24 79(1)

D5 I6

D13 I16

R1

R10

N11

N11

E1 R1 G1 V3 S1 V1

R16 G2 N2 V3 S2 V1

R2

R2

P4 R3

R4

R1

R3 Q1

Q2

The indicated FMDV populations were used to infect BHK-21 cells in the absence or presence of mAb SD6. At the indicated times post-infection biological clones were obtained at random, viral RNA extracted, and the region encoding VP1 amino acids 138–147 sequenced. The origin of viral populations and infection conditions are described in the Introduction, Table 1 and Materials and Methods. a The number of clones with the wild-type C-S8c1 sequence is given in parenthesis. b VP1 amino acids of the epitope defined by mAb SD6 (sequence in Figure 1); amino acids in which no substitutions have been found are not listed; amino acids of motif RGDL are underlined. Subindices indicate the number of clones with a given amino acid substitution with respect to the residue given in the first column. Asterisks indicate that the S139/R replacement was due to mutation U3624A, which differs from the mutation U3624G that led to the same replacement in the other escape mutants and in those studied previously (Figure 1). Procedures for nucleotide sequencing are detailed in Materials and Methods. c These mutants originated from C-S8c1, as revealed by residue 3797 (see the text).

456 it occurred after several serial passages in the absence of antibody selection, under our conditions of infection of BHK-21 cells.20,41 This course of events would be even more unlikely for the six mutants (out of the 116 analysed) that derived from C-S8c1, since their generation from a single founder genome common to 116 mutants would necessitate implausible mutational events affecting genomic position 3797, and in three of them another mutation leading to a unique triplet encoding R at position 139 (Table 2). It must be indicated that replacement A3797G, which serves as a marker for memory populations of the FMDV RED lineage,20–22 is not required for selection of mAb SD6resistant mutants. This has been shown previously by selection of high fitness mAb SD6-escape mutants devoid of the A3797G replacement.42,43 In the present study, the six mAb-escape mutants derived from C-S8c1 (footnote c in Table 2) also lacked substitution A3797G. The demonstration that a mutant ensemble was selected by a highly specific selective constraint such as that mediated by a mAb is in agreement with the interpretation of natural selection as a “condensation phenomenon”, meaning “the condensation or localisation of a sequence distribution in a limited area of sequence space”.2 It adds to a number of observations that beg considering RNA virus populations as mutant distributions. The main observations are: (i) Suppressive effects of mutant spectra on phenotypic expression of individual mutants contained in them that show that the target of selection is not a single species.12–16 (ii) Quasispecies memory, a property of quasispecies distributions by which a subset of components of the mutant spectra reflects the genomes that were dominant at an earlier phase of evolution of the same viral lineage. Since memory is erased by population bottlenecks, that limit the portion of the mutant spectrum that participates in replication,20,21 memory is a property of the quasispecies as a whole (review by Domingo44). (iii) Mutations present in individual, minority components of a mutant spectrum can contribute to increase the fitness both of the entire population, and of the dominant genomes when the mutations are introduced in them.40 Thus, mutant spectra participate decisively in conforming the phenotype of the virus. Therefore, considering RNA virus populations as defined merely by a consensus genomic nucleotide sequence is a severe simplification of reality that may obscure the understanding of virus behaviour. That the target of selection is an ensemble of mutants has important implications for RNA virus adaptability. Indeed, when a selective pressure acts on a quasispecies distribution of genomes, an ensemble of genomes harbouring the selectable marker will originate (and their progeny replenish) a new quasispecies. This means that a mutant cloud, broader than the one that would have been generated from a single individual, will be formed as a result of selection.2 Broader clouds are more

Target Selection in Viruses

likely to find adaptive pathways, as evidenced in many viral systems by compensatory mutations that produce fitness gains of genomes either subjected to Muller ’s ratchet 45–49 or selected by inhibitors when selection entails a fitness cost.50 A possible alternative outcome of our experiment could have been the selection of one out of the 19 mutants present, either because a single genome rapidly could dominate the others, once the individual was selected (a type of random drift among selectable entities), or because of fitness differences among the components of the mutant spectra. Differences in relative fitness among the 19 mutants used to reconstruct quasispecies C-S8c1 (C19m) can be estimated as 2.7–4.0 fitness units, since fitness increases in 0.06–0.09 fitness units per passage18 (detailed in Materials and Methods). Among the mutants selected from C-S8c1 (C19m) by mAb SD6 (Table 2), four mutants originated from passage 45, five from passage 50, and one from passage 60. The absence of mutants from passages 15 and 25 (those with the lowest relative fitness in these series) suggests a modulating effect of fitness on the selectable repertoire. However, since mutants from passages 45 and 60 were selected, relative fitness differences of 0.9–1.3 fitness units were not determinant of selection. By using mutants with larger and carefully quantified fitness differences, our quasispecies reconstruction system will permit us to address to which extent fitness can modulate the repertoire of selected variants. These experiments are now in progress. Selection of mutant ensembles acquires additional significance following the demonstration that memory genomes in viral quasispecies19–22 can gain fitness at a similar pace to the ensemble of genomes to which they belong.18 Therefore, selective forces acting on viral quasispecies are expected to produce a broad repertoire of mutants reflecting present and past evolutionary history of the viral lineage, offering a rich substrate of genomes for subsequent selection events.

Materials and Methods Cells and viruses The origin of BHK-21 cells and FMDV C-S8c1 has been described.29 FMDV C-S8c1p100 was obtained after 100 serial cytolytic passages of C-S8c1 in BHK-21 cells at a multiplicity of infection (moi) of 2–4 pfu/cell. Procedures for the isolation of mutants resistant to neutralisation by mAb SD6 have been described.20,21,51 To monitor that no cross-contamination among cultures occurred, mockinfected cultures were treated in parallel and maintained throughout the experiments. No evidence of infection of control cultures was obtained. Likewise, none of the total of 629 nucleotide sequences determined throughout this study suggested any cross-contamination.

457

Target Selection in Viruses

RNA extraction, cDNA synthesis, PCR amplification and DNA sequencing RNA was extracted using Trizol, as described.45 Procedures and primers used for RT-PCR amplification and nucleotide sequencing have been described.20,21 Reverse transcription was carried out using avian myeloblastosis virus RT (Promega), and PCR amplification was performed using Expand High Fidelity polymerase (Roche) as specified by the manufacturers. Reconstruction of quasispecies A FMDV quasispecies (FMDV (C19m)) was reconstructed by mixing about 10 pfu from each of 19 mAb SD6-escape mutant viruses with 4!106 pfu from C-S8c1 as majority genomes. The 19 mutants used to reconstruct quasispecies C-S8c1 (C19m) originated from FMDV RED memory lineage at passages 15, 25, 45, 50 and 60,22 as follows: G142/E plus D143/G (double mutant), L147/P (from passage 15); H146/P (from passage 25); A138/D, S139 / I, S139/G, G142/E, D 143/V, D143/E, H146/Q (from passage 45); S139/N, G142/R, D143/G, D143/N, L144/V, L144/S, H146/R (from passage 50); and S139/R, S139/T (from passage 60). The locations of these substitutions in the epitope recognised by mAb SD6 are indicated in Figure 1. Relative fitness values for the individual mAb SD6escape mutants used to reconstruct the C-S8c1 (C19m) quasispecies have not been determined. However, the increase in relative fitness of clonal populations, derived also from FMDV C-S8c1 passaged under the same conditions as the FMDV RED lineage, is in the range of 0.06–0.09 fitness units per passage.18 Therefore, the 19 mutants used to reconstruct the quasispecies may vary by 2.7–4.0 fitness units. In parallel, C-S8c1 (4!106 pfu) and 19m (a total of about 200 pfu from 19 mAb SD6-resistant mutants) were used as controls. Viral infections in the presence or absence of mAb SD6 Infections of BHK-21 cell monolayers (3!106 cells infected with either 4!106 pfu of FMDV C-S8c1 or FMDV C-S8c1 (C19m), or 200 pfu of 19m) and plaque assays in semisolid agar medium were carried out as described.52 After virus adsorption for one hour at 37 8C, monolayers were washed once with 0.1 M phosphate buffer (pH 6.0) to inactivate unadsorbed virus, extensively with Dulbecco Modified Eagle’s Medium (DMEM), and further incubated in DMEM plus 1% (v/v) foetal calf serum. mAb SD6 (supernatant of SD6 hybridoma culture) was used at 1 : 5 dilution (a concentration that caused neutralisation of about 99% of wild-type virus and allowed survival of a broad repertoire of escape mutants). At different times after infection (0, 3, 6, 9, 12 and 24 hours), samples were taken for titration of infectivity. In the agar overlay a 1 : 20 dilution of mAb SD6 was added for titration of progeny produced in the presence of mAb SD6. Procedures used for infections with FMDV in liquid culture medium and in semisolid agar medium for plaque assays have been described.29,45,52,53

Acknowledgements We thank C. Escarmı´s and A. Arias for valuable

comments, and M. Da´vila for technical assistance. This work was supported by grants BMC 20011823-C02-01, CAM 08.2/0015/2001.1, PROFIT 2003 awarded to Genetrix S.L. (FIT 010000-2002-38) and by Fundacio´n Ramo´n Areces.

References 1. Eigen, M. & Schuster, P. (1979). The hypercycle. A principle of natural self-organization, Springer, Berlin. 2. Eigen, M. & Biebricher, C. K. (1988). Sequence space and quasispecies distribution. In RNA Genetics (Domingo, E., Ahlquist, P. & Holland, J. J., eds), Vol. 3, pp. 211–245, CRC Press, Boca Raton, FL. 3. Domingo, E., Biebricher, C., Eigen, M. & Holland, J. J. (2001). Quasispecies and RNA virus evolution: principles and consequences, Landes Bioscience, Austin. 4. Eigen, M., McCaskill, J. & Schuster, P. (1988). Molecular quasi-species. J. Phys. Chem. 92, 6881–6891. 5. Schuster, P. & Stadler, P. F. (1999). Nature and evolution of early replicons. In Origin and Evolution of Viruses (Domingo, E., Webster, R. G. & Holland, J. J., eds), pp. 1–24, Academic Press, San Diego. 6. Eigen, M. (2000). Natural selection: a phase transition? Biophys. Chem. 85, 101–123. 7. Wilke, C. O., Ronnewinkel, C. & Martinetz, T. (2001). Dynamic fitness landscapes in molecular evolution. Phys. Rep. 349, 395–446. 8. Domingo, E. (1999). Quasispecies. In Encyclopedia of Virology (Granoff, A. & Webster, R. G., eds), pp. 1431– 1436, Academic Press, London. 9. Duarte, E. A., Novella, I. S., Weaver, S. C., Domingo, E., Wain-Hobson, S., Clarke, D. K. et al. (1994). RNA virus quasispecies: significance for viral disease and epidemiology. Infect. Agents Dis. 3, 201–214. 10. Domingo, E. (2003). Quasispecies and the development of new antiviral strategies. Prog. Drug Res. 60, 133–158. 11. Baranowski, E., Ruı´z-Jarabo, C. M., Pariente, N., Verdaguer, N. & Domingo, E. (2003). Evolution of cell recognition by viruses: a source of biological novelty with medical implications. Advan. Virus Res. 62, 19–111. 12. de la Torre, J. C. & Holland, J. J. (1990). RNA virus quasispecies populations can suppress vastly superior mutant progeny. J. Virol. 64, 6278–6281. 13. Chumakov, K. M., Powers, L. B., Noonan, K. E., Roninson, I. B. & Levenbook, I. S. (1991). Correlation between amount of virus with altered nucleotide sequence and the monkey test for acceptability of oral poliovirus vaccine. Proc. Natl Acad. Sci. USA, 88, 199–203. 14. Borrego, B., Novella, I. S., Giralt, E., Andreu, D. & Domingo, E. (1993). Distinct repertoire of antigenic variants of foot-and-mouth disease virus in the presence or absence of immune selection. J. Virol. 67, 6071–6079. 15. Teng, M. N., Oldstone, M. B. & de la Torre, J. C. (1996). Suppression of lymphocytic choriomeningitis virus— induced growth hormone deficiency syndrome by disease-negative virus variants. Virology, 223, 113–119. 16. Gonza´lez-Lo´pez, C., Arias, A., Pariente, N., Go´mezMariano, G. & Domingo, E. (2004). Preextinction viral RNA can interfere with infectivity. J. Virol. 78, 3319–3324. 17. Arias, A., La´zaro, E., Escarmı´s, C. & Domingo, E. (2001). Molecular intermediates of fitness gain of an

458

18.

19.

20. 21.

22.

23.

24. 25. 26. 27.

28. 29.

30.

31.

32.

33. 34.

35.

RNA virus: characterization of a mutant spectrum by biological and molecular cloning. J. Gen. Virol. 82, 1049–1060. Arias, A., Ruiz-Jarabo, C. M., Escarmis, C. & Domingo, E. (2004). Fitness increase of memory genomes in a viral quasispecies. J. Mol. Biol. 339, 405–412. Briones, C., Domingo, E. & Molina-Parı´s, C. (2003). Memory in retroviral quasispecies: experimental evidence and theoretical model for human immunodeficiency virus. J. Mol. Biol. 331, 213–229. Ruı´z-Jarabo, C. M., Arias, A., Baranowski, E., Escarmı´s, C. & Domingo, E. (2000). Memory in viral quasispecies. J. Virol. 74, 3543–3547. Ruı´z-Jarabo, C. M., Arias, A., Molina-Parı´s, C., Briones, C., Baranowski, E., Escarmı´s, C. & Domingo, E. (2002). Duration and fitness dependence of quasispecies memory. J. Mol. Biol. 315, 285–296. Ruı´z-Jarabo, C. M., Miller, E., Go´mez-Mariano, G. & Domingo, E. (2003). Synchronous loss of quasispecies memory in parallel viral lineages: a deterministic feature of viral quasispecies. J. Mol. Biol. 333, 553–563. Batschelet, E., Domingo, E. & Weissmann, C. (1976). The proportion of revertant and mutant phage in a growing population, as a function of mutation and growth rate. Gene, 1, 27–32. Drake, J. W. & Holland, J. J. (1999). Mutation rates among RNA viruses. Proc. Natl Acad. Sci. USA, 96, 13910–13913. Rowlands, D. J., ed. (2003). Foot-and-mouth disease. Virus Res. 91, 1–161. Sobrino, F. & Domingo, E., eds (2004). Foot and Mouth Disease: Current Perspectives, Horizon Bioscience, Wymondham. Acharya, R., Fry, E., Stuart, D., Fox, G., Rowlands, D. & Brown, F. (1989). The three-dimensional structure of ˚ resolution. foot-and-mouth disease virus at 2.9 A Nature, 337, 709–716. Mateu, M. G. (1995). Antibody recognition of picornaviruses and escape from neutralization: a structural view. Virus Res. 38, 1–24. Sobrino, F., Da´vila, M., Ortı´n, J. & Domingo, E. (1983). Multiple genetic variants arise in the course of replication of foot-and-mouth disease virus in cell culture. Virology, 128, 310–318. Baranowski, E., Sevilla, N., Verdaguer, N., Ruı´zJarabo, C. M., Beck, E. & Domingo, E. (1998). Multiple virulence determinants of foot-and-mouth disease virus in cell culture. J. Virol. 72, 6362–6372. Baranowski, E., Ruı´z-Jarabo, C. M., Sevilla, N., Andreu, D., Beck, E. & Domingo, E. (2000). Cell recognition by foot-and-mouth disease virus that lacks the RGD integrin-binding motif: flexibility in aphthovirus receptor usage. J. Virol. 74, 1641–1647. Ruı´z-Jarabo, C. M., Pariente, N., Baranowski, E., Da´vila, M., Go´mez-Mariano, G. & Domingo, E. (2004). Expansion of host-cell tropism of foot-andmouth disease virus despite replication in a constant environment. J. Gen. Virol. 85, 2289–2297. Baranowski, E., Ruiz-Jarabo, C. M. & Domingo, E. (2001). Evolution of cell recognition by viruses. Science, 292, 1102–1105. Baxt, B., Morgan, D. O., Robertson, B. H. & Timpone, C. A. (1984). Epitopes on foot-and-mouth disease virus outer capsid protein VP1 involved in neutralization and cell attachment. J. Virol. 51, 298–305. Berinstein, A., Roivainen, M., Hovi, T., Mason, P. W. & Baxt, B. (1995). Antibodies to the vitronectin receptor

Target Selection in Viruses

36. 37.

38.

39.

40.

41.

42.

43.

44. 45.

46.

47.

48.

49.

50.

(integrin aV b3) inhibit binding and infection of footand-mouth disease virus to cultured cells. J. Virol. 69, 2664–2666. Jackson, T., King, A. M., Stuart, D. I. & Fry, E. (2003). Structure and receptor binding. Virus Res. 91, 33–46. Mateu, M. G., Martı´nez, M. A., Capucci, L., Andreu, D., Giralt, E., Sobrino, F. et al. (1990). A single amino acid substitution affects multiple overlapping epitopes in the major antigenic site of foot-and-mouth disease virus of serotype C. J. Gen. Virol. 71, 629–637. Verdaguer, N., Mateu, M. G., Andreu, D., Giralt, E., Domingo, E. & Fita, I. (1995). Structure of the major antigenic loop of foot-and-mouth disease virus complexed with a neutralizing antibody: direct involvement of the Arg-Gly-Asp motif in the interaction. EMBO J. 14, 1690–1696. Mateu, M. G., Valero, M. L., Andreu, D. & Domingo, E. (1996). Systematic replacement of amino acid residues within an Arg-Gly-Asp-containing loop of foot-and-mouth disease virus and effect on cell recognition. J. Biol. Chem. 271, 12814–12819. Baranowski, E., Ruı´z-Jarabo, C. M., Lim, P. & Domingo, E. (2001). Foot-and-mouth disease virus lacking the VP1 G-H loop: the mutant spectrum uncovers interactions among antigenic sites for fitness gain. Virology, 288, 192–202. Ruı´z-Jarabo, C. M., Sevilla, N., Da´vila, M., Go´mezMariano, G., Baranowski, E. & Domingo, E. (1999). Antigenic properties and population stability of a foot-and-mouth disease virus with an altered ArgGly-Asp receptor-recognition motif. J. Gen. Virol. 80, 1899–1909. Martı´nez, M. A., Carrillo, C., Gonza´lez-Candelas, F., Moya, A., Domingo, E. & Sobrino, F. (1991). Fitness alteration of foot-and-mouth disease virus mutants: measurement of adaptability of viral quasispecies. J. Virol. 65, 3954–3957. Holguı´n, A., Herna´ndez, J., Martı´nez, M. A., Mateu, M. G. & Domingo, E. (1997). Differential restrictions on antigenic variation among antigenic sites of footand-mouth disease virus in the absence of antibody selection. J. Gen. Virol. 78, 601–609. Domingo, E. (2000). Viruses at the edge of adaptation. Virology, 270, 251–253. Escarmı´s, C., Da´vila, M. & 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. Escarmı´s, C., Go´mez-Mariano, G., Da´vila, M., La´zaro, E. & Domingo, E. (2002). Resistance to extinction of low fitness virus subjected to plaque-to-plaque transfers: diversification by mutation clustering. J. Mol. Biol. 315, 647–661. La´zaro, E., Escarmı´s, C., Domingo, E. & Manrubia, S. C. (2002). Modeling viral genome fitness evolution associated with serial bottleneck events: evidence of stationary states of fitness. J. Virol. 76, 8675–8681. Lazaro, E., Escarmis, C., Perez-Mercader, J., Manrubia, S. C. & Domingo, E. (2003). Resistance of virus to extinction on bottleneck passages: study of a decaying and fluctuating pattern of fitness loss. Proc. Natl Acad. Sci. USA, 100, 10830–10835. Maisnier-Patin, S. & Andersson, D. I. (2004). Adaptation to the deleterious effects of antimicrobial drug resistance mutations by compensatory evolution. Res. Microbiol. 155, 360–369. Nijhuis, M., Schuurman, R., de Jong, D., Erickson, J., Gustchina, E., Albert, J., Schipper, P. et al. (1999).

Target Selection in Viruses

Increased fitness of drug resistant HIV-1 protease as a result of acquisition of compensatory mutations during suboptimal therapy. AIDS, 13, 2349–2359. 51. Martı´nez, M. A., Verdaguer, N., Mateu, M. G. & Domingo, E. (1997). Evolution subverting essentiality: dispensability of the cell attachment Arg-Gly-Asp motif in multiply passaged foot-and-mouth disease virus. Proc. Natl Acad. Sci. USA, 94, 6798–6802. 52. Domingo, E., Da´vila, M. & Ortı´n, J. (1980). Nucleotide sequence heterogeneity of the RNA from a natural population of foot-and-mouth-disease virus. Gene, 11, 333–346. 53. Escarmı´s, C., Da´vila, M., Charpentier, N., Bracho, A.,

459 Moya, A. & Domingo, E. (1996). Genetic lesions associated with Muller’s ratchet in an RNA virus. J. Mol. Biol. 264, 255–267. 54. Toja, M., Escarmis, C. & Domingo, E. (1999). Genomic nucleotide sequence of a foot-and-mouth disease virus clone and its persistent derivatives. Implications for the evolution of viral quasispecies during a persistent infection. Virus Res. 64, 161–171. 55. Charpentier, N., Da´vila, M., Domingo, E. & Escarmı´s, C. (1996). Long-term, large-population passage of aphthovirus can generate and amplify defective noninterfering particles deleted in the leader protease gene. Virology, 223, 10–18.

Edited by J. Karn (Received 22 September 2004; received in revised form 22 October 2004; accepted 22 October 2004)