We have conducted a comprehensive study of the molecular basis of cystic fibrosis (CF) in 350 German CF patients. A screening approach based on single-strand conformation analysis and direct sequencing of genomic polymerase chain reaction products has allowed us to detect the molecular defects on 95.4% of the CF chromosomes within the coding region and splice sites of the cystic fibrosis transmembrane conductance regulator (CFTR) gene. The spectrum of sequence changes comprises 54 different mutations, including 17 missense mutations, 14 nonsense mutations, 11 frameshift mutations, 10 splice site variants and two amino acid deletions. Eleven of these mutations have not previously been described. Our results reflect the marked mutational heterogeneity of CF in a large sample of patients from a non-isolated population.

P-glycoprotein, the overexpression of which is a major cause for the failure of cancer chemotherapy in man, recognizes and transports a broad range of structurally unrelated amphiphilic compounds. This study reports on the localization of the binding site of P-glycoprotein for iodomycin, the Bolton-Hunter derivative of the anthracycline daunomycin. 125 I]iodomycin had been predominantly incorporated into the fragment 230 -312 of isoform I of hamster P-glycoprotein. According to models based on hydropathy plots, the amino acid sequence 230 -312 forms the distal part of transmembrane segment 4, the second cytoplasmic loop, and the proximal part of transmembrane segment 5 in the Nterminal half of P-glycoprotein. The binding site for iodomycin is recognized with high affinity by vinblastine and cyclosporin A. P-glycoprotein, which belongs to the large family of ABC 1 transporters (1), binds and transports a broad range of structurally unrelated compounds (2). Its overexpression may cause the phenomenon of multidrug resistance (MDR) during cancer chemotherapy, whereby the tumor cells become resistant to a variety of antineoplastic agents due to a reduced intracellular accumulation of drugs (2-4). The MDR phenotype can be overcome by modulators, i.e. substances that are bound by P-glycoprotein and inhibit its drug excluding function (3). The medically important substrates of P-glycoprotein comprise anticancer drugs such as Vinca alkaloids and anthracyclines (2-4) and approved drugs that turned out to be potent modulators such as verapamil (5, 6), calcium antagonists (7), and cyclosporins (8).The major issue of how P-glycoprotein can handle so many substrates has been mainly approached by photoaffinity labeling and mutagenesis studies. Mutants that arose spontaneously during drug selection of cells and thereby changed their resistance profile were detected in the first cytoplasmic loop of human MDR1 protein (Gly 185 3 Val) (9) and in transmembrane segment (TM) 6 of hamster P-glycoprotein (Gly 338 3 Ala/Ala 339 3 Pro) (10). Active mutagenesis identified further motifs and positions in the N-and C-terminal half of P-glycoprotein which influence substrate specificity (11, 12). For example, the exchange of amino acids in the first cytoplasmic loop (11), TM5, TM6, TM10, TM12 (13-16), and the cytosolic linker peptide (17) resulted in all in an altered multidrug resistance phenotype, suggesting that during binding and transport the substrate is recognized by multiple residues located either in the cytosolic, membraneous, or ectoplasmic domains.Hydropathy plots deduced from the primary sequence predict that P-glycoprotein consists principally of two symmetrical halves, each of which contain a membrane domain with six membrane-spanning segments and a subsequent cytosolic nucleotide binding fold (18 -20). The topology of P-glycoprotein in vivo, however, may be more variable than predicted, e.g. under certain experimental conditions, TM segments were detected outside the membrane and loops postulated to face the cytosol were foun...

We have developed a primer set for a prokaryotic differential display of mRNA in the Enterobacteriaceae group. Each combination of ten 10mer and ten 11mer primers generates up to 85 bands from total Escherichia coli RNA, thus covering expressed sequences of a complete bacterial genome. Due to the lack of polyadenylation in prokaryotic RNA the type T11VN anchored oligonucleotides for the reverse transcriptase reaction had to be replaced with respect to the original method described by Liang and Pardee [ Science , 257, 967-971 (1992)]. Therefore, the sequences of both the 10mer and the new 11mer oligonucleotides were determined by a statistical evaluation of species-specific coding regions extracted from the EMBL database. The 11mer primers used for reverse transcription were selected for localization in the 3'-region of the bacterial RNA. The 10mer primers preferentially bind to the 5'-end of the RNA. None of the primers show homology to rRNA or other abundant small RNA species. Randomly sampled cDNA bands were checked for their bacterial origin either by re-amplification, cloning and sequencing or by re-amplification and direct sequencing with 10mer and 11mer primers after asymmetric PCR.

The alternatively spliced exon 9 of the cystic fibrosis transmembrane conductance regulator (CFTR) gene codes for the initial part of the amino-terminal nucleotide-binding fold of CFTR. A unique feature of the acceptor splice site preceding this exon is a variable length polymorphism within the polypyrimidine tract influencing the extent of exon 9 skipping in CFTR mRNA. We investigated this repeat for its relationship to CFTR mutations and intragenic markers on 200 chromosomes from German patients with cystic fibrosis (CF). Four frequent length variations were strongly associated with the four predominant haplotypes previously defined by intragenic marker dimorphisms. One of these alleles displayed absolute linkage disequilibrium to the major CF mutation delta F508. Other frequent CFTR mutations were linked to one particular splice site haplotype indicating that differential exon 9 skipping contributes little to the clinical heterogeneity among CF patients with an identical mutation. We also identified a novel missense mutation (V456F) and a novel nonsense mutation (Q414X) within the coding region of exon 9. The missense mutation V456F adjacent to Walker motif A was present in a pancreas-sufficient CF patient. In contrast, the pancreas-insufficient Q414X/delta F508 compound heterozygote suffered from a severe form of the disease, indicating that alternative splicing of exon 9 does not overcome the deleterious effect of a stop codon with this exon.