VIROLOGY i 90, 815-823 (19%‘)

A Cellular Promoter-Based Expression Cassette for Generating Recombinant Baculoviruses Directing Rapid Expression of Passenger Genes in Infected Insects

RUSSELL JOHNSON, ROY G. MEIDINGER, AND KOSTAS IATROU’

Department of Medical Biochemistry, Faculty of Medicine, The University of Calgary, 3330 Hospital Drive NW, Calgary, Alberta, Canada T2N 4Nl

Received May 20, 1992; accepted June 30, 1992

We have developed an expression cassette which allows the generation of recombinant baculoviruses that can express passenger genes under the control of a constitutive cellular promoter derived from the cytoplasmic actin gene of the silkmoth Bombyx mori. Silkmoth tissue culture cells which were infected with a recombinant B. mori nuclear polyhedrosis virus (BmNPV) containing the gene-encoding chloramphenicol acetyl transferase (CAT) under the control of this expression cassette expressed significant CAT activity beginning 5 hr postinfection (p.i.). Cells infected with a recombinant BmNPV containing the cat gene under the control of the polyhedrin gene promoter did not express CAT activity until 20 hr p.i. Silkworm larvae were also infected with the two recombinant viruses by hemocelic injections and all larval tissues examined were found to express the cat gene. While significant actin-cassette-driven CAT expression in vivo was first seen at 24 hr p.i., expression from the polyhedrin promoter was not seen until 48 hr p.i. By 60 hr p.i., tissues of larvae infected with the recombinant virus expressing cat under polyhedrin promoter control were found to exhibit sixfold higher CAT activity than those infected with recombinant virus expressing the cat gene under the control of the actin promoter. The 24-hr temporal advantage in expression of a passenger gene in infected larvae indicates that the actin-promoter-based expression cassette or other analogous cellular promoter-based cassettes could be used for generating recombinant baculovirus insecticides which could incapacitate pest insects more quickly than viruses employing the polyhedrin or other late viral promoters for expressing insect-incapacitating proteins. o ISSZ Academic

Press, Inc

INTRODUCTION

Nuclear polyhedrosis viruses (NPVs), a subgroup of the family Baculoviridae whose virions are embedded into proteinaceous polyhedra in the nucleus of host cells, infect a wide variety of insect species (Blissard and Rohrman, 1990). NPV-based gene expression systems, in which a nonessential viral gene is replaced by a particular gene of interest, have proved to be ex- tremely useful for high-level production of properly pro- cessed eukaryotic proteins in infected insect cells or larvae (Miller, 1988, 1989; Maeda, 1989; Summers, 1992).

NPVs also have much potential utility for biological control of economically important pest insects. Be- cause most NPVs have a host range restricted to only a few closely related species, they can be used without disrupting the balance of other insect and noninsect species (e.g., important predators) in the agricultural ecosystem. To date, however, baculoviruses have met with only limited commercial success as control agents, due to difficulties with virus stability and, most

’ To whom reprint requests should be addressed.

importantly, slower speed of action than that achieved with chemical insecticides (Huber, 1986; Wood and Granados, 199 1).

In order to increase the speed of action of baculovi- ruses, recombinant viruses have been generated which express proteins intended to be toxic to the in- fected insect. The underlying premise for the creation of such recombinants has been that the expressed for- eign protein should be able to incapacitate the infected larvae before they would normally succumb to viral in- fection (Wood and Granados, 1991; latrou, 1992). This strategy has met with varying degrees of success (Car- bonell et al., 1988; Maeda, 1989; Merryweather et al., 1990) and some recombinant viruses have been devel- oped recently which significantly increased the speed of host incapacitation (Stewart et a/., 1991; Tomalski and Miller, 1991; Maeda et a/., 1991; McCutchen et al., 1991).

The recombinant viruses listed above employed the polyhedrin promoter, the pl0 promoter, or a synthetic promoter based on the previous two (Wang et a/., 1991) to express the desired foreign protein. While these promoters can direct expression of high levels of protein, they are not expressed until the very late stages of infection. Furthermore, there is some doubt

815 0042.6822/92 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction r any form reserved.

816 JOHNSON, MEIDINGER, AND IATROU

as to their level of expression in the epithelial cells of the gut (latrou, 1992) which represent the primary site of infection.

It should be possible to achieve more rapid control of insect larvae using recombinant baculoviruses which employ strong constitutive promoters, so that expres- sion of the foreign protein can begin immediately upon infection. Promoters from immediate-early viral genes could, in principle, serve this purpose, but since these genes are essential for viral function and may not be deleted, a recombinant virus utilizing such a promoter would necessarily contain a duplication which could prove to be unstable upon multiple passages. Further- more, while early viral promoters are active during the initial phase of infection, most become inactive after viral DNA replication (Kovacs et a/., 1991, and refer- ences therein). To circumvent such difficulties, we have set out to investigate the possible advantages that a promoter element derived from a constitutively and ubiquitously expressed insect gene may offer over viral promoters with respect to speed of transgene ex- pression. Our decision to utilize a cellular promoter ele- ment as a driver of transgene expression in recombi- nant baculoviruses has been based on the demonstra- tion that silkworm chromosomal genes introduced into a baculovirus genome can be expressed correctly in insect cells in vivo under the control of their own pro- moter following infection of larvae with the corre- sponding recombinant baculovirus (latrou and Mei- dinger, 1990).

We now report on the utilization of the promoter from the cytoplasmic actin gene of Bombyx mori, which is known to be active in a wide variety of tissues (Mounier and Prudhomme, 1986, 1991) to construct an expres- sion cassette which, when incorporated into a viral ge- nome, should be able to direct an immediate expres- sion of genes under its control in all infected tissues. A recombinant virus expressing CAT protein as a re- porter was used to analyze the timing of expression in infected Bm5 cells and in various tissues of infected larvae relative to expression directed by the polyhedrin promoter. Our results demonstrate that this actin-pro- moter-based expression cassette can be used to gen- erate recombinant baculoviruses which express pas- senger genes more rapidly than currently available re- combinant baculovirus systems which utilize late viral promoters.

MATERIALS AND METHODS

Plasmid constructions

The cytoplasmic actin expression cassette, termed pBmA (Fig. 1 A), is a pBluescript (Stratagene) derivative of clone pA3-5500 (Mounier and Proudhomme, 1986)

which contains the A3 cytoplasmic actin gene of B. mob. This cassette contains 1.5 kb of the A3 gene Y-flanking sequences, part of its first exon to position +67 (relative to transcription initiation), a polylinker re- gion derived from plasmid pBluescript for insertion of foreign gene sequences, and an additional 1.05 kb of the A3 gene sequences encompassing part of the third exon of the gene from position +836 and adjacent 3’- flanking sequences which contain signals required for polyadenylation. The ATG translation initiation codon of the actin gene, originally present at positions +36 to +38, has been inactivated by mutation into AGG (Kun- kel, 1985) therefore allowing translation initiation from any gene inserted into the cassette polylinker from the first ATG triplet of the insert.

A 900-bp Xholl fragment containing the cat open reading frame (ORF) was excised from pCARCAT-1 (Mitsialis et a/., 1987) and inserted into the BamHl site of pBmA to create pBmA.cat (Fig. 1 B). To produce an appropriate viral transfer vector, plasmid pBmp2 (la- trou and Meidinger, 1989), which contains a copy of the BmNPV polyhedrin gene whose promoter and part of the coding sequence have been deleted, was di- gested with Xbal and this restriction site was con- verted to an Sstl site by addition of linkers and religa- tion. The resulting plasmid was designated pBmp2s (Fig. 1 C). The Sstl fragment of pBmA.cat containing the A3-cat chimeric gene was then inserted into the Sstl site of pBmp2s to generate pBmp2s/A.cat (Fig. 1 D). Plasmid pBmp26cat (Fig. 1 E) contains a polyhedrin gene whose translation initiation codon ATG has been mutatated to ATT and sequences from +27 to +146 (relative to translation initiation) removed and replaced by the cat gene ORF. The polyhedrin promoter se- quences are retained in this gene and direct the tran- scription of the cat reporter gene. The construction of this transfer vector will be described in detail else- where (Johnson et a/., in preparation).

Cell culture and viruses

Bm5 silkworm tissue culture cells were maintained in IPL-41 medium containing 10% fetal calf serum as previously described (latrou et al., 1985). Recombinant viruses were obtained by cotransfection of Bm5 cells with pBmp2s/A.cat or pBmp26cat and wild type BmNPV DNA. Cells were transfected in six-well micro- titer plates as described by latrou and Meidinger (1989) except that the transfection solution (500 ~1) consisted of 30 pg/ml Lipofectin (Bethesda Research Laborato- ries) in basal IPL-41 containing 5 pg/ml of the transfer vector and 0.2 pg/ml BmNPV DNA. After incubation for 5 hr, the transfection solution was replaced by 2 ml of complete medium containing 50 pg/ml of gentamycin.

BACULOVIRUS EXPRESSION CASSETTE 817

The medium, containing crude recombinant virus, was collected after 7 days. A purified recombinant virus was obtained by serial dilution as described previously (Pen er al., 1989; Goswami and Glazer, 1991) and its structure confirmed by Southern blotting.

Viral inoculum for all experiments except those shown in Fig. 2 was prepared by collecting the medium (supernatant) from cells 5 days after infection. The titer of extracellular virus (ECV) present in this inoculum was determined by infecting cells in 96-well microtiter plates with diluted virus and counting the number of wells showing infection. For the experiments shown in Fig. 2, ECV was pelleted from the medium by centrifu- gation at 50,000 rpm in a Beckman 100.2 rotor for 1 hr. After removal of the supernatant, the viral pellet was rinsed with 1 ml H,O, resuspended in 1.6 ml H,O, and recentrifuged as before. The final pellet was again rinsed, resuspended in H,O, and used to infect cells.

Infections

Bm5 cells to be infected with virus were seeded into 24-well microtiter plates at a density of 2 X 1 O5 cells (in 500 ,uI medium) per well. After 3 days, 100 ~1 of viral inoculum [l O6 plaque-forming units (PFU)] was added to each well. Time after infection was counted from the time that the virus was added to the medium. Cells were removed from the wells at the appropriate time by repeated pipeting.

Silkworm larvae were reared on a diet of fresh mul- berry leaves and were injected with virus at the begin- ning of the fifth instar. After incubating the animals at 0” for 1 hr, 10 ~1 of viral inoculum (1 O5 PFU) was in- jected into the hemocel using a 26-gauge needle. The infected animals were maintained as before and col- lected at the appropriate time for dissection.

CAT assays

Cells to be assayed for CAT activity were pelleted from the medium at 1000 g for 5 min, suspended in 1 ml PBS (10 mn/l KH,PO,, 2 m/W NaH,PO,, 140 mM NaCI, 40 mM KCI), repelleted, and stored at -70”. The cells were then resuspended in 200 ~1 of 0.25 MTris- HCI, pH 7.8, freeze-thawed three times using dry ice to disrupt the cells, and the supernatants retained for CAT assays.

Larvae to be assayed for CAT activity were dissected in cold PBS and the appropriate tissues were removed. Midguts were cut open longitudinally to allow removal of the gut contents and all tissues were rinsed exten- sively with several changes of PBS. The tissue samples were ground in 0.25 M Tris-HCI, pH 7.8, with a small pestle in a microcentrifuge tube, freeze-thawed, and centrifuged as described above. All larval extracts

were heated to 65” for 5 min to inactivate cellular de- acetylase activity before use in CAT assays. This treat- ment was not necessary for Bm5 cell extracts as they did not contain deacetylase activity. Assays for protein content and CAT activity of the extracts from cells and larvae were performed as previously described (latrou and Meidinger, 1989). One unit of CAT activity cata- lyzes the acetylation of 1 nmol of chloramphenicol per minute at 37”.

Hybridization

Cell suspensions were dot blotted onto Hybond N+ membrane (Amersham) and treated with 0.5 M NaOH followed by 0.5 MTris-HCI, pH 7.5. Labeling of linear- ized PBS/SK+ DNA with [(r-32P]dCTP and hybridization of this probe to the membrane (at 65”) was carried out as described by Fotaki and latrou (1988). After hybrid- ization, the membrane was washed at 65” in 0.1 X SSC, 0.1% SDS and exposed to X-ray film.

RESULTS

Development of vectors and recombinant viruses

Vectors were constructed that would allow genera- tion of recombinant BmNPV which could express any desired passenger gene under the control of the cyto- plasmic actin promoter of B. mori. Plasmid pBmA con- tains a multiple cloning site (MCS) downstream of the A3 gene 5’-untranslated sequences into which any coding sequence may be inserted (Fig. 1A). The plas- mid pBmp2s (Fig. 1 C) contains a 1 0-kb Pstl fragment of BmNPV DNA including a partially deleted polyhedrin gene. Part of the coding sequence and 5’ flanking se- quences have been excised from the polyhedrin gene, removing all promoter activity (latrou and Meidinger, 1989). Any chimeric gene constructed within pBmA may, thus, be excised with Sstl and inserted into the unique Sstl site in the deleted polyhedrin gene of pBmp2s to produce a transfer vector which may be used to generate a recombinant virus by homologous recombination with the BmNPV genome. The resulting recombinant can then be selected on the basis of its polyhedrin-minus phenotype.

In order to characterize the expression of a foreign protein from a passenger gene under the control of the actin promoter in infected cells, we chose to create a recombinant virus containing the cat reporter gene. Plasmid pBmA.cat (Fig. 1 B), which contained the cat ORF inserted into the MCS of pBmA, was used to gen- erate a transfer vector pBmp2s/A.cat (Fig. 1 D) as out- lined above. A recombinant virus, BmNPV/A.cat, was obtained using this transfer vector and purified as de- scribed under Materials and Methods. In order to com-

818 JOHNSON, MEIDINGER, AND IATROU

PstI

FIG. 1. Plasmid vectors used in the generation of recombinant viruses. (A) The 5’. and 3’-flanking sequences of the cytoplasmic actin gene are represented by the unfilled regions and transcribed sequences are indicated by the shaded region. The line represents sequences from pBS/SK+ including the beta-lactamase (bla) gene which confers ampicillin resistance. All of the restriction enzyme sites shown in the multiple cloning site are unique in this plasmid. (6) The vertically striped box represents the cat gene ORF. All other designations are as in A. (C) The horizontally striped box represents the partially deleted polyhedrin (poly) gene containing no promoter sequences. Black regions represent sequences flanking the polyhedrin gene in the BmNPV genome. The line represents pUC DNA. (D) The line represents DNA from pf3YSK-t. All other designations are as described in A and C. (E)The region containing horizontal lines represents the polyhedrin gene including all sequences necessary for strong promoter activity. All other designations are as described in C.

pare activity derived from the actin promoter to that obtained from the polyhedrin promoter, a recombinant virus containing the cat-gene-coding sequence in- serted downstream of the polyhedrin promoter was also used. This virus, BmNPV/P26.cat, was obtained using the transfer vector pBmp26cat (Fig. 1 E).

Presence of CAT activity in recombinant viruses

It was our intention to use these two recombinant viruses to infect cells and compare the time course of CAT expression from the two promoters. Before doing so, however, it was necessary to determine whether the infected cells would contain any background level of CAT activity as a result of the encapsulation of the extremely stable CAT protein within the recombinant extracellular virus (ECV) and its release within the cell immediately upon infection. Medium collected from in- fected cells 5 days after infection (to be used as inocu-

lum) was found to contain high levels of CAT activity (Fig. 2A, right). ECV isolated from this medium was lysed by sonication and was also found to contain a significant amount of CAT activity even after extensive washing (Fig. 2A, left). Ceils infected with either recom- binant virus were found to contain CAT activity 5 min after addition of the viral inoculum. The level of CAT activity in these cells reached a plateau within 20 min (Fig. 2B) and this level was maintained for several hours (not shown). When cells were infected with the standard viral inoculum, which contained both virus and medium, a small portion of this CAT activity could be removed by washing the cells with PBS, but the majority remained even after extensive washing. When cells were infected with isolated virus, none of the activ- ity could be removed by washing (Fig. 2C). It was con- cluded, therefore, that this remaining activity repre- sented CAT enzyme attached to or introduced into the cells during viral infection and that this background

BACULOVIRUS EXPRESSION CASSETTE 819

AC-C

VIRUS 1 MEDIUM

A.cat P26.cat

(50 pg protein9 hr) (10 pg protein-l hr)

C

-- ____ _ IM+VI V 1 I I

Ix 5x lx 5x lx 5x

> AC-C

FIG. 2. Analysis of CAT activity in ECV. (A) ECVfrom BmNPV/A.cat or BmNPV/P26,cat was separated from the medium as described under Materials and Methods. Twenty microliters of isolated ECV or of medium containing ECV (viral inoculum) was assayed for 1 hr as described. The upper spots (AC-C) represent acetyl chlorampheni- co/, the products of the CAT reaction, while the lower spot in each assay (C) represents the substrate, chloramphenicol. (B) Bm5 cells were assayed for CAT activity at the indicated times after addition of BmNPV/A.cat or BmNPV/P26.cat inoculum. Assays were performed using the amount of cell protein and length of incubation indicated. (C) Bm5 cells were infected with the standard viral inoculum which contains medium and virus (M + V), with ECV isolated from this inoculum (V), or mock infected (----). The cells were collected 1 hr after infection and were washed once (1 X) or five times (5X) with 1 ml PBS.

value must be subtracted from all subsequent mea- surements in order to quantify the amount of CAT en- zyme present due to promoter activity during time course experiments. As indicated in Fig. 2C, unin- fected Bm5 cells do not contain any detectable CAT activity nor do cells infected with wild type virus (data not shown).

Time course of gene expression in Bm5 cells

Bm5 cells were infected with recombinant virus BmNPV/A.cat or BmNPVIP26.cat and cell extracts

were assayed at various times after infection. CAT ac- tivity above the background level was first detected in cells infected with BmNPV/A.cat at 5 hr p.i. (Fig. 3A), while in cells infected with BmNPV/P26.cat, activity above background was not observed until the 20-hr time point (Fig. 3B). The cell extracts were then diluted appropriately to obtain quantitative CAT assays for all time points and the background values were sub- tracted from each point to generate the curves shown in Fig. 3C. At early times, from 5-20 hr p.i., the actin promoter was significantly more active than the poly- hedrin promoter, resulting in greater CAT activity. At 25-30 hr p.i. CAT activity from the two promoters was roughly equal, and after 30 hr p.i. expression from the polyhedrin promoter was higher than that from the ac-

0 2 5 8 12 20 30 40 50

P TIME p.i. (hours)

0 2 5 8 12 20 30 40 50

TIME p.i. (hours)

C

100 L--

10 20 30 40 50

TIME p.i. (hours)

FIG. 3. Time course of CAT activity in infected Bm5 cells. Cells were infected and assayed as described under Materials and Meth- ods. (A) Cells were infected with BmNPV/A.cat inoculum and CAT assays were performed using 5 pg cell protein incubated for 1 hr. (B) Cells were infected with BmNPVIP26.cat inoculum and assayed us- ing 1 pg cell protein incubated for 1 hr. (C) Appropriate amounts of cell extract were assayed to quantitate CAT assays at each time point. The background observed at l-2 hr p.i. was subtracted from each point and values were plotted on a logarithmic scale.

820 JOHNSON, MEIDINGER, AND IATROU

B HMSG BHMSG

A.cat P26.cat

FIG. 4. Analysis of CAT activity in infected larvae. Fifth instar larvae were injected with either BmNPV/A.cat or BmNPV/P26.cat inoculum and after 2 days were assayed for CAT activity in body wall (B), head (H), midgut (M), silk gland (S), and gonad (G). The assays were per- formed using 10 pg protein incubated for 1 hr. Tissues from three larvae were pooled for each assay.

tin promoter (by a factor of 3 at 50 hr p.i., the last point of the time course).

Gene expression in infected larvae

Fifth instar B. mob larvae were injected with recombi- nant virus BmNPV/A.cat or BmNPV/P26.cat and various tissues were initially assayed for CAT activity 2 days p.i. (Fig. 4). Larvae infected with either virus con- tained high levels of CAT activity in both the head (H) and body wall (B), a lower level in the midgut (M), and still lower expression in the gonads (G). Although de- tectable, only very low levels of CAT activity could be seen in the silk glands (S). Uninfected larvae, or larvae infected with wild typevirus, did not contain anydetect- able CAT activity (data not shown).

In a second set of experiments, larvae at the begin- ning of fifth instar were injected with the recombinant viruses, tissues were collected at different times p.i. and assayed for CAT activity. Tissues from the injected larvae contained a low level of background CAT activity at early times p.i. In larvae infected with BmNPV/A.cat, CAT activity above background was first observed in body wall tissues at 24 hr p.i. (Fig. 5A), while larvae infected with BmNPV/P26.cat did not show any CAT activity above background levels until 48 hr p.i. (Fig. 5B). A quantitative analysis of the results obtained from all time points following subtraction of the background values is shown in Fig. 5C. At all time points up to 48 hr p.i., the actin promoter was significantly more active than the polyhedrin promoter. At 60 hr p.i., the activity from the polyhedrin promoter was found to be higher than that from the actin promoter but the difference in expression levels at that time point was only sixfold. Similar results were obtained from the other tissues that were examined (not shown). In these experiments, activity from the actin promoter was seen 24 hr earlier

than from the polyhedrin promoter. Although the times at which CAT activity was first expressed from the two viruses in infected larvae was later than that observed in infected Bm5 cells, the pattern of expression in viva was similar to that observed in vitro.

Expression in other lepidopteran species

To determine the level of activity of the B. mori cyto- plasmic actin promoter in cells of other insect species, we transfected plasmid pBmA.cat into tissue culture cells of both B. mori (Bm5) and Spodoptera frugiperda (Sf21). Cells were collected 2 days after transfection and assayed both for CAT activity and for the amount of plasmid DNA present in the cells. The Bm5 cells contained about four times as much CAT activity as an equivalent volume of Sf21 cells (Fig. 6A). DNA hybrid-

A

TIME p.i. (hours)

TIME p.i. (hours)

TIME p.i. (hours)

FIG. 5. Time course of CAT activity in infected larvae. Early fifth instar larvae were injected with recombinant viruses and analyzed as described under Materials and Methods. (A) Larvae were injected with BmNPV/A.cat and CAT assays performed using 3 gg of protein incubated for 1 hr. Body wall tissues from two larvae were pooled for each assay. (B) Larvae were injected with BmNPVIP26.cat and CAT assays performed as in A. (C)Appropriate amounts of tissue extract were assayed to quantitate CAT activity at each point, the back- ground level observed at 2-4 hr p.i. was subtracted and values plot- ted on a logarithmic scale. Each point represents the average of two larvae. Solid circles, BmNPV/A.cat; open circles, BmNPV/P26cat.

BACULOVIRUS EXPRESSION CASSETTE 821

5 25 5 25 pl cells

Bm5 1 Sf21

25 25 pl cells

Bm5 I

Sf21

FIG. 6. CAT activity in transfected cells. Tissue culture cells (Bm5 or Sf21) were transfected with pBmA.cat in a 24.well microtiter plate as described under Materials and Methods. Wells were seeded with either 2 x 1 O5 Bm5 cells or 4 x 1 O5 Sf2 1 cells and the cells incubated in 200 ~1 transfection solution containing 5 pg/ml of pBmA.cat. (A) Two days after transfection cells from each well were collected, washed with PBS, lysed into 200 ~1 0.25 M Tris-HCI, pH 7.8, and CAT assays were performed using either 5 ~1 or 25 ~1 of cell extracts. (B) Cells were collected, washed as above, and suspended in 200 ~1 of PBS. The cell suspensions (2 ~1 or 5 ~1) were dot blotted and probed with pBS/SK+ DNA as described under Materials and Meth- ods.

ization, however, indicated that, under the transfection conditions used, the Bm5 cells had taken up at least twice as much plasmid DNA as equivalent volumes of Sf21 cells (Fig. 6B). These experiments, therefore, indi- cate that the B. mori actin promoter is active in S. frugi- perda cells at levels comparable to those observed in B. mori cells. Similar results were observed when pBmA.cat was transfected into cells of the spruce bud- worm, Choristoneura fumiferana (Cfl24; not shown).

DISCUSSION

We have developed a prototype expression cassette which can be used to generate recombinant baculovi- ruses that can express foreign proteins under the con- trol of a constitutive and ubiquitous cellular promoter. A recombinant BmNPV containing the cat reporter gene under the control of the cytoplasmic actin pro- moter directed expression of CAT protein in both in- fected Bm5 cells and in all examined tissues of infected B. mori larvae. Placement of the cat gene under the control of the actin promoter resulted in significantly faster induction of expression relative to that observed with the cat gene under the control of the polyhedrin promoter, both in infected Bm5 cells and in infected larvae. In infected larvae, the constitutive actin pro-

moter was active 24 hr earlier than the polyhedrin pro- moter. It is likely that the actin promoter is activated immediately upon entrance of the viral genome into the nuclei of infected cells. However, in our experiments, low levels of CAT which might have been produced in infected cells sooner than 5 hr p.i. would have been masked by the background activity. Significant amounts of protein would not be expected until several rounds of viral DNA replication have occurred to am- plify the number of gene copies present in the infected cells.

At early times p.i., the actin promoter directed sub- stantially more CAT accumulation in Bm5 cells than did the polyhedrin promoter. At 30 hr p.i., the activity of the polyhedrin promoter was equal to that of the actin pro- moter and, by 50 hr p.i., the polyhedrin promoter was three times more active. When larvae were infected by hemocelic injection, activity from the actin promoter was substantially higher than that from the polyhedrin promoter at all times up to 48 hr p.i. Even at 60 hr p.i., when polyhedrin-promoter-derived CAT activity was found to be higher than that from the actin cassette, the quantitative advantage of the polyhedrin promoter was only sixfold. Thus, at times which are relevant for the development of fast-acting baculovirus insecti- cides, the polyhedrin promoter does not reach a level of expression very much higher than the actin pro- moter. The earlier expression obtained from the actin- promoter-based vector could give a substantial advan- tage in the control of insect pests over currently avail- able baculovirus vectors based on late viral promoters. Recombinant viruses utilizing late promoters to ex- press insect-specific toxins have been found to kill lar- vae more rapidly than wild type viruses (Stewart et a/., 1991; Tomalski and Miller, 1991). If such toxic proteins were to be expressed using a constitutive cellular pro- moter such as that of the cytoplasmic actin gene, it should be possible to achieve even more rapid control, particularly if the toxin used is active at relatively low concentrations.

When larvae are infected orally with relatively low doses of polyhedra, as would normally occur under field conditions, the first cells to be infected are the columnar and regenerative cells of the midgut epithe- lium, while the connective layer of the midgut, which contains the muscle cells involved in gut contraction, becomes infected only a few hours after the epithelial cells (Keddie et al., 1989). In contrast, the generalized spread of the virus to other tissues of an infected larva through the circulation does not occur until 36 hr after viruses are observed in the gut epithelium. The very late stage of infection, in which the polyhedrin and pl0 promoters are active, would not begin until an even later time.

822 JOHNSON, MEIDINGER, AND IATROU

Since the cells of the midgut epithelium and connec- tive layer are infected only 13-l 8 hr after delivery of the virus, it would make sense to design recombinant bac- uloviruses which express proteins that would disrupt crucial functions of the gut, e.g., contraction, under the control of a promoterwhich is active in these cells. This strategy may be able to give more rapid control than that achieved from vectors that rely on control during the secondary spread of the virus to other tissues through the hemolymph.

It is unclear how active the polyhedrin and p10 pro- moters are in the midgut epithelium, as normal produc- tion of polyhedra is not obsetved in these ceils (Grana- dos and Lawler, 1981, and references therein). In these initially infected epithelial cells, very few of the newly produced virions are occluded into polyhedra, perhaps so that the viral progeny produced in these cells can be available for further infection of other tis- sues. The cytoplasmic actin promoter, on the other hand, has been found to be active in all B. mori tissues tested, including gut (Mounier and Prudhomme, 1991), and is presumably a constitutive housekeeping gene that is active in all cell types. This promoter, therefore, should be well suited for expression of foreign proteins in the gut immediately upon infection.

When silkworm larvae were infected with BmNPV/ A.cat or BmNPV/P26.cat by injection of ECV into the hemocel, CAT activity was observed in all tissues tested, indicating that all of these tissues can be in- fected during the secondary spread of the virus through the hemolymph and that both the actin and polyhedrin promoters are active in at least some cells in each of these tissues. Since we infected larvae by hemocelic injection rather than orally, these experi- ments do not address the question of promoter activity in the initially infected columnar and regenerative cells of the gut epithelium or the midgut connective layer after oral infection. To investigate transgene expres- sion in gut cells immediately after oral infection and compare the activities of the actin and polyhedrin pro- moters during the initial phase of infection, it will be necessary to employ recombinant viruses that are oc- cluded within polyhedra. We are now preparing such viruses, so that we can determine if the actin expres- sion cassette provides additional advantages over the polyhedrin promoter in terms of expression in gut cells during the initial phase of infection.

Finally, we have also determined that the actin-cat expression cassette is expressed in cells of S. frugi- perda (Sf21) and C. fumiferana (Cf124) at levels compa- rable to those in Bm5 cells, suggesting that it may be possible to utilize this expression cassette for con- structing recombinant baculoviruses to be used for control of a variety of lepidopteran species. An alterna-

tive approach would, of course, be to utilize a constitu- tive cellular promoter from the particular species for which control is desired. Alternatively, it might also prove possible to find and/or design a stronger consti- tutive or tissue-specific promoter that could allow for even higher levels of immediate expression and still better control.

ACKNOWLEDGMENTS

We thank N. Mounier and I. C. Prudhomme for providing us with the B. mori A3 gene clone, W. Qiu for maintaining the Bm5 cell cultures, and S. Mahalingam for maintaining the B. mori larvae. This work has been supported by Insect Biotech Canada, one of the Ca- nadian Networks of Centres of Excellence, and by a postdoctoral fellowship from the Alberta Heritage Foundation for Medical Re- search (to R.J.).

REFERENCES

BLISSARD, G. W., and ROHRMAN, G. F. (1990). Baculovirus diversity and molecular biology. Annu. Rev. Entomol. 35, 127-l 55.

GARBONELL. L. F., HODGE, M. R., TOMALSKI, M. D., and MILLER, L. K. (1988). Synthesis of a gene coding for an insect-specific scorpion neurotoxin and attempts to express it using baculovirus vectors. Gene 73,409-418.

FOTAKI, M. E., and IATROU, K. (1988). Identification of a transcription- ally active pseudogene in the chorion locus of the silkmoth Bom- byx mori. J. Mol. Biol. 203, 849-860.

GOSWAMI. B. B., and GLAZER, R. I. (1991). A simplified method for the production of recombinant baculoviruses. Biotechniques 10, 626- 630.

GRANADOS, R. R., and BAWLER, K. A. (1981). In viva pathway of Auto- grapha californica baculovirus invasion and infection. J. \/irol. 108, 297-308.

HUEIER, J. (1986). Use of baculoviruses in pest management pro- grams. In “The Biology of Baculoviruses” (R. R. Granados and B. A. Federici, Eds.), CRC Press, Boca Raton, FL.

IATROU, K. (1992). Engineered baculoviruses: Molecular tools for lep- idopteran developmental biology and physiology and potential agents for insect pest control. In “Molecular Biology and Molecu- lar Genetics of Lepidoptera” (M. R. Goldsmith and A. S. Wilkins, Eds.). Cambridge Univ. Press, U.K. In press.

IATROU, K., ITO, K., and WITKIEWICZ, H. (1985). Polyhedrin gene of Bombyx mori nuclear polyhedrosis virus. J. viral. 54, 436-445.

IATROU, K., and MEIDINGER. R. G. (1989). Bombyx mori nuclear poly- hedrosis virus-based vectors for expressing passenger genes in silkmoth cells under viral or cellular promoter control. Gene 75, 59-7 1.

IATROU, K., and MEIDINGER, R. G. (1990). Tissue-specific expression of silkmoth chorion genes in viva using Bombyxmori nuclear poly- hedrosis virus as a transducing vector. Proc. Nat/. Acad. Sci. USA 87, 3650-3654.

KEDDIE, B. A., APONTE, G. W., and VOLKMAN, L. E. (1989). The path- way of infection of Autographa californica nuclear polyhedrosis virus in an insect host. Science 243, 1728-1730.

KOVACS, G. R., GUARINO, L. A., and SUMMERS, M. D. (1991). Novel regulatory properties of the IEl and IEO transactivators encoded by the baculovirus. Autographa californica multicapsid nuclear polyhedrosis virus. 1. L’irol. 65, 5281-5288.

KUNKEL, T. A. (1985). Rapid and efficient site-specific mutagenesis without phenotypic selection. froc. Nat/. Aced. Sci. USA 82,488- 492.

BACULOVIRUS EXPRESSION CASSETTE 823

MAEDA, S. (1989). Expression of foreign genes in insects using bacu- loviruses vectors. Annu. Rev. Entomol. 34, 35 l-372.

MAEDA, S. (1989). Increased insecticidal effect by a recombinant baculovirus carrying a synthetic diuretic hormone gene. Biochem. Biophys. Res. Commun. 165, 1177-I 183.

MAEDA, S. eta/. (1991). Insecticidal effects of an insect-specific neu- rotoxin expressed by a recombinant baculovirus. Virology 184, 777-780.

MAW\M, A. M., and GILBERT, W. (1977). A new method for sequenc- ing DNA. Proc. Natl. Acad. Sci. USA 74, 560-564.

MCCUTCHEON, B. F. et al. (1991). Development of a recombinant baculovirus expressing an insect-selective neurotoxin: Potential for pest control. &o/Technology 9, 848-852.

MERRYWEATHER, A. T. et al. (1990). Construction of genetically engi- neered baculovirus insecticides containing the Bacillus thuringien- sis subsp. kurstaki HD-73 delta endotoxin. J. Gen. Viral. 71, 1535- 1544.

MILLER, L. K. (1988). Baculoviruses as gene expression vectors. Annu. Rev. Microbial. 42, 177-l 99.

MILLER, L. K. (1989). Insect baculoviruses: Powerful gene expression vectors. Bioessays 11, 91-95.

MITSIALIS, S. A., SPOEREZ, N., LEVITEN, M., and KAFATOS, F. C. (1987). A short DNA region sufficient for developmentally correct expres- sion of moth chorion genes in Drosophila. Proc. Nat/. Acad. Sci. USA 84, 7987-7991.

MOLINIER, N., and PRUDHOMME, J. C. (1986). Isolation of actin genes in Bombyxmori; The coding sequence of a cytoplasmic actin gene expressed in the silk gland is interrupted by a single intron in an unusual position. Biochimie 68, 1053-l 06 1.

MOUNIER, N., and PRUDHOMME, J. C. (1991). Differential expression of muscle and cytoplasmic actin genes during development of Bom- byx mori. Insect Biochem. 21, 523-533.

PEN, J., WELLING, G. W.. and WELLING-WESTER, S. (1989). An efficient procedure for the isolation of recombinant baculovirus. Nucleic Acids Res. 17, 45 1.

STEWART, L. M. D. et a/. (1991). Construction of an improved baculo- virus insecticide containing an insect-specific toxin gene. Nature 352,85-88.

SUMMERS, M. D. (1992). Baculovirus directed foreign gene expres- sion. ln “Insect Neuropeptides” (J. Menn, T. Kelly, and E. Masler, Eds.), pp. 237-251. American Chemical Society Symposium Se- ries, No. 453.

TOMALSKI, M. D., and MILLER, L. K. (1991). Insect paralysis by baculo- virus-mediated expression of a mite gene. Nature 352, 82-85.

WANG, X., 001, 6. G., and MILLER, L. K. (1991). Baculovirus vectors for multiple gene expression and for occluded virus production. Gene 100, 131-137.

WOOD, H. A., and GRANADOS, R. R. (1991). Genetically engineered baculoviruses as agents for pest control. Annu. Rev. Microbial. 45, 69-87.