Insecticidal Properties of Genetically Engineered Baculoviruses ...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, May 1992, p. 1583-1591

Vol. 58, No. 5

0099-2240/92/051583-09$02.00/0 Copyright ©3 1992, American Society for Microbiology

Insecticidal Properties of Genetically Engineered Baculoviruses Expressing an Insect Juvenile Hormone Esterase Gene RUSS ELDRIDGE,' DAVID R. O'REILLY,l12t* BRUCE D. HAMMOCK,3 AND LOIS K. MILLER"12 Department of Entomology' and Department of Genetics, 2 University of Georgia, Athens, Georgia 30602, and Departments of Entomology and Environmental Toxicology, University of California, Davis, California 956163 Received 15 October 1991/Accepted 29 February 1992

Exploring the possibility of enhancing the properties of baculoviruses as biological control agents of insect we tested the effect of expressing an insect gene (ihe) encoding juvenile hormone esterase. Juvenile hormone esterase inactivates juvenile hormone, which regulates the outcome of an insect molt. A cDNA encoding the juvenile hormone esterase of Heliothis virescens was inserted into the genome of Autographa californica nuclear polyhedrosis virus such that the gene was expressed under the control of a strong, modified viral promoter. This virus, however, naturally encodes an ecdysteroid UDP-glucosyltransferase which inactivates ecdysone, the hormone which initiates molting. Since ecdysteroid UDP-glucosyltransferase could mask the effects ofjhe expression by blocking molting entirely, jhe-expressing viruses in which the ecdysteroid UDP-glucosyltransferase gene was deleted or disrupted were constructed. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of proteins from infected cells revealed several intracellular proteins and two major secreted proteins which reacted with antibodies to authentic juvenile hormone esterase. Western blot analysis coupled with tunicamycin treatment indicated that differential glycosylation was responsible for the multiple products. Hemolymph of recombinant virus-infected fourth-instar Trichoplusia ni larvae contained levels of juvenile hormone esterase activity 40-fold higher than maximal levels found in uninfected larvae. However, little or no difference in developmental characteristics, weight gain, or time of mortality was observed between insects infected with the jhe-expressing viruses and control viruses. pests,

The potential of using baculoviruses as alternatives to the chemical control of insects has long been recognized (16). Unlike many synthetic chemical pesticides, baculoviruses have minimal environmental impact, have shown no mammalian toxicity, and have high target specificity (8). However, a primary disadvantage of baculoviruses is their inability to kill target insects rapidly. In contrast to the rapid effects observed with many synthetic chemical pesticides, insects infected with baculoviruses continue to live and feed for many days after infection. An important goal of genetically engineering these viruses is to shorten this postinfection feeding period. Using the insects' own hormones or hormone-regulatory proteins represents a possible route for enhancing the control properties of baculoviruses. Juvenile hormone (JH) plays a pivotal role in regulation of insect development (7, 32). In response to ecdysteroids, the presence of JH during larval instars regulates the expression of specific genes so that a larval-larval molt ensues (27). During the last larval instar of members of the order Lepidoptera, JH levels in the hemolymph must be reduced for ecdysone to trigger a pupal molt. This decrease in JH titer occurs by reduced JH biosynthesis and by active degradation of circulating JH by the hydrolyzing enzyme JH esterase (JHE) (6, 10). Recently a cDNA encoding JHE from the moth Heliothis virescens (Lepidoptera: Noctuidae) was cloned and expressed by using a baculovirus expression vector (11). The vector produced active JHE, and effects of the expressed enzyme on the size and coloration of infected larvae were reported. While relatively high levels of expression of active

JHE were detected in culture fluid from infected cells, only nominally increased levels were found in the hemolymph of young larvae of Trichoplusia ni (Lepidoptera: Noctuidae) larvae fed occlusion-negative recombinant virus. Furthermore, no metamorphic abnormalities were evident in infected insects, although a reduction in weight gain was observed in larvae that were fed budded virus. The Autographa californica nuclear polyhedrosis virus (AcMNPV) expression vector used in these studies encoded ecdysteroid UDP-glucosyltransferase (EGT), which inactivates ecdysone by conjugating the hydroxyl group at C-22 with a sugar (23, 24). Thus, insects infected with a virus containing the gene encoding EGT (egt) do not molt because of a lack of active ecdysone. This inactivation of ecdysone would be expected to mask the metamorphic effects of JHE, since ecdysone is required to initiate a molt and JH influences only the outcome of the molt. Furthermore, the deletion of egt enhances the properties of AcMNPV as a pesticide because insects infected with egt- viruses attempt to molt and feed less during the infection (25). Thus, we were interested in determining whether a recombinant virus lacking egt and overexpressing the gene encoding JHE (jhe) would induce a premature pupal molt. Premature pupation would be expected to limit feeding damage by infected insects more than larval-larval molts would. Thus, an egt-, jhe-expressing virus was constructed. High levels of active JHE were observed in both infected-cell culture and the hemolymph of infected T. ni larvae. However, premature pupation did not occur in infected larvae, and little or no reduction in feeding was observed relative to larvae infected with the control virus, either with neonates or with late-instar larvae. Thus, the expression of jhe provides no significant improvement in the properties of AcMNPV as a biopesticide by using these assays.

* Corresponding author. t Present address: Department of Biology, Imperial College, London SW7 2BB, England.

1583

1584

APPL. ENVIRON. MICROBIOL.

ELDRIDGE ET AL.

MATERIALS AND METHODS

EcoRI I GAAGTCATGTCTGTTGGAATTCGATAGATCTATT

Cells and virus. AcMNPV L-1 (19) was propagated and its titer was determined in Spodoptera frugiperda (Lepidoptera: Noctuidae) IPLB-SF21 cells (SF21 [30]) maintained in complete TC-100 medium (GIBCO Laboratories, Gaithersburg, Md.). Complete TC-100 medium is TC-100 medium supplemented with tryptose broth (GIBCO) and 10% fetal bovine serum (21). For some experiments, SF21 cells maintained in SF-900 serum-free medium (GIBCO) were used. Three genetically modified AcMNPVs were used for construction of jhe expression vectors. The first virus, designated vEGTZ, is an AcMNPV L-1 derivative in which the Eschenichia coli lacZ gene is inserted in frame in the viral egt gene (24). The second virus, designated vEGTDEL, is a mutant which has a 1.1-kb deletion in the egt open reading frame (ORF) (23). The third virus, designated vEGTSyngal+, has the same egt deletion as vEGTDEL but also carries a lacZ gene under the control of a synthetic viral promoter in place of the polyhedrin gene. This virus facilitated screening for occlusionpositive recombinants expressing JHE (see below). Construction of transplacement plasmids. Two different transplacement plasmids were constructed for the insertion of jhe into AcMNPV. In the first, a 1.75-kb cDNA fragment encoding JHE was excised from pJHE16B (11, 13) by digestion with EcoRI and KpnI and gel purified. This fragment was then ligated into the EcoRI and KpnI sites of the multiple-cloning region of plasmid vector pEVmXIV (31) by standard molecular biological techniques (20). pEVmXIV contains 2.3 kb of AcMNPV sequences flanking a multiplecloning site which is immediately downstream of the XIVmodified polyhedrin promoter (31). The flanking regions of this plasmid facilitate allelic replacement into the AcMNPV genome, and the XIV promoter yields levels of expression higher than that of the wild-type (wt) polyhedrin promoter (22, 26). The plasmid containing jhe was designated

CTTCAGTACAGACAACCTTAAGCTATCTAGATAA

pXIVJHE. The second transplacement plasmid, designated pSPX IVJHE (Fig. 1) was constructed by digesting pJHE16B with KpnI, blunt ending the 3' overhanging ends with T4 DNA polymerase, and then cutting the linearized plasmid with EcoRI to give the 1.75-kb JHE cDNA. This fragment was gel purified and then ligated into the EcoRI and SmaI sites of the multiple-cloning site of the transplacement plasmid pSynXIVVI+X3. This transplacement plasmid has two modified viral promoters placed in tandem which have been shown to drive very high levels of expression of genes placed under their control (31). One of the promoters is XIV (described above), while the other is a synthetic promoter and consists of consensus sequences from a number of AcMNPV late and very late promoters (31). Flanking the MCS and promoter regions of this plasmid are the AcMNPV 603 ORF on one side and the wt polyhedrin promoter and ORF on the other side (Fig. 1). These flanking sequences facilitated allelic replacement during cotransfection with viral genomic DNA. The presence of the polyhedrin promoter and ORF results in an occlusion-positive recombinant virus. The junctions between the promoters and the cDNA of the plasmids were confirmed by DNA sequencing by the method of Chen and Seeburg (4). Construction of recombinant viruses. Three jhe-expressing recombinant viruses were generated by allelic replacement (21). The first virus, designated vJHEwt, is a recombinant in which jhe is expressed in an egt+ background and was generated by cotransfection of SF21 cells with pXIVJHE and wt AcMNPV DNA. This virus was identified by an

I

MetI

JH~~~~~F SaclI

Xho I SmalI/Kpn I

Xho I

Polyhedrin Gene>

EcoRV HindM HindM

Kpn I

FIG. 1. Context of the jhe gene in the transplacement plasmid pSPXIVJHE and in vJHEEGTD. The direction of the JHE cDNA ORF is indicated by the arrow labeled JHE. Small arrows labeled XIV and SP are modified viral promoters that drive expression of jhe. The 603 ORF and the polyhedrin gene flank the JHE cDNA and promoter region. Polyhedrin gene expression results in an occlusion-positive recombinant virus. The sequence of the junction between the JHE cDNA and pSynXIVVI+X3 is shown above the schematic diagram. The EcoRI site used for cloning and the initiating Met codon are indicated.

occlusion-negative phenotype caused by the loss of the polyhedrin gene due to allelic replacement with jhe. The second viral derivative, designated vJHEEGTZ, is a jheexpressing recombinant virus in which egt has been disrupted by the E. coli lacZ gene. This virus was identified by an occlusion-negative, blue-plaque phenotype when screened in the presence of a chromogenic indicator of lacZ, 5-bromo-4-chloro-3-indolyl-p-galactoside (X-Gal). The third virus, designated vJHEEGTD, is an occlusion-positive, jheexpressing recombinant which contains the egt deletion. This virus was generated by cotransfection of pSPXIVJHE and vEGTSyngal+ and was isolated by screening for white occlusion-positive recombinant plaques against the blue occlusion-negative plaques of the vEGTSyngal+ background. After plaque purification of occlusion-positive plaques, a second screen for vJHEEGTD consisted of assaying JHE levels in the tissue culture fluid from infected cells by the partition assay of Hammock and Roe (12). Only those recombinant viruses that yielded high levels of JHE activity were included for further characterization. All viral constructs were plaque purified three times, and their genotypes were confirmed by restriction digestion and probing of viral genomic Southern blots. Viral inocula. High-titer stocks of budded virus for injection bioassays as well as JHE activity assays were generated by infecting 100-ml spinner flasks of SF21 cells (106 cells per ml) at a multiplicity of infection of 0.5 PFU per cell and harvesting culture fluid at 4 days postinfection (p.i.). Polyhedral inclusion bodies (PIBs) for per os bioassays were obtained by injecting fifth-instar T. ni larvae with 5 x 104 PFU of budded vEGTDEL, vJHEEGTD, or wt virus and collecting cadavers soon after death. Each insect was homogenized in 1 ml of distilled water with a high-speed homogenizer. The number of PIBs per milliliter of homogenate was determined by using a hemocytometer. Radioactive labeling and analysis of proteins. SF21 cells (106/35-mm-diameter plate) were infected with either vJHEEGTD or vEGTDEL at a multiplicity of infection of 20 PFU per cell. After a 1-h adsorption period, the inocula were removed and replaced with complete TC-100 medium. Two hours before the appropriate time point, the medium was replaced with incomplete TC-100 medium which also lacked cysteine. The cells were labeled 1 h later by replacing the cysteine-free medium with incomplete medium containing 25 VCi of L-[35S]cysteine (New England Nuclear, Boston,

VOL. 58, 1992

INSECTICIDAL PROPERTIES OF ENGINEERED BACULOVIRUSES

Mass.; 1,000 Ci/mmol). At the end of a 1-h room temperature incubation period, the labeled medium was removed and replaced with 500 p.l of phosphate-buffered saline (PBS), pH 6.2 (1 mM Na2HPO4 7H20, 10.5 mM KH2PO4, 140 mM NaCl, 40 mM KCl), and the cells were incubated for 1 h. The PBS containing secreted products was then removed and saved by freezing at -80°C. After removal of the PBS, cells were lysed by incubation on ice in 50 ,ul of lysis buffer (1% Nonidet P-40, 150 mM NaCl, 50 mM Tris-HCl [pH 8.0]). The cell lysates were collected after 30 min and stored at -80°C. The cell lysates and supematants were boiled for 5 min in a loading buffer consisting of 2% sodium dodecyl sulfate (SDS), 62.5 mM Tris-HCl (pH 6.7), 15% glycerol, and 100 mM dithiothreitol, and the proteins were analyzed by electrophoresis on vertical 10% polyacrylamide gels in the presence of SDS (18). Protein standards were visualized by staining the gels with Coomassie brilliant blue. The gels were prepared for fluorography by the method of Bonner and Laskey (2). Western blot analysis. Western blot (immunoblot) analysis was done by a modification of the methods of Burnette (3) and Blake et al. (1). Briefly, SF21 cells (2 x 106/60-mmdiameter plate) that had been maintained in SF-900 serumfree medium were infected at a multiplicity of infection of 20 PFU per cell with either vJHEwt, vJHEEGTZ, AcMNPV L-1 (wt), or vEGTZ for 1 h. After adsorption, the inocula were removed and the cells were incubated in SF-900 medium for 12 h. After incubation, the medium was replaced with 1 ml of incomplete TC-100 medium either with or without 5 p.g of tunicamycin (Sigma Chemical Co., St. Louis, Mo.). The cells were rocked gently for 24 h at 27°C, and the tissue culture fluid and cell lysates were collected. Proteins were separated by SDS-polyacrylamide gel electrophoresis (PAGE) and transferred to an Immobilon nylon blotting membrane (Millipore, Bedford, Mass.) with a Transblot apparatus (Bio-Rad Laboratories, Richmond, Calif.). The membrane was blocked for 2 h at 37°C with 3% bovine serum albumin in 20 mM Tris-HCl (pH 7.5)-500 mM NaCl and then probed at 37°C with diluted (1:1,500) polyclonal antiserum against JHE (14). Bound antibodies were detected with alkaline phosphatase-conjugated anti-rabbit immunoglobulin G diluted 1:3,000 (Sigma). As a control, an identical blot was treated similarly and probed with nonimmune serum. Protein size markers transferred to the membrane were visualized separately by staining with Coomassie brilliant blue. JHE activity bioassay. T. ni eggs were provided by Beth Gray of Abbott Laboratories (Chicago, Ill.). The insects were reared on an alfalfa diet as previously described (9). T. ni larvae that were approximately 24 h into the fourth instar were injected with S x 104 pfu of budded vJHEEGTD, wt, or vEGTDEL virus. At 24-h intervals p.i., the insects were bled, and the hemolymph was assayed for JHE activity with [10-3H(N)]JH III (New England Nuclear; 17 Ci/mmol) as a substrate by the partition assay method of Hammock and Roe (12). Insect bioassays. Bioassays of late-instar T. ni were carried out in two ways. First, to obtain a synchronous infection, larvae (25 to 30 individuals) approximately 24 h into the fourth or fifth instar were injected with 5 x 104 PFU of budded wt, vEGTDEL, or vJHEEGTD virus or mock infected in a volume of 2 p.l. After infection, the insects were weighed and observed for developmental abnormalities at 24-h intervals p.i. A per os bioassay was also carried out with fourth- and fifth-instar T. ni larvae. Insects beginning to molt, as deter-

1585

mined by head capsule slippage, were transferred to separate containers without food or water and examined every 5 h for completion of molting. The newly molted insects were starved for up to 18 h and then allowed to feed on 1 p.l (fourth instar) or 2 p.l (fifth instar) of a 10% sucrose solution containing 5 x 103 viral PIBs or, as a mock infection control, a 10% sucrose solution without viral PIBs. The larvae (30 per treatment) were transferred to an uninfected diet, weighed every 24 h p.i., and observed for mortality and developmental phenotype. Weight means for both bioassays were computed and separated by Duncan's multiple-range test. Standard statistical software (SAS Institute, 1982 version 6.04) was used for analysis. For determination of the 50% lethal concentration (LC50) and 50% survival time (ST50), T. ni neonate larvae were infected by allowing them to feed on a virus-contaminated diet for 24 h before transfer to an uncontaminated diet. Each virus, in doses of 500, 2,000, 8,000, 16,000, 32,000, 70,000, 150,000, and 300,000 PIBs per ml of diet, was fed to groups of 60 insects. Treatments and doses in ST50 and LC50 experiments were performed by a blind testing method. Insects were examined at 8-h intervals for 7 days p.i. for mortality. Data were subjected to probit analysis with standard software (5). LC50s and ST5Os were considered significantly different if their 95% fiducial limits did not overlap.

RESULTS Virus constructs. Three recombinant viruses which express jhe under the control of the XIV promoter (31), a modified polyhedrin promoter with enhanced expression properties (22, 26), were constructed: vJHEwt, vJHEEGTZ, and vJHEEGTD. The recombinant virus vJHEwt contains jhe in a wild-type (egt+) viral background and is occlusion negative. vJHEEGTZ is an occlusion-negative virus and has an E. coli lacZ insertion in frame with the viral egt gene, while in vJHEEGTD, the egt ORF has a 1.1-kb internal deletion. The latter recombinant is also occlusion positive because of the presence of the intact polyhedrin gene. The control viral construct for vJHEEGTZ is vEGTZ. This virus contains the same lacZ-egt fusion but lacks jhe (24). The virus that served as a control for vJHEEGTD, vEGTDEL (23), has a deletion in egt, is occlusion positive, and lacks jhe. JHE synthesis. To monitor synthesis of JHE in SF21 cells, cells were infected with vJHEEGTD or vEGTDEL (control) and radiolabeled with [35S]cysteine at various times p.i. Several protein bands, ranging in size from 60 to 66 kDa, were observed only in cell lysates from vJHEEGTD-infected cells (Fig. 2A). These proteins were expressed from 24 through 48 h p.i., as would be expected for products of a gene under the control of the very late XIV promoter. A single protein band of approximately 62 kDa could be seen in lysates of control-infected cells, but this protein did not follow the temporal expression expected for a product of a gene placed under control of a polyhedrin-related promoter. Proteins secreted from infected cells that were pulselabeled with [35S]cysteine were analyzed by SDS-PAGE. Two closely migrating protein bands at approximately 64 and 65 kDa were observed in the fluid collected from cells infected with vJHEEGTD (Fig. 2B). These proteins were not seen in control (vEGTDEL-infected) cells. Secretion of the expressed protein is to be expected, since the JHE cDNA contains 19 amino acids at the NH2 terminus which match well with consensus sequences for export signal peptides (14). The presence of the protein in the extracellular medium

1586


ELDRIDGE ET AL.

A

B

1HEE4 TD 16 12 24 36 48116 12 24 36 481

vJHEEGTD 16 12 24 36

481

6

vEGTDEL 122436481 -116 - 97

66

JHE[

-

45

-

29

40 AML

..

&,A&

FIG. 2. Autoradiograph of proteins synthesized in SF21 cells infected with vJHEEGTD or vEGTDEL. (A) Cells were infected with the indicated viruses and then pulse-labeled with [35S]cysteine at 6, 12, 24, 36, and 48 h pi. The cells were lysed and harvested, and 5-pJl aliquots were fractionated by SDS-PAGE and examined by fluorography. (B) The procedure was as described for panel A, except that after being pulse-labeled, the cells were incubated in 500 ,ul of PBS for an additional 1 h, the supernatant was harvested, and 25 ,ul was fractionated by SDS-PAGE and visualized by fluorography. Size standards are indicated in kilodaltons, and JHE is indicated by the bracket.

was first observed at 12 h p.i. Synthesis of secreted protein was maximal at 24 h p.i. and decreased at 36 and 48 h p.i. The apparent molecular mass of the expressed proteins corresponds closely to the 61-kDa size predicted for the mature JHE protein on the basis of the nucleotide sequence of the cloned cDNA fragment (14).

Immunoblotting. To determine whether these proteins were related to authentic JHE, we used an antiserum specific for H. virescens JHE (14) to probe lysates and supernatants of cells infected with vJHEwt and vJHEEGTZ. Lysates and supernatants of mock-infected cells or cells infected with wt virus or vEGTZ served as controls. A selective crossreaction was found in lysates of those cells infected with vJHEwt and vJHEEGTZ, while bands of similar size were not found in lysates from those cells that were mock infected or infected with the control viruses (Fig. 3). A slight crossreaction was observed with the abundant polyhedrin protein at approximately 30 kDa in the control virus-infected lane; this is likely to be a nonspecific reaction due to the large amount of protein present. The patterns and sizes of the immunopositive proteins found in lysates from the JHE

recombinant virus-infected cells are identical to that observed in lysates of the pulse-labeled proteins from cells infected with vJHEEGTD (compare Fig. 2 with Fig. 3). Similarly, two protein bands of approximately 64 and 65 kDa were recognized by the antisera in supernatants of cells infected with the JHE-expressing viruses. Proteins of similar sizes were not recognized in supernatants from cells that were mock infected or infected with the two control viruses. Furthermore, no bands were detected on the blot probed with the nonimmune serum (data not shown). These data show that the multiple bands observed in Fig. 2 are antigenically related to authentic JHE. To determine whether glycosylation accounts for the heterogeneity of the gene products, we compared gene expression in the vJHEwt- and wt-infected cells in the presence or absence of tunicamycin from 12 to 36 h p.i. (Fig. 3, panel on right). Western blot analysis using antisera directed against JHE showed a pattern of immunopositive bands similar to that observed previously in lysates and supernatants of vJHEwt-infected cells which were not treated with tunicamycin (Fig. 3, lanes -). In those recom-


VOL. 58, 1992

Lysate

r IIn 0 iw 1d-

.

o6(

1587

vJHEwt

Supematant

m

I I

4D + + E 09 a LC UL 6=: (

Tunicamycin

971

66-

-

] JHE

45-

29-

FIG.

3.

SF21 cells

Western blot were

analysis

mock infected

or

of

proteins

from

jhe-expressing

and control virus-infected cells and effects of

infected with either vJHEwt, vJHEEGTZ, vEGTZ,

or

tunicamycin

wt virus. After incubation for 12

on

expression.

h, the serum-free

lysates were harvested, and 25 and 5 p1l, separated by SDS-PAGE, electroblotted onto a filter, and probed with JHE-specific antisera. Some cells infected with vJHEwt (indicated by the vJHEwt bracket) were treated as described above except that at 12 h p.i., the media were replaced with fresh media which either did (+) or did not (-) contain 5 mg of tunicamycin per ml. Cell lysates and supernatants (Super.) were blotted and probed media

were

respectively,

as

replaced of each

with fresh media and incubated for another 24 h. The media and cell

were

described above. Size standards

are

indicated in kilodaltons.

binant cells treated with tunicamycin (lanes +), the apparent size of the immunopositive bands decreased to 60 kDa, the size of a single band which approximates the size predicted for unglycosylated JHE on the basis of the cDNA sequence. This corresponds to the smallest JHE-specific band observed in infected lysates (Fig. 3, left panel). Tunicamycin is a strong inhibitor of N-linked glycosylation (28), and the results are consistent with the interpretation that the larger proteins from untreated cells are glycosylated forms of JHE. The data also suggest that glycosylation is important for secretion of the recombinant JHE from cells in vitro. Supernatants from infected, untreated cells showed much more immunoreactive protein than did supernatants from treated, infected cells, even though equivalent volumes of supernatant were loaded onto the gel (rightmost two lanes of Fig. 3). This suggests that little of the nonglycosylated protein is released from the cells, a conclusion supported by the observation of a very strong immunopositive band in lysates from those infected cells treated with tunicamycin. Previous studies have also shown that glycosylation is important for proper secretion of some proteins expressed in baculovirusinfected cells (17). In vivo activity assays. To determine whether vJHEEGTD expresses active JHE in vivo, T. ni larvae approximately 24

h into the fourth instar were injected with 5 x 104 PFU of budded vJHEEGTD, vEGTDEL, or wt virus. At 24-h intervals p.i., different groups of larvae were bled and the JHE activity levels in the hemolymph were determined. The results are shown in Table 1. As early as 24 h p.i., a higher level of JHE activity was detected in those insects infected with the jhe-expressing recombinant virus than in insects that were mock infected or infected with the wt or vEGTDEL control viruses. By 48 h p.i., there was a massive

TABLE 1. JHE activity in hemolymph of fourth-instar T. ni larvae infected with control or recombinant virus JHE activity (meana + SEM) (nM/min/ml)

at indicated time p.i.

Virus

None (mock

infection) wt

vEGTDEL vJHEEGTD

24 h

48 h

72 h

7.5 ± 5.7

43.8 ± 8.1

68.7 ± 21.6

7.9 ± 5.5 1.2 ± 0.7 39.1 ± 11.0

1.3 ± 0.0 1.1 ± 0.2

1.1 ± 0.6 3.2 ± 1.4 1,210 ± 500

a Mean of three replicates.

3,040 - 1,320

1588


ELDRIDGE ET AL. 40 0o

B 30 )o E Z 20 )o

E

C0

02)

// / sL

t

/began

a)

10

)o

o 3

1 2 Time pi (days)

1 2 Time pi (days)

3

FIG. 4. Effect of infection with wt anId recombinant viruses on weight of fourth-instar (A) or fifth-insta r (B) T. ni. Larvae were injected with the budded form of each viruis (5 x 104 PFU per insect) and weighed at 24-h intervals p.i. Each pc)int represents the average weight of 25 to 30 insects. Vertical lines i ndicate the standard error mock infe of the mean. Symbols: ,

....,

vEGTDEL infection;

--

-, vJHEEGTD infection,

increase in JHE activity, to an ave-rage of approximately 3,000 nM/min/ml, in those insects infc-cted with vJHEEGTD. Levels seen in control virus-infectedI insects remained negligible. The mean JHE activity levels of mock-infected insects, which by 48 h were in the fifth instar, climbed to approximately 40 nM/min/ml. At 72 1h p.i., the activity level in insects infected with the JHE reco: mbinant virus appeared to be lower than that seen at 48 h. However, it was still approximately 20-fold higher than th( e level in mock infected insects at this stage, which had clinnbed to 70 nM/min/ml. This level of JHE activity in mocd k-infected insects is in agreement with previous measuremeints of JHE activity in T. ni (11, 15). Low levels were still obsserved in control virusinfected insects. The consistently lo' w JHE levels observed in vEGTDEL and wt-infected larvae are apparently due to the adverse effects of virus infectiion. Most vEGTDELinfected insects died while molting to the fifth instar, and wt-infected insects did not molt; hen ce, JHE activity in wtor vEGTDEL-infected insects nev er increased to levels found in normal fifth-instar larvae. B' y 4 days p.i., all insects were dead except those that were mlock infected. Analysis of larval weight after infection by injection. Weights of fourth- and fifth-instar T. ni larvae infected with vJHEEGTD, vEGTDEL, or wt virus were monitored at 24-h intervals p.i. to determine whether J[HE expression confers any advantages with regard to the uise of baculoviruses as pest control agents. These stages of laarval development were chosen because fifth-instar larvae, because of preparation for metamorphosis, exhibit an endlocrine milieu different from that of larvae of previous instarsa, while the fourth instar is more representative of the previouw s developmental stages. Larvae 24 h into their respective inst;ars were injected with 5 x 104 pfu of budded virus and examiried at 24-h intervals p.i. Injection of virus allowed us to adnninister a defined dose over a relatively short period, thus synchronizing the infection at a specific time in larval dev4 elopment. The average weights of fourth-instar-infected larvae are shown in Fig. 4A. No significant weight difference tbetween larvae infected with vJHEEGTD and those infecte(d with vEGTDEL was observed at any time point p.i. (P:> 0.05). Infection with either virus resulted in larvae that g9 ained approximately 30 mg through the first 2 days p.i.; afteir this time, the insects' weights either decreased or did nc)t change as the viral -

infection reached its late stages. Furthermore, there was no difference in time of mortality or in developmental phenotype between insects infected with vJHEEGTD and insects infected with vEGTDEL during the course of the study (data not shown). By day 2 p.i., insects infected with either virus to die, while most individuals died during day 3 p.i.; only a few survived to day 4 p.i. The presence of active ecdysone in insects infected with either vJHEEGTD or vEGTDEL was apparent because infected larvae died in the process of molting or immediately before molting (as determined by head capsule slippage). The average weight of insects infected with either recombinant virus, however, was significantly lower by day 1 p.i. than that of mock-infected or wt-infected insects. Mortality was observed 1 day earlier for vEGTDEL- and vJHEEGTD-infected larvae than for wtinfected larvae. Previous studies with S. fiugiperda have reported similar observations of reduced weight and earlier mortality after injection with egt- viral derivatives (25). Unlike all of the egt-deleted recombinant virus-infected insects, insects infected with wt virus never showed head capsule slippage before death, as would be expected for insects in which the ecdysone had been inactivated by virus-encoded EGT. This developmental arrest due to EGT has been demonstrated previously with other insects infected with wt AcMNPV (24). By day 3 p.i., mock-infected insects had completed their molt into the fifth instar and quickly gained weight, whereas the weight gains of wtinfected insects slowed drastically as the larvae succumbed to infection. In studies of fifth-instar T. ni larvae, little effect on weight gain was observed for insects infected with vJHEEGTD in comparison with insects infected with the vEGTDEL control virus (Fig. 4B). Larvae infected with either virus gained weight through the first 24 h and then lost weight starting on day 2 p.i. This weight loss parallels the weight loss of mock-infected larvae, indicating the cessation of feeding due to preparation for the larva-pupa molt. Note that the wtinfected insects continued to gain weight through day 2, as would be expected for larvae in which the ecdysone had been inactivated by EGT. Only by day 2 p.i. were the weights of the larvae infected with the JHE recombinant significantly lower than those of the larvae infected with the egt-deleted control virus. This difference, however, was small. As with the infected fourth-instar larvae, there was also no difference in time of mortality or in developmental phenotype (data not shown). The weights of insects infected with either vJHEEGTD or vEGTDEL were significantly lower than those of insects infected with wt virus from day 2 until day 4, when all insects had died (P < 0.05). The effects of the egt deletion were most striking on days 1 and 2 for the wt-infected fifth-instar larvae. While mock-infected insects had ceased feeding and started to spin cocoons by day 2 p.i., the wt-infected insects showed neither a reduction in feeding nor any spinning behavior. Instead, they continued to feed until they became moribund from viral infection. The times to 100% mortality of the test populations were similar; all of the insects were dead by day 4 p.i. in all three virus-infected populations. Analysis of larval weight following per os infections. To complement the injection studies described above, we infected fourth- and fifth-instar T. ni larvae with virus per os. The results of the fourth-instar per os infections were very similar to those described above for the injection studies, although the weight at the time of infection was lower because of the preinfection 18-h starvation period (compare Fig. 5A with Fig. 4A). No significant difference in average


VOL. 58, 1992

1501

3001

1589

TABLE 2. Dose-mortality response of neonate T. ni infected with wt or recombinant virus 95% fiducial limit

100

Virus

E

.a

wt vEGTDEL vJHEEGTD

50

1

2 Time pi (days)

2 3 Time pi (days)

FIG. 5. Effect of per os infection on weight of fourth-instar (A) or fifth-instar (B) T. ni. After an 18-h period without food, larvae were infected by feeding on a 1-,ul (fourth instar) or 2-,ul (fifth instar) droplet of sucrose solution containing 5 x 103 PIBs. Larvae were weighed at 24-h intervals p.i. Each point represents the average of 30 insects. Vertical lines indicate the standard error of the mean. Symbols are as in Fig. 4.

(PIB/ml)

(

(PIB/ml) 8,320 8,510 8,910

Slope

Upper

Lower

10,000 10,300 10,900

6,760 7,010 7,160

1.69 1.80 1.57

a Sixty neonates per dose were fed a contaminated diet for 24 h and then transferred to an uncontaminated diet.

viruses were significantly shorter than that observed for wt-infected insects by approximately 20 h. A similar reduction in median survival time was observed previously in time-mortality studies on S. frugiperda infected with wt AcMNPV or egt- AcMNPV derivatives (25).

DISCUSSION weight between insects infected with vJHEEGTD and vEGTDEL was observed at any of the times sampled. There were also no major differences detected in developmental phenotype or in time of mortality (data not shown). As with the injection bioassays, most of the insects infected with the two egt-deleted viruses died in the process of molting to the fifth instar. While the weight curves for fourth-instar larvae fed vEGTDEL or vJHEEGTD were essentially identical, those for fifth-instar larvae showed some differences (Fig. SB). vEGTDEL- and vJHEEGTD-infected insects showed no significant differences in weight at either 1 or 2 days p.i., and they did not show significant differences in weight from insects which were infected with wt virus (P > 0.05). Irrespective of the treatment, insects gained approximately five times their initial weight by day 2 p.i. At 3 and 4 days p.i., however, vJHEEGTD-infected insects weighed a small but statistically significant amount less than those larvae infected with vEGTDEL (P < 0.05). However, this disparity in weight may be due to differences in rate of weight loss during preparation for pupation. The differences are not great enough for vJHEEGTD to be considered an improved viral pesticide. The average weights of the mock-infected insects did not differ significantly from the average weights of either vEGTDEL- or vJHEEGTD-infected insects on day 3 p.i., and all mock-infected larvae had pupated by day 4 p.i. Even though the average weights of larvae infected with vJHEEGTD were significantly lower than the average weights of larvae infected with vEGTDEL, no developmental differences were detected. Insects infected with either virus started to spin cocoons at similar times and in similar numbers (data not shown). Dose- and time-mortality bioassay. Even though there was little evidence of an effect of JHE in feeding or development of late-instar T. ni larvae, the possibility of more pronounced effects in neonates existed. To test this, we determined the LC50 and ST50 for a jhe-expressing virus in neonate T. ni (Tables 2 and 3). No significant differences in LC50 between neonates allowed to feed on a diet contaminated with PIBs of vJHEEGTD, vEGTDEL, or wt virus were observed (Table 2). Furthermore, there was no significant difference in the ST50s of neonate larvae fed an LCioo dose of PIBs of vJHEEGTD or vEGTDEL (Table 3). However, the median times of mortality for insects infected with both egt-deleted

The possibility of using jhe expression as a means of improving the efficacy of baculoviruses as biopesticides was considered previously, but few or no differences were observed in either developmental characteristics or weight gain (11). However, those studies did not use vectors which were specifically defective in egt expression. On the basis of known hormonal regulation in insects, the presence of EGT and its subsequent inactivation of the molting hormone ecdysone (24) would be expected to circumvent the disruptive effects of premature clearing of JH by overexpressed JHE. Our experiments addressed this possibility directly by comparing jhe expression in an egt- genetic background. Although the ST50 of vJHEEGTD-infected neonates was reduced approximately 20 h compared with that of wt virus-infected neonates, the same reduction was observed with vEGTDEL-infected neonates, and thus it is due solely to the egt deletion. A reduction in ST50 for egt-disrupted virus in S. fiugiperda has been described elsewhere (25). In addition, no differences in developmental characteristics between neonates or late-instar larvae infected with vJHEEGTD and those infected with vEGTDEL were observed. Active ecdysone was present in recombinant virus-infected insects, as demonstrated by head capsule slippage in infected fourth-instar larvae and prepupation behavior in infected fifth-instar larvae. However, even with active ecdysone and high titers of active recombinant JHE, neither instar infected with vJHEEGTD showed any signs of premature molting. Since ester hydrolysis appears to be a major route of JH metabolism in insects (10), we thought that viruses expressing the H. virescens jhe cDNA would adversely affect lepidopteran species such as T. ni. However, our studies of TABLE 3. Time-mortality response of neonate T. ni infected with wt or recombinant virus Virus

STS(,a (h)

wt vEGTDEL vJHEEGTD

102 82 78

95% fiducial limit (h) Lower Upper

106 86 82

98 75 73

Slope

15.4 10.7 9.9

a Sixty neonates per dose were fed a contaminated diet for 24 h and then transferred to an uncontaminated diet

1590


ELDRIDGE ET AL.

both T. ni and fourth- and fifth-instar S. frugiperda showed no such effect (data not shown). We also tested fourth-instar H. virescens, and no difference between vJHEEGTZ- and vEGTZ-infected larvae in weight, development, or time of mortality was observed (data not shown). Although our study indicated no difference in weight gain between T. ni larvae infected with control andjhe-expressing virus, Hammock et al. (11) reported reduced weight gain in neonates infected per os with the budded form of a jheexpressing virus. We did not compare weights of the neonates, but no overt difference in size was observed before the death of the insects. The effect of the recombinant virus that Hammock et al. reported might be due to the method of infection, since the effects of feeding of budded virus are not well characterized. The reason for the lack of increased efficacy of the egtdeleted, jhe-expressing recombinant virus is unknown. Although previous studies (11) suggested that JHE is unstable in vivo, we have observed that JHE can accumulate to levels in recombinant virus-infected insects that are 20- to 40-fold higher than those found naturally. Thus, JHE levels should be high enough to exert a significant effect. However, insects may have a strong homeostatic response to inappropriate levels of JHE and may compensate by synthesizing and/or releasing more JH. A recombinant virus that expresses both jhe and an allatostatin to down-regulate JH synthesis by the corpora allata (6, 29) may overcome this problem. The lack of JHE effect suggests that our understanding of JHE function is still incomplete or that AcMNPV alters the hormonal milieu in ways additional to those previously defined. Further research in these areas is necessary before this gene can become useful in improving baculovirus pesticide efficacy. ACKNOWLEDGMENTS We thank Karen Liljebjelke for western blot analysis and Steve Hilliard for cell culture and bioassay assistance. We also thank Judith Willis and Arden Lea for helpful comments on the manuscript. This work was supported in part by Public Health Service grant Al 23719 from the National Institute of Allergy and Infectious Diseases to L.K.M. and with partial support from NSF grant DCB-904753 and USDA grant 88-37234-4006 to B.D.H.

8.

9.

10.

11. 12. 13.

14.

15. 16.

17. 18.

19. 20. 21.

1.

2.

3.

4. 5.

6.

7.

REFERENCES Blake, M. S., K. H. Johnson, G. J. Russell-Jones, and E. C. Gotschlich. 1984. A rapid, sensitive method for detection of alkaline phosphatase-conjugated antibody on western blots. Anal. Biochem. 136:175-179. Bonner, W. M., and R. A. Laskey. 1974. A film detection method for tritium labeled proteins and nucleic acids in polyacrylamide gels. Eur. J. Biochem. 46:83-88. Burnette, W. N. 1981. "Western blotting": electrophoretic transfer of proteins from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A. Anal. Biochem. 112:195-203. Chen, E. Y., and P. H. Seeburg. 1985. Supercoil sequencing: a fast and simple method for sequencing plasmid DNA. DNA 4:165-170. Daum, R. J. 1970. A revision of two computer programs for probit analysis. Bull. Entomol. Soc. Am. 16:10-15. Feyereisen, R. 1985. Regulation of juvenile hormone: synthesis, p. 391-430. In G. A. Kerkut and L. I. Gilbert (ed.), Comprehensive insect physiology, biochemistry and pharmacology, vol. 7. Pergamon Press, Inc., New York. Gilbert, L. I., and D. S. King. 1973. Physiology of growth and development: endocrine aspects, p. 299-370. In M. Rockstein

22. 23.

24.

25. 26.

(ed.), The physiology of insects, vol. 1. Academic Press, Inc., New York. Groner, A. 1986. Specificity and safety of baculoviruses, p. 177-202. In R. R. Granados and B. A. Federici (ed.), The biology of baculoviruses, vol. 1. CRC Press, Inc., Boca Raton, Fla. Guy, R. H., N. C. Leppla, J. R. Rye, C. W. Green, S. L. Barrett, and K. A. Hollien. 1985. Trichoplusia ni, p. 487-494. In P. Singh and R. F. Moore (ed.), Handbook of insect rearing, vol. 2. Elsevier, New York. Hammock, B. D. 1985. Regulation of juvenile hormone titer: degradation, p. 431-472. In G. A. Kerkut and L. I. Gilbert (ed.), Comprehensive insect physiology, biochemistry and pharmacology, vol. 7. Pergamon Press, Inc., New York. Hammock, B. D., B. Bonning, R. D. Possee, T. N. Hanzlik, and S. Maeda. 1990. Expression and effects of the juvenile hormone esterase in a baculovirus vector. Nature (London) 344:458-461. Hammock, B. D., and R. M. Roe. 1985. Analysis of juvenile hormone esterase activity. Methods Enzymol. 111:487-493. Hammock, B. D., V. Wroblewski, L. Harshman, T. Hanzlik, S. Maeda, M. Philpott, B. Bonning, and R. Possee. 1990. Cloning, expression and biological activity of the juvenile hormone esterase from Heliothis virescens, p. 49-56. In H. H. Hagedom, J. G. Hildebrand, M. G. Kidwell, and J. H. Law (ed.), Molecular insect science. Plenum Press, New York. Hanzlik, T. N., Y. A. I. Abdel-Aal, L. G. Harshman, and B. D. Hammock. 1989. Isolation and sequencing of cDNA clones coding for juvenile hormone esterase from Heliothis virescens. J. Biol. Chem. 264:12419-12425. Hanzlik, T. N., and B. D. Hammock. 1988. Characterization of juvenile hormone hydrolysis in early larval development of Tnchoplusia ni. Arch. Insect Biochem. Physiol. 9:135-156. Huber, J. 1986. Use of baculoviruses in pest management programs, p. 182-202. In R. R. Granados and B. A. Federici (ed.), The biology of baculoviruses, vol. 2. CRC Press, Inc., Boca Raton, Fla. Jarvis, D. L., and M. D. Summers. 1989. Glycosylation and secretion of human tissue plasminogen activator in recombinant baculovirus-infected insect cells. Mol. Cell. Biol. 9:214-223. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680-685. Lee, H. H., and L. K. Miller. 1978. Isolation of genotypic variants of Autographa califomica nuclear polyhedrosis virus. J. Virol. 27:754-767. Maniatis, T., E. F. Fritsch, and J. Sambroolk 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Miller, D. W., P. Safer, and L. K. Miller. 1986. An insect baculovirus host-vector system for high level expression of foreign genes, p. 277-298. In J. K. Setlow and A. Hollander (ed.), Genetic engineering. Plenum Press, New York. Ooi, B. G., C. Rankin, and L. K. Miller. 1989. Downstream sequences augment transcription from the essential initiation site of a baculovirus polyhedrin gene. J. Mol. Biol. 210:721-736. O'Reilly, D. R., 0. W. Howarth, H. W. Rees, and L. K. Miller. 1992. Structure of the ecdysone glucoside formed by a baculovirus ecdysteroid UDP-glucosyltransferase. Insect Biochem. 21:795-801. O'Reilly, D. R., and L. K. Miller. 1989. A baculovirus blocks insect molting by producing ecdysteroid UDP-glucosyl transferase. Science 245:1110-1112. O'Reilly, D. R., and L. K. Miller. 1992. Improvement of a baculovirus pesticide by deletion of the egt gene. Bio/Technology 9:1086-1089. Rankin, C., B. G. Ooi, and L. K. Miller. 1988. Eight base pairs encompassing the transcriptional start point are the major determinant for baculovirus polyhedrin gene expression. Gene

70:39-49. 27. Riddiford, L. M., and S. R. Palli. 1990. Molecular aspects of juvenile hormone reception and action, p. 77-81. In H. H. Hagedorn, J. G. Hildebrand, M. G. Kidwell, and J. H. Law (ed.), Molecular insect science. Plenum Press, New York.

VOL. 58, 1992


28. Takatsuki, A., and K. Kohno. 1975. Inhibition of biosynthesis of polyisoprenol sugars in chick embryo microsomes by tunicamycin. Agric. Biol. Chem. 39:2089-2091. 29. Tobe, S. S., and B. Stay. 1985. Structure and regulation of the corpus allatum. Adv. Insect Physiol. 18:305-342. 30. Vaughn, J. L., R. H. Goodwin, G. L. Thompkins, and P. McCawley. 1977. The establishment of two cell lines from the insect Spodoptera frugiperda (Lepidoptera: Noctuidae). In

1591

Vitro (Rockville) 13:213-217. 31. Wang, W., B. G. Ooi, and L. K. Miller. 1991. Baculovirus vectors for multiple gene expression and for occluded virus production. Gene 100:131-137. 32. Willis, J. H. 1990. Regulating genes for metamorphosis: concepts and results, p. 91-98. In H. H. Hagedorn, J. G. Hildebrand, M. G. Kidwell, and J. H. Law (ed.), Molecular insect science. Plenum Press, New York.