Adenovirus Serotype Evolution Is Driven by Illegitimate Recombination in the Hypervariable Regions of the Hexon Protein

VIROLOGY

224, 357–367 (1996) 0543

ARTICLE NO.

Adenovirus Serotype Evolution Is Driven by Illegitimate Recombination in the Hypervariable Regions of the Hexon Protein LETA K. CRAWFORD-MIKSZA*,†,1 and DAVID P. SCHNURR* *Viral and Rickettsial Disease Laboratory, Division of Communicable Disease Control, California Department of Health Services, Berkeley, California 94704; and †School of Public Health, Program in Infectious Disease, University of California at Berkeley, Berkeley, California Received May 20, 1996; accepted August 7, 1996 The origin of AIDS-associated adenoviruses (AV 43–AV 49) was investigated by examining evolutionary relationships among 18 serologically related subgenus D serotypes and 3 intermediates and determining the mutation rate of a single serotype, AV 48, among clinical isolates from AIDS patients over a 6-year period. Nucleotide sequence of conserved and seven hypervariable regions (HVRs) of the hexon protein, the pVI core protein signal peptide, and noncoding region between the two genes was determined. Among AV 48 isolates the base misincorporation rate was 3.2 per 10,000 bases over 6 years. A 6-bp deletion occurred in one isolate between short direct repeats in HVR 7. Among subgenus D serotypes mutation rates were extremely low in the pVI peptide, the 5* hexon noncoding region, and first 187 bases of hexon protein. Small 2and 3-bp deletions between short direct repeats in a polypurine stretch in the noncoding region occurred in 3 strains. Mutation increased with proximity to the HVRs. Within HVR 1, 2, 4, 5, and 7 variability consisted of extensive intrachromosomal illegitimate recombination, including deletions between short direct repeats, insertions and duplications in repetitive polypurine stretches, and numerous base substitutions. All serotypes and intermediates differed by at least one illegitimate recombination event, with one exception. We conclude that AV serotype evolution is driven by illegitimate recombination events (antigenic shift), concommitant with single base mutation (antigenic drift), and that the HVRs are ‘‘hot spots’’ for both. These events could be explained by slippage-misalignment of the AV DNA polymerase in repetitive polypurine stretches during single-strand DNA replication. This mutability in the surface regions of the major viral coat protein confers a distinct survival advantage to this family of viruses. q 1996 Academic Press, Inc.

INTRODUCTION

acid homology, fiber protein characteristics, and biological properties (Horwitz, 1990). Within several subgenera low-level cross-reactivities occur in reproducible patterns. The serotypes involved have been designated ‘‘clusters’’ (Hierholzer et al., 1991). This clustering may represent a common heritage or progenitor for members of the group. The emergence of ‘‘new’’ viruses is the result of recognition of organisms that existed previously, but have only recently been detected, and those that have newly evolved by mutation from previously recognized entities. The mechanisms that produce these mutational changes have been extensively studied in RNA viruses (Domingo and Holland, 1988). Less is known about the evolution of DNA viruses. The same factors do not appear to be operative on DNA genomes. Viral DNA polymerases have greater fidelity due to proofreading capability (Tamanoi, 1986), and viral DNA genomes remain stable under a variety of conditions (Adrian et al., 1986). The appearance of multiple serotypes, or genotypes, among AVs and papilloma viruses argues that some evolutionary mechanism is operative on DNA viruses. Among human AVs subgenus D is the largest and fastest growing group, containing 31 of the 49 serotypes and 12 of the latest 14 to be characterized (AV 36 to AV 39,

The adenoviruses (AV) are a large family of viruses that infect mammals, marsupials, birds, and amphibians and share a common structure and genome organization, although infectivity is species-specific, even among primates (Wigand and Adrian, 1986). They are nonenveloped with an icosahedral capsid structure composed of three surface proteins: hexon, fiber, and penton. The linear double-stranded DNA genome of approximately 36 kbp is replicated unidirectionally by a viral alpha-type DNA polymerase that copies a single strand at a time (Challberg, 1989; Wang et al., 1989). The human AVs currently number 49 serotypes (Hierholzer et al., 1991; Schnurr and Dondero, 1993). Serotype designation is based on (i) neutralization of infectivity, which is type-specific and directed against epitopes on the hexon protein, and (ii) hemagglutination-inhibition, which is directed against epitopes on the fiber protein (Norrby, 1969; Willcox and Mautner, 1976). Serotypes are divided into 6 subgenera (A–F) on the basis of nucleic 1 To whom correspondence and reprint requests should be addressed at Viral and Rickettsial Disease Laboratory, California Dept. of Health Services, 2151 Berkeley Way, Berkeley CA 94704. Fax: (510) 540-3305. E-mail: [email protected].

0042-6822/96 $18.00

357

AID

VY 8172

/

6A1F$$$901

09-14-96 07:42:01

Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

viras

AP: Virology

358

CRAWFORD-MIKSZA AND SCHNURR

AV 42 to AV 49). The last 7 have been isolated almost exclusively from AIDS patients since the mid-1980’s (Hierholzer et al., 1988b; Schnurr and Dondero, 1993; Schnurr et al., 1995). In addition to new serotypes, intertypic strains, called intermediates, have been described in increasing numbers, particularly from AIDS patients (Hierholzer et al., 1988a; Hierholzer et al., 1988b; Wigand and Adrian, 1989; Hierholzer, 1991). It appears that the process of evolution among the human AVs is moving at the greatest rate within subgenus D and that this rate may have been affected by the AIDS epidemic. Homologous recombination has been proposed most consistently as the primary evolutionary mechanism for AVs and has been extensively studied in vitro (Williams and Ustacelebi, 1971; Takemori, 1972; Sambrook et al., 1975; Williams et al., 1975; Young and Silverstein, 1980; Boursnell and Mautner, 1981; Munz et al., 1983; Volkert and Young, 1983; Munz and Young, 1984). It was found to be restricted to closely related strains or serotypes within the same subgenus, in regions of greatest homology. We have recently determined that serotype-specificity of the hexon protein is located in seven hypervariable regions (HVRs) (CrawfordMiksza and Schnurr, 1996). The hypervariable nature of these regions does not provide the required sequence homology for homologous recombination to take place. Minimum sequence homology requirements for efficient homologous recombination have been estimated to be from 20 to 50 matched base pairs in bacteria and yeast (Ehrlich, 1989; Mezard et al., 1992) to more than 200 bp in mammalian cells (Liskay et al., 1987). Crossover fine-mapping studies in AVs have noted that recombination did not occur in the variable regions of the hexon protein (Williams et al., 1975; Boursnell and Mautner, 1981). Homologous recombination does not explain the evolution of novel antigenic specificities. Illegitimate recombination is a collection of intrachromosomal, extrachromosomal, and site-specific mutational events that result in the joining of two pieces of DNA with limited homology, usually involving short direct repeats at the breakpoint junctions of the joined pieces of DNA. These events result in deletions from which one of the direct repeats is deleted, as well as insertions, duplications, and translocations (Meuth, 1989). A number of viruses, including AVs, have been associated with illegitimate recombination events in bacteria and mammalian cells (Ruley and Fried, 1983; Bullock et al., 1985; Champoux and Bullock, 1988; Coffin, 1990; Goodrich and Duesberg, 1990; Junejo et al., 1991; Kumagai and Ikeda, 1991; Wang and Rogler, 1991; Stary and Sarasin, 1992; Schorr and Doerfler, 1993). A mechanism which involved slipped mispairing of the DNA polymerase during replication was first proposed for frameshift mutations in phage T4 by Streisinger et al. (1966), and has since been extended to include illegitimate recombination events in bacteria (Farabaugh et al., 1978; Allgood and Silhavy, 1988; Brunier et al., 1989) and mammalian cells

AID

VY 8172

/

6A1F$$$902

09-14-96 07:42:01

(Efstratidis et al., 1980; Krawczak and Cooper, 1991; Kimura et al., 1994). Most of these observations have noted the importance of the DNA context on the frequency of illegitimate recombination events in the creation of mutational ‘‘hot spots.’’ In this study we examined the possibility that subgenus D AVs may be mutating at an accelerated rate giving rise to the new serotypes that have appeared recently, and looked for evolutionary relationships among serologically related clusters that might identify the origins of these new serotypes, and elucidate the mechanism by which they arose. We examined the mutation rate for AIDS-associated serotype AV 48, from isolates collected from AIDS patients over a 6-year period. In addition, we investigated the genetic relatedness among 18 subgenus D AVs and 3 intermediates in conserved regions and in the serotype-specific HVRs of the hexon protein, in the conserved carboxy terminus of the pVI core protein and in the noncoding region between the two genes. We found that the base substitution rate for AV 48 DNA polymerase was very low. Variation among the subgenus D serotypes and intermediates was the result of multiple illegitimate recombination events in the HVRs, as well as single-base substitution. The extent of single base variation in the HVRs between serotypes was profound, and given the low rate of base substitution of the polymerase, of ancient origin. MATERIALS AND METHODS DNA template preparation Prototype serotype strains were from the collection of the Viral and Rickettsial Disease Laboratory, California Department of Health Services (Berkeley, CA). Clinical patient isolates were sent directly to the VRDL for identification by local hospitals or were forwarded by county public health departments. These included clinical isolates of AV 48, collected from 11 different AIDS patients in the San Francisco Bay Area from 1985 to 1990 (L. Crawford-Miksza and D. P. Schnurr, in press), two intermediates for which no serotype designation was assignable but demonstrated low-level neutralization with prototype antisera, and one ‘‘intermediate’’ which was unneutralizable. Serologic clusters included were: AV 8 and AV 9 AV 10, AV 19 and AV 37 AV 26, AV 36, AV 38, AV 39 and AV 43 AV 17, AV 45 and intermediate J-87 AV 47 and AV 48 AV 30, AV 49 and intermediate R-83 AV 44 which has a one-way cross with AV 48 AV 46 9 9 9 9 9 9 9 AV 49 Intermediate V-92, not neutralized by prototype antisera

viras

AP: Virology

ADENOVIRUS EVOLUTION BY ILLEGIMATE RECOMBINATION

(Kemp et al., 1983; Hierholzer et al., 1991; Schnurr and Dondero, 1993; Crawford-Miksza and Schnurr, 1994). Full-length AV template was prepared by Hirt extraction of infected A549 cells (Hirt, 1967). PCR and direct DNA sequencing PCR and direct sequencing was performed as described (Crawford-Miksza and Schnurr, 1996). Overlapping PCR products were generated using consensus primers with deoxyinosine in positions of ambiguity and then sequenced in both directions using direct cycle sequencing with internal and template primers. Primers that framed the carboxy terminus of the pVI core protein, the hexon 5* noncoding region, and the first 187 bases of the hexon protein were (AV 48 codon numbering): UP, 59-AACAGCATIGTGGGTITGGGIGTG-3* (pVI) and 1L, 5*TCCAGCACGCCGCGGATGTCAAAGTA-3* (codons 106 to 98); primers for HVR 1 to 6 were: 1R, 5*-TACTTTGACATCCGCGGCGTGCTGGA-3* (98 to 106) and 3L, 5*-CTGTCIACIGCCTGITTCCACAT-3* (388 to 381); for HVR 4 and 5: 2Rm, 5*-CCITGCTATGGITCITTTGC-3* (220 to 226) and 3L; for HVR 7: 3R, 5*-ATGTGGAAICAGGCIGTIGACAG-3* (381 to 388) and 5L, 5*-CGGTGGTGITTIAAIGGITTIACITTGTCCAT-3* (533–523); AV 48-specific HVR 2 primer: 1.5R, 5*-CAAATTGGAATTGATGCAACCAAA-3* (191 to 198). Standard PCR conditions consisted of 100 pmol of each primer, 50 mM KCL, 10 mM Tris, 1.5 mM MgCl2 , 0.01% gelatin, 2.5 U of Tfl heat-stable DNA polymerase (Epicenter Technologies, Madison, WI), 250 mM each deoxynucleotide triphosphate, 6% glycerol, 1 mg singlestranded DNA binding protein (Bind-Aid, Amersham Life Science, Arlington Heights, IL), and 1 to 5 mg of DNA template. PCR cycle consisted of initial denaturation at 947 for 2 min, followed by 30 cycles of 947 for 1 min, 457 for 1 min, and 727 for 2 min. PCR products were isolated

359

directly or from agarose gels by adsorption to silica (Prep-A-Gene, Bio-Rad Laboratories, Hercules, CA). Cycle sequencing by direct incorporation of [33P]dATP was carried out with the ‘‘fmol’’ DNA Sequencing System according to the product manual (Promega, Madison, WI), using 10 pmol of primer and 1 to 5 ml of PCR product. PCR cycle for the sequencing was an initial denaturation at 947 for 2 min, followed by 40 cycles of 947 for 30 sec, annealing for 30 sec, and extension at 727 for 1 min. Annealing temperatures were set at 27 below the calculated Tm for each primer. Gene sequences were aligned and analyzed with Align Plus (Scientific and Educational Software, State Line, PA).

RESULTS AV 48 mutation rate A total of 1404 bases per AV 48 isolate were sequenced from 11 clinical isolates, for a total of 15,444 bases (Fig. 1). Five single base mutations (SBMs) were observed, four of which resulted in coding changes, with no specific pattern of substitution. Two were in first position of the codon, two in the second, and one in the wobble position. Three of the coding changes occurred in HVRs: two in HVR 1 and one in HVR 7. The one coding change that occurred in a conserved region between HVR 2 and HVR 3 was a nonconservative substitution of lysine for isoleucine (ATA ú AAA). There was one A ú T and one T ú A transversion, two G ú A and one A ú G transitions. In addition, there was a six-base deletion in HVR 7 between two short direct repeats of two bases in which one repeat was deleted (Fig. 2a). Base substitution frequency was 3.2 per 10,000 bases. Deletion frequency was 1 per 15,000 bases.

FIG. 1. Gene map of the regions surrounding the AV hexon protein showing structural loops, regions sequenced, and primers. HVRs 1, 2, 4, 5, and 7 included in this analysis are darkly shaded. HVRs 3 and 6, which do not contribute to serotype specificity within subgenus D are lightly shaded. The conserved pVI localization peptide is stipled.

AID

VY 8172

/

6A1F$$$902

09-14-96 07:42:01

viras

AP: Virology

360

CRAWFORD-MIKSZA AND SCHNURR

FIG. 2. Illegitimate recombination events in the noncoding region and HVRs of subgenus D AVs. Short direct repeats involved in deletion junctions are shaded. Single base mutations, duplications, insertions, and palindromes are labeled and indicated in bold face.

Comparison of subgenus D serotypes and intermediates In the conserved upstream 280 bases that included the pVI core protein nuclear localization signal peptide, the hexon 5* noncoding region, and the first 187 bases of the hexon protein, there were twelve SBMs among

AID

VY 8172

/

6A1F$$$902

09-14-96 07:42:01

the 18 serotypes and 3 intermediates, and four small deletions in three strains. A mutation that occurred in more than one serotype was scored as a single event. All four deletions and four of the SBMs occurred in a repetitive stretch of 18 polypurines in the noncoding region (Fig. 2b). Four SBMs in conserved coding regions were silent, and two coded for conservative

viras

AP: Virology

ADENOVIRUS EVOLUTION BY ILLEGIMATE RECOMBINATION

changes, a serine for threonine substitution (ACG ú TCG), and a phenylalanine for tyrosine (TAT ú TTT). Two were nonconservative, a threonine for alanine substitution (GCG ú ACG) and a serine for proline (CCG ú TCG). There were two A ú T, two G ú T and one G ú C transversions, and two A ú G, two G ú A, and three C ú T transitions. SBM increased with proximity to the HVRs, and was almost exclusively silent. Conservative substitutions occurred in regions between the HVRs, valine for isoleucine, tyrosine for phenylalanine, aspartic acid for glutamic acid, and lysine for arginine. Within the HVRs a complex pattern of illegitimate recombination involving deletion, duplication, and insertion was evident (Figs. 2c–2j). All recombination events preserved the reading frame and involved purine bases almost exclusively, in runs of alternating guanines and adenines or multiple adenines. Short direct repeats of 1 to 3 bases were visible at the breakpoints of the junctions, with deletion of one of the repeats. In one case the deletion of a palindromic sequence was evident (Fig. 2e). Duplication of an entire codon was common (Figs. 2c, 2f, and 2i). Insertions of 3 or 6 bases were observed containing little homology with adjacent sequence (Figs. 2h and 2j). SBM among the matrix of conserved hydrophobic residues within the HVRs was extensive, and contained multiple silent mutations, utilizing as many as five different codons for leucine residues, and all possible codons for glycine, isoleucine, valine, proline, and threonine. In order to evaluate whether the HVRs had higher intrinsic

361

rates of SBM, neutral silent mutations in conserved and HVRs were compared. Figure 3 demonstrates the extent of silent SBM in the HVRs as opposed to the highly conserved first 62 codons of the hexon protein. Comparison of codon usage for the 14 conserved glycine, proline, and valine residues in the HVRs with 11 such residues in the conserved region reveals a greatly increased rate of silent SBM in the HVRs. Base substitution rate in the conserved region was 1.5 per 1000 bases; in the HVRs it was 44 per 1000 bases, a 30-fold increase in the rate of substitution. In order to establish a geneology for the AIDS-associated AVs the conserved residues of the HVRs were aligned and the DNA sequences compared. HVR 1 was the longest, varying from 24 to 31 codons (72 to 93 bp) between serotypes (Fig. 4). HVR 1 appeared to have two distinct lineages based on the alignment of conserved matrix residues: glycine, lysine, asparagine, histidine, and valine. This alignment revealed at least 15 different illegitimate recombination events among 16 strains represented. All of the serotypes and intermediates differed from each other by at least one illegitimate recombination event in HVR 1, except AV 30 and AV 37. HVR 5 also appeared to have two distinct lineages, based on conservation of glycine, asparagine, proline, alanine, and isoleucine residues (Fig. 5). It also contained the greatest variability in length, differing by as much as 11 codons (33 bp) between serotypes. Among 20 strains analyzed there were at least 15 illegitimate recombination events represented.

FIG. 3. Silent single base mutation in HVRs versus conserved structural regions among 20 serotypes and intermediates. Codon usage variability was compared for 11 conserved hydrophobic glycine, proline, and valine residues in the conserved first 62 codons of the hexon and 14 conserved residues in the HVRs. SBMs resulting in coding changes are indicated but not scored.

AID

VY 8172

/

6A1F$$$902

09-14-96 07:42:01

viras

AP: Virology

362

CRAWFORD-MIKSZA AND SCHNURR

FIG. 4. Alignment of the first twenty-one codon positions of HVR 1. Two distinct lineages are indicated (the top eight sequences, the bottom eight sequences) based on the location and type of conserved framework and matrix residues (shaded). Serotype variation in length in this region was from 13 to 21 codons.

HVR 7 contained the fewest number of recombination events, with at least 10 events represented among 19 strains (Fig. 6). The length of HVR 7 varied from 7 to 14 codons. Comparison of serologically clustered serotypes revealed that serologic relatedness did not equate with genetic relatedness (Table 1). There were three genetic clusters discernible. AV 39 and AV 43 were very closely related, differing only by the deletion of two codons in HVR 1 of AV 39 and 41 SBMs over the 1400 bp sequenced. Twenty-one of these were silent, mostly in conserved regions between the HVRs. Of the 20 that produced coding changes, 17 were in HVRs. AV 36 was very similar to AV 39 and AV 43 in HVRs 4 and 5, but diverged in HVRs 1, 2, and 7 by multiple illegitimate recombination events. AV 47 and AV 48 differed only slightly in HVRs 2, 4, and 5, but by at least 3 recombination events in HVRs 1 and 7. Establishment of an unequivocal genetic relationship was possible with AV 30 and AV 37, which do not share serologic cross-reactivity. The structure of all HVRs indicates that they have evolved from the same recombination events, and variability in all HVRs was entirely from SBM. Among the 1400 bp sequenced there were 55 SBMs, 27 of which were in HVRs, and 22 in the framework regions between HVRs. In the

AID

VY 8172

/

6A1F$$$903

09-14-96 07:42:01

HVRs, 22 of the 27 resulted in coding changes, in the framework region 17 of the 22 were silent. The remaining 6 SBMs, 4 of which were silent, were randomly distributed over the conserved regions. Several clustered pairs of viruses shared homology in one or two HVRs, but differed by complex patterns of deletion or insertion in others. AV 8 and AV 9 were identical in HVR 2, but completely different in all other HVRs. Other serologically related pairs showed no genetic relatedness. HVRs 3 and 6 were formerly defined in comparisons of AVs from other subgenera and animal species (Crawford-Miksza and Schnurr, 1996), but did not contribute to serotype specificity within subgenus D. Variability was confined to SBMs in three and four codons, respectively, which were shared by multiple serotypes. DISCUSSION The observed error rates of 3.2 substitutions per 10,000 bases and 1 deletion per 15,000 bases in AV 48 over a 6-year period is extremely low. In a series of studies on eukaryotic alpha DNA polymerases, Kunkel and associates, using an M13 vector system, found that in a single cycle of replication, base misincorporation frequency for alpha polymerases averaged 1 per 4000

viras

AP: Virology

ADENOVIRUS EVOLUTION BY ILLEGIMATE RECOMBINATION

363

FIG. 5. Alignment of HVR 5. Two lineages are indicated (the top 12 sequences, the bottom eight sequences). Conserved framework and matrix residues are shaded. Variation in length was from 6 to 17 codons.

bases (Kunkel and Alexander, 1986), and deletion or insertion frameshift errors occurred at 1 error per 10,000– 30,000 bases polymerized (Kunkel, 1986). A significant portion of base substitutions, as well as frameshift mutations, were proposed to result from transient misalignment of template and daughter strand (Kunkel and Soni, 1988). Frameshift mutations occurred most frequently in or near a run of repeated bases with the insertion or deletion of a single base (Kunkel, 1986). For AVs, any loss of reading frame in the hexon protein would be a lethal mutation, so only those mutational events that preserve the reading frame were fixed and observable. Those that occurred in the noncoding region were not so constrained. It is difficult to directly compare in vitro data from a single round of replication with data from viruses circulating in the human population representing numerous replication cycles. However, the fact that the observed numbers and kinds of mutations for AV 48 were in the same ranges as those of other alpha-type DNA polymerases confirms the fidelity of the AV 48 polymerase and rules out our premise that the AIDS-associated AVs might have an accelerated mutation rate. We also found no support for our second hypothesis, that the AIDS-associated AVs arose recently and their

AID

VY 8172

/

6A1F$$$903

09-14-96 07:42:01

appearance was affected by the AIDS epidemic, with the exception of AV 43. AV 43 differs from AV 39 by only one deletion and 41 single base mutations among the 1400 bases sequenced. Given a molecular clock of 5 SBMs and 1 deletion in 6 years among AV 48 isolates, the divergence of these two serotypes was probably within the last 10 to 20 years. The divergences of AV 47 from AV 48 and AV 30 from AV 37 predate that, but are also of relatively recent origin. All other subgenus D serotypes, as well as all the intermediates, differed profoundly from each other, divergence being of ancient rather than recent origin given the slow rate of AV mutation. Phylogenetic analysis of human papillomaviruses (HPVs) has revealed that divergence among HPVs may have been coevolutionary with their host species, and consequently of very ancient origin (Chan et al., 1992; Ho et al., 1993; Ong et al., 1993). The apparent recent association of the higher numbered AV serotypes with concurrent HIV infection may have other biological factors, such as disease manifestation, susceptibility, and transmission (L. Crawford-Miksza and D. P. Schnurr, in press), rather than an evolutionary relationship. The neutralization epitopes of AVs are complex and conformational and are composed of serotype-specific

viras

AP: Virology

364

CRAWFORD-MIKSZA AND SCHNURR

FIG. 6. Alignment of HVR 7. Conserved framework residues are shaded. Variation in length was from 7 to 14 codons.

residues in the HVRs, although the number and locations of these residues are unknown (Crawford-Miksza and Schnurr, 1996). Based on the crossreactivity data (Table 1), five HVRs (1, 2, 4, 5, and 7) may contain parts of the neutralization epitopes, since serologically related strains shared homology in each of these regions. Several serologically unrelated serotypes contained homology in HVR 7, making its contribution to the neutralization epitopes less certain. Multiple illegitimate recombination events in the HVRs, as well as extensive SBM, have given rise to 49 unique serotypes identified to date. The ultimate number of unique serotypes is controlled by the number, location, and biologically acceptable variation of residues that make up the neutralization epitopes. Based on the complexity and mutability of the regions responsible for serotype specificity this may ultimately be a very large number. The lack of genetic relatedness among serotypes and intermediates that are serologically related which we observed for AVs has also been demonstrated for Hepatitis B strains (Ohba et al., 1995). Phylogenetic analysis of conserved regions of the AV genome has confirmed subgenus classification previously defined by biological parameters. Serotypes within the same subgenus clustered in all of the regions analyzed (Bailey and Mautner, 1994). Within a given subgenus, however, the establishment of relationships be-

AID

VY 8172

/

6A1F$$$903

09-14-96 07:42:01

tween serotypes or strains may not be as clear. In each of the individual HVRs, lineages arising from the same illegitimate recombination event were discernible for these subgenus D AVs (Figs. 4, 5, and 6). However, the clustering was not the same from one HVR to the next. Homologous recombination within a subgenus has been repeatedly demonstrated in vitro (Williams et al., 1975; Boursnell and Mautner, 1981; Munz et al., 1983), and sufficient homology exists between HVRs 3 and 4, and between HVRs 6 and 7, to make this possible. Relationship between any two given serotypes may therefore be discernible in only limited regions of the genome. The slippage misalignment model for deletion formation involves forward slippage of the polymerase during DNA replication between two regions with short microhomologies, causing looping-out of the parental strand, and deletion of one of the repeats and the intervening sequence in the daughter strand (Efstratiadis et al., 1980; Allgood and Silhavy, 1989; Krawczak and Cooper, 1991). A backward slippage of the polymerase would account for duplications of adjacent bases. It is more difficult to explain the insertion of bases without adjacent homology, although this has been observed elsewhere (Jego et al., 1993). It seems clear that the HVRs are contextual ‘‘hotspots’’ for both illegitimate recombination and an intrinsically higher rate of SBM. Mutation rate for eucary-

viras

AP: Virology

ADENOVIRUS EVOLUTION BY ILLEGIMATE RECOMBINATION TABLE 1 Serological Relationships versus Sequence Relationships among Subgenus D Serotypes and Intermediates Regions of identitya HVR: Serologically related Genetic relatedness Closely related AV 39, 43 AV 47, 48 Distantly related AV 36, 39 AV 36, 38 AV 8, 9 AV 10, 19 AV 44, 48 AV 17, J-89 AV 17, 45 Unrelated AV 30, 49 AV 49, R-83 AV 19, 37 AV 26, 36 AV 46, 49 Serologically unrelated Genetic relatedness Closely related AV 30, 37 Distantly related AV 44, V-92 AV 10/19, 45 AV 9, 10, 17, 19, 26, 45

1

2

4

5

7

/ /

/ /

/ /

/ Å

/ Å

/

/ Å

Å

/

/

Å

/ Å Å

Å / Å

/

/

/

/

6A1F$$$904

The authors thank Dr. Loy Volkman and Dr. Gertrude Buehring for critical reading of the manuscript and Dr. Carl Hanson, Dr. Mike Botchan, and Dr. James Hardy for helpful discussions.

REFERENCES

otic DNA polymerases has been shown to be inversely related to processivity (Kunkel, 1985). Pausing of the Escherichia coli DNA polymerase I stimulated illegitimate recombination (Papanicolaou and Ripley, 1991). Nearly all of the illegitimate recombination events we observed, as well as many of the SBMs, were found in repetitive polypurine stretches. Purine-rich repetitive regions have previously been described as hotspots of genetic instability (Knapp et al., 1994). Sections of DNA in which one strand is polypurine and the other strand polypyrimidine show a characteristic distortion of the helix (Watson et al., 1987). These polypurine regions may be a sequence-specific determinant of mutation frequency in the HVRs of the hexon protein. There is evidence that several organisms possess endogenous mechanisms of mutation, including illegitimate recombination, that confer survival advantage (Cairns et al., 1988; Dimpfl and Echols, 1989; Hall, 1990; Haggard-Ljungquist et al., 1992). A recent large-scale search among sequence databases for genes on which positive selection may operate turned up 17 families of genes, 9 of which

VY 8172

were the surface antigens of parasites and viruses (Endo et al., 1996). This appears to be true of AVs as well. It seems clear that the AVs have evolved an adaptive mechanism that confers a mutability in regions that allow escape from the host immune system and insure their own long-term survival. ACKNOWLEDGMENTS

a /, sharing 75% or greater homology in a given region. Å, sharing 50 to 74% homology.

AID

365

09-14-96 07:42:01

Adrian, Th., Wadell, G., Hierholzer, J. C., and Wigand, R. (1986). DNA restriction analysis of adenovirus prototypes 1 to 41. Arch. Virol. 91, 277–290. Allgood, N. D., and Silhavy, T. J. (1988). Illegitimate Recombination in Bacteria. In ‘‘Genetic Recombination’’ (R. Kucherlapati and G. R. Smith, Eds.) Am. Soc. Microbiol., Washington, DC. Bailey, A., and Mautner, V. (1994). Phylogenetic Relationships among adenovirus serotypes. Virology 205, 438–452. Boursnell, M. E. G., and Mautner, V. (1981). Recombination in adenovirus: Crossover sites in intertypic recombinants are located in regions of homology. Virology 112, 198–209. Bullock, P., Champoux, J. J., and Botchan, M. (1985). Association of crossover points with topoisomerase I cleavage sites: a model for nonhomologous recombination. Science 230, 954–958. Brunier, D., Peeters, B. P. H., Bron, S., and Ehrlich, S. D. (1989). Breakage-reunuion and copy choice mechanisms of recombination between short homologous sequences. EMBO J. 8, 3127–3133. Cairns, J., Overbaugh, J., and Miller, S. (1988). The origin of mutants. Nature 335, 142–145. Challberg, M. D. (1989). Animal virus DNA replication. Annu. Rev. Biochem. 58, 671–717. Champoux, J. J., and Bullock, P. A. (1988). Possible role for the eucaryotic type I topoisomerase in illegitimate recombination. In ‘‘Genetic Recombination’’ (R. Kucherlapati and G. R. Smith, Eds.) Am. Soc. Microbiol., Washington, DC. Chan, S.-Y., Bernard, H.-U., Ong, C.-K., Chan, S.-P., Hofmann, B., and Delius, H. (1992). Phylogenetic analysis of 48 papillomavirus types and 28 subtypes and variants: A showcase for the molecular evolution of DNA viruses. J. Virol. 66, 5714–5725. Coffin, J. M. (1990). Molecular mechanisms of nucleic acid integration. J. Med. Virol. 31, 43–49. Crawford-Miksza, L., and Schnurr, D. P. (1994). A quantitative spectrophotometric microneutralization assay for the characterization of adenoviruses. J. Clin. Microbiol. 32, 2231–2234. Crawford-Miksza, L., and Schnurr, D. P. (1996). Analysis of 15 adenovirus hexon proteins reveals the location and structure of seven hypervariable regions containing serotype-specific residues. J. Virol. 70, 1836–1844. Crawford-Miksza, L., and Schnurr, D. P. Seroepidemiology of AIDSassociated adenoviruses among the San Francisco Men’s Health Study. J. Med. Virol., in press. Dimpfl, J., and Echols, H. (1989). Duplication mutagenesis as an SOS response in E. coli: Enhanced duplication formation by a constitutively-activated RecA. Genetics 123, 255–260. Domingo, E., and Holland, J. J. (1988). High error rates, population equilibrium, and evolution of RNA replication systems. In ‘‘RNA Genetics’’ (E. Domingo, J. J. Holland, and P. Ahlquist, Eds.), Vol. III. CRC Press, Boca Raton, FL. Efstratiadis, A., Posakony, J. W., Maniatis, T., Lawn, R. M., O’Connell,

viras

AP: Virology

366

CRAWFORD-MIKSZA AND SCHNURR

C., Spritz, R. A., DeRiel, J. K., Forget, B. G., Weissman, S. M., Slightom, J. L., Blechl, A. E., Baralle, F. E., Shoulders, C. C., and Proudfoot, N. J. (1980). The structure and evolution of the human beta-globin gene family. Cell 21, 653–668. Ehrlich, S. D. (1989). Illegitimate recombination in bacteria. In ‘‘Mobile DNA’’ (D. E. Berg and M. M. Howe, Eds.) Am. Soc. Microbiol., Washington, DC. Endo, T., Ikeo, K., and Gojobori, T. (1996). Large-scale search for genes on which positive selection may operate. Mol. Biol. Evol. 13, 685– 690. Farabaugh, P. J., Schmeissner, U., Hofer, M., and Miller, J. H. (1987). Genetic studies on the lac repressor VII. On the molecular nature of spontaneous hotspots in the lacI gene of Escherichia coli. J. Mol. Biol. 126, 847–863. Goodrich, D. W., and Duesberg, P. H. (1990). Evidence that retroviral transduction is mediated by DNA, not by RNA. Proc. Natl. Acad. Sci. USA 87, 3604–3608. Haggard-Ljungquist, E., Halling, C., and Calendar, R. (1992). DNA sequences of the tail fiber genes of bacteriophage P2: evidence for horizontal transfer of tail fiber genes among unrelated bacteriophages. J. Bacteriol. 174, 1462–1477. Hall, B. G. (1990). Spontaneous point mutations that occur more often when advantageous than when neutral. Genetics 126, 5–16. Hierholzer, J. C., Adrian, Th., Anderson, L. J., Wigand, R., and Gold, J. W. M. (1988a). Analysis of antigenically intermediate strains of subgenus B and D—adenoviruses from AIDS patients. Arch. Virol. 103, 99–115. Hierholzer, J. C., Wigand, R., Anderson, L. J., Adrian, Th., and Gold, J. W. M. (1988b). Adenoviruses from patients with AIDS: A plethora of serotypes and a description of five new serotypes of subgenus D (types 43–47). J. Infect. Dis. 158, 804–813. Hierholzer, J. C., Stone, Y. O., and Broderson, J. R. (1991). Antigenic relationships between the 47 human adenoviruses determined in reference horse antisera. Arch. Virol. 121, 179–197. Hirt, B. (1967). Selective extraction of polyoma DNA from infected mouse cell cultures. J. Mol. Biol. 26, 365–369. Ho, L., Chan, S.-Y., Burk, R. D., Das, B. C., Fujinaga, K., Icenogle, J. P., Kahn, T., Kiviat, N., Lancaster, W., Mavromara-Nazos, P., Labropoulou, V., Mitrani-Rosenbaum, S., Norrild, B., Pillai, M. R., Stoerker, J., Syrjaenen, K., Syrjaenen, S., Tay, S.-K., Villa, L. L., Wheeler, C. M., Williamson, A. L., Bernard, H.-U. (1993). The genetic drift of human papillomavirus type 16 is a means of reconstructing prehistoric viral spread and the movement of ancient human populations. J. Virol. 67, 6413–6423. Horwitz, M. S. (1990). Adenoviridae and their replication. In ‘‘Fields’ Virology’’ (B. N. Fields, D. M. Knipe et al., Eds.), Vol. 2, 2nd ed. Raven Press, New York, NY. Jego, N., Thomas, G., and Hamelin, R. (1993). Short direct repeats flanking deletions, and duplicating insertions in p53 gene in human cancers. Oncogene 8, 209–213. Junejo, F., MacLean, A. R., and Brown, S. M. (1991). Sequence analysis of the herpes simplex virus type 1 strain 17 variants 1704, 1705, and 1706 with respect to their origin and effect on the latency-associated transcript sequence. J. Gen. Virol. 72, 2311–2315. Kemp, M. C., Hierholzer, J. C., Cabradilla, C. P., and Obijeski, J. F. (1983). The changing etiology of epidemic keratoconjunctivitis: antigenic and restriction enzyme analyses of adenovirus types 19 and 37 isolated over a 10-year period. J. Infect. Dis. 148, 24–33. Kimura, H., Iyehara-Ogawa, H., and Kato, T. (1994). Slippage-misalignment: to what extent does it contribute to mammalian cell mutagenesis? Mutagenesis 9, 395–400. Knapp, S., Karth, G. D., Friedl, J., Purtscher, B., Mitterbauer, G., Domer, P. H., and Jaeger, U. (1994). Guanine-rich (GGNNGG) elements at chromosomal breakpoints interact with a loop-forming, singlestranded DNA-binding protein. Oncogene 9, 1501–1505. Krawczak, M., and Cooper, D. N. (1991). Gene deletions causing human

AID

VY 8172

/

6A1F$$$904

09-14-96 07:42:01

genetic disease: mechanisms of mutagenesis and the role of the local DNA sequence environment. Hum. Genet. 86, 425–441. Kumagai, M., and Ikeda, H. (1991). Molecular analysis of the recombination junctions of lambda-bio transducing phages. Mol. Gen. Genet. 230, 60–64. Kunkel, T. A. (1985). Mutational specificity of DNA polymerases-alpha and -gamma during in vitro DNA synthesis. J. Biol. Chem. 260, 12866– 12874. Kunkel, T. A., and Alexander, P. S. (1986). The base substitution fidelity of eucaryotic DNA polymerases. J. Biol. Chem. 261, 160–166. Kunkel, T. A. (1986). Frameshift mutagenesis by eucaryotic DNA polymerases in vitro. J. Biol. Chem. 261, 13581–13587. Kunkel, T. A., and Soni, A. (1988). Mutagenesis by transient misalignment. J. Biol. Chem. 263, 14784–14789. Liskay, R. M., Letsou, A., and Stachelek, J. L. (1987). Homology requirements for efficient gene conversion between duplicated chromosomal sequences in mammalian cells. Genetics 115, 161–167. Meuth, M. (1989). Illegitimate recombination in mammalian cells. In ‘‘Mobile DNA’’ (D. E. Berg and M. M. Howe, Eds.) Am. Soc. Microbiol., Washington, DC. Mezard, C., Pompon, D., and Nicholas, A. (1992). Recombination between similar but not identical DNA sequences during yeast transformation occurs within short stretches of identity. Cell 70, 659–670. Munz, P. L., Young, C., and Young, C. S. H. (1983). The genetic analysis of adenovirus recombination in triparental and superinfection crosses. Virology 126, 576–586. Munz, P. L., and Young, C. S. H. (1984). Polarity in adenovirus recombination. Virology 135, 503–514. Norrby, E. (1969). The structural and functional diversity of adenovirus capsid components. J. Gen. Virol. 5, 221–236. Ohba, K.-I., Mizokami, M., Ohno, T., Suzuki, K., Orito, E., Lau, J. Y. N., Ina, Y., Ikeo, K., and Gojobori, T. (1995). Relationships between serotypes and genotypes of hepatitis B virus: Genetic classification of HBV by use of surface genes. Virus Res. 39, 25–34. Ong, C.-K., Chan, S.-Y., Campo, M. S., Fujinaga, K., Mavromara-Nazos, P., Labropoulou, V., Pfister, H., Tay, S.-K., ter Meulen, J., Villa, L. L., and Bernard, H.-U. (1993). Evolution of human papillomavirus type 18: an ancient phylogenetic root in Africa and intratype diversity reflect coevolution with human genetic groups. J. Virol. 67, 6424– 6431. Papanicolaou, C., and Ripley, L. S. (1991). An in vitro approach to identifying specificity determinants of mutagenesis mediated by DNA misalignments. J. Mol. Biol. 221, 805–821. Ruley, H. E., and Fried, M. (1983). Clustered illegitimate recombination events in mammalian cells involving very short sequence homologies. Nature 304, 181–184. Sambrook, J., Williams, J., Sharp, P. A., and Grodzicker, T. (1975). Physical mapping of temperature-sensitive mutations of adenoviruses. J. Mol. Biol. 97, 369–390. Schnurr, D. P., and Dondero, M. E. (1993). Two new candidate adenovirus serotypes. Intervirol. 36, 79–83. Schnurr, D. P., Bollen, A., Crawford-Miksza, L., Dondero, M. E., and Yagi, S. (1995). Adenovirus mixture isolated from the brain of an AIDS patient with encephalitis. J. Med. Virol. 47, 168–171. Schorr, J., and Doerfler, W. (1993). Non-homologous recombination between adenovirus and AcNPV DNA fragments in cell-free extracts from insect Spodoptera frugiperda nuclei. Virus Res. 28, 153–170. Stary, A., and Sarasin, A. (1992). Simian virus 40 (SV40) large T antigendependent amplification of an Epstein-Barr virus-SV40 hybrid shuttle vector integrated into the human HeLa cell genome. J. Gen. Virol. 73, 1679–1685. Streisinger, G., Okada, Y., Emrich, J., Newton, J., Tsugita, A., Terzaghi, E., and Inouye, M. (1966). Frameshift mutations and the genetic code. Cold Spring Harbor Symp. Quant. Biol. 31, 77–84. Takemori, N. (1972). Genetic studies with tumorigenic adenoviruses III. Recombination in adenovirus type 12. Virology 47, 157–167.

viras

AP: Virology

ADENOVIRUS EVOLUTION BY ILLEGIMATE RECOMBINATION Tamanoi, F. (1986). On the mechanism of adenovirus DNA replication. In ‘‘Adenovirus DNA’’ (W. Doerfler, Ed.) Nijhoff, Boston, MA. Volkert, F. C., and Young, C. S. H. (1983). The genetic analysis of recombination using adenovirus overlapping terminal DNA fragments. Virology 125, 175–193. Wang, H.-P., and Rogler, C. E. (1991). Topoisomerase I-mediated integration of hepadnavirus DNA in vitro. J. Virol. 65, 2381–2392. Wang, T. S.-F., Wong, S. W., and Korn, D. (1989). Human DNA polymerase alpha: predicted functional domains and relationships with viral DNA polymerases. FASEB J. 3, 14–21. Watson, J. D., Hopkins, N. H., Roberts, J. W., Steitz, J. A., and Weiner, A. M. (1987). ‘‘The Molecular Biology of the Gene,’’ 4th ed., Benjamin/ Cummings Publishing Co., Menlo Park, CA. Wigand, R., and Adrian, Th. (1986). Classification and epidemiology of adenoviruses. In ‘‘Adenovirus DNA’’ (W. Doerfler, Ed.) Martinus Nijhoff Publishing, Boston, MA.

AID

VY 8172

/

6A1F$$$904

09-14-96 07:42:01

367

Wigand, R., and Adrian, Th. (1989). Intermediate adenovirus strains of subgenus D occur in extensive variety. Med. Microbiol. Immunol. 178, 37–44. Willcox, N., and Mautner, V. (1976). Antigenic determinants of adenovirus capsids I. Measurement of antibody cross-reactivity. J. Immunol. 116, 19–24. Williams, J., and Ustacelebi, S. (1971). Complementation and recombination with temperature-sensitive mutants of adenovirus type 5. J. Gen. Virol. 13, 345–348. Williams, J., Grodzicker, T., Sharp, P., and Sambrook, J. (1975). Adenovirus recombination: physical mapping of crossover events. Cell 4, 113–119. Young, C. S. H., and Silverstein, S. J. (1980). The kinetics of adenovirus recombination in homotypic and heterotypic genetic crosses. Virology 101, 503–515.

viras

AP: Virology