BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers,academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research.

Mamestra brassicae Nucleopolyhedrovirus Infection andEnhancing Effect of Proteins Derived from Xestia c-nigrumgranulovirus in Larvae of Mamestra brassicae and Helicoverpaarmigera (Lepidoptera: Noctuidae) on CabbageAuthor(s): Shigeyuki Mukawa and Chie GotoSource: Journal of Economic Entomology, 103(2):257-264. 2010.Published By: Entomological Society of AmericaDOI: http://dx.doi.org/10.1603/EC09211URL: http://www.bioone.org/doi/full/10.1603/EC09211

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in thebiological, ecological, and environmental sciences. BioOne provides a sustainable onlineplatform for over 170 journals and books published by nonprofit societies, associations,museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated content indicatesyour acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercialuse. Commercial inquiries or rights and permissions requests should be directed to theindividual publisher as copyright holder.

BIOLOGICAL AND MICROBIAL CONTROL

Mamestra brassicae Nucleopolyhedrovirus Infection and EnhancingEffect of Proteins Derived From Xestia c-nigrum granulovirus inLarvae of Mamestra brassicae and Helicoverpa armigera

(Lepidoptera: Noctuidae) on Cabbage

SHIGEYUKI MUKAWA AND CHIE GOTO1

Insect Pest Management Research Team, National Agricultural Research Center, Kannondai,Tsukuba, Ibaraki 305-8666, Japan

J. Econ. Entomol. 103(2): 257Ð264 (2010); DOI: 10.1603/EC09211

ABSTRACT The insecticidal effect of Mamestra brassicae nucleopolyhedrovirus (MabrNPV) T5against Mamestra brassicae (L.) and Helicoverpa armigera (Hubner) (Lepidoptera: Noctuidae),important pests of various vegetables and ornamental crops in Japan as well as many other countries,and the enhancing activity of proteins derived from occlusion bodies of Xestia c-nigrum granulovirus(XecnGV) �-4, which was named GVPs, on the infectivity of MabrNPV were evaluated in a bioassaywith second-instar larvae fed on virus-applied cabbage, Brassica oleracea L. var. capitata, plants. Thelethal concentrations of MabrNPV achieving 95% mortality (LC95) were estimated to be 7.7 � 105 and1.8 � 105 OBs per ml forM. brassicae and H. armigera, respectively, with MabrNPV-alone treatment.Because the mean areas of cabbage leaf consumed by one larva in 60 h were not signiÞcantly differentbetweenM. brassicae and H. armigera, we conclude that MabrNPV is more infectious to H. armigerathan toM. brassicae. When larvae were fed on cabbage plants treated with 104 OBs per ml MabrNPVand various concentrations of the GVPs, the mortality of the two noctuid larvae increased in relationto GVP concentration. The requisite concentrations of the GVPs achieving 95% mortality with theMabrNPV for M. brassicae and H. armigera were estimated to be 5.93Ð8.30 and 1.94Ð3.48 �g/ml,respectively. In a comparison of the MabrNPV-alone treatment with equivalent 95% mortality,addition of GVPs increased the rate of larval death at younger instars, especially inM. brassicae. Ourresults indicate that GVPs are a potentially useful additive for improving the insecticidal efÞcacy ofMabrNPV.

KEYWORDS Mamestra brassicae nucleopolyhedrovirus, granulovirus proteins,Mamestra brassicae,Helicoverpa armigera, viral enhancement

Baculoviruses, including nucleopolyhedrovirus(NPV) and granulovirus (GV), infect only inverte-brate hosts and are effective and safe tools for in-tegrated pest management. There are several lim-iting factors on the practical use of baculoviruses,such as narrow spectrum and high production cost.From the commercial point of view, it is desirable todevelop new virus insecticides suitable for use on awider range of major pests.Mamestra brassicae (L.) is an important insect pest

of vegetables and ornamental plants in Europe andAsia. The bertha armyworm, Mamestra configurataWalker, a species closely related toM. brassicae, is alsoknown to be an important pest of cruciferous oilseedcrops in North America. An NPV isolate (MabrNPVT) obtained from M. brassicae in Japan has been ge-netically identiÞed as a variant of one NPV includingM. brassicae nucleopolyhedrovirus (MabrNPV) Ox-

ford strain and M. configurata NPV B (Mukawa andGoto 2006). MabrNPV Oxford is known to be infec-tious to many species belonging to the Noctuidae(Doyle et al. 1990), so it is highly possible thatMabrNPV T has a host range as wide as the Oxfordstrain. Helicoverpa armigera (Hubner) is also a majornoctuid pest of various crops in Asia, the eastern Pa-ciÞc Islands, Australia, southern Europe, and Africa(King 1994), infesting many common host plants withM. brassicae, such as cabbage, lettuce, and sweet pep-per (AEZ 2006). European isolates of MabrNPV, in-cluding the Oxford strain (�MbMNPVD), are highlyinfectious to H. armigera (Figueiredo et al. 1999,Rovesti et al. 2000). Our artiÞcial diet bioassay re-vealed that MabrNPV T is a potential viral insecticideagainst M. brassicae, H. armigera (Mukawa and Goto2007, Mukawa et al. 2008), and Autographa nigrisigna(Walker) (Goto et al. 2009). It would be a great ad-vantage to have one virus that can control severalimportant pests.1 Corresponding author, e-mail: cgoto@affrc.go.jp.

0022-0493/10/0257Ð0264$04.00/0 � 2010 Entomological Society of America

Some GVs and NPVs produce proteins named en-hancins that promote the viral infectivity of NPVs(Corsaro et al. 1993, Liu et al. 2006). Enhancins facil-itate the arrival of virions to host midgut cells viadisruption of the midgut peritrophic matrix by deg-radation of peritrophic matrix proteins and/or medi-ate fusion between the virion envelope and the cellplasma membrane (Corsaro et al. 1993). Four ho-mologs of the enhancin gene are found in the genomeof Xestia c-nigrum GV (XecnGV) (Hayakawa et al.1999) and H. armigera GV (HearGV) (Harrison andPopham 2008). An early study showed that intactocclusion bodies (OBs) of XecnGV act as an NPVenhancer in Þfth-instar larvae of X. c-nigrum (L.) butinterfere with the NPV infection in fourth instar larvae(Goto 1990). In addition, coinfection of HearGV withH. zeaNPV inHelicoverpa zea (Boddie) larvae causesinhibition of NPV replication in the host (Hackett etal. 2000). To avoid these interfering effects of GV onNPV infection, we have prepared GV proteins (GVPs)by removing XecnGV virions from alkaline-dissolvedOBs. Our bioassay revealed that adding GVPs at 100�g/g artiÞcial diet enhances the infectivity ofMabrNPV T to larvae of M. brassicae as effectively asadding the well-known NPV enhancer Tinopal at 1mg/g diet (Mukawa and Goto 2007). We also con-Þrmed the enhancing effect of GVPs on the infectivityof the NPV to H. armigera (Mukawa et al. 2008).However, the additional cost for production of GVPsalso should be taken into account. Gallo et al. (1991)reported that the probit mortality of Trichoplusia ni(Hubner) increased linearly with the log-dose ofenhancin of T. ni GV, when larvae were inoculatedwith a combination of Autographa californica NPV(AcMNPV) and various doses of the puriÞed en-hancin. The optimum concentrations of GVPs for NPVenhancement may vary depending on the target pests.

Chemicals and enzymes contained in food plantsaffect the infectivity of NPVs by reducing the OBsolubility in the host midgut (Hunter-Fujita et al.1998) and/or by accelerating the sloughing of NPV-infected midgut cells (Hoover et al. 2000). Beforestarting Þeld application of the virus, it is thereforenecessary to examine the viral efÞcacy against the pestspecies by bioassay using larvae fed on their hostplants. In this study, we evaluated the effect of GVPsat different concentrations on the infectivity ofMabrNPV by a bioassay using virus-applied cabbage,with the aim of developing a low-cost and effectiveformulation of MabrNPV T for controllingM.brassicaeand H. armigera.

Materials and Methods

Insects, Plant, Virus, andAdditives.M.brassicae andH. armigerawere collected in Tsukuba, Ibaraki, Japan,and maintained continuously in our laboratory for�10 and 4 yr, respectively. Larvae were reared on anartiÞcial diet containing mulberry leaf powder anddefatted soybean as the main ingredients (InsectaLFS, Nihon Nosan Kogyo Co., Ltd., Yokohama, Japan)

and maintained at 25�C under a photoperiod of 16:8(L:D) h.

Cabbage, Brassica oleracea L. var. capitata (ÔKinkei201�; Sakata Seed Co. Ltd., Yokohama, Japan) wassown in 36-well seedling trays (40 by 40 mm, height 45mm per well) containing fertilized soil (Kureha Co.Ltd., Tokyo, Japan) and grown in a growth chamberat 24�C in 14 h of light and 21�C in 10 h of darkness.In all the experiments, cabbage plants with three tofour fully opened true leaves were used at 26Ð31 dafter sowing.

We used theM.brassicaeNPV (MabrNPV) T5 clone(Mukawa and Goto 2006). The concentration ofMabrNPV in the viral stock suspension was deter-mined by counting the number of occlusion bodies(OBs) using a Petroff-Hausser bacteria counter underphase-contrast microscopy. Proteins derived from theOBs of Xestia c-nigrum GV �-4 clone (GVPs) wereprepared as described by Mukawa and Goto (2007).Quantification of Consumption Area of CabbageLeaf. The leaves of the cabbage plant were cut offexcept for one leaf (mean area, 34.9 cm2; range, 26.7Ð46.7 cm2). The leaf was immersed in 400 ml of 0.03%(vol:vol) spreader (Gramin S, Sankyo Agro Co. Ltd.,Tokyo, Japan) solution for 30 s and was dried for 1 h.The treated leafwasplaced ina200-mlplastic cupwitha cut for the petiole to pass through. The petiole waswrapped with a small piece of cotton to plug the cup.A larva in the pharate Þrst-instar was placed in the cupand allowed to ingest the leaf for 60 h. After removalof the larva, digital images of both sides of the leaf wereobtained using a scanner at a resolution of 300 dpi. Abinary format image was created by painting the feed-ing trace in black. Only one side of the feeding tracein the image was painted when the larva grazed oneside of the leaf surface. If the larvae had eaten holesin the leaf, both sides of the leaf were painted. Feedingarea was calculated using Ryushikaiseki III image anal-ysis software (SMT-Kashima Co. Ltd., Ibaraki, Japan).The experiment was replicated 32 and 35 times withM.brassicae and H. armigera, respectively.Calculation of Weight of Solution Applied on Cab-bage.The mean volume of the inoculum applied to theleaf surface was calculated based on the increase inweight after the treatment. Each cabbage leaf (meanarea, 39.1 cm2; range, 24.4Ð51.8 cm2) was cut from thebase of the petiole and weighed. Immediately afterdipping in a solution containing 0.03% (vol:vol) of aspreader for 30 s, the leaf was again weighed to cal-culate the amount of the solution on its surface. Thisexperiment was replicated 30 times.Bioassay with Cabbage Plants. The root ball of the

cabbage was wrapped individually in a plastic bag. Thecabbage shoot was immersed in 400 ml of MabrNPVsuspension containing 0.03% (vol:vol) of a spreaderwith or without GVPs for 30 s. The cabbage plant wasthen hung upside-down at room temperature for 1 hto dry the inoculum. The treated cabbage was then setinto a container consisting of two large clear plasticcups (126 by 126 mm, 91 mm in height) attachedtogether (top to top) with four small binder clips. Thecontainer had a ventilation hole (55 mm in diameter)

258 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 103, no. 2

covered by Þlter paper on the top. Twelve Þrst-instarlarvae at the molting stage, identiÞed by head capsuleslippage, were placed in the container. The newlymolted larvae were free to climb up the plant or theinside of the container, and eat the cabbage leavesduring the second stadium at 24 � 2�C under a pho-toperiod of 11:13 (L:D) h. After 64Ð66 h, the larvaewere transferred to 21-ml plastic cups for individualrearing on a virus-free artiÞcial diet at 25 � 1�C undera photoperiod of 16:8 (L:D) h. To examine the infec-tivity of MabrNPV alone, the concentrations ofMabrNPV used were 103, 104, 105, 106, and 107 OBs perml. Control larvae were fed on a cabbage treated witha 0.03% solution of the spreader. To determine theoptimum concentration of the GVPs, we prepared 104

OBs per ml of MabrNPV suspension containing dif-ferent concentrations of the GVPs: 0.562, 1, 1.78, 3.16,10, 31.6, and 100 �g/ml in the assay for M. brassicae;and 1, 3.16, 10, 31.6, and 100 �g/ml in the assay for H.armigera. The control larvae were fed on a cabbagetreated with 104 OBs per ml of MabrNPV alone. Wealso set up a negative control, in which larvae were fedon a cabbage treated with 100 �g/ml GVPs withoutMabrNPV. We prepared three plants for each treat-ment. Each experiment was replicated three times(nine plants with 108 larvae in total per individualconcentration of MabrNPV or GVPs). Larvae wereobserved daily for mortality until 13 d after inocula-tion. Larvae with typical symptoms of NPV infection,such as a waxy appearance and liquefaction of thecadaver, were recorded as infected. In the case ofdeath within 13 d of inoculation without typical viralsymptoms, tissue smears of the larva were preparedand examined under a phase-contrast microscope tocheck for the presence of OBs.Data Analysis. Area of cabbage leaf consumed per

larva was compared using a t-test for unequal vari-ances between M. brassicae and H. armigera, becauseLeveneÕs test for equality of variances was signiÞcant(F1,65 � 14.05; P � 0.01), although the ShapiroÐWilktest showed the data to be normally distributed (M.brassicae:W � 0.983, P� 0.893; andH. armigera:W �0.984, P � 0.911).

Mortality data were analyzed using a logistic re-gression against a common logarithm of concentra-tions of MabrNPV or GVPs. Data for the larval instarsat death by MabrNPV infection were analyzed usingan ordered logistic regression. Data were log-trans-formed after adding one to GVPs concentration, toinclude the data of the control of MabrNPV-alonetreatment as 0 �g/ml GVPs to the analysis. All statis-tical analyses were performed using the computersoftware statistics program JMP version 5.0.1 (SASInstitute 2002).

Results

Quantitative Measurement of Inoculum Ingestedby Second-Instar Larvae. The mean leaf areas of cab-bage plants consumed per larva ofM. brassicae andH.armigera were estimated to be 120.1 and 130.2 mm2

(with a standard deviation of 24.2 and 49.1 mm2),

respectively. The leaf area consumption of M. brassi-cae was not signiÞcantly different from that of H.armigera (t-test: t� 1.079, df � 50.57, P� 0.286). Themean volume of inoculum applied to the leaf surfacewas estimated to be 0.0252 �l/mm2 (range, 0.0115Ð0.0389 �l/mm2). Therefore, the amounts of inoculumingested by one larva of M. brassicae and H. armigerawere estimated to be 3.03 and 3.28 �l, respectively.Pathogenicity ofMabrNPVAlone. In the bioassay of

the MabrNPV treatment in M. brassicae and H. ar-migera, 98.8 and 95.2% of larvae, respectively, werecollected from the treated cabbage plants. When lar-vae were collected from the cabbage, 7.2% ofM. bras-sicae larvae were at the feeding second instar, 92.3%of them were at the pharate third instar, and theremainder (0.5%) had molted to the third instar. Theproportion of the larvae in each instar was not signif-icantly different between the control and theMabrNPV treatment at the maximum concentration of107 OBs per ml (likelihood ratio test: �2 � 0.168, df �1,P� 0.682). InH. armigera, 56.4% were at the feedingsecond instar, 34.8% at the pharate third instar, and8.6% at the feeding third instar at the collection time.The proportion of larvae at younger instars was sig-niÞcantly more in the treatment at the concentrationof 107 OBs per ml than in the control (likelihood ratiotest: �2 � 8.131, df � 2, P � 0.017), whereas theproportion of larvae at each instar was not signiÞcantlydifferent between the MabrNPV treatment at the con-centration of 106 OBs per ml and the control (likeli-hood ratio test: �2 � 2.664, df � 2, P � 0.264).

Among the collected larvae in the control, three of106 M. brassicae larvae and Þve of 100 H. armigeralarvae, respectively, died. No NPV OBs were detectedin any of the tissue smears of these larvae. AfterMabrNPV inoculation, nine of 534M. brassicae larvaeand 33 of 517 H. armigera larvae tested were consid-ered to have died without NPV infection after micro-scopic examination of their tissue smears. Becausethese nonviral mortality rates in MabrNPV treatmentwere not signiÞcantly different from those in the con-trols by each insect species (likelihood ratio test: M.brassicae,�2 � 0.563, df � 1,P� 0.453; andH.armigera,�2 � 0.292, df � 1, P� 0.589), the larvae that had diedfor reasons other than NPV infection were excludedfrom the after analysis.

ConcentrationÐmortality responses of M. brassicaeandH. armigera to MabrNPV were estimated by larvalmortality against all concentrations of MabrNPV byusing the logistic regression (Table 1). The medianlethal concentration (LC50) of MabrNPV was esti-mated to be 2.0 � 104 to 6.0 � 104 OBs per ml in M.brassicae and 4.4 � 103 to 1.1 � 104 OBs per ml in H.armigera, respectively. To test the effect of host spe-cies on the concentrationÐmortality response, we re-analyzed the pooled data of the three trials by a logisticregression against MabrNPV concentration and hostspecies. By the likelihood ratio test for the modeleffects, the larval mortality was signiÞcantly corre-lated with the MabrNPV concentration (�2 � 657.50,df�1,P�0.01)anddifferedbetweenM.brassicaeandH. armigera (�2 � 21.30, df � 1, P � 0.01). No inter-

April 2010 MUKAWA AND GOTO: MabrNPV ENHANCEMENT BY XecnGV PROTEINS 259

action between MabrNPV concentration and the hostspecies was detected (�2 � 0.19, df � 1, P � 0.660).

Most of the larvae infected with MabrNPV diedduring the third or fourth stadium in bothM. brassicaeand H. armigera (Fig. 1). After inoculation withMabrNPV in H. armigera, 258 larvae died in the thirdinstar in total, whereas this data included 56 individ-

uals that died at the pharate fourth instar or halfwaythrough ecdysis. In contrast, no visible sign indicatingthat larvae had reached the molting stage, such as headcapsule slippage, was observed in any of the 171 larvaeof M. brassicae that died in the third instar due toMabrNPV infection. An ordered logistic regressionanalysis was performed to test the effects of MabrNPVconcentration on the larval instar at death fromMabrNPV infection by host species (Table 2). A sig-niÞcant relationship with MabrNPV concentrationwas detected in both species, indicating that an in-crease in viral concentration causes larval death atyounger instars. Heterogeneity among the trials wasdetected, whereas the difference among cabbageplants was insigniÞcant.Effect of GVPs on Pathogenicity of MabrNPV. In

the bioassay of the MabrNPV plus GVPs treatment inM. brassicae and H. armigera, 99.28 and 94.97% oflarvae, respectively, were collected from the treatedcabbage plants. After 104 OBs per ml of MabrNPVinoculation without GVPs, inM.brassicae 4.6% were atthe feeding second instar, 83.3% at the pharate thirdinstar, and 12.0% at the feeding third instar at collec-

Table 1. The concentration–mortality response of M. brassicae and H. armigera to MabrNPV, estimated by logistic regression afterinoculation with MabrNPV alone

Host n Slope � SE LC50 (OBs per ml)95% CI

(OBs per ml)LC95 (OBs per ml)

95% CI(OBs per ml)

Goodness of Þttest

df �2 P

M. brassicaeTrial 1 175 2.05 � 0.29 6.0 � 104 3.4 � 104Ð1.0 � 105 1.6 � 106 6.9 � 105Ð6.9 � 106 3 0.811 0.847Trial 2 174 2.72 � 0.42 2.8 � 104 1.7 � 104Ð4.5 � 104 3.3 � 105 1.6 � 105Ð1.2 � 106 3 1.203 0.752Trial 3 176 2.01 � 0.30 2.0 � 104 1.1 � 104Ð3.5 � 104 5.8 � 105 2.4 � 105Ð2.6 � 106 3 0.725 0.867Poola 525 2.13 � 0.18 3.2 � 104 2.3 � 104Ð4.3 � 104 7.7 � 105 4.5 � 105Ð1.5 � 106 3 1.277 0.735H. armigera

Trial 1 164 1.65 � 0.28 4.4 � 103 1.9 � 103Ð8.6 � 103 2.7 � 105 9.5 � 104Ð1.9 � 106 3 3.756 0.289Trial 2 155 2.62 � 0.47 4.7 � 103 2.6 � 103Ð8.3 � 103 6.2 � 104 2.7 � 104Ð2.9 � 105 3 0.170 0.982Trial 3 165 2.15 � 0.34 1.1 � 104 6.2 � 103Ð2.0 � 104 2.6 � 105 1.1 � 105Ð1.2 � 106 3 3.584 0.310Poola 484 2.02 � 0.19 6.4 � 103 4.4 � 103Ð9.0 � 103 1.8 � 105 1.0 � 105Ð4.1 � 105 3 0.995 0.802

a To check the heterogeneity of the experiments among the trials and among the cabbages, the logistic regression for each host was performedagainst log10 (MabrNPV concn), the three trials, and the three cabbages with data for cabbages nested in the trials. Statistics from the pooleddata are shown, since the likelihood ratio tests for the effects did not detect heterogeneity among the trials (M. brassicae: �2 � 9.164, df � 2,P � 0.010; H. armigera: �2 � 6.522, df � 2, P � 0.038) and among the cabbages (M. brassicae: �2 � 6.170, df � 6, P � 0.404; H. armigera: �2 �16.667, df � 6, P � 0.018) at 1% signiÞcance level.

Fig. 1. Mortality rates by instar of M. brassicae (A) andH. armigera (B) inoculated with MabrNPV alone at thesecond instar. The three columns show the data for the Þrst,second, and third trials, respectively, at each concentration.Cadavers ofH. armigerawhich failed to shed their exuvia andremained in the pharate fourth instar were counted as larvaethat died in the third instar.

Table 2. Likelihood ratio test of the ordered logistic regressionfor the effect of MabrNPV concentration on the larval instar of M.brassicae and H. armigera at death from MabrNPV infectiona

Factor df �2 P

M. brassicaeMabrNPV concn 1 193.351 �0.01Trials 2 10.752 �0.01Cabbageb 6 0.505 0.998H. armigera

MabrNPV concn 1 149.863 �0.01Trials 2 7.073 0.029Cabbageb 6 4.278 0.639

a Effect of the whole model estimated by Þtting the ordered logisticmodel was signiÞcant (M. brassicae: �2 � 197.27, df � 9, P � 0.01; H.armigera: �2 � 156.35, df � 9, P� 0.01). Goodness of Þt test supportedthe null hypothesis of the model Þt (M. brassicae: �2 � 69.25, df � 105,P � 0.997; H. armigera: �2 � 84.78, df � 120, P � 0.994).b Three cabbages were designated as being nested in the trial.

260 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 103, no. 2

tion time from the cabbage, and in H. armigera 46.5,36.6, and 16.8%, respectively. The proportion of larvaeat the younger instars increased with increasing GVPsconcentration in both species. After inoculation withMabrNPV and 100 �g/ml GVPs, inM. brassicae 24.3%were at the feeding second instar, 63.6% at the pharatethird instar, and 12.2% at the feeding third instar; andinH. armigera 74.5, 18.6, and 6.9%, respectively. Whenan ordered logistic regression was used to test log10

(GVPs 1) for an order of 1Ð3 assigned to the feedingsecond instar, the pharate third instar, and the feedingthird instar, a signiÞcant effect of GVPs concentrationwas detected in both species (M. brassicae: �2 � 32.44,df � 1, P � 0.01; H. armigera: �2 � 28.17, df � 1, P �0.01).

Among the larvae in the negative control, fed oncabbage plants treated with 100 �g/ml GVPs withoutMabrNPV, two of 107M. brassicae larvae and 10 of 104H. armigera larvae died, respectively. No OBs of theNPV were detected in any tissue smear of these larvae.After 104 OBs per ml of MabrNPV inoculation with orwithout GVPs, Þve of 858M. brassicae larvae and 44 of615H. armigera larvae tested were concluded to havedied without NPV infection after microscopic exam-ination of their tissue smears. Because these nonviralmortality rates in MabrNPV treatment were not sig-niÞcantly different from those in the controls for eachspecies (likelihood ratio test:M. brassicae, �2 � 1.613,df � 1, P� 0.204; H. armigera, �2 � 0.727, df � 1, P�0.394), larvae that had died for reasons other than NPVinfection were excluded from the after analysis.

Concentration-mortality responses to GVPs wereestimatedby larvalmortality against all concentrations(M. brassicae) and 0Ð10 �g/ml concentrations (H.armigera) of GVPs using the logistic regression (Table3). The effective concentration of GVPs giving 95%mortality (EC95) was estimated to be 5.93Ð8.30 �g/mlin M. brassicae and 1.94Ð3.48 �g/ml in H. armigera,respectively. We did not carry out any statistical anal-ysis to test the difference between the host species,because heterogeneity among the trials inM. brassicaewas detected by the logistic regression (Table 3).

Most of the larvae infected with MabrNPV diedduring the third or fourth stadium in bothM. brassicaeand H. armigera (Fig. 2). After inoculation withMabrNPV in H. armigera, 413 larvae died in the thirdinstar and of that number 159 died at the pharatefourth instar or halfway through ecdysis. However,none of 435 larvae that died from MabrNPV-infection

Table 3. Effect of GVPs concentration on the mortality of larvae in M. brassicae and H. armigera, estimated by logistic regressionafter inoculation with a combination of MabrNPV (104 OBs per ml) and GVPs

Host n Slope � SE EC95 (�g/ml) 95% CI (�g/ml)Goodness of Þt test

df �2 P

M. brassicaeTrial 1 285 4.57 � 0.67 8.30 5.55Ð16.13 6 17.432 0.008Trial 2 283 4.90 � 0.76 5.93 4.03Ð11.38 6 14.563 0.024Trial 3 285 6.38 � 0.89 6.31 4.58Ð10.54 6 9.189 0.163Poola

H. armigeraTrial 1 115 3.59 � 1.11 3.48 1.75Ð23.93 2 4.382 0.112Trial 2 128 5.28 � 1.32 2.38 1.37Ð7.57 2 0.275 0.872Trial 3 139 5.79 � 1.43 1.94 1.14Ð5.62 2 0.203 0.904Poola 382 4.84 � 0.73 2.49 1.74Ð4.30 2 1.267 0.531

a In order to check the heterogeneity of the experiments among the trials and among the cabbages, the logistic regression for each host wasperformed against log10 (GVPs concn 1), the three trials, and the three cabbages with data for cabbages nested in the trial. Statistics fromthe pooled data were shown if the likelihood ratio tests for the effects did not detect heterogeneity among the trials (M. brassicae: �2 � 13.447,df � 2, P� 0.01; H. armigera: �2 � 0.230, df � 2, P� 0.892) and among the cabbages (M. brassicae: �2 � 12.438, df � 6, P� 0.053; H. armigera:�2 � 8.039, df � 6, P � 0.235) at 1% signiÞcance level.

Fig. 2. Mortality rates by instar of M. brassicae (A) andH. armigera (B) inoculated with a combination of MabrNPVand GVPs at the second instar. MabrNPV was applied alone(NPV alone) or in combination at 104 OBs per ml. The threecolumns show the data for the Þrst, second, and third trials,respectively, at each concentration. Cadavers ofH. armigerawhich failed to shed their exuvia and remained in the pharatefourth instar were counted as larvae that had died in the thirdinstar. NT indicates that the concentration was not tested.

April 2010 MUKAWA AND GOTO: MabrNPV ENHANCEMENT BY XecnGV PROTEINS 261

in the third instar developed to the pharate fourthinstar in M. brassicae. An ordered logistic regressionanalysis was performed to test the effects of GVPsconcentration on the larval instar at death fromMabrNPV infection by host species (Table 4). A sig-niÞcant relationship with GVPs concentration was de-tected in both species, indicating that an increase inGVPs concentration causes larval death at youngerinstars. Heterogeneity among the trials was detected,whereas the difference among cabbage plants wasinsigniÞcant.

Discussion

The concentration-mortality response of M. brassi-caewas signiÞcantly different from that ofH. armigeraby the logistic regression. Because the consumption ofcabbage leaf was not signiÞcantly different betweenM. brassicae and H. armigera, the lower LC50 of H.armigera indicates that this species is more susceptibleto MabrNPV than M. brassicae. Based on the LC50

values and the amounts of inoculum on the leaf surfaceingested by a larva, the LD50 values are estimated tobe 61Ð182 and 14Ð36 OBs per larva inM. brassicae andH. armigera, respectively. Our LD50 values forMabrNPV infection inH. armigera are similar to thosein a previous study (Rovesti et al. 2000). However, ourLD50 values inM.brassicaewere about nine-fold lowerthan those in the previous assays performed by leafdisk method (Evans 1981, Rovesti et al. 2000). Al-though, it is difÞcult to identify reasons for the dif-ference between LD50s of our results and those ofprevious studies for M. brassicae, the susceptibility ofthe larvae for the NPV may vary between the strainsof the host species in Japan and England. In ourpresent bioassay, larvae continuously fed on virus-contaminated leaves during the second stadium with-out any starvation, whereas in the previous studieslarvae probably experienced starvation for some timeafter they consumed leaf disks. These differences inhost strains and virus inoculation method may havemore of an effect on the susceptibility of larvae in M.brassicae than in H. armigera.

It is important to know the activity of the virus onthe target crop for pest management before perform-ing Þeld experiments, because some food plants de-crease the susceptibility of host insects to NPVs (Aliet al. 2002, Hunter-Fujita et al. 1998). For example,larvae of Heliothis virescens (F.) fed on cotton, Gos-sypiumhirsutumL., proved more resistant to infectionby AcMNPV than those fed on an artiÞcial diet(Hoover et al. 2000). In our previous study usingvirus-contaminated artiÞcial diet, the mean weight ofartiÞcial diet consumed per larva of H. armigera (9.87mg) during the second stadium was 1.54-fold higherthan that ofM. brassicae (6.41 mg), whereas the LD50

values of MabrNPV in H. armigera (2.4 � 102 to 9.5 �102 OBs per larva) do not differ from those in M.brassicae (3.3 � 102 to 1.1 � 103 OBs per larva)(Mukawa and Goto 2007, Mukawa et al. 2008). Com-parison of the results of the two bioassays using theartiÞcial diet and cabbage plants suggested that theformer method may not always be a good alternativeto the latter for the examination of the applicability ofnew viral insecticides. The rather higher LD50 valuesin the artiÞcial diet bioassay than in the cabbage bio-assay might have been caused by antiviral activitiesof some ingredients, such as mulberry leaf powderand/or preservatives contained in the ready-made ar-tiÞcial diet Insecta LFS. The assay using an artiÞcialdiet provides a useful tool in screening for synergisticactivity of some chemicals or biological compounds oninfection of NPVs, whereas the assay using host plantsis indispensable to estimate a feasible concentration ofMabrNPV for future Þeld study.

After enhancins were identiÞed from some GVs andNPVs, researchers have focused their attention on themode of action of the viral enhancement (Corsaro etal. 1993, Liu et al. 2006), and few attempts have beenmade toward the practical use of the proteins for pestmanagement. Our previous study has shown that theinfectivity of MabrNPV toM.brassicaeandH.armigeraincreases 70.7- to 81.5-fold and 13.8- to 38.5-fold re-spectivelybyadding1mg/ml(Þnal concentration, 100�g/g diet) of the GVPs to the virus-contaminated diet(Mukawa and Goto 2007, Mukawa et al. 2008). As thenext step in development of new viral insecticidesusing GVPs, we tried to Þnd the optimum concentra-tion of GVPs by adding various concentrations ofGVPs to 104 OBs per ml of MabrNPV which has aconcentration 77-fold lower than the LC95 from thepooled data of the MabrNPV-alone treatment againstM.brassicae.Concentrations of the GVPs for achieving95% mortality of larvae inoculated with 104 OBs per mlof MabrNPV (EC95) were estimated to be under 10�g/ml in bothM. brassicae andH. armigera.This resultindicates that the GVPs can be used at much lowerconcentrations than in our previous studies (Mukawaand Goto 2007, Mukawa et al. 2008) with sufÞcienteffect on MabrNPV enhancement.

When the larvae were collected from the virus-treated cabbage, the proportion ofM. brassicae larvaein each instar was not different among all NPV con-centration in the MabrNPV alone treatment. How-ever, the proportion of H. armigera larvae at younger

Table 4. Likelihood ratio test of the ordered logistic regressionfor the effect of GVPs concentration on the larval instar of M.brassicae and H. armigera at death from MabrNPV infectiona

Factor df �2 P

M. brassicaeGVPs concn 1 281.502 �0.01Trials 2 11.222 �0.01Cabbageb 6 12.619 0.050H. armigera

GVPs concn 1 81.215 �0.01Trials 2 23.378 �0.01Cabbageb 6 3.337 0.766

a Effect of the whole model estimated by Þtting the ordered logisticmodel was signiÞcant (M. brassicae: �2 � 286.66, df � 9, P � 0.01; H.armigera: �2 � 104.19, df � 9, P� 0.01). Goodness of Þt test supportedthe null hypothesis of the model Þt (M. brassicae: �2 � 121.20, df �195, P � 1.000; H. armigera: �2 � 115.64, df � 150, P � 0.983).b Three cabbages were designated as being nested in the trial.

262 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 103, no. 2

instars increased in the treatment at the highest con-centration of 107 OBs per ml. The inhibition of larvalgrowth after the inoculation with MabrNPV at 100-fold of LC95 in H. armigera might be caused by somedefensive response of the host such as sloughing of theinfected midgut cells into the gut lumen (Volkman2007), but inoculation of MabrNPV at lower dosesinduced no apparent response of larvae. However, theproportion of larvae in the earlier instars when thelarvae were collected from the virus-treated cabbagewas increased in relation with the GVPs concentrationin the MabrNPV plus GVPs treatment. Enhancins inGV occlusion bodies have a protease activity and dis-rupt peritrophic matrix by digesting the peritrophicmatrix proteins in the host midgut (Lepore et al. 1996).Because the peritrophic matrix plays important rolesin protection of midgut cells from mechanical damageand invasion of microorganisms (Terra 2001), it mightbe possible that the GVPs concentration reßects in-duction of host defense mechanism and prevention ofthe larval ingestion.

The concentration of MabrNPV mixed into the ar-tiÞcial diet affects the larval instar at death in M.brassicae (Mukawa and Goto 2007), whereas no sim-ilar pattern is clear in H. armigera (Mukawa et al.2008). Our current study using cabbage plants showedthat larvae died at signiÞcantly younger instars withincreasingconcentrationofMabrNPVorGVPs inbothM.brassicae andH. armigera.WhenM.brassicae larvaewere inoculated with approximately a 95% mortalityconcentration of MabrNPV alone (106 OBs per ml) orMabrNPV(104 OBsperml)plusGVPs(10 �g/ml), therates of the larvae which died before reaching thefourth instar were higher in the latter treatment thanin the former treatment. This result indicates that theaddition of GVPs improves the virulence of MabrNPVby killing larvae at younger instars in M. brassicae.Because progress of larval instars causes an increase incrop losses, addition of GVPs will reduce not only theoptimal effective concentration of MabrNPV, but alsothe foliar damage, by killing pests at younger stadia.

AmongH. armigera larvae killed by MabrNPV at thethird instar after inoculation with MabrNPV alone andMabrNPV plus GVPs, 21.7 and 38.5%, respectively, hadreached the pharate fourth instar. However, no signsof incipient molting, such as head capsule slippage,were observed in any of the M. brassicae larvae withapparent viral symptoms. A similar phenomenon wasseen in our previous studies (Mukawa and Goto 2007,Mukawa et al. 2008). The cuticle ofH. armigera larvae,died at the pharate fourth instar, did not easily ruptureafter the liquefaction of the internal tissues, and theOBs produced in the host remained in the cadavers.This phenomenon suggests that the spread ofMabrNPV inH. armigeramight be less than that inM.brassicae, in spite of higher susceptibility of the formerspecies to the virus.

Our results reveal the potential for simultaneouscontrol of M. brassicae and H. armigera with 104 OBsper ml of MabrNPV by adding GVPs at concentrations�10 �g/ml, whereas �106 OBs per ml of MabrNPVwill be necessary to control both host species without

GVPs. However, a higher concentration of MabrNPVor GVPs may be necessary in the Þeld, because envi-ronmental factors such as UV radiation inactivate bac-ulovirus insecticides (Shapiro and Domek 2002).However, lower concentrations of these agents maybe sufÞcient, because we can expect additional pestdensity suppression by indigenous natural enemies,which as nontarget organisms are immune to this mi-crobial insecticide (van Driesche and Bellows 1996).Further Þeld tests are needed based on our estimateof effective combinations of MabrNPV and GVPs, todetermine practical concentrations of MabrNPV andGVPs.

Acknowledgments

We are grateful to Youichi Kobori for valuable suggestionson the bioassay using cabbage plants and to Yoshito Suzukifor critical reading of the manuscript. We thank TakayukiMitsunaga for helpful advice on the statistical analysis. Wealso thank Keiko Ozawa and Yumiko Togashi for insect rear-ing and plant cultivation. This research was supported by aGrant-in-Aid “Development of new biorational techniquesfor sustainable agriculture” from the Ministry of Agriculture,Forestry and Fisheries, Japan.

References Cited

[AEZ]JapaneseSocietyofAppliedEntomologyandZoology.2006. Major insect and other pests of economic plants inJapan. Rev. ed. AEZ, Tokyo, Japan.

Ali,M. I., S. Y. Young,G.W.Felton, andR.W.McNew. 2002.Inßuence of the host plant on occluded virus productionand lethal infectivity of a baculovirus. J. Invertebr. Pathol.81: 158Ð165.

Corsaro,B.G.,M.Gijzen,P.Wang, andR.R.Granados. 1993.Baculovirus enhancing proteins as determinants of viralpathogenesis, pp. 127Ð145. InN. E. Beckage, S. N. Thomp-son, and B. A. Federici [eds.], Parasites and pathogens ofinsects, vol. 2. Pathogens. Academic, New York.

Doyle, C. J., M. L. Hirst, J. S. Cory, and P. F. Entwistle. 1990.Risk assessment studies: detailed host range testing ofwild-type cabbage moth, Mamestra brassicae (Lepidop-tera: Noctuidae), nuclear polyhedrosis virus. Appl. En-viron. Microbiol. 56: 2704Ð2710.

Evans, H. F. 1981. Quantitative assessment of the relation-ships between dosage and response of the nuclear poly-hedrosis virus ofMamestra brassicae. J. Invertebr. Pathol.37: 101Ð109.

Figueiredo, E., D.Munoz, A. Escribano, A.Mexia, J.M. Vlak,and P. Caballero. 1999. Biochemical identiÞcation andcomparative insecticidal activity of nucleopolyhedrovi-rus isolates pathogenic forHeliothis armigera (Lep., Noc-tuidae) larvae. J. Appl. Entomol. 123: 165Ð169.

Gallo, L. G., G. B. Corsaro, P. R.Hughes, andR. R.Granados.1991. In vivo enhancement of baculovirus infection bythe viral enhancing factor of a granulosis virus of thecabbage looper, Trichoplusia ni (Lepidoptera: Noctu-idae). J. Invertebr. Pathol. 58: 203Ð210.

Goto,C. 1990. Enhancement of a nuclear polyhedrosis virus(NPV) infection by a granulosis virus (GV) isolated fromthe spotted cutworm, Xestia c-nigrum L. (Lepidoptera:Noctuidae). Appl. Entomol. Zool. 25: 135Ð137.

Goto, C., S. Mukawa, and Y. Suzuki. 2009. Virus control ofnoctuid pests, pp. 133Ð136. In S. Miyai, A. Kawai, H.Hagiwara, K. Takahashi, K. Yoshida, H. Yoshitomi, T.

April 2010 MUKAWA AND GOTO: MabrNPV ENHANCEMENT BY XecnGV PROTEINS 263

Kimura, and S. Moriya [eds.], The latest biorational tech-niques for integrated pest and weed management andreduction of fertilizer. National Agricultural ResearchCenter, Tsukuba, Japan.

Hackett, K. J., A. Boore, C. Deming, E. Buckley, M. Camp,and M. Shapiro. 2000. Helicoverpa armigera granulovi-rus interference with progression of H. zea nucleopoly-hedrovirus disease in H. zea larvae. J. Invertebr. Pathol.75: 99Ð106.

Harrison, R. L., andH. J. Popham. 2008. Genomic sequenceanalysis of a granulovirus isolated from the Old Worldbollworm, Helicoverpa armigera. Virus Genes. 36: 565Ð581.

Hayakawa, T., R. Ko, K. Okano, S. I. Seong, C. Goto, and S.Maeda. 1999. Sequence analysis of the Xestia c-nigrumgranulovirus genome. Virology 262: 277Ð297.

Hoover, K., J. O.Washburn, and L. E. Volkman. 2000. Mid-gut-based resistance of Heliothis virescens to baculovirusinfection mediated by phytochemicals in cotton. J. InsectPhysiol. 46: 999Ð1007.

Hunter-Fujita, F. R., P. F. Entwistle, H. F. Evans, and N. E.Crook. 1998. Insect viruses and pest management.Wiley, Chichester, United Kingdom.

King, A.B.S. 1994. Heliothis/Helicoverpa (Lepidoptera: Noctu-idae), pp. 39Ð106. In G. A. Matthews and J. P. Tunstall[eds.], Insect pests of cotton. CAB International, Cam-bridge, United Kingdom.

Lepore, L. S., P. R. Roelvink, and R. R. Granados. 1996.Enhancin, the granulosis virus protein that facilitatesnucleopolyhedrovirus (NPV) infections, is a metallopro-tease. J. Invertebr. Pathol. 68: 131Ð140.

Liu, S., H. Li, S. Sivakumar, and B. C. Bonning. 2006. Virus-derived genes for insect-resistant transgenic plants, pp.427Ð457. In B. C. Bonning [ed.], Advances in virus re-

search, vol. 68. Insect viruses: biotechnological applica-tions. Elsevier, San Diego, CA.

Mukawa, S., and C. Goto. 2006. In vivo characterization of agroup II nucleopolyhedrovirus isolated from Mamestrabrassicae (Lepidoptera: Noctuidae) in Japan. J. Gen. Vi-rol. 87: 1491Ð1500.

Mukawa, S., and C. Goto. 2007. Enhancement of nucle-opolyhedrovirus infectivity against Mamestra brassicae(Lepidoptera: Noctuidae) by proteins derived fromgranulovirus and a ßuorescent brightener. J. Econ. En-tomol. 100: 1075Ð1083.

Mukawa, S., C. Goto, and Y. Suzuki. 2008. Effect of proteinsderived from granulovirus on the infectivity ofMamestrabrassicae nucleopolyhedrovirus in Helicoverpa armigera(Lepidoptera: Noctuidae). Appl. Entomol. Zool. 43: 323Ð329.

Rovesti,L.,N.E.Crook, andD.Winstanley. 2000. Biologicaland biochemical relationships between the nucleopoly-hedroviruses of Mamestra brassicae and Heliothis ar-migera. J. Invertebr. Pathol. 75: 2Ð8.

SAS Institute. 2002. JMP statistics and graphics guide, ver-sion 5. SAS Institute, Cary, NC.

Shapiro, M., and J. Domek. 2002. Relative effects of ultra-violet and visible light on the activities of corn earwormand beet armyworm (Lepidoptera: Noctuidae) nucle-opolyhedroviruses. J. Econ. Entomol. 95: 261Ð268.

Terra, W. R. 2001. The origin and functions of the insectperitrophic membrane and peritrophic gel. Arch. InsectBiochem. Physiol. 47: 47Ð61.

van Driesche, R. G., and T. S. Bellows, Jr. 1996. Biologicalcontrol. Chapman & Hall, New York.

Volkman, L. E. 2007. Baculovirus infectivity and the actincytoskeleton. Curr. Drug Targets 8: 1075Ð1083.

Received 26 June 2009; accepted 7 December 2009.

264 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 103, no. 2