Search

Neurofibromatosis type 1 (NFl) is a common genetic disorder characterized primarily by the development of multiple neurofibromas and pigmentary changes. The recent identification of contiguous gene deletions in NF1 a previously unrecognized molecular basis for this disorder, raises important questions regarding deletion frequency in the patient population and the role that contiguous genes may play in the physical manifestations of NF1 patients. To facilitate the identification of patients with large NFl deletions, we have isolated clones carrying large genomic segments from the NF1 locus and tested their efficacy as probes for fluorescence in situ hybridization (FISH). Clone P1-9 spans approximately 65 kb of the NFl gene, including exons 2–11, and clone P1-12 carries ∼55 kb of NFl intron 27B. FISH studies performed with P1-9, P1-12, and a set of overlapping 1F10 cosmid clones mapping telomeric to the NF1 locus identified large deletions in two new neurofibromatosis type 1 patients who, like previously characterized deletion patients, had mildly dysmorphic facial features and large numbers of cutaneous neurofibromas.

Using fluorescence in situ hybridization (FISH), we have identified seven NF1-related loci, two separate loci on chromosome 2, at bands 2q21 and 2q33-q34, and one locus each on five other chromosomes at bands 14q11.2, 15q11.2, 18p11.2, 21q11.2-q21, and 22q11.2. Application of PCR using NF1 primer pairs and genomic DNA from somatic cell hybrids confirmed the above loci, identified additional loci on chromosomes 12 and 15, and showed that the various loci do not share homology beyond NF1 exon 27b. Sequenced PCR products representing segments corresponding to NF1 exons from these loci demonstrated greater than 95% sequence identity with the NF1 locus. We used sequence differences between bona fide NF1 and NF1-homologous loci to strategically design primer sets to specifically amplify 30 of 36 exons within the 5' end of the NF1 gene. These developments have facilitated mutation analysis at the NF1 locus using genomic DNA as template.

Neurofibromatosis 1 maps to chromosome band 17q11.2, and the NF1 locus has been partially characterized. Even though the full-length NF1 cDNA has been sequenced, the complete genomic structure of the NF1 gene has not been elucidated. The 5' end of NF1 is embedded in a CpG island containing a NotI restriction site, and the remainder of the gene lies in the adjacent 350-kb NotI fragment. In our efforts to develop a comprehensive screen for NF1 mutations, we have isolated genomic DNA clones that together harbor the entire NF1 cDNA sequence. We have identified all intron-exon boundaries of the coding region and established that it is composed of 59 exons. Furthermore, we have defined the 3'-untranslated region (3'-UTR) of the NF1 gene; it spans approximately 3.5 kb of genomic DNA sequence and is continuous with the stop codon. Oligonucleotide primer pairs synthesized from exon-flanking DNA sequences were used in the polymerase chain reaction with cloned, chromosome 17-specific genomic DNA as template to amplify NF1 exons 1 through 27b and the exon containing the 3'-UTR separately. This information should be useful for implementing a comprehensive NF1 mutation screen using genomic DNA as template.

The specificity and properties of the group of enzymes present in extracts of Lactobacillus casei which catalyzed phosphoribosyltransfer from 5-phosphoribosyl 1-pyrophosphate to purine bases were markedly different from those in extracts of Escherichia coli. A comparison of the levels of hypoxanthine, guanine, and xanthine phosphoribosyltransfer activities in extracts of a wild strain and a 6-mercaptopurine-resistant strain of L. casei suggested the presence of a xanthine phosphoribosyltransferase which was distinct from the enzyme specific for hypoxanthine and guanine. This suggestion was substantiated by the finding that the activity toward xanthine was less stable at -21° and 60° than the activities toward hypoxanthine and guanine, and by its separation from these activities on Sephadex G-100. The hypoxanthine-guanine phosphoribosyltransferase present in the 6-mercaptopurine-resistant strain differed from that of the wild strain in its lower rate of ribonucleotide synthesis with hypoxanthine relative to that with guanine, and in its somewhat lower affinities for hypoxanthine and 6-mercaptopurine. When extracts of E. coli were treated at 60°, the phosphoribosyltransfer activities toward guanine and xanthine decreased at the same rate, whereas the activity toward hypoxanthine decreased more rapidly. Chromatography on Sephadex G-100 did not resolve the activities toward hypoxanthine, guanine, and xanthine in extracts of E. coli. However, on Ecteola-cellulose, these activities were separated into two distinct peaks, the relative specificities of which indicated the presence of two different enzymes or enzyme forms. With one, hypoxanthine was the most efficient phosphoribosyl acceptor, whereas with the other, guanine was the most efficient acceptor. No distinct xanthine phosphoribosyltransferase was detected in extracts of E. coli. In extracts of both L. casei and E. coli, the activity toward adenine was distinct from the activities toward hypoxanthine, guanine, and xanthine. This was shown by separations on Sephadex G-100 columns and by the differences in rates of heat inactivation.

The structural requirements for the binding of purines to adenine phosphoribosyltransferase from monkey liver were explored by the determination of the Ki values for 67 purines and purine analogues. Adenine was bound 40 times better than the most effectively bound analogues. Substitutions on the 6-amino group of adenine decreased binding to varying degrees. The hydroxylamino- and hydrazinopurines bound well; the binding of alkylaminopurines was moderately good only for small alkyl groups. With the exception of purine-6-aldoxime, purine, and 6-methylpurine, other 6-substituted purines showed Ki values g10-3 m. All 2-substituted derivatives of adenine tested were poorly bound, except 2-fluoroadenine. Ring nitrogen methylation of adenine markedly reduced binding. Adenine derivatives, substituted in position 8 with methyl, mercapto, or m-nitrophenyl, were effectively bound. Although poorly bound, 2,6-diaminopurine and 5-aminoimidazole-4-carboxamide reacted at rates which indicated a high maximal velocity. Conversely, 4-aminopyrazolo[3,4-d]pyrimidine was effectively bound but reacted slowly. Experiments with radioactive 5-aminoimidazole-4-carboxamide are consistent with earlier work which suggested that this substrate is converted to the nucleotide level by adenine phosphoribosyltransferase but not by the corresponding hypoxanthine enzyme. The kinetics of the magnesium activation of AMP synthesis was analogous to those previously reported for human hypoxanthine phosphoribosyltransferase. Although the apparent affinity (Km value = 0.037 mm) of adenine phosphoribosyltransferase for 5-phosphoribosyl 1-pyrophosphate was an order of magnitude greater than that of the human hypoxanthine enzyme, similar values for the dissociation constant of the dimagnesium salt of 5-phosphoribosyl 1-pyrophosphate (10-3 m) were estimated from the data obtained with both enzymes. This is consistent with the hypothesis that magnesium ions activate both these enzymatic reactions by forming a complex with the substrate, 5-phosphoribosyl 1-pyrophosphate. The adenine phosphoribosyltransfer activity of the monkey liver preparation was not decreased by sulfhydryl or serine modifiers, nor was it activated by sulfhydryl reagents; the residual phosphoribosyltransfer activity toward hypoxanthine was also insensitive to p-chloromercuribenzoate and dithiothreitol.

This study directly compares the specificities of the structurally similar hydroxylating enzymes, aldehyde oxidase and xanthine oxidase. Michaelis-Menten constants for a variety of substrates of xanthine oxidase were in general lower than those of aldehyde oxidase. With respect to the rates of oxidation, the basic similarity was a preference for compounds having a substituted pyrimidine ring structure. Outstanding among the differences were the effects of the number and position of the ring substituents. Both enzymes readily oxidized a variety of unsubstituted and C-monosubstituted heterocycles, but only xanthine oxidase readily oxidized C-disubstituted derivatives. Certain N-substitutions, however, enhanced substrate activity with aldehyde oxidase, but markedly decreased it with xanthine oxidase. Although both enzymes preferred oxo over amino substituents, there were some specificity differences with respect to the chemical nature of substituents. Aldehyde oxidase, but not xanthine oxidase, tolerated 6-substitution of purine by alkyl, halogeno, cyano, or methylthio groups, while 6-hydroxyl or 6-methylamino substituents were tolerated only by xanthine oxidase. The position at which oxidation occurred was influenced by both the chemical nature and the positions of substituents. With some purines a different site was initially hydroxylated by each enzyme.