No abstract

BackgroundWe applied crosslinking techniques as a first step in preparation of stable avian sarcoma virus (ASV) integrase (IN)-DNA complexes for crystallographic investigations. These results were then compared with the crystal structures of the prototype foamy virus (PFV) intasome and with published data for other retroviral IN proteins.Methodology/ResultsPhotoaffinity crosslinking and site-directed chemical crosslinking were used to localize the sites of contacts with DNA substrates on the surface of ASV IN. Sulfhydryl groups of cysteines engineered into ASV IN and amino-modified nucleotides in DNA substrates were used for attachment of photocrosslinkers. Analysis of photocrosslinking data revealed several specific DNA-protein contacts. To confirm contact sites, thiol-modified nucleotides were introduced into oligo-DNA substrates at suggested points of contact and chemically crosslinked to the cysteines via formation of disulfide bridges. Cysteines incorporated in positions 124 and 146 in the ASV IN core domain were shown to interact directly with host and viral portions of the Y-mer DNA substrate, respectively. Crosslinking of an R244C ASV IN derivative identified contacts at positions 11 and 12 on both strands of viral DNA. The most efficient disulfide crosslinking was observed for complexes of the ASV IN E157C and D64C derivatives with linear viral DNA substrate carrying a thiol-modified scissile phosphate.ConclusionAnalysis of our crosslinking results as well as published results of retroviral IN protein from other laboratories shows good agreement with the structure of PFV IN and derived ASV, HIV, and MuLV models for the core domain, but only partial agreement for the N- and C-terminal domains. These differences might be explained by structural variations and evolutionary selection for residues at alternate positions to perform analogous functions, and by methodological differences: i.e., a static picture of a particular assembly from crystallography vs. a variety of interactions that might occur during formation of functional IN complexes in solution.

A synthetic ribooligonucleotide, r(CCAGACUGm-AAGAUCUGG), corresponding to the unmodified yeast tRNA(Phe) anticodon arm is shown to bind to poly(U) programmed small ribosomal subunits of both E. coli and rabbit liver with affinity two order less than that of a natural anticodon arm. Its deoxyriboanalogs d(CCAGACTGAAGATCTGG) and d(CCAGA)r(CUGm-AAGA)d(TCTGG), are used to study the influence of sugar-phosphate modification on the interaction of tRNA with programmed small ribosomal subunits. The deoxyribooligonucleotide is shown to adopt a hairpin structure. Nevertheless, as well as oligonucleotide with deoxyriboses in stem region, it is not able to bind to 30S or 40S ribosomal subunits in the presence of ribo-(poly(U] or deoxyribo-(poly (dT) template. The deoxyribooligonucleotide also has no inhibitory effect on tRNA(Phe) binding to 30S ribosomes at 10-fold excess over tRNA. Neomycin does not influence binding of tRNA anticodon arm analogs used. Complete tRNA molecule and natural modifications of anticodon arm are considered to stabilize the arm structure needed for its interaction with a programmed ribosome.

No abstract

No abstract