Recombinant human adenoviruses containing hybrid adenovirus type 5 (Ad5)/Ad12 E1A genes: characterization of hybrid E1A proteins and analysis of transforming activity and host range

Abstract

Hybrid adenovirus type 12 (Ad12)/Ad5 E1A genes were constructed by homologous recombination in Escherichia coli, a technique which offers several advantages over conventional mutagenesis for genetic analysis of proteins. In particular, functional differences between the proteins can be mapped by correlating the replacement of specific sequences with the acquisition of new properties, and there is no requirement for common unique restriction sites or polymerase chain reaction strategies to construct the hybrids. Recombinant adenoviruses expressing these hybrid E1A proteins were capable of replicating efficiently in HeLa cells, with the exception of one construct which contained a hybrid transactivation domain. The transforming activity of the hybrid E1A constructs was assayed by DNA transfection of primary baby rat kidney cells. Plasmids containing Ad12 E1 were approximately 20-fold less efficient at transformation than those with E1 of Ad5, and it was found that two regions in exon 1 of E1A mediate this difference. No differences were found in the abilities of any hybrid E1A proteins to bind to cellular proteins previously determined to be important for transformation by E1A.

E1A 12S and 13S of the transformation-defective adenovirus type 12 strain CS-1 inactivate proteins of the RB family, permitting transactivation of the E2F-dependent promoter

“…As another possible RB-independent mechanism for E1A proteins to transform cells, the binding of Ad5 E1A to the coactivator protein p300 was recently described to correlate with E1A-induced transformation (51). Moreover, it has been shown that the large E1A proteins of Ad12 bind p300 directly and with the same affinity as those of Ad5 (15,27,58). Therefore, we tested the ability of the 13S protein of the transformation-defective Ad12-CS-1 mutant expressed as a GST fusion protein to bind p300 using A549 nuclear extracts; for comparison, we also tested wild-type 13S and ⌬CR2 mutant 13S proteins.…”

“…occurs independent of the RB binding sites, suggesting a different, RB-independent mechanism of transcription regulation and E1A-induced transformation. It has been shown before that despite some variation in the p300 binding sequences of Ad5 E1A and Ad12 E1A, the large E1A proteins of Ad12 bind p300 directly and with the same affinity as those of Ad5 (15,27,58). Since in Ad12 both wild-type E1A and CS-1 mutant E1A are identical in the p300 binding sequence, it is unlikely that the mutant E1A protein does not bind p300.…”

“…Previous studies by Jelinek et al (28) indicated that cells transformed by hybrid E1A genes encoding the amino-terminal Ad12 residues up to amino acid 80, including CR1, were nontumorigenic, whereas cells transformed by hybrids encoding Ad12 E1A sequences from the amino terminus up to the leftmost border of CR3 or beyond were tumorigenic. Moreover, their findings implied that CR2 of Ad12 E1A, which is critical for the transforming activity of both E1A proteins, is not involved in the differential transforming efficiencies of Ad5 and Ad12 and that cellular proteins whose interactions with CR2 of Ad5 E1A are critical for transformation interact equally well with CR2 of Ad5 E1A and Ad12 E1A (27), further supporting our idea that the E1A transforming function may be localized in E1A regions different from known cellular protein binding sites (28). In this context, the Ad12 spacer region bordered by CR2 and CR3 in concert with Ad12 E1A sequences to the left of CR2 is of particular interest and was discussed as an important determinant of viral tumorigenicity (28,55).…”

“…In primary rodent cells, stable E1A protein expression leads to immortalization (18,25) and, in cooperation with the Ad early region 1B (E1B) gene product (53,57) or a cellular oncoprotein, such as Ras (8), to complete oncogenic transformation. Mutational analysis of the sequences of E1A proteins from different serotypes revealed two highly conserved regions, 1 and 2 (CR1 and CR2), which are common in both E1A proteins and the amino terminus responsible for the transforming activity (27,29,54,60). A third conserved region (CR3), which is unique to the 13S product, is not essential for transformation (41).…”

“…The regulatory functions of both E1A oncogene products appear to derive largely from physical interactions with cellular proteins such as the "pocket-protein" family members p105-RB, p107, and p130 (19,21,63), and the same regions of the E1A proteins which are involved in transformation are also essential for binding to these host cell proteins (16,27,39,60,61). p105-RB, p107, and p130 all interact primarily with a conserved motif located in transforming domain 2 of adenovirus type 5 (Ad5), which is required for growth activation (26,39), whereas sequences within CR1 and the amino terminus are capable of weak interactions and appear to function cooperatively with CR2 to bind these proteins (2).…”