scispace - formally typeset
Search or ask a question

Showing papers in "Cold Spring Harbor Symposia on Quantitative Biology in 1986"


Journal ArticleDOI
TL;DR: An alternative method for the synthesis of specific DNA sequences is explored that involves the reciprocal interaction of two oligonucleotides and the DNA polymerase extension products whose synthesis they prime, when they are hybridized to different strands of a DNA template in a relative orientation such that their extension products overlap.
Abstract: The discovery of specific restriction endonucleases (Smith and Wilcox 1970) made possible the isolation of discrete molecular fragments of naturally occurring DNA for the first time. This capability was crucial to the development of molecular cloning (Cohen et al. 1973); and the combination of molecular cloning and endonuclease restriction allowed the synthesis and isolation of any naturally occurring DNA sequence that could be cloned into a useful vector and, on the basis of flanking restriction sites, excised from it. The availability of a large variety of restriction enzymes (Roberts 1985) has significantly extended the utility of these methods. The de novo organic synthesis of oligonucleotides and the development of methods for their assembly into long double-stranded DNA molecules (Davies and Gassen 1983) have removed, at least theoretically, the minor limitations imposed by the availability of natural sequences with fortuitously unique flanking restriction sites. However, de novo synthesis, even with automated equipment, is not easy; it is often fraught with peril due to the inevitable indelicacy of chemical reagents (Urdea et al. 1985; Watt et al. 1985; Mullenbach et al. 1986), and it is not capable of producing, intentionally, a sequence that is not yet fully known. We have been exploring an alternative method for the synthesis of specific DNA sequences (Fig. 1). It involves the reciprocal interaction of two oligonucleotides and the DNA polymerase extension products whose synthesis they prime, when they are hybridized to different strands of a DNA template in a relative orientation such that their extension products overlap. The method consists of repetitive cycles of denaturation, hybridization, and polymerase extension and seems not a little boring until the realization occurs that this procedure is catalyzing a doubling with each cycle in the amount of the fragment defined by the positions of the 5' ends of the two primers on the template DNA, that this fragment is therefore increasing in concentration exponentially, and that the process can be continued for many cycles and is inherently very specific. The original template DNA molecule could have been a relatively small amount of the sequence to be synthesized (in a pure form and as a discrete molecule) or it could have been the same sequence embedded in a much larger molecule in a complex mixture as in the case of a fragment of a single-copy gene in whole human DNA. It could also have been a single-stranded DNA molecule or, with a minor modification in the technique, it could have been an RNA molecule. In any case, the product of the reaction will be a discrete double-stranded DNA molecule with termini corresponding to the 5' ends of the oligonucleotides employed. We have called this process polymerase chain reaction or (inevitably) PCR. Several embodiments have been devised that enable one not only to extract a specific sequence from a complex template and amplify it, but also to increase the inherent specificity of this process by using nested primer sets, or to append sequence information to one or both ends of the sequence as it is being amplified, or to construct a sequence entirely from synthetic fragments.

3,721 citations





Journal ArticleDOI
TL;DR: It is shown that RFLPs can be found linked to any common human disease that shows simple Mendelian transmission and is caused by a single genetic locus, and that traditional methods for genetic linkage analysis are quite inefficient at mapping genetically complex traits.
Abstract: It has been clear since the rediscovery of Mendel that humans obey laws of heredity identical with those of other organisms. The central features of Mendelism were observable in humans by following simply inherited common traits, including some diseases. However, the systematic study of human heredity using the standard concepts (complementation and recombination, tests of epistasis, etc.) has been infeasible in humans for two reasons: (1) because Homo sapiens is not an experimental animal that can be manipulated at will and (2) because few genetic markers had been found that were heterozygous often enough to allow random matings to be informative. The advent of recombinant DNA technology led to the suggestion (Botstein et al. 1980) that common polymorphisms in DNA sequence (conveniently observed as restriction fragment length polymorphisms, or RFLPs) could be used as generally informative genetic markers allowing the systematic study of heredity in humans, including the construction of a true linkage map of the entire human genome. The application of RFLP technology to simple Mendelian diseases has proceeded rapidly. RFLPs have been found closely linked to the autosomal dominant Huntington's disease (Gusella et al. 1983) and polycystic kidney disease (Reeder et al. 1985), the autosomal recessive cystic fibrosis (Knowlton et al. 1985; Tsui et al. 1985; Wainwright et al. 1985; White et al. 1985), and the X-linked recessive Duchenne muscular dystrophy (DMD) (Davies et al. 1983). In turn, closely linked DNA markers have made it possible to localize the disease genes to chromosomal regions (Gusella et al. 1983; Knowlton et al. 1985; Reeder et al. 1985; Wainwright et al. 1985; White et al. 1985), to undertake prenatal diagnosis of fetuses known to be at risk, and to begin chromosome walks in an attempt to clone the disease genes (Kunkel et al.; Worton et al.; all this volume). In addition, the RFLP linkages have shed light on the formal genetics of the disorders-showing, for example, that mutations at a single locus account for all (or almost all) cases of cystic fibrosis (Donis-Keller et al.; Tsui et al.; Williamson et al.; White et al.; all this volume) or of Huntington's disease (Gusella et al., this volume). These successes make clear that RFLPs can be found linked to any common human disease that shows simple Mendelian transmission and is caused by a single genetic locus. Much of what we want to know about human heredity, however, concerns traits whose underlying genetics is less favorable for analysis. Some apparently identical clinical conditions can result from mutations at any one of several g e n e s a circumstance called genetic heterogeneity. Some traits have incompletepenetrance, with only a fraction of those carrying the appropriate mutant genotype actually displaying the trait. Conversely, some genotypes predispose individuals to a disease, but those of normal genotype may be affected as well, just at a lower rate. Environment may play an important role in the expression of the trait. In addition, gene interactions can occur, in which a phenotype results from the interaction of alleles at more than one locus. These complexities can underlie even the most clearcut phenotypes. These complex modes of inheritance are common in genetically well-studied organisms, such as bacteria, yeast, nematodes, and fruit flies, and evidence is accumulating that humans are no different. Geneticists usually surmount the problems by isolating purebreeding stocks, each carrying a mutation at a single locus, and then arranging crosses at will. In the case of human genetics, however, we must take crosses as we find them. Unfortunately, even with RFLPs as markers, traditional methods for genetic linkage analysis are quite inefficient at mapping genetically complex traits: A prohibitively large sample may be required before linkage can be detected or other genetic properties tested. Furthermore, traditional linkage analysis absolutely requires families with two or more affected individuals. Yet, many traits of biological or medical interest are quite rare, with most cases being sporadic. Collecting enough pedigrees with multiple affected subjects may be impossible. It may thus appear, at first sight, that the vast majority of human heredity must remain refractory to genetic mapping, due to complexity or rarity. On the contrary, it is the main thesis of this paper that this need not be so. The important point is that studying the segregation patterns of a large number of mapped RFLPs simultaneously (rather than one at a time) can make it feasible to map many traits that are genetically complex and/or rare. We discuss below new mathematical techniques, unusual genetic resources, and clinical up-

190 citations











Journal ArticleDOI
Walter F. Bodmer1
TL;DR: The 1986 Cold Spring Harbor Symposium on the subject of human genetics as discussed by the authors was the first one devoted to human genetics and discussed the case for a definitive characterization and sequencing of the human genome.
Abstract: The 1986 Cold Spring Harbor Symposium was on the subject of human genetics; it was the first symposium at Cold Spring Harbor on this topic since 1964. In the opening remarks for the conference, Walter F. Bodmer first summarized the progress in this field since 1964. He then described what is presently known about the functional complexity of the human genome and discussed the case for a definitive characterization and sequencing of the human genome. The following is an abridged and slightly adapted version of this talk; it is reproduced courtesy of the Cold Spring Harbor Laboratory © 1987.





Journal ArticleDOI
TL;DR: A variety of DNA markers for apolipoprotein genes were examined among patients with angiocardiographically proven heart disease and among a variety of normal individuals with various lipid values, finding higher levels of cholesterol and lower levels of HDL among carriers of the common apoAII MspI and the rare apoB PvuII variants.
Abstract: A variety of DNA markers for apolipoprotein genes were examined among patients with angiocardiographically proven heart disease and among a variety of normal individuals with various lipid values. An increased frequency of an apoAI-CIII SstI RFLP and an apoB minisatellite (allele 5) was found among patients with CHD. Higher levels of cholesterol were found among carriers of the rare apoB TaqI and the common apoCII TaqI variants, whereas higher levels of triglycerides were found in carriers of the common apoAII MspI and the rare apoB XbaI variants. Lower levels of HDL were found among carriers of the common apoAII MspI and the rare apoB PvuII variants. The biological significance of these results and those of other investigators for the pathogenesis of CHD and hyperlipidemia is suggestive but not yet fully clarified. Additional genetic epidemiologic studies and family investigations will be required. Currently used statistical methodology may lead to false inferences regarding the genetic equilibrium or disequilibrium status of closely linked DNA variants. Conclusions regarding the presence of genetic equilibrium if closely linked flanking markers are in disequilibrium may be faulty.



Journal ArticleDOI
TL;DR: The analysis of four large deletions has revealed an unexpectedly universal involvement of Alu repeats in their generation, and studies indicate that repetitive DNAs can destabilize a gene through homologous recombination.
Abstract: Since the discovery of the LDL receptor 13 years ago, a multidisciplinary approach to its study has revealed much about this important cell-surface protein Most recently, we have developed tools in the form of full-length cDNAs and cloned genomic DNAs necessary to understand the molecular genetics of this locus The frequent occurrence of mutations in the LDL receptor gene in patients with FH provides a fertile ground on which to explore the parts of the receptor that are necessary for its function The analysis of four large deletions has revealed an unexpectedly universal involvement of Alu repeats in their generation These studies indicate that repetitive DNAs can destabilize a gene through homologous recombination Inasmuch as the LDL receptor gene is a mosaic of exons shared with at least five other proteins, it is possible that early exon-shuffling events involved recombination between these repetitive elements Is it possible that the very plasticity that permitted evolution of the LDL receptor also accounts for its frequent disruption by mutation? Further study may help to answer this question Mutations that disrupt the structure of the protein have been identified The biochemical and cellular consequences of these mutations reveal crucial aspects of receptor structure The receptor is clearly divided into quasi-independent domains with discrete functions Mutations that disrupt the cytoplasmic domain alter the ability of the LDL receptor to cluster in coated pits, but they do not disrupt ligand binding or produce major effects on intracellular transport Some of the mutations in the external domain disrupt binding but do not affect transport or internalization(ABSTRACT TRUNCATED AT 250 WORDS)