vLIP, a Viral Lipase Homologue, Is a Virulence Factor of Marek's Disease Virus

Sequence analysis and secondary-structure prediction.

BLAST and PSI-BLAST analyses of vLIP protein sequences were performed using National Center for Biotechnology Information servers (6, 7). Prediction of vLIP protein structure was performed using the GenTHREADER program via the PSIPRED server (45). For the prediction of signal peptides and signal peptide cleavage sites, the SignalP prediction server (version 1.1) was used (57, 58). Routine amino acid alignments were performed using MacVector 7.2.2 software (Accelrys, San Diego, CA) using default settings.

Virus and cells.

The MDCC MSB-1 tumor cell line harboring latent MDV (4) was maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) at 37°C. Lytic virus replication was induced by the addition of sodium butryate to a final concentration of 3 mM. In some experiments, 0.5 mM foscarnet (trisodium phosphonoformate hexahydrate [PFA]; Sigma-Aldrich, St. Louis, MO) was used at the time of induction to determine kinetic classes of transcripts. For the animal study and for growth of lytic MDV in culture, bacterial artificial chromosome (BAC) clones derived from the RB-1B strain of MDV (pRB-1B) were reconstituted by standard calcium phosphate transfection (72) and were passaged no more than three times for in vivo studies and no more than five times for in vitro studies. Transfection and propagation of pRB-1B-based parental, mutant, and revertant MDVs were performed on secondary chicken embryo cells (CEC), which were cultivated in a 1:1 mixture of EMEM and HMEM supplemented with 2 to 10% FBS (72).

Trichoplusia ni insect cells, used for the expression of recombinant vLIP, were grown in 250-ml to 1-liter polycarbonate Erlenmeyer flasks (Nalgene) at 140 rpm in shaker culture at 28°C in ExCell 405 medium (JRH Biosciences, Lenexa, KS). Spodoptera frugiperda (Sf9) insect cells were grown in a similar manner to the T. ni cells, except that ExCell 420 medium (JRH Biosciences) was used.

Cloning and characterization of vLIP mRNA.

Total RNA was isolated from 1 × 10⁸ MSB-1 cells using Trizol (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. For cDNA synthesis, 20 μg of total RNA was used for first-strand synthesis primed with the SS-dT oligonucleotide (Table 1). Reverse transcription (RT) was carried out using avian myeloblastosis virus reverse transcriptase (Roche, Indianapolis, IN) at 42°C for 1 h using the conditions specified by the manufacturer. Upon completion of RT, reactions were extracted against phenol-chloroform and passed over Chromaspin TE 200 columns (Stratagene Inc., La Jolla, CA).

Primers (200 nM each) used in rapid amplification of 3′ cDNA ends were SS and sp.LIP5′Cpo (Table 1). Thirty PCR cycles (93°C for 45 s, 56°C for 60 s, and 68°C for 300 s) were performed, using Expand High-Fidelity Taq polymerase (Roche) at 2.3 U (0.7 μl) per 50-μl reaction volume. PCR products were cloned into the pCR2.1-TopoTA vector (Invitrogen). The LIP.RC(AS) and sp.LIP5′CPO primers were used to screen bacterial colonies for plasmids with spliced vLIP cDNA inserts (Table 1). DNA sequencing was performed using vector-specific primers, along with several vLIP-specific primers designed from MDV genomic DNA sequences (Davis Sequencing LLC, Davis, CA). All oligonucleotides were custom synthesized (Integrated DNA Technologies, Coralville, IA).

For Northern blot analysis, 10 μg of total RNA was denatured for 15 min at 65°C, loaded onto 1% agarose-6% formaldehyde gels, and run at 75 V in 50 mM morpholinepropanesulfonic acid (MOPS) buffer (pH 7.2). Resolved RNAs were transferred overnight onto Hybond-N membranes (Amersham Biosciences Corp., Piscataway, NJ) by the capillary method in 20× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate, pH 7.0) (68). Transferred RNA was UV cross-linked using the Stratalinker apparatus (Stratagene). A ³²P-labeled 376-bp single-stranded RNA probe antisense to vLIP mRNA was transcribed in vitro from a pSP72 (Promega, Madison, WI)-based plasmid using the MaxiScript T7 kit (Ambion) after linearization of the construct. As a negative control, a sense single-stranded RNA probe corresponding to residues coding for amino acids 158 to 255 of vLIP was prepared and hybridized in parallel to a duplicate membrane. Unincorporated radioactivity was removed using NICK columns (Amersham Biosciences). Membranes were hybridized overnight at 68°C in UltraHyb solution (Ambion), and 1.5 × 10⁶ cpm of probe per ml of hybridization solution was used in each experiment. Blots were washed twice for 5 min each at room temperature in 2× SSC-0.1% sodium dodecyl sulfate (SDS) and twice for 15 min each at 68°C in 0.1× SSC-0.1% SDS; then they were exposed at −80°C to BioMax MR film (Kodak, Rochester, NY) in the presence of an intensifying screen. The Millenium RNA ladder (Ambion), in conjunction with a ³²P-labeled Millenium probe (Ambion), was used to determine RNA sizes.

Construction of plasmids.

The QuikChange kit (Stratagene) was used to introduce an S307A change at the predicted nucleophile site of vLIP. According to the manufacturer's instructions, the LipMutS and LipMutAS primers (Table 1) were used to introduce a single-base-pair change from thymidine to guanine, changing a TCC serine codon to a GCC alanine codon. For baculovirus expression, wild-type (wt) vLIP cDNA or S307A vLIP cDNA, starting at residue Val31 (a predicted signal peptide comprises residues 1 to 30) was PCR adapted with BamHI and XhoI sites at the 5′ and 3′ ends, using primers BamHILipS and XhoILipAS (Table 1), and then inserted into pBAC11 (Novagen, Madison, WI). This cloning strategy placed the N terminus of vLIP cDNA in frame with the gp64 signal peptide from pBAC11 and placed the C terminus of vLIP in frame with a polyhistidine tag also encoded by the vector. All clones were verified by DNA sequencing.

To incorporate a FLAG epitope tag (DYKDDDDK) at the carboxy terminus of vLIP, overlap extension PCR was used. Separate PCRs were performed with the following pairs of primers: LIPFLAGS-LIPFLAGS2 and LIPFLAGR-LIPF2 (Table 1). The resulting PCR products were purified by agarose gel electrophoresis and combined in an overlap extension PCR performed using the external primers (LIPFLAGS2 and LIPF2). The PCR product was then cloned into pCR2.1-TopoTA (Invitrogen) and sequenced (Cornell BioResource Center). A sequence-verified clone was digested with BsiWI and ClaI to release a FLAG-tagged fragment of the vLIP cDNA. The BsiWI-ClaI fragment was then ligated into pEco4.268, yielding pEco4.268_vLIP* (see Fig. 5). The pEco4.268 plasmid was originally constructed by inserting the EcoRI 4,268-bp subfragment of the Md5 genome derived from the sn5 cosmid (65) (a gift of Sanjay Reddy) into pUC18 (Fig. 5).

In order to construct a shuttle vector to incorporate an in-frame deletion of amino acids 256 to 428 of the vLIP open reading frame, a 519-bp fragment was released from the pEco4.268 plasmid using MluI and SpeI, and overhangs were filled in using T4 DNA polymerase (NEB). Finally, the plasmid was closed using T4 DNA ligase (NEB) and transformed into Escherichia coli. Individual plasmids bearing the deletion were sequenced, and a clone bearing an in-frame deletion was obtained and digested with the BamHI and PshAI enzymes (NEB). The 3,307-bp BamHI-PshAI fragment was inserted into pST76K-SR (1), a vector for performing RecA-based shuttle mutagenesis, which had been opened with BamHI and SmaI. To prepare a similar shuttle vector that would be used to restore the wt vLIP gene to the deletion mutant BAC, pEco4.268 was digested with BamHI and PshAI, and the 3,826-bp fragment was ligated into pST76K-SR as described above. In order to prepare a shuttle vector to transfer the S307A mutation into the pRB-1B BAC, a pCR2.1-vLIP cDNA clone bearing the S307A mutation was digested with MluI and SpeI, and the 516-bp MluI-SpeI fragment was inserted into the pEco4.268 plasmid. After confirmation of the presence of the S307A mutation by DNA sequencing, the 3,826-bp BamHI-PshAI fragment was transferred into pST76K-SR. Finally, to prepare a shuttle vector for incorporation of a FLAG-tagged wt vLIP into the pRB-1B BAC, the pEco4.268_vLIP* plasmid was opened with BamHI and PshAI, and the 3,850-bp fragment was transferred to pST76K-SR (Fig. 5).

Shuttle mutagenesis and BAC DNA preparation.

The pST76K-SR plasmid (a gift of Martin Messerle), which has been previously described (1, 11), contains a temperature-sensitive origin of replication, as well as genes encoding the bacterial recombinase RecA, the negative selection marker SacB, and kanamycin resistance. The pST76K-SR shuttle constructs described above were electroporated into DH10B E. coli harboring the pRB-1B BAC, plated on Luria Bertani (LB) agar plates containing kanamycin (50 μg/ml) and chloramphenicol (30 μg/ml), and incubated at 42°C. For the construction of a vLIP deletion mutant, minipreparations of BAC DNA from cointegrates were prepared from overnight cultures grown at 42°C, and DNA was analyzed by restriction enzyme analysis to verify cointegration of the shuttle construct. Verified cointegrates were then grown in LB broth at 37°C for 6 h in the presence of chloramphenicol but without kanamycin, so as to allow for the loss of the shuttle vector through a second RecA-mediated homologous recombination. Cultures were then streaked out on LB chloramphenicol plates containing 10% (vol/vol) sucrose to allow for negative selection by SacB. Colonies on these plates were then replica plated on kanamycin and chloramphenicol plates. Colonies that grew only on the chloramphenicol plates were screened by PCR for the presence of the deletion and then were verified by restriction enzyme analysis. To repair the vLIP deletion mutant, pST76K-SR shuttle vectors bearing the wt vLIP or the FLAG-tagged version were electroporated into DH10B E. coli harboring the pRB-1B vLIPΔMluI-SpelI BAC. Shuttle mutagenesis was performed as described above, with the exception that the PCR screening procedures were adjusted accordingly. Large-scale purification of BAC DNA was performed using a QIAGEN Maxi-prep kit, as per the manufacturer's instructions for purification of very low copy number plasmids and cosmids.

Baculovirus expression and protein purification.

Recombinant plasmid pBAC11-vLIP WT or pBAC11-vLIP S307A was cotransfected with BacVector3000 DNA into Sf9 insect cells using reagents from the BacVector kit (Novagen). Recombinant viruses were plaque purified twice and verified for expression by Western blotting with antibodies against vLIP.

For the purification of polyhistidine-tagged vLIP from supernatants of baculovirus-infected T. ni insect cells (multiplicity of infection, 1.0), TALON resin (Clontech, La Jolla, CA) was used. Briefly, supernatants were adjusted to a final concentration of 50 mM sodium phosphate (pH 7.0), 300 mM NaCl, 5 mM imizadole, and 10% glycerol and were mixed gently on a magnetic stirrer with preequilibrated TALON resin (Clontech, La Jolla, CA) for 20 min. The TALON beads were then spun down at 700 × g, washed twice with 20 volumes of TALON phosphate buffer, loaded onto a 10-ml TALON column, and washed with 5 volumes of TALON phosphate buffer. Recombinant proteins were eluted with TALON elution buffer containing 150 mM imizadole and dialyzed against HEPES-buffered saline overnight at 4°C to remove imizadole.

Immunoprecipitation and Western blot analysis.

Protein concentrations were measured using the bicinchoninic acid reagent (Pierce Biotechnology Inc., Rockford, IL). Routinely, vLIP proteins were analyzed by SDS-8% polyacrylamide gel electrophoresis (PAGE), run in a Miniprotean III device (Bio-Rad Laboratories Inc., Hercules CA). For Western blotting, resolved proteins were transferred to Immobilon-P polyvinylidene difluoride (PVDF) membranes (Millipore, Billerica, MA) using a Bio-Rad semidry transfer apparatus. PVDF membranes were blocked using 5% powdered milk in 1× Tris-buffered saline-Tween 20 (5% PM-TBST). Affinity-purified polyclonal rabbit antibodies raised against the peptide TKTTNENGHEKDSKD (amino acids 108 to 122 of vLIP) were used at a 1:1,000 dilution in 5% PM-TBST. The anti-FLAG antibody M2 (Stratagene) was purchased to detect epitope-tagged vLIP expressed from chicken cells. Horseradish peroxidase-conjugated donkey anti-rabbit immunoglobulin G (Amersham Biosciences) was employed as a secondary antibody at a 1:2,000 dilution in 5% PM-TBST, and horseradish peroxidase-conjugated goat anti-mouse immunoglobulin G (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) was used at 1:5,000 in 5% PM-TBST. ECL (Amersham Biosciences) chemiluminescent detection reagents were used according to the manufacturer's instructions to image bound secondary antibodies, and blots were exposed to BioMax MR film (Kodak). Digitized images were processed with Adobe Photoshop 7.0 software for Macintosh.

Immunoprecipitations were carried out overnight at 4°C in the presence of Complete Mini protease inhibitor (Roche) using 15 ml of cell culture supernatants of either pRB-1B- or pRB-1B-vLIP*-infected CEC. Debris was removed by centrifugation at 3,000 × g for 10 min, and 40 μl of EZview ANTI-FLAG M2 affinity gel (Sigma-Aldrich) was added. Beads were washed three times in TBST and eluted by incubation with 3× FLAG peptide (Sigma-Aldrich) as per the manufacturer's instructions. Concentration of cell culture supernatants for Western blot analysis was achieved using 20-kDa MWCO iCON devices (Pierce).

For deglycoslylation reactions, peptide N-glycosidase F enzyme (PNGase F) (NEB) was used according to the manufacturer's instructions. Briefly, purified vLIP protein was incubated first for 10 min at 100°C in 1× denaturation buffer (NEB) and then in the presence of PNGase F enzyme for 1 h at 37°C in 1× G7 buffer (NEB) containing NP-40.

Phospholipase assays.

Phospholipase assays on [¹⁴C]dioleoylphosphatidylcholine ([1,2-¹⁴C]oleate) (New England Nuclear, Boston, MA) were performed as follows. Unlabeled dioleoylphosphatidylcholine (160 μg; Avanti Polar Lipids, Alabaster, AL) was combined with 40 μg of unlabeled phosphatidic acid and 5 μCi of [¹⁴C]dioleoylphosphatidylcholine, and organic storage solvents were dried off under a stream of N₂. Four hundred microliters of lipase assay buffer (100 mM Tris-HCl, pH 8.0, 150 mM NaCl, 2 mM CaCl₂) was added to the dry lipid film. Lipids were sonicated, and 40 μl of lipid micelles was transferred to microcentrifuge tubes on ice. Assays were initiated by adding 10 μl containing 10 ng of either recombinant purified wt vLIP protein, S307A vLIP protein, or chicken K60 chemokine (baculovirus expressed) as a negative control to the 40 μl of assay solution. Bee venom phospholipase A₂ (Sigma-Aldrich) (10 ng) was used as a positive control. Reaction mixtures were incubated at 37°C for 1 h, and reactions were stopped by the addition of 187.5 μl of MeOH:CHCl₃ (3:1). Then 62.5 μl of 0.5 M HCl and 62.5 μl of CHCl₃ were added, and reaction products were vortexed. Centrifugation at 2,300 × g was applied to separate phases, and 100 μl of the organic phase was removed, dried under a stream of N₂, and resuspended in 10 μl of developing solvent (90:10:10, CHCl₃:MeOH:CH₃COOH) (8). Lipids were then loaded onto activated silica gel 60A TLC plates (Sigma-Aldrich) and developed in a preequilibrated TLC chamber at room temperature. After the solvent front reached within 1.5 cm of the top of the plate, the plate was removed, allowed to air dry, and then exposed to BioMax MR film (Kodak).

Triacylglycerol lipase assays.

Lipase assays on the radiolabeled triacylglyceride substrate [³H]triolein were performed essentially as described for the measurement of lipoprotein lipase activity by following the protocol of Briquet-Laugier et al. (16) using 12.8 μCi of freshly purified glycerol tri[9,10(n)-³H]oleate (Amersham). Radioactive lipid was mixed with 75 μl of a 200-mg/ml solution of cold triolein (Nu-Chek-Prep, Inc. Elysian, MN) and 10 μl of a 100-mg/ml solution of lecithin, and the mixture was dried under a stream of N₂. Then 2.2 ml of 0.2 M Tris-Cl, pH 8.0, and 0.1 ml of 10% Pentex bovine serum albumin (Serologicals, Norcross, GA) were added to the dry lipid film and sonicated. Then 0.45 ml of 10% bovine serum albumin and 0.25 ml of heat-inactivated rat serum (a source of apolipoprotein-CII) were added to the substrate and mixed. Assays were performed in silica glass test tubes (Fisher Scientific, Pittsburgh, PA) and were initiated by adding recombinant protein to 100 μl of the substrate. Samples were incubated for 60 min at 37°C, and reactions were stopped by the addition of 3.25 ml of MeOH:CHCl₃:heptane (1.41:1.25:1, vol/vol/vol). Liberated fatty acids were extracted into the aqueous phase by the addition of 1 ml of 0.1 M potassium carbonate-borate buffer (pH 10.5) followed by vigorous vortexing for 15 seconds. The tubes were then centrifuged at 500 × g for 10 min at room temperature to separate the organic and aqueous phases. Finally, 1 ml of the aqueous phase was removed from each tube, mixed with 10 ml of scintillation fluid, and measured for radioactive counts in a scintillation counter.

Animal experiments.

Specific-pathogen-free P_2a line chickens (1 day old) were infected by intramuscular injection with 1,000 PFU of various viruses. Mock-infected birds received intramuscular injection of uninfected CEC in culture medium supplemented with 10% FBS. Experiment 1 included five groups of 15 birds infected with viruses reconstituted from pRB-1B, as well as an additional mock-infected group. The specific viruses tested were parental pRB-1B, pRB-1B vLIP-rev (vLIP-rev), pRB-1B vLIP*-rev (vLIP*-rev), pRB-1B vLIPΔMluI-SpeI (ΔvLIP), and pRB-1B-vLIP S307A (vLIP S307A). Experiments 2 and 3 included vLIP*-rev, ΔvLIP, and vLIP S307A. In experiment 2, 8 birds per group were infected, and in experiment 3, 10 birds per group were used. Experiments were allowed to proceed for 8 weeks, after which all surviving birds were sacrificed and necropsies were performed in a randomized, blind manner. Birds that died within the first week of the experiments were excluded from the studies. During the course of the experiments, birds that presented with Marek's disease symptoms were sacrificed, and necropsies were performed. Tumor incidence, morbidity, and mortality were recorded. In the third experiment, blood samples were obtained from all birds at days 4, 12, 15, and 19 postinfection (p.i.) by wing vein puncture. Data were analyzed using the SAS package for Windows, v8.2 (SAS Institute, Cary, NC), using the chi-square and Fisher's exact tests.

Real-time qPCR.

Real-time quantitative PCR (qPCR) was performed on an ABI PRISM 7700 Sequence Detector (Applied Biosystems), and data were analyzed using ABI PRISM SDS 1.9.1 software on a Macintosh computer. For all reactions, TaqMan Universal PCR Master Mix, No AmpErase UNG (Applied Biosystems), was used. Reactions were run in MicroAmp 96-well plates (Applied Biosystems) in a final volume of 25 μl per reaction. Reaction conditions were 50°C for 2 min, 95°C for 10 min, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C. Samples were measured in duplicate on DNA samples obtained from 10 μl of whole chicken blood. DNA was prepared using the QIAGEN DNeasy tissue kit using a final elution volume of 400 μl, and 5 μl was used in each qPCR. The qPCR probes and primer sets used in this study were configured to detect the MDV-1 gD gene or the chicken genomic inducible nitric oxide synthase (iNOS) gene exactly as described previously (Table 1) (36, 43). Probes for detection of gD and iNOS were labeled with 6-carboxyfluorescein (FAM) as the reporter and with Black Hole Quencher-1 (BHQ1) as the quencher dye (Eurogentec N.A., San Diego, CA). The primer pairs used were ck-iNOS F-ck-iNOS R and gD F-gD R (Table 1).

Standards of known copy numbers were run in triplicate to generate a standard curve for each qPCR and consisted of a log₁₀ serial dilution ranging from 5 × 10⁶ to 5 × 10¹ copies in the case of the ck-iNOS target plasmid and from 5 × 10⁶ to 5 × 10⁰ copies in the case of MDV gD. The plasmid target used for generation of a standard curve for quantitation of chicken DNA copies in the unknown samples was plasmid pBS-iNOS, which harbors cDNA of the chicken iNOS gene (obtained from Keith Jarosinski) (50). The target DNA used for the generation of the gD standard curve was the MDV BAC20 clone (72). Probes were used at a final concentration of 250 nM, and primers were used at a final concentration of 900 nM. Threshold cycle number values (C_T) were calculated to correspond to the upper 10-fold standard deviation of the background fluorescence signal.

Characterization of the vLIP transcript.

The vLIP ORF had been identified in close proximity to the terminal repeat long (TR_L) region in the U_L region of the MDV genome (Fig. 1a). The lymphoblastoid MSB-1 cell line was used as a source of RNA samples for cDNA cloning and Northern blot analysis. A cDNA of the vLIP transcript spanning the first coding sequences to the poly(A) site was obtained by RT-PCR on RNA from sodium butyrate-induced MSB-1 cells. In contrast, vLIP could not be amplified from RNA samples derived from noninduced cells, indicating that vLIP is not transcribed during latency, and Northern blot analysis confirmed such an assertion (see below). Comparison of the vLIP cDNA to the corresponding genomic DNA sequence, shown in Fig. 1a, revealed that a 70-bp intron was removed from the vLIP transcript by splicing, which brings sequences encoding the signal peptide in frame with the rest of the vLIP ORF.

In order to characterize the size and kinetic class of the vLIP transcript, RNAs from MSB-1 cells induced with butyrate were collected after treatment for 30 h with PFA, an inhibitor of viral DNA polymerase. A ³²P-labeled riboprobe antisense to a 376-bp region of the vLIP reading frame, corresponding to vLIP amino acids 476 to 601, resulted in the detection of a single sodium butyrate-inducible, foscarnet-sensitive transcript of ∼2.4 kb (Fig. 1b). RNA integrity was monitored by UV imaging of ethidium bromide-stained RNAs, both prior to and after capillary transfer; levels of intact rRNA were comparable across all samples (Fig. 1b). From the above experiments we concluded that the MDV vLIP ORF is 2,271 bp in length and is expressed with late kinetics from one major spliced mRNA transcript of 2.4 kb.

Analysis of vLIP amino acid similarity to cellular lipases.

As a first approach to elucidate the function of the protein encoded by vLIP, comprehensive structural predictions were performed. The homology between vLIP and cellular lipases is especially significant in that structural residues essential to lipase activity are—for the most part—well conserved in the viral protein. Figure 2a depicts an amino acid sequence alignment of vLIP against homologous residues from a wasp venom phospholipase and the human lipoprotein lipase. A secondary-structure prediction for vLIP, generated by GenTHREADER analysis (45), is also included in Fig. 2a to demonstrate the placement of catalytic-triad positions in relation to conserved secondary-structure elements found among pancreatic lipases. The oxyanion hole residues, Phe/Trp 77 and Leu153 (by HPL numbering), which serve to stabilize tetrahedral reaction intermediates during enzyme activity, are conserved in vLIP and its adenovirus homologues. The serine residue at position 307 of vLIP (HPL numbering 152), representing the triad nucleophile (15, 79, 81), is also well conserved and is found in a typical HPL-like GHSLG motif.

In contrast, the aspartic acid position of the catalytic triad, located at position 335 on the vLIP chain (HPL numbering 176), bears an asparagine substitution in each of the MDV strains for which sequence data are available (Fig. 2a) (49, 78). The equivalent HPL mutation, D176N, has been shown to lead to a loss of enzyme activity in HPL (52). Furthermore, sequence homology of vLIP to cellular lipases is much lower in the region of the catalytic-triad histidine residue (Fig. 2b). In this region, pancreatic lipases conserve a “lid” or “flap” domain, which is delimited by a pair of cysteine residues that form the structure by participating in a disulfide bond (21, 81, 82). In some lipases, the lid regulates access to the enzyme active site and accounts for an observed preference for water-insoluble substrates known as “interfacial activation” (44). The triad histidine residue is normally located 2 amino acid positions downstream of the second cysteine residue of the pair that forms the lid. While a pair of cysteines and a histidine are present in the alignment of vLIP with relevant cellular lipases, the second cysteine involved in lid formation does not align favorably and the putative triad histidine residue would be positioned on the “lid,” instead of just after it (Fig. 2b). Thus, based on the structural criteria, it is not clear a priori whether MDV vLIP would behave like a prototypic (or conventional) lipase.

MDV vLIP lacks lipase activity.

Lipoprotein lipases are a subset of the pancreatic lipase gene family; thus, the asparagine substitution observed in vLIP at the acid position (D335N [Fig. 2a]) would most likely render the protein incapable of lipase activity (52). To eliminate the possibility that vLIP is a functional lipase enzyme whose triad acid residue has been shifted from the preferred location in strand β6 to a position following strand β7 (71), functional assays were performed to demonstrate or exclude enzymatic activity.

In order to obtain significant amounts of recombinant protein, wt vLIP and the vLIP serine nucleophile site mutant S307A, which would be unable to make nucleophilic attacks on lipid substrates, were purified from insect cell culture supernatants. A Coomassie-stained SDS-PAGE gel and Western blot analysis of the steps performed during the purification of the recombinant proteins are shown in Fig. 3a. Highly purified recombinant vLIP protein was tested on ¹⁴C-labeled phospholipid and ³H-labeled triacylglyceride substrates in vitro. Experiments were repeated several times, and representative results are shown as thin-layer chromatography demonstrating separation of substrate and reaction products for a phospholipid substrate, 1,2-dioleoylphosphatidylcholine ([1,2-¹⁴C]oleate), in which each fatty acid chain is radiolabeled (Fig. 3b) and as a phase separation-based assay for the triacylglyceride substrate glycerol tri[9,10(n)-³H]oleate (triolein) (Fig. 3c). In each reaction, 10 ng of insect cell-derived recombinant vLIP was used. Recombinant wt vLIP did not yield activity on phospholipid substrates above background levels observed in buffer alone, with a nonspecific protein control (K60), or with the S307A vLIP mutant (Fig. 3b). In contrast, phospholipid hydrolysis could readily be observed in samples to which 10 ng of bee venom phospholipase was added (Fig. 3b). As shown in Fig. 3c, vLIP did not exhibit hydrolytic activity on a triacylgyceride substrate above background levels, either. From the results of the (phospho)lipase assays, we concluded that vLIP, consistent with the unfavorable D335N substitution, lacked detectable hydrolytic activity on (phospho)lipid substrates.

The vLIP gene product is secreted and undergoes N glycosylation.

When vLIP cDNA was expressed in mammalian 293T cells (data not shown) or in chicken cells during virus replication (see below), the size of the mature, secreted protein was observed to be approximately 120 kDa, although the calculated molecular mass of vLIP (residues 31 to 756) is 81.8 kDa, taking into account signal peptide cleavage, which is predicted to remove the first 30 amino acids of the vLIP polypeptide (57, 58). To analyze putative vLIP glycosylation in more detail, vLIP was expressed in T. ni insect cells using a baculovirus expression system. SDS-PAGE, followed by Western blotting with a peptide-specific rabbit antibody, was used to compare the electrophoretic mobilities of purified vLIP protein before and after treatment with PNGase F. Untreated vLIP had an apparent molecular size of approximately 110 kDa, but treatment with PNGase F increased the electrophoretic mobility of vLIP, and the apparent molecular size was reduced to approximately 76 kDa (Fig. 4). These findings clearly indicated that the discrepancy between the 81.8-kDa predicted size of the vLIP protein chain and the size observed when vLIP was expressed in eukaryotic cells can be accounted for by N glycosylation.

Construction and in vitro characterization of vLIP MDV mutants.

The results obtained from the sequence analyses and the lipase assays suggested that vLIP may not serve as a conventional lipase enzyme. In addition, previous results had suggested that vLIP was not essential for viral replication in culture (39). In order to test whether the lipase homology of vLIP had any measurable role during in vivo replication of MDV, vLIP mutant MDVs were constructed from pathogenic strain RB-1B cloned as a BAC (pRB-1B) (64). As outlined in Fig. 5, by using shuttle mutagenesis techniques in E. coli, two vLIP mutants were prepared from pRB-1B: an S307A point mutant of the nucleophile site (vLIP S307A) and an in-frame deletion of 173 amino acids in the lipase homologous region, spanning positions 256 to 426 (ΔvLIP). Native (vLIP-rev) and C-terminally FLAG epitope tagged (vLIP*-rev) vLIP revertant clones were prepared from the ΔvLIP mutant BAC, also by shuttle mutagenesis in E. coli (Fig. 5).

The DNA of each pRB-1B mutant was analyzed by several restriction enzymes in order to confirm that the appropriate changes were present and that no spurious rearrangements took place during shuttle mutagenesis. The results of an EcoRI digest are shown in Fig. 6a, where the DNA fragment encoding the vLIP gene migrated at 3.743 kb in the digest of the ΔvLIP mutant owing to the loss of 519 bp, instead of the usual 4.262 kb observed in the digests of parental pRB-1B or vLIP-rev. The vLIP*-rev mutant exhibited a slight decrease in mobility in the fragment encoding vLIP, causing it to run as a doublet with the next largest band in the digest, which is 4.524 kb (Fig. 6a). This shift reflects the incorporation of an additional 24 bp of sequence encoding the 8-amino-acid FLAG epitope. DNA sequencing of a 4-kbp region of each mutant and revertant BAC was performed to further confirm the genetic makeup of the vLIP region of each mutant BAC; it demonstrated that only the intended modifications were introduced (data not shown).

Virus was recovered from BAC DNA after transfection of CEC, and growth kinetics demonstrated that there were no significant defects in virus replication of either vLIP mutant in vitro (Fig. 6c). To confirm that a FLAG-tagged vLIP was indeed expressed in virus-infected cells, the FLAG epitope-tagged revertant, vLIP*-rev, was grown on CEC, and supernatants and cell lysates were compared to those from parental pRB-1B virus. Anti-FLAG immunoprecipitation, as visualized by Western blotting using both anti-FLAG antibody and the vLIP-specific rabbit polyclonal antibody, clearly demonstrated a distinct band of approximately 120 kDa specific to vLIP*-rev supernatants (Fig. 6b). However, vLIP was not detected in infected-cell lysates or in concentrated supernatants by either antibody. The latter result suggested that, consistent with the presence of a suboptimal cytosine at +4 in the Kozak sequence at the translation initiation site of vLIP (ACAATGC), vLIP is not highly expressed during MDV replication in cultured cells (47-49, 78). From these results we concluded that vLIP was secreted upon lytic infection of chicken cells and that deletion or mutation of vLIP sequences did not have a detrimental effect on MDV growth properties in vitro.

vLIP is a determinant of virulence.

In a preliminary experiment, each of the mutant and revertant viruses was used to infect a group of 15 chickens of the susceptible P_2a line. Mock-infected birds and animals infected with parental virus reconstituted from pRB-1B were used as negative and positive controls, respectively. In this first experiment, increased survival and reduced incidence of Marek's disease, characterized by wasting, visceral tumors, and paralysis, were observed in the vLIP mutant virus groups relative to those groups infected with the parental or revertant virus (Tables 2 and 3). Because the observed differences in mortality in the first experiment amounted only to about a 30% reduction from the levels observed for parental and revertant viruses, two further animal studies were conducted to confirm the observed effects. In order to minimize the number of experimental animals, the pRB-1B and the untagged vLIP revertant (vLIP-rev) group were not included in any further experiments. The reduction in the number of experimental animals was possible because the first trial had demonstrated that the parental pRB-1B, the FLAG-tagged revertant vLIP*-rev, and the native vLIP revertant vLIP-rev were equally virulent, inducing characteristic gross T-cell lymphoma at rates greater than 90% (Table 2). Therefore, the notion that the incorporation of a FLAG epitope-tag at the C terminus of the vLIP reading frame might abrogate pathogenesis in vivo was no longer a concern.

The results of the second and third experiments suggested that the reduction in virulence observed in the first study was a repeatable phenotype for both the vLIP deletion and nucleophile mutant viruses. In Table 2, the observed defect of vLIP mutant viruses in causing Marek's disease is summarized, as measured by incidence of gross lymphoma in visceral organs or musculature at necropsy. In the course of the three experiments, the vLIP*-rev revertant virus was observed to cause gross lymphoma at rates of 92 to 100%, but ΔvLIP caused gross tumors in only 44 to 64% of birds, while the vLIP S307A point mutant virus caused tumors in 58 to 71% of birds. Statistical analyses using Fisher’s exact and χ² tests demonstrated that the reductions in tumor incidence in both groups were statistically significant (P < 0.004).

In order to elucidate whether any measurable defect in early lytic replication of vLIP mutant viruses could be identified, samples of whole blood were taken at days 4, 7, 12, 15, and 19 p.i. from every bird in the third experiment, which included 9 to 10 birds per group. Blood samples (10 μl) were subjected to direct DNA purification, and aliquots of DNA, each equivalent to 0.125 μl of whole blood, were analyzed by real-time PCR. This small quantity of DNA sample, though it included erythrocyte DNA, routinely was measured to contain 7 × 10⁵ to 1.6 × 10⁶ chicken genome copies, and copies detected per sample were consistent within each time point.

While virus DNA could not be detected in every sample collected on day 4 p.i., virus was reliably detected in all samples from day 7 and all later time points. The in vivo data on levels in blood shown in Fig. 7a indicate that, compared to chickens infected with revertant virus, vLIP mutant virus-infected chickens experienced reduced levels of viral DNA in peripheral blood at late, but not early, time points during lytic virus infection. Average levels of virus DNA in peripheral blood appeared nearly indistinguishable between all three groups at days 7 and 12, but vLIP*-rev showed two- to threefold-higher average levels of viral DNA at days 15 and 19 than the mutants, which remained indistinguishable from each other (Fig. 7a).

Notably, the trend seen in Fig. 7a is echoed in the survival curves in Fig. 7b, which shows the percent survival per week using the combined data from experiments 2 and 3. The revertant virus was observed to cause onset of Marek's disease more rapidly than the vLIP mutants, which each showed increased numbers of asymptomatic birds surviving the full 8 weeks to the end of the experiment (Fig. 7b). Although a subset of birds surviving to the end of the experiment invariably presented with characteristic Marek's disease tumors upon necropsy, over the three experiments the vLIP deletion and nucleophile substitution mutant groups each consistently contained a number of birds that survived to full term without any apparent Marek's disease pathogenesis at necropsy (Table 3). The percentages of birds surviving to full term without gross lymphoma or other MD symptoms at necropsy ranged from 20 to 29% for vLIP S307A and from 25 to 44% for ΔvLIP. In contrast, in the vLIP*-rev and vLIP-rev groups, completely asymptomatic birds were never observed among the few that survived to full term, and the scarce examples of birds free of gross lymphoma (n = 2) represented cases where birds were sacrificed prior to the end of the experiment because they had presented with severe paralytic symptoms. From the experiments summarized in Table 2, Table 3, and Fig. 7, we concluded that deletion of vLIP or mutation of the protein's serine nucleophile resulted in reduced lytic virus replication in vivo and significantly lower numbers of chickens that developed and/or succumbed to Marek's disease.