Transcription mapping of mouse adenovirus type 1 early region 3
Post on 30-Aug-2016
Embed Size (px)
VIROLOGY 175,81-90 (1999)
Transcription Mapping of Mouse Adenovirus Type 1 Early Region 3
CLAYTON W. BEARD, AMY OBERHAUSER BALL, E. HANNAH WOOLEY, AND KATHERINE R. SPINDLER
Department of Genetics, University of Georgia, Athens, Georgia 30602
Received September 6, 1989; accepted November 10, 1989
Early region 3 (E3) of mouse adenovirus type 1 was analyzed using Sl nuciease protection and primer extension assays, cDNA sequencing, and genomic sequencing. We present the genomic sequence from 79 to 83 map units of the viral genome, the precise ends and splice sites of the E3 mRNAs, and the predicted protein sequence encoded by the mRNAs. Three major classes of early mRNAs were identified; all were approximately 1 kb long, consisted of three exons, and shared 5 and 3 ends. The three classes had alternative splicing at the junction between the second and third exon. The three proteins predicted by the three mRNAs were slightly similar to the E3 19K glycoprotein of human adenovirus type 3; the longest of the three was the most similar. Open reading frames corresponding to late proteins were also identified in the translated mouse adenovirus type 1 DNA sequence. In mouse adenovirus, as in the human adenoviruses, L4 overlaps E3, and L5 starts just downstream of the E3 region. 0 1990 Academic Press, Inc.
Early region 3 (E3) of human adenovirus types 2 and 5 (Ad2/5) is transcribed rightward from 75.9 to 86 map units (m.u.) on the genome (Berk and Sharp, 1978; Chow et al., 1979; Pettersson et a/., 1976). E3 tran- scription is complex, consisting of at least nine overlap- ping mRNAs generated by alternative splicing (Chow et al., 1979; Cladaras et al., 1985). Heteroduplex map- ping indicates considerable sequence divergence in E3, even within subgroups of adenoviruses (Bartok et al., 1974; Belak et al., 1986; Garon et a/., 1973). E3 is nonessential for viral replication in vitro (Kelly and Lewis, 1973). However, since the E3 region has been evolutionarily conserved, it is believed that E3 proteins are important for natural infections. It has recently been shown that E3 moderates adenovirus infection of cot- ton rats (Ginsberg et a/., 1989) and hamsters (Morin et a/., 1987). Viral mutants deleted for E3 replicate like wild-type virus in cotton rat lungs, but elicit an in- creased inflammatory response. In addition, studies of E3 proteins and mutants in cell culture suggest that this region is involved in virus-host interactions.
The human E3 region contains at least nine open reading frames (offs) long enough to encode polypep- tides 6K or longer. Proteins have been identified in Ad2/ 5-infected cells corresponding to six of these orfs (Tol- lefson and Wold, 1988; Wang et al., 1988; Wold et a/., 1984; W. Wold, personal communication). The Ad2/5 19K glycoprotein (gpl9K) is the best-characterized of these and it is the most abundant viral protein at early times after infection (Persson e2 a/., 1979; Wold et a/.,
To whom requests for reprints should be addressed.
1985). The gpl9K is a transmembrane protein that can bind class I major histocompatibility (MHC) antigens (Kampe et a/., 1983; Kvist et a/., 1978; Paabo et al., 1983; Severinsson and Peterson, 1985; Signas et al., 1982) and this binding inhibits the glycosylation of the antigens and prevents their efficient transport to the cell surface (Andersson et al., 1985; Burgert and Kvist, 1985; Severinsson and Peterson, 1985). Despite se- quence divergence in the E3 regions, the ability of the 19K glycoprotein to inhibit transport of the class I MHC antigen is conserved in all human adenovirus sub- groups except subgroup A (Paabo eta/., 1986). The E3 10.4K protein has been shown to down-regulate the expression of epidermal growth factor receptors on in- fected cells (Carlin et al., 1989) and it has been shown that this may involve a complex of the 10.4K protein and the recently identified E3 14.5K protein (W. Wold, personal communication). In cultured cells infected by Ad2/5, an E3 14.7K protein prevents lysis by tumor ne- crosis factor (Gooding et a/., 1988). No function has been identified for the 1 1.6K and 6.7K proteins found in infected cells. The known biochemical activities of the 19K, 14.7K, 14.5K, and 10.4K Ad2/5 E3 proteins in cell culture infections all involve interactions with host cell components and are probably important for the bi- ology of adenovirus infection.
The role of E3 in in viva adenovirus infections has been difficult to study because of the species-specific- ity of human adenoviruses. Although there has been success in using cotton rats (Ginsberg et a/., 1989) and hamsters (Hjor-th eta/., 1988) to studyin vivo pathogen- esis, development of a model using an adenovirus in its natural host is desirable. The molecular genetics of mouse adenovirus type 1 (MAV-1) are being character-
81 0042.6822/90 $3.00 CopyrIght 0 1990 by Academic Press, Inc All rights of reproduction ,n any form reserved
82 BEARD ET AL.
ized, and MAV-1 should provide a useful animal adeno- virus model. In this work we have focused on the E3 region of MAV-1. Ball et a/. identified early region 1 (El) at the left end of the MAV-1 genome (Ball et al., 1989, 1988); on the basis of this orientation of MAV-1, we predicted that a region like that of Ad2/5 E3 would be found near 80 m.u. on the MAV-1 genome.
We present the genomic sequence, transcription mapping, and the predicted proteins of the E3 region of MAV-1. Three major alternatively spliced E3 mRNAs were identified in a region corresponding to 79 to 83 m.u. on the MAV-1 genome, a region approximately 1000 bp in length. Analysis of the predicted polypep- tides of these mRNAs revealed one with a slight similar- ity to human adenovirus E3 gpl9K. Other ot-fs with sim- ilarity to the late proteins, 33K phosphoprotein, pVIII, and fiber, were identified; however, no other similarities to human adenovirus E3 proteins or orfs were noted. The significance of these findings and their relevance to the role of E3 in adenovirus pathogenesis is dis- cussed.
MATERIALS AND METHODS
Plasmids and genomic sequencing
The construction of plasmids containing MAV-1 ge- nomic fragments used in this work was described pre- viously (Ball et al., 1989). For sequencing genomic DNA, the HindIll-C fragment was cloned into Blue- scribe+ (Stratagene) in both orientations. A series of deletion clones was made using the Exolll nuclease method (Henikoff, 1984) and the clones were se- quenced by the dideoxy chain-termination method (Sanger et al., 1977) using Sequenase enzyme (U.S. Biochemical Corp.). The sequence was compared with that of Raviprakash et a/. (1989) and found to be identi- cal. Computer analysis was performed using IntelliGe- netics, Inc. and International Biotechnologies, Inc./ Pustell software as described (Ball et a/., 1988).
Mouse L cells in monolayers and MAV-1 were grown as previously described (Ball et al., 1989). Suspension L cells were seeded from monolayers at 2 X 1 O4 cells/ ml in a-MEM containing 10% heat-inactivated calf se- rum, and infected with 1 to 5 PFU/cell at a density of 5 X lo5 cells/ml in (u-MEM containing 2% heat-inacti- vated calf serum. Total RNA was isolated from mock- or MAV-1 -infected L cell monolayers or suspension L cells as previously described (Ball et al., 1989). Early RNA was collected at 22 hr post infection (p.i.) and late RNA was collected at 45 or 48 hr p.i. (Ball et a/., 1989). Northern blots, Sl nuclease protection assays, and
primer extension analyses were performed as pre- viously described (Ball et a/., 1989).
The cDNA library was generated from 22 hr MAV-l- infected-cell poly(A)+ RNA (Ball eta/., 1989) and 2 X 1 O6 plaques were screened with the isolated MAV-1 HindIll-C fragment. The 29 cDNA clones which hybrid- ized were isolated and sequenced by the dideoxy chain-termination method (Sanger et al., 1977) using Sequenase enzyme (U.S. Biochemical Corp.). Oligonu- cleotide primers used for sequencing cDNAs and for primer extension assays were synthesized by Operon Technologies.
Genomic sequence of E3
The MAV-1 genomic sequence of the HindIll-C frag- ment (77 to 89 m.u.; nt 1 is at the HindIll site at 77 m.u.) was obtained by sequencing clones from a deletion se- ries as described under Materials and Methods. The transcription analysis below indicated that E3 was tran- scribed within the segment from 79 to 83 m.u., and only this segment is shown here. The sequence was determined from both strands for a majority of this seg- ment. Where sequence for only one strand was ob- tained, it was compared and found to be identical to the HindIll-C fragment sequence determined by Ravi- prakash et al. (1989). Figure 1 presents the genomic sequence from nt 701 to 1780. Important features are indicated in the figure legend, including the TATA box, transcription and translation start sites, splice donor and acceptor sites, stop codons, and polyadenylation signals and sites.
Identification of multiple E3 transcripts
Northern blots of poly(A)+ RNA isolated at various times p.i. were probed with the HindIll-C fragment of MAV-1. Only a single class of mRNAs, approximately 1 kb in length, was detected at early times (Fig. 2, lane 2). The 1 -kb mRNAs could first be detected at 14 hr p.i. with a probe made from an E3 cDNA clone (see below) (Fig. 2, lane 7). Smaller probes from this region of the genome were used to more precisely locate the 1-kb RNAs within the HindIll-C fragment (data not shown). Assays with strand-specific probes indicated that both at early and late times after infection mRNAs from this region were only transcribed in the rightward direction (data not shown). Taken together these experiments identified a mRNA or family of mRNAs transcribed at early times in the rightward direction near 80 m.u., con- sistent with an MAV-1 E3 transcription unit.
Figure 3 presents the structures of the E3 messages determined in the experiments described below. Sl
MOUSE ADENOVIRUS E3 TRANSCRIPTION 83
790 GACACCTTTG ACGCCGCCCT AACAAGCAAC GGAGCGCAAT TAGCTGGAGG GGCGTGGATA AACW ACGGTAGTGT TCGCTACGAA
TATA box 880
GCGCCCTTGC AGCTGGCCGA GGAACAGGTC GGTGGACCGC TAAACGCCTT TGCTATAAAA CATCAGCTAC AACTAGCAGG AGGAGCTCTT 1 1 transcription start Bites
970 TCTGCTTCTDTCCGAAAT GAGCGGGGCG CCCAGAATCC CGCGCAGCGG AGGTATTGGG TCGTGGCAAT TTTCTCGAGA ATTCCCCCCT
start codon I intron #l donor site 1060
ACTGTTTACC TTAACCCTTT TTCCGGCAGT CCTGACACTT TTCCTCATCA ATTTCTTTCT AACTATGACT CTTTCTCTCA CACGGTGGAC
1150 GGGTATGACBTTCACCGT CCAGATCGGC TGCGCTTCCT GTGCCTGCTT CTACTCGTAT TGGGTTGGTG TTTGCCCGTG ACCGGTCATC
intron #l acceptor site 1 pVII1 atop codon
1240 CTCTCAAAGG GGTTCAACCA TCGCAGTGTC AGTGCCCTGC TAGTCCCCCG TGGACTAATT CTTCTGTTAC TTCCTTCGCC CAGAAAACAA
1330 AATGGGAAAA CTCACGGTAT GTACAAGTAA GCCGTACTTA AATTTTTCTC GTGCTATACG TACGTACCTG TGCGGCTCCA AATGCGATAA
41 4A I intron #2 donor sites 1420
CGCTATCTAT TTTACACCCC AGAAAATTGT TATCGAGCTG GTGCAGGAAA AAAAAACCAC TCAGTTACTC CTTTTGCTTG CAGCCAGTAT intron #2 acceptor oites 5 1 5Al
1510 TGCCCTGTAC CTTC-GTC CTCAACTCGG GGCAAGAATG CTGTTCGAAC TGGTGCAGGC CCGGACGACG AGTGTTTC-CAGCAGCGT
clam 2 otop codon class 1 atop codon 1600
GGCTGCTGCC CTGTTTGCCT GCGCCGGAGA GGAAATAATC AACCCAGCAA TTTTTCTGTT TCTGCATGTT CTCACACTTG TGATCGTTCT
1690 GGCTATGGCC GCTGAAGTAA TCTATAATCG CTGCCGTCGT ACTACTCGAC CTACTGCACC CCCACCCCCT GTCAACAATG CTGATTTTAA
1780 CCTGGCAGAT GCCTTAGATG AAACTTAC ALUAATAAAAA TTTGCAACAC GTACTCCGGC TCGCCTCCTUTTTTCTTT GCAGAAGGAC
polyadenan oignals I polyadenylation site class 3 stop codon fiber start codon
FIG. 1. MAV-1 genomic DNA sequence from 79 to 83 mu. Important features are underlined, including transcription and translation start sites, alternate splice donor sites 4 and 4A, alternate splice acceptor sites 5 and 5A, polyadenylation signals, and polyadenylation sites. Stop codons for the class 1, 2, and 3 E3 mRNAs (see text) are indicated. All features are those of E3 messages unless otherwise noted. Nucleotide 1 corresponds to the HindIll site at 77 m.u.; for the complete sequence of the HindIll-C fragment, see Raviprakash et a/. (1989).
nuclease protection and primer extension analyses were used to map the 5 and 3 ends and splice sites. Direct sequencing of cDNA clones was used to deter- mine the precise splice sites and the 3polyadenylation site.
For 5 end mapping, both the oligonucleotide for primer extension and the Sl nuclease probe were 5 end-labeled at nt 873. The Sl nuclease and primer ex- tension products were compared on a sequencing gel (Fig. 4, lanes 5-7 and 8-10, respectively); both pro- duced fragments which mapped two closely spaced 5 ends to nt 793 and 796 in early RNA. Other larger fragments which extended up to several hundred nu- cleotides upstream were protected by the S 1 nuclease probe in late RNA (data not shown), and were not mapped precisely. A series of bands in late RNA corre- sponding to 5 ends around nt 770, the predicted early transcription TATA box, were seen in the Sl nuclease
analysis (Fig. 4, lane 7). These bands probably arise from Sl nuclease digestion at A-T-rich regions in het- eroduplexes (Hansen et al., 1981). The additional primer extension product seen at late times, which would map an end to nt 754 (Fig. 4, lane lo), is proba- bly a reverse transcriptase-induced artifact, since sim- ilar 5 ends were not seen in the Sl nuclease analysis. Strong stops were seen near nt 754 when DNA was sequenced with either avian myeloblastosis virus or Moloney murine leukemia virus reverse transcriptase, but were not seen when sequencing was performed using Sequenase (data not shown). The 3 end of E3 mRNAs was mapped to ant 1740 using a probe 3 end-labeled at nt 1450; this probe protected one 3end in both early and late RNA (Fig. 5 and data not shown).
Two introns were identified using the probes dia- gramed in Fig. 6B. For the first intron the splice donor site (site 2) was mapped using Sl nuclease analysis
84 BEARD ET AL.
:: 22 45 % 8 11141822 45hr
123 456769 10
FIG. 2. Northern blot analysis of MAV-1 transcription from 79 to 83 m.u. Poly(A)+ RNA was isolated from mock-infected or MAV-l- infected ceils from 8 to 45 hr p.i. and analyzed on Northern blots as described under Materials and Methods. Hours p.i. are indicated across the top of the figure. Lanes l-3 were probed with random primer-labeled HindIll-C fragment of MAV-1. Lanes 4-10 were probed with a random primer-labeled E3 cDNA clone of the class 1 type (see Fig. 3). Numbers between lanes 3 and 4 represent posi- tions of Ad5 DNA HindIll fragments.
and a probe 3 end-labeled at nt 827. This probe pro- tected 100 and 110 nt fragments in early RNA, map- ping a splice donor site to approximately nt 930 (Fig. 6A, lane 2). To map the first splice acceptor site (site 3), a probe was 5 end-labeled at nt 1 144. This probe protected two fragments, 65 and 70 nt in length, in early RNA (Fig. 6A, lane 5) mapping a splice acceptor site to ant 1080.
The second splice donor site (si...