---

Original Paper	Pdf file on Synergy OPEN omments
Acta Biochim Biophys Sin 2006, 38: 577-585
doi:10.1111/j.1745-7270.2006.00195.x

Disruption Effect of Microplitis bicoloratus Polydnavirus EGF-like Protein, MbCRP, on Actin Cytoskeleton in Lepidopteran Insect Hemocytes

Kai-Jun LUO^1,2 and Yi PANG¹*

¹ State Key Laboratory of Biocontrol and Institute of Entomology, Sun Yat-Sen University, Guangzhou 510275, China;

² Agricultural Environment and Resource Research Institute, Yunnan Academy of Agricultural Sciences, Kunming 650205, China

Received: March 23, 2006�������

Accepted: April 29, 2006

This work was supported by the grants from the Major State Basic Research Development Program of China (No. G2000016209), the National Natural Science Foundation of China (No. 30530540) and Open-funding of the State Key Laboratory of Biocontrol at Sun Yat-Sen University (No. 0404)

*Corresponding author: Tel, 86-20-84113860; Fax, 86-20-84037472; E-mail, [email protected]

Abstract������� Microplitis bicoloratus is a braconid endoparasitic wasp associated with the polydnavirus named Microplitis bicoloratus bracovirus (MbBV). Parasitism of Spodoptera litura larvae leads to an impaired� cellular immune response and to the disappearance of the 42 kDa actin in host hemocytes. In this work, we investigated if the absence of actin in blood cells was related to MbBV infection. An MbBV gene similar to egf-like genes identified in another bracovirus was partially cloned and named Mbcrp1. The full-length gene, named Mbcrp, is transcribed throughout the course of parasitism in host hemocytes and the 30 kDa MbCRP protein was detected in hemocytes 6-7 d post-parasitization. The Mbcrp1 gene contains the cysteine-rich trypsin inhibitor-like (TIL) domain coding sequence and the expression of recombinant MbCRP1 inhibited the expression of the 42 kDa actin in Hi5 cells. The 34.1 kDa MbCRP1-green fluorescent protein fusion protein locate specifically in the cytoplasm. These results suggest that expression of MbCRP in lepidopteran insect cells is related to the disruption of the actin cytoskeleton.

Key words������� actin cytoskeleton; polydnavirus; Microplitis bicoloratus; cellular immune response; trypsin inhibitor-like domain; MbCRP1

The Polydnaviridae family consists of segmented, double-stranded DNA viruses that are specifically associated� with certain types of parasitoid wasps [1]. Polydnaviruses (PDVs) are divided into two families, bracoviruses (BVs) and ichnoviruses (IVs) that coexist with wasps from the families Braconidae and Ichneumonidae, respectively [2]. The virion morphologies of BVs and IVs differ significantly. BVs have cylindrical nucleocapsids of various lengths, with virions made of one or more nucleocapsids surrounded by a single membrane, whereas those of IVs have lenticular nucleocapsids of uniform size, individually surrounded by two membranes [2,3]. Nevertheless, for both BVs and IVs, virion assembly takes place in the nuclei of calyx cells situated between the ovarioles and lateral oviducts, and viral particles are injected in the lepidopteran along with the wasp's egg. The expression of a viral gene causes the suppression of the host's immune system that is essential for the survival of the parasitoid's progeny [1,4-6].

PDVs can infect most larval tissues, particularly hemocytes [7,8]. In parasitized host larvae, hemocytes might become apoptotic [5,9] or lose their ability to adhere� to a substrate [4,9-12]. Some polydnaviral genes are reportedly� involved in hemocyte function disruption. For example, Glc1.8, a protein encoded by the Microplitis demolitor bracovirus (MdBV), induces a loss of adhesion and phagocytosis in insect Hi5 and S2 cells [13]. Cr1, a protein encoded by the Cotesia congregata bracovirus, prevents cell surface exposure of lectin-binding sites and microparticle formation during immune stimulation of hemocytes. Cr1 is a secreted glycoprotein that has been implicated in depolymerization of the actin cytoskeleton of hemocytes, resulting in hemocyte inactivation [14]. Despite evidence that PDV infection is correlated with hemocyte cytoskeleton disruption, the mechanisms by which PDV proteins affect the cytoskeleton remain unclear.

MdBV encodes three genes, egf1.0, egf1.5 and egf0.4 (GenBank accession No. U76034, U76033 and DQ000240, respectively), which are characterized by an epidermal growth factor-like cysteine-rich motif and an amino-terminal trypsin inhibitor-like (TIL) domain [15,16]. The MdBV egf genes are transcribed 12-24 h post-parasitization (p.p.) and the peak of their expression coincides with physiological changes observed in host hemocytes. This suggests that the proteins encoded by egf genes are related� to immune suppression [15].

The closely related species Microplitis bicoloratus is a parasitoid of Spodoptera litura larvae and is associated with the Microplitis bicoloratus bracovirus (MbBV). As for other parasitoids, parasitization of S. litura larvae by M. bicoloratus leads to suppression of the immune response� and developmental arrest of the host larvae. Interestingly, we found that parasitism was correlated with an absence of 42 kDa actin in S. litura hemocytes.

In the present study, we isolated a sequence from MbBV using primers designed according to MdBV egf sequences which was named Mbcrp1. The recombinant MbCRP1 contains a TIL domain, and is concomitant with actin decrease in infected Hi5 cells. We also detected a 30 kDa protein, MbCRP, in hemocytes 6-7 d post-parasitization using antiserum of the recombinant MbCRP1.

Materials and Methods

Cell lines, polydnavirus and parasitoid rearing

Spodoptera frugiperda (Sf9) and Trichoplusia ni (Hi5) cells were maintained in Grace's medium (Gibco BRL, Gaithersburg, USA) supplemented with 10% fetal bovine serum.

The parasitoid colony was maintained on S. litura larvae� reared in the laboratory; adults were provided honey as a dietary supplement. MbBV DNA was extracted and analyzed� according to established methods [17].

Electron microscopy

Reproductive tracts of M. bicoloratus females were pulled out in Pringle's solution [18]. Isolated ovaries were fixed in 2.5% glutaraldehyde overnight at 4 �C, washed three times with phosphate-buffered saline (PBS) buffer (pH 7.2) and post-fixed in 1% osmium tetroxide in PBS for 1 h at 4 �C. After fixation, ovaries were washed three times with PBS buffer, dehydrated in graded ethanol and soaked in acetone. Samples were embedded in Spurr's resin. Ultra-thin sections were stained with uranyl acetate and lead citrate and observed under a JEM-100CXII transmission electron microscope (Jeol, Tokyo, Japan) operating at 100 kV.

For negative staining, ovaries isolated from three female wasps were placed in 50 ml ice-cold Pringle's solution and gently dissected to allow the calyx fluid to diffuse. The extract was then centrifuged at 1000 g for 3 min to remove cellular debris and eggs. The supernatant was adsorbed on a Formvar-coated (Nisshin EM) electron microscopy grid stained with 2% phosphotungstic acid, dried on filter paper and observed with the JEM-100CXII microscope.

egf-like gene cloning

Two primers were designed based on sequence alignment of MdBV egf genes. These primers, P1S, 5'-CACCTGCTTTTCCTGTTTGCATTTT-3' and P1A, 5'-CCGTTTTGAAGAATCATTGTTGGC-3', correspond to the N-terminal cysteine-rich conversed region of MdBV egf genes. To isolate an egf-like gene from MbBV, a polymerase� chain reaction (PCR) was carried out using MbBV DNA as a template. The amplified fragment was cloned into pMD18-T (TaKaRa, Dalian, China) vector and sequenced. The sequence was named Mbcrp1 and submitted� to GenBank with accession No. DQ286649.�

Sequence analysis

Nucleotide BLAST searches were carried out against the GenBank nr database at the National Centre for Biotechnology Information website (www.ncbi.nlm.nih.gov/blast). Protein analysis tools were found at the ExPASy molecular biology server (http://au.expasy.org/tools) and EMBL-EBI (http://www.ebi.ac.uk).

Plasmid construction and preparation of MbCRP1 antiserum

Two primers were designed to amplify the coding� sequence of Mbcrp1 with open reading frame� using MbBV DNA as a template: P2S�, 5'-GGATCCATGCCTACTAAAGAAAGTGA-3' (BamHI site underlined); and P2A, 5'-GTCGACTTAGTTGTAATAGCAGTAAA-3' (SalI site underlined). The amplified fragment was cloned into pMD18-T vector, then into the expression vector pQE30 (Qiagen, Hilden, Germany) with a 6�His-tag sequence. The resulting plasmid pQE30-MbCRP1 was checked by sequencing and transformed into Escherichia coli M15.

The Mbcrp1 expression under control of bacteriophage T5 promoter in E. coli M15 was induced with 1 mM isopropyl-b-D-thiogalactopyranoside as recommended by the manufacturer (Qiagen). Expressed 6�His-MbCRP1 was purified from E. coli M15 by Ni-NTA agarose under denaturing� conditions as described in the handbook (QIAexpressionist; Qiagen). Purified 6�His-MbCRP1 protein� was used to raise antiserum in rabbits according to the method described by Sambrook et al. [19] using Freund's adjuvant (Gibco BRL).

Reverse transcription-polymerase chain reaction (RT-PCR)

Total RNA was extracted using the Qiagen RNA Kit (Invitrogen, Valencia, USA). For S. litura larvae, total RNA was isolated� from 100 ml hemocytes recovered from non-parasitized and parasitized S. litura larvae. For Hi5 cells, total RNA of Hi5 cells was isolated from 5105 mock-infected and recombinant baculovirus (see below) infected Hi5 cells at multiplicity of infection (MOI) of 5 at various time points post-infection (p.i.).

RT-PCR was carried out using an RNA-PCR-AMV Kit (Version 2.1; TaKaRa) with 3 mg of total RNA as the template. First-strand cDNA was synthesized using avian myeloblastosis virus (AMV) reverse transcriptase and random 9-mer according to the manufacturer's instructions. The cDNA mixture was amplified with MbCRP1-specific primers� P2S and P3A-5'-TCTAGAGTTGTAATAGCAGT�AAA-3' (XbaI site underlined). The obtained PCR products� were analyzed in 1.0% agarose gel.

Construction and identification of the recombinant viruses�

Transfer-vector plasmids were constructed to generate� recombinant AcMNPV expressing MbCRP1 protein. MbCRP1 was cloned into the BamHI and XbaI sites of the pFastBacI vector (Invitrogen, Rockville, USA). Green fluorescent protein (GFP) gene was also cloned into the XbaI and PstI sites of pFastBac-MbCRP1 vector as a C-terminus fusion to MbCRP1. pFastBacI vector was used as a control. The pFastBac-MbCRP1-GFP and pFastBac-GFP vectors were transformed into E. coli DH10Bac cells (Gibco BRL) and positive colonies were selected as described� by the manufacturer. To obtain recombinant viruses, high molecular mass DNA was purified from the selected colonies as recommended by the manufacturer and transfected into Sf9 cells. The recombinant baculo�viruses Ac-MbCRP1-GFP and Ac-GFP, named reAcMG and reAcG, respectively, were amplified to passage� 3 in Sf9 cells. Titers of recombinant viruses were determined by tissue culture infectious dose 50 assay with infected Sf9 cells.

Collection of protein samples in hemocytes and Western� blot analysis

S. litura larvae were bled from a proleg on ice. A total of 50-100 ml of hemocytes was pooled at various� time points and centrifuged at 3000 g for 3 min. Supernatant (cell-free hemolymph) was collected and one volume of 2�sodium dodecylsulfate (SDS) loading buffer was added. The cellular pellet was resuspended in 50 ml of double distilled water and one volume of 2�SDS loading buffer was added. Equal non-heated protein samples were electrophoresed on denaturing 12% SDS-polyacrylamide gels [20]. Protein bands were either stained with Coomassie brilliant blue or transferred to a nitrocellulose membrane (Schleicher and Schuell, Dassel, Germany) as described earlier [19]. The membranes were blocked with 1% blocking� solution (Roche, Mannheim, Germany) then probed with anti-MbCRP1 antiserum at a dilution of 1:200. Immunoreactive proteins were visualized using goat anti-rabbit immunoglobulin G and Ap-conjugate (Roche) following� the protocol provided by the manufacturer. Actin� blots were probed with a 1:5000 dilution of anti-actin (Calbiochem, San Diego, USA) and visualized with Ap-conjugated goat anti-mouse secondary antibodies at a 1:10,000 dilution (Roche).

Assessment of transcription and expression of MbCRP1 in infected Hi5 cells

Transcription and expression of MbCRP1 in infected Hi5 cells was assessed by RT-PCR and Western blot. Hi5 cells were seeded at a density of 5�105 cells per well in a 35 mm plate (Corning, Corning, USA). Cells were infected with recombinant virus at MOI of 5. Cells were harvested 6-72 h p.i.. RT-PCR and Western blot were carried out as described above.

Fluorescence microscopy

Hi5 cells (1�105) were grown on glass cover slips in 35 mm Petri dishes (Corning) and infected with reAcMG and reAcG at MOI of 5. At 1248 h p.i., the cells were examined� with a confocal-laser scanning fluorescence microscope (Leica SDK; Leica, Heidelberg, Germany).

Results

The virion morphology and DNA analysis of MbBV

The polydnavirus associated with M. bicoloratus is a typical bracovirus, as indicated by virion morphology and genome characteristics. The negative staining of MbBV virions showed evident tail-like appendages [Fig. 1(A), arrows] and heads [Fig. 1(A), asterisks]. The particles consisted primarily of one or more nucleocapsids surrounded by a single envelope [Fig. 1(B), arrows]. The MbBV genome had estimated 11 segments ranging from 8000 to 50,000 bp [Fig. 1(C)].

The cloning of a partial egf-like gene in MbBV

Due to the congeneric nature of MbBV and MdBV, we expected that the MbBV genome can also encode members of the egf-like family. Therefore we used primers designed in conserved regions of MdBV egf genes and obtained a 652 nt amplifier using MbBV DNA as a template. Sequencing confirmed that the amplified fragment was the 5' region of an egf-like gene (BlastN e-value=5e-131; BlastX e-value=4e-11). The full egf-like gene was named Mbcrp, and the fragment we obtained was named Mbcrp1. Alignment with MdBV genomic sequences indicated that the Mbcrp1 contains partial exon sequence of Mbcrp encoding� the TIL domain (data not shown). This was confirmed by InterPro Scan and MotifScan analysis of MbCRP1 that MbCRP1 consists of 61 amino acids and contains a TIL domain (Fig. 2) encompassing a cysteine-rich repeat, located in position 35-51. And egf motif is absent in the Mbcrp1.

MbCRP is expressed in hemocytes of S. litura larvae naturally parasitized by M. bicoloratus

Transcription of Mbcrp1 in hemocytes of naturally parasitized S. litura larvae was examined by RT-PCR and Western� blot (Fig. 3). A band of 195 bp was detectable 1-7 d p.p., although the bands were very faint 1-3 d p.p. [Fig. 3(A)]. This indicated that the Mbcrp gene is transcribed in S. litura hemocytes and transcription appears to be persistent during parasitoid larval development.

This result was confirmed by Western blot analysis. Indeed, in parasitized S. litura, the MbCRP1 antiserum recognized a specific protein of approximately 30 kDa (putative MbCRP protein) in larvae hemocyte lysate. The 30 kDa band was observed 6 and 7 d p.p. in hemocyte lysate [Fig. 3(B)] but was absent from cell-free hemolymph (data not shown). This indicated that the putative MbCRP protein was present only within the cell.

Effects of M. bicoloratus parasitism on actin in S. litura hemocytes

To test the effects of M. bicoloratus parasitization on actin in hemocytes, we analyzed expression of the 42 kDa actin in hemocyte lysate collected from parasitized S. litura larvae. In our experiments, the monoclonal anti-actin antibody� recognized the 42 kDa actin. Western blot analysis� showed that this monoclonal antibody recognized the 42 kDa band only in hemocytes from non-parasitized larvae [Fig. 3(C), lane U]. Indeed, no signal was detected in hemocytes of parasitized larvae 3-7 d p.p.. This result suggests that the expression of the 42 kDa actin is inhibited� by parasitism, which might lead to the disappearance of functional actin filaments in S. litura hemocytes.

Transcription and expression of MbCRP1 in Hi5 cells infected by reAcMG and effects of infection on actin of Hi5 cells

To further investigate if there is a link between absence of actin and expression of MbCRP in infected lepidopteran cells, we constructed a recombinant baculovirus expressing the MbCRP1 TIL domain of MbCRP and analyzed expression� of actin in the baculovirus-infected Hi5 insect cell line.

After infection of the Hi5 cell line with the baculovirus reAcMG, transcription of MbCRP was analyzed by RT-PCR. The 195 bp PCR-amplified fragment corresponding to Mbcrp1 was detected as early as 6 h p.i. and continued� to be present until 72 h p.i. [Fig. 4(A)]. Hi5 cells were thus successfully infected by recombinant baculovirus reAcMG and MbCRP1 was persistently transcribed in infected� cells. Western blot analysis using the anti-MbCRP1 antiserum confirmed the presence of the 34.1 kDa MbCRP1-GFP fusion protein in infected Hi5 cell lysate [Fig. 4(B)]. The 34.1 kDa band was first detected� 12 h p.i., with a peak of expression 48 h p.i., then expression clearly declined 72 h p.i. [Fig. 4(B)].

Expression of 42 kDa actin in Hi5 cells infected by reAcMG and reAcG was analyzed by Western blot. The 42 kDa band was detected in mock-infected Hi5 cells [Fig. 4(C,D), lane Mi] and in Hi5 cells infected by reAcG recombinant virus 6-72 h p.i. [Fig. 4(D)]. In contrast, in Hi5 cells infected by reAcMG the 42 kDa band was detected 6-24 h p.i., but was absent 48-72 h p.i. [Fig. 4(C)]. Therefore, in baculovirus-infected Hi5 cells, expression of MCRP1 seems to be correlated with an absence of 42 kDa actin expression.

At the time when MbCRP1 shows a peak of expression and 42 kDa actin is no longer expressed, approximately 48 h p.i., Hi5 cells infected with reAcMG had lost their adhesive properties and floated in the culture medium. In contrast, cells infected with reAcG remained attached to the dishes. This observation confirms the hypothesis that MbCRP1 is in some way related to disruption of the actin cytoskeleton of Hi5 cells.

The subcellular localization of the MbCRP1 protein was investigated using confocal laser scanning microscopy. Hi5 cells were infected with reAcG and reAcMG, incubated� and examined for fluorescence at various time points p.i. (Fig. 5). GFP alone showed homogeneous fluorescence in the cytoplasm and nucleus 12-48 h p.i. The MbCRP1-GFP fusion protein was localized predominantly in the cytoplasm.

Discussion

Our study indicates that the presumed cysteine-rich EGF-like protein� MbCRP of MbBV is probably related to the disruption of the actin cytoskeleton observed in the hemocytes of S. litura larvae parasitized by M. bicoloratus. Indeed, we found that infection of lepidopteran cell lines with a baculovirus expressing the MbCRP TIL domain resulted in inhibition of expression of the 42 kDa actin. Moreover, in infected cell lines, the peak expression of MbCRP1 coincided with the complete disappearance of the 42 kDa actin [Fig. 4(B,C)].

In parasitized S. litura larvae, Mbcrp1 is transcribed in host hemocytes throughout the course of parasitism [Fig. 3(A)] and the MbCRP protein is detected in the cytoplasm of polydnavirus-infected hemocytes 6-7 d p.p. using MbCRP antiserum [Fig. 3(B)]. By this time, the 42 kDa actin is no longer expressed in hemocytes [Fig. 3(C)].

Cytoskeletal proteins are found in highly organized arrays� within the cytoplasm of higher eukaryotic cells. These proteins are intimately involved in functions such as cell and intracellular organelle movement, maintenance of cell shape and endocytosis [21,22]. The cytoskeleton of most eukaryotic cells includes the microfilaments composed� of actin, the microtubules composed of tubulin� and the intermediate filaments composed of vimentin or desmin.

The absence of functional actin filaments in MbCRP-expressed cells was probably the primary cause of cells or hemocyte inactivation. Although little is known about the molecular mechanisms leading to hemocyte activation, processes such as attachment, aggregation [23,24] and coagulation reaction [25] involve complex rearrangement of the actin cytoskeleton [26]. In vitro, the expression of the MdBV glc1.8 causes the disruption of F-actin in the cytoplasm and induces the accumulation of F-actin at the periphery of cells leading to a loss of adhesion by hemocyte-like cell lines, namely, Hi5 [14]. In our study, the expression of recombinant MbCRP1 also greatly reduced� the ability of these cells to adhere to a substrate, which was probably linked to the absence of the 42 kDa actin (Fig. 4). Some other PDV genes have been implicated� in disruption of the actin cytoskeleton but their subcellular� localizations were distinctly different: Glc1.8 protein localized� to the surfaces of infected cells [13] and Cr1 protein was secreted [14], whereas MbCRP localized in the cytoplasm (Fig. 5).

The unique intracellular location of MbCRP protein suggested� it is probably involved in specific roles within the actin cytoskeleton. Further studies would be necessary� to investigate whether MbCRP protein is an actin-binding protein or not, through protein-protein interactions. In our experiments, we also found that the 42 kDa actin declined in Hi5 cells infected by reAcG at 48 and 72 h p.i. [Fig. 4(D)]. We presume that this side-effect might cause due to Bac-to-Bac expression system. To test this hypothesis, other insect cell expression systems should be used in further research.

References

 1�� Webb BA, Strand MR. The biology and genomics of polydnaviruses. In: Gilbert LI, Iatrou K, Gill SS eds. Comprehensive Molecular Insect Science. San Diego: Elsevier Press 2004

 2�� Webb BA, Beckage NE, Hayakawa Y, Krell PJ, Lanzrein B, Strand MR, Stoltz DB et al. Polydnaviridae. In: van Regenmortel MHV, Fauquet CM, Bishop DHL, Carstens EB, Estes MK, Lemon SM, Maniloff J et al. eds. Virus Taxonomy. San Diego: Academic Press 2000

 3�� Stoltz DB. The polydnavirus life cycle. In: Beckage NE, Thompson SN, Federici BA eds. Parasites and Pathogens of Insects. San Diego: Academic Press 1993

 4�� Schmidt O, Theopold U, Strand MR. Innate immunity and its evasion and suppression by hymenopteran endoparasitoids. Bioessays 2001, 23: 344-351

 5�� Strand MR, Pech LL. Immunological basis for compatibility in parasitoid-host relationships. Annu Rev Entomol 1995, 40: 31-56

 6�� Webb BA. Polydnavirus biology, genome structure and evolution. In: Miller LK, Ball AL eds. The Insect Viruses. New York: Plenum Press 1998

 7�� Stoltz DB, Vinson SB. Penetration into caterpillar cells of virus-like particles injected during oviposition by parasitoid ichneumonid wasps. Can J Microbiol 1979, 25: 207-216

 8�� Wyder S, Blank F, Lanzrein B. Fate of polydnavirus DNA of the egg-larval parasitoid Chelonus inanitus in the host Spodoptera littoralis. J Insect Physiol 2003, 49: 491-500

 9�� Amaya KE, Asgari S, Jung R, Hongskula M, Beckage NE. Parasitization of Manduca sexta larvae by the parasitoid wasp Cotesia congregata induces an impaired host immune response. J Insect Physiol 2005, 51: 505-512

10� Lavine MD, Strand MR. Insect hemocytes and their role in immunity. Insect Biochem Mol Biol 2002, 32: 1295-1309

11� Lavine MD, Beckage NE. Polydnaviruses: Potent mediators of host insect immune dysfunction. Parasitol Today 1995, 11: 368-378

12� Asgari S, Hellers M, Schmidt O. Host haemocyte inactivation by an insect parasitoid: Transient expression of a polydnavirus gene. J Gen Virol 1996, 77: 2653-2662

13� Beck M, Strand MR. Glc1.8 from Microplitis demolitor bracovirus induces a loss of adhesion and phagocytosis in insect high five and S2 cells. J Virol 2005, 79: 1861-1870

14� Asgari S, Schmidt O, Theopold U. A polydnavirus-encoded protein of an endoparasitoid wasp is an immune suppressor. J Gen Virol 1997, 78: 3061-3070

15� Strand MR, Witherell RA, Trudeau D. Two Microplitis demolitor polydnavirus mRNAs expressed in hemocytes of Pseudoplusia includens contain a common cysteine-rich domain. J Virol 1997, 71: 2146-2156

16� Trudeau D, Witherell RA, Strand MR. Characterization of two novel Microplitis demolitor polydnavirus mRNAs expressed in Pseudoplusia includens haemocytes. J Gen Virol 2000, 81: 3049-3058

17� Albrecht U, Wyler T, Pfister-Wilhelm R, Gruber A, Stettler P, Heiniger P, Kurt E et al. Polydnavirus of the parasitic wasp Chelonus inanitus (Braconidae): Characterization, genome organization and time point of replication. J Gen Virol 1994, 75: 3353-3363

18� Pringle JHW. Proprioception in insects. J Exp Biol 1938, 15: 101-103

19� Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd ed. New York: Cold Spring Harbor Laboratory Press 1989

20� Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970, 227: 680-685

21� Mitchison TJ, Cramer LP. Actin-based cell motility and cell locomotion. Cell 1996, 84: 371-379

22� Yeaman C, Grindstaff KK, Nelson WJ. New perspectives on mechanisms involved in generating epithelial cell polarity. Physiol Rev 1999, 79: 73-98

23� Gupta AP. Insect immunocytes and other hemocytes: Roles in humoral and cellular immunity. In: Gupta AP ed. Immunology of Insects and Other Arthropods. Boca Raton: CRC Press 1991

24� Ratcliffe NA. Cellular defense responses of insects: Unresolved problems. In: Beckage NE, Thompson SN, Federici BA eds. Parasites and Pathogens of Insect Species. San Diego: Academic Press 1993

25� Bohn H. Hemolymph clotting in insects. In: Berlin MB, ed. Immunity in Invertebrates. Berlin/Heidelberg: Springer-Verlag Press 1986

26� Gruenheid S, Finlay BB. Microbial pathogenesis and cytoskeletal function. Nature 2003, 422: 775-781