Enzymatic processing of uracil glycol, a major oxidative product of DNA cytosine.

A major stable oxidation product of DNA cytosine is uracil glycol (Ug). Because of the potential of Ug to be a strong premutagenic lesion, it is important to assess whether it is a blocking lesion to DNA polymerase as is its structural counterpart, thymine glycol (Tg), and to evaluate its pairing properties. Here, a series of oligonucleotides containing Ug or Tg were prepared and used as templates for a model enzyme, Escherichia coli DNA polymerase I Klenow fragment (exo-). During translesion DNA synthesis, Ug was bypassed more efficiently than Tg in all sequence contexts examined. Furthermore, only dAMP was incorporated opposite template Ug and Tg and the kinetic parameters of incorporation showed that dAMP was inserted opposite Ug more efficiently than opposite Tg. Ug opposite G and A was also recognized and removed in vitro by the E. coli DNA repair glycosylases, endonuclease III (endo III), endonuclease VIII (endo VIII), and formamidopyrimidine DNA glycosylase. The steady state kinetic parameters indicated that Ug was a better substrate for endo III and formamidopyrimidine DNA glycosylase than Tg; for endonuclease VIII, however, Tg was a better substrate.

A common mechanism of action for the N-glycosylase activity of DNA N-glycosylase/AP lyases from E. coli and T4.

Duplex oligonucleotides containing the base lesion analogs, O-methylhydroxylamine- and O-benzylhydroxylamine-modified abasic (AP) sites, were substrates for the DNA N-glycosylases endonuclease III, formamidopyrimidine DNA N-glycosylase and T4 endonuclease V. These N-glycosylases are known to have associated AP lyase activities. In contrast, uracil DNA N-glycosylase, a simple N-glycosylase which does not have an associated AP lyase activity, was unable to recognize the modified AP sites. Endonuclease III, formamidopyrimidine DNA N-glycosylase and T4 endonuclease V recognized the base lesion analogs as N-glycosylases generating intermediary AP sites which were subsequently cleaved by the enzyme-associated AP lyase activities. Kinetic measurements showed that O-alkoxyamine-modified AP sites were poorer substrates than the presumed physiological substrates. For endonuclease III, DNA containing O-methylhydroxyl-amine or O-benzylhydroxylamine was recognized at 12 and 9% of the rate of DNA containing thymine glycol, respectively, under subsaturating substrate concentrations (as determined by relative Vmax/K(m)). Similarly, with formamidopyrimidine DNA N-glycosylase and T4 endonuclease V. DNA containing O-methylhydroxylamine or O-benzylhydroxylamine was recognized at 4-9% of the efficiency of DNA containing N7-methyl formamidopyrimidine or pyrimidine cyclobutane dimers, respectively. Based on the known structures of these base lesion analogs and the substrate specificities of the N-glycosylases, a common mechanism of action is proposed for DNA N-glycosylases with an associated AP lyase activity.

Uracil DNA N-glycosylase distributively interacts with duplex polynucleotides containing repeating units of either TGGCCAAGCU or TGGCCAAGCTTGGCCAAGCU.

Uracil DNA N-glycosylase (UDG) has been used as a model enzyme to test a novel universal approach to discriminate between two possible enzymatic mechanisms of specific site location in DNA, processive (DNA-scanning mechanism) and distributive (random diffusion-mediated mechanism). Two double-stranded concatemeric polynucleotides of defined length (440-480 nucleotides) containing deoxyuridine at either every 10th or 20th nucleotide in the DNA chain were prepared by the ligation of self-complementary 10- or 20-mer oligodeoxyribonucleotides. Incubation of these polynucleotides with Escherichia coli UDG, followed by thermal breakage of the abasic sites, formed fragments that were multiples of either the 10- or the 20-mer. Since the processive and distributive mechanisms of uracil removal by UDG would be very different, the fragment distribution, generated at each time interval during the UDG reaction, should be unique. To show this, we developed a computer model illustrating both possible mechanisms of UDG functioning. The distribution of DNA fragments experimentally generated during the time course of the UDG reaction was compared with the results of the computer programs that modeled the distributive and processive mechanisms. The data indicated that uracil removal, catalyzed by UDG, is consistent with a distributive model.

Amino acid sequence of a platelet-binding human anti-DNA monoclonal autoantibody.

The complete amino acid sequences of the variable regions of the heavy and light chains of a human IgM monoclonal platelet-binding autoantibody have been determined. This antibody, HF2-1/17, produced by a human x human hybridoma prepared from lymphocytes of a patient with systemic lupus erythematosus and thrombocytopenia, is polyreactive with single-stranded DNA, synthetic polynucleotides, sulfated carbohydrates, and acidic glycolipids isolated from platelet membranes. The heavy chain is of the VHIII subgroup, and the light chain is of the VKI subgroup. The heavy chain is the expression product of the VH26 germline gene. The light chain bears significant homology to other immunoglobulins of known primary structure, including WEA, GAL, HAU, HK101, and DEE. These results suggest that HF2-1/17 may be an autoantibody derived with little or no modification from germline genes. A model of the antibody combining site suggests that arginine 24 and arginine 30 in the light chain (CDR1) interact with a surface defined by phosphate or sulfate groups of the antigen.

Homology of the NH2-terminal amino acid sequences of the heavy and light chains of human monoclonal lupus autoantibodies containing the dominant 16/6 idiotype.

The NH2-terminal amino acid sequences have been determined by automated Edman degradation for the heavy and light chains of five monoclonal IgM anti-DNA autoantibodies that were produced by human-human hybridomas derived from lymphocytes of two patients with systemic lupus erythematosus. Four of the antibodies were closely related to the idiotype system 16/6, whereas the fifth antibody was unrelated idiotypically. The light chains of the 16/6 idiotype-positive autoantibodies (HF2-1/13b, HF2-1/17, HF2-18/2, and HF3-16/6) had identical amino acid sequences from residues 1 to 40. Their framework structures were characteristic of VKI light chains. The light chain of the 16/6 idiotype-negative autoantibody HF6-21/28 was characteristic of the VKII subgroup. The heavy chains of the 16/6 idiotype-positive autoantibodies had nearly identical amino acid sequences from residues 1 to 40. The framework structures were characteristic of the VHIII subgroup. In contrast, the GM4672 fusion partner of the hybridoma produced small quantities of an IgG with a VHI heavy chain and a VKI light chain. The heavy chains of the lupus autoantibodies and the light chains of those autoantibodies that were idiotypically related to the 16/6 system had marked sequence homology with WEA, a Waldenstrom IgM that binds to Klebsiella polysaccharides and expresses the 16/6 idiotype. These results indicate a striking homology in the amino termini of the heavy and light chains of the lupus autoantibodies studied and suggest that the V regions of the heavy and light chains of the 16/6 idiotype-positive DNA-binding lupus auto-antibodies are each encoded by a single germ line gene.

Primary structure of macromomycin, an antitumor antibiotic protein.

The antitumor protein macromomycin is a single chain polypeptide of 112 amino acid residues cross-linked by two intramolecular disulfide bonds. The protein was reduced and S-alkylated with 2-mercaptoethanol in 8 M urea followed by treatment with iodoacetic acid. Tryptic digestion of tetra-S-carboxymethyl macromomycin gave four tryptic peptides which were fractionated by gel permeation on Sephadex G-50. The amino acid sequence of the tryptic peptides and the overlap sequences were determined by a combination of automated Edman degradation analysis, gas chromatographic mass spectrometry, and fast atom bombardment mass spectrometry. A comparison of the structures of macromomycin, actinoxanthin, and neocarzinostatin suggests that they belong to a family of related proteins.

Metabolism of parathyroid hormone by Kupffer cells: analysis by reverse-phase high-performance liquid chromatography.

Parathyroid hormone (PTH) undergoes rapid proteolysis in the liver, which results in the appearance of multiple COOH-terminal fragments in plasma. Using reverse-phase high-performance liquid chromatographic (HPLC) techniques, we have shown that biologically active bovine PTH (bPTH) internally labeled with [3H]tyrosine is, like 125I-labeled bPTH, rapidly metabolized by isolated rat Kupffer cells in vitro to multiple COOH-terminal fragments that are chemically identical with those previously found in plasma after metabolism in vivo. Quantitation of specific carboxyl fragments in crude mixtures is achieved rapidly by direct HPLC analysis and is as precise as that achieved by Edman degradation. In addition, several different carboxyl fragments with identical NH2 termini were resolved, revealing a complexity not apparent in previous studies employing direct Edman degradation of such mixtures. Parallel studies with [[35S]Met]bPTH show the generation, by the Kupffer cells in vitro, of several labeled NH2-terminal fragments which undergo rapid further degradation in vitro. Thus, hepatic metabolism of PTH by Kupffer cells proceeds by an initial endopeptidase cleavage within the hormonal sequence in a manner compatible with the generation of biologically active NH2 fragments.

Primary sequence of two regions of mouse pro-adrenocorticotropin/endorphin.

The amino acid sequences of two previously uncharacterized regions of the mouse anterior pituitary common precursor to adrenocorticotropin (ACTH) and beta-endorphin (pro-ACTH/endorphin) were determined. Portions of the NH2-terminal region of pro-ACTH/endorphin (called the 16K fragment) and the region between ACTH and beta-endorphin (called gamma-lipotropin) were sequenced by Edman degradations of biosynthetically labeled immunoprecipitated proteins and by Edman degradations of purified 16K fragment and beta-lipotropin. With a combination of these two approaches, 29 of the first 34 residues at the NH2-terminal end of the mouse 16K fragment were determined. The NH2-terminal region of the mouse 16K fragment was found to be nearly identical with the homologous porcine and bovine molecules. The complete amino acid sequence of the NH2-terminal region of gamma-lipotropin was determined. In contrast to the highly conserved nature of the 16K fragment, mouse gamma-lipotropin was found to differ substantially from the gamma-lipotropins of other species. Although the NH2-terminal and beta-melanotropin-like regions of the mouse gamma-lipotropin are similar to the corresponding regions of other gamma-lipotropins, the intervening region of mouse gamma-lipotropin is substantially shorter than it is in other gamma-lipotropins. In addition, mouse gamma-lipotropin lacks the pair of basic amino acids that normally mark the proteolytic cleavage site used to produce beta-melanotropin from gamma-lipotropin.