Lacanobia oleracea nucleopolyhedrovirus (LaolNPV): A new European species of alphabaculovirus with a narrow host range

During an insect sampling program in alfalfa crops near Montpellier, France in 2011, Lacanobia oleracea larvae were collected that died due to nucleopolyhedrovirus infection (LaolNPV). This virus was subjected to molecular and biological characterization. The virus was a multiple nucleocapsid NPV that showed similar restriction profiles to Mamestra configurata NPV-A (MacoNPV-A) but with significant differences. Polypeptide analysis demonstrated similar proteins in occlusion bodies and occlusion derived virions, to those observed in NPVs from Mamestra spp. Terminal sequencing revealed that the genome organization shared similarity with that of MacoNPV-A. The most homologous virus was MacoNPV-A 90/2 isolate (95.63% identity and 96.47% similarity), followed by MacoNPV-A 90/4 strain (95.37% and 96.26%), MacoNPV-B (89.21% and 93.53%) and M. brassicae MNPV (89.42% and 93.74%). Phylogenetic analysis performed with lef-8, lef-9, polh and a concatenated set of genes showed that LaolNPV and the Mamestra spp. NPVs clustered together with HaMNPV, but with a closer genetic distance to MacoNPV-A strains. The Kimura 2-parameter (K-2-P) distances of the complete genes were greater than 0.05 between LaolNPV and the MbMNPV/MacoNPV-B/HaMNPV complex, which indicates that LaolNPV is a distinct species. K-2-P distances were in the range 0.015–0.050 for comparisons of LaolNPV with MacoNPV-A strains, such that additional biological characteristics should be evaluated to determine species status. While MacoNPV-A was pathogenic to seven lepidopteran species tested, LaolNPV was only pathogenic to Chrysodeixis chalcites. Given these findings, Lacanobia oleracea nucleopolyhedrovirus should be considered as a new species in the Alphabaculovirus genus.


Virus isolates, insect source and rearing
In April 2011, lepidopteran larvae were collected from alfalfa crops and the weeds present (mainly grasses and wild brassicas) near Montpellier, France. Two sites were examined for the presence of Lepidoptera; one located beside the village of Lattes (43˚33'13.9"N, 3˚56'10.0"E) and the other beside the village of Marsillargues (43˚39'15.0"N 4˚11'25.9"E) in southern France. These fields had been used for many years for the production of pasture and alfalfa. No specific permissions were required for access to the land and the field studies did not involve endangered or protected species.
The noctuids Agrochola lychnidis, Lacanobia oleracea and Xylena exsoleta were identified as the most abundant species present, whereas Aporophyla australis occurred in low numbers. In this region L. oleracea is common, especially during warm summers (M. Frayssenet, pers. obs.), although the larvae and adults of this species can be confused with A. lychnidis.
A total of 26 larvae were collected and individually reared on semisynthetic diet at 23-25˚C. During laboratory rearing seven larvae showed the typical signs of lethal polyhedrosis disease. It was difficult to identify the host species; we believe these insects were L. oleracea, but we cannot exclude the possibility that one or more of them was A. lychnidis. Consequently, we will describe the novel isolate as Lacanobia oleracea NPV (LaolNPV) in the following text.

Molecular characterization
OB purification, DNA extraction and restriction endonuclease analysis. Occlusion bodies (OBs) of the different isolates were extracted from dead larvae by homogenizing the cadavers in 500 μl of 0.1% (wt/vol) sodium dodecyl sulfate (SDS), filtered through muslin and centrifuged at 2,500 x g for 5 min. Pellets were resuspended twice in 500 μl of 0.1% SDS and centrifuged for 5 min at 2,500 x g. The resulting pellets were washed twice in distilled water, resuspended in~200 μl distilled water and OBs were stored at 4˚C until required.
For DNA extraction, virions were released from OBs by dissolving the polyhedrin matrix by mixing 100 μl of purified OB suspensions, comprising~10 8 OBs, with 100 μl of 0.5 M Na 2 CO 3 and 50 μl of 10% SDS in a final volume of 500 μl followed by incubation at 60˚C for 10 min. Undissolved OBs and other debris were pelleted at 6,000 x g for 5 min. The virion-containing supernatant was transferred to sterile 1.5 ml vials and incubated at 50˚C with 25 μl proteinase K (20 mg/ml) for 1 h. Viral DNA was extracted twice with phenol (pH 7.8):chloroform (1:1). DNA was precipitated by addition of 10% (v/v) 3 M sodium acetate (pH 5.2) and 2.5 volumes of ice-cold 96% ethanol at 12,000 x g for 10 min. The DNA pellet was washed with 70% cold ethanol and centrifuged for 5 min. The ethanol was discarded and DNA was dried at room temperature for 5 min. Finally, DNA pellets were resuspended in 50-100 μl of 0.1X TE buffer (10 mM Tris, 1 mM EDTA), and kept at 4˚C until use. For comparison, DNA was also extracted from OBs of ChchNPV, HearNPV, SpliNPV, SeMNPV, MbMNPV, MacoNPV-A and MacoNPV-B.
For restriction endonuclease (REN) digestion, a sample of viral DNA (1-2 μg) was incubated with one of the following enzymes BamHI, BglII, EcoRI or PstI (Takara) (10 U) at 37˚C for 4-12 h. Each reaction was stopped by the addition of 4 μl of 6x loading buffer (0.25% w/v bromophenol blue and 40% w/v sucrose). Fragments were separated by electrophoresis using horizontal 1% agarose gels in TAE buffer (0.004 M Tris-acetate, 0.001 M EDTA, pH 8.0) at 16 V for 14 h. DNA fragments were stained with ethidum bromide and visualized on a UV transilluminator (GeneSnap, Syngene). DNA fragment sizes were estimated by comparison to a standard molecular weight marker (Tandem ladder, Lonza, Rockland, USA).
Nucleocapsid packaging. To determine whether the virus under study was a single or multiple nucleocapsid NPV, occlusion derived virions (ODVs) were harvested by treating 5x10 8 OBs with an equal volume of 0.1 M Na 2 CO 3 , for 30 min at 28˚C. Undissolved OBs and other debris were pelleted and discarded by low-speed centrifugation (2,500 x g, 2 min). The resulting suspensions containing the ODVs were placed on the top of continuous 30-60% (w/w) sucrose gradients and centrifuged at 30,000 x g for 1 h at 4˚C in a Beckman Ti28 rotor. The banding pattern was visually inspected, photographed and compared with other viruses.
Polypeptide analysis. Structural polypeptides from the purified OBs and occlusion derived virions (ODVs) were analyzed on 11% SDS-polyacrylamide slab gels (SDS-PAGE). ODVs were harvested from OBs by mixing 10 μl of purified OBs at 10 10 OBs/ml with an equal volume of 0.1 M Na 2 CO 3 and incubating at 28˚C for 30 min. Undissolved OBs and other debris were pelleted at 6,000 x g for 5 min. A 10 μl volume of purified OBs at a concentration of 10 10 OBs/ml and 10 μl of harvested ODVs were solubilized with an equal volume of 2x Laemmli sample buffer (65.8 mM Tris-HCl, pH 6.8, 2.1% SDS, 26.3% (w/v) glycerol, 0.01% bromophenol blue, BioRad) by heating to 100˚C for 5 min prior to electrophoresis. Electrophoresis was performed at 50 mA during 2 h. Finally, the gels were stained in Comassie Brilliant Blue R solution (0.1% Comassie R Brilliant Blue, 10% v/v acetic acid and 50% v/v ethanol) for 30 min and destained with a bleaching solution (9.45% v/v ethanol and 6.75% v/v acetic acid). The polypeptide profiles were visually inspected and photographed.
DNA cloning, sequence analysis and phylogenetic analysis. In order to obtain genomic sequence information four genomic fragments were terminally sequenced. For this, two genomic libraries were constructed in pUC19 (New England Biolabs) using the EcoRI or PstI digested DNAs. Briefly, 2 μg of viral DNA extracted from purified OBs were digested with EcoRI or PstI overnight at 37˚C and then heated to 65˚C during 15 min to inactivate enzymes. A 5 μg sample of pUC19 vector was digested with EcoRI or PstI in the same conditions. The digested vector (20 μl) was dephosphorylated during 2h at 37˚C using an alkaline phosphatase (Roche Life Science, Basel, Switzerland). After dephosphorylation, the DNA (100 μl) was purified in low melting agarose gel using commercial kit for gel extraction (PCR clean-up, gel extraction, Macherey-Nagel, Düren, Germany) following manufacture's instructions. A 50 ng sample of the purified vector was ligated with 5 μl of the digested viral DNA (100 ng/μl of total fragmented DNA) using T4 DNA ligase (New England Biolabs) at 16˚C overnight. After ligation, DH5α cells were transformed with the recombinant plasmids and plated on LB agar containing 100 μg/ml ampicillin, 1 μM IPTG and 80 μg/ml X-Gal. A total of 50 white colonies for each ligation were amplified in LB broth containing 100 μg/ml ampicillin. Plasmid DNAs were purified by alkaline lysis and screened for the presence of EcoRI or PstI inserts by digestion with the respective enzymes followed by electrophoresis in 1% agarose gel. Inserts were authenticated by comparing their migration in agarose gels with the fragments of the viral DNA generated by the digestion with the same enzymes.
Two EcoRI and two PstI fragments (EcoRI-4Kb, EcoRI-6kb, PstI-5kb and PstI-6kb) were selected for terminal sequencing. Terminal nucleotide sequences were determined by Sanger Sequencing method performed by Stab Vida Company (Caparica, Portugal), employing standard M13 forward and M13 reverse primers. Sequence information was analyzed for the presence of open reading frames (ORFs) and for domain prediction using Clone Manager 9.0 (Scientific and Educational Software Server). Homology searches were performed both at the nucleotide and deduced amino acid levels, for all putative ORFs. DNA and protein comparisons with entries in the updated GenBank/EMBL, SWISS-PROT and PIR databases were performed using BLATn, BLASTp and FASTA programs [28,29].
For phylogenetic analyses, the DNA sequences within the coding regions of three highly conserved genes, the late expression factor 8 (lef-8), late expression factor 9 (lef-9) and polyhedrin (polh), were used [4,30,31]. As recommended by Jehle et al. [4], for viruses with Kimura two-parameter (K-2-P) distances of more than 0.015 in the marker genes, the complete sequences of these three marker genes were determined and used in phylogenetic analyses. For this, primers were designed in the polh, lef-8 and lef-9 genes based on the genomic sequence of the most homologous NPV. The primer pairs used were MacoA-polh-F (ATGTATACCCGTTATAGTTA)-MacoA-polh-R (TTAGTAAGCCGGTCCGTTGTA), MacoA-lef8-F (ATGACGGACGTGATTGACGA)-MacoA-lef8-R (TCATCGAACCACTGTGTTGTG), and MacoA-lef9-F (ATGACCTTTAGCGGTCATTC)-MacoA-lef9-R (CTAGTCCAAAAACATGTC GA). The resulting fragments were cloned into pGEM-T Easy vector (Promega) following manufacturer's instructions and transformed as previously described. Two clones for each gene were selected and nucleotide sequences were determined by Sanger Sequencing method performed by Stab Vida Company (Caparica, Portugal), employing standard SP6 and T7 reverse primers.
The K-2-P distances were calculated for single and concatenated lef-8, lef-9 and polh genes. Concatenated sequences from the same virus were treated as a single sequence. Multiple sequence alignments were performed using MEGA6.06 software [32], and the K-2-P nucleotide substitution model was used for the analysis. Maximum parsimony (MP) phylogenetic trees (1,000 bootstrap replicates) were inferred from the nucleotide sequence alignments using MEGA 6.06. Introduced gaps were treated as missing data.

Biological characterization
Host range determination. The host range of LaolNPV was determined by oral inoculation bioassays with the following lepidopteran species: M. brassicae, M. configurata, T. ni, C. chalcites, S. littoralis, S. frugiperda, S. exigua and H. armigera. The results of the oral inoculation studies were compared with those of MbMNPV, MacoNPV-A and MacoNPV-B. For this, second-instar larvae from the laboratory colonies were starved for 8-12 h at 26˚C and then allowed to drink from an aqueous suspension containing 10% sucrose (w/v) and Fluorella blue and OBs at concentrations of 10 5 and 10 7 OBs/ml, following the droplet feeding method [48]. These concentrations were selected to permit determination of the susceptible host species based on preliminary tests (data not shown) and previous studies using MacoNPV [13,15,16] or MbMNPV [49,50]. Larvae that ingested the suspension within 10 min, which were identified by the blue color of their intestine, were transferred to individual wells of a 25-well tissue culture plate with a cube (1 cm 3 ) of semisynthetic diet [26]. Host range tests were performed using 25 larvae per virus concentration and 25 larvae treated with aqueous suspension containing sucrose and Fluorella blue without OBs as controls. Each assay comprised three replicates. Larvae were reared individually at 25˚C and mortality was recorded every 24 h until the insects had either died or pupated. Dead larvae were examined microscopically to determine the presence of OBs. When present, OBs were purified from each group of dead larvae and viral DNA was extracted as previously described and subjected to REN analysis to determine the identity of the virus. The percentage of larval mortality was calculated for each virus and concentration and subjected to univariante analysis of variance (ANOVA) in SPSS v23 (IBM SPSS Statistics, Softtonic). The significance of treatments was determined for comparisons among the estimated means by Tukey test (p<0.05).

Molecular characterization
REN profiles indicated the proximity of LaolNPV to MacoNPV-A. The BglII, EcoRI and PstI profiles of the isolates from each of the seven insects showed very similar patterns with the different enzymes tested. No submolar bands were observed in the different REN profiles, indicating that each isolate comprised a single or majority genotype ( Fig 1A). Therefore, the first isolate was selected as the prototype and its DNA and OBs were used for subsequent studies.
When comparing the REN profiles with those of ChchNPV, HearNPV, SeMNPV or SpliNPV, marked differences were observed, suggesting that the novel virus was genetically distinct from the other virus species tested ( Fig 1B). Additionally comparison with the REN profiles generated in silico with AcMNPV, LdMNPV or TnSNPV also indicated that LaolNPV was genetically distinct from these viruses (data not shown).
The REN profiles were then compared with those of MbMNPV, MacoNPV-A and MacoNPV-B. As observed in the BglII profiles ( Fig 1C) the LaolNPV isolate from France might be more similar to MacoNPV-A than to MbMNPV or MacoNPV-B, although clear differences among these viruses were observed in the EcoRI and PstI profiles, suggesting that LaolNPV might represent a novel virus species.
LaolNPV is a multiple nucleocapsid NPV. The banding pattern observed following ODV density gradient centrifugation revealed that the ODVs of LaolNPV contained multiple nucleocapsids, as indicated by the multiple bands visible in the sucrose gradient (Fig 2). Based on visual inspection, the banding patterns of MbMNPV, MacoNPV-A and MacoNPV-B were quite similar, and comprised between 1 and 9 nucleocapsids/ODV in MbMNPV and OB and ODV structural polypetides in LaolNPV. The SDS-PAGE profiles of OB and ODV polypeptides were generally similar between LaolNPV and the other three NPVs (Fig 3). There were, however, clear differences in the molecular weights of specific major proteins and the presence or absence of particular proteins. In the OB polypeptide profiles, the virus under study presented a band at 75 kDa that was absent in MbMNPV, MacoNPV-A and MacoNPV-B. Additionally, the intensity of the band at 20 kDa was much reduced in LaolNPV compared to that of the other three viruses. Similarly, in the ODV profiles, LaolNPV presented a clear band of~30 kDa that was absent in the other three viruses, while the band below the 25 kDa marker showed a lower molecular weight and the band of~10 kDa had a stronger intensity in LaolNPV than in the other three viruses.
Terminal sequencing and phylogenetic analysis indicates that LaolNPV evolved from the same ancestor as MacoNPV-A. The terminal sequence information obtained from the four cloned fragments showed homology with 18 ORFs. However, although they appeared distinct in agarose gels, the two PstI fragments that were terminal sequenced were in fact identical fragments (corresponding to the 5,575 bp [nt 6,124 to 11,699] PstI fragment in the MacoNPV-A 90/2 genome [14]), showing the same ORFs. Therefore, just a total of 14 different ORFs were identified as being homologues of known baculovirus ORFs and all were assigned on the basis of their similarity to previously identified ORFs in the MacoNPV-A 90/2 genome [14]. The direction of transcription, its position in the sequenced fragment relative to that in the MacoNPV-A 90/2 genome, as well as its identity to MacoNPV-A 90/2 homologs are summarized in Table 1. Within the 5' end of PstI-5kb (M13-Rv sequence) two ORFs were identified, namely orf5 and odpv-6e that showed 93% and 95% nt sequence identity with the homologous ORFs of MacoNPV-A 90/2, while the 3' end (M13-Fw sequence) included efp and orf10, with 97 and 99% nt identity to the corresponding homologs in MacoNPV-A 90/2, respectively (Table 1). Similarly, within the PstI-6kb 5'-terminal sequence (M13-Fw) orf5 and odpv-6e were littoralis (Spli), Spodopera exigua MNPV (Se) and the virus under study (LaolNPV) with BamHI, BglII, EcoRI and PstI, and C) genomic DNAs of Mamestra brassicae MNPV (Mb), Mamestra configurata NPV-A (MacoA), Mamestra configurata NPV-B (MacoB) and LaolNPV with BglII, EcoRI and PstI.
The amino acid sequence identity and similarity of the 14 ORFs were also compared with those of other NPVs ( Table 2). The most homologous NPV to LaolNPV was MacoNPV-A 90/2 with 95.93% identity and 96.64% similarity at the amino acid level, followed by MacoNPV-A 90/4 with 95.78% identity and 96.50% similarity. The identity and similarity values to MbMNPV (90.00 and 93.79%) and MacoNPV-B (89.78 and 93.64%), were lower, but were of a similar magnitude to the identity and similarity values (90.42 and 95.00%) for comparisons of MacoNPV-A 90/2 and MacoNPV-B or MbMNPV (data not shown). In contrast, the identity and similarity values for comparisons of the MacoNPV-A 90/2 and MacoNPV-A 90/4 isolates, using the same sequences, increased to 99.53 and 99.74%, respectively (data not shown). Similar results were obtained at the nucleotide level: the nt sequence identity between LaolNPV and the MacoNPV-A isolates was intermediate (95.89%) between that found between the same species (isolates of MacoNPV-A; 99.68%), and different species (MacoNPV-A 90/2 and MacoNPV-B or MbMNPV; 90.42 and 88.95%, respectively).
To further determine the relationship between LaolNPV and the other NPVs and to establish whether this is a novel virus or a variant of MacoNPV-A, a phylogenetic study was performed using a total of 19 NPV sequences and the complete nucleotide sequences of single or concatenated lef-8, lef-9 and polh using maximum parsimony (MP) and MEGA6.06 distance        [4]. The K-2-P model was used to calculate the genetic distances between the different viruses. The phylogenetic trees using complete genes sequences showed that in the four cases Mamestra NPVs formed a separate branch together with HearMNPV and LaolNPV, although with different bootstrap values between branches depending on the gene (Fig 4). MbMNPV shares close sequence identity with MacoNPV-B and HearMNPV, whereas MacoNPV-A strains formed a separate branch. In all cases, except the lef-8 gene, LaolNPV clustered on the same branch as MacoNPV-A isolates, clearly separated from the MbMNPV/MacoNPV-B/ HaMNPV complex. MacoNPV-A 90/2 and MacoNPV-A 90/4 invariably clustered together, whereas LaolNPV, although clearly proximal, was separated on another branch.
The K-2-P values between LaolNPV and the MbMNPV/MacoNPV-B/HearMNPV complex for lef-9, polh and concatenated sequences were greater than 0.05 in all cases (underlined in Table 3 and Table 4). Similarly, the K-2-P values for MacoNPV-A and MacoNPV-B, considered as distinct species, also exceeded 0.05 (dashed line underlined in Table 3 and Table 4). In contrast, K-2-P distances between the two viruses previously considered to be the same species, MbMNPV and MacoNPV-B, were consistent below 0.015 (shown in bold in Table 3 and Table 4). In contrast, when LaolNPV was compared with MacoNPV-A strains the K-2-P distances were intermediate, between 0.05 and 0.015, being of 0.049, 0.028, 0.026 and 0.039 for lef-8, lef-9, polh and the concatenated sequences, respectively (in bold and underlined in Table 3 and Table 4). In this case other characteristics, such as the host range need to be evaluated to establish species status.

Biological characterization
LaolNPV has a narrow host range compared to MacoNPV-A. The host range of LaolNPV was compared with that of MacoNPV-A, MacoNPV-B and MbMNPV at high (1x10 7 OBs/ml) ( Fig 5A) and moderate (1x10 5 OBs/ml) (Fig 5B) concentrations in eight lepidopteran species; M. brassicae, M. configurata, T. ni, C. chalcites, S. littoralis, S. exigua, S. frugiperda and H. armigera. In all cases dead larvae died due to the inoculated virus as showed in the REN profiles obtained with the OBs from cadavers (data not shown). The REN profile results indicate that cross-contamination did not occur during host range bioassays and OB treatments did not activate sublethal infections by homologous NPVs.
Although the mortality across species produced by MbMNPV was quite similar to those produced by MacoNPV-A and MacoNPV-B at high and moderate OB concentrations, MbMNPV showed a greater host range as it was pathogenic at the moderate OB concentration to species such as H. armigera. In general, MbMNPV was more pathogenic to the homologous host M. brassicae at high (100%) and moderate (89%) concentrations than to the heterologous hosts (Tukey, p<0.05). At the moderate OB concentration MbMNPV produced 100% mortality in M. brassicae and mortalities lower than 100% in all heterologous species, although these differences were not significant in the case of C. chalcites or S. exigua (Tukey, p>0.05).
LaolNPV showed a markedly different host range compared with MacoNPV-A. LaolNPV was not infective to M. brassicae or M. configurata even at the high OB concentration. LaolNPV was only pathogenic to C. chalcites at moderate and high concentrations, which resulted in 19 and 50% mortality, respectively. The host range of LaolNPV was completely different, therefore, to that of MacoNPV-A, indicating that these viruses occupy distinct ecological niches. Regrettably, we were unable to determine the pathogenicity of LaolNPV OBs in L. oleracea or A. australis as laboratory or field populations have not been available for testing.

Discussion
The present study aimed to characterize a novel NPV from diseased larvae of L. oleracea collected in 2011 from a field of alfalfa near Montpellier, France, although because of difficulties in differentiation of noctuid species in the larval stage we could not be certain of the identity of each and every diseased larva collected.
Restriction endonuclease analyses suggested that the baculovirus under study, LaolNPV, and MacoNPV-A diverged to a greater degree than MbMNPV and MacoNPV-B, that were previously considered to be the same viral species isolated from different hosts [4,15]. The differences in REN profiles were of a similar magnitude to those found between two different viruses, namely MacoNPV-A and MacoNPV-B. Sucrose density gradients demonstrated that, like MbMNPV, MacoNPV-A and MacoNPV-B [16], LaolNPV is a multiple nucleocapsid virus with a banding pattern quite similar to that of MacoNPV-A, suggesting a similar distribution of multiple nucleocapsids within the ODVs. Additionally, OB and ODV structural polypeptide profiles were also similar to MbMNPV, MacoNPV-A and MacoNPV-B, but with clear differences in the presence and molecular weights of several proteins. These results were consistent with the hypothesis that LaolNPV was closely related to, but distinct from the other NPVs that infect Mamestra spp.
Terminal sequencing information revealed that the most homologous NPV to the virus under study was MacoNPV-A 90/2 [14] closely followed by MacoNPV-A 90/4 [16]. The percentage of amino acid sequence identity between LaolNPV and MacoNPV-A was intermediate to that found between strains of MacoNPV-A or MbMNPV/MacoNPV-B, viruses of the same viral species, or between viruses of different species, such as MacoNPV-A and MbMNPV/ MacoNPV-B. When comparing with MbMNPV, the identity clearly decreased to a values corresponding to different species. Consequently, it appears that LaolNPV is more similar to the New World MacoNPV-A than to MbMNPV, a species from the Old World (Europe).
Phylogenetic analysis demonstrated that LaolNPV formed a separate branch within a MacoNPV-A clade, with high bootstrap support, as a distinct lineage from MbMNPV, MacoNPV-B and HearMNPV. Additionally, the baculovirus under study differed with respect to the MbMNPV/MacoNPV-B/HaMNPV group with a K-2-P distance greater than 0.05 in all cases, supporting the concept that LaolNPV is phylogenetically different from MbMNPV, HaMNPV or MacoNPV-B. In contrast, the MbMNPV/MacoNPV-B/HaMNPV group had K-2-P distances among the different viruses of less than 0.015.
In many cases the same virus isolated from different hosts has been assigned different names. This was the case with Anagrapha falcifera NPV (AnfaNPV) and Rachioplusia ou MNPV (RoMNPV) [43,51] and HearSNPV and HzSNPV [40]. Alternatively, different viruses isolated from the same species have received the same name [52]. Although both MacoNPV-A and MacoNPV-B were isolated from the same host and their genomes are closely related, MacoNPV-B represents a separately evolving virus from MacoNPV-A, as the K-2-P distances estimated in the present study were consistently higher than 0.05 (0.088), and exceed the genetic distance between RoMNPV and BmNPV (0.055), which are also thought to be different species [4,43]. MacoNPV-A is much more prevalent and infective to M. configurata populations than MacoNPV-B, however MacoNPV-B has higher pathogenicity to a wider range of host species. It is likely that there are alternative hosts for MacoNPV-B, so that infections in M. configurata are only commonly observed during outbreak years of high population density. It is intriguing that these two viruses could evolve divergently while infecting the same host, M. configurata [4,14,15]. This type of situation highlights the present drawbacks in baculovirus nomenclature as virus names are based on the host from which the virus was first isolated. However, it is also clear that host range can prove to be an informative aspect among the ecological characteristics that contribute to the baculovirus species definition criteria.
Baculovirus isolates from separate populations of a particular host species, or from closely related species, often differ in pathogenicity and the origin of a particular NPV is not a useful predictor of its virulence in a specific host, although local isolates tend to be more effective in controlling local insect populations than geographically distant isolates [6,13,21,53]. Surprisingly, although the baculovirus under study was phylogenetically close to MacoNPV-A, LaolNPV showed particular host range differences and was not pathogenic to M. brassicae or M. configurata, even at moderately high OB concentrations. In contrast, MacoNPV-A produced mortality across a variety of species tested. Therefore the effective host range of MacoNPV-A appears to be notably broader than that of LaolNPV. This suggests that these viruses occupy different ecological niches, and therefore, in combination with the phylogenetic analyses, are likely to represent different species. Surprisingly, the divergence in pathogenicity between viruses within related insect populations was also observed with the MacoNPV viruses. MacoNPV-B was more pathogenic than MacoNPV-A for several species. Additionally both MacoNPV-A and -B viruses were more pathogenic to M. brassicae than to the homologous host, M. configurata. This surprised us, as NPVs tend to be more pathogenic to homologous hosts rather than heterologous species [21,54,55]. However, in the present study we observed that MbMNPV was more pathogenic to M. configurata than MacoNPV-A or -B, which was also reported by Erlandson [6]. MbMNPV and MacoNPV-B represent variants of the same viral species, with distributions that include southern Canada and Europe [15]. Both viruses infect M. configurata, and as MbMNPV is common in Europe, it is possible that MacoNPV-B-like viruses may be found worldwide.
We conclude that Lacanobia oleracea nucleopolyhedrovirus (LaolNPV) should be considered as a new species of multiple nucleocapsid NPV in the Alphabaculovirus genus on the basis of REN profiles, gene sequence features and host range properties, in addition to the original insect host species and geographical point of isolation. Future studies should examine intraspecific variation in LaolNPV and establish the dose or concentration mortality response of this virus in its homologous host and sympatric noctuid species, such as Agrochola lychnidis, that are present in the alfalfa agroecosystem of southern Europe. Such information could prove valuable in evaluating the virus' potential as a biological insecticide for control of L. oleracea and related noctuid pests in this region.