[This retracts the article DOI: 10.1093/ve/vez043.].

Zika virus (ZIKV) is a mosquito-borne flavivirus first isolated in Uganda in 1947. Although entomological and virologic surveillance have reported ZIKV enzootic activity in diverse countries of Africa and Asia, few human cases were reported until 2007, when a Zika fever epidemic took place in Micronesia. In the context of West Africa, the WHO Collaborating Centre for Arboviruses and Hemorrhagic Fever at Institut Pasteur of Dakar (http://www.pasteur.fr/recherche/banques/CRORA/) reports the periodic circulation of ZIKV since 1968. Despite several reports on ZIKV, the genetic relationships among viral strains from West Africa remain poorly understood. To evaluate the viral spread and its molecular epidemiology, we investigated 37 ZIKV isolates collected from 1968 to 2002 in six localities in Senegal and Côte d'Ivoire. In addition, we included strains from six other countries. Our results suggested that these two countries in West Africa experienced at least two independent introductions of ZIKV during the 20th century, and that apparently these viral lineages were not restricted by mosquito vector species. Moreover, we present evidence that ZIKV has possibly undergone recombination in nature and that a loss of the N154 glycosylation site in the envelope protein was a possible adaptive response to the Aedes dalzieli vector.

Within the Haloferax genus, both the surface (S)-layer protein, and the glycans that can decorate it, vary between species, which can potentially result in many different surface types, analogous to bacterial serotypes. This variation may mediate phenotypes, such as sensitivity to different viruses and mating preferences. Here, we describe S-layer glycoproteins found in multiple Haloferax strains and perform comparative genomics analyses of major and alternative glycosylation clusters of isolates from two coastal sites. We analyze the phylogeny of individual glycosylation genes and demonstrate that while the major glycosylation cluster tends to be conserved among closely related strains, the alternative cluster is highly variable. Thus, geographically- and genetically-related strains may exhibit diverse surface structures to such an extent that no two isolates present an identical surface profile.

With the development of reliable recombination detection tools and an increasing number of available genome sequences, many studies have reported evidence of recombination in a wide range of virus genera. Recombination is apparently a major mechanism in virus evolution, allowing viruses to evolve more quickly by providing immediate direct access to many more areas of a sequence space than are accessible by mutation alone. Recombination has been widely described amongst the insect-transmitted plant viruses in the genus Begomovirus (family Geminiviridae), with potential recombination hot- and cold-spots also having been identified. Nevertheless, because very little is understood about either the biochemical predispositions of different genomic regions to recombine or what makes some recombinants more viable than others, the sources of the evolutionary and biochemical forces shaping distinctive recombination patterns observed in nature remain obscure. Here we present a detailed analysis of unique recombination events detectable in the DNA-A and DNA-A-like genome components of bipartite and monopartite begomoviruses. We demonstrate both that recombination breakpoint hot- and cold-spots are conserved between the two groups of viruses, and that patterns of sequence exchange amongst the genomes are obviously non-random. Using a computational technique designed to predict structural perturbations in chimaeric proteins, we demonstrate that observed recombination events tend to be less disruptive than sets of simulated ones. Purifying selection acting against natural recombinants expressing improperly folded chimaeric proteins is therefore a major determinant of natural recombination patterns in begomoviruses.

We used a comparative genomics approach to investigate the evolution of a complex nucleotide-binding (NB)-leucine-rich repeat (LRR) gene cluster found in soybean (Glycine max) and common bean (Phaseolus vulgaris) that is associated with several disease resistance (R) genes of known function, including Rpg1b (for Resistance to Pseudomonas glycinea1b), an R gene effective against specific races of bacterial blight. Analysis of domains revealed that the amino-terminal coiled-coil (CC) domain, central nucleotide-binding domain (NB-ARC [for APAF1, Resistance genes, and CED4]), and carboxyl-terminal LRR domain have undergone distinct evolutionary paths. Sequence exchanges within the NB-ARC domain were rare. In contrast, interparalogue exchanges involving the CC and LRR domains were common, consistent with both of these regions coevolving with pathogens. Residues under positive selection were overrepresented within the predicted solvent-exposed face of the LRR domain, although several also were detected within the CC and NB-ARC domains. Superimposition of these latter residues onto predicted tertiary structures revealed that the majority are located on the surface, suggestive of a role in interactions with other domains or proteins. Following polyploidy in the Glycine lineage, NB-LRR genes have been preferentially lost from one of the duplicated chromosomes (homeologues found in soybean), and there has been partitioning of NB-LRR clades between the two homeologues. The single orthologous region in common bean contains approximately the same number of paralogues as found in the two soybean homeologues combined. We conclude that while polyploidization in Glycine has not driven a stable increase in family size for NB-LRR genes, it has generated two recombinationally isolated clusters, one of which appears to be in the process of decay.

We present the new and fast method Recco for analyzing a multiple alignment regarding recombination. Recco is based on a dynamic program that explains one sequence in the alignment with the other sequences using mutation and recombination. The dynamic program allows for an intuitive visualization of the optimal solution and also introduces a parameter α controlling the number of recombinations in the solution. Recco performs a parametric analysis regarding α and orders all pareto-optimal solutions by increasing number of recombinations. α is also directly related to the Savings value, a quantitative and intuitive measure for the preference of recombination in the solution. The Savings value and the solutions have a simple interpretation regarding the ancestry of the sequences in the alignment and it is usually easy to understand the output of the method. The distribution of the Savings value for non-recombining alignments is estimated by processing column permutations of the alignment and p-values are provided for recombination in the alignment, in a sequence and at a breakpoint position. Recco also uses the p-values to suggest a single solution, or recombinant structure, for the explained sequence. Recco is validated on a large set of simulated alignments and has a recombination detection performance superior to all current methods. The analysis of real alignments confirmed that Recco is among the best methods for recombination analysis and further supported that Recco is very intuitive compared to other methods. Wir prasentieren Recco, eine neue und schnelle Methode zur Analyse von Rekombinationen in multiplen Alignments. Recco basiert auf einem dynamischen Programm, welches eine Sequenz im Alignment durch die anderen Sequenzen im Alignment rekonstruiert, wobei die Operatoren Mutation und Rekombination erlaubt sind. Das dynamische Programm ermoglicht eine intuitive Visualisierung der optimalen Losung und besitzt einen Parameter α, welcher die Anzahl der Rekombinationsereignisse in der optimalen Losung steuert. Recco fuhrt eine parametrische Analyse bezuglich des Parameters α durch, so dass alle pareto-optimalen Losungen nach der Anzahl ihrer Rekombinationsereignisse sortiert werden konnen. α steht auch direkt in Beziehung mit dem sogenannten Savings-Wert, der die Neigung zum Einfugen von Rekombinationsereignissen in die optimale Losung quantitativ und intuitiv bemisst. Der Savings-Wert und die optimalen Losungen haben eine einfache Interpretation bezuglich der Historie der Sequenzen im Alignment, so dass es in der Regel leicht fallt, die Ausgabe von Recco zu verstehen. Recco schatzt die Verteilung des Savings-Werts fur Alignments ohne Rekombinationen durch einen Permutationstest, der auf Spaltenpermutationen basiert. Dieses Verfahren resultiert in p-Werten fur Rekombination im Alignment, in einer Sequenz und an jeder Position im Alignment. Basierend auf diesen p-Werten schlagt Recco eine optimale Losung vor, als Schatzer fur die rekombinante Struktur der erklarten Sequenz. Recco wurde auf einem grosen Datensatz simulierter Alignments getestet und erzielte auf diesem Datensatz eine bessere Vorhersagegute in Bezug auf das Erkennen von Alignments mit Rekombination als alle anderen aktuellen Verfahren. Die Analyse von realen Datensatzen bestatigte, dass Recco zu den besten Methoden fur die Rekombinationsanalyse gehort und im Vergleich zu anderen Methoden oft leichter verstandliche Resultate liefert.

We have studied the evolution of a type IV secretion system (T4SS), in Bartonella, which is thought to have changed function from conjugation to erythrocyte adherence following a recent horizontal gene transfer event. The system, called Trw, is unique among T4SSs in that genes encoding both exo- and intracellular components are located within the same duplicated fragment. This provides an opportunity to study the influence of selection on proteins involved in host-pathogen interactions. We sequenced the trw locus from several strains of Bartonella henselae and investigated its evolutionary history by comparisons to other Bartonella species. Several instances of recombination and gene conversion events where detected in the 2- to 5-fold duplicated gene fragments encompassing trwJIH, explaining the homogenization of the anchoring protein TrwI and the divergence of the minor pilus protein TrwJ. A phylogenetic analysis of the 7- to 8-fold duplicated gene coding for the major pilus protein TrwL displayed 2 distinct clades, likely representing a subfunctionalization event. The analyses of the B. henselae strains also identified a recent horizontal transfer event of almost the complete trwL region. We suggest that the switch in function of the T4SS was mediated by the duplication of the genes encoding pilus components and their diversification by combinatorial sequence shuffling within and among genomes. We suggest that the pilus proteins have evolved by diversifying selection to match a divergent set of erythrocyte surface structures, consistent with the trench warfare coevolutionary model.

Viruses are the most abundant and the smallest organisms, which are relatively simple to sequence. Genome sequence data of viruses for individual species to populations out‐ number that of other species. Although this offers an opportunity to study viral diversity at varying levels of taxonomic hierarchy, it also poses challenges for systematic and struc‐ tured organization of data and its downstream processing. Extensive computational anal‐ yses using a number of algorithms and programs have opened exciting opportunities for virus discovery and diagnostics, apart from augmenting our understanding of the intri‐ guing world of viruses. Unravelling evolutionary dynamics of viruses permits improved understanding of phenomena such as quasispecies diversity, role of mutations in host switching and drug resistance, which enables the tangible measurements of genotype and phenotype of viruses. Improved understanding of geno-/serotype diversity in corre‐ lation with antigenic diversity will facilitate rational design and development of effica‐ cious vaccines against emerging and re-emerging viruses. Mathematical models developed using the genomic data could be used to predict the spread of viruses due to vector switching and the (re)emergence due to host switching and, thereby, contribute to‐ wards designing public health policies for disease management and control.

Viking smallpox diversity Humans have a notable capacity to withstand the ravages of infectious diseases. Smallpox killed millions of people but drove Jenner's invention of vaccination, which eventually led to the annihilation of this virus, declared in 1980. To investigate the history of smallpox, Mühlemann et al. obtained high-throughput shotgun sequencing data from 1867 human remains ranging from >31,000 to 150 years ago (see the Perspective by Alcamí). Thirteen positive samples emerged, 11 of which were northern European Viking Age people (6th to 7th century CE). Although the sequences were patchy and incomplete, four could be used to infer a phylogenetic tree. This showed distinct Viking Age lineages with multiple gene inactivations. The analysis pushes back the date of the earliest variola infection in humans by ∼1000 years and reveals the existence of a previously unknown virus clade. Science, this issue p. eaaw8977; see also p. 376 Variola virus DNA sequences isolated from ancient human remains reveal the presence of smallpox in 7th-century northern Europe. INTRODUCTION Variola virus (VARV), the causative agent of smallpox, is estimated to have killed between 300 million and 500 million people in the 20th century and was responsible for widespread mortality and suffering for at least several preceding centuries. Humans are the only known host of VARV, and smallpox was declared eradicated in 1980. The timeline of the emergence of smallpox in humans is unclear. Based on sequence data up to 360 years old, the most recent common ancestor of VARV has been dated to the 16th or 17th century. This contrasts with written records of possible smallpox infections dating back at least 3000 years and mummified remains suggestive of smallpox dating to 3570 years ago. RATIONALE Ancient virus sequences recovered from archaeological remains provide direct molecular evidence of past infections, give detail of genetic changes that have occurred during the evolution of the virus, and can reveal viable virus sequence diversity not currently present in modern viruses. In the case of VARV, ancient sequences may also reduce the gap between the written historical record of possible early smallpox infections and the dating of the oldest available VARV sequences. We therefore screened high-throughput shotgun sequencing data from skeletal and dental remains of 1867 humans living in Eurasia and the Americas between ~31,630 and ~150 years ago for the presence of sequences matching VARV. RESULTS VARV sequences were recovered from 13 northern European individuals, including 11 dated to ~600–1050 CE, overlapping the Viking Age, and we reconstructed near-complete VARV genomes for four of them. The samples predate the earliest confirmed smallpox cases by ~1000 years. Eleven of the recovered sequences fall into a now-extinct sister clade of the modern VARVs in circulation prior to the eradication of smallpox, while two sequences from the 19th century group with modern VARV. The inferred date of the most recent common ancestor of VARV is ~1700 years ago. The number of functional genes is generally reduced in orthopoxviruses with narrow host ranges. A comparison of the gene content of the Viking Age sequences shows great contrast with that of modern VARV. Three genes that are active in all modern VARV sequences were inactive over 1000 years ago in some or all ancient VARV. Among 10 genes inactive in modern and Viking Age VARV, the mutations causing the inactivations are different and the genes are predicted to be active in the ancestor of both clades, suggesting parallel evolution. Fourteen genes inactivated in modern VARV are active in some or all of the ancient sequences, eight of which encode known virulence factors or immunomodulators. The active gene counts of the four higher-coverage Viking Age viral genomes provide snapshots from an ~350-year period, showing the reduction of gene content during the evolution of VARV. These genomes support suggestions that orthopoxvirus species derive from a common ancestor containing all genes present in orthopoxviruses today, with the reduction in active gene count conjectured to be the result of long-term adaptation within host species. CONCLUSION The Viking Age sequences reported here push the definitive date of the earliest VARV infection in humans back by ~1000 years. These sequences, combined with early written records of VARV epidemics in southern and western Europe, suggest a pan-European presence of smallpox from the late 6th century. The ancient viruses are part of a previously unknown, now-extinct virus clade and were following a genotypic evolutionary path that differs from modern VARV. The reduction in gene content shows that multiple combinations of active genes have led to variola viruses capable of circulating widely within the human population. Sample locations and genome evolution of Viking Age variola virus. (Top left) Sample VK533 with dagger, in situ. (Bottom left) Positive sample locations. Higher- and lower-coverage samples have solid and open circles, respectively. (Right) Diverse gene inactivation patterns in the four higher-coverage 600–1000 CE Viking Age sequences, within clade and as compared to sequences from other human VARVs and the phylogenetically nearest animal poxviruses, camelpox and taterapox. Open circles show intact genes, circle colors (independent for each gene) indicate different inactivations. Gene names at the bottom. Smallpox, one of the most devastating human diseases, killed between 300 million and 500 million people in the 20th century alone. We recovered viral sequences from 13 northern European individuals, including 11 dated to ~600–1050 CE, overlapping the Viking Age, and reconstructed near-complete variola virus genomes for four of them. The samples predate the earliest confirmed smallpox cases by ~1000 years, and the sequences reveal a now-extinct sister clade of the modern variola viruses that were in circulation before the eradication of smallpox. We date the most recent common ancestor of variola virus to ~1700 years ago. Distinct patterns of gene inactivation in the four near-complete sequences show that different evolutionary paths of genotypic host adaptation resulted in variola viruses that circulated widely among humans.

UNLABELLED RDP2 is a Windows 95/XP program that examines nucleotide sequence alignments and attempts to identify recombinant sequences and recombination breakpoints using 10 published recombination detection methods, including GENECONV, BOOTSCAN, MAXIMUM chi(2), CHIMAERA and SISTER SCANNING. The program enables fast automated analysis of large alignments (up to 300 sequences containing 13 000 sites), and interactive exploration, management and verification of results with different recombination detection and tree drawing methods. AVAILABILITY RDP2 is available free from the RDP2 website (http://darwin.uvigo.es/rdp/rdp.html) CONTACT darren@science.uct.ac.za SUPPLEMENTARY INFORMATION Detailed descriptions of RDP2 and the methods it implements are included in the program manual, which can be downloaded from the RDP2 website.

Tuberculosis (TB) and leprosy are ancient human diseases, as demonstrated by an abundant osteoarchaeological record and by molecular analyses of prehistoric and historic human remains affected by the disease. Unfortunately, mycobacterial infections (which include TB and leprosy) are not diseases of the past. Today, mycobacteria are still an enormous burden to humans, with an estimated two billion individuals currently being affected by the disease (in a latent or active stage) worldwide. It has become clear that these successful human pathogens can rapidly adapt to their changing environment and host. Evolutionary theory provides an excellent framework to test hypotheses that aim to explain how these pathogens are able to sustain such changes. Understanding the evolutionary mechanisms of the genus Mycobacterium will improve our understanding, and perhaps our ability to control, the processes that yield to these pathogens' evolution and acquisition of drug resistance. Among these adaptive mechanisms, recombination is without doubt an effective way to acquire foreign genetic material into the mycobacterial genome. The scientific community has not reached a consensus on the occurrence of recombination in the genus Mycobacterium. This dissertation aimed to screen a data set comprising 2354 homologous genome sections from publicly available mycobacterial sequences representing 18 strains from 13 species for the occurrence of recombination. Homologous genomic sets were aligned and concatenated in the chromosomal order of one mycobacterial reference species. The concatenated sequence was screened for recombination using a variety of statistical methods, with each event being evaluated by comparing maximum likelihood phylogenies of each recombinant section with the non-recombining portion of the dataset. Incongruent phylogenies were identified by comparing the site-wise log-likelihoods of each tree using multiple tests. The analyses described here indicate that the hypothesis of recombination in mycobacterial homologues cannot be rejected. While an analysis of non-homologous mycobacterial regions will be required to assess the occurrence of nonhomologous recombination, this study shows that we cannot assume mycobacterial clonality, even in homologous genomic sections.

To monitor the relative prevalence and evolutionary trends of HIV-1 in Uganda, we conducted a retrospective study of pregnant women over the time period 1989-2000. From a total of 300 women sampled, we defined subtypes by heteroduplex mobility assay for 230 subjects and by partial sequencing and phylogenetic analyses of the env gene for 216 subjects. Subtypes A and D were most prevalent, and there were no significant trends in relative frequencies of subtypes A (45%), D (41%), C (5%), or recombinants (9%) over the 11 years sampled. There was also no phylogenetic clustering of subtypes related to geography (clinic location) or year of collection. Mean pairwise nucleotide diversity of subtype A (pi = 0.163) and subtype D (pi =0.156) samples did not differ significantly between subtypes, nor did these levels change over the period of the study. This report suggests that among pregnant women in Uganda A and D subtypes are transmitted without geographic constraints, and are not associated with significantly different transmission rates.

This study reports the findings of a distinct Potato virus Y (PVY) isolate found in Northeast China. One hundred and ten samples (leaves and tubers) were collected from potato plants showing mosaic symptoms around the city of Harbin in Heilongjiang province of China. The collected tubers were planted and let to grow in a greenhouse. New potato plants generated from these tubers showed similar symptoms, except for one plant. Subsequent serological analyses revealed PVY as the causing agent of the disease. A novel PVY isolate (referred to as HLJ-C-44 in this study) was isolated from this sample showing unique mild mosaic and crisped leaf margin symptoms. The complete genome of this isolate was analyzed and determined. The results showed that HLJ-C-44 is a typical PVY isolate. Phylogenetic analysis indicated that this isolate belongs to the N-Wi strain group of PVY recombinants (PVYN-Wi) and also shared the highest overall sequence identity (nucleotide and amino acid) with other members of this strain group. However, recombination analysis of isolate HLJ-C-44 revealed a recombination pattern that differed from that of other PVYN-Wi isolates. Moreover, biological assays in four different potato cultivars and in Nicotiana tabacum also revealed a different phenotypic response than that of a typical PVYN-Wi isolate. This data, combined, suggest that HLJ-C-44 is a novel PVY recombinant with distinct biological properties.

This case underscores the risk for spillover infections to humans who work with exotic squirrels.

The phylogenetic relationship between the complete genomic sequences of ten Japanese and one French isolate of TT virus-like mini virus (TLMV) was investigated. Analysis of the variability of the nucleotide sequences and the detection of signature patterns for overlapping genes suggested that ORFs 1 and 2 are probably functional. However, this was not the case for a putative third ORF, ORF3. Throughout the viral genome, several nucleotide or amino acid motifs that are conserved in circoviruses such as TT virus (TTV) and chicken anaemia virus were identified. Phylogenetic analysis distinguished three main groups of TLMV and allowed the identification of putative recombination breakpoints in the untranslated region. TLMV genomes were detected by PCR in the plasma of 38/50 French blood donors tested and were also identified in peripheral blood mononuclear cells, faeces and saliva. A phylogenetic study of 37 TLMV strains originating from France, Japan and Brazil showed that groupings were not related to geographical origin.

The performance of 14 different recombination detection methods was evaluated by analyzing several empirical data sets where the presence of recombination has been suggested or where recombination is assumed to be absent. In general, recombination methods seem to be more powerful with increasing levels of divergence, but different methods showed distinct performance. Substitution methods using summary statistics gave more accurate inferences than most phylogenetic methods. However, definitive conclusions about the presence of recombination should not be derived on the basis of a single method. Performance patterns observed from the analysis of real data sets coincided very well with previous computer simulation results. Previous recombination inferences from some of the data sets analyzed here should be reconsidered. In particular, recombination in HIV-1 seems to be much more widespread than previously thought. This finding might have serious implications on vaccine development and on the reliability of previous inferences of HIV-1 evolutionary history and dynamics.

The genotyping of highly polymorphic multigene families across many individuals used to be a particularly challenging task because of methodological limitations associated with traditional approaches. Next‐generation sequencing (NGS) can overcome most of these limitations, and it is increasingly being applied in population genetic studies of multigene families. Here, we critically review NGS bioinformatic approaches that have been used to genotype the major histocompatibility complex (MHC) immune genes, and we discuss how the significant advances made in this field are applicable to population genetic studies of gene families. Increasingly, approaches are introduced that apply thresholds of sequencing depth and sequence similarity to separate alleles from methodological artefacts. We explain why these approaches are particularly sensitive to methodological biases by violating fundamental genotyping assumptions. An alternative strategy that utilizes ultra‐deep sequencing (hundreds to thousands of sequences per amplicon) to reconstruct genotypes and applies statistical methods on the sequencing depth to separate alleles from artefacts appears to be more robust. Importantly, the ‘degree of change’ (DOC) method avoids using arbitrary cut‐off thresholds by looking for statistical boundaries between the sequencing depth for alleles and artefacts, and hence, it is entirely repeatable across studies. Although the advances made in generating NGS data are still far ahead of our ability to perform reliable processing, analysis and interpretation, the community is developing statistically rigorous protocols that will allow us to address novel questions in evolution, ecology and genetics of multigene families. Future developments in third‐generation single molecule sequencing may potentially help overcome problems that still persist in de novo multigene amplicon genotyping when using current second‐generation sequencing approaches.

The genomes of most, if not all, flowering plants have undergone whole genome duplication events during their evolution. The impact of such polyploidy events is poorly understood, as is the fate of most duplicated genes. We sequenced an approximately 1 million-bp region in soybean (Glycine max) centered on the Rpg1-b disease resistance gene and compared this region with a region duplicated 10 to 14 million years ago. These two regions were also compared with homologous regions in several related legume species (a second soybean genotype, Glycine tomentella, Phaseolus vulgaris, and Medicago truncatula), which enabled us to determine how each of the duplicated regions (homoeologues) in soybean has changed following polyploidy. The biggest change was in retroelement content, with homoeologue 2 having expanded to 3-fold the size of homoeologue 1. Despite this accumulation of retroelements, over 77% of the duplicated low-copy genes have been retained in the same order and appear to be functional. This finding contrasts with recent analyses of the maize (Zea mays) genome, in which only about one-third of duplicated genes appear to have been retained over a similar time period. Fluorescent in situ hybridization revealed that the homoeologue 2 region is located very near a centromere. Thus, pericentromeric localization, per se, does not result in a high rate of gene inactivation, despite greatly accelerated retrotransposon accumulation. In contrast to low-copy genes, nucleotide-binding-leucine-rich repeat disease resistance gene clusters have undergone dramatic species/homoeologue-specific duplications and losses, with some evidence for partitioning of subfamilies between homoeologues.

The evolution of the mitochondrial (mt) genome is far from being fully understood. Systematic investigations into the modes of inheritance, rates and patterns of recombination, nucleotide substitution, and structural changes in the mt genome are still lacking in many groups of plants. In this study, we sequenced >11kbp mtDNA segments from multiple accessions of 36 pine species to characterize the evolutionary patterns of mtDNA in the genus Pinus. We found extremely low substitution rates and complex repetitive sequences scattered across different genome regions, as well as chimeric structures that were probably generated by multiple intergenomic recombinations. The mtDNA-based phylogeny of the genus differed from that based on chloroplast and nuclear DNA in the placement of several groups of species. Such discordances suggest a series of mtDNA capture events during past range shifts of the pine species and that both vertical and horizontal inheritance are implicated in the evolution of mtDNA in Pinus. MtDNA dating revealed that most extant lineages of the genus originated during Oligocene-Miocene radiation and subgenus Strobus diversified earlier than subgenus Pinus. Our findings illustrate a reticular evolutionary pathway for the mt genome through capture and recombination in the genus Pinus, and provide new insights into the evolution of the genus.

The biological and molecular characterization of a virus recognized as a distinct begomovirus species, Tomato curly stunt virus (ToCSV), first observed in South Africa in 1997, is reported here. Whitefly-transmission and host-range studies were carried out using a Bemisia tabaci colony identified as the B-biotype. The experimental host range of ToCSV spanned primarily species in the Solanaceae and Fabaceae. The complete ToCSV genome (2·766 kb) was amplified by PCR, cloned, and the DNA sequence determined. Phylogenetic analysis revealed that ToCSV was most closely related to Tobacco leaf curl Zimbabwe virus (TbLCZV), at 84% nucleotide identity, indicating that ToCSV is a new species in the genus Begomovirus that is probably endemic to southern Africa. The ToCSV genome sequence contained all of the hallmark coding and non-coding features characteristic of other previously recognized monopartite begomoviruses. ToCSV is only the second begomovirus described from southern Africa that infects solanaceous species. Neither a begomoviral DNA-B component nor a satellite-like DNA molecule was detected by PCR in extracts of ToCSV-infected plants.