Construction of a Genetic Linkage Map in Man Using Restriction
Contents
1. 330 30 31 32 33 34 35 36 37 38 39 40 41 42 43 45 46 47 48 49 50 51 52 53 BOTSTEIN ET AL Mourant AE Kopec AC DoMANIEWSKA SOBCZAK The Distribution of Human Blood Groups and Other Polymorphisms st ed London Oxford Univ Press 1954 GIBLETT ER Genetic Marker Human Blood st ed Philadelphia FA Davis 1969 KompF J BISSPORT S GUSSMAN S RITTER H Polymorphism of red cell glyoxalase I EC 4 4 1 5 A new genetic marker in man Hum Genet 27 141 143 1975 SAMLOFF IM Townes PL Pepsinogens genetic polymorphism in man Science 168 144 145 1970 TENG Y S ANDERSON JE GIBLETT ER Cytidine deaminase a new genetic polymorphism demonstrated in human granulocytes Am J Hum Genet 27 492 497 1975 ALPER CA BOENISCH T WATSON L Genetic polymorphism in human glycine rich beta glycoprotein J Exp Med 135 68 1972 HIRSCHFELD J The Gce system Prog Allergy 6 155 186 1962 HOPKINSON DA Harris H A third phosphoglucomutase locus in man Ann Hum Genet 31 359 367 1968 HOPKINSON DA Harris H Rare phosphoglucomutase phenotypes Ann Hum Genet 30 167 181 1966 TEISBERG P High voltage agarose gel electrophoresis in the study of C3 polymorphism Vox Sang 19 47 56 1970 MERRIT DA Rivas M BAXLER D NEWELL R Salivary and pancreatic amylase electrophoretic characterizations and genetic studies Am J Hum Genet 25 510 520 1973 HOPKINSON DA MESTRINER2. 60 1978 Morton NE Segregation and linkage in Methodology in Human Genetics edited by BURDETTE WJ San Francisco Holden Day 1962 pp 17 52 THOMPSON EA Kravitz K HILL J SKOLNICK M Linkage and the power of a pedigree structure in Genetic Epidemiology edited by MORTON NE CHUNG CS New York Academic Press 1978 pp 247 253 THoMas G BODMER WF BopMER J The HLA system as a model for studying the interaction between selection migration and linkage in Population Genetics and Ecology edited by KARLIN S NEvo E New York Academic Press 1976 pp 465 497 CLEGHORN TE MNSs gene frequencies in English blood donors Nature 187 701 1960 RACE RR TAYLOR GL CAPPELL DF MCFARLANE MN Recognition of a further common Rh genotype in man Nature 153 52 53 1944 ROBSON EB Harris H Further studies on the genetics of placental alkaline phosphatase Ann Hum Genet 30 219 232 1967 DAIGER SP LABROWE ML Parsons M WANG L CAVALLI SFORZA LL Detection of genetic variation with radioactive ligands III Genetic polymorphism of transcobalamin II in human plasma Am J Hum Genet 30 202 214 1978 CHEN S H GIBLETT ER ANDERSON J FOSSEN BL Genetics of glutamic pyruvate transaminase its inheritance common and rare variants population distribution and dif ferences in catalytic activity Ann Hum Genet 35 401 409 1972 Race RR HoLt HA THOMPSON JS Th inheritance and distribution of the Duffy blood groups Heredity 5 103 110 1951
3. RUBIN GM Isolation of a telomeric DNA sequence from Drosophila melanogaster Cold Spring Harbor Symp Quant Biol 42 1041 1046 1978 SCHMID C DEININGER P Sequence organization of the human genome Cell 6 345 358 1975 WYMAN A WHITE RL Restriction fragment length polymorphism in human DNA In preparation MoRrTON NE CHUNG CS EDS Genetic Epidemiology New York Academic Press 1978 SING CF SKOLNICK M EDS Genetic Analysis of Common Diseases Applications to Predictive Factors in Coronary Disease New York Academic Press 1978 CARTWRIGHT GE EDWARDS CQ KRAVITZ K ET AL Hereditary hemochromatosis phenotypic expressions of the disease N Engl J Med 301 175 179 1979 RUDDLE F McKuSICcK V The status of the gene map of the human chromosomes Science 196 390 405 1977 ORKIN S ALTER B ALTAY C ET AL Application of endonuclease mapping to the analysis and prenatal diagnosis of thalassemias caused by globin gene deletion N Engl J Med 299 166 171 1978
4. origin B globin genes were also found on 7 0 and 13 0 kb Hpal fragments Of individuals with the hemoglobin S allele 87 were found to have the 13 0 kb variant fragment This linked RFLP has been used to diagnose sickle cell trait in utero 17 More recently the abundance of RFLPs associated with the B and y globin regions have been examined 18 19 Among 60 individuals Jeffreys found one rare and two frequent polymorphisms GENETIC MAPPING WITHOUT SPECIFIC GENE ISOLATION The above RFLPs were recognized by their relationship to the DNA sequence for a gene of interest These experiments could not have proceeded without prior isolation of the specific gene s DNA We note however that for the more general purpose of mapping genetic loci RFLPs need not encode the gene of interest but only be sufficiently nearby to display genetic linkage This is extremely important because a recombinant DNA probe need not reveal a restriction fragment containing part of the gene of interest to be useful that is one does not have to isolate the gene to map it If a substantial number of polymorphic regions can be identified and we estimate that 150 will eventually be required then all genes will be linked to one or another of the regions containing RFLPs and can thereby be mapped MAKING THE RFLP MAP The demonstration of a heritable and detectable RFLP at the B globin locus implies that the genetic mapping scheme based on linkage in f
5. Cancer edited by MULVIHILL JJ MILLER RW FRAUMENI FJ JR New York Raven Press 1977 pp 19 25 WILLIAMS R SKOLNICK M CARMELLI D ET AL Utah pedigree studies design and preliminary data for premature male CHD deaths in The Genetic Analysis of Common Diseases Applications to Predictive Factors in Coronary Heart Disease edited by SING CF SKOLNICK M New York Alan R Liss 1979 pp 711 732 B sHoP DT SKOLNICK M Numerical considerations for mapping DNA polymorphic markers in Proceedings Conference on Human Health Data from Defined Populations Banbury Center Cold Spring Harbor Laboratory October 1979 Banbury Report No 4 Cold Spring Harbor Laboratory In press 1980 HARRIS H HOPKINSON D Average heterozygosity per locus in man an estimate based on 54 55 56 57 58 59 60 61 62 63 64 65 GENETIC MAP USING RESTRICTION FRAGMENTS 331 the incidence of enzyme polymorphisms Ann Hum Genet 36 9 20 1972 ROSBASH M CAMPO M GUMMERSON K Conservation of cytoplasmic poly A containing RNA in mouse and rats Nature 258 682 686 1975 BRITTEN R CETTA A DAVIDSON E The single copy DNA sequence polymorphism of the sea urchin Strongylocentrotus purpuratus Cell 10 509 519 1977 LEWONTIN RC The Genetic Basis of Evolutionary Change New York Columbia Univ Press 1974 UPHOLT W Estimation of DNA sequence divergence from comparison of restriction endonuclease digests Nucleic Acids Res 4 1257 1265 1977
6. MA CORTNER J HARRIS H Esterase D a new human polymorphism Ann Hum Genet 37 119 137 1973 HOPKINSON DA Cook PJL HArris H Further data on the adenosine deaminase ADA polymorphism and a report of a new phenotype Ann Hum Genet 32 361 367 1969 BISSPORT S Kompr J Red cell galactose 1 p uridyl transferase formal genetics and linkage relations Humangenetik 18 93 94 1973 GIBLETT ER ANDERSON J CHEN S H TENG Y S COHEN F Uridine monophosphate kinase a new genetic polymorphism with possible clinical implications Am J Hum Genet 26 627 635 1974 CLEGHORN TE The occurrence of certain rare blood group factors in Britain PhD thesis England Univ of Sheffield 1961 RAPLEY S ROBSON EB HARRIS H MAYNARD SMITH S Data on the incidence segregation and linkage relations of the adenylate kinase AK polymorphism Ann Hum Genet 31 237 242 1967 CARTER ND FILDES RA FITCH LI PARR CW Genetically determined electrophoresis variations of human phosphogluconate dehydrogenase Acta Genet Stat Med Basel 18 109 122 1968 Lewis WHP Harris H Human red cell peptidases Nature 215 351 355 1967 CHAKRABORTY R FUERST PA FERRELL RE Potential information in family studies of linkage in Genetic Analysis of Common Diseases Applications to Predictive Factors in Coronary Disease edited by SING CF SKOLNICK M New York Alan R Liss 1979 pp 297 303 SKOLNICK M Prospects for population oncogenetics in Genetics of Human
7. for loci likely to show polymorphism The mobile sequence could be cloned and used to screen the human genomic library Isolation and analysis of the sister DNA segment clones should reveal the presence of adjacent single copy sequences These then could be used as probes of DNA from individuals to determine whether the repetitious segment is mobile from the new locus thereby defining a new polymorphic site Although the number of fragments examined is the important parameter in discovering RFLPs due to single base changes the length of DNA screened is the most important factor in the case of insertions and deletions since many different restriction fragments defined by different enzymes will reveal the same insertion or deletion SOURCE OF RECOMBINANT DNA PROBES DEFINING RFLP LOCI To find polymorphisms we must be able to identify specific and unique regions of the genome The recent development of recombinant DNA technology permits the cloning in E coli of individual segments of the human genome These DNAs can be labeled with radioisotopes to high specific activities and used as hybridization probes of DNA from individuals being digested with various restriction fragments from specific regions Two libraries of human recombinant DNA have been constructed by two quite different methods each with properties relevant to this purpose One library was constructed from the DNA of the entire human genome 17 using a bacteriophage vector The in
8. i denotes the linkage group chromosome if known j the locus within the chromosome and k the specific fragment at the locus Many fragments at a locus made with different restriction enzymes will be investigated In addition neighboring clones will be studied to expand polymorphism to an amount sufficient for linkage studies Before RFLPs become a useful tool for mapping disease loci the number of members of set R will be reduced Some R will be unnecessary for initial screening and efficiency dictates that a primary set R be established eliminating some of the j sites and for each site some of the fragments If linkage is not detected the secondary set R containing additional fragments at R loci and additional loci will be sequentially examined until linkage is established Once linkage is suspected neighboring loci will be explored until the closest locus is found One may expand the set of fragments at this locus by seeking neighboring fragments to increase polymorphism For some counseling purposes we wish to find the closest RFLP to the disease locus which has unique alleles in both parents Another possible approach is to try to find two RFLP loci which flank the disease locus on the genetic map Such flanking markers will greatly increase the certainty of genotype determinations used as the basis for counseling Estimates of the Frequency of Polymorphism in DNA The above considerations lead us to the question
9. of how much polymorphism to expect in human DNA Although little is known about the frequency of DNA polymorphism in humans several sources of information can be used to provide order of magnitude estimates including determinations of base sequence divergence based on melting temperature depressions of reannealed DNA and measurements of amino acid sequence divergence among proteins Estimates from the frequency of protein polymorphism The frequency of polymorphism among 71 protein studies in European populations 53 was 0 28 suggesting that the minimum DNA sequence polymorphism must be at least 0 28 base changes per 1 000 base pairs the size of the average gene Assuming that visible electromorphs represent about one third of the nucleic acid sequence changes and that many third base changes are not reflected in corresponding amino acid changes we can 324 BOTSTEIN ET AL estimate the frequency of DNA sequence polymorphism to be about 0 001 per base pair among protein coding sequences This sets a lower limit for the amount of base sequence polymorphism since sequences found as mRNA have been shown to diverge more slowly than the bulk of the single copy DNA 54 Estimates from the depression of the melting temperature of reassociated DNA Divergence of nucleic acid sequence between related species has also been determined by measurement of the relative depression of the denaturation temperature of DNA that has been melted and reassociated
10. 1 536 1974 POTTER S NEWBOLD J HUTCHINSON C III EDGELL M Specific cleavage analysis of mammalian mitochondrial DNA Proc Natl Acad Sci USA 72 4496 4500 1975 PETES TK BOTSTEIN D Simple Mendelian inheritance of the reiterated ribosomal DNA of yeast Proc Natl Acad Sci USA 74 5091 5095 1977 GoopMAN H OLSON M HALL B Nucleotide sequence of a mutant eukaryotic gene the yeast tyrosine inserting ochre suppressor SUP4 0 Proc Natl Acad Sci USA 14 5453 5457 1977 CAMERON JR LOH EY Davis RW Evidence for transposition of dispersed repetitive DNA families in yeast Cell 16 739 751 1979 OLSON MV HALL BD CAMERON JR DAvis RW Cloning of the yeast tyrosine transfer RNA genes in Bacteriophage lambda J Mol Biol 127 285 295 1979 MANIATIS T HARDISON R Lacy E ET AL The isolation of structural genes from libraries of eucaryotic DNA Cell 15 687 701 1978 KAN Y Dozy A Antenatal diagnosis of sickle cell anaemia by DNA analysis of amniotic fluid cells Lancet 2 910 912 1978 JEFFREYS AF DNA sequence variants in the Gy Ay o and B globin genes of man Cell 18 1 10 1979 Tuan D Biro PB DE RIEL JK Lazarus H ForGET BG Restriction endonuclease mapping of the human y globin gene loci Nucleic Acids Res 6 2519 2544 1979 BirD AP Use of restriction enzymes to study eukaryotic DNA methylation II The symmetry of methylated sites supports semi conservative copying of the methylation pattern J Mol Biol 118 49
11. 31 GLO ig So cele erin EEG 363 2 S W German 169 32 Be Co reat ieee ten i tate ater be aD 359 2 Caucasian 931 33 CDA 5 occ t td ranting oe ae neck 352 2 Caucasian 189 34 BY ces Picts ei oR eh eee ek 342 3 Caucasian 158 35 GO nome anaa ea a wee aN ae 325 2 English 49 36 POM E E EENT E E E 308 2 English 583 37 POM raer A EEDS A 286 6 English 2115 38 CS AONE TEEN ETET 277 2 Norwegian 400 39 M 5 1s oe Sas cea amp SiS cw 2 171 2 Caucasian U S 673 40 ESD EEEE E EARE 162 2 European 867 41 ADA tiets ECEE NANNE 089 2 English 1353 42 GALT isona ene eee 087 2 S W German 195 43 UMPK cenere i Ri toe a kieke 086 3 Caucasian 386 44 KELD see th Oia Oe hee bi 084 2 British 8767 45 AR EET A AEE EEEN 082 2 British N Ireland 1887 46 EDD ok gh NS ee Bearings E E E 058 2 English 1166 30 OPGD SR ES a Eee ee 042 3 English London 4558 47 PEPA oe hui eke Samu ak he 002 3 British 2283 48 Data compiled from Mourant AC Kopec AC Domaniewska Sobczak The Distribution of the Human Blood Groups and Other Polymorphisms 2nd ed London Oxford Univ Press 1976 322 BOTSTEIN ET AL well known principles multigenerational pedigrees are more useful than nuclear families and large families give considerably more information than smaller families The expected LOD score also varies with the recombination fraction Figure 3 from 22 shows that an expected LOD score of 0 1 spans both the range of reasona
12. Am J Hum Genet 32 314 331 1980 Construction of a Genetic Linkage Map in Man Using Restriction Fragment Length Polymorphisms DAVID BOTSTEIN RAYMOND L WHITE MARK SKOLNICK AND RONALD W Davis SUMMARY We describe a new basis for the construction of a genetic linkage map of the human genome The basic principle of the mapping scheme is to develop by recombinant DNA techniques random single copy DNA probes capable of detecting DNA sequence polymorphisms when hybridized to restriction digests of an individual s DNA Each of these probes will define a locus Loci can be expanded or contracted to include more or less polymorphism by further application of recombinant DNA technology Suitably polymorphic loci can be tested for linkage relationships in human pedigrees by established methods and loci can be arranged into linkage groups to form a true genetic map of DNA marker loci Pedigrees in which inherited traits are known to be segregating can then be analyzed making possible the mapping of the gene s responsible for the trait with respect to the DNA marker loci without requiring direct access to a specified gene s DNA For inherited diseases mapped in this way linked DNA marker loci can be used predictively for genetic counseling INTRODUCTION Although it is possible to detect linkage among simple Mendelian traits in humans no method of systematically mapping human genes has been devised largely because of th
13. POLYMORPHISM RFLP DEFINITION DNA restriction enzymes see 6 for review recognize specific sequences in DNA and catalyze endonucleolytic cleavages yielding fragments of defined lengths Restriction fragments may be displayed by electrophoresis in agarose gels separating the fragments according to their molecular size Differences among individuals in the lengths of a particular restriction fragment could result from many kinds of genotypic differences one or more individual bases could differ resulting in loss of a cleavage site or formation of a new one alternatively insertion or deletion of blocks of DNA within a fragment could alter its size These genotypic changes can all be recognized by the altered mobility of restriction fragments on agarose gel electrophoresis Fragments encoding specific sequences from within a large and complex population of DNA fragments can be detected by hybridization using the method of Southern 7 The DNA from an agarose gel is transferred onto nitrocellulose paper and hybridized with radioactive probe sequences Recently the use of probe sequences of very high specific radioactivity has permitted the identification by this method of single copy sequences in mammalian DNA 8 This powerful new technology makes possible the identification of variants from within a specific region of the genome using digest of total human DNA 316 BOTSTEIN ET AL To illustrate figure la shows a schematic representation
14. Tim Helentjaris REFERENCES 1 ELSTON RC STEWART J A general model for the genetic analysis of pedigree data Hum Hered 21 523 542 1971 2 CANNINGS C SKOLNICK MH DE NEVERS K SRIDHARAN R Calculation of risk factors and likelihoods for familial diseases Comput Biomed Res 9 393 407 1976 3 CANNINGS C THOMPSON EA SKOLNICK M Probability functions on complex pedigrees Adv Appl Prob 10 26 61 1978 4 Renwick JH Progress in mapping human autosomes Br Med Bull 25 65 73 1969 KRAVITZ K SKOLNICK M CANNINGS C ET AL Genetic linkage between hereditary hemochromatosis and HLA Am J Hum Genet 31 601 619 1979 n 13 14 15 16 17 20 2I 22 23 24 25 26 27 28 29 GENETIC MAP USING RESTRICTION FRAGMENTS 329 NATHANS D SMITH H Restriction endonucleases in the analysis and restructuring of DNA molecules Annu Rev Biochem 44 273 293 1975 SOUTHERN EM Detection of specific sequences among DNA fragments separated by gel electrophoresis J Mol Biol 98 503 517 1975 JEFFREYS A FLAVELL R A physical map of the DNA regions flanking the rabbit globin gene Cell 12 429 439 1977 GRODZICKER T WILLIAMS J SHARP P SAMBROOK J Physical mapping of temperature sensitive mutations of adenoviruses Cold Spring Harbor Symp Quant Biol 39 439 446 1974 HUTCHINSON CII NEWBOLD J POTTER S EDGELL M Maternal inheritance of mammalian mitochondrial DNA Nature 25
15. ag ments from a specified longer primary sequence containing no mid repeats constitutes a usable probe for that region REFINING MODELS OF INHERITANCE USING THE RFLP MAP The resolution of genetic and environmental components of disease has evolved in the last decade into a well formed discipline of genetic epidemiology While the methods of analysis are still widely disputed 61 62 many of the problems at least are agreed upon A description of the genetic epidemiology of a disease must involve unraveling the model for the underlying genetic predisposition endogenous factors understanding the environmental contributions exogenous factors including the effects of chance development and age and finally understanding the variability of expression of the phenotype In principle linked marker loci can allow one to establish with high certainty the genotype of an individual and consequently assess much more precisely the contribution of modifying factors such as secondary genes penetrance and environ ment Tightly linked markers such as HLA have been used to clarify genetic models GENETIC MAP USING RESTRICTION FRAGMENTS 327 Cy Awm Pia Pi 2 c maA probes Pi P11 4 clon ed DNA segments cee e Cy WwW 2 w single copy DNA i 2 WAM repetitive DNA Fic 4 Expansion of probe p with neighboring probes p 2 Pi 1s Pitis and Pi 2 Probes are diagrammed in an intact DNA region with restric
16. amily studies of randomly derived RFLPs is possible in principle It is less clear however how difficult it will be to find usefully polymorphic loci and to establish linkage relationships in practice To address these issues we must answer a series of questions many of which can have interdependent and complex answers Four interrelated questions must be answered 318 BOTSTEIN ET AL 1 How many markers are needed 2 How polymorphic must each marker be 3 How many families are needed to establish linkage and 4 How much polymorphism can we expect in the human genome For counseling purposes the number of markers needed and the degree of polymorphism desired are unlimited more and higher are always better Although loose linkage to a slightly polymorphic marker is better than no linkage the tighter the linkage the more accurately one can predict genotypes and the higher the degree of polymorphism the more often one will have informative families necessary for counseling For the purpose of construction of a genetic linkage map we can be less demanding and more precise although here too the need for polymorphism and useful marker loci can never be totally fulfilled Individuals and lineages spanning multiple generations will be selected from large pedigrees The restriction fragment fingerprint resulting for each probe from each individual will be recorded and analyzed for linkage The pedigrees under study will already have bee
17. bility offspring Genotype of affected parent Genotype of unaffected mate informative Frequency of mating A A a ele ge Sw ew cece ees A A 1 0 2P p l pi p AGA Trea LETEN eos A A or Adj 1 0 2p P 1 pi PiP 2p n BAS os cpus en arte Rhea All homozygotes 1 0 2P P 2 p A EEEE P ETET A A J 2Pi P X 2pi P All homozygotes All genotypes 0 p One can evaluate each marker locus for its polymorphism information content PIC by summing in table 1 the mating frequencies multiplied by the probability that an offspring will be informative Under our assumptions the expected value of PIC can be calculated as n n 1 n CERPI gt 2 2 2pepe i 1 i 1 j i 1 A similar index can be found in 49 Both index and marker loci are considered in nuclear families ignoring both the advantages and complexities of multigenerational studies which allow one to know the coupling or repulsion of disease and marker loci and on occasion the genotypes for GENETIC MAP USING RESTRICTION FRAGMENTS 321 loci with recessive or silent alleles Multigenerational studies can increase the amount of information which can be derived from a marker locus but they should not alter our basic impression of the value of a locus for linkage Complications in the genetic model can make detection of linkage much more difficult and can reduce the PIC Table 2 presents the PIC and number of alleles for 29 established codominant chromosomal markers The calcu
18. ble sibship sizes 2 to 8 and recombination fractions 3 to 0001 To establish a LOD score of 3 0 we would need 30 of these units this translates into 300 individuals to establish a linkage at 0 3 and as few as 21 individuals to establish linkage in three 5 child families where tight linkage between the index and marker loci exists These numbers can be decreased further if the nuclear families are linked in a multigenerational pedigree The advantages of large genealogies can also be exploited further if the genealogy of untested family members is known and principles of sequential sampling are employed to determine which members should be studied One excellent source of large pedigrees of known genealogical structure is the computerized data base created by the genetics group at the University of Utah a file that contains seven generations of descendants of the Utah Mormon pioneers who because of large family sizes often have descendants numbering in the thousands 50 51 This data base is also useful in the search for new markers in pedigrees where the known markers have been assayed The numbers cited above 15 to 300 individuals must be multiplied by a function of the reciprocal of the PIC of the marker locus used For this reason we are initiating our search for RFLPs in large Mormon pedigrees with genetic defects which are already being analyzed for known chromosomal markers The above discussion emphasizes the interplay between sam
19. characteristics rearrange ments giving codominance and modifying enzymes being either dominant or reces sive Other even more complex possibilities may exist Before considering our questions we must briefly address the issue of how linkage data in humans are analyzed As RFLPs are identified they must be examined for a number of statistical properties by pedigree analysis 1 3 To establish a map of markers pedigree analysis will be used with qualitative traits such as allele 1 2 n As diseases are studied later quantitative traits such as cholesterol weight time to an event or a combination of the above will also be considered Models are fit by maximizing the likelihood of the parameters given the pedigree structure and the observations Geneticists compare likelihoods of different models by calculating likelihood ratios When the likelihood ratio is 1000 1 that is the odds of one model are 1 000 fold greater than the odds of another then the first is accepted compared to the GENETIC MAP USING RESTRICTION FRAGMENTS 319 latter The log of the likelihood ratio LOD score as described by Morton 21 is usually reported Thus a LOD score of 1 to 2 is interesting 2 to 3 suggestive and gt 3 proof of the phenomenon in question How Many Markers Are Needed To answer this question we must know how large the human genome is in genetic terms and how far one can be from a marker and st
20. e To simplify this discussion we will imagine that we are examining for linkage a pair of loci which we will call the index locus and the marker locus We will assume that the index locus contains a rare dominant allele which segregates in the family under study The allele could be a disease trait or a rare allele of an RFLP locus The marker locus is a polymorphic locus for which information for linkage studies is to be determined We assume that all alleles at the marker locus are codominant which is our expectation for RFLPs Our assumption of rarity of an allele at the index locus is made only to simplify the discussion and is not a necessity for our analysis The lineage of the rare allele at the index locus in the family under study is always clear by our assumptions The question to be answered is How does the degree of polymorphism at the marker locus influence the probability of detection of linkage to the index locus Informativeness in this context is represented by the probability that a given offspring of a parent carrying the rare allele at the index locus will allow deduction of the parental genotype at the marker locus Table 1 gives the mating types their frequencies and the probability of informativeness of an offspring for a marker locus A with n alleles and frequencies pj TABLE 1 MATING TYPES PROBABILITY THAT OFFSPRING ARE INFORMATIVE AND FREQUENCY OF EACH TYPE FOR THE LINKAGE MODELS Proba
21. e paucity of highly polymorphic marker loci Received November 20 1979 revised January 4 1980 This project was supported by grants CA 16573 from the National Institutes of Health and NP 193B NP 286 and VC 245 from the American Cancer Society 1 Department of Biology Massachusetts Institute of Technology Cambridge Mass 2 Department of Microbiology University of Massachusetts Medical School Worcester 3 Department of Medical Biophysics amp Computing University of Utah College of Medicine and LDS Hospital Salt Lake City 4 Department of Biochemistry Stanford University School of Medicine Stanford Calif 1980 by the American Society of Human Genetics 0002 9297 80 3203 0013 01 58 314 GENETIC MAP USING RESTRICTION FRAGMENTS 315 The advent of recombinant DNA technology has suggested a theoretically possible way to define an arbitrarily large number of arbitrarily polymorphic marker loci In this paper we discuss this general approach to the construction of a human genetic linkage map using recombinant DNA probes to define marker loci which are polymorphic in DNA sequences a subset of such DNA polymorphisms can readily be detected directly as differences in the length of DNA fragments after digestion with DNA sequence specific restriction endonucleases These restriction fragment length polymorphisms RFLPs can be easily assayed in individuals facilitating large population studies Small volumes of peripheral blood sho
22. e the amount of DNA polymorphism In yeast and GENETIC MAP USING RESTRICTION FRAGMENTS 325 Drosophila RFLPs have been discovered which are due to such alterations In these species the mobile DNA segments are moderately repetitive 14 58 The frequency in humans of DNA polymorphism from this source cannot yet even be estimated Insertion and deletion of such mobile DNA segments could however also be a major mechanism for the generation of restriction fragment length variation in humans The change in mobility of a restriction fragment due to insertion or deletion is unpredictable in advance Insertions would increase the length of a restriction fragment if the inserted segment does not contain a restriction site or possibly but not necessarily decrease the fragment length if it does contain a restriction site Similarly deletions completely internal to the restriction fragment will decrease its length whereas deletions including one of the restriction sites could either increase or decrease the fragment length depending on the size of the deletion and location of the next nearest restriction site Deletion of the entire probed sequence would not be revealed and would have a recessive phenotype The most diagnostic feature of insertions and deletions would be their pleiotropic effect on some restriction fragments generated by several restriction enzymes If mobile repetitive DNA segments were found in human DNA these could serve as a probe
23. he disease Moreover when a specific linkage is found further refinements should be able to increase the available polymorphism and tighten the linkage until a valid tool for preventive medicine is established A large set of RFLPs will have many other uses in human population genetics A large set of truly neutral markers will reopen questions of genetic distance between geographical isolates and may shed more light on population structure and selection as competing mechanisms of allelic variation and linkage disequilibrium Multilocus theory will have abundant data to test its predictions Human cytogenetic analysis will also benefit from this new technology Often birth defects phenotypically resemble known chromosomal deletions but show no missing bands on the suspect chromosome With a map of RFLPs one could look for areas where the alleles of one parent were entirely missing and infer a deletion too small to be detected by standard techniques Similarly trisomies would produce unique patterns which could reveal small translocations ACKNOWLEDGMENTS We are indebted to numerous individuals for valuable suggestions during the development of this project We wish to thank the graduate students and faculty who participated in the Alta genetics retreat April 1978 for their part in the discussions which originated this project Specifically we would like to acknowledge Maurice S Fox Raymond F Gesteland D Timothy Bishop Arlene Wyman and
24. ill detect linkage Renwick 4 has estimated that the human genome is 33 Morgans in length Figure 2 from 22 shows that the expected LOD score for two linked markers in small family structures falls by an order of magnitude between 6 1 and 0 3 where is the recombination fraction Therefore a search for linkage will be very inefficient at greater values of 0 For this discussion an equivalence between Morgans and recombination fraction is assumed which is approximately true within this range of values We will pick 6 2 as the maximum distance desired between RFLP marker loci which will still allow linkage to be detected To map a new RFLP a dominant gene for a disease or a gene for a carrier trait RFLPs spaced 0 2 Morgans apart will be sufficient to guarantee that a gene is detectably linked to an RFLP locus at a distance no greater than 0 1 At this distance and without further dividing the 33 Morgans into individual chromosome mee EXPECTED LOD 0 3 Ce i d 10 5 1072 RECOMBINATION FRACTION Fic 2 Expected LOD score as a function of recombination fraction for two 4 member pedigrees o7 5 320 BOTSTEIN ET AL sizes about 150 RFLP markers would be required To create a map more than 150 polymorphic probes marker loci must be isolated and tested and occasionally slightly greater than optimal distances will have to be accepted How Polymorphic Must Each Marker B
25. in the single copy sequences 59 Since the mid repeat DNAs will hybridize to many fragments presumably from many parts of the genome the pattern of the Southern transfer becomes too complex for ready interpretation However the available evidence suggests that not all the single copy human DNA contains short period interspersed mid repeat sequences a fraction of the single copy DNA should contain quite long more than 10 kb segments of DNA with no interspersed mid repeat material Preliminary results indicate that a reasonable propor tion 1 3 of the phage clones contains no mid repeat sequences and the clones are usable as probes without further alteration 60 These were discovered by screening the phage library with radiolabeled mid repeat human DNA selecting clones which do not hybridize Figure 4 shows that a locus can be substantially expanded by using the initial recombinant DNA sequence to probe the phage library for adjacent sequences from which a new composite probe can be constructed The procedure is called walking and consists of identifying successive adjacent clones of recombinant DNA orienting them by their restriction enzyme site pattern with the primary clone and excising the mid repeat sequences as indicated above Additional steps can be taken the new Sequences each having their associated polymorphisms until the desired degree of polymorphism is achieved The combined set of subcloned single copy only fr
26. lations use gene frequencies in Caucasians Under this formulation there are four codominant loci which are highly informative PIC gt 5 fifteen are reasonably informative 5 gt PIC gt 25 and ten are only slightly informative PIC lt 25 Loci with many alleles and a PIC near 1 are most desirable Such loci will probably be essential if the kind of mapping method we propose is to become a practical reality How Many Pedigrees The number of informative families required can only be estimated The precise pedigree structure and family size are relevant in determining this number An analysis of this problem 22 has shown that the amount of information in a family study varies considerably with the precise structure of the pedigree Thompson et al quantified two TABLE 2 PIC OF 29 CODOMINANT HUMAN CHROMOSOME MARKER LOCI IN CAUCASIANS Locus PIC No alleles Population No subjects Source HLA A ice pati Hes bee reas 976 12 European vee 23 HLA B 5 505 wii ER 976 16 European an 23 MNSS lt oc24 02 cre ea a ee 638 4 English 1000 24 Rh nesae ken e ta lass 580 6 English 154 25 AP oeae E EE deh baled 448 4 English 597 26 TCU rarus iird eeek a 400 4 Caucasian 131 27 GPT ee the h ee wioeeas 382 3 Caucasian 528 28 DIC te ot eas ha tcc Ba tere eel nae 375 2 English 201 29 EAA E E ere eT ee 375 2 English 1166 30 A SRO Re A ee ee SE 368 2 English 1166 30 19 111 eari a EA 367 2 English 218
27. n of enzymes A and B given cuts of a GENETIC MAP USING RESTRICTION FRAGMENTS 317 Restriction endonuclease fragment polymorphisms have also proven useful in the genetic analysis of chromosomal markers in yeast Three groups 12 14 discovered RFLPs in their systems of study Six of eight EcoR1 restriction fragments encoding tyrosine tRNA were mapped using RFLP analysis 13 15 An RFLP at a particular tyrosine locus SUP4 was shown to be generated by the presence or absence of a 5 8 kb repetitive DNA segment 14 By use of a restriction site polymorphism recombination was examined and the RFLP was subsequently used to map the DNA locus 12 RFLP analysis has also been applied to genetic mapping in the nematode worm Cenorhabditis elegans Hirsch personal communication 1979 RFLPs IN HUMAN DNA More recently RFLPs have been discovered in human systems In the region of the human globin genes a heterozygosity for a PstI restriction site has been found within the intervening sequence of the y globin gene among recombinant DNA clones 16 Apparently the two chromosomes of the individual from which the DNA was isolated were heterozygous for this marker This was confirmed by examination of a Southern transfer of DNA from the individual Hpal fragment length polymorphic variants associated with sickle cell traits have also been found 17 Although most human B globin genes were present on a 7 6 kb Hpal fragment among individuals of African
28. n typed for HLA red cell antigens and isoenzymes Standard LOD score analysis in pedigrees explained below will determine linkages between the probes and between each probe and other chromosomal markers Mostly we expect RFLPs to be inherited as Mendelian codominant alleles These might simply be the result of single base pair changes although other possibilities can be imagined Deletions additions and other local rearrangements should also manifest Mendelian inheritance RFLPs within distant translocations however will reveal unusual inheritance since the probe may be detecting sequences that although homologous are genetically unlinked Another interesting possibility is that RFLPs could result from the activity of DNA modifying enzymes 20 In this case pedigree analysis would show us the linkage relationship of the modification genes Furthermore these polymorphisms might not give the expected codominant phenotype but will be expressed as either true dominants or recessives depending on whether the gene is structural for a modifying enzyme or regulatory and if regulatory whether positive or negative Furthermore the activity of such a system might affect RFLPs detectable with several probes possibly known by mapped RFLPs to reflect sequences imbedded in separate linkage groups In this case again linkage relationships might yield apparently disparate results The two described possibilities are distinguished by their dominance
29. of the chromosomal arrangement of sites for two restriction enzymes for a pair of homologous chromo somes Restriction enzyme B cleaves at identical locations b and 5 in both chromosomes yielding only one fragment length homologous to the probe sequence which is revealed as a single band of hybridization as indicated in figure 1b Restriction enzyme A on the other hand is seen to be cleaving the homologous DNA sequences at one identical site a but the location of the next enzyme A site differs between the two homologs This difference in location between a 4 and a generates the pattern fig 1b of two bands when hybridized with the probe RFLPs AS GENETIC MARKERS RFLPs were first used as a tool for genetic analysis in 1974 Linkage of temperature sensitive mutations of adenovirus to specific restriction fragment length differences was used to locate the mutations on a physical map of the restriction fragments 9 Other studies have shown the maternal inheritance of mammalian mitochondrial DNA 10 and the existence of RFLP within mitochondrial DNAs from two individual humans as well as from several human cell lines 11 g s Enzyme Enzyme Molecular Weight i i a Chromosomal Arrangement b Hybridization Pattern restriction endonuclease A restriction endonuclease B TTI probed single copy region Fic 1 a Cuts made in pair of homologous chromosomes by enzyme A and enzyme B b hybridization patter
30. ple size degree of polymorphism of each marker locus and recombination distance between the loci in question In summary we expect that a set of about 150 markers each 0 2 Morgans apart will be required and that if their average information content PIC is 5 several L EXPECTED LOD 1072 1073 2 4 6 NUMBER OF OFFSPRING FIG 3 Expected LOD score for two recombination fractions r as a function of sibship size GENETIC MAP USING RESTRICTION FRAGMENTS 323 hundred individuals at most will be necessary to establish linkage to any given index locus During the process of mapping RFLPs we need to make more extensive pedigree studies using RFLPs which are less polymorphic If the number of subjects available for study is limited a larger set of markers could allow one to reduce the number of individuals required An expected LOD score of 3 will result from about 6 000 analyses individuals X markers over a wide range of marker densities 52 The most efficient procedure in establishing RFLPs will be to study a small set of large pedigrees which have been genotyped for all known polymorphic markers red cell antigens isoenzymes HLA and chromosome banding polymorphisms All probes showing usable PIC values will be studied further to establish segregation and define linkage groups RFLP markers will also be studied further for linkage with known markers As the study develops a set of RFLPs denoted Rix will emerge where
31. serted segments in these recombinant phage range in size from 15 to 20 kb A second library of human DNA sequences was constructed from messenger RNA L Villa Komaroff personal communication 1978 To prepare this human library messenger RNA was enzymatically copied into DNA cDNA inserted into a plasmid vector These segments of messenger sequences are each about 500 base pairs long The phage library contains segments representing the DNA sequences of the entire human genome and the cDNA library reflects those human sequences transcribed into messenger RNA providing an Opportunity to characterize each set respecting polymorphism 326 BOTSTEIN ET AL The phage clones with their large inserts will reveal many more restriction fragments than the much smaller inserts of the cDNA clones Most of the long segments however are likely to contain interspersed repetitive sequences 59 which would hybridize to fragments from many different regions of the genome The cDNA clones should mostly reflect gene sequences present in only one copy or a few see below per haploid genome but many represent DNA sequences which are more highly conserved and reveal fewer polymorphisms The liability of the large phage probes is that they generally will contain moderately repetitious sequences interspersed within single copy There is evidence that the human genome like most eukaryotic genomes is organized with moderately repetitious Sequences interspersed with
32. that a given restriction site is not polymorphic then 1 p Taking into account such parameters as average restriction fragment length AT to GC ratio and appearance of new Sites as well as disappearance of old sites Upholt 57 was able to show that f the fraction of restriction fragments showing no polymorphism will be approximately equal to 2 S If we assume a nucleotide site polymorphism of 1 1 the upper limit suggested by the mouse study described above then the probability that a restriction fragment defined by a four base enzyme would also be polymorphic is 12 3 Using the lower limit for polymorphism suggested by the protein electromorph studies 0 23 the probability of restriction fragment polymorphism becomes 0 34 The corresponding probabilities for fragments defined by restriction enzymes which recognize six base sequences are 17 7 and 0 50 Even though the probability per fragment of polymorphism is somewhat less for the fragments defined by the four base enzymes they should be more useful in revealing polymorphism due to single base changes in DNA since there are approximately eight times as many base fragments produced per unit length of DNA Polymorphism due to insertions and deletions Another possible source of polymorphism other than single base pair changes is the insertion or deletion of blocks of DNA These DNA polymorphisms would be underrepresented in the above estimates but will serve to increas
33. tion cleavage sites indicated and also in homologous cloned DNA segments C C2 C3 and C4 which may contain parts of some probes and improve recurrence risks 5 This analysis simultaneously established the recessive genetic model and demonstrated linkage to HLA Furthermore studies of additional pedigrees failed to demonstrate heterogeneity in the etiology of the disease and suggested a disease frequency approximately 40 times greater than previously estimated 63 The analysis also showed that one large informative pedigree gave twice the information available in nine smaller ones However since not all disease loci can be expected to be found in the HLA region and since other markers only slightly polymorphic would not have allowed the discrimination necessary to resolve the genetic system this approach has been severely limited The intrinsic liabilities to the marker systems are compounded by hetero geneity late age of onset variable expressivity environmental interactions and other complications found associated with many important diseases In fact known markers have failed to provide linkage to important disease loci such as Huntington disease Because of these difficulties only 2 3 of known Mendelian disorders have been mapped As an alternate approach substantial progress has been and is being made in determining the chromosomal location of many genes by the method of segregating human chromosomes from h
34. uld provide sufficient lymphocyte DNA for analysis RFLPs should be inherited as simple Mendelian codominant markers this can be verified in family studies Linkage relationships among RFLPs can be established using pedigree analysis 1 3 Evaluation of many DNA marker loci should allow the establishment of a set of well spaced highly polymorphic genetic markers covering the entire human genome Since the RFLPs are being used simply as genetic markers any trait caused wholly or partially by a major locus segregating in a pedigree can be mapped Such a procedure would not require any knowledge of the biochemical nature of the trait or of the nature of the alterations in the DNA responsible for the trait No specific gene isolation is required and the RFLPs can be random sequences functionally unrelated and physically distant from the DNA encoding the locus of interest In fact on the average the physical distance represented by 1 cM 01 recombination fraction is about 1 000 000 base pairs 1 000 kilobases kb 4 we suggest the need for linked markers about 20 cM apart i e 20 million base pairs Such a mapping procedure will also allow better definition and substantiation of models of inheritance for the many familial traits which have been refractory to simple genetic analysis in humans The usefulness of a tightly linked polymorphic locus in resolving modes of inheritance has recently been demonstrated 5 RESTRICTION FRAGMENT LENGTH
35. uman mouse fusion hybrids see 64 for review However to map disease loci this method depends upon knowledge of the biochemi cal defects at the cellular level Since the biochemical basis for most genetic disease remains obscure their loci are refractory to localization by the fusion approach or its Variations 328 BOTSTEIN ET AL DISCUSSION The application of a set of probes for DNA polymorphism to DNA available to us from large pedigrees should provide a new horizon in human genetics Recent applications of these methods specifically to hemoglobinopathies 18 19 have already provided major results Once established DNA polymorphisms will be useful for in utero assays as has been shown for hemoglobin 18 65 This is important since even when successful current linkage studies provide markers which are generally unusable for preventive medicine because they cannot be assayed in utero or because of only marginally informative polymorphism With linkage based on DNA markers parents whose pedigrees might indicate the possibility of their carrying a deleterious allele could determine prior to pregnancy whether or not they actually carry the allele and consequently whether amniocentesis might be necessary Analogously individuals potentially at risk for heritable disorders including some forms of cancer of which expression occurs late in life ought to be able to determine in advance of obvious symptoms whether they are at risk for t
36. with DNA from a heterologous species However only one attempt has been made to determine the extent of polymorphism within a mammalian species by heteroannealing DNA from separate individuals of the same species 55 This study indicates that the DNA sequence polymorphism due to single base changes between an individual wild mouse and an inbred C57B1 mouse is low In fact no melting temperature depression was observed from which one can conclude that the resultant heterozygosity was probably less than 5 Britten personal communication 1979 corresponding to a DNA polymorphism frequency of 1 1 This conversion from degree of heterozygosity to degree of polymorphism is approximate and conservative approximate in that it uses a formula 56 which assumes no Selection and conservative in that we have defined polymorphism as requiring a minor allele with a frequency of at least 10 as less polymorphic loci are not of interest for our purposes This suggests an upper limit to the extent of DNA polymorphism in mammals due to base pair changes RFLP calculation Assuming an average of the above estimates of polymorphism for human DNA due to base pair changes we can calculate the probability that a restriction fragment will be polymorphic If we let p be the fraction of nucleotide sites showing significant polymorphism frequency of the minor allele greater than 10 n the number of nucleotides in the restriction enzyme site and S the probability
Download Pdf Manuals
Related Search