Analysis of the genome of Spodoptera frugiperda nucleopolyhedrovirus (SfMNPV-19) and of the high genomic heterogeneity in group II nucleopolyhedroviruses

Analysis of the genome of Spodoptera frugiperdanucleopolyhedrovirus (SfMNPV-19) and of the highgenomic heterogeneity in group IInucleopolyhedroviruses

Jose Luiz Caldas Wolff,1 Fernando Hercos Valicente,2 Renata Martins,1

Juliana Velasco de Castro Oliveira3 and Paolo Marinho de Andrade Zanotto3

Correspondence

Jose Luiz Caldas Wolff

[email protected]

1Laboratorio de Virologia Molecular, Nucleo Integrado de Biotecnologia, Universidade de Mogi dasCruzes, Mogi das Cruzes, SP, Brazil

2Embrapa Milho e Sorgo, Sete Lagoas, MG, Brazil

3Laboratorio de Evolucao Molecular e Bionformatica, Departamento de Microbiologia, Instituto deCiencias Biomedicas, Universidade de Sao Paulo, SP, Brazil

Received 10 November 2007

Accepted 25 January 2008

The genome of the most virulent among 22 Brazilian geographical isolates of Spodoptera

frugiperda nucleopolyhedrovirus, isolate 19 (SfMNPV-19), was completely sequenced and shown

to comprise 132 565 bp and 141 open reading frames (ORFs). A total of 11 ORFs with no

homology to genes in the GenBank database were found. Of those, four had typical baculovirus

promoter motifs and polyadenylation sites. Computer-simulated restriction enzyme cleavage

patterns of SfMNPV-19 were compared with published physical maps of other SfMNPV isolates.

Differences were observed in terms of the restriction profiles and genome size. Comparison of

SfMNPV-19 with the sequence of the SfMNPV isolate 3AP2 indicated that they differed due to a

1427 bp deletion, as well as by a series of smaller deletions and point mutations. The majority

of genes of SfMNPV-19 were conserved in the closely related Spodoptera exigua NPV

(SeMNPV) and Agrotis segetum NPV (AgseMNPV-A), but a few regions experienced major

changes and rearrangements. Synthenic maps for the genomes of group II NPVs revealed that

gene collinearity was observed only within certain clusters. Analysis of the dynamics of gene gain

and loss along the phylogenetic tree of the NPVs showed that group II had only five defining

genes and supported the hypothesis that these viruses form ten highly divergent ancient lineages.

Crucially, more than 60 % of the gene gain events followed a power-law relation to genetic

distance among baculoviruses, indicative of temporal organization in the gene accretion process.

INTRODUCTION

The Baculoviridae is a large family of insect-specific viruseswhose major morphological trait is the presence of anocclusion body, a protein matrix that protects the infectiveparticles and is critical for transmission to insect larvae(Tanada & Kaya, 1992; Blissard & Rohrmann, 1990). Thisfamily of virus is divided into two genera: Nucleopoly-hedrovirus (NPV) and Granulovirus (GV) (Blissard et al.,2000). Phylogenetic analyses of baculovirus genes con-firmed the division of the two genera and indicated thatNPVs could possibly be separated into two groups, namedgroup I and II (Zanotto et al., 1993). Baculoviruseshave enveloped rod-shaped virions containing a circular

supercoiled double-stranded DNA molecule ranging in sizefrom 81 756 bp for Neodiprion lecontei NPV (Lauzon et al.,2004) to 178 733 bp for Xestia c-nigrun GV (Hayakawaet al., 1999). Their genomes comprise from 90 to 181 genesthat are closely packaged, sometimes even partiallyoverlapped, in both strands of the DNA molecule.

The first baculovirus to be completely sequenced was theAutographa californica NPV (AcNPV; Ayres et al., 1994).Three years later, the genome of another NPV, the Orgyiapseudotsugata NPV (OpMNPV), was published (Ahrenset al., 1997). Today, genomes from over 40 baculoviruseshave been determined. Most of these genomes are fromNPVs and GVs that infect Lepidoptera. However, in 2001the genome of a NPV that infects Culex nigripalpus(Diptera) was determined (Afonso et al., 2001). Morerecently, genomes of NPVs that infect hymenopteranspecies were also sequenced (Garcia-Maruniak et al.,

The GenBank accession number of the sequence reported in this paperis EU258200.

Supplementary Tables are available with the online version of this paper.

Journal of General Virology (2008), 89, 1202–1211 DOI 10.1099/vir.0.83581-0

1202 0008-3581 G 2008 SGM Printed in Great Britain

2004; Lauzon et al., 2004; Duffy et al., 2006). Analysis of theavailable genomes has shown that 29 genes are conserved inall baculoviruses sequenced so far (Herniou et al., 2003;Garcia-Maruniak et al., 2004; Lauzon et al., 2005). Thisgroup of core genes represent a relatively small part of theover 800 different orthologous gene groups found in allbaculovirus genomes up to the present date (Jehle et al.,2006). The phylogenetic investigation of these genomesrevealed that gene acquisition and loss are importantfactors in the evolution of baculoviruses (Herniou et al.,2001, 2003; Oliveira et al., 2006). Moreover, genomicanalysis suggests that NPVs infecting members of distinctinvertebrate Orders are sufficiently different to arguablymerit classification into separate virus genera (Jehle et al.,2006).

Spodoptera frugiperda is a severe pest of corn and othercrops. In Brazil, where infestations can result in losses of upto 34 % in corn yield, the control of S. frugiperda is oftendifficult and requires the undesirable use of repeatedchemical control. For that reason, the application ofalternative methods of control, such as baculoviruses, hasbecome a priority. Isolates of S. frugiperda NPV (SfMNPV)have been identified from S. frugiperda larvae collectedfrom various regions of the Americas (Maruniak et al.,1984; Shapiro et al., 1991; Barreto et al., 2005; Loh et al.,1981, 1982; Knell & Summers, 1981). Several studies haveshown that SfMNPV isolates often display genetic andbiological differences (Loh et al., 1982; Knell & Summers,1981; Berretta et al., 1998). Recently, the analysis of 22SfMNPV geographical isolates collected in Brazil identifiedan isolate, named SfMNPV-19, which had a lower LC50

(Barreto et al., 2005). This geographical isolate was used inthis investigation.

Sequencing studies with SfMNPV started with thepolyhedrin gene, one of the first baculovirus genes to besequenced (Gonzalez et al., 1989). Six years later, the gp41gene sequence was determined and its expression patternanalysed (Liu & Maruniak, 1995). In 2003, the sequence ofa 5122 bp region that encompasses the gp37, ptp-2 and egtgenes and three other ORFs was determined in a study thatshowed that SfMNPV genome expresses a functionalecdysteroid UDP-glucosyltransferase (EGT) protein(Tumilasci et al., 2003). The sequencing of this genomicregion revealed that genes located in the proximity of theSfMNPV egt are highly conserved and collinear with genesfrom Spodoptera exigua NPV (SeMNPV) and Mamestraconfigurata NPV (MacoMNPV) (Tumilasci et al., 2003). Inthe past few years, the sequences from the p74 gene(GenBank accession no. AY174684), the immediate earlygene (ie-0) (AY846365) and the pif gene (Simon et al.,2005b) have also became available. Finally, 38 ORFs locatedin termini of cloned restriction fragments were partiallysequenced in a study with a Nicaraguan isolate (Simonet al., 2005a). The distribution of these putative genes alongthe genome and the level of sequence similarity observedconfirmed that the SfMNPV genome is indeed highlysimilar to SeMNPV and MacoMNPV (Simon et al., 2005a).

As we were completing the sequence of SfMNPV-19, thesequence of another SfMNPV isolate, named isolate 3AP2,was deposited in the database (GenBank accession no.EF035042). The sequence of this genome was included inour analysis. Herein we report the complete genomesequence of SfMNPV-19, an SfMNPV isolate that has goodpotential to be used as a biological control agent due to itshigh infectivity. The analysis of the genome is presented aswell as a comparison with closely and distantly relatedmembers of the Baculoviridae.

METHODS

SfMNPV purification and DNA preparation. SfMNPV geograph-

ical isolate 19 was obtained from a single larval cadaver collected inParana State, Southern Brazil. It was amplified by infection in vivo

and was subjected to several passages in S. frugiperda larvae reared in

the laboratory. Polyhedra obtained from infected S. frugiperda larvaewere purified and DNA extracted using standard procedures with

minor adjustments (O’Reilly et al., 1992). The DNA was digested with

the restriction enzymes HindIII and BamHI, which did not appear toyield submolar bands.

Construction of a genomic library representing the entire

genome. Purified viral DNA was digested with the restrictionenzymes BglII, EcoRV, HincII, HindIII, KpnI, PstI and SalI and

digested simultaneously with BamHI and BglII. DNA fragments were

cloned into pGEM-3Z (Promega) following standard procedures(Sambrook et al., 1989). A total of 66 unique clones, covering

approximately 90 % of the SfMNPV-19 genome, were chosen as

templates for DNA sequencing. The selection of the unique clones wasdone in accordance with the sequences identified on the edges of the

cloned fragments. The fact that most genes in the genome of

SfMNPV-19 were collinear to those in the closely related SeMNPVand Agrotis segetum NPV (AgseMNPV-A) helped to find possible gaps

in the genomic library. Missing genomic segments were amplified byPCR using oligonucleotides derived from the cloned flanks of each

gap.

Sequence determination and analysis. DNA templates were

sequenced initially using universal primers and oligonucleotides ofthe PCR reactions. The internal sequences were determined by

primer-walking using specific primers. Sanger reactions were

performed by cycle sequencing using Eppendorf thermocyclermachines with the ABI PRISM Big Dye Terminator sequencing ready

reaction kit v. 3 (Applied Biosystems). Electrophoreses were done on

an ABI 377 DNA sequencer (Applied Biosystems). Aligner version1.3.4 (CodonCode) was used for contig assembly using default Phrap

and Phred parameters. The complete genome was checked manually

in order to detect possible errors. Alternative oligonucleotides wereused for resequencing regions that repeatedly displayed ambiguity

and/or poor quality.

Sequence analysis and genome annotation. The release 8 ofArtemis (Rutherford et al., 2000) (The Sanger Institute; http://

www.sanger.ac.uk/Software/Artemis/) and ORF Finder (http://

www.ncbi.nlm.nih.gov/projects/gorf/) were used to find and analyseopen reading frames of over 50 aa, except for those that were

completely overlapped by other ORFs. Identity values of ORFs were

obtained using BLASTP (Altschul et al., 1990), also available at theNCBI (http://www.ncbi.nlm.nih.gov/blast/). Regions of the genome

that differed (presence or absence of ORFs as well as any major

differences in size) from the closely related genomes were checked indetail to verify possible sequencing or assembly errors. The complete

High genomic heterogeneity in group II NPVs

http://vir.sgmjournals.org 1203

genome was compared with other baculovirus genomes using theArtemis comparative tool (ACT) release 5 (Carver et al., 2005) (TheSanger Institute; http://www.sanger.ac.uk/Software/ACT/). The Dotterprogram (Sonnhammer & Durbin, 1995) was used to locatehomologous regions (hrs). Palindromes present in the hrs wereidentified using the program PALINDROME from the EMBOSS package(http://emboss.bioinformatics.nl/cgi-bin/emboss/palindrome). Anno-tation of the genome was done with Artemis (Rutherford et al., 2000).

Phylogenomics. Since the reciprocal value of bit score (1/S9)obtained with the TBLASTX program (Altschul et al., 1990) available inthe NCBI BLAST distribution (http://www.ncbi.nlm.nih.gov/blast/download.shtml) is a direct measure of evolutionary distance(Oliveira et al., 2006), pairwise distance matrices of median valuesof 1/S9 from complete genomes were clustered using the weightedneighbour-joining (weighbor) (Bruno et al., 2000), neighbour-joining, unweighted pair group mean average (UPGMA) and Fitschand Kitsch methods from the PHYLIP package (Felsenstein, 1989).Pairwise distance matrices were calculated with the BlastPhen pipeline(Oliveira et al., 2006) to establish the relationship of SfMNPV-19 to42 other sequenced baculoviruses. For each pair of genomes, medianvalues from bit scores S9 (Altschul et al., 1990; Pertsemlidis &Fondon, 2001) were calculated for all local high-score pairs (HSP),generated by the TBLASTX program.

Gene content and genomic organization. We generated a binarygene-content matrix, expanded from Herniou et al. (2001, 2003) andOliveira et al. (2006) to analyse the gene content of the baculoviruses.This matrix scored the presence (1) or absence (0) of 862 knownbaculovirus genes from 43 baculovirus genomes. The most parsimo-nious reconstruction (MPR) of each gene gain or loss event wasestimated along the branches of a complete genome-based pheno-gram for the Baculoviridae obtained with the BlastPhen algorithm(Oliveira et al., 2006). We also included a synthetic outgroup (i.e. a‘basal’ root in which all genes are absent), in order better to definegain and loss of genes among baculoviruses and other cellular andviral organisms (including prokaryotes, eukaryotes, DNA and RNAviruses) (Herniou et al., 2001, 2003). Complete genome alignmentswere built in order to study genome structure and conservation (i.e.synteny) among the baculoviruses. Syntenic maps for genomes frommembers of group II were built with the Artemis comparative tool, byplotting the high similarity scoring pairs along pairs of genomes thatwere obtained with TBLASTX (using default settings) using the proteinsencoded in all six reading frames.

RESULTS AND DISCUSSION

Nucleotide sequence analysis

The complete SfMNPV-19 genome comprises 132 565 bpand a total of 141 ORFs encoding putative proteins of morethan 50 aa (Fig. 1). In agreement with the convention forbaculoviruses (Vlak & Smith, 1982), the polyhedrin genewas designated ORF number 1 and the putative codingregions were numbered sequentially in the clockwiseorientation. The adenine of the start codon of thepolyhedrin gene was assigned base number 1. Seventy-fourORFs (52.5 %) were in the clockwise direction and 67(47.5 %) in the opposite direction (Fig. 1).

The 29 genes shared by all baculovirus genomes (Herniouet al., 2003; Garcia-Maruniak et al., 2004) were also foundin the SfMNPV-19 genome. The phenogram shown inFig. 2 summarizes the relationship of SfMNPV-19 to 42

other complete genomes of baculoviruses. The topology ofthe genome phenogram was congruent with trees obtainedfrom individual genes (Zanotto et al., 1993) or trees forconcatemers of the 29 baculovirus conserved genes(Garcia-Maruniak et al., 2004) and was obtained with allthe clustering methods used. Nevertheless, we preferred toconsider the weighbor tree, since this method allows formore precise branch length estimates (Bruno et al., 2000).The phenogram had the SfMNPV-19 and the SfMNPV-3AP2 clustering with the NPVs that infect S. exigua(SeMNPV), A. segetum (AgseMNPV-A) and, to a lesserextent, to the two NPVs that infect M. configurata(MacoNPV-A and B) (Fig. 2). Moreover, as observed inprevious gene-based phylogenies (Zanotto et al., 1993;Bulach et al., 1999; Herniou et al., 2003; Garcia-Maruniaket al., 2004; Jehle et al., 2006) and genome-basedphylogenies (Oliveira et al., 2006), branch lengths leadingto group II NPV in Fig. 2 were longer than those for groupI NPV. As a consequence, the monophyly of group IIappeared to be supported by a limited set of shared traits.

The G+C content of the SfMNPV-19 genome was40.26 mol%. Similar G+C content values were observedfor the genomes of other members of the SfMNPV-19cluster. In these viruses the G+C values range from40 mol% in MacoNPV-B (Li et al., 2002) to 44 mol% inSeMNPV (IJkel et al., 1999). The variation of the G+Ccontent along the SfMNPV-19 genome is shown in Fig. 1.Computer-generated cleavage patterns for HindIII andBamHI were in agreement with the restriction patterngenerated in the laboratory (data not shown). Nevertheless,the analysis of 6596 bases (approx. 5 % of the totalgenome) whose sequence was derived from two or moredifferent cloned fragments revealed that a total of onlyeight bases (0.12 %) had nucleotide polymorphism, whichwas within an acceptable sequencing error level (99.88 %unambiguous nucleotides) and suggested that our assem-bly reflects good depiction of a stable and homogeneousSfMNPV genome. Moreover, the comparison of genecontent with that of closely related NPVs did not showdiscrepancies that could be explained by genetic diversityin the DNA source.

Homologous repeat sequences

Canonical homologous regions (hrs) are made of directrepeat sequences with imperfect palindromes. The baculo-virus hrs have been implicated in important functions suchas origins of DNA replication (Leisy & Rohrmann, 1993;Leisy et al., 1995; Kool et al., 1995) and transcriptionalenhancement (Guarino et al., 1986; Theilmann & Stewart,1992). A total of eight hrs were found in the SfMNPV-19genome (Supplementary Table S1, available with the onlineversion of the paper). The main feature of the hrs wasimperfect palindromic repeats of 44 bp, which were similarto palindromic repeats found in the hrs of SeMNPV andAgseMNPV (Broer et al., 1998; Jakubowska et al., 2006).The SfMNPV-19 hrs harboured from one up to four of

J. L. C. Wolff and others

1204 Journal of General Virology 89

these imperfect palindromes. The consensus sequence ofthe 18 palindromes found in the hrs was: CCATGTTTG-CTTTCGGCGAAGTGTTTCGCTGAAAGCAAACATCA.

A non-hr origin of replication has been identified inSeMNPV (Heldens et al., 1997). Sequences that resemblethe SeMNPV non-hr origins of replication (ORI) werefound in the intergenic region between ORFs 83 and 84 ofSfMNPV-19.

Comparison of SfMPV-19 with other SfMNPVisolates

The first physical map of an SfMNPV isolate was done by Lohet al. (1981). A more detailed physical map was presented in1984 in a study that analysed seven plaque-purified variants

(Maruniak et al., 1984). The computer-generated digestionthe SfMNPV-19 genome revealed a profile that differed fromthose of the previous investigations in terms of number andsize of fragments generated (data not shown). Moreover, thegenome of SfMNPV-19 was 10.9 kbp larger than theestimated size of the SfMNPV-2 clone (Maruniak et al.,1984). More recently, the physical and partial genetic map ofa Nicaraguan SfMNPV isolate was constructed and the dataobtained showed a genome of approximately 129.3 kbp(Simon et al., 2005a). Comparison with SfMNPV-19revealed differences in terms of number and position ofrestriction sites along the genomes. Finally, the sequence ofSfMNPV isolate 3AP2 (GenBank accession no. EF035042)became available as we were completing the genome ofSfMNPV-19. Alignment of the genomic sequence fromisolates SfMNPV-19 and 3AP2 with CLUSTAL W confirmed

Fig. 1. Circular map of the SfMNPV-19 genome. Arrows indicate direction of transcription and relative size of the ORFs. hrs

sequences and their positions are indicated by grey rectangles. The inner black line shows the G+C content (mol%) wherepeaks above the centre line correspond to regions that have more than 50 mol% G+C content. The figure was generated usingthe CGView program (Stothard & Wishart, 2005).



that the two viruses were variants of the same species (datanot shown). Major differences were limited to regions thathad deletions and insertions. The largest of these wereobserved around the egt gene. In the 3AP2 isolate, there wasa 1427 nt deletion that removed the amino terminal end ofthe putative EGT protein and the carboxy terminal end ofthe ORF that corresponds to SfMNPV-19 ORF 26 (data notshown).

Another evident difference was observed in the putativedutpase gene. In the SfMNPV-19 genome, ORF 53 encodesa putative dUTPase protein that is 21 aa shorter than thatencoded by isolate 3AP2, due to differences in the 59 end ofthe genes. A comparison with the putative dUTPase

proteins encoded by SeMNPV and AgseMNPV-A showedthat the longer version of the gene was shared by the mostclosely related members of the SfMNPV cluster (data notshown). It is noteworthy that the non-coding regionlocated immediately upstream to the SfMNPV-19 dutpasegene was poorly conserved, suggestive of a mutation hotspot. Other differences were also observed in the hr5 regionthat was longer in isolate 3AP2 by one repeat. Moreover,these two genomes differed in a series of point mutationsand short deletions or insertions that added or removedcodons. Computer-generated restriction profiles forSfMNNPV isolates 19 and 3AP2 were distinct, since theformer had 19 HindIII sites and the latter had only 17 (datanot shown).

Fig. 2. Tree based on the percentage of the reciprocal of the median bit-score values (S9) between all pairs drawn from 43complete baculovirus genomes (i.e. pairwise distances, see text). The dendrogram, obtained with the weighbor program (Brunoet al., 2000), was congruent to those obtained with other clustering methods (neighbour-joining, UPGMA and Fitsch and Kitschmethods from the PHYLIP package). Lepidopteran baculoviruses are split into three groups: groups I and II NPVs and GVs. TheHymenopteran and Dipteran baculoviruses formed their own branches. Abbreviations of baculovirus names are given in the textor in Supplementary Table S2 (available with the online version of this paper). The numbers of independent changes in terms ofgene gain and loss events are shown at each node and also at the final ramification of the tree. The ten different lineagesidentified in group II are indicated.



Comparison of SfMNPV-19 genome with those ofSeMNPV and AgseNPV-A

SfMNPV-19, SeMNPV, AgseNPV-A, MacoNPV-A andMacoNPV-B formed a closely related cluster in ourphylogenetic tree (Fig. 2). Within this group, SfMNPV-19, SeMNPV and AgseNPV-A were very similar in terms ofgene content, genomic organization and conservation. Forinstance, SfMNPV-19 had 125 and 129 orthologues in thegenomes of SeMNPV and AgseMNPV-A, respectively.

The average value for the identities between genes fromSfMNPV-19 and SeMNPV was 67 % (SD516 %). However,several structural proteins (polyhedrin, P45/P48, ODV-E25, ODV-EC27, GP41, ODV-E18, VP39 capsid, F protein,P74, PIF-2 and PIF-3) had identity values above 83 %.These results were in agreement with those obtained forother closely related baculoviruses (Gomi et al., 1999; Li etal., 2002; Oliveira et al., 2006). The conservation ofstructural proteins may indicate that functional constraintsimpose strong stabilizing (i.e. negative) selection on thesegenes. Similarly, the proteins that make the DNA-dependent RNA polymerase (LEF-4, LEF-8, LEF-9 andP47) and two proteins involved in viral DNA replication(DNA polymerase and helicase) were also highly conserved.It is also noteworthy that several non-essential proteinswith auxiliary functions (cathepsin, chitinase, superoxidedismutase, EGT and 38K hydrolase) displayed levels ofsequence identity above 83 %, suggesting that their genesmay be under strong negative selection.

SfMNPV-19 lacked 14 ORFs present in SeMNPV(Supplementary Table S1, available with the online versionof this paper). In general, the SfMNPV-19 genome haddifferent genes in place, suggesting that SfMNPV-relatedviruses present considerable genomic plasticity via genereplacement. However, ORFs Se39 and 49 (SupplementaryTable S1) were missing and were not replaced in SfMNPV-19. Nine of the unique SeMNPV ORFs were found in twoORF clusters, one from Se20 to 24 and another from Se83to 86. The SfMNPV-19 putative lef-7 gene (Sf20) washomologous with two SeMNPV ORFs, Se17 and Se18(Supplementary Table S1). One bro gene (Sf68) wasidentified in SfMNPV-19 (Supplementary Table S1),whereas the SeMNPV genome did not have a single brogene (IJkel et al., 1999). Likewise, the SfMNPV-19 genomecontained 24 fewer ORFs than the AgseMNPV-A genome(Supplementary Table S1). Three of these (Agse75, Agse76and Agse128) are putative vef genes, which encode virus-enhancing factors (Jakubowska et al., 2006). The AgseNPV-A genome has four copies of the bro genes, named bro-a tobro-d (Jakubowska et al., 2006). Of these, only bro-b had ahomologue in the SfMNPV-19 genome (SupplementaryTable S1). The genomes of both SfMNPV-19 and SeMNPVhad two copies of the odv-e66 gene (Sf56 and Sf114), as hasbeen observed previously (Simon et al., 2005a). However,AgseMNPV-A had only one complete copy of odv-e66(Agse125) and this gene had 49 % sequence identity to itsSfMNPV-19 equivalent (Supplementary Table S1). The

odv-66 gene encodes a structural protein involved in theformation of occluded virus envelope (Hong et al., 1997).Two copies of the p26 gene were found in SfMNPV-19(Sf85 and Sf129) and in the genomes of SeMNPV andAgseMNPV-A (Supplementary Table S1).

Interestingly, the genomes of both SeMNPV andAgseMNPV-A carry two ribonucleotide reductase genes,rr1 and rr2b (IJkel et al., 1999; Jakubowska et al., 2006).The rr1 gene is usually linked to the presence of either rr2aor rr2b (Herniou et al., 2003). However, only the rr1 genewas found in SfMNPV-19 and it had 34 and 33 % identityto the rr1 genes of SeMNPV and AgseMNPV-A, respect-ively. This value is considerably lower than the identitiesobserved for other closely related viruses that carryribonucleotide reductase genes (Li et al., 2002;Jakubowska et al., 2006). For instance, the identity levelfor the rr1 gene of AgseMNPV-A and those of SeMNPVand MacoNPV-B are 58 and 60 %, respectively(Jakubowska et al., 2006). It is possible that the loss ofone of the two copies of rr2 caused rr1 to experienceelevated rates of substitution.

ORFs with no homology to baculovirus genes

Most unique genes identified in SfMNPV-19 were located inregions that also had unique genes in the other viruses of thecluster, suggesting that these regions experience increasedgenetic variation. A total of 11 unique ORFs were identifiedin the genome of SfMNPV-19 (Sf5, Sf6, Sf7, Sf10, Sf31, Sf42,Sf43, Sf46, Sf54, Sf84 and Sf95). The size of these ORFsvaried considerably, ranging from 52 to 484 aa, and seven ofthem were larger than 100 aa (Supplementary Table S1).ORFs Sf31 (121 putative aa), Sf42 (395 putative aa), Sf43(263 putative aa) and Sf95 (166 putative aa) had knownbaculovirus promoter motifs upstream from their putativeinitiation codon and polyadenylation signals at their 39 end(data not shown). Sf5 was 88 aa long and displayed somehomology to the putative HOAR protein gene fromMacoNPV-A (with a BLAST e-value of 0.059). TheSfMNPV-19 hoar gene (Sf6) encoded a putative proteinthat was at least 130 aa shorter than the HOAR proteinsfrom other baculoviruses. As these two ORFs were conservedin both SfMNPV-19 and the 3AP2 isolate, it seems that thehoar gene was split into two in SfMNPV. It also came to ourattention that the region between the hoar and the odv-e56(Sf08) putative genes was highly variable among themembers of group II NPV. In SeMNPV, AgseMNPV-Aand Trichoplusia ni SNPV (TnSNPV), the genomic regionbetween the hoar and the odv-e56 is occupied either byunique ORFs or by poorly conserved ORFs (IJkel et al., 1999;Willis et al., 2005; Jakubowska et al., 2006).

Genomic organization, gene content analysis andevolution in group II NPVs

The tree for 43 complete genomes of baculoviruses isolatedfrom Lepidoptera (39), Hymenoptera (3) and Diptera (1)



(Fig. 2) was similar to others obtained previously for singlegenes (Zanotto et al., 1993), for a set of conservedorthologues (Herniou et al., 2001, 2003) and for completegenomes (Garcia-Maruniak et al., 2004; Oliveira et al.,

2006). As expected, SfMNPV was placed within group IINPV (Fig. 2). The subdivisions of group II into two(Zanotto et al., 1993) or three subgroups (Bulach et al.,1999) has been proposed based on the analysis of the

Fig. 3. Syntenic maps of the genomes from the SfMNPV cluster (a) and of the genomes representing ten different lineages ofgroup II NPV (b). Each genome is represented by a solid line within which the positions of the nucleotides are shown. Thegenomes’ starting positions are the polyhedrin genes, as in their GenBank files. The viruses compared are shown on the left sideof the figure. Syntenic regions, indicated by stripes connecting the genomes, are areas of sequence conservation, irrespectiveof the actual coding regions, as found with the TBLASTX program (minimum cut-off size of 20 bp with a minimum percentageidentity of 50 %). The map was obtained with the Artemis comparative tool (version 8.0).



polyhedrin and the DNA polymerase genes. Thesesubdivisions proved to be insufficient given the genomicinformation now available. The syntenic maps generated inthis study showed extensive collinearity and gene contentconservation between the genomes of SfMNPV, SeMNPV,AgseMNPV-A and, to a lesser extent, to both MacoNPV-Aand B (Fig. 3a). Similar results were observed in syntenicmaps for members of two other individual clusters withingroup II [one including the viruses of the Heliothis and theother including Chrysodeixis chalcites NPV (ChchNPV) andTnSNPV, data not shown]. On the other hand, theinclusion of one member of each of these three clustersand the other seven viruses of the group always resulted inextensive lack of collinearity and gene conservation(Fig. 3b). This lack of conservation of gene order andcontent between members of the different lineages of groupII contrasted markedly with group I NPV (Oliveira et al.,2006) and invited further analyses.

The topology of the weighbor tree was also used to inferthe dynamics of gene gain and loss as the mostparsimonious unequivocal reconstruction (MPR) alongthe radiation of the baculoviridae. The striking result wasthat group II had only five shared defining genes (i.e.synapomorphies) while group I had 25 and GV had 27(Fig. 2). The NPV had 22 synapomorphies and theLepidopteran baculovirus (NPV and GV together) had 18synapomorphies. These critical differences in the numberof shared MPR are not detected in phylogenetic analysesthat include only the core set of shared baculovirus genes(Herniou et al., 2003; Garcia-Maruniak et al., 2004; Jehleet al., 2006; Oliveira et al., 2006). Our current analysesindicated, for instance, that the LdMNPV had 52apomorphies (i.e. genes unique to that lineage), whichquestions its inclusion in any of the current NPV groups. Asimilar argument could be extended to the Leucaniaseparata NPV (LeseNPV), with 51 apomorphies, to theSpltNPV, with 49 apomorphies, and possibly to the threeHeliothis NPVs, with 29 synapomorphies (Fig. 2). Theseanalyses led to the conclusion that, within group II, therewere at least ten deep lineages that were as divergent fromeach other as from group I (Fig. 2).

Given the complex pattern of gene gain and loss, it becameinteresting to ask how the gene gain and loss processcorrelated along the genome tree that had branch lengthsestimated as median values of 1/S9 (i.e. reciprocal of themedian values of the bit-scores on pairwise genomecomparisons with TBLASTX) (Oliveira et al., 2006).Interestingly, more than 60 % of the data had a power-law relation expressing the positive dependence of thenumber of gene gain (g) on branch length (b) in theweighbor tree (g52537.11b10.870, r250.54), whereas lossevents (l) were less frequent and had less dependence(12 %) on branch lengths (l529.96b10.409, r250.12). Forthe gain process, the signature of the power-law relationwas also obtained by a good linear fitting of a log-log plot(logg51.063logb+3.871, r250.61). In this regard, it isnecessary to point out that BlastPhen comparisons

considered only traits shared among pairs of genomes,and not gain and loss events used to estimate MPRs.Therefore, the dependency of values of MPRs on theestimated branch lengths did not entail circular reasoning,since both sets of values were obtained from fairlyindependent methods of measurement. Nevertheless, itneeds to be acknowledged that our method could inflatedistances among closely related viruses, due to the largeramount of high-score pairs and polymorphisms found innon-coding regions during the pairwise TBLASTX search.This effect tends to disappear as genomes lose similarity intheir non-coding regions (i.e. among less related genomes).However, we do not find that our main results are wrongbecause (i) the tree topology in Fig. 2 is congruent withthat obtained previously from the conserved core set ofgenes and (ii) there was no change in the topology whenthree viruses closely related to AcMNPV and SfMNPV-19(i.e. PlxyMNPV, RoMNPV and SfMNPV-3AP2) weretaken out from the MPR/branch length comparison andthe power-law signature was altered by only 6.7 %, which isalmost exactly the amount of data (6.9 %) taken out of theanalyses.

In summary, an implication of our findings is that, to aconsiderable extent, the process of gene gain in baculo-viruses appears to be temporally organized (since branchlengths increase in time) following a power-law, whereasgenes loss events appear to be less frequent and notcorrelated with time. Nevertheless, most baculoviruslineages we studied are apparently increasing in genomecomplexity with time, since gene losses are less frequentand more irregular. Similar accretion processes of auxiliarygene functions in large DNA viruses were observed for poxviruses (McLysaght et al., 2003) and herpesviruses(Montague & Hutchison, 2000). These findings positrenewed challenge to our understanding of the evolutionof baculovirus genome organization and taxonomy.

ACKNOWLEDGEMENTS

This research work was funded by Fundacao de Amparo a Pesquisa do

Estado de Sao Paulo (FAPESP), project 04/09389-9. J. L. C. W. holds

an FAEP scholarship (Fundacao de Ampara ao Ensino e a Pesquisa da

Universidade de Mogi das Cruzes). P. M. de A. Z. holds a CNPq PQ

Research Scholarship and was funded by FAPESP, project 00/04205-6.

J. V. de C. O. holds an FAPESP scholarship, project 04/12456-0. We

thank Dr Elisabeth Herniou, Dr Alejandra Garcia-Maruniak and Dr

Hilary Lauzon for providing updates for the baculovirus gene content

table and Dr Jenny Cory for allowing the use of the available

sequences on the McplvNPV genome.

REFERENCES

Afonso, C. L., Tulman, E. R., Lu, Z., Balinsky, C. A., Moser, B. A.,Becnel, J. J., Rock, D. L. & Kutish, G. F. (2001). Genome sequence of a

baculovirus pathogenic for Culex nigripalpus. J Virol 75, 11157–11165.

Ahrens, C. H., Russell, R. L., Funk, C. J., Evans, J. T., Harwood, S. H. &Rohrmann, G. F. (1997). The sequence of the Orgyia pseudotsugata



multinucleocapsid nuclear polyhedrosis virus genome. Virology 229,

381–399.

Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J.(1990). Basic local alignment search tool. J Mol Biol 215, 403–410.

Ayres, M. D., Howard, S. C., Kuzio, J., Lopez-Ferber, M. & Possee,

R. D. (1994). The complete DNA sequence of Autographa californica

nuclear polyhedrosis virus. Virology 202, 586–605.

Barreto, M. E., Guimaraes, C. T., Teixeira, F. F., Paiva, E. & Valicente,

F. H. (2005). Effect of Baculovirus spodoptera isolates in Spodoptera

frugiperda (J. E. Smith) (Lepidoptera: Noctuidae) larvae and their

characterization by RAPD. Neotrop Entomol 34, 67–75.

Berretta, M. F., Rios, M. L. & de Sciocco, C. A. (1998).Characterization of a nuclear polyhedrosis virus of Spodoptera

frugiperda from Argentina. J Invertebr Pathol 71, 280–282.

Blissard, G. W. & Rohrmann, G. F. (1990). Baculovirus diversity and

molecular biology. Annu Rev Entomol 35, 127–155.

Blissard, G. W., Black, B., Crook, N. E., Keddie, B. A., Possee, R. D.,Rohrmann, G. F., Theilmann, D. A. & Volkman, L. E. (2000). Family

Baculoviridae. In Virus Taxonomy: Seventh Report of the International

Committee on Taxonomy of Viruses, pp. 195–202. Edited by M. H. V.

Van Regenmortel, C. M. Fauquet, D. H. L. Bishop, E. B. Carstens, M.

K. Estes, S. M. Lemon, J. Maniloff, M. A. Mayo, D. J. McGeoch, C. R.

Pringle & E. B. Wickner. San Diego, CA: Academic Press.

Broer, R., Heldens, J. G. M., Strien, E. A., Zuidema, D. & Vlak, J. M.(1998). Specificity of multiple homologous genomic regions in

Spodoptera exigua nucleopolyhedrovirus DNA replication. J Gen Virol

79, 1563–1572.

Bruno, W. J., Socci, N. D. & Halpern, A. L. (2000). Weighted neighbor

joining: a likelihood-based approach to distance-based phylogeny

reconstruction. Mol Biol Evol 17, 189–197.

Bulach, D. M., Kumar, C. A., Zaia, A., Liang, B. & Tribe, D. E. (1999).

Group II nucleopolyhedrovirus subgroups revealed by phylogenetic

analysis of polyhedrin and DNA polymerase gene sequences.

J Invertebr Pathol 73, 59–73.

Carver, T. J., Rutherford, K. M., Berriman, M., Rajandream, M. A.,Barrell, B. G. & Parkhill, J. (2005). ACT: the Artemis comparison tool.

Bioinformatics 21, 3422–3423.

Duffy, S. P., Young, A. M., Morin, B., Lucarotti, C. J., Koop, B. F. &Levin, D. B. (2006). Sequence analysis and organization of the

Neodiprion abietis nucleopolyhedrovirus genome. J Virol 80, 6952–6963.

Felsenstein, J. (1989). PHYLIP – phylogeny inference package (version

3.2). Cladistics 5, 164–166.

Garcia-Maruniak, A., Maruniak, J. E., Zanotto, P. M. A., Doumbouya,

A. E., Liu, J. C., Merritt, T. M. & Lanoie, J. S. (2004). Sequence analysis

of the genome of the Neodiprion sertifer nucleopolyhedrovirus. J Virol

78, 7036–7051.

Gomi, S., Majima, K. & Maeda, S. (1999). Sequence analysis of

the genome of Bombyx mori nucleopolyhedrovirus. J Gen Virol 80,

1323–1337.

Gonzalez, M. A., Smith, G. E. & Summers, M. D. (1989). Insertion of

the SfMNPV polyhedrin gene into an AcMNPV polyhedrin deletion

mutant during viral infection. Virology 170, 160–175.

Guarino, L. A., Gonzalez, M. A. & Summers, M. D. (1986). Complete

sequence and enhancer function of the homologous DNA regions of

Autographa californica nuclear polyhedrosis virus. J Virol 60, 224–229.

Hayakawa, T., Ko, R., Okano, K., Seong, S. I., Goto, C. & Maeda, S.

(1999). Sequence analysis of the Xestia c-nigrum granulovirus

genome. Virology 262, 277–297.

Heldens, J. G., Broer, R., Zuidema, D., Goldbach, R. W. & Vlak, J. M.

(1997). Identification and functional analysis of a non-hr origin of

DNA replication in the genome of Spodoptera exigua multicapsid

nucleopolyhedrovirus. J Gen Virol 78, 1497–1506.

Herniou, E. A., Luque, T., Chen, X., Vlak, J. M., Winstanley, D., Cory,J. S. & O’Reilly, D. R. (2001). Use of whole genome sequence data toinfer baculovirus phylogeny. J Virol 75, 8117–8126.

Herniou, E. A., Olszewski, J. A., Cory, J. S. & O’Reilly, D. R. (2003).The genome sequence and evolution of baculoviruses. Annu RevEntomol 48, 211–234.

Hong, T., Summers, M. D. & Braunagel, S. C. (1997). N-terminalsequences from Autographa californica nuclear polyhedrosis virus

envelope proteins ODV-E66 and ODV-E25 are sufficient to direct

reporter proteins to the nuclear envelope, intranuclear microvesiclesand the envelope of occlusion derived virus. Proc Natl Acad Sci U S A

94, 4050–4055.

IJkel, W. F., van Strien, E. A., Heldens, J. G., Broer, R., Zuidema, D.,Goldbach, R. W. & Vlak, J. M. (1999). Sequence and organization of

the Spodoptera exigua multicapsid nucleopolyhedrovirus genome.

J Gen Virol 80, 3289–3304.

Jakubowska, A. K., Peters, S. A., Ziemnicka, J., Vlak, J. M. & vanOers, M. M. (2006). Genome sequence of an enhancin gene-rich

nucleopolyhedrovirus (NPV) from Agrotis segetum: collinearity withSpodoptera exigua multiple NPV. J Gen Virol 87, 537–551.

Jehle, J. A., Blissard, G. W., Bonning, B. C., Cory, J. S., Herniou, E. A.,Rohrmann, G. F., Theilmann, D. A., Thiem, S. M. & Vlak, J. M. (2006).On the classification and nomenclature of baculoviruses: a proposal

for revision. Arch Virol 151, 1257–1266.

Knell, J. D. & Summers, M. D. (1981). Investigation of genetic

heterogeneity in wild isolates of Spodoptera frugiperda nuclear

polyhedrosis virus by restriction endonuclease analysis of plaque-purified variants. Virology 112, 190–197.

Kool, M., Ahrens, C. H., Vlak, J. M. & Rohrmann, G. F. (1995).Replication of baculovirus DNA. J Gen Virol 76, 2103–2118.

Lauzon, H. A. M., Lucarotti, C. J., Krell, P. J., Feng, Q., Retnakaran, A.& Arif, B. M. (2004). Sequence and organization of the Neodiprionlecontei nucleopolyhedrovirus genome. J Virol 78, 7023–7035.

Lauzon, H. A. M., Jamieson, P. B., Krell, P. J. & Arif, B. M. (2005). Geneorganization and sequencing of the Choristoneura fumiferana

defective nucleopolyhedrovirus genome. J Gen Virol 86, 945–961.

Leisy, D. J. & Rohrmann, G. F. (1993). Characterization of thereplication of plasmids containing hr sequences in baculovirus-

infected Spodoptera frugiperda cells. Virology 196, 722–730.

Leisy, D. J., Rasmussen, C., Kim, H. T. & Rohrmann, G. F. (1995). The

Autographa californica nuclear polyhedrosis virus homologous region

1a: identical sequences are essential for DNA replication activity andtranscriptional enhancer function. Virology 208, 742–752.

Li, L., Donly, C., Li, Q., Willis, L. G., Keddie, B. A., Erlandson, M. A. &Theilmann, D. A. (2002). Identification and genomic analysis of asecond species of nucleopolyhedrovirus isolated from Mamestra

configurata. Virology 297, 226–244.

Liu, J. C. & Maruniak, J. E. (1995). Nucleotide sequence and

transcriptional analysis of the gp41 gene of Spodoptera frugiperda

nuclear polyhedrosis virus. J Gen Virol 76, 1443–1450.

Loh, L. C., Hamm, J. J. & Huang, E. S. (1981). Spodoptera frugiperda

nuclear polyhedrosis virus genome: physical maps for restriction

endonucleases BamHI and HindIII. J Virol 38, 922–931.

Loh, L. C., Hamm, J. J., Kawanishi, C. & Huang, E. S. (1982). Analysis of

the Spodoptera frugiperda nuclear polyhedrosis virus genome byrestriction endonucleases and electron microscopy. J Virol 44, 747–751.

Maruniak, J. E., Brown, S. E. & Knudson, D. L. (1984). Physical maps

of SfMNPV baculovirus DNA and its genomic variants. Virology 136,221–234.



McLysaght, A., Baldi, P. F. & Gaut, B. S. (2003). Extensive gene gainassociated with adaptive evolution of poxviruses. Proc Natl Acad SciU S A 100, 15655–15660.

Montague, M. G. & Hutchison, C. A., III (2000). Gene contentphylogeny of herpesviruses. Proc Natl Acad Sci U S A 97, 5334–5339.

Oliveira, J. V. de C., Wolff, J. L. C., Garcia-Maruniak, A., Ribeiro, B. M.,de Castro, M. E. B., de Souza, M. L., Moscardi, F., Maruniak, J. E. &Zanotto, P. M. de A. (2006). Genome of the most widely used viralbiopesticide: Anticarsia gemmatalis multiple nucleopolyhedrovirus.J Gen Virol 87, 3233–3250.

O’Reilly, D. R., Miller, L. K. & Luckow, V. A. (1992). BaculovirusExpression Vectors: a Laboratory Manual. Salt Lake City, UT: W. H.Freeman.

Pertsemlidis, A. & Fondon, J. W., III (2001). Having a BLAST withbioinformatics (and avoiding BLASTphemy). Genome Biol 2,reviews2002.1–reviews2002.10.

Rutherford, K., Parkhill, J., Crook, J., Horsnell, T., Rice, P.,Rajandream, M. A. & Barrell, B. (2000). Artemis: sequencevisualization and annotation. Bioinformatics 16, 944–945.

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning:a Laboratory Manual, 2nd edition. Cold Spring Harbor, NY: ColdSpring Harbor Laboratory.

Shapiro, D. I., Fuxa, J. R., Braymer, H. D. & Pashley, D. P. (1991).DNA restriction polymorphism in wild isolates of Spodopterafrugiperda nuclear polyhedrosis virus. J Invertebr Pathol 58,96–105.

Simon, O., Chevenet, F., Williams, T., Caballero, P. & Lopez-Ferber, M.(2005a). Physical and partial genetic map of Spodopterafrugiperda nucleopolyhedrovirus (SfMNPV) genome. Virus Genes 30,403–417.

Simon, O., Gutierrez, S., Williams, T., Caballero, P. & Lopez-Ferber, M.(2005b). Nucleotide sequence and transcriptional analysis of the pifgene of Spodoptera frugiperda nucleopolyhedrovirus (SfMNPV). VirusRes 108, 213–220.

Sonnhammer, E. L. L. & Durbin, R. (1995). A dot-matrix programwith dynamic threshold control suited for genomic DNA and proteinsequence analysis. Gene 167, GC1–GC10.

Stothard, P. & Wishart, D. S. (2005). Circular genome visualizationand exploration using CGView. Bioinformatics 21, 537–539.

Tanada, Y. & Kaya, H. K. (1992). Insect Pathology. San Diego, CA:Academic Press.

Theilmann, D. A. & Stewart, S. (1992). Tandemly repeated sequenceat the 39 end of the IE-2 gene of the baculovirus Orgyia pseudotsugatamulticapsid nuclear polyhedrosis virus is an enhancer element.Virology 187, 97–106.

Tumilasci, V. F., Leal, E., Zanotto, P. M., Luque, T. & Wolff, J. L. (2003).Sequence analysis of a 5.1 kbp region of the Spodoptera frugiperdamulticapsid nucleopolyhedrovirus genome that comprises a func-tional ecdysteroid UDP-glucosyltransferase (egt) gene. Virus Genes 27,137–144.

Vlak, J. M. & Smith, G. E. (1982). Orientation of the genome ofAutographa californica nuclear polyhedrosis virus: a proposal. J Virol41, 1118–1121.

Willis, L. G., Siepp, R., Stewart, T. M., Erlandson, M. A. & Theilmann,D. A. (2005). Sequence analysis of the complete genome ofTrichoplusia ni single nucleopolyhedrovirus and the identification ofa baculoviral photolyase gene. Virology 338, 209–226.

Zanotto, P. M., Kessing, B. D. & Maruniak, J. E. (1993). Phylogeneticinterrelationships among baculoviruses: evolutionary rates and hostassociations. J Invertebr Pathol 62, 147–164.



Analysis of the genome of Spodoptera frugiperda nucleopolyhedrovirus (SfMNPV-19) and of the high genomic heterogeneity in group II nucleopolyhedroviruses

Documents