Previous Article | Next Article ![]()
Journal of Virology, December 2004, p. 13090-13103, Vol. 78, No. 23
0022-538X/04/$08.00+0 DOI: 10.1128/JVI.78.23.13090-13103.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Patrizia Falabella,4
Elisabeth Huguet,1
Raffaella La Scaleia,4
Laurence Cattolico,5
Marylène Poirié,1
Carla Malva,2
Julie A. Olszewski,3
Francesco Pennacchio,4 and
Jean-Michel Drezen1*
Institut de Recherche sur la Biologie de l'Insecte, UMR CNRS 6035, Faculté des Sciences et Techniques, Tours,1 Genoscope, Centre National de Séquençage, Evry, France,5 Istituto di Genetica e Biofisica, Consiglio Nazionale della Richerche, Naples,2 Dipartimento di Biologia, Difesa e Biotecnologie Agro-Forestali, Università della Basilicata-Macchia Romana, Potenza, Italy,4 Department of Biological Sciences, Imperial College, London, United Kingdom3
Received 21 April 2004/ Accepted 19 July 2004
|
|
|---|
|
|
|---|
The relationship between wasps and polydnaviruses constitutes one of the few known cases of mutualism between viruses and eukaryotes (15), in addition to retroviruses contributing to human placenta organogenesis (41). Viral replication is restricted to specialized cells of the wasp ovary, and the virus is transmitted exclusively by Mendelian inheritance (55) through a proviral form, chromosomally integrated in the wasp genome (10, 21, 51). Virus particles are injected during wasp oviposition in the hemocele of the host, typically a lepidopteran larva. In the host cells, viral genes are expressed, but there is no viral replication (58, 70). Polydnavirus gene products are responsible for several alterations of host physiology induced by parasitism, comprising suppression of immunity and control of the caterpillar developmental program (61).
Polydnaviruses harbored by wasps from the Braconidae family belong to the genus bracovirus and are associated with wasp species forming a monophyletic group, the microgastroid complex (which comprises seven braconid subfamilies: Cheloninae, Dirrhopinae, Mendesellinae, Khoikhoiinae, Cardiochilinae, Miracinae, and Microgastrinae) (69). All wasp-bracovirus associations might thus have originated from a unique event, the integration of an ancestral bracovirus in the chromosome of the ancestor of the microgastroid complex living 73.7 ± 10 million years ago (66). Since this integration event, the bracovirus is likely to have contributed to the tremendous diversification of the microgastroid lineage, comprising 17,500 described species (37). Some genes inherited from the putative ancestral virus are expected to be common to all bracoviruses. However, while numerous bracovirus genes have been characterized in the last few years (12, 13, 31, 63), conserved genes have been identified so far only in closely related species of the Cotesia genus (39, 67).
In this paper, we present the characterization of a gene family shared by two subfamilies of bracovirus-bearing wasps, Microgastrinae and Cardiochilinae, that diverged more than 40 million years ago (66). Twenty-seven members of this gene family were identified in the genomes of CcBV, the symbiotic virus of Cotesia congregata (Microgastrinae), and 13 members in TnBV, the bracovirus associated with Toxoneuron nigriceps (Cardiochilinae). These genes encode proteins belonging to the protein tyrosine phosphatase family which, together with protein tyrosine kinases, are known to play a key role in the control of signal transduction pathways in vertebrates (3).
The protein tyrosine phosphatase family is divided into three groups of enzymes: classical, dual-specificity, and low-molecular-weight protein tyrosine phosphatases. The bracovirus protein tyrosine phosphatases identified belong to the classical protein tyrosine phosphatases, characterized by the domain composed of 10 conserved motifs whose functions are well known from structural and functional analyses in vertebrates (3). Although it has been proposed that the ancestral bracovirus was a baculovirus (15, 19), the bracovirus protein tyrosine phosphatases are not related to baculovirus or poxvirus protein tyrosine phosphatases, which belong to the dual-specificity protein tyrosine phosphatases.
To initiate functional characterization of protein tyrosine phosphatases and to assess the role they might play in host-parasite interactions, their expression profile was analyzed. Furthermore, two CcBV protein tyrosine phosphatase proteins were produced, with recombinant baculoviruses, in order to determine whether they display tyrosine phosphatase activity.
|
|
|---|
Viral DNA purification and sequencing. Wasp ovaries were dissected and viral DNA was extracted from virus particles purified by filtration or sucrose gradient as previously described (8, 18). CcBV DNA segments were sequenced with a shot gun strategy (17a). Sequencing of TnBV DNA was carried out at the TIGEM-IGB sequencing core service, Naples, Italy.
The coding DNA sequences were identified, and the potential genes were designated CcBV PTP for protein tyrosine phosphatase followed by a letter (A to Z and
to
) and TnBV PTP followed by a number (1 to 13).
RNA isolation. RNA isolation was performed on different M. sexta tissues, which were dissected from 10 fifth-instar larvae and separately rinsed several times in Ringer solution before their use. For parasitized host larvae, dissections were carried out 12 and/or 24 h postoviposition. Parasitization was checked visually, and only host larvae stung twice by C. congregata females were processed. Synchronous nonparasitized host larvae served as controls. Hemocytes were collected by hemolymph centrifugation at 800 x g for 10 min. All the tissues were crushed for 1 min at 4°C with an Ultraturrax T8 (Ika-Verke-Staufen) homogenizer. RNA extractions were carried out with the RNeasy plant kit (Qiagen) according to the manufacturer's instructions. The quality of samples was checked by electrophoresis on an ethidium bromide-stained 1% agarose gel. To eliminate trace amounts of viral DNA present in total RNA, polyadenylated RNAs were purified with the Oligotex kit (Qiagen). H. virescens RNA extraction from parasitized and nonparasitized host tissues and polyadenylated RNA purification were carried out as previously described (18).
Preparation and screening of genomic and cDNA libraries. Viral genomic libraries and cDNA libraries from parasitized host tissues, such as hemocytes and the head and thorax region of H. virescens final-instar larvae 48 h postoviposition, were prepared and screened according to protocols reported in detail in previous papers (18, 63).
A cDNA library was constructed from total RNA isolated from M. sexta fifth-instar larvae fat body 24 h postoviposition, with a SMART cDNA library construction kit (Clontech) according to the manufacturer's instructions. The library was screened with the CcBV PTPI and PTPM probes with standard methods (50). Individual positive plaques from the secondary screening were isolated in agar plugs, eluted in SM buffer, and then converted into a plasmid (pTriplEx2) in Escherichia coli (BM25.8) by in vivo excision and circularization. DNA sequencing reactions of selected clones were carried out according to the manufacturer's instructions on an ABI Prism 3100 Avant sequencer (Applied Biosystems). The DNA and protein sequences were subjected to computer BLAST analysis (2).
PTPA and PTPN cDNAs were isolated by reverse transcription-PCR from mRNA extracted from the fat body and hemocytes of M. sexta fifth-instar larvae 24 h postoviposition, with Omniscript reverse transcriptase (Qiagen) and high-fidelity ExTaq (Takara). The primers were designed from the annotated genomic sequence to amplify the whole predicted coding DNA sequences (PTPIatg, TGACGTAGATGTCAAATAAGTG; PTPIstop, CGGAGTCTGATCGATTAAACTC; PTPMatg, GCGGACGTCCACATATTGTCTA; and PTPMstop, TGAGTCCCTAGCCGTAAAATGA)
Field inversion gel electrophoresis and Southern blot mapping. Undigested CcBV or TnBV DNA (250 ng) was separated by electrophoresis (with 1% pulse field certified agarose gels; Bio-Rad) on a field inversion gel electrophoresis apparatus (Bio-Rad), delivering alternative 120-V forward and 180-V reverse pulses for 20 h. The duration of the pulses ranged from 0.4 s at the beginning to 0.7 s at the end of the run. After migration, the DNA was transferred to a nylon membrane (50) and hybridized to 32P-labeled protein tyrosine phosphatase-specific probes. The probes were PCR amplification products obtained with CcBV virus DNA and the primers indicated in Table 1, purified with the Qiaquik kit purification (Qiagen), or TnBV protein tyrosine phosphatase cDNAs.
|
View this table: [in a new window] |
TABLE 1. Primersa
|
Membranes were prehybridized in Church buffer (46) at 65°C for 3 h and washed in 2x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-0.1% sodium dodecyl sulfate (SDS) twice for 15 min and in 0.2x SSC-0.1% SDS twice for 15 min at 65°C.
Reverse transcription multiplex PCR. Polyadenylated mRNA purified from M. sexta larvae was reverse transcribed (Omniscript reverse transcriptase; Qiagen) and reverse transcription multiplex PCRs (Qiagen) were performed following the manufacturer's instructions. The 22 CcBV protein tyrosine phosphatase products were amplified in four different reactions with the primers listed in Table 1, and single-stranded cDNAs were obtained from 15 ng of polyadenylated mRNA. PCRs were also performed on 15 ng of mRNA (without reverse transcription) to exclude viral DNA contamination. The multiplex PCR program included a denaturation step (95°C for 15 min), followed by 30 amplification cycles (denaturation at 94°C for 30 s, annealing at 57°C for 90 s, and extension at 72°C for 90 s) and a final extension step (72°C for 10 min).
Northern blot analysis. Northern blot analyses were performed with polyadenylated RNA with standard techniques (50) and separated on a formaldehyde gel. The RNA samples used in the experiments were extracted from hemocytes of H. virescens fifth-instar larvae at different times after T. nigriceps parasitization and from synchronous nonparasitized controls.
Phylogenetic analysis. Alignments of bracovirus protein tyrosine phosphatases with human and insect protein tyrosine phosphatase domains were performed with ClustalX (59). The alignment optimization was guided by the localization of the motifs described in vertebrate protein tyrosine phosphatase domains (3). Distance and parsimony analyses were performed with the PAUP4 program (56). The trees were checked by bootstrap analysis (20). Maximum-likelihood analysis was performed with Tree-puzzle 5.1 (52).
Construction of recombinant baculoviruses expressing CcBV protein tyrosine phosphatases. Recombinant baculoviruses were generated from a novel Autographa californica nucleopolyhedrovirus engineered genome, maintained as a bacterial artificial chromosome, which facilitates the rapid creation of recombinant baculovirus expression vectors (30). We used a second generation of the reported engineered genome, a polyhedrin-positive variant of bAcGOZA called bApGOZA, in which the polyhedrin gene is expressed from the p10 locus (Y. H. Je et al., unpublished data). Briefly, bApGOZA DNA purified from bacteria was cotransfected into Spodoptera frugiperda IPLB-SF21 (Sf 21) cells (64) with the polyhedrin-based transfer vector pBacPAK8 (Clontech), in which the PTPA or PTPM gene from CcBV was cloned (pBacPAK8-PTPA or pBacPAK8-PTPM). The transfection was performed with DOTAP liposomal transfection reagent (Roche), according to the manufacturer's protocol. After 6 days, the culture medium containing the resulting viral progeny was collected. Recombinants were subjected to one round of plaque purification as previously described (45). Viruses were propagated to passage number 3 in monolayers of Sf21 cells, grown at 28°C in TC-100 (Gibco-BRL) containing 10% fetal bovine serum (M. B. Meldrum).
Analysis of recombinant baculovirus mRNA expression. Sf21 cells seeded in 60-mm dishes (2 x 106 cells per dish) were infected at a multiplicity of infection of 10 PFU/cell with wild-type or recombinant baculoviruses expressing CcBV PTPA (Rec-PTPA) or CcBV PTPM (Rec-PTPM). Three days postinfection, cells were harvested (45) and resuspended in RLT buffer (Qiagen) as per the manufacturer's instructions. The samples were homogenized by loading them into a QIAshredder column (Qiagen) and spinning for 2 min at 14,000 rpm. Total RNA extraction was performed from these lysates with RNeasy (Qiagen) according to the manufacturer's protocol.
For reverse transcription-PCR, contaminant DNA was removed by digestion of eluted RNA with 1 U of RNase-free DNase RQ1 (Promega) for 10 min at 37°C. Reverse transcription was carried out according to the manufacturer's instructions (Promega), and amplification was done with primer pairs ARTF (5'-ACTGGCCCGACAACAGTATC-3') and ARTR (5'-CTTGACAGATTTTTGGCTTC-3') to detect PTPA transcripts and RTMF (5'-TGTTGTACATGGCAATGCTGG-3') and RTMR (5'-TGAGTCCCTAG CCGTAAAATG-3') to detect PTPM transcripts. PCR conditions were 2 min at 94°C, followed by 35 cycles of 15 s at 94°C, 15 s at 45°C, and 45 s at 72°C.
Characterization of recombinant gene expression by SDS-PAGE. Sf21 cells seeded in 60-mm dishes (2 x 106) were infected with the viral stocks with an multiplicity of infection of 10 PFU/cell. Samples were collected at 24, 48, and 72 h postinfection. Cell monolayers were harvested by resuspension in the supernatant medium and recovered by low-speed centrifugation. The pellet was rinsed with phosphate-buffered saline twice and resuspended in 50 µl of NP-40 lysis buffer (45). Following incubation on ice for 30 min, one volume of 1x SDS-polyacrylamide gel electrophoresis (PAGE) loading buffer was added to the samples, and 7.5 µl of each sample was separated by SDS-PAGE (12% polyacrylamide gel) and stained with Coomassie brilliant blue. For metabolic labeling of proteins, cultures were methionine starved for 1 h, pulse labeled for 1 h with 25 µCi of Trans35S label (ICN Biomedicals, Inc.), and resuspended in 500 µl of medium without methionine. Samples were collected, resolved by SDS-PAGE (12% polyacrylamide gel), and stained with Coomassie brilliant blue. The dried gels were exposed to X-ray film.
Protein tyrosine phosphatase assay. We infected 5 x 106 Sf21 cells with the viral stocks with a multiplicity of infection of 10. They were harvested at 5 days postinfection by scraping into the culture medium and recovered by centrifugation. The pellets were lysed with 150 µl of Tris-buffered saline (50) supplemented with 1% NP-40, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, and protease inhibitor cocktail (BD Biosciences Pharmingen). The samples were incubated for 30 min on ice and homogenized by loading them into QIAshredder columns. The lysates were brought to 2.5 ml with Tris-buffered saline supplemented with 1 mM dithiothreitol and 1 mM phenylmethylsulfonyl fluoride, precleared to remove endogenous free phosphates by loading them into PD-10 desalting columns containing Sephadex G-25 (Amersham Biosciences).
The measurement of protein tyrosine phosphatase activity was performed with the tyrosine phosphatase assay system (Promega), according to the manufacturer's instructions, in 96-well plates with a flat bottom. Briefly, 85 µl of the cleared lysates was incubated with Tyr phosphopeptide 1, END(pY)INASL (0.13 mM), in the presence and absence of 1 mM sodium vanadate. The reaction buffer used was Tris-buffered saline supplemented with 1 mM dithiothreitol and 1 mM phenylmethylsulfonyl fluoride. After 30 min at 30°C, the enzymatic reaction was stopped with 115 µl of a molybdate dye-additive mixture, which was used for visualization of liberated phosphate by the formation of a colored complex. The optical density of the samples was read at 600 to 650 nm range with a Metertech
960 plate reader; background absorbance due to the reaction mix was subtracted from the reported values and converted to a concentration. Two to five independent phosphatase assays (from independent infections) for each infected cell lysate were analyzed statistically by a paired Student t test (Microsoft), and data were graphed with the standard error of the mean with Microsoft Excel.
Accession numbers. The CcBV and TnBV protein tyrosine phosphatase sequences have been deposited in GenBank under accession numbers AJ634653 to AJ634660 and AJ640087 to AJ640113, respectively.
|
|
|---|
Protein tyrosine phosphatase genes were found on seven segments of the CcBV genome (17a) and on 38 out of 194 genomic overlapping TnBV clones. They constituted the major gene family of the CcBV genome in terms of number of genes (17a). Unlike many CcBV and TnBV genes, protein tyrosine phosphatase genes are not predicted to contain introns. This was confirmed by the isolation of four cDNAs from CcBV (PTPA, PTPI, PTPM, and PTPN) and 2 cDNAs from TnBV (PTP5 and PTP7). The organization of the different putative proteins was similar and resembled that of vertebrate PTP1B (3) with a single domain not associated with other conserved protein domains. In the coding DNA sequences encoding a truncated product (CcBV PTP
, CcBV PTPD, TnBV PTP12, and TnBV PTP13), the premature interruption was due to a frameshift or to the presence of a stop codon. The missing part of the protein tyrosine phosphatase domain was found downstream of the stop codon, suggesting that a single mutation occurred recently. In contrast, two CcBV genes displayed more complex rearrangements, suggesting that they constitute pseudogenes (CcBV PTPT and CcBV PTP
).
Comparison and relationships between bracovirus, invertebrate, and vertebrate protein tyrosine phosphatases. The amino acid sequences of bracovirus protein tyrosine phosphatases were deduced from the gene sequences by assuming that the cDNAs and the genes have the same sequence as predicted by Genscan analysis (http://genes.mit.edu/GENSCAN.html), as is the case for the six protein tyrosine phosphatase genes for which the cDNA was isolated. Bracovirus sequences were aligned to the profile of the protein tyrosine phosphatase domain of vertebrates (http:/science.novonordisk.com/ptp) and to several insect protein tyrosine phosphatase domains with the ClustalX program.
The 10 conserved motifs that define the protein tyrosine phosphatase family (3) were identified in bracovirus protein tyrosine phosphatases (Fig. 1). Among these motifs, the protein tyrosine phosphatase loop, which has a cysteine residue in a conserved position, has been shown by mutational analysis to be critical for protein tyrosine phosphatase activity (22, 62). Interestingly, some of the bracovirus protein tyrosine phosphatases lack this cysteine residue. Moreover, several bracovirus protein tyrosine phosphatases lack other less conserved motifs (see Fig. 1).
![]() ![]() View larger version (209K): [in a new window] |
FIG. 1. Sequence comparison of bracovirus protein tyrosine phosphatases with human and insect protein tyrosine phosphatase domains. The localization of alpha-helices and beta-strands based on the X-ray crystal structure of PTP1B are shown above the alignment (4). The 22 invariant residues (underscored) and the 42 highly conserved residues (>80% identity) of vertebrate protein tyrosine phosphatase D1 domains are indicated at the top of the alignment with a brief description of the function of the motif (3). Proteins: human HsMEG2, HsPTP1B, HsPTPgD1 and HsPTPgD2 (for the gamma D1 and D2 domains, respectively) (GenBank accession numbers M83738, M33689, and L09247, respectively); AmPTP, Apis mellifera protein tyrosine phosphatase domain (sequence 2044722913BCM from the Apis mellifera genome sequence); DmPTP and AgPTP, protein tyrosine phosphatase domains from Drosophila melanogaster and Anopheles gambiae protein tyrosine phosphatases, respectively (EMBLAE003447 and GenBank XM322055, respectively). Bracovirus protein tyrosine phosphatases: CcPTPs, TnPTPs, CgPTP, and GiPTP, bracovirus proteins from C. congregata, T. nigriceps, Cotesia glomerata (GenBank AY466396), and Glyptapanteles indiensis (GenBank AAP37630), respectively.
|
, CcBV PTPE and PTPX, and TnBV PTP5 and PTP7, showing 96%, 94%, and 84% amino acid similarity, respectively). Bracovirus protein tyrosine phosphatases are more divergent than vertebrate protein tyrosine phosphatase domains (http://www.univ-tours.fr/irbi/ptp) having different subcellular localizations and biological functions (3). Bracoviruses were originally described as unusual baculoviruses because of similarities in their particle morphology (36). In order to study the relationship between bracovirus and baculovirus protein tyrosine phosphatases, comparisons were also performed between protein tyrosine phosphatases from bracovirus and other viruses (data not shown). Poxvirus and baculovirus protein tyrosine phosphatases are much shorter and belong to the dual-specificity protein tyrosine phosphatases. As generally observed when comparing classical and dual-specificity protein tyrosine phosphatases, some homology was detected only in the phosphatase catalytic sites (protein tyrosine phosphatase loop).
The phylogenetic relationship between bracovirus, vertebrate, and invertebrate protein tyrosine phosphatases was analyzed with the methods of distance, parsimony, and maximum likelihood. Due to the high divergence of the bracovirus sequences, it was not possible to reconstruct a complete history of their phylogeny (the trees were not completely resolved) or to demonstrate that bracovirus proteins originated from a particular type of invertebrate or vertebrate protein tyrosine phosphatase (data not shown). However, several clades of protein tyrosine phosphatase genes were consistently obtained with the different methods with significant bootstrap values (Fig. 2), indicating that the corresponding genes originated from a common ancestor, probably through several rounds of gene duplication.
![]() View larger version (30K): [in a new window] |
FIG. 2. Phylogenetic analysis of bracovirus protein tyrosine phosphatases. The neighbor-joining tree and one of the most parsimonious trees (parsimony) were generated with PAUP4 from the alignment shown in Fig. 1. (The more divergent sequences [TnBV PTP2, PTP 4, and PTP 6] were not included in the analysis.) The trees were rooted with human and insect protein tyrosine phosphatases of the MEG2 subtype (Hs, Homo sapiens; Dm, Drosophila melanogaster; Ag, Aedes aegypti) as outgroups. Bootstrap values of >50% are indicated. The circles designate the most internal nodes supported by bootstrap values of >75% which define six monophyletic groups of bracovirus protein tyrosine phosphatases consistently found with different phylogenetic methods. On the parsimony tree for the CcBV protein tyrosine phosphatases, the PDV segments containing the corresponding genes are indicated in brackets (C1 to C26).
|
![]() View larger version (52K): [in a new window] |
FIG. 3. Hybridization of CcBV and TnBV protein tyrosine phosphatase probes to DNAs (250 ng) extracted from viral particles separated by field inversion gel electrophoresis (exposure time, 16 h). Ethidium bromide-stained CcBV and TnBV DNA segments are visualized in lanes Cc and Tn, respectively (the average size of TnBV segments is much smaller than that of CcBV segments). The sizes of a set of double-stranded DNA segments determined from genome assembly and identified by hybridization of non-protein tyrosine phosphatase probes are indicated on the left (circular DNA sizes), and a linear size marker (2.5 kb ladder from Bio-Rad) is shown (lane MW). Seven probes were used to hybridize to CcBV DNA (lanes A, N, X, O, R, I, and W hybridized with the CcBV PTPA, PTPN, PTPX, PTPO, PTPR, PTPI, and PTPW probes, respectively), and one probe was used to hybridize to TnBV DNA (lane 7, TnBV PTP7). A major signal was obtained with each CcBV protein tyrosine phosphatase probe, corresponding to a molecule of the expected size (indicated below each lane in base pairs with the name of the corresponding CcBV segment); the intensity of the signal varied according to the relative abundance of the different segments in the viral DNA and the specific activity of the probe. The upper signals visualized in lanes N and I correspond to the linear sizes of segments C10 and C1, suggesting that a small fraction of the molecules were damaged during viral DNA extraction, as previously observed for the EP1 segment (46). Two signals were obtained with the TnBV probe, one corresponding to hybridization with the segment containing PTP7, and the other corresponding to hybridization with another segment harboring the homologous PTP5.
|
Protein tyrosine phosphatase gene expression in parasitized host tissues. The occurrence of pseudogenes in bracovirus protein tyrosine phosphatase gene families and of potential proteins lacking the catalytic cysteine residue in the protein tyrosine phosphatase loop suggests that only a subset of protein tyrosine phosphatases might be involved in the regulation of host-parasite interactions. Detecting the expression of different protein tyrosine phosphatase genes during parasitism could constitute a first indication of their involvement in bracovirus-induced alterations of host physiology. To obtain a large picture of CcBV protein tyrosine phosphatase gene expression during parasitism, a multiplex PCR approach was chosen (72) with primers designed to amplify specific regions of the cDNAs of 22 protein tyrosine phosphatase genes. To validate the technique, reverse transcription multiplex PCR analysis was first performed with purified viral DNA; PCR products of the expected sizes were obtained for the 22 genes tested (Fig. 4).
![]() View larger version (48K): [in a new window] |
FIG. 4. Analysis of CcBV protein tyrosine phosphatase gene expression in tissues of parasitized M. sexta larvae with reverse transcription multiplex PCR (see Table 2 for a summary of the results). For CcBV, the ethidium bromide-stained electrophoresis gel of PCR products obtained with purified viral DNA amplified in four separate reactions (1 to 4) (the primers used are listed in Table 1) is shown. On the right are shown the positions of the products corresponding to the different genes analyzed. For the nervous system, midgut, Malpighian tubules, fat body, and hemocytes, the ethidium bromide-stained gel electrophoresis of PCR products obtained with cDNA extracted from the tissues of parasitized M. sexta larvae dissected either 24 h postoviposition (hemocytes and fat body) or 12 h postoviposition (other tissues) is shown. Lanes C, amplification performed to assess viral DNA contamination in the mRNA sample (without reverse transcription).
|
For each gene analyzed, a product of the expected size was obtained with mRNAs extracted from one or several tissues (Fig. 4). The number of protein tyrosine phosphatase mRNAs detected ranged from nine, in mRNA extracted from Malpighian tubules, to 20, in the nervous system (Table 2). Several genes displayed a ubiquitous expression profile (CcBV PTPA, PTPK, PTPI, PTPL, PTPM, PTPP, PTPR, PTPS, and PTPX), while expression of the other genes was restricted to different subsets of the tissues analyzed (Table 2).
|
View this table: [in a new window] |
TABLE 2. Protein tyrosine phosphatase mRNA expression in parasitized M. sexta larvae and in the parasitoid was C. congregata
|
![]() View larger version (111K): [in a new window] |
FIG. 5. Northern blot analysis of TnBV protein tyrosine phosphatase mRNA expression in parasitized H. virescens larvae. Upper panel: hybridization of PTP7 cDNA to RNA extracted from H. virescens hemocytes collected at different times (from 3 to 48 h) following T. nigriceps parasitization and from nonparasitized control larvae (NP). Lower panel: control of the amount and quality of the RNA samples by hybridization of a cDNA of H. virescens, coding for a putative protein showing high similarity with M. sexta ribosomal protein S3 (63), as a probe.
|
Expression of CcBV PTPA and PTPM in baculovirus-infected insect cells. To evaluate the potential role of bracovirus protein tyrosine phosphatases in parasitism, we had to determine if they displayed protein tyrosine phosphatase activity. To assess the biochemical activity of bracovirus protein tyrosine phosphatases, a baculovirus expression system was used to generate two different CcBV protein tyrosine phosphatases, one with a regular protein tyrosine phosphatase loop (CcBV PTPA) and the other without the cysteine in the loop (CcBV PTPM). Recombinant baculoviruses Rec-PTPA and Rec-PTPM were produced, with the genes expressed under the control of the very late baculovirus polh promoter. After infection of Sf21 cells, expression of each protein tyrosine phosphatase was checked by reverse transcription-PCR. Specific bands were obtained for PTPA (380 bp) and PTPM (259 bp) after Rec-PTPA and Rec-PTPM infections, but not with wild-type infections (Fig. 6A and C). Metabolic labeling of proteins produced during baculovirus infection of Sf21 cells and subsequent SDS-PAGE analysis demonstrated the production of a novel protein following Rec-PTPA and Rec-PTPM infections at very late times (Fig. 6B and D). The migrations of the baculovirus-expressed PTPA and PTPM were consistent with their predicted sizes (37.3 and 34.5 kDa, respectively).
![]() View larger version (31K): [in a new window] |
FIG. 6. Expression of CcBV PTPA and PTPM from baculovirus infections and screen for protein tyrosine phosphatase activity. Sf21 insect cells were infected with wild-type A. californica nucleopolyhedrovirus (lane WT) or recombinant baculovirus Rec-PTPA or Rec-PTPM. Infected cell RNA was collected and subjected to reverse transcription-PCR, and the products were analyzed by electrophoresis on 1.5% agarose gels (A and C). PTPA- and PTPM-specific bands (A and C, respectively) are indicated with arrows. DNA markers (lane M), template-free reaction (lane ), and reaction lacking reverse transcriptase (lane no RT) are indicated. Panel B shows protein metabolic labeling and SDS-PAGE of proteins from Rec-PTPA-infected cells, and panel D shows SDS-PAGE and Coomassie brilliant blue staining of proteins from Rec-PTPM-infected cells at various times postinfection (hours). Pol, A. californica nucleopolyhedrovirus very late structural protein. Infected cell lysates (E), precleared of free phosphate, were exposed to a synthetic tyrosine phosphopeptide (Promega), and liberated phosphate was measured spectrophotometrically at 600 nm. +I indicates the presence of a protein tyrosine phosphatase inhibitor. The asterisk denotes significantly more protein tyrosine phosphatase activity (P = 0.02) compared to wild-type-infected extracts.
|
As shown in Fig. 6E, wild-type virus-infected cell extracts showed some protein tyrosine phosphatase activity, as previously reported (26, 34, 53) and as expected from the presence of dual-specificity protein tyrosine phosphatases in the baculovirus. However, the Rec-PTPA virus-infected cell lysates released significantly more free phosphate than wild-type virus-infected cell lysates (P = 0.02). In contrast, the Rec-PTPM virus-infected cell lysates generated the same amount of free phosphate as wild-type-infected cells, indicating that PTPM overexpression does not lead to additional protein tyrosine phosphatase activity in baculovirus-infected cells. Incubation of either type of infected cell lysate with the general tyrosine phosphatase inhibitor sodium vanadate resulted in a decrease in the amount of free phosphate generated. Overall, our results indicate that CcBV PTPA is a functional tyrosine phosphatase, while CcBV PTPM might have a different activity.
|
|
|---|
It would be of great interest to determine if protein tyrosine phosphatase genes are also encoded by bracoviruses from other braconid subfamilies of the monophyletic microgastroid complex (66). Conceivably, some protein tyrosine phosphatase genes might have been present in the gene set of the nonintegrated ancestor of the bracoviruses, and indeed viruses are known to pick up cellular genes, including protein tyrosine phosphatases, that are beneficial for their life cycle in infected hosts (29). The structure of the bracovirus protein tyrosine phosphatases, with 10 motifs conserved in classical protein tyrosine phosphatases, strongly suggests that they are indeed of cellular origin.
Based on morphological similarities between bracovirus and baculovirus particles, it was proposed that the ancestral bracovirus was a baculovirus (15, 19). However, bracovirus and baculovirus protein tyrosine phosphatases are unrelated; baculoviruses contain genes for dual-specificity protein tyrosine phosphatases (dephosphorylating tyrosine and serine residues) (26). Thus, the identification of the bracovirus protein tyrosine phosphatases does not support the hypothesis of a baculovirus origin of bracoviruses.
The considerable divergence of bracovirus protein tyrosine phosphatase sequences might be explained by the fact that the two subfamilies to which the wasp species examined here belong have been separated for tens of millions of years (66) and that these genes, like other polydnavirus (PDV) genes involved in the host-parasitoid interaction, may have undergone fast evolutionary changes (16). However, the hypothesis that protein tyrosine phosphatases were acquired independently in the two lineages through a convergent evolution process cannot be rejected. Independent acquisitions from wasp chromosomes might have occurred through recombination or transposition events, resulting in the integration of wasp genes in the chromosomal form of the virus genome.
Bracovirus protein tyrosine phosphatase organization in a large gene family. A characteristic feature common to the bracovirus protein tyrosine phosphatases is their organization into large gene families. Such an organization provides an efficient mechanism for generating genetic diversity (44). For example, animal toxins are often encoded by gene families (35) and appear to have diversified by gene duplication and adaptive evolution. However, gene families are rare among free-replicating viruses compared to cellular genomes. This paucity has been attributed to the limitation in size of the genome to one that can be efficiently encapsidated in virus particles (29). Accordingly, in viruses that have large genomes, some genes, such as the bro genes of baculoviruses, are organized in gene families (29).
Polydnaviruses have some of the largest virus genomes (568 kb for CcBV) (17a), and gene families are common in both ichnoviruses (9, 23, 37, 65) and bracoviruses (18, 37, 57, 60). The organization of bracovirus genomes in large double-stranded DNA segments encapsidated in distinct nucleocapsids (1) and the fact that they are exclusively chromosomally transmitted (in the absence of virus replication within the host tissues) probably relieved constraints on viral genome size, allowing the formation of gene families. It is noteworthy that closely related genes are not necessarily encoded by the same virus segment (Fig. 3), suggesting the occurrence of duplications, likely by unequal crossing-over events, of the chromosome regions where the integrated forms of the virus sequences are clustered (10, 71). The high sequence similarity (see Results) of some protein tyrosine phosphatase genes isolated from the bracovirus suggests that they originated from recent duplication events.
The expansion of the protein tyrosine phosphatase gene family is particularly striking. Indeed, the CcBV genome contains 27 protein tyrosine phosphatase genes, while other CcBV gene families contains at most half a dozen (17a). Protein tyrosine phosphatases genes are also numerous in the TnBV genome, which contains more than 13. This suggests either that duplication of protein tyrosine phosphatase sequences is a particularly frequent phenomenon or that duplicated protein tyrosine phosphatase genes have been selectively conserved during the evolution of the bracovirus genome, probably because of the strong selection pressures associated with their role in successful parasitism.
Diversification of bracovirus protein tyrosine phosphatases. Protein tyrosine phosphatases are almost as abundant in CcBV DNA as in eukaryotic genomes (the human genome contains 37 protein tyrosine phosphatase genes), but their organization is comparatively rudimentary. Bracovirus protein tyrosine phosphatases consist essentially of a protein tyrosine phosphatase catalytic domain, while in cellular protein tyrosine phosphatases, the protein tyrosine phosphatase domain is generally associated with other conserved protein domains involved in modulating the function of the protein (3). Interestingly, dual-specificity protein tyrosine phosphatases from poxviruses and baculoviruses, including VH1 from vaccinia virus, which inhibits the interferon signaling pathway by dephosphorylating transcription factor Stat1 (42), are also composed of a single protein tyrosine phosphatase domain that might be sufficient for viral products to act as inhibitors.
In contrast to their homogeneous organization, bracovirus protein tyrosine phosphatases show considerable diversity in their amino acid sequences; they are much more divergent than the protein tyrosine phosphatase domains of vertebrate proteins with different biological functions. Protein tyrosine phosphatases are a key group of signal transduction enzymes that, together with protein tyrosine kinase, control the levels of cellular protein phosphorylation (43), playing a pivotal role in cellular signaling. Each protein tyrosine phosphatase dephosphorylate phosphotyrosine residues on a specific substrate. Since the protein tyrosine phosphatase domain itself confers the target specificity of protein tyrosine phosphatases (3), the bracovirus protein tyrosine phosphatases might interact with distinct host target proteins to interfere with different signaling processes of the host. Furthermore, as some wasps parasitize several host species (28), the diversity of bracovirus protein tyrosine phosphatases might also mirror the diversity of molecular targets in different hosts. Bracovirus protein tyrosine phosphatase diversification might have thus contributed to wasp adaptive radiation, allowing the colonization of new ecological niches. It might have also helped parasitoids to deal with new resistance genes arising from the evolutionary arms race between host and parasites, as suggested for other PDV genes (16).
Potential role of bracovirus protein tyrosine phosphatases during parasitism. The analysis of protein tyrosine phosphatase gene expression in parasitized hosts suggests that most protein tyrosine phosphatases have a function in contemporary host-parasitoid interactions. Indeed, the mRNAs of the 22 CcBV protein tyrosine phosphatase genes tested were detected in tissues of the parasitized host larvae with reverse transcription multiplex PCR. Similarly, a protein tyrosine phosphatase gene characterized recently from another wasp of the Microgastrinae subfamily (a gene from Glyptatanteles indiensis resembling CcBV PTPQ) is ubiquitously expressed in parasitized hosts (12, 13). TnBV protein tyrosine phosphatase expression was also detected during the host-parasite interaction, the expression of TnBV PTP5 and/or PTP7 being observed by Northern blot analysis in the hemocytes and thorax of the parasitized host. Preliminary experiments suggest that some protein tyrosine phosphatases are also expressed in adult wasps, and more analyses are currently being performed to determine if expression is restricted to particular tissues, such as ovaries, where virus particles are produced.
Surprisingly, the protein tyrosine phosphatase genes do not necessarily encode a product with protein phosphatase activity. Indeed, several CcBV protein tyrosine phosphatases (CcBV PTPM, PTPS, PTPH, PTPD, PTPU, PTPO, PTPE, and PTPX) are mutated in the protein tyrosine phosphatase loop at a site critical for activity (the catalytic center cysteine residue). Only 15 out of 27 CcBV protein tyrosine phosphatases have a regular catalytic site. To determine if bracovirus protein tyrosine phosphatases actually comprised both functional and inactive phosphatases, a protein of each type (with a regular or mutated protein tyrosine phosphatase loop) was produced with the baculovirus expression system. Interestingly, protein tyrosine phosphatase activity was characterized from baculoviruses producing PTPA (with a regular protein tyrosine phosphatase loop) and not from those producing PTPM (mutated in the protein tyrosine phosphatase loop). The fact that CcBV PTPA is active as a protein tyrosine phosphatase strongly suggests that bracovirus protein tyrosine phosphatases have a biological function during host-parasite interactions. A protein tyrosine phosphatase without phosphatase activity might have a different biochemical function, such as trapping phosphorylated tyrosine proteins, as suggested by the example of the D2 domain of the human protein tyrosine phosphatase CD45, which is similarly altered in the protein tyrosine phosphatase loop and involved in CD45 substrate (T-cell receptor
) recruitment (33).
The diversity of the bracovirus protein tyrosine phosphatases, together with the exquisite specificity of characterized protein tyrosine phosphatases for their substrates conferred by the protein tyrosine phosphatase domain, theoretically provides the parasitoid with the opportunity to interfere with multiple pathways, in particular with those controlling host development and immunity.
The prothoracic glands of mature H. virescens larvae show a dramatic reduction in biosynthetic activity when parasitized by T. nigriceps, and this alteration is reproduced by TnBV infection (40, 48). The reduced ecdysone biosynthesis is associated to the underphosphorylation of key regulatory proteins of the prothoracicotropic hormone signal transduction pathway, which culminates with a translational block of host protein synthesis (48). The possibility that this alteration is induced by TnBV-encoded protein tyrosine phosphatases remains to be experimentally tested. Because prothoracicotropic hormone stimulates a number of rapid protein phosphorylations in the prothoracic glands via protein kinase A, mitogen-activated protein kinases, and other uncharacterized kinases (24), it is reasonable to speculate that these pathways could be targeted by virulence factors such as protein tyrosine phosphatases. This might also prove to be true in other host-parasitoid systems, such as M. sexta parasitized by C. congregata (54), that become refractory to prothoracicotropic hormone stimulation (6, 7) even though prothoracic gland degeneration is not observed (5).
Bracovirus protein tyrosine phosphatases might also be able to disrupt the host signal transduction pathways controlling hemocyte cytoskeleton dynamics during encapsulation and phagocytosis, mimicking bacterial strategies of phagocytosis inhibition (17). Indeed, certain bacterial pathogens use protein tyrosine phosphatases as virulence factors to interfere directly with the host signal transduction pathway, mediated by ß1-integrin receptors, which control actin rearrangements. In particular, the protein tyrosine phosphatase virulence factor YopH and the cytotoxin YopE of Yersinia pestis (the agent that causes bubonic plague), which are injected by the type III secretion system into human macrophages, have dramatic effects on actin rearrangements (14), resulting in inhibition of phagocytosis (11). ß-Integrin subunits were recently characterized in Pseudoplusia includens, strongly suggesting that this signaling pathway leading to actin rearrangements is conserved in lepidopteran hosts and might thus constitute a target for bracovirus protein tyrosine phosphatases (38).
Future work will be directed toward characterization of the molecular targets of the different bracovirus protein tyrosine phosphatases to identify which host signaling pathways they effectively alter and thus to determine how they are involved in the alteration of host immunity and development.
We thank J. Dérisson and C. Ménoret for their involvement in M. sexta and C. congregata rearing in Tours and E. T. Caprioli for H. virescens and T. nigriceps rearing in Potenza.
Present address: Institut de Génétique et Microbiologie, UMR CNRS 8621, Université Paris Sud, Orsay Cedex, France. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»