Human DNA helicase II (HDH II) is a novel ATP‐dependent DNA unwinding enzyme, purified to apparent homogeneity from HeLa cells, which (i) unwinds exclusively DNA duplexes, (ii) prefers partially unwound substrates and (iii) proceeds in the 3′ to 5′ direction on the bound strand. HDH II is a heterodimer of 72 and 87 kDa polypeptides. It shows single‐stranded DNA‐dependent ATPase activity, as well as double‐stranded DNA binding capacity. All these activities comigrate in gel filtration and glycerol gradients, giving a sedimentation coefficient of 7.4S and a Stokes radius of approximately 46 A, corresponding to a native molecular weight of 158 kDa. The antibodies raised in rabbit against either polypeptide can remove from the solution all the activities of HDH II. Photoaffinity labelling with [alpha‐32P]ATP labelled both polypeptides. Microsequencing of the separate polypeptides of HDH II and cross‐reaction with specific antibodies showed that this enzyme is identical to Ku, an autoantigen recognized by the sera of scleroderma and lupus erythematosus patients, which binds specifically to duplex DNA ends and is regulator of a DNA‐dependent protein kinase. Recombinant HDH II/Ku protein expressed in and purified from Escherichia coli cells showed DNA binding and helicase activities indistinguishable from those of the isolated protein. The exclusively nuclear location of HDH II/Ku antigen, its highly specific affinity for double‐stranded DNA, its abundance and its newly demonstrated ability to unwind exclusively DNA duplexes, point to an additional, if still unclear, role for this molecule in DNA metabolism.

No abstract

The repetitive sequence (AGGGCCCTAGAGGGGCCC-TAG)n was previously shown to be curved by gel mobility assays. Here we show, using hydroxy radical/DNase I digestion and differential helical phasing experiments that the curvature is directed towards the major groove and is located in the GGGCCC, but not the CTAGAG segments. The effect of the GC step in the context of the GGGCCC motif is apparently about as large as that of AA/TT, i.e. enough to cancel the macroscopic curvature of helically phased A-tracts. These data are in agreement with positive roll-like curvature of the GCC/GGC motif, predicted from nucleosome packing data and the 3D structure of the GGGGCCCC octamer, but they are not in agreement with the dinucleotide-based roll angle values predicted for AG/CT, TA, GG/CC and GC steps. Our results thus indicate the importance of interactions beyond the dinucleotide steps in predictive models of DNA curvature.

No abstract

No abstract