The DNA binding domain of a papillomavirus E2 protein programs a chimeric nuclease to cleave integrated human papillomavirus DNA in HeLa cervical carcinoma cells

Viral DNA binding proteins that direct nucleases or other protein domains to viral DNA in lytically or latently infected cells may provide a novel approach to modulate viral gene expression or replication. Cervical carcinogenesis is initiated by high-risk human papillomavirus (HPV) infection, and viral DNA persists in the cancer cells. To test whether a DNA binding domain of a papillomavirus protein can direct a nuclease domain to cleave HPV DNA in cervical cancer cells, we fused the DNA binding domain of the bovine papillomavirus type 1 (BPV1) E2 protein to the catalytic domain of the FokI restriction endonuclease, generating a BPV1 E2-FokI chimeric nuclease (BEF). BEF introduced DNA double-strand breaks on both sides of an E2 binding site in vitro, whereas DNA binding or catalytic mutants of BEF did not. After expression of BEF in HeLa cervical carcinoma cells, we detected cleavage at E2 binding sites in the integrated HPV18 DNA in these cells and also at an E2 binding site in cellular DNA. BEF-expressing cells underwent senescence, which required the DNA binding activity of BEF, but not its nuclease activity. These results demonstrate that DNA binding domains of viral proteins can target effector molecules to cognate binding sites in virally infected cells.     

Sequence-specific deoxyribonucleic acid (DNA) recognition by steroidogenic factor 1: a helix at the carboxy terminus of the DNA binding domain is necessary for complex stability

Steroidogenic factor 1 (SF1) is a member of the NR5A subfamily of nuclear hormone receptors and is considered a master regulator of reproduction because it regulates a number of genes encoding reproductive hormones and enzymes involved in steroid hormone biosynthesis. Like other NR5A members, SF1 harbors a highly conserved approximately 30-residue segment called the FTZ-F1 box C-terminal to the core DNA binding domain (DBD) common to all nuclear receptors and binds to 9-bp DNA sequences as a monomer. Here we describe the solution structure of the SF1 DBD in complex with an atypical sequence in the proximal promoter region of the inhibin-alpha gene that encodes a subunit of a reproductive hormone. SF1 forms a specific complex with the DNA through a bipartite motif binding to the major and minor grooves through the core DBD and the N-terminal segment of the FTZ-F1 box, respectively, in a manner previously described for two other monomeric receptors, nerve growth factor-induced-B and estrogen-related receptor 2. However, unlike these receptors, SF1 harbors a helix in the C-terminal segment of the FTZ-F1 box that interacts with both the core DBD and DNA and serves as an important determinant of stability of the complex. We propose that the FTZ-F1 helix along with the core DBD serves as a platform for interactions with coactivators and other DNA-bound factors in the vicinity.     

Isolation of a cDNA encoding the adenovirus E1A enhancer binding protein: a new human member of the ets oncogene family.

The cDNA encoding adenovirus E1A enhancer-binding protein E1A-F was isolated by screening a HeLa cell lambda gt11 expression library for E1A-F site-specific DNA binding. One cDNA clone produced recombinant E1A-F protein with the same DNA binding specificity as that endogenous to HeLa cells. Sequence analysis of the cDNA showed homology with the ETS-domain, a region required for sequence-specific DNA binding and common to all ets oncogene members. Analysis of the longest cDNA revealed about a 94% identity in amino acids between human E1A-F and mouse PEA3 (polyomavirus enhancer activator 3), a recently characterized ets oncogene member. E1A-F was encoded by a 2.5kb mRNA in HeLa cells, which was found to increase during the early period of adenovirus infection. In contrast, ets-2 mRNA was significantly reduced in infected HeLa cells. The results indicate that E1A enhancer binding protein E1A-F is a member of the ets oncogene family and is probably a human homologue of mouse PEA3.     

Identification and partial purification of a protein binding to the human immunoglobulin kappa enhancer kappa E2 site.

We previously described a domain in the 5'' half of the human immunoglobulin kappa enhancer which could bind nuclear proteins in vitro, as detected by a lambda exonuclease protection assay. A second more 3'' binding domain in the enhancer has now been detected by a similar assay employing a different exonuclease, the T7 gene 6 exonuclease. Using this assay and starting with a pig spleen nuclear extract, we have purified 5000-fold a protein that binds to the 3'' domain. In a DNase I footprint experiment the partially purified protein protects a 27 bp segment in the enhancer centered around the sequence CAGGTGGC, which corresponds to the kappa E2 sequence motif described in the mouse kappa enhancer. The protein, designated NF-kappa E2, also appears to bind at a position downstream of kappa E2, at or near the kappa E3 site. Proteins capable of binding at kappa E2 are found in several mammalian species and are expressed in both lymphoid and non-lymphoid tissues.     

The protein Id: a negative regulator of helix-loop-helix DNA binding proteins

We have isolated a cDNA clone encoding a novel helix-loop-helix (HLH) protein, Id. Id is missing the basic region adjacent to the HLH domain that is essential for specific DNA binding in another HLH protein, MyoD. An in vitro translation product of Id can associate specifically with at least three HLH proteins (MyoD, E12, and E47) and attenuate their ability to bind DNA as homodimeric or heterodimeric complexes. Id is expressed at varying levels in all cell lines tested. In three cell lines that can be induced to undergo terminal differentiation, Id RNA levels decrease upon induction. Transfection experiments indicate that over-expression of Id inhibits the trans-activation of the muscle creatine kinase enhancer by MyoD. Based on these findings, we propose that HLH proteins lacking a basic region may negatively regulate other HLH proteins through the formation of nonfunctional heterodimeric complexes.     

Amyloid-like properties of a synthetic peptide corresponding to the carboxy terminus of beta-amyloid protein precursor.

A synthetic peptide whose sequence corresponds to the 20 carboxy-terminal amino acids of beta-amyloid protein precursor (APP) was found to form fibrils in vitro. These fibrils showed birefringence in polarized light when stained with Congo red, fluoresced when bound with thioflavin S, were resistant to proteases, and had a cross-beta conformation. By contrast, peptides with other sequences from the intracellular domain of APP and a peptide corresponding to this entire domain did not exhibit the full range of beta-amyloid properties. These results suggest that a fragment from the C-terminus of the beta-amyloid protein precursor could bind to intraneuronal paired helical filaments and account for some of its amyloid-like properties.     

Identification of the human beta-actin enhancer and its binding factor.

An enhancer of the human beta-actin gene and a factor that specifically interacts with it were detected. A mobility shift assay showed that the factor bound to the 25-base-pair sequence (between +759 and +783 downstream from the cap site) with high specificity. This finding correlated with those of DNase I protection and exonuclease III digestion assays. This binding region of the beta-actin enhancer contained a hyphenated dyad symmetry and an enhancer core-like sequence. In vitro competition experiments indicated that the factor did not bind to the simian virus 40 enhancer core region.     

Lysine-dependent multipoint binding of the Borrelia burgdorferi virulence factor outer surface protein E to the C terminus of factor H

Serum resistance, an important virulence determinant of Borrelia burgdorferi sensu lato strains belonging to the Borrelia afzelii and B. burgdorferi sensu stricto genotypes, is related to binding of the complement inhibitor factor H to the spirochete surface protein outer surface protein E (OspE) and its homologues. In this study, we show that the C-terminal short consensus repeats 18-20 of both human and mouse factor H bind to OspE. Analogously, factor H-related protein 1, a distinct plasma protein with three short consensus repeat domains homologous to those in factor H, bound to OspE. Deleting 15-aa residues (region V) from the C terminus of the OspE paralog P21 (a 20.7-kDa OspE-paralogous surface lipoprotein in the B. burgdorferi sensu stricto 297 strain) abolished factor H binding. However, C-terminal peptides from OspE, P21, or OspEF-related protein P alone and the C-terminal deletion mutants of P21 inhibited factor H binding to OspE only partially when compared with full-length P21 or its N-terminal mutant. Alanine substitution of amino acids in peptides from the key binding regions of the OspE family indicated that several lysine residues are required for factor H binding. Thus, the borrelial OspE family proteins bind the C inhibitor factor H via multiple sites in a lysine-dependent manner. The C-terminal site V (Ala(151)-Lys(166)) is necessary, but not sufficient, for factor H binding in both rodents and humans. Identification of the necessary binding sites forms a basis for the development of vaccines that block the factor H-OspE interaction and thereby promote the killing of Borreliae.     

Specific DNA binding to a major histocompatibility complex enhancer sequence by a synthetic 57-residue double zinc finger peptide from a human enhancer binding protein

Two 57-residue peptides containing one pair of "zinc fingers" from a human enhancer binding protein were prepared by solid-phase peptide synthesis. One peptide (MBP-DF) contained the native sequence, while the second peptide ([Abu11]MBP-DF) has an alpha-aminobutyric acid residue substituted for a nonconserved cysteine residue at position 11. The peptides were characterized by several chemical and physical methods, and their DNA binding properties were evaluated using gel retardation experiments. Spectroscopic studies demonstrated that addition of metal ions such as zinc and cobalt resulted in specific conformational changes in both peptides, indicating that cysteine-11 does not appear to be involved in metal chelation. One-dimensional 1H NMR studies indicate that a stable folded structure is formed upon addition of zinc, and the chemical shift pattern is consistent with that previously observed for one constituent single finger (Omichinski, J., Clore, G. M., Appella, E., Sakaguchi, K., and Gronenborn, A. M. (1990) Biochemistry 29, 9324-9334). Gel retardation experiments demonstrate that the peptides are capable of interacting with a 15-mer oligonucleotide comprising a portion of the major histocompatibility complex enhancer sequence and that the interaction is zinc-dependent. The dissociation constant for the [Abu11]MBP-DF peptide is 1.4 x 10(-7) M with maximal binding occurring at a zinc-to-peptide ratio of 2 to 1. The binding specificity observed with respect to related enhancer sequences exhibits the same relative order as noted previously for the whole protein. Studies with point mutants of the major histocompatibility complex enhancer binding sequence indicate that the last GC base pair in a four-guanine stretch plays a pivotal role in the interaction between the peptide and DNA.