The adenovirus L4 33-kilodalton protein binds to intragenic sequences of the major late promoter required for late phase-specific stimulation of transcription

Abstract

The adenovirus late IVa2 protein is required for maximally efficient transcription from the viral major late (ML) promoter, and hence, the synthesis of the majority of viral late proteins. This protein is a sequence-specific DNA-binding protein that also promotes the assembly of progeny virus particles. Previous studies have established that a IVa2 protein dimer (DEF-B) binds specifically to an intragenic ML promoter sequence necessary for late phase-specific stimulation of ML transcription. However, activation of transcription from the ML promoter correlates with binding of at least one additional infected-cell-specific protein, termed DEF-A, to the promoter. Using an assay for the DNA-binding activity of DEF-A, we identified the unknown protein by using conventional purification methods, purification of FLAG-tagged IVa2-protein-containing complexes, and transient synthesis of viral late proteins. The results of these experiments established that the viral L4 33-kDa protein is the only component of DEF-A: the IVa2 and L4 33-kDa proteins are necessary and sufficient for formation of all previously described complexes in the intragenic control region of the ML promoter. Furthermore, the L4 33-kDa protein binds to the promoter with the specificity characteristic of DEF-A and stimulates transcription from the ML promoter in transient-expression assays.

FIG. 1.

Electrophoretic mobility shift assay for IVa₂ and DEF-A. (A) The intragenic ML binding sites for the IVa₂ protein dimer (DEF-B) and DEF-A are illustrated schematically but to scale. (B) A ³²P-labeled double-stranded DNA containing these binding sites was incubated with no protein (0), whole-cell extracts prepared from mock-infected HeLa cells (U), or Ad5-infected cells harvested 24 h after infection (I). The protein-DNA complexes were examined as described in Materials and Methods. The infected-cell-specific a, b, and c complexes (44) are indicated on the right.

FIG. 2.

Purification of DEF-A. (A) Summary of the purification scheme. (B) Electrophoretic mobility shift assay of fractions recovered from the Superdex S-200 column. The positions of the infected-cell-specific complexes are indicated on the right. The positions of the two peaks of DEF-A activity described in the text are indicated below the fraction numbers. Fractions 62 to 69 were pooled and purified by affinity chromatography on DNA containing the ML intragenic promoter sequences. (C) Electrophoretic mobility shift assay of the pooled S-200 fractions (S) and fractions recovered during DNA affinity chromatography (U, unbound proteins; W, proteins recovered in 0.1 M wash buffer; E, proteins recovered in 1.0 M NaCl). (D) Silver staining of the DNA affinity chromatography fractions, indicated as in panel C, following separation by electrophoresis in a 10% SDS-polyacrylamide gel. Prestained molecular mass markers, as listed at the left, were loaded in the lane marked M. The positions of the IVa₂ and enriched 33-kDa proteins are indicated on the right. (E) The same DNA affinity chromatography fractions were examined by immunoblotting using monoclonal and polyclonal antibodies specific for the IVa₂ and L4 33-kDa proteins, respectively.

FIG. 3.

Copurification of DEF-A and the L4 33-kDa protein by binding to a FLAG-tagged IVa₂ protein. Nuclear extracts prepared from mock- or Ad5-infected 293FT cells stably synthesizing the IVa₂-FLAG protein were subjected to anti-FLAG immunoaffinity chromatography as described in Materials and Methods. (A) The Ad5-infected cell extract (NE), wash (W), and immunoprecipitated proteins (IP) were assayed for proteins that bound to the ML DE sequence. The positions of the b and c complexes are indicated on the right. (B) The nuclear extract and immunoprecipitated proteins (IP) from the experiment shown in panel A (Ad5) or from mock-infected cells (Mock) were examined by immunoblotting them with polyclonal antibodies against the L4 33-kDa protein.

FIG. 4.

Inhibition of binding of DEF-A to the ML promoter by antibodies against the L4 33-kDa protein. A whole-cell extract prepared from HeLa cells 24 h after Ad5 infection was preincubated on ice for 15 min with no (−) antibody or with (+) the antibody indicated at the top. Binding of proteins to the ML DE sequence was then assayed. The positions of the b and c complexes are indicated on the right.

FIG. 5.

The IVa₂ and L4 33-kDa proteins reconstitute DEF-A DNA-binding activity. (A, top) The structures of unspliced and spliced L4 33-kDa RNAs are depicted to scale, with the positions of the RT-PCR primers (not to scale) and the predicted products indicated below. As these two primers were also used for cloning L4 coding sequences, they added 37 bp to the predicted RT-PCR products. (Bottom) DNA products made in reactions that contained Ad2 DNA (Ad2) or cytoplasmic (C) or nuclear (N) RNA purified from 293FT cells containing the L4 33-kDa expression vector. Reactions that contained or lacked reverse transcriptase are indicated by + and −, respectively. A 100-bp DNA ladder (Invitrogen) was run in lane 1. The product of spliced L4 33-kDa RNA is indicated on the right. (B) Nuclear extracts were prepared from 293FT cells (lane 3) or from these cells 24 h after the introduction of vectors directing synthesis of the IVa₂ protein (lane 4), the L4 33-kDa protein (lane 5), or both vectors (lane 6) and assayed for binding to the intragenic ML promoter sequence. The reactions whose products are shown in lanes 2 and 7 contained an Ad5-infected cell nuclear extract (I) and both the IVa₂-containing and the L4 33-kDa-protein-containing extracts, respectively. (C) Reactions contained the extract isolated from cells producing both the IVa₂ and the L4 33-kDa proteins and no competitor (lane 1), a 30-fold molar excess of DE m1 competitor that retained the wild-type DE2 sequence (lane 2), or a competitor (DE m12ab) in which all three DE sequences were mutated (lane 3). (D) Binding reactions contained nuclear-extract proteins from 293FT cells synthesizing the viral proteins indicated at the top (lanes 3 to 6), no protein (lane 1), or Ad5-infected nuclear extract (I; lane 2) and ³²P-labeled DEm1 DNA. This probe was identical to the wild-type probe, except for the presence of DE1 (Fig. 1A) substitutions that block binding of DEF-A (37).

FIG. 6.

DEF-A DNA-binding activity is associated with a hypophosphorylated form of the L4 33-kDa protein. Whole-cell extracts were prepared from HeLa cells infected with Ad5 for the times indicated or mock infected (M) and examined for proteins that bind to the ML DE sequence (A) or by immunoblotting for the L4 33-kDa protein (B). The positions of the a, b, and c complexes are indicated on the right of panel A. Extracts prepared from 293FT cells transiently synthesizing the L4 33-kDa protein were loaded onto the lanes marked L4 in panel B before (−) or after (+) treatment with 400 units phosphatase (PPase) for 1 h at 30°C, followed by incubation for 1 h at 65°C. (C) Proteins present in the 24-h-infected cell extract (I) and that prepared from L4 33-kDa protein-producing 293FT cells (L4) shown in panel A were separated by electrophoresis in 15% SDS-polyacrylamide gel electrophoresis and examined by immunoblotting them with polyclonal antibodies against the L4 33-kDa protein. (D) The infected cell extract (I) was examined as described above before (−) and after (+) treatment with phosphatase. p.i., postinfection.

FIG. 7.

The L4 33-kDa protein stimulates transcription from the ML promoter. (A) The organization of the reporter gene is shown schematically, with the ML promoter and EGFP coding sequences represented by the solid line and open box, respectively. The positions of the site of initiation of ML transcription (gray arrow), the DE sequence (gray box), and the primer for reverse transcription (dashed arrow) are indicated. (B) Ten micrograms of the plasmid containing the ML-GFP reporter gene were introduced into 293FT alone (−) or with (+) 10 μg of IVa₂ or L4 33-kDa expression plasmid, as indicated at the top. Total RNA was purified 27 h later and analyzed by RT-PCR with primers specific for the ML sequence of the reporter gene (ML) or human β-actin as a control for cell number and RNA recovery.

FIG. 8.

The L4 33-kDa protein interacts with a packaging-signal repeat sequence only in the presence of the IVa₂ protein. Binding reactions contained no protein (0), nuclear extract prepared from HeLa cells infected with Ad5 for 24 h (I), or extracts isolated from 293FT cells synthesizing the viral proteins indicated at the top and 1 ng of the AI-AII packaging-repeat sequence labeled as described in Materials and Methods. A 30-fold molar excess of the unlabeled DE m1 and DE m12ab competitor DNAs described in the legend to Fig. 5 were included in the reactions whose products are shown in lanes 7 and 8, respectively.