Edward B. Ziff

The initiation sites for RNA transcription in Ad2 DNA

Six restriction fragments of Ad2 DNA which contain sites for RNA initiation have been identified by their ability to hybridize nascent labeled RNA less than 1 kb in length. Four RNA initiation sites early in infection are identified in regions where previous work (Flint, 1977) had mapped mRNAs. The major late RNA initiation site is the origin of a giant nuclear transcript extending from approximately 16.3 map units to (or close to) 100 units at the end of the genome. This transcription unit encompasses at least four or five mRNA sites; processing of this long transcript appears necessary to generate the mRNA.

Promoters and heterogeneous 5′ termini of the messenger RNAs of adenovirus serotype 2☆

Abstract Adenovirus serotype 2 messenger RNAs are transcribed from nine separate transcription units, which function during early, intermediate and late stages of infection. We have sequenced capped 5′ termini of messenger RNAs from each of these transcription units, and located the templates for each capped terminus in the sequence of the Ad2† genome. These results define the putative initiation sites and promoters for all major Ad2 transcripts. By comparing the termini and promoter DNA sequences, we conclude that seven of nine initiation sites including EIa, EIb, PIX, major late, EIII, EIV and the late form of region IIa mRNA have clearly defined “TATA” box homologies. With early region IIa and IVa2, this homology is absent. For most messengers, the mRNA terminus is microheterogeneous, and encoded within a two to seven-base cap region located ~30 nucleotides downstream from the TATA box. However, the two transcription units that lack TATA boxes display equivalent microheterogeneity. Purine termini are greatly favored over pyrimidine termini. We propose a model for Ad2 cap site recognition in which RNA polymerase scans short DNA segments for purine starts, and discuss the model in the context of EIV mRNAs of Ad2 and Ad5, which differ in their promoter region DNA sequence.

The major late adenovirus type-2 transcription unit: Termination is downstream from the last poly(A) site☆

Abstract Late in adenovirus type-2 infection of HeLa cells the majority of viral RNA synthesis occurs from a transcription unit that extends between 16% and 100% on the physical map of the Ad-2 † genome. Poly(A) is added to RNA at five distinct sites within the transcription unit (Fraser & Ziff, 1978; Nevins & Darnell, 1978a,b). As many as 13 different messenger RNAs can be processed from the primary RNA transcripts of this region of the genome (Chow et al., 1977a,b; Nevins & Darnell, 1978a). Examination of nuclear RNA in several different types of experiments (RNA “fingerprint” studies and hybridization of pulse-labeled “nascent” RNA and labeled nuclear RNA synthesized after ultraviolet irradiation to an ordered set of restriction endonuclease-generated DNA fragments) indicates that the mRNA 3′ terminus most distant from the promoter does not correspond to the RNA polymerase termination site for the transcription unit. Transcription, continues ~2500 to 3000 residues beyond the 3′ terminus of the most promoter distal mRNA (localized between co-ordinates 89.7 and 91.9 ) and into the region 98.2–100 on the Ad-2 physical map. The occurrence of RNA termination sites distal to the poly(A) site is discussed as a general feature of transcriptional unit design in animal cells.

RNA structures near poly(A) of adenovirus-2 late messenger RNAs.

Abstract RNA sequences near poly(A) of adenovirus-2 late messenger RNAs have been mapped within the Ad2 genome and compared by T 1 ribonuclease fingerprinting. Polyadenylated viral mRNAs transcribed 14 to 17 hours post-infection were prepared highly labeled with 32 P, and then fractionated by hybridization to restriction fragments of Ad2 DNA or by gel electrophoresis. The 3′-terminal regions of purified mRNAs were isolated by partially degrading the mRNA with alkali or T 1 ribonuclease to an average length of ~100 to 200 residues, and selecting fragments with poly(A) by poly(U)-Sepharose chromatography. Six different 3′-terminal structures, all derived from rightward transcripts, were identified. The map positions of these structures were initially deduced by establishing the fingerprints of sequences near poly(A) of mRNAs hybridized to DNA restriction fragments with known map co-ordinates. In four cases the DNA templates for the termini were located more precisely by identifying short restriction fragments of known map position which encode, in vitro , large T 1 oligo-nucleotides found in vivo near poly(A). Five 3′ termini map within restriction fragments Sma1 I/ Hin III J (co-ordinates 37.3 to 40.5), Sma1 D/ Kpn D (47.4 to 52.6), Eco RI B (58.5 to 70.7), Eco RI D (75.9 to 83.4), and Sma1 C/ Eco R1 C (89.7 to 91.9). These 3′ termini fall within the boundaries of the major Ad2 late transcription unit (co-ordinates ~16 to 100; Evans et al. , 1977). The sixth terminus maps outside this region in Sma1 M (11.2 to 11.6) and thus is transcribed from a second transcription unit which also functions late in infection. Each of the six 3′ termini yielded large T 1 oligonucleotides which were specific to the particular 3′-terminal region, rather than common to different structures. The hexanucleotide sequence A-A-U-A-A-A, which is located near poly(A) of several previously studied eukaryotic messengers (Proudfoot & Brownlee, 1976), is present in at least two of the Ad2 3′-terminal structures, and potentially in all six. The RNA sequences surrounding the hexanucleotide are not identical in the different viral cases, although these mRNAs are transcribed from a single DNA genome. The location of Ad2 mRNA 3′-terminal sequences within the viral genome is discussed in terms of the Ad2 late mRNA map and a hypothesis for the role of 3′ terminus formation in Ad2 mRNA synthesis.

Poly(A) sites of adenovirus serotype 2 transcription units

Abstract During productive infection of human cells, adenovirus serotype 2 transcription proceeds through early and late stages of transcription, which are separated by the onset of DNA replication. In this paper we analyze the 3′-terminal sequences of messenger RNAs from five early transcription units, and messenger RNA from the L2, L3 and L4 families of the major late transcription unit. By correlating RNAase T1 oligonucleotides near poly(A) with the DNA sequence of Ad-2 ∥ we have defined the poly(A) acceptor sequences of these messengers. Our principal conclusions are as follows. 1. (1)The A-A-U-A-A-A hexanucleotide homology common to many cellular mRNA 3′ ends is present near poly(A) of all but one of the mRNAs studied. Notably, it is present within the major late transcription unit at the L2, L3 and L4 coterminal family poly(A) sites, and potentially at all five poly(A) sites of the late transcription unit. Because the latter mRNA 3′ ends are formed by cleavage, the hexanucleotide probably does not specify a RNA polymerase II termination site. 2. (2) An early region III poly(A) site apparently lacks the A-A-U-A-A-A homology and may be an exception to the above generalizations. 3. (3) Region EIb and protein IX mRNA share a common poly(A) site, although their precursors are initiated at separate promoters. Also, the same poly(A) sites at the left end function during the early and late stages of infection. Thus, these poly(A) sites are not regulated in a stage-specific manner, and may function for precursors initiated at more than one promoter. 4. (4) Residues near poly(A) of mRNA belonging to the 3′ coterminal families of the late transcription unit form ribonuclease-resistant hybrids to Ad-2 DNA which encode mRNA main body sequences. Thus late mRNA 3′ ends are locally encoded. We present a model for sequence recognition during formation of 3′ ends, and discuss autonomy of poly(A) site function.

TRANSCRIPTION OF ADENOVIRUS DNA IN INFECTED CELL EXTRACTS1

ABSTRACT The activity of the nine known adenovirus promoter sites has been studied in vitro in a whole cell extract system. Six sites (four early, one intermediate (PIX) and the major late promoters) functioned with comparable efficiency in uninfected extracts. The other early promoter (for early region II) was utilized only poorly. Two intermediate promoters (leftward at 15.9 and 72.0 map units) which function in vivo only at intermediate or late stages in the lytic cycle, were inactive in uninfected extracts. Although extracts prepared from late infected cells did not show transcription of these two promoters, these extracts did show an early to late shift in vitro . In late infected extracts the late and PIX promoters were enhanced five-to ten-fold relative to early promoters. DNA titration and extract mixing experiments indicated the presence in uninfected and infected extracts of factors which could distinguish between early and late promoter classes.