Virology 283, 197–206 (2001) doi:10.1006/viro.2000.0890, available online at http://www.idealibrary.com on The Long Repeat Region Is Dispensable for Fowl Adenovirus Replication in Vitro Davor Ojkic and Éva Nagy 1 Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, Ontario N1G 2W1, Canada Received December 8, 2000; returned to author for revision January 8, 2001; accepted February 26, 2001 Two regions containing tandemly repeated sequences are present in the fowl adenovirus 9 (FAdV-9) genome. The longer repeat region (TR-2) is composed of 13 contiguous 135-bp-long direct repeats, the function of which is unknown. An infectious FAdV-9 genomic clone, constructed by homologous recombination in Escherichia coli, was used for engineering of recombinant viruses. The enhanced green fluorescence protein (EGFP) coding sequence was cloned in both rightward and leftward orientations so as to replace TR-2. Replication-competent recombinant FAdVs were recovered, demonstrating that TR-2 was dispensable for FAdV-9 propagation in vitro. The expression of EGFP in infected cells was demonstrated by fluorescence microscopy, immunoprecipitation, and RT-PCR. © 2001 Academic Press Key Words: fowl adenovirus; tandem repeats; recombinant virus. AdV replication in mammalian AdV-infected cells could not be determined because mammalian AdV genes implicated in the process have a role in other essential steps of virus replication. However, it was possible to demonstrate that the induction of a heat-shock response in FAdV-infected cells is an essential step in AdV replication since CELO virus utilizes a single gene, Gam1, to induce the heat-shock response in infected cells (Glotzer et al., 2000). Early regions 1, 3, and 4 are the common sites to accommodate foreign genes in recombinant mammalian AdVs. Since FAdVs lack these regions, alternative strategies had to be explored for the construction of FAdVbased vectors. Mutational analysis demonstrated that CELO virus could tolerate deletions/insertions into the right end of its genome, which allowed the generation of replication-competent recombinant viruses (Michou et al., 1999). However, deletions in the right end of FAdV genomes can have a negative impact on virus growth, and depending on the location and the extent of deletions, these may result in a severe drop in titers of recombinant viruses (Johnson et al., 2000). The left end of the CELO virus genome contains regions that could be deleted and supplied in trans, but since no complementing cell lines are available to support propagation of replication-defective FAdVs, only replication-competent recombinant FAdVs have been characterized in more detail. An unusual feature of the FAdV-9 genome is the presence of two regions of tandemly repeated sequences (Cao et al., 1998). Imperfect repeats found in the early region 4 of mouse adenovirus type 1 appear to have an impact on virus pathogenicity (Ball et al., 1991). Tandem reiterations have also been described INTRODUCTION The family Adenoviridae includes viruses that have been isolated from many mammalian (genus Mastadenovirus) and avian (genus Aviadenovirus) species. Avian adenoviruses (AAdVs) are further subdivided into three serological groups (McFerran, 1997). Fowl adenoviruses (FAdV 1–12) belong to group I AAdVs and share a common group antigen with viruses isolated from geese, ducks, and turkeys. Several features distinguish FAdVs from their mammalian counterparts. Two fibers protruding from the penton base were observed when 14 FAdV strains, representing 11 serotypes, were examined by electron microscopy (Gelderblom and Maichle-Lauppe, 1982), whereas almost all mastadenoviruses, with the exception of subgroup F human adenoviruses (HAdVs), have only one fiber per penton base (Kidd et al., 1993; Yeh et al., 1994). FAdV genomes are much larger than those of other adenoviruses (AdVs), and FAdV-9 has the longest AdV genome whose complete nucleotide sequence has been determined so far, 45,063 bp. Moreover, FAdV-1 (CELO virus) and FAdV-9 do not have recognizable early regions 1, 3, and 4, and although most late genes are well conserved, protein V coding sequences could also not be identified (Chiocca et al., 1996; Ojkic and Nagy, 2000). Certain FAdV genes have very specialized functions and one of these facilitated the evaluation of an important aspect of the AdV replication cycle. The importance of heat-shock response to 1 To whom correspondence and reprint requests should be addressed at Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, Ontario, N1G 2W1, Canada. Fax: 519-824– 5930. E-mail: enagy@ovc.uoguelph.ca. 197 0042-6822/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved. 198 OJKIC AND NAGY FIG. 1. (A) The strategy for construction of pPacFAdV9. FAdV-9 terminal fragments in pWE-PacITR were extended by inserting the SmaI–SmaI fragment from pF-ApaI into SmaI-digested pWE-PacITR so that it now contains the complete ApaI left 4.9-kb and right 6.7-kb terminal fragments, generating pFAP2. Plasmid pFAP2 was then linearized with ApaI and cotransformed into competent BJ5183 E. coli along with FAdV-9 DNA, generating pPacFAdV9. Thick black lines represent FAdV-9 sequences and hatched blocks represent TR-2. Thin and other gray shaded lines represent vector sequences. (B) The strategy for construction of pFDTR2-EGFP and pFDTR2-EGFPinv. pAC-EGFP and pAC-EGFPinv were generated by inserting the EGFP-coding sequence (excised from pEGFP-N1 with Bsp120I and NotI) into Eco52I-digested pAC8, deleting TR-2. ApaI–XbaI fragment in pFDSal was replaced with ApaI–XbaI fragments from pAC-EGFP and pAC-EGFPinv, respectively. Recombinant plasmids containing modified viral genomes (pFDTR2-EGFP and pFDTR2-EGFPinv) were generated by recombination between SalI-linearized pFDSalEGFPinv, pFDSal-EGFP, and FAdV-9 DNA. Thick black lines represent FAdV-9 sequences and the hatched blocks are TR-2 sequences. Thin and other gray shaded lines represent vector sequences. (C) Tandem repeat region as EGFP insertion site. TR-2 contains 13 repeated subunits and each subunit has two Eco52I recognition sites. TR-2 was deleted by digestion with Eco52I, and EGFP coding sequences were inserted in place of deleted TR-2. The recombinant virus containing EGFP sequence in leftward orientation was designated rFDTR2-EGFP, whereas the recombinant virus containing the EGFP sequence in rightward orientation was designated rFDTR2-EGFPinv. in DNA from canine adenovirus 1 vaccine strain CLL (Sira et al., 1987) and a mutant human adenovirus 34 (Chen and Horwitz, 1990) as part of the inverted terminal repeats. On the other hand, FAdV-9 tandem repeats are located on the right end of the viral genome and are incorporated within a region rich in open reading frames. The shorter repeat region (TR-1) is composed of five 33-bp-long direct repeats located between nt 37,648 and 37,812, whereas the longer repeat region (TR-2), positioned between nt 38,807 and 40,561, contains 13 tandemly repeated 135-bp-long subunits (Ojkic and Nagy, 2000). Interestingly, sequences similar to FAdV-9 TR-2 were detected in field isolates of fowl adenoviruses from Ontario, India (unpub- TR-2 IS NOT REQUIRED FOR FAdV REPLICATION 199 FIG. 1—Continued lished data), and Australia (Johnson et al., 2000), but the function(s) of these repeats, if any, is unknown. The objectives of this study were to determine whether TR-2 was dispensable for FAdV-9 propagation in vitro and to examine the suitability of TR-2 as a site for insertion of foreign genes. We inserted the enhanced green fluorescence protein (EGFP) coding sequences in both rightward and leftward orientations as marker gene in place of the deleted TR-2, developed a system for the construction of TR-2-deleted recombinant FAdVs, generated recombinant FAdVs, and examined their potential as gene expression vectors in vitro. 200 OJKIC AND NAGY FIG. 2. Restriction enzyme digestion and Southern blot analysis of recombinant viruses and FAdV-9. (A) BamHI digestion of DNA extracted from FAdV-9 (lanes 1), rFDTR2-EGFP (lanes 2), and rFDTR2-EGFPinv (lanes 3). Lanes 4: 1-kb ladder. (B) DNA from the gel shown in A was transferred onto a Nytran membrane and probed with a DIG-labeled EGFP probe. RESULTS Construction of recombinant FAdV-9 viruses The construction of a plasmid containing the entire FAdV-9 genome was central to the overall strategy since generation and rescue of recombinant FAdV-9 viruses depended on a full-length, infectious genomic clone of the viral DNA. The construction of the FAdV-9 genomic clone, designated pPacFAdV9, was carried out in a series of cloning steps coupled with recombination between a plasmid containing cloned FAdV-9 end fragments (pFAP2) and FAdV-9 DNA in Escherichia coli, as depicted in Fig. 1A. When the PacI-digested pPacFAdV9 DNA was transfected into CH-SAH cells, infectious viruses were recovered. A three-step cloning/recombination approach (see Materials and Methods for detailed explanations) was utilized to construct recombinant plasmids containing modified viral genomes. The EGFP coding sequence was inserted in rightward (pFDTR2-EGFPinv) and leftward (pFDTR2-EGFP) orientations in place of deleted TR-2, as shown in Figs. 1B and 1C. A cytopathic effect was observed between 5 and 7 days after transfections of CH-SAH cells with PacI-digested pFDTR2-EGFPinv and pFDTR2-EGFP DNA. The recombinant virus generated after transfection with the former was designated rFDTR2-EGFPinv, whereas the virus generated after transfection with the latter was designated rFDTR2EGFP. The DNA extracted from rFDTR2-EGFP and rFDTR2EGFPinv were digested with BamHI and separated by agarose gel electrophoresis. The estimated sizes of BamHI fragments concurred with those predicted by computer sequence analysis of FAdV-9 DNA (Fig. 2). The presence of EGFP coding sequences in DNA extracted from recombinant viruses was also confirmed by South- ern blot analysis with an EGFP probe. The difference in sizes of BamHI DNA fragments that contained EGFP sequences (1.4 kb in rFDTR2-EGFP; 1.8 kb in rFDTR2EGFPinv) is due to different orientations of the EGFP coding sequences (Fig. 1C). The integrity of the junctions between viral sequences and cloned EGFP DNA was also confirmed by sequencing of the 1.5-kb ApaI/XbaI fragments obtained from recombinant viruses. To examine the stability of the modified viral genomes, recombinant viruses were plaque-purified through four passages in cultured cells. DNA was extracted from 12 plaque-purified recombinant viruses and subjected to restriction enzyme analysis. No detectable differences in the restriction enzyme pattern were observed between DNA extracted from recombinant viruses representing the first and fourth passages. The plaque morphology of the rFDTR2-EGFP and rFDTR2-EGFPinv was also indistinguishable from that of FAdV-9 (data not shown). Analysis of EGFP expression Fluorescence microscopy was utilized for rapid detection of EGFP expression in CH-SAH and QT-35 cells that were infected with recombinant viruses. Green fluorescence was observed in cells infected with both rFDTR2EGFP and rFDTR2-EGFPinv. However, rFDTR2-EGFPinv induced much higher levels of EGFP expression than rFDTR2-EGFP, as judged by the intensity of green fluorescence in infected cells (Fig. 3), and was used in all subsequent experiments. EGFP expression in rFDTR2-EGFPinv-infected CHSAH cells was also evaluated by immunoprecipitation. CH-SAH cells were labeled with [ 35S]methionine at vari- FIG. 3. Detection of EGFP expression 24 h.p.i. by fluorescence microscopy in cells infected at an m.o.i. of 1. (A) CH-SAH cells infected with FAdV-9. (B) rFDTR2-EGFP-infected CH-SAH cells. (C) rFDTR2EGFPinv-infected CH-SAH cells. (D) rFDTR2-EGFPinv-infected QT-35 cells. TR-2 IS NOT REQUIRED FOR FAdV REPLICATION 201 lication in cultured cells was lower than that of its wildtype counterpart between 12 and 30 h.p.i. However, between 36 and 48 h.p.i., rFDTR2-EGFPinv titers were similar to those obtained for FAdV-9. The yield of recombinant virus was routinely 1–2 3 10 8 plaque-forming units/ml (p.f.u./ml), which is similar to titers normally obtained for FAdV-9 (3 3 10 8 p.f.u./ml). Temporal analysis of EGFP mRNA expression FIG. 4. Detection of EGFP by immunoprecipitation. CH-SAH cells were either mock-infected (lane 1) or infected with rFDTR2-EGFPinv (lanes 2, 3, and 5) or FAdV-9 (lane 4) and labeled with [ 35S]methionine. Cell lysates were obtained from infected cells at different times p.i. (lanes 1 and 2: 18 h; lane 3: 24 h; lanes 4 and 5: 36 h) and immunoprecipitated with an EFGP-specific monoclonal antibody. Immunoprecipitated proteins were separated by SDS–PAGE and visualized by fluorography. Lane M: marker. ous times postinfection (p.i.). Proteins from lysates of radiolabeled cells were immunoprecipitated with an EGFP-specific monoclonal antibody, separated by SDS– PAGE, and visualized by fluorography. A major band at a molecular mass of 27 kDa was observed in cells infected with rFDTR2-EGFPinv, but not in mock-infected or FAdV9-infected CH-SAH cells and was in agreement with the expected molecular mass of EGFP, 27 kDa (Fig. 4). EGFP expression in infected cells could be detected by fluorescence microscopy and immunoprecipitation only during the late times of rFDTR2-EGFPinv infection. One-step growth curve Importantly, the one-step growth curve experiment results indicated that rFDTR2-EGFPinv grew as efficiently as FAdV-9 (Fig. 5). The kinetics of rFDTR2-EGFPinv rep- The temporal transcription profile of EGFP mRNA in rFDTR2-EGFPinv-infected cells was examined by RT-PCR at various times p.i. Although both fluorescence microscopy and immunoprecipitation results suggested that EGFP expression occurred during the late phase of infection, mRNA transcription may have started earlier. A low level of EGFP mRNA was detected as early as 2 h.p.i. but then decreased to barely detectable levels between 4 and 10 h.p.i. The level of EGFP mRNA expression increased detectably between 10 and 12 h.p.i. and continued throughout the late phase of infection (Fig. 6). Identification of FAdV-9 leader sequences, promoters, and poly(A) site 59-RACE analysis was carried out with EGFP-specific primers to examine the structure of EGFP mRNA 24 h.p.i. and to eventually identify the promoter used to initiate EGFP expression in cells infected with rFDTR2-EGFPinv. Sequence analysis of the 59-RACE product of the untranslated 59 portion of EGFP mRNA revealed the presence of bipartite leader sequences, which were also found in the untranslated 59 regions of CELO virus late transcripts (Payet et al., 1998). The first leader was located between nt 8385 and 8414 and the second leader FIG. 5. One-step growth curves for rFDTR2-EGFPinv and FAdV-9. Confluent monolayers of CH-SAH cells in six-well dishes were infected with rFDTR2-EGFPinv and FAdV-9 at an m.o.i. of 5. Supernatants and cells were harvested at various times p.i. and the production of total virus was monitored by plaque assays. The data represents two replicates at each time point. 202 OJKIC AND NAGY FIG. 6. Kinetics of EGFP mRNA expression in rFDTR2-EGFPinv infected cells. CH-SAH cells were either mock-infected (lane N) or infected with rFDTR2-EGFPinv or FAdV-9. Total RNA was extracted at various times p.i., indicated by the numbers above the lanes, reverse transcribed, and amplified by PCR with primers internal to EGFP mRNA. PCR products were separated by agarose gel electrophoresis and visualized by ethidium bromide staining. Lane M: 1-kb ladder. was between nt 12,299 and 12,421. Following the bipartite leader the first nucleotide upstream of the EGFP was at nt 38,503, 287 nts from the EGFP ATG initiation codon. The presence of these leader sequences on the EGFP mRNA suggested that EGFP was expressed from the major late promoter (MLP). The location of FAdV-9 MLP was predicted between nt 8173 and 8361, with a putative TATA box located between nt 8345 and 8351 (Ojkic and Nagy, 2000). 39-RACE with EGFP-specific primers was also carried out and the nucleotide sequence of the cloned 39-RACE product was analyzed. The EGFP mRNA was polyadenylated and the terminal end was at nt 40,178. This suggests that the polyadenylation signal used for EGFP mRNA was the one located between nt 40,155 and 40,159 (Fig. 7). DISCUSSION Our results demonstrated that TR-2 was dispensable for in vitro virus replication since deletion of TR-2 and insertion of the EGFP coding sequence in either orientation did not have a significant impact on FAdV-9 growth characteristics. In contrast to HAdV vectors where expression cassettes containing exogenous promoters and polyadenylation sites are often used, the EGFP gene inserted in place of TR-2 lacked additional transcriptional elements. Nevertheless, EGFP expression was under the control of the endogenous FAdV-9 MLP. Moreover, late in infection, high levels of transgene expression were de- tected by RT-PCR, fluorescence microscopy, and radioimmunoprecipitation. The role of TR-2, if any, during the natural course of FAdV-9 infection is unknown. The hypervirulent FAdV-8 strain CFA40 genome has seven 145-bp-long tandem repeats, showing 80.4% similarity to FAdV-9 TR-2 and in a similar location as in FAdV-9. Although the development of FAdV-8 strain CFA40-based vectors has been reported, the role of these sequences was not investigated (Johnson et al., 2000). Interestingly, in our laboratory Southern hybridization revealed sequences similar to FAdV-9 TR-2 in numerous FAdV field isolates associated with outbreaks of inclusion body hepatitis in chickens in Ontario. Partial DNA sequence analysis of two field isolates demonstrated that, although other parts of the genomes of these viruses showed relatively weak similarities to FAdV-9, they also contained repeats with 98% identity to FAdV-9 TR-2. Furthermore, the DNA obtained from an FAdV field isolate from India contained sequences which were identical to FAdV-9 TR-2 (data not shown). The presence of TR-2-like sequences in FAdV field isolates originating from three continents suggests that TR-2-related sequences might play an important role during the natural course of infection. Since several different FAdV serotypes have been associated with inclusion body hepatitis, the genotype might play a more important role than the serotype in the pathogenicity of the particular group I AAdV strain (McFerran, 1997). Inverted and direct repeat sequences, playing various roles in viral replication cycles, have also been found in retroviruses, poxviruses, and herpesviruses. In most cases, the exact role of tandem repeats is still elusive and their overall implications and precise role in virus replication are not well understood. However, it has been reported that tandemly repeated sequences could be involved in modulation of viral virulence and oncogenicity. For example, Marek’s disease virus (MDV) contains a variable number of copies of a 132-bp-long direct repeat. An increase in the copy number of this repeat has been positively correlated with viral attenuation, but it is not the only determinant of MDV virulence (Tulman et al., 2000). The sequence similarity between FAdV-9 TR-2 and the MDV 132-bp repeat is relatively low, only 35.9% at the FIG. 7. The structure of EGFP mRNA in cells infected with rFDTR2-EGFPinv. Total RNA extracted from cells infected with rFDTR2-EGFPinv was reverse transcribed and used for 59- and 39-RACE. The sequence analysis of cloned cDNAs revealed the presence of bipartite leaders. TR-2 IS NOT REQUIRED FOR FAdV REPLICATION nucleotide level, and these repeats do not appear to be related. However, a possible role of TR-2 as a genotypeassociated factor which might be involved in modulating FAdV virulence is intriguing and future studies are required to examine the role of TR-2 in vivo. Since TR-2 is dispensable, it can be used as an alternative site for insertions of foreign genes. The development of transfer plasmids in which TR-2 is replaced with a gene of interest, and our strategy to generate a recombinant virus in two cloning steps followed by recombination in E. coli, simplifies the time-consuming and laborintensive process of production and purification of recombinant viruses. These FAdV-9 vectors will be developed as recombinant vaccines for poultry, where use of replication-competent viruses could be advantageous in inducing a protective immune response. Wellcharacterized HAdVs have been extensively used in recombinant vector development for both humans and animals. Nonetheless, we note that FAdV-9-based vectors have several characteristics that could make them even more attractive for applications in mammalian species. The genome of FAdVs is much larger (;45 kb) than genomes from mammalian AdVs (;30–36 kb). This property might allow for insertion of larger genes into FAdVbased vectors compared to HAdV-based vectors. While the presence of HAdV neutralizing antibodies can severely limit their usefulness for human applications, it is less likely that a preexisting immune response to FAdVs exists. HAdV infections are common in humans, so that even helper-dependent HAdV vectors might revert to replication-competent viruses by in vivo recombination with naturally occurring HAdVs. FAdVs are naturally replication-defective in human cells, which provides an additional safety feature (Cotten et al., 1993). However, they can still be used for gene delivery and expression in nonpermissive mammalian systems. For example, CELO virus is capable of transducing several types of human cells in culture at levels comparable to HAdV-5 (Michou et al., 1999). We have found that rFDTR2-EGFPinv could also transduce human lung carcinoma cells (A549) and bovine kidney (MDBK) cells (data not shown). However, transduction efficiency (judged by the number of infected cells expressing EGFP) and EGFP expression levels in the mammalian cells (judged by the intensity of green fluorescence in transduced cells) were lower than those observed in cells of avian origin, even though the cells were infected with the same m.o.i. It is clear that more detailed characterization of FAdV molecular biology and evaluation of viral gene expression are required before FAdV vectors could be used in nonpermissive systems. The work reported here lays the foundation for further structural and functional studies of FAdV-9 and facilitates our efforts in developing this virus as a recombinant vector. 203 MATERIALS AND METHODS Virus FAdV-9 was obtained from ATCC as avian adenovirus type 8, strain A-2A (ATCC VR-833). In the older North American literature this strain is often called avian adenovirus type 8, but in the more recent North American literature the same strain is called fowl adenovirus type 8 and fowl adenovirus serotype 9. In the older European literature the strain A2 is called fowl adenovirus serotype 9, whereas in the newer European literature the same strain is sometimes designated serotype 10. In our previous publications (Clavijo et al., 1996; Alexander et al., 1998; Cao et al., 1998, 2000; Ojkic and Nagy, 2000) the strain A-2A is called avian/fowl adenovirus type 8. However, in the Seventh Report of the International Committee on Taxonomy of Viruses (Benkö et al., 2000) strain A-2A is now listed as FAdV-9 and therefore in this paper we followed the official designation. Cell lines The virus was propagated in a chicken hepatoma cell line (CH-SAH), kindly provided by Solway Animal Health (Mendota Heights, MN). Quail tumor cells (QT-35) were a generous gift from C. Moscovici. The cells were maintained in DMEM-F12 medium (Life Technologies) supplemented with 10% fetal bovine serum (FBS; Cansera), 2 mM L-glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin. Viral DNA extraction The viral DNA was extracted from purified virus. Infected cells were harvested at maximum CPE, freezethawed 3 times, and centrifuged at 6000 rpm in a Beckman 16.250 rotor for 30 min at 4°C to remove cellular debris. The virus was pelleted by ultracentrifugation and resuspended in TE buffer (pH 8.0). The 30-ml centrifuge tube was filled to 2/3 of its volume with the virus suspension, and 1/3 of the volume of ice-cold sucrose solution (30% w/w in TE buffer) was then layered beneath the virus suspension. The tube was centrifuged at 24,000 rpm in a Beckman SW-28 rotor for 90 min at 4°C; the supernatant was removed and the pellet was resuspended in 1 ml TE buffer. The virus was disrupted by addition of an equal volume of 23 lysis buffer [0.01 M Tris–HCl (pH 7.5) 0.01 M EDTA, 1% SDS, 1 mg/ml proteinase K in 0.01 M Tris–HCl (pH 8)] and incubated overnight at 37°C. The lysed virus was then extracted with phenol, the aqueous phase was transferred to a fresh centrifuge tube, and viral DNA was recovered by ethanol precipitation. Construction of FAdV-9 genomic clones The strategy for the construction of a full-length FAdV-9 genomic plasmid clone is depicted in Fig. 1A. A plasmid 204 OJKIC AND NAGY (pITR) containing the FAdV-9 left 2.5-kb and right 1.5-kb SmaI-terminal fragments in opposite orientation was initially constructed. The terminal fragments, contained in a BamHI fragment, were then transferred into pWE-Amp, a modified cosmid vector derived from pWE-15 (Clontech) and PacI-restriction sites were introduced into the ends of the cloned viral DNA by linker addition (pWE-PacITR). FAdV-9 DNA lacks PacI sites and PacI sites were introduced to enable linearization of cloned viral DNA from the plasmid prior to transfections. The FAdV-9 terminal fragments in pWE-PacITR were then extended by inserting the SmaI–SmaI fragment from pF-ApaI so that it now contains the complete ApaI left 4.9-kb and right 6.7-kb terminal fragments (pFAP2). Recombination in E. coli was carried out as first described by Chartier et al. (1996). pFAP2 was linearized with ApaI and cotransformed into competent BJ5183 E. coli (kindly provided by D. Hanahan) along with FAdV-9 DNA, allowing for recombination between overlapping fragments to occur and generating pPacFAdV9, a plasmid containing the entire cloned FAdV-9 genome. Restriction enzyme analysis verified the correct structure of the plasmid. into CH-SAH cells by Lipofectamin (Life Technologies) as suggested by the manufacturer. Approximately 1 mg of DNA and 5 ml of Lipofectamin were used for each transfection. Cells were exposed to the DNA–liposome complexes overnight and 1 ml of DMEM-F12 containing 20% FBS was added the next morning. The medium was replaced with fresh, complete DMEM-F12 24 h after transfection. CPE was typically observed between 5 and 7 days after transfections. Generation of recombinant FAdV-9 viruses Fluorescence microscopy A three-step strategy was developed for the construction of two recombinant viruses. Plasmid pAC-Not(2) contains ApaI FAdV-9 terminal fragment between nt 38,288 and 45,063, including the TR-2. In the first step the EGFP coding sequence was excised from pEGFP-N1 (Clontech) with Bsp120I and NotI and cloned into Eco52Idigested pAC8 (see Fig. 1A) in both rightward (pACEGFPinv) and leftward (pAC-EGFP) orientations, deleting TR-2. In the second step the ApaI–XbaI fragments (containing the EGFP coding sequence in place of the deleted TR-2) were excised from pAC-EGFP and pAC-EGFPinv and used to replace the ApaI–XbaI fragment in the plasmid pFDSal (constructed by digesting pPacFAdV9 with SalI to completion followed by religation resulting in deletion of all internal SalI fragments from pPacFAdV9). Two transfer vectors were constructed, containing the EGFP coding sequence in rightward (pFDSal-EGFPinv) and leftward (pFDSal-EGFP) orientations. In the third step, the transfer vectors pFDSal-EGFPinv and pFDSalEGFP were linearized with SalI and cotransformed into competent BJ5183 E. coli along with FAdV-9 DNA. Resultant recombinant plasmids contained modified viral genomes with EGFP coding sequences replacing deleted TR-2 in rightward (pFDTR2-EGFPinv) and leftward (pFDTR2-EGFP) orientations (Fig. 1B). Restriction enzyme analysis verified the correct structure of the cloned viral DNAs. For the fluorescence microscopy analysis cells were grown in chamber slides and infected with FAdV-9, rFDTR2-EGFP, and rFDTR2-EGFPinv at a multiplicity of infection (m.o.i.) of 1. The cells were washed twice with PBS and covered with a coverslip. EGFP expression in infected cells was observed by an Olympus Provis AX-70 microscope equipped with FITC optics. Transfections PacI-digested cloned viral DNAs (pPacFAdV9, pFDTR2-EGFP, and pFDTR2-EGFPinv) were transfected Southern blotting Viral DNA extracted from purified virions was digested with restriction enzymes and separated by agarose gel electrophoresis. Transfer of DNA from agarose gels to nylon membranes (Schleicher and Schuell) was carried out as described elsewhere (Southern, 1975). DIG DNA Labeling and Detection Kit (Roche) was used for the labeling of gel-purified EGFP DNA (BamHI–NotI fragment from pEGFP-N1). The manufacturer’s instructions were followed for the hybridization of DIG-labeled EGFP probe to immobilized nucleic acids as well as for the colorimetric detection of hybridized DNA sequences. Immunoprecipitation CH-SAH cells were grown in 60-mm dishes, infected with FAdV-9 and rFDTR2-EGFPinv at an m.o.i. of 5. The cells were starved for 1 h in a methionine-deficient medium for 1 h before labeling. At various times p.i. the cells were labeled with 50 mCi/ml of Easy Tag Express [ 35S]methionine (NEN) for 2 h. After labeling, the cells were washed twice with cold PBS and disrupted in 500 ml of lysis buffer (PBS containing 0.5% Triton X-100 and 0.1 mM PMSF). The lysate was cleared by centrifugation at 14,000 rpm in an Eppendorf 5415C microcentrifuge for 20 min at 4°C and the supernatant was transferred to a clean tube. Fifty-microliter aliquots of the lysate was used for immunoprecipitations. In the preclearing step to remove nonspecifically bound material the lysate was mixed with 30 ml of Pansorbin cells (Calbiochem), incubated for 1 h on ice, and centrifuged at 14,000 rpm in an Eppendorf 5415C microcentrifuge for 1 min at 4°C. The supernatant was transferred into a fresh tube and 2 ml of Living Colors (JL-8) monoclonal antibody (Clontech) was added and kept on ice for 2 h, while mixing regularly. In the precipitation step, 30 ml of Pansorbin cells were added and kept on ice, mixing regularly. Immune com- TR-2 IS NOT REQUIRED FOR FAdV REPLICATION plexes were pelleted by centrifugation at 14,000 rpm in an Eppendorf 5415C microcentrifuge for 1 min at 4°C. The immunoprecipitates were solubilized in SDS sample buffer, subjected to electrophoresis in 10% SDS–polyacrylamide gels, and analyzed by fluorography. One-step growth curve CH-SAH cells were grown in 35-mm tissue culture plates and infected with FAdV-9 and rFDTR2-EGFPinv at an m.o.i. of 5. At various times p.i. the medium was removed and frozen at 270°C, the cells were washed 3 times with PBS, 1 ml of fresh medium was added to each plate, and the plates were frozen at 270°C. The cells were freeze-thawed 3 times before titrations. Plaque assays were carried out for both intracellular and extracellular virus as described elsewhere (Alexander et al., 1998). RT-PCR Total RNA from CH-SAH cells infected with FAdV-9 and rFDTR2-EGFPinv at an m.o.i. of 5 was extracted at various times p.i. by Trizol (Life Technologies). The total RNA was first treated with DNaseI (Life Technologies) to remove any traces of viral DNA. First-strand cDNA synthesis was carried out in a total reaction volume of 20 ml. For the first-strand cDNA synthesis 2 mg of total RNA was mixed with 100 ng of random primers (Life Technologies), heated at 70°C, and then chilled on ice. After addition of 1 ml of 10 mM dNTP mix, 4 ml of 53 first-strand buffer [250 mM Tris–HCl (pH 8.3), 375 mM KCl, 15 mM MgCl 2 ], and 2 ml of 0.1 mM DTT, the reactions were incubated for 10 min at 25°C. Two-hundred units of SuperScript II RNaseH 2 reverse transcriptase (Life Technologies) was added, mixed gently, and incubated at 45°C for 45 min. After heat inactivation (70°C for 15 min) the reactions were treated with 2 U of E. coli RNaseH (Life Technologies) at 37°C for 20 min to remove the RNA. Two-microliter aliquots of the first-strand reactions were taken for PCR amplification in a total reaction volume of 50 ml. Reactions without reverse transcriptase were the negative control used to ascertain that no residual viral DNA was present. Two sets of primers were used for PCR amplification to evaluate EGFP mRNA expression: TR-F2 (59-GTTACGCTCCACCCCGAATCCAG-39) and TR-R2 (59-GGGACAAGGTTTACACGGACGAGA-39), flanking the EGFP ORF. The results of RT-PCR analysis obtained by using internal EGFP primers, 3EGFP (59AAGCAGAAGAACGGCATCAAGGTG-39), and EGFP-R (59-CACGAACTCCAGCAGGACCATG-39) are not shown. 59 and 39 rapid amplification of cDNA ends (59-RACE) 59- and 39-RACE kits (Life Technologies) were used according to the manufacturer’s recommendations. Total RNA was extracted from cells infected with rFDTR2- 205 EGFPinv at 8 and 24 h.p.i., pooled, and reverse transcribed as described above. 59-RACE was done with two EGFP specific primers: 5EGFP (59GGGTCTTGTAGTTGCCGTCGTCCT39) and nested EGFP-E (59TTGAATTCATCGCCCTCGCCCTCGCCG39). 39-RACE was also carried out with two EGFP specific primers: 3EGFP (59AAGCAGAAGAACGGCATCAAGGTG39) and nested EGFP-C (59CATGGTCCTGCTGGAGTTCGTG39). The 59and 39-RACE amplification products were gel purified and cloned in pUC19. The nucleotide sequence of cloned cDNA was determined as described below. Sequencing and nucleotide sequence analysis All sequencing reactions were carried out with an Applied Biosystems 377 automated sequencer located at the University of Guelph Molecular Supercentre. Sequence analysis, plasmid map drawing, and oligonucleotide design were done by the Vector NTI 5.0 software package (Informax Inc.). ACKNOWLEDGMENTS The authors acknowledge the editorial comments on this manuscript by Peter J. Krell. This work was supported by the Natural Sciences and Engineering Research Council of Canada and the Ontario Ministry of Agriculture, Food, and Rural Affairs. REFERENCES Alexander, H. S., Huber, P., Cao, J. X., Krell, P. J., and Nagy, É. (1998). Growth characteristics of fowl adenovirus type 8 in a chicken hepatoma cell line. J. Virol. Meth. 74, 9–14. Ball, O. B., Clayton, W. B., Villegas, P., and Spindler, K. R. (1991). Early region 4 sequence and biological comparison of two isolates of mouse adenovirus type 1. Virology 180, 257–265. Benkö, M., Harrach, B., and Russell, W. C. (2000). Family Adenoviridae. In “Virus Taxonomy: Classification and Nomenclature of Viruses. Seventh Report of the International Committee on Taxonomy of Viruses” (M. H. V. van Regenmortel, C. M. Fauquet, D. H. L. Bishop, E. B. Carstens, M. K. Estes, S. M. Lemon, J. Maniloff, M. A. Mayo, D. J. McGeoch, C. R. Pringle, and R. B. Wickner, Eds.), pp. 227–238. Academic Press, San Diego, CA. Cao, J. X., Krell, P. J., and Nagy, É. (1998). Sequence and transcriptional analysis of terminal regions of the fowl adenovirus type 8 genome. J. Gen. Virol. 79, 2507–2516. Cao, J. X., Krell, P. J., and Nagy, É. (2000). The ORF RTL1 transcript of fowl adenovirus type-8 is spliced and truncated at late stages of the virus replication cycle. Virus Genes 20, 135–137. Chartier, C., Degryse, E., Gantzer, M., Dieterle, A., Pavirani, A., and Mehtali, M. (1996). Efficient generation of recombinant adenovirus vectors by homologous recombination in Escherichia coli. J. Virol. 70, 4805–4810. Chen, M., and Horwitz, M. S. (1990). Replication of an adenovirus type 34 mutant DNA containing tandem reiterations of the inverted terminal repeat. Virology 179, 567–575. Chiocca, S., Kurzbauer, R., Schaffner, G., Baker, A., Mautner, V., and Cotten, M. (1996). The complete DNA sequence and genomic organization of the avian adenovirus CELO. J. Virol. 70, 2939–2949. Clavijo, A., Krell P. J., and Nagy, É. (1996). Molecular cloning and restriction enzyme mapping of avian adenovirus type 8 DNA. Virus Res. 45, 93–99. Cotten, M., Wagner, E., Zatloukal, K., and Birnstiel, M. L. (1993). Chicken 206 OJKIC AND NAGY adenovirus (CELO virus) particles augment receptor-mediated DNA delivery to mammalian cells and yield exceptional levels of stable transformants. J. Virol. 67, 3777–3785. Gelderblom, H., and Maichle-Lauppe, I. (1982). The fibers of fowl adenoviruses. Arch. Virol. 72, 289–298. Glotzer, J. B., Saltik, M., Chiocca, S., Michou, A. I., Moseley, P., and Cotten, M. (2000). Activation of heat-shock response by an adenovirus is essential for virus replication. Nature 407, 207–211. Johnson, M. A., Pooley, C., and Lowenthal, J. W. (2000). Delivery of avian cytokines by adenovirus vectors. Dev. Comp. Immunol. 24, 343–354. Kidd, A. H., Chroboczek, J., Cusack, S., and Ruigrok, R. W. (1993). Adenovirus type 40 virions contain two distinct fibers. Virology 192, 73–84. McFerran, J. B. (1997). Group I adenovirus infections. In “Diseases of Poultry” (B. W. Calnek, Ed.), pp. 608–620. Iowa State Univ. Press, Ames, IA. Michou, A. I., Lehrmann, H., Saltik, M., and Cotten, M. (1999). Mutational analysis of the avian adenovirus CELO, which provides a basis for gene delivery vectors. J. Virol. 73, 1399–1410. Ojkic, D., and Nagy, É. (2000). The complete nucleotide sequence of fowl adenovirus type 8. J. Gen. Virol. 81, 1833–1837. Payet, V., Arnauld, C., Picault, J. P., Jestin, A., and Langlois, P. (1998). Transcriptional analysis of avian adenovirus CELO. J. Virol. 72, 9278– 9285. Sira, S., Abouhaidar, M. G., Liu, Y. C., and Campbell, J. B. (1987). Multiple reiteration of a 40-bp nucleotide sequence in the inverted terminal repeat of the genome of a canine adenovirus. Virology 159, 76–83. Southern, E. M. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98, 503–517. Tulman, E. R., Afonso, C. L., Lu, Z., Zsák, L., Rock, D. L., and Kutish, G. F. (2000). The genome of a very virulent Marek’s disease virus. J. Virol. 74, 7980–7988. Yeh, H. Y., Pieniazek, N., Pieniazek, D., Gelderblom, H., and Luftig, R. B. (1994). Human adenovirus type 41 contains two fibers. Virus Res. 33, 179–198.