Sequence Analysis of the Genome of the Neodiprion sertifer Nucleopolyhedrovirus

The genome of the Neodiprion sertifer nucleopolyhedrovirus (NeseNPV),which infects the European pine sawfly, N. sertifer (Hymenoptera:Diprionidae), was sequenced and analyzed. The genome was 86,462bp in size. The C+G content of 34% was lower than that of themajority of baculoviruses. A total of 90 methionine-initiatedopen reading frames (ORFs) with more than 50 amino acids andminimal overlapping were found. From those, 43 ORFs were homologousto other baculovirus ORFs, and 29 of these were from the 30conserved core genes among all baculoviruses. A NeseNPV homologto the ld130 gene, which is present in all other baculovirusgenomes sequenced to date, could not be identified. Six NeseNPVORFs were similar to non-baculovirus-related genes, one of whichwas a trypsin-like gene. Only one iap gene, containing a singleBIR motif and a RING finger, was found in NeseNPV. Two NeseNPVORFs (nese18 and nese19) were duplicates transcribed in oppositeorientations from each other. NeseNPV did not have an AcMNPVORF 2 homolog characterized as the baculovirus repeat ORF (bro).Six homologous regions (hrs) were located within the NeseNPVgenome, each containing small palindromes embedded within directrepeats. A phylogenetic analysis was done to root the tree basedupon the sequences of DNA polymerase genes of NeseNPV, 23 otherbaculoviruses, and other phyla. Baculovirus phylogeny was thenconstructed with 29 conserved genes from 24 baculovirus genomes.Culex nigripalpus nucleopolyhedrovirus (CuniNPV) was the mostdistantly related baculovirus, branching to the hymenopteranNeseNPV and the lepidopteran nucleopolyhedroviruses and granuloviruses.

INTRODUCTION

Top

Introduction

Viruses in the family Baculoviridae are large, rod-shaped, andenveloped and have circular double-stranded DNA genomes. Theyare the largest and most widely studied group of insect viruses.This family is subdivided into two genera, the granuloviruses(GVs), with small occlusion bodies that generally contain asingle virion, and the nucleopolyhedroviruses (NPVs), with occlusionbodies that include multiple virions. The most widely studiedgroup of baculoviruses is the lepidopteran NPV, which has beendivided into groups I and II since early phylogenetic studiesusing the polyhedrin gene (81). The subdivision of NPVs I andII has been maintained when all the conserved genes in the baculoviruseshave been used in phylogenetic analysis (32, 60). The Neodiprionsertifer nucleopolyhedrovirus (NeseNPV) infects the Europeanpine sawfly from the order Hymenoptera. It has single nucleocapsidswithin an envelope, like the other NPVs that infect nonlepidopteraninsects (22).

To date, a total of 23 baculovirus genomes are available inGenBank, some of which are close variants. From these, 22 infectinsects from the order Lepidoptera and 1 infects mosquitoesfrom the order Diptera. The comparative study of these baculovirusgenomes reflects their diversity, with a size range from 99,657bp in Adoxophyes orana GV (AdorGV) (78) to 178,733 bp in Xestiac-nigrum GV (XcGV) (30); number of open reading frames (ORFs)from 109 in Culex nigripalpus NPV (CuniNPV) (1) to 181 in XcGV;number of homologous regions (hrs) from 17 in Spodoptera lituraNPV (SpltMNPV) (65) to none in Cydia pomenella GV (CpGV) andAdoxophyes orana GV (AdorGV) (53, 78). The number of conservedgenes reported among lepidopteran baculoviruses varies, as morebaculovirus genomes became available (31). However, when CuniNPVis included in the analysis, the number is reduced to 30 genes(32).

NeseNPV has been used to control the European pine sawfly, N.sertifer, one of the most harmful insects to the coniferousforests of the northern hemisphere (41). Sawfly larvae livegregariously and resemble caterpillars. Originally from Europe,N. sertifer was accidentally brought to North America around1925. Before the concern of viral registration, NeseNPV wassuccessfully used by foresters, Christmas tree growers, andprivate landowners. Later, the virus was registered as Neochek-Sand became available in the United States for use against theEuropean pine sawfly (34). Nucleopolyhedroviruses of sawflieshave been reported in 25 species (34). These viruses are highlyhost specific and, in contrast to most lepidopteran NPVs, replicateonly in the epithelial cells of the larval midgut (22). Infectionby the virus results in a reduction of larval feeding and, eventually,complete suppression of appetite long before death.

Few genes of NeseNPV had been previously sequenced (Liu andMaruniak, unpublished data; GenBank accession no. AF121349).Rohrmann et al. (72) published a description of the N-terminalsequence of the NeseNPV polyhedrin protein. The complete polyhedrinamino acid sequence was later obtained (Harris and Possee, unpublisheddata) and used in phylogenetic studies that placed NeseNPV asthe outgroup baculovirus (11, 81).

Insight into the origin and subsequent evolution of the familyBaculoviridae can be obtained by comparing the similaritiesand differences in the genomic contents and gene distributionof distantly related viruses. The comparison may also elucidatesome properties of the baculovirus morphogenesis, structure,or genomic information that are essential for effective propagationand infectivity. In this work, we describe the genome sequenceand the genetic organization of the hymenopteran baculovirusNeodiprion sertifer NPV.

MATERIALS AND METHODS

Top

Materials and Methods

Virus, DNA purification, cloning, and sequencing. Neodiprion sertifer NPV was obtained as a crude preparationof Neochek-S stocks from James Slavicek and John Podgewaite.The composition of this preparation was approximately 0.05%polyhedral inclusion bodies of NeseNPV and 99.95% milled N.sertifer insect parts. NeseNPV DNA was purified from this preparationas described by Maruniak (55). The dry preparation was homogenizedin a solution containing 10 mM Tris, 1 mM EDTA, 2% sodium dodecylsulfate (SDS), and 1% ascorbic acid. After centrifugation at10,000 x g for 10 min, the pellet was resuspended in 0.5 M NaCland then rinsed with water by another cycle of homogenizationand centrifugation. The sample, resuspended in water, was layeredon a 40 to 63% continuous sucrose gradient. The polyhedra thatbanded at approximately 55% was disrupted with a dilute alkalinesolution (0.1 M sodium carbonate, 0.17 M NaCl, 0.01 M EDTA)and layered on a 40 to 56% continuous sucrose gradient. Theband, formed at the top quarter of the tube, containing thesingle nucleocapsid virions, was treated with 1% SDS and 1 mg/mlproteinase K. After two phenol and three ether extractions,the viral DNA was dialyzed against TE (0.01 M Tris, 0.001 MEDTA). Approximately 4 µg of DNA was obtained per gramof Neochek-S preparation. The viral DNA was digested with HindIIIand EcoRI separately and cloned in pGEM 7Z plasmid (Promega).This resulted in the construction of NeseNPV HindIII and a partialEcoRI libraries. Initially, the sequencing of each NeseNPV clonewas done using the T7 and SP6 primers from the vector. Then,EZ::TN transposons (Epicentre) were inserted in the clones torandomly distribute primer binding sites (forward and reverse)present in the transposon. Colonies carrying the antibioticresistance trait from the transposons were screened using severalrestriction endonucleases (RENs) to ensure that only one transposonwas inserted per clone and to map its location within the viralfragment. This approach facilitated the assembly of problematicregions, such as highly repetitive sequences or those suspectedof having secondary structures. Specific primers were synthesized,and additional sequences were obtained by primer walking tofill in gaps. Other strategies to sequence the genome comprisedsubcloning and PCR amplification of some regions. Sequencingreactions were done in a ThermoCycler machine (MJ Research)with the ABI PRISM BigDye Terminator sequencing ready reactionkit (Applied Biosystems) and sent to the University of FloridaInterdisciplinary Center for Biotechnology Research to be runin an automated DNA sequencer.

Multiplex PCR. Because no clonal isolate of NeseNPV was available, a multiplexPCR procedure was used to arrange the cloned fragments, to determineif the whole genome was cloned, and to detect genome variants.The physical map of the NeseNPV DNA was assembled with a modificationof the optimized multiplex PCR method (74). The genomic viralDNA was used as a template for PCRs that would allow the correctorder of the clones in the genome to be determined after sequencingthe PCR products. Using the initial sequence data from all theHindIII clones, unique primers were synthesized within the first300 bp that annealed with the DNA fragment toward the end ofthe fragment (end-out primers). The primers were first testedby sequencing the corresponding clone. The primers were thencombined in groups of three per pool in such a way that primersfrom opposite ends of the same fragment were together in thesame pool. All possible pairs of pool combination were usedin PCRs using the genomic NeseNPV DNA as the template underthe following conditions: 1.5 mM MgCl₂, 0.5 mM deoxynucleosidetriphosphates, 1 ng of genomic NeseNPV DNA, 10 pmol of eachprimer, 1 U of Elongase (Invitrogen), for a final 25-µlreaction volume. Cycling conditions were those described byMoraes and Maruniak (59). Two microliters of each PCR productwere electrophoresed on a 0.7% agarose gel. When no PCR productswere obtained, it meant that none of the six primers from thetwo pools tested amplified the DNA under those conditions and,hence, their corresponding fragments were in the wrong orientationor too distantly separated in the genome to amplify the DNA.When one DNA band was obtained, the PCR product was sequencedusing the pooled primers previously used for the PCR amplification.When one DNA band was seen in all the lanes containing a particularpool of primers, the common pool of primers was used to sequencethat PCR product. If more than one DNA band was obtained ina single lane, the pools involved were further divided in groupsof two primers each and the PCR was repeated. The sequencesobtained were compared to previous data from the HindIII library.Overlapping sequences permitted the location of the correctorder of the HindIII fragments on the NeseNPV physical map.The order of the HindIII fragments was further confirmed byrepeating the PCR amplification using the genomic DNA as templateand only the primer pair from adjacent fragments identifiedby the multiplex method.

DNA sequence analysis. The complete NeseNPV genome was assembled with the Sequencher4.1 program (Gene Codes) using the sequence data from everyindividual clone and the multiplex PCR experiments. Every baseof each DNA strand was sequenced a minimum of three times; however,due to the fact that the location of every sequence had beenpreviously determined, either by transposon mapping or by primerwalking, the assembly process was not random. The ORFs encoding50 amino acids or more and showing minimum overlapping, as hasbeen the criteria for baculoviruses (5), were found using theORF Finder program (http://www.ncbi.nlm.nih.gov/gorf/gorf.html)(77). BLASTP 2.2.6 (http://www.ncbi.nlm.nih.gov/BLAST) (3),conserved domain database (CDD) (http://www.ncbi.nlm.nih.gov/Structure/cdd/cdd.shtml)search, run by default in parallel with BLASTP and PSI-BLAST(4) programs and Smith-Waterman similarity search (http://coe02.ucalgary.ca/algo-sw/sw_aa.shtml),were used to compare all of the possible ORFs to the amino acidsequence database. The following criteria were used to acceptNeseNPV ORFs as baculovirus homologs: (i) During BLASTP comparison,an E value of 0.1 or less was accepted; (ii) when the BLASTlist showed similarity with related ORFs from several baculoviruses,it was considered a NeseNPV homolog even if the E value of somemembers of the list were higher than 0.1; (iii) if an hr overlappedwith an ORF, that ORF was not considered for the analysis. ForNeseNPV ORFs that showed random matches with other baculovirusORFs (represented by an E value over 0.1), forced alignmentswere performed to reveal if there was significant homology notevident by BLAST or Smith-Waterman similarity search (http://coe02.ucalgary.ca/).This was also done with NeseNPV ORFs that had signal peptidesand transmembrane domains in order to search for an ld130 homologcandidate. CLUSTALW (http://www.ebi.ac.uk/clustalw/) was usedfor this purpose and for several other amino acid alignments.Some ORF overlapping was permitted, especially when both ORFswere homologous to other baculoviruses. The putative codingregions were numbered as NeseNPV ORFs. The sequences 160 nucleotidesupstream of each ORF start codon were analyzed for early andlate transcriptional motifs. Late motifs contained the sequence(A/G/T)TAAG (71). Early motifs were defined as either TATA alone(designated "e" in Table 1) or TATA with the CA(T/G)T sequence,20 to 40 nucleotides downstream ("E" in Table 1). To find possibletransmembrane domains in the NeseNPV putative proteins, theTMHMM server 2.0 was used (http://www.cbs.dtu.dk/services/TMHMM-2.0).To examine the NeseNPV ORFs for N-terminal signal peptides,SignalP 2.0 prediction server (61) was used (http://www.cbs.dtu.dk/services/SignalP).Tandem Repeats Finder (http://c3.biomath.mssm.edu/trf.html)was used to locate the hr's (6). The alignment parameters usedwere 2, 7, and 7 for match, mismatch, and indels, respectively,a minimum alignment score of 50 to report repeats, and a maximumperiod size (repeat size) of 500. Several programs from theGenetics Computer Group (Wisconsin Package Version 10.2) (18)were used to locate the palindrome sequences of NeseNPV hr'sand to determine the G+C ratio of each hr and of the entireNeseNPV genome

View this table:
[in this window]
[in a new window]
TABLE 1. NeseNPV ORFs^a

Phylogenetic analysis. In order to root a tree of the 24 baculoviruses sequenced todate, a DNA polymerase phylogeny was inferred using the dna-polgenes from those baculoviruses and from 23 other organisms representingdifferent Phyla which had a high score and the lowest e-valuesin tblastx searches against baculovirus DNA polymerases. Theunclassified double-stranded DNA Heliothis zea virus 1 (Hz-1)(GenBank accession no. NC_004156) and several other DNA virusDNA polymerases were included for comparison. A maximum-likelihoodtree from a data set with 2,397 amino acid sites including 47taxa was calculated with the parallel version 5.2 of the TREE-PUZZLEprogram (http://www.tree-puzzle.de/) using default settingsbut assuming a variable rate of evolution among sites followinga gamma distribution approximated by 8 categories, with theshape parameter (alpha) estimated from the data. Likewise, tohave the maximum evidence for a baculovirus phylogeny, a singledata set with 16,803 amino acid sites (proteon) comprising 29conserved genes found in 24 baculovirus genomes was also analyzedwith TREE-PUZZLE. The DNA polymerase tree adjacency patternwas then used to orient in time (by rooting) the protein tree.

GeneParity analysis. The gene order of the NeseNPV genome was compared to those ofsix other baculovirus genomes: AcMNPV (representing group Ilepidopteran NPVs), LdMNPV, MacoMNPV-A and HzSNPV (group IIlepidopteran NPVs), CpGV (lepidopteran GV), and CuniNPV (dipteranNPV). For this, a modified version of the GeneParity plot method(33) was used. Each complete genome was compared individuallyto that of NeseNPV. Minor modifications were done to two ofthe genomes, relative to the original publication. For AcMNPV(5), the comparison was made starting with the polyhedrin gene.For CuniNPV (1), the order of lef-5, 38K, ac96, and helicasein the genome was used, not the ORF number. A gene cluster wasdefined as any group of genes (two or more) that appeared consecutivelyin NeseNPV and in the comparative genome. However, some virusesdo not have all of the genes in a cluster because they may bepresent in another location of the genome or just absent. Therefore,relaxed clusters were also determined as those in which at leasttwo genes were consecutive in both genomes compared, but a certainnumber of genes (a maximum of five in one case for CuniNPV)may have been inserted between genes that appear consecutivelyin NeseNPV.

Nucleotide sequence accession number. The NeseNPV genome sequence has been deposited in GenBank underaccession no. AY430810.

RESULTS AND DISCUSSION

Top

Results and Discussion

Multiplex PCR. Due to the lack of an insect colony or insect cell culture susceptibleto NeseNPV, the DNA obtained for this project was not from ahomogeneous isolate. The approach to circumvent this problemwas to sequence the ends of all the NeseNPV HindIII and EcoRIclones obtained. After confirming that the terminal sequencesof the submolar bands were nearly the same as those of the majorbands, we continued sequencing the major bands. Some singlenucleotide polymorphisms were detected while sequencing severalclones of the same fragment and also when comparing the sequencesobtained from the multiplex PCR adjacent to the HindIII recognitionsites. From a total of 10,900 bases (13% of the total genome)analyzed for nucleotide differences from the consensus sequence,only 68 bases (0.6%) had nucleotide polymorphisms (data notshown). The colinear position and orientation of the HindIIIclones in the NeseNPV physical map were assembled using theMultiplex PCR data (Fig. 1). A 330-bp fragment (HindIII-M) thatwas not originally cloned was sequenced and mapped during themultiplex PCR procedure. In addition, a 4.5-kb region foundto be unclonable, located inside HindIII-D fragment, was PCRamplified and then sequenced. Multiplex PCR has been originallyused to rapidly close gaps in whole-genome shotgun sequencingprojects (74). It proved to be a very convenient and reliablemethod to ensure the correct location and position of the NeseNPVcloned fragments. It also allowed us to detect genetic variation,which is present in NeseNPV.

View larger version (22K):
[in this window]
[in a new window]
FIG. 1. Linear map of the HindIII sites of the NeseNPV genome. The number and transcriptional direction of each ORF are labeled as a black arrow. Known baculovirus homologs are labeled in bold by the NeseNPV ORF number. Nonbaculovirus homologs are indicated in parentheses. The homologous regions (hr) are shown as triangles, and the direct repeats (dr) are shown as circles (•).

Nucleotide sequence analysis. The genome of NeseNPV was 86,462 bp in size. Since baculovirusgenomes range in size from 99,657 bp in AdorGV (78) to 178,733bp for XcGV (30), NeseNPV, and other Neodiprion baculoviruses(B. M. Arif, personal communication), have the smallest genomesof baculoviruses sequenced thus far. Nucleotide number 1 ofthe genome was considered to be the A of the polyhedrin gene(polh) ATG start codon as has been conventionally adopted (30,36, 75), and the successive nucleotides were numbered in thedirection of the polh gene (Table 1). A total of 90 methionine-initiatedORFs with 50 or more amino acids and minimal overlapping ofadjacent ORFs were found in NeseNPV. This is the smallest numberof ORFs found in baculoviruses to date (Table 2). A total of14 ORFs overlapped adjacent ORFs, and the maximum overlappingfound was between NeseNPV ORF 40 (nese40 = lef-9) and nese41(ac68) with 58 nucleotides shared. The distribution of the orientationsfor the predicted genes was uneven, with 54 ORFs (60.7%) orientedclockwise and 35 ORFs (39.3%) in the opposite direction. Thisis different from the other baculoviruses that present a moreeven distribution. The complete nucleotide sequence of NeseNPVwas determined to have a G+C content of 34%. Currently, AdorGVpresents the lowest G+C content of the sequenced baculoviruseswith 34.5% (78). Therefore, the NeseNPV genome has the lowestG+C content of baculoviruses sequenced to this date (Table 2).

View this table:
[in this window]
[in a new window]
TABLE 2. Characteristics of baculovirus genomes^a

From the 90 NeseNPV ORFs, 43 were homologous to ORFs from otherbaculoviruses (see "best match" column in Tables 1 and 3), and29 of them belonged to the core of 30 conserved genes amongall baculoviruses (32). The only gene not identified in NeseNPVfrom the 30 conserved genes was the ld130 gene, which codesfor the fusion or F protein. The homology between NeseNPV ORFsand other baculoviruses ORFs was low, ranging from 18 to 56%,and the average amino acid identity between the NeseNPV ORFsand their best match homologs was 32.7% (Table 1). Even thepolyhedrin gene, which is a highly conserved gene among baculoviruses,had an amino acid identity of 38 to 53%. The highest homologywas found with the polyhedrin gene of another Neodiprion baculovirus,the Neodiprion abietis nucleopolyhedrovirus (NeabNPV) (GenBankaccession no. AAM95580) with a 92% amino acid identity. SixNeseNPV ORFs were highly similar to non-baculovirus-relatedgenes, and the other 41 ORFs were not considered homologs toother genes because of low amino acid sequence similarity andtherefore are considered unique to NeseNPV. A similar rangeof low amino acid conservation was observed by Afonso et al.(1) when comparing the CuniNPV genome with the those of Lepidopterabaculoviruses, although a higher number of conserved baculovirusORFs was present in NeseNPV. In these viruses, which are notclosely related, similar functional genes may be present thatare simply not identified by the amino acid identity. Additionalresearch will be necessary to reveal their function. To facilitatethe cross-reference of conserved genes, Table 1 presents theamino acid identity of the baculovirus genes with homology toNeseNPV. Also shown are the corresponding ORFs of other baculovirusesthat were indirectly homologous to NeseNPV by homology betweenother baculoviruses; however, no amino acid identity informationis shown under those circumstances.

View this table:
[in this window]
[in a new window]
TABLE 3. Baculovirus genes present in NeseNPV genome

DNA replication genes. NeseNPV had the following genes considered essential for baculovirusDNA replication by Lu et al. (49): late expression factors lef-1(nese68), lef-2 (nese57), helicase (nese61), and dnapol (nese28);but it lacked the immediate-early gene ie-1 and the late expressionfactor lef-3, also absent in CuniNPV (1). The ie-1 gene, however,is poorly conserved even among lepidopteran baculoviruses (36).Those four DNA replication genes present in NeseNPV are foundin all baculoviruses sequenced (32), indicating that they arevery important for DNA replication. Other genes identified inNeseNPV have been characterized as being involved in the processingof mature DNA (45, 56, 80) are vlf-1 (nese45) and alk-exo (nese31);and in DNA binding (57, 76), dbp (nese22) and p6.9 (nese36).

Regulatory genes. Gene expression in baculoviruses follows a temporal cascade.Four genes capable of transregulating early viral gene expressionthat occurs prior to viral DNA replication have been identified:ie-0, ie-1, ie-2, and pe38 (23). NeseNPV does not have significanthomology to any of these early transregulating genes of baculoviruses.On the other hand, it had homology with various late transcriptionactivator genes (50, 51), including the late expression factorslef-4 (nese62), lef-5 (nese58), lef-8 (nese81), lef-9 (nese40),lef-11 (nese23), p47 (nese49), and the very late factor vlf-1(nese45). From these, the four genes encoding the viral RNApolymerase subunits (lef-4, lef-8, lef-9, and p47), formingthe simplest eukaryotic DNA-directed RNA polymerase (27), arepresent in all the sequenced baculoviruses, including CuniNPV(1, 32). Another gene involved in regulation of late expressionwas nese5, homologous to ac69 recently named mtase1 for itsmethyltransferase activity (79). Although mtase1 has no demonstrableeffect on viral replication, it has been shown to stimulatetransient late gene expression (44). NeseNPV lacks lef-6, lef-10,lef-12, and 39K that are late regulatory genes (51). However,Lin and Blissard (47), using a lef-6 null AcMNPV virus, haveshown that lef-6 is not essential for viral DNA replicationnor for late gene transcription but probably is important inaccelerating the infection cycle of AcMNPV. In support of thisfinding, NeseNPV does not have lef-6, indicating that it isnot needed for replication of this and perhaps other viruses.The two other lef genes absent from NeseNPV, lef-10 and lef-12,are not conserved in all baculoviruses (32). The protein kinase(pk-1) gene homolog, involved in very late gene expression (20),was not found in NeseNPV.

Structural genes. From the 15 conserved structural genes of lepidopteran baculoviruses(32), only 12 were found to have homology with NeseNPV (Table3). The structural polyhedrin or granulin gene homologs arepresent in all baculoviruses with the exception of CuniNPV.NeseNPV polyhedrin (nese1) had the highest amino acid homologyto the Neodiprion baculovirus NeabNPV polyhedrin with a 92%amino acid identity, but when compared to the lepidopteran polyhedrins/granulins,the homology ranged from 53% with SpltMNPV to 38% with CpGV(Table 1). The NeseNPV basic DNA binding protein p6.9 (nese36)had a 56% amino acid identity with AdorGV p6.9 but was not associatedby BLAST with any other baculovirus p6.9 gene, even though ithad a high arginine (29%) and serine (27%) concentration andthe RRPGRPR conserved motif identified in other p6.9 proteins(76). The size of the predicted NeseNPV p6.9 protein was 86amino acids (aa), bigger that all other p6.9 proteins exceptfor those of HaSNPV (76) and HzSNPV (14). Nese36, however, hada high amino acid identity with splicing factor proteins fromorganisms such as insects, mouse, and human. The occlusion-derivedviral envelope odv-e18 (nese65), odv-e56 (nese38), and odv-ec27(nese66) homologs in NeseNPV had amino acid identities of 34%(with XcGV), 39% (with LdMNPV) and 23% (with BmNPV), respectively(Table 1). The capsid-associated protein genes vp39 (nese89)and vp91 capsid (nese84) were also present. Although the identityof the NeseNPV vp39 homolog averaged only 25% with other vp39of baculoviruses, it contained a total of six cysteines of theeight conserved in AcMNPV and OpMNPV (24). From those, fourwere in conserved locations that could generate disulfide bondsimportant for the protein structure. The vp1054 (nese85), aviral structural protein present in both occlusion-derived virus(ODV) and budded virus (BV) required for nucleocapsid assembly(63), and the glycoprotein gp41 (nese47), found only in ODVbut required for the egress of nucleocapsids from the nucleusduring the BV synthesis (64), were found in NeseNPV. The averageidentity with other vp1054 and gp41 proteins was 24% and 25%,respectively. The final NeseNPV structural gene found in alllepidopteran baculoviruses was p74 (nese50) with an averageamino acid identity of 38%. This gene product, found associatedwith ODVs, is essential for the primary infection of midgutcells of insect larvae during oral infection (21). Another gene,associated with the ODV envelope of Spodoptera littoralis MNPV(SpliNPV), having a similar function as p74, is the ac119 homologcalled the per os infectivity factor or pif gene (39). Nese79had a 33% amino acid identity with ac119 and 34% with SpliNPVpif. A second pif gene (pif-2 = SeMNPV ORF 35), has recentlybeen described for SeMNPV (70). Nese55 had a 35% amino acididentity with Se35 and 46% with its AcMNPV homolog, ac22. Bothpif and pif-2 genes have been conserved in all the baculovirusessequenced (32). Genes not identified in NeseNPV were p10, pk-1,odv-e66, odv-e25, and neither ld130 nor gp64. The p10 gene,responsible for the formation of fibrillar structures and associatedwith the occlusion body, is also absent in CpGV (53) and CuniNPV(1). CuniNPV also lacks odv-e66 and pk-1 homologs. A major differencebetween NeseNPV and other baculoviruses is the lack of a definitivegp64 or ld130 homolog that code for envelope fusion proteins.These functionally analogous proteins (37) present in BV envelopesare capable of inducing cell fusion and are involved in viralattachment (52, 68). All the baculoviruses sequenced to datehave the ld130 homolog (32), which codes for the F protein,and the members of the group I NPVs also have gp64 (8, 9, 58).NeseNPV does not have either of these gene homologs. NeseNPVinfections of sawfly larvae are restricted to the insect midgutepithelial cells, as happens with baculoviruses that infectother Hymenoptera, Diptera, Coleoptera, Thysanura and Trichopterahosts (22). In NeseNPV, the function of viral transmission betweenthe epithelial cells may lie in another gene. To find probablecandidates for such role, all the NeseNPV ORFs were tested fortransmembrane domains and N-terminal signal peptides. The followingNeseNPV ORFs had both a signal peptide and a transmembrane domain:Nese26, nese55 (pif-2), nese60 (ac96 homolog), nese65 (odv-e18),nese67 (ac145 homolog), nese69 (ac115 homolog), nese71, nese79(pif) and nese84 (vp91 capsid). These and other NeseNPV ORFsthat had a transmembrane domain, an N-terminal signal peptideor a similar size to baculovirus ld130 homologs were alignedindividually with the ld130 homologs from several baculoviruses.However in all searches, the amino acid identity was never higherthan 15% and considered random amino acid matches.

Inhibitors of apoptosis. Apoptosis or programmed cell death is known to happen in vertebratesand invertebrate organisms as a response to a series of inductions,including pathogen attacks (15). Baculoviruses have genes capableof suppressing cellular apoptosis to maintain viral replication.The family of genes called inhibitor of apoptosis, iap, whichwas first identified in CpGV (17), is found in most baculovirusessequenced to date with the exception of CuniNPV. NeseNPV containsan iap homolog (nese17) with an amino acid identity of 38% toCpGV iap-3. Several baculoviruses have more than one iap genein their genomes, while NeseNPV has only one. Homologs to iapusually have two baculovirus iap repeats (BIR) and a C-terminalzinc finger (7). Nese17 has only one BIR motif and the zincfinger motif, similar to what is found in SpltMNPV (65) andLdMNPV (42), and also in some iap-2s such as MacoMNPV and RoMNPV(28, 46). NeseNPV iap had a higher homology with the iap-3 genefamily proposed by Luque et al. (53) based on their iap phylogeneticanalysis. Until biological activity is determined for nese17,it will simply be called iap. Other apoptosis inhibitor genesinclude p35, found originally in AcMNPV (16), and Slp49 of SpliNPV,which is a p35 homolog that blocks apoptosis induced by infectionof Sf9 cells with an AcMNPV p35-deficient mutant (19, 82, 69).NeseNPV did not have any homology with either p35 or Slp49 genes,indicating that either its iap is the only gene responsiblefor any apoptotic suppressor activity or a very low amino acidhomology has made it difficult to locate a p35 or a Slp49 homolog.However, since it has been shown that not all iap's are capableof preventing cell death (54), the biological activity of theNeseNPV iap homolog will have to be experimentally confirmed.

Other conserved genes. Other NeseNPV ORFs had homology with baculovirus genes, thebiological functions of which have not yet been determined.These are listed in Table 3. Nese67 was homologous to ac145(30%) and ac150 (25%).

Duplicated genes. Two NeseNPV ORFs (nese18 and nese19) were duplicates, positionedin opposite orientation with a homologous region (hr4) and asmall direct repeat (dr2) separating them (Fig. 1). NeseNPVdoes not have a copy of the AcMNPV ORF 2 homolog, characterizedas the baculovirus repeat ORF (bro) (38). Other baculovirusesusually have one or more copies of bro genes with the exceptionof SeMNPV (36), PxGV (29), and AdorGV (78), which, like NeseNPV,have no bro homologs.

Nonbaculovirus homologs. Six NeseNPV ORFs had high homology with known proteins not relatedto baculoviruses (Table 1). Amino acid identity as high as 40%was obtained between nese7 and the family of trypsin-like proteasesfrom insects belonging to the orders Lepidoptera, Diptera, Coleoptera,and Siphonaptera. Homology was also obtained with the trypsinof other arthropods and mammals. Figure 2, shows the alignmentof nese7 with trypsins of other insects and a mammal (cow).Conserved amino acid sequences are shown in boxes. The trypsincatalytic triad (histidine, aspartic acid, and serine) and theconserved cysteines from the trypsin-like proteins (73) werepresent in nese7 and are indicated in Fig. 2. The size of thepredicted protein (258 aa) was comparable to that of other trypsins.The maximum-likelihood quartets method suggests that the trypsin-likegene of the NeseNPV may be of invertebrate origin and in 35%of the time its sister taxa was the trypsin of a Coleoptera,Diaprepes abbreviata (data not shown). This putative gene hasnot been reported in any DNA virus including baculoviruses.Neodiprion baculoviruses, however, seem to have it present,since Neodiprion lecontei (Basil Arif, personal communication)also has this gene. It will be interesting to determine if thisgene was derived from the insect host, since no hymenopterantrypsin was found in the database search to make a comparison.Nese52 had four C2H2 zinc finger domains and significant homologywith C2H2 zinc finger proteins from insects and mammals. Althoughon average, only the first half of this ORF presented the homology,this type of protein has not been previously described in baculoviruses.Two NeseNPV ORFs, nese72 and nese73, had homology with the regulatorof chromosome condensation (RCC1) proteins from a wide groupof organisms such as microsporidia, nematodes, fruit flies,frogs, rodents and humans. Nese83 had up to 39% amino acid identitywith the capsid protein of several insect densoviruses. Finally,nese90 was homologous to a large number of phosphotransferases.

View larger version (58K):
[in this window]
[in a new window]
FIG. 2. Alignment of NeseNPV ORF 7 (nese7) and trypsins from other organisms. Bos taurus, cow; Hypoderma lineatum, early cattle grub; Sarcophaga bullata, grey fleshfly; Drosophila erecta, a fruit fly; Aedes aegypti, yellow fever mosquito; Bombyx mori, silkworm. Regions highlighted as conserved amino acids by the conserved domain database (CDD) are boxed with a consensus sequence indicated on the top line. # represents the trypsin catalytic triad, histidine, aspartic acid, and serine. * indicates the conserved cysteines.

Absent genes. NeseNPV did not present any of the known baculovirus genes involvedin nucleotide metabolism or auxiliary genes such as the ribonucleotidereductases (rr1, rr2), dUTPase (dutpase), proliferating cellnuclear antigen (pcna), protein tyrosine phosphatase (ptp),ubiquitin, superoxidase dismutase (sod), conotoxin-like peptide(ctl), cathepsin, chitinase, ecdysteroid UDP-glucosyltransferase(egt), fibroblast growth factor (fgf), actin rearrangement-inducingfactor-1 (arif-1), and enhancin. However, these genes are notconserved among baculoviruses. The exception is the ld130 gene,which has been shown to be present in all baculoviruses (32)and was not found in NeseNPV. Seven NeseNPV ORFs had a verylow homology to baculovirus genes even after forcing alignmentwith the suspected baculovirus homolog, and so for the presentanalysis, they were considered random similarities between theviruses. However, since some of these NeseNPV ORFs had predictedamino acid sizes very close to those for the baculovirus geneproducts, and since some of these ORFs may have real functionalsimilarities that the amino acid sequence identity did not support,a listing of such ORFs was considered useful. Under that groupwere ie-2 (ac151 = 408 aa) (nese9 = 416 aa), cg30 (ac88 = 264aa) (nese12 = 130 aa), odv-e25 (ac94 = 228 aa) (nese26 = 218aa), ie-1 (ac147 = 582 aa) (nese82 = 209 aa) and lef-3 (ac67= 385 aa) (nese87 = 311 aa). Two AcMNPV homologs with unknownfunctions were also under this group: ac66 (808 aa) (nese29= 792 aa), and ac75 (133 aa) (nese43 = 125 aa).

hrs. Six hrs were located within the NeseNPV genome (Fig. 3), eachcontaining small (5- or 6-bp) perfect palindromes, which canbe extended up to 19-bp imperfect palindromes, embedded withina series of direct repeats. The nucleotide sequence has beenaligned to show the size of the repeat units, and when needed,gaps were introduced to optimize the alignment (Fig. 3A). Thenumber of hrs in other baculoviruses ranges from four in CuniNPV,PxGV and MacoNPV (1, 29, 46) to 17 in SpltNPV (65). Baculovirusesreported without typical hrs are CpGV (53) and AdorGV (78).NeseNPV hrs varied in length from 396 (hr5) to 794 (hr2) nucleotides(Table 4), and the combined size of the six hrs was 3,669 bpor 4.3% of the genome. This percentage was significantly higherthan those for the hrs from CuniNPV (0.8%) and AdhoNPV (1.1%)(1, 60) but lower than those for several other baculovirus hrs,including SpltMNPV (65), HzSNPV (14), HaSNPV (13), and PxGV(29), which comprise approximately 6% of their genomes. NeseNPVhrs consist of repeats of 45 bp (hr5 and hr6), 65 bp (hr2),or alternating 65 bp and 45 bp units (hr1, hr3, and hr4) (Fig.3B). The number of repeat units in each hr ranged from 9 to14 and the length and number of times they occur are presentedin Table 4. The average G+C content of the hrs was 49.8%, significantlyhigher than the 34% for the complete NeseNPV genome.

View larger version (64K):
[in this window]
[in a new window]
FIG. 3. (A) hrs of NeseNPV. The nucleotide sequence of three NeseNPV hrs are aligned to exemplify the direct repeats found in units of 65 bp (hr2), 45 bp (hr5), or alternating 65 and 45 bp (hr1). The consensus sequence of the hrs is shown in bold, with uppercase letters representing 75% or more conserved nucleotides (60). Gaps (–) were inserted to optimize the alignment of the repeat units. The numbers on the left hand side of each hr indicate the nucleotide position in the genome. For each hr, the nucleotides forming a perfect palindrome (10 to 12 nucleotides) are marked with a solid arrow. Imperfect palindromes, formed by 70% or more complementary bases, are indicated by dots that extend 31 to 40 bases. The late promoter motif ATAAG and the potential GATA factor-binding site are double underlined and boxed, respectively. (B) Schematic representation of the repeats units of either 65 bp (solid arrows) or 45 bp (double-line arrows) found in each hr. (C) Comparison of the consensus sequences of the six NeseNPV hrs with the perfect palindrome region underlined and the extended imperfect palindrome dotted. An asterisk in hr6 indicates that the repeat sequence is different from the other hrs.

View this table:
[in this window]
[in a new window]
TABLE 4. Structure of NeseNPV homologous regions (hr's) and direct repeats (dr's)

In addition to the hrs of NeseNPV, four small repetitive regionsnot recognized as hrs were identified. These regions were calleddirect repeats (drs), with dr1 consisting of two direct repeatsof 120 bp, dr2 of three repeats of 29 bp, dr3 of two repeatsof 13 bp, and dr4 of two repeats of 48 bp (Table 4). When comparedwith the hr nucleotide composition, the dr's were found to beA-T rich, especially dr3 (70.4%) and dr4 (65.7%). Two of thedr's (dr1, dr4) were located inside NeseNPV ORFs, nese16 andnese88, respectively. The same has been described for AdorGV(78), wherein the repeats may cause these putative ORFs to notbe functional.

Baculovirus hrs have been characterized to be enhancers of RNAII-mediated transcription of early genes (26, 48, 62). hrs havealso been characterized as origins of DNA replication (66, 67)and are typically dispersed around the genome. This dispersionis suggested to speed the rate at which replication occurs byinitiation at several sites of the genome (67). However, inNeseNPV the hrs were located predominantly in the first halfof the circular genome, with hr1 at 8.6 to 9.1 kb (9.95 to 10.58map units) and hr6 at 49.0 to 49.5 kb (56.7 to 57.2 map units).Furthermore, the distance between hr5 and hr6 was 26.9 kb, whichreveals a further clustering of hrs in the first quarter ofthe genome (Fig. 1; Table 4). Clustering of the hrs has alsobeen observed in CuniNPV, with its hrs concentrated in 35% ofthe genome.

A unique feature observed in NeseNPV hrs was the presence ofmultiple copies of the sequences ATAAG and GATA (Fig. 3A and3C). Baculovirus late promoter motifs are (A/G/T)TAAG; however,TAAG is the essential element (51). Although the late promoterelement TAAG is found in baculoviruses, it is generally notlocated inside of an hr. Interestingly, a GATA element thatbinds host transcription factors (40) is located close to, andin hr1 and hr2 overlapping, each of the TAAG late promoter motifswithin the hrs. Multiple GATA elements have been found in theHz-1 PAT1 persistence-associated transcript, which does notencode a protein (12). The sequence of PAT1 also displays abundantdirect and inverted repeats. No ORFs are predicted for the NeseNPVhrs either, indicating that they could have a yet-uncharacterizedregulatory function similar to that of the GATA elements ofHz-1 PAT1. hrs are a distinctive feature of the majority ofbaculovirus genomes and have been shown to be essential forregulatory function (26, 62, 66, 67). The function of the NeseNPVhrs as origins of replication or enhancers of transcriptionhas yet to be determined experimentally.

GeneParity. The comparison of the gene order of NeseNPV and six other genomesshowed that from the 43 NeseNPV baculovirus homolog genes, 42were in AcMNPV, 41 in LdMNPV, 43 in MacoNPV-A, 43 in HzSNPV,41 in CpGV, and 29 in CuniNPV. Eight clusters of two or moregenes were found; however, not all of the genomes had all theseclusters (Fig. 4). Clusters 1, 2, 3, and 5 were absent in CuniNPV,while clusters 6, 7, and 8 were present only in CuniNPV. Relaxedclusters were determined to try to recognize regions where genesmay have been added or deleted between genes that were consecutivein a more ancestral lineage of baculovirus. For example, cluster2 (p45, p40, and p6.9) was considered relaxed because in NeseNPVas well as in the other genomes there is an ORF between p45and p40. Cluster 3 (vlf-1, ac78, gp41, and ac81) was consideredrelaxed only in AcMNPV because there is an extra ORF betweenac78 and gp41 not found in any of the other baculovirus genomesanalyzed. Cluster 4 (lef-5, 38K, ac96, helicase, and lef-4)was the most conserved cluster. It was considered relaxed toaccommodate the addition of lef-4 to a cluster that has alreadybeen shown to be conserved (32). Five genes exist between helicaseand lef-4 in CuniNPV, while the other baculoviruses have eitherthree or four genes. However, not only was the order of thegenes in cluster 4 conserved, but also their relative transcriptionaldirection. Cluster 5 (ac142, odv-e18, odv-ec27, and ac148) wasrelaxed because in NeseNPV there is an extra ORF between ac142and odv-e18, which does not occur in the five other genomes.

View larger version (33K):
[in this window]
[in a new window]
FIG. 4. GeneParity plots comparing the NeseNPV gene order to six other baculoviruses: AcMNPV (A), LdMNPV (B), MacoMNPV-A (C), HzSNPV (D), CpGV (E), and CuniNPV (F). Eight clusters are underlined and/or boxed in the graphs, which include the following genes or AcMNPV homologs: 1, ac92, ac93; 2, p45, p40, p6.9; 3, vlf-1, ac78, gp41, ac81; 4, lef-5, 38K, ac96, helicase and lef-4; 5, ac142, odv-e18, odv-ec27*, ac148; 6, lef-9, ac68; 7, p47, p74; and 8, lef-1, ac115. Asterisks represent genes that are not part of a specific cluster in certain genomes. Straight lines mark clusters of two or more genes (underlined above) that are sequential in both NeseNPV and the genome being compared. Relaxed clusters that include genes that are not sequential in the compared genome are indicated with a box.

GeneParity plots have been used to compare conservation of genomicorganization. The GV genomes maintain a strong colinearity betweenthem (29, 53, 78), but they are not strongly colinear with NPVssuch as AcMNPV. The comparison between lepidopteran NPVs showsome genome arrangements, such as inversions. However, the orderof a significant number of baculovirus genes is conserved (33,36, 46, 65). In contrast, NeseNPV had few common clusters withother baculoviruses even after accepting relaxed clusters.

The presence of conserved gene clusters can result from a physicalconstraint preventing their separation, as suggested by Herniouet al. (32) about lef-5, 38K, ac96, and helicase (cluster 4in the present analysis). Another explanation for the conservationof clusters could be gene overlapping. Cluster 3 had genes withoverlapping coding regions. In genes transcribed on oppositestrands and away from each other, such as the genes from clusters1 and 4, the promoter region or start codon of one gene couldbe included in the continuous gene and so the physical separationof these two genes may cause their inactivation. Therefore,the continuity of the gene cluster would be maintained in theevolution of the baculovirus lineages.

Baculovirus phylogeny. The DNA polymerase was used previously to investigate the relationshipamong distantly related organisms and DNA viruses (10). Therefore,the DNA polymerase of the baculovirus was used in order to helpestablish the position of the root for a baculovirus tree byrooting it with 23 DNA polymerases from widely diverse phylogeneticorigins. It was necessary to include these diverse taxa to determinewhether there were different DNA polymerase lineages for baculovirusesand the Hz-1 virus. The results (Fig. 5a) indicated that 75%of the time the DNA polymerases did split into viral and nonviral.Moreover, the NeseNPV DNA polymerase was more related to thosefrom the lepidopteran NPV and GV than the DNA polymerase ofthe CuniNPV from Diptera. The Hz-1 virus DNA polymerase possiblydoes not belong to a baculovirus monophyletic group includingCuniNPV, NeseNPV, and the lepidopteran NPVs and GVs. These resultssuggested that possibly CuniNPV was the appropriate rootingoutgroup for the monophyletic baculovirus group, which was alsosupported by the overall level of genomic similarity to NeseNPVand to the other NPVs and GVs (e.g., CuniNPV does not have apolyhedrin/granulin ortholog). The same adjacency pattern, indicatingthat the CuniNPV is the most ancient lineage, was obtained bymidpoint rooting of the maximum-likelihood tree or by UPGMAclustering after excluding all nonbaculovirus DNA polymerasesfrom the analysis (data not shown). Once the cladogenetic eventswere oriented in time by using the DNA polymerase gene, indicatingthe split of the CuniNPV lineage from the ancestral stem leadingto the NeseNPV and the other lepidopteran NPVs and GVs, a maximum-likelihoodtree was estimated for the proteons including 29 conserved genesfor 24 baculoviruses and used the CuniNPV as outgroup. Figure5b shows a tree for the proteons which agrees with the DNA polymeraseplacement of CuniNPV and NeseNPV. As for the DNA polymerasegene phylogeny, the same adjacency pattern was also obtainedwithout an explicit choice of outgroup by midpoint rooting andUPGMA clustering. Additionally, the group I NPV, group II NPV,and GV clades are delineated using both proteons and the DNApolymerase in the phylogenetic analysis.

View larger version (51K):
[in this window]
[in a new window]
FIG. 5. DNA polymerase phylogeny (A) indicates distinct lineages when all sequenced baculoviruses and the Hz-1 DNA polymerases are included. The root position of the DNA polymerase gene suggested that the lepidopteran NPV and GV share a common ancestral lineage with the NeseNPV after its split from the one leading to the CuniNPV. (B) A global maximum-likelihood phylogeny for 29 conserved genes from 24 baculoviruses supports the DNA polymerase tree in its separation of NeseNPV and CuniNPV. The numbers near nodes indicate the percentage of time a partition was found by the quartet method and is considered the level of support for the tree. Only nodes appearing more than 50% of the time were resolved.

NeseNPV is different from other baculoviruses in several notableareas: genome size, smaller number of ORFs, low GC content,hr location, and a general low homology with other baculoviruses.An indirect way of finding important genes is by their conservationamong the baculoviruses genomes. In that regard, genes and ORFsalso present in distantly related members of the Baculoviridaefamily, such as those infecting insects from the orders Dipteraand Hymenoptera, can be used to predict genes essential forthe virus replication cycle. However, there can be a seriesof other genes used by those viruses that differ from the morestudied group of lepidopteran baculoviruses, since the virus-hostrelationship is very important in the evolutionary history ofseveral organisms, especially for viruses that use the hostmachinery for their propagation. It will be very informativeto compare NeseNPV, which was isolated from Old World N. sertiferinsects, to NeleNPV, which infects New World N. lecontei insects,and also other sawfly baculoviruses. This will give a betterunderstanding of the evolutionary history of host-derived genes,such as that for trypsin, and others found in the genomes ofhymenopteran baculoviruses. We may also obtain information onthe mutation rates of baculoviruses that have diverged overan extensive amount of time. The value of the genome informationfrom hymenopteran baculoviruses will provide a foundation forthe evolutionary history of the lepidopteran and dipteran baculovirusesas well. Ultimately this new information and comparison couldresult in the establishment of a new genus within the familyBaculoviridae.

ACKNOWLEDGMENTS

This work was supported by United States Department of Agriculturegrant 2001-35302-11006.

We thank Ronit Keisari for technical help, Basil Arif and HilaryLauzon for sharing information about the Neodiprion leconteiNPV, and D. Boucias, Q. Li, and Pamela Howell for comments andcorrections on the manuscript.

FOOTNOTES

* Corresponding author. Mailing address: Entomology and Nematology Department, University of Florida, Bldg. 970, Natural Area Dr., Gainesville, FL 32611-0620. Phone: (352) 392-1901, ext. 148. Fax: (352) 392-0290. E-mail: jem{at}mail.ifas.ufl.edu.

Florida Agricultural Experiment Station journal series no. R-09762.

Present address: LEMB—Instituto de Ciências Biomédicas—USP,CEP 05508-900, São Paulo, SP, Brazil.

Present address: Department of Molecular and Cellular Oncology,University of Texas, M. D. Anderson Cancer Center, Houston,TX 77054.

^|| Present address: Department of Pediatrics, Health Science Centerat Houston Graduate School of Biomedical Sciences, Universityof Texas, Houston, TX 77030.

REFERENCES

Top