Vol. 285, No. 1
`
`ISSN 0027-5107
`MUREAV 285 (1) 1- 144 (1993)
`
`~TION
`____ SEARCH
`
`HEALTH SCIENCES LIBRARY
`University of Wisconsin
`
`JAN 2 6 1993
`
`1305 Linden Drive
`Madfson, WI 53706
`
`tiona! journal on mutagenesis,
`orne breakage and related subjects
`
`1or-m-Chief: F.H. Sobcls (Leiden)
`
`ard of Managing Editors
`I \shb~. Macclesfield ; S.M. Galloway, West Point, PA (Mutation Research Letters); .I.M. Gentile,
`I /land, Mf; B.W. Glickman, Sidney, B.C.; P.C. Hanawalt, Stanford, CA (DNA Repair); P.H.M.
`l'hman. Leiden (DNA Repair); K. Sankaranarayanan, Leiden; F.J. de Serres, Research Triangle Park,
`
`I. R.B. Setlow, Upton, NY (DNAging); M.D. Shelby, Research Triangle Park, NC; T. Sugimura, Tokyo
`~~Aging); H. Takebe, Kyoto (DNA Repair); J. Vijg, Leiden (DNAging); E. Vogel, Leiden; J.S.
`1 ''Om, Oak Ridge, TN
`I
`
`f'undamental and Molecular Mechanisms
`.f Mutagenesis
`
`lSevier
`
`Special Issue
`
`In Memory of Max Clark.
`a Pioneer in Fundamental Mutation Research
`edtft•d by Donald G. MacPhee
`
`GeneDX 1020, pg. 1
`
`

`

`Mu111110n Research, Fundamemal and Molecular MeclwniSIIl\ of Mlllagenesi.\ (ISSN 0027-5107) is published
`US mailing notice -
`monthly by Elsevier Science Publishers (Molenwerf I. P.O. Box 211. 1000 AE Amsterdam. The Netherlands). Annual subscription
`price in the U.S.A. US$ 1624.00 (subject to change). including air speed delivery. Second class postage paid at Jamaica, NY 11431.
`
`USA POSTMASTERS: Send address changes to Mma1ion Research, Fundamental and Molecular Meclzamfms of Mmagenesis,
`Publications Expediting, Inc., 200 Meacham Avenue, Elmont, NY 11003.
`
`Airfreight and mailing in the U.S.A. by Publications Expediting.
`
`Advertising Advertising orders and enquiries can be sent to the Advertising Manager, Elsevier Science Publishers. Advertising
`Department, P.O. Box 211, 1000 AE Amsterdam (The Netherlands). Tel.: 20-515.3220; FAX: 20-683.30~1. attn. Advertising
`Dept.; Great Britain: T.G. Scott and Son Ltd., Portland House, 21 Narborough Road, Cosby, Leicestershire LE9 5TA. Tel.:
`0533-753.333: FAX: 0533-750.522, attn. Tim Blake; U.S.A. and Canada: Weston Media Associates. Daniel Lipner, P.O. Box 1110,
`Greens Farms, cr 06436-1110, Tel.: 203-261.2500; FAX: 203-261.0101.
`
`© 1993, ELSEVIER SCICNCE PUBLISHERS B.V. All rights reserved.
`
`No part of this publication may be reproduced, stored in a retrieval system. or transmiued. 10 any form or by any means,
`electronic, mechanical, photocopying, recording or otherwise. without the prior wrillen permission of the Publisher, Elsevier
`Science Publishers B.V., Copyright and Permissions Department, P.O. Box 521, 1000 AM Amsterdam (The Netherlands).
`
`This journal is printed on acid-free paper.
`
`No responsibility is assumed by the Publisher for any injury and/or damage to rerson~ or property as a maucr of products
`liability, negligence or otherwise, or from any use or operation of any methods, products, inst•·uctions or ideas contained in the
`material herein. Because of the rapid advances in the medical sciences, the Publisher recommends that independent verification
`of diagnoses and drug dosages should he made.
`
`Although all advertising material is expected to conform to ethical (medical) stand<mls, inclusion in this publication docs not
`constitute a gm;rantee or endorsement of the quality or value of such product or of the claims made of ir by its manufacturer.
`
`Special regulation\ for awhors. Upon acceptance of an article by the journal, the author(s) will be asked to transfer copyright of
`the article to the Publisher. This transfer will ensure the widest possible dissemination of information.
`
`Special regulations for readers in the U.S.A. This journal has been registered with the Copyright Clearance Center, Inc. Consent is
`given for copying of articles for personal or internal use, or for the personal usc of specific clients. This consent i\ given on the
`condition that the copier pays through the Center the per-copy fee stated in the code on the first page of each article for copying
`beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. The appropriate fee should be forwarded with a copy of
`the first page of the article to the Copyright Clearance Center. Inc., 27 Congress Street, Salem, MA 01970 (U.S.A.). If no code
`appears in an article the author has not given broad consent to copy and permbsion to copy mu't be obtained directly from the
`author. All articles publbhed prior to 1980 may be copied for a per-copy fee of U.S.$ 2.25, also payable through 1he Center. This
`consent does not extend to other kinds of copying such as for general distribution, resale, advertising and promotion purposes. or
`for creating new collective works. Special written permission must be ob1aincd from the Publisher for such copying.
`
`PRINTED IN THE NETHERLANDS
`
`GeneDX 1020, pg. 2
`
`

`

`\(ul/ltion Research, 285 (1993) 125-144
`·~ 1993 Elsevier Science Publishers B.V. All rights reserved 0027-SI07j93j$06.00
`
`125
`
`MUT 00047
`
`Current methods of mutation detection
`
`R.G.H. Cotton
`Olite Miller Laboratory, Murdoch lnstitllle, Royal Children's Hos(Jital, Flemington Road, Parkrille, Vic. 3052, Australia
`
`(Received 3 July 1992)
`(Revision received 17 July 1992)
`(Accepted 24 July 1992)
`
`Keywords: Detection of mutations; Mutation detection; Mismatch; Hetero<.luplcx
`
`Summary
`
`Mutation detection is important in all areas of biology. Detection of unknown mutations can involve
`sequencing of kilobases of DNA, often in many patients. This has lead to the development of methods to
`screen DNA for mutations as well as methods to detect previously described mutations. This review
`discusses current methods used for such purposes with special emphasis on genetic diseases of humans.
`However, savings can be made by sin1ilar means in other areas of biology where repetitive or extensive
`sequencing for comparative purposes needs to be done. This review covers the methods used for
`detection of unknown mutations, namely the ribonuclease, denaturing gradient-gel electrophoresis,
`carbodiimide, chemical cleavage, single-strand conformation polymorphism, heteroduplex and sequenc(cid:173)
`ing methods. Once mutations have been defined they can be searched for repeatedly by methods
`referred to as diagnostic methods. Such methods include allele-specific oligonucleotide hybridization,
`allele-specific amplification, ligation, primer extension and the artificial introduction of restriction sites.
`We can now choose from a range of excellent methods, but the choice will usually depend on the
`background of the laboratory andjor the application in hand. Screening methods are evolving to more
`satisfactory forms, and the diagnostic methods can be automated to screen whole populations inexpen(cid:173)
`sively.
`
`The rate of identification and characterisation
`of genes which cause specific inherited diseases
`in humans is rapidly increasing. Also the numbers
`of mutations identified in a particular gene as
`causative in disease arc increasing rapidly. This is
`likely to gain further impetus from the Human
`
`Correspondence: Dr. R.G.H. Cotton, Olive Miller Labora(cid:173)
`tory, Murdoch Institute, Royal Children's Hospital. Fleming(cid:173)
`ton Road, Parkville, Vic. 3052, Australia.
`
`Genome Project. Thus methods to detect un(cid:173)
`known mutations and previously described muta(cid:173)
`tions are assuming increasing prominence and
`usc. The importance of such studies is enhanced
`as knowledge of the molecular basis of cancer has
`increased, given that mutations in oncogenes and
`tumour suppressor genes are now well-docu(cid:173)
`mented as causes of cancer. Changes of such
`magnitude have not occurred in genetics other
`than the area of human disease, but nevertheless
`the methods to be discussed are potentially ad(cid:173)
`vantageous in these areas.
`
`GeneDX 1020, pg. 3
`
`

`

`126
`
`There are 3 main areas where such methods
`are used in human disease. (a) the research labo(cid:173)
`ratory attempting to find mutations in a gene
`which causes a specific disease, (b) the clinical
`diagnostic laboratory which needs to look for
`known and unknown mutations causing a specific
`disease for prenatal or other diagnostic purposes,
`as well as polymorphic harmless mutations for
`linkage studies, and (c) the screening of popula(cid:173)
`tions for specific mutations (such as has occurred
`with Tay Sachs disease and as is beginning to
`occur in cystic fibrosis).
`For this review, methods will be divided into
`screening methods and diagnostic methods. The
`former are usually used to detect unknown muta(cid:173)
`tions, but there is an increasing tendency to use
`screening methods to screen for a number of
`known mutations together with any unknown mu(cid:173)
`tations in the diagnostic setting (see below) (Fig.
`1).
`Methods of detecting mutations have been re(cid:173)
`viewed several times in recent years (Caskey,
`1987; Grompe et al., 1989; Rossiter and Caskey,
`1990; Cotton, 1989, 1991, 1992). One of these was
`particularly detailed (Cotton, 1991) and reviewed
`the area up to the end of 1990 and a subsequent
`review (Cotton, 1992) is essentially an update of
`this review to near the end of 1991. The field is
`evolving so rapidly that frequent reviews are nec(cid:173)
`essary to monitor important new methods and
`modifications of older methods; it is also impor(cid:173)
`tant to assess the effectiveness of methods after a
`time in operation.
`This review aims to provide a brief description
`of the principles and practice of methods avail(cid:173)
`able at this time together with their variants, and
`a discussion of their advantages and disadvan(cid:173)
`tages. Key illustrative applications will be given.
`For more detail (and further examples) the reader
`is referred to an earlier review (Cotton, 1991).
`Only those methods used actively at present or
`those described in the last few years will be
`covered. Methods to detect the more obvious
`deletion/ insertion mutations have been covered
`earlier (Rossiter and Caskey, 1990) and will not
`be covered here, where detection of point muta(cid:173)
`tions will be emphasized. It should be noted that
`some of the methods mentioned below will be
`reviewed in more detail in another issue of Muta(cid:173)
`tion Research.
`
`Screening methods
`
`The screening methods can be divided into
`two types: (a) those simple methods which rely on
`differences in electrophoretic properties being
`generated between mutant and wild-type nucleic
`acid by point mutations (these methods cannot,
`as currently used, detect all mutations, do not
`localize them within the fragment, and can only
`be applied to DNA fragments hundreds of bases
`long), (b) and another group which
`includes
`cleavage methods and the carbodiimide method
`(which can screen kilobase lengths and localise
`the mutations to within 10 bases in the fragment
`examined). The subcategOJy of chemical methods
`have the potential to detect all mutations. Se(cid:173)
`quencing is more frequently used to detect un(cid:173)
`known mutations than it is for diagnostic pur(cid:173)
`poses.
`
`Ribonuclease cleavage (RNAase)
`Many ribonucleases cleave single-stranded
`RNA after pyrimidine residues. This finding was
`exploited when it was found that single base-pair
`mismatches in RNA: RNA heteroduplexes were
`cleaved by ribonuclease (Freeman and Huang.
`1981; Winter et al., 1985) as well as in RNA: DNA
`heteroduplexes (Myers et al., 1985a).
`The method was given considerable impetus
`when uniformly labelled probes could be conve(cid:173)
`niently produced as described in 1984 using the
`SP6 system (Melton et al., .1984). Application of
`the method directly to unamplified genomic DNA
`has been reported (Myers et al., 1985a; Kaufman
`et al., 1990). Cleavage of the DNA to which the
`cleaved RNA was hybridized is possible via S1
`nuclease (Atweh et al. , 1988). Cleaved RNA bas
`been detected after transfer to a membrane and
`hybridization with probe (Genovese et al., 1989).
`The main advantage of the method is that it is
`a simple single-step reaction which locates the
`mutations within the fragment. This is, however,
`offset by the fact that special RNA probe has to
`be prepared and that only about 70% of aU
`mutations are detected (Myers et al., 1985a). This
`is because when purines appear in the probe at
`the mismatch most mismatches are not cleaved.
`Despite the aforementioned disadvantages the
`method has been used until the present day. For
`
`GeneDX 1020, pg. 4
`
`

`

`example, variation in HIV isolates have been
`studied (Lopez-Galindez et al., 1991), the pattern
`generated after digestion by RNAase being in(cid:173)
`dicative of geographical distribution and tempo(cid:173)
`ral appearance of resistance to AZT. In addition,
`a number of mutations in the ape gene were
`identified with ribonuclease (Nishisho et al.,
`1991). The method has also been applied to the
`intensively studied p53 gene in tumours and cell
`lines (Kim et al., 1991).
`The fact that around 30% of mutations are
`missed with this method is a considerable short(cid:173)
`coming, if a simple single-step screening method
`capable of detecting 100% of mutations becomes
`available the RNAase method is bound to de(cid:173)
`crease in both use and value.
`
`Denaturing gradient-gel electrophoresis (DGGE)
`and related techniques
`When double-stranded DNA is electropho(cid:173)
`resed into a gradient of increasing denaturant a
`portion of a given strand separates but the strands
`are anchored together by the portion (higher
`melting domain) which has not melted at this
`point. This split in the duplex suddenly arrests
`the movement of the molecule in the gel. If a
`single-base change is present in a similar duplex
`in the split portion, the denaturant concentration
`for strand separation is usually different, thus the
`arrest of movement occurs at a different position
`in the gel and a mutation can be detected by the
`differential positions of arrest (Myers et al.,
`1985b). The difference between the positions of
`arrest is greater if hcteroduplex molecules (be(cid:173)
`tween mutant and wild-type) are used. The gel is
`poured with an increasing gradient of denaturant
`(formamide) and run at 60°C in a special appara(cid:173)
`tus needed to keep the temperature constant.
`The length screened is 50-500 bp and it is possi(cid:173)
`ble to use unlabelled DNA.
`There has been considerable evolution of the
`method since it was first described, and also there
`arc a number of variants. Changes have been
`directed either to increasing the percentage of
`mutations detected or to simplifying the method(cid:173)
`ology. The most important modification has been
`the placing of a high melting point 40-base GC
`rich sequence (the GC clamp) at one end of the
`fragment to be screened. Most recently this has
`
`127
`
`been achieved using PCR technology, with spe(cid:173)
`cial primers being synthesized with a clamp at(cid:173)
`tached (Sheffield et al., 1989, 1992a). This means
`the whole area to be screened is in a low melting
`point domain and that ''almost all'' mutations,
`instead of about 50%, can be detected. In the
`practical situation it was found that mutant sam(cid:173)
`ples had to be mixed with normal DNA to ensure
`heteroduplexes were formed in order to ensure
`detection of a maximal number of mutations (Cai
`and Kan, 1990; Higuchi et al., 1990). Kilobase
`lengths of genomic DNA can be screened for
`polymorphisms (60% of any base changes) by
`digestion with restriction enzymes, separation by
`DGGE, blotting onto a membrane and then
`probing with relevant genes (Gray, 1992).
`Further modifications have attempted to avoid
`the use of the special apparatus altogether. Smith
`et al. (1988) melted the duplexes in solution con(cid:173)
`taining stepwise increases in denaturant and
`analysed them by standard polyacrylamide gel
`electrophoresis. Another variation has been to
`usc a
`temperature rather
`than a
`liquid-de(cid:173)
`naturant gradient (Rosenbaum and Rcissuer,
`1987). The most recent modification leading to
`greater simplicity has been the constant denatu(cid:173)
`rant gel electrophoresis (CDGE) method (Hovig
`et al., 1991). Here separation is undertaken at
`that concentration of denaturant which corre(cid:173)
`sponds to that of the melting domain of the
`fragment being analysed. The authors reported
`detection of 6 of 7 mutations at a particular locus
`whereas 3 of 7 were found with conventional
`DGGE. This low detection rate with conventional
`DGGE was despite the use of a GC clamp which
`is rather surprising.
`One of the special and important advantages
`of the above method (and other methods separat(cid:173)
`ing intact mutant and wild-type molecules during
`analysis (see below)) is
`the fact that mutant
`molecules can be isolated from gels for further
`analysis such as sequencing. This feature was
`exploited in the study of errors during PCR am(cid:173)
`plification (Keohavong and Thilley, 1989). Other
`advantages are the fact that it can be used in
`unlabelled mode, it can be used directly on on(cid:173)
`amplified genomic DNA, and a result can be
`obtained in 24 h. A particular disadvantage of
`almost all variants is that either preliminary ex-
`
`GeneDX 1020, pg. 5
`
`

`

`128
`
`periments or computer analysis has to be per(cid:173)
`formed before screening different DNA frag(cid:173)
`ments. Also for the original DGGE method spe(cid:173)
`cial apparatus has to be used. The method cannot
`guarantee detection of all mutations.
`The method has been frequently used since its
`(irst description until the present day. Most appli(cid:173)
`cations have been in two major categories (a)
`screening a newly characterized gene for muta(cid:173)
`tions and (b) screening a well-studied disease
`gene to detect known, and possible unknown,
`mutations in the clinical setting. Examples of the
`former include application to Factor VIII defi(cid:173)
`ciency (Higuchi et al., 1991), Rhodopsin muta(cid:173)
`tions (Sung et al., 1991) and mitochondrial DNA
`(Yoon et al., 1991) and the latter include B(cid:173)
`thalassaernia (Loosekoot et al., 1990), a-t-anti(cid:173)
`trypsin deficiency (Johnston et al., 1991) and por(cid:173)
`phyria carriers (Bourgeois et al., 1992). Applica(cid:173)
`tion of the CDGE method to p53 has been de(cid:173)
`scribed (Borresen et al., 1991).
`Because of its ability to detect "almost all"
`mutations this method and its variants are likely
`to be used for some time yet. However, its rela(cid:173)
`tive complexity (apparatus and the preliminary
`studies which are required) is likely to see it lose
`ground to SSCP (see below) due to the latter's all
`round simplicity, even although SSCP is likely tu
`detect fewer mutations than DGGE and its vari(cid:173)
`ants.
`
`Carbodiimide modification (CD!)
`Carbodiimide is a bulky reagent that reacts
`with the imino sites of T and G bases and reacts
`more rapidly with unpaired bases than with paired
`bases (e.g. Kelly and Maden, 1980). It was re(cid:173)
`cently found that mismatched T and G bases in
`DNA heteroduplexes react more rapidly than
`matched T and G bases (Novak et al., 1986).
`Thus when reacted with heteroduplexes between
`mutant and wild-type DNA, any base change
`(and therefore any mutation) should be poten(cid:173)
`tially detectable. For example, a C ~ G change
`will give rise to heteroduplexes with C · C and
`G · G mismatches and both and bases should be
`detectable. With a T ~ G change the mismatches
`generated, T · C and G · A, should both be de(cid:173)
`tectable.
`
`The original method (Novak et al., 1986), de(cid:173)
`scribed detection of the reaction of CO l with the
`mismatches by slowing of the electrophoretic mo(cid:173)
`bility of the DNA fragment during electrophore(cid:173)
`sis. Later, detection by cleavage of the hcterodu(cid:173)
`plex at the derivatized mismatch point by ABC
`excinuclease was used (Thomas et al., 1988). Nei(cid:173)
`ther of these methods were applied to a practical
`situation. Detection of the derivatized mismatch
`by electron microscopy was made possible by the
`production of antibodies to the carbodiimide
`(Ganguly et al., 1989). A more practical protocol
`was established when it was shown that the
`derivatized T and G bases would block primer
`extension (Ganguly and Prockop, 1990) and this
`is the protocol that has mainly been applied to
`the practical situation. Thus detecting elec(cid:173)
`trophoresis of products shorter than the full seg(cid:173)
`ment length, signals the presence of a mismatch.
`The main advantage of the most recent proto(cid:173)
`col is that no probes have to be produced, since
`the label is incorporated during primer extension;
`also, the one chemical which is used is relatively
`non-toxic. Some specific minor problems are
`mentioned
`in
`the description (Ganguly and
`Prockop, 1990) but it seems that two sets of
`incubation conditions are needed to guarantee
`t:omplete detection. Because two detectable mis(cid:173)
`matched bases arc generated from any muta(cid:173)
`tional change any mutation has two chances of
`detection. A theoretical objection is that when a
`T · G mismatch is generated, as it is the most
`stable mismatch (Cotton, 1989), both chances of
`detection may be negative. While examples of all
`possible T and G mismatches have been shown to
`be detectable, the method has not seen wide
`enough application for potential problems to be
`discovered.
`All practical applications so far have been
`from the laboratory where the method originated.
`Point mutations in the collagen genes COLIAl
`(Zhuang et al., 1991) and COLIA2 (Ganguly et
`al., 1991; Spotilla et al., 1991) were detected and
`positioned by the carbodiimide method. It should
`be noted that in one of these cases (Ganguly et
`a!., 1991) the position of a splice site mutation
`was ascertained in plasmid inserts 3-4.5 kb long
`by immuno-electron-m icroscopy.
`
`GeneDX 1020, pg. 6
`
`

`

`This illustrates a special feature of the carbodi(cid:173)
`irnide method, which is the ability to screen for
`mismatches in very large heteroduplexes (7.2 and
`4.9 kb) by immuno-electron-microscopy (Ganguly
`et al., 1989). This capability, for those with ready
`access to immuno-electron-microscopy, is unique
`to the carbodiimide method and could be of
`enormous value in future. The limited application
`so far has not allowed specific problems of the
`CDI method to be identified and possibly elimi(cid:173)
`nated. Other than this, it remains a viable alter(cid:173)
`native to the chemical cleavage method but it is
`not yet available in an unlabelled form.
`
`Chemical cleauage of mismalch (CCM)
`This method is based on the fact that mis(cid:173)
`matched C bases and mismatched T bases in
`heteroduplcxes were found to be more reactive
`with hydroxylamine and osmium tetroxide respec(cid:173)
`tively than equivalent matched base-pairs (Cotton
`et al., 1988). The point of reaction can be readily
`ascertained by further reaction with piperidine,
`which cleaves the strand containing the mis(cid:173)
`matched base (Cotton et al., 1988). Mismatched
`G and A bases are ascertained by use of the same
`probe but of opposite sense which transposes
`them to mismatched C and T bases respectively.
`Use of mutant probe as well as wild-type probe
`ensures each mutation has two chances of being
`detected making it unlikely that any mutation is
`missed (see below).
`There have been many variations and improve(cid:173)
`ments described since the initial publication. The
`method has been shown to be readily applicable
`to mRNA (Dahl et al., 1989) and viral RNA
`(Cotton and Wright, 1989). In this latter applica(cid:173)
`tion, end-labelled probe and incomplete reaction
`allowed a pattern of difference or fingerprint to
`be generated between viral strains. Detection of
`mutations can be indirect, since mismatches can
`have a destabilizing influence on nearby matched
`T and C bases such that they become reactive
`with the two chemicals (Cotton and Campbell,
`1989). To determine if a mutation is present in
`the homozygous or heterozygous state after it has
`been detected, mutant probe can be made for
`hybridization with mutant sample. Thus homozy(cid:173)
`gous samples will give no signal whereas het(cid:173)
`erozygous samples will give a detectable mis-
`
`129
`
`match (Dianzani et al., 199la). A class of T · G
`mismatches which were urreactive with osmium
`tetroxide (Theophilus et al., 1989; Forrest et al.,
`1991; Anderson et al., 1992) led to a modification
`in the strategy recommended for mutation detec(cid:173)
`tion using CCM, when it was suggested that mu(cid:173)
`tant probe be made for hydridization with wild(cid:173)
`type DNA (Forrest et al., 1991). This would en(cid:173)
`sure detection of the complementory C · A mis(cid:173)
`match with hydroxylamine. It was first expected
`that this would generate twice as much work but
`it was found that by hybridizing mutant and
`wild-type probe together in equimolar quantities
`and reacting them in the one tube mutations
`could be readily detected (J. Saleeba, unpub(cid:173)
`lished). Earlier protocols recommend 10 x excess
`unlabelled target. This protocol was indepen(cid:173)
`dently published by others (Han and Sternberg,
`1990). Labelling the probe with 35S has been
`shown to give sharper bands (Saleeba and Cot(cid:173)
`ton, 1991). Most recently an unlabelled variant
`has been developed (Saleeba et al., 1992) where
`mutant and wild-type DNA are mixed in equimo(cid:173)
`lar quantities, reacted with the chemicals and the
`gels are stained with silver after electrophoretic
`separation. This procedure is thought best ap(cid:173)
`plied to heteroduplexes less than 600 bp long.
`Permanganate has been suggested as an alterna(cid:173)
`tive chemical
`to osmium
`tetroxide for mis(cid:173)
`matched T bases (Gogos et al., 1990), but confir(cid:173)
`matory studies have not been published and pre(cid:173)
`liminary experiments do not support this sugges(cid:173)
`tion (R. Cotton, unpublished).
`The main advantage of this method is its abil(cid:173)
`ity to detect 100% of mutations and it must
`therefore be compared with using direct sequenc(cid:173)
`ing as a screening method. Its ability to simulta(cid:173)
`neously screen I -2 kb lengths of DNA is also a
`significant advantage over direct sequencing with
`an 11 kb screen taking only a person week (Ro(cid:173)
`berts et al., 1992). Its major drawback is the need
`for many manipulations to be performed in a
`fumehood due to the toxic chemicals, and also
`the need for the two-step reaction and treatment
`with 3 chemicals.
`Applications have been steadily
`increasing
`since the first description. Most of these applica(cid:173)
`tions have been searches for mutations in genes
`in order to define those causing disease. These
`
`GeneDX 1020, pg. 7
`
`

`

`130
`
`have recently included p53 mutations in colorec(cid:173)
`tal cancer (Rodrigues et al., 1990), simultaneous
`screening for mutations in ~-thalassaemia (Di(cid:173)
`anzani et al., 199Ib), ornithine transcarbamylase
`(Grompe et al., 1991a), collagen (Pro a 1(1))
`(Valli et al., 199 J), BCL2 oncogene (Tanaka et
`al., 1992), ~-hexosaminidase (Akli et al., 1991),
`topoisomerase II (Bugg et al., 1991) and cystic
`fibrosis (Strong et al., 1991; Jones et al., 1992).
`Mutations in a bacterial gene have also been
`detected (Grompe et al., 1991 b). Large lengths of
`DNA have been screened for mutations in indi(cid:173)
`vidual patients. These include factor VIII in a
`screen of about 8 kb (Naylor et al., 1991) and
`dystrophin eDNA in a screen of about 11 kb
`(Roberts et al., 1992). The fact that large tracts of
`DNA can be screened is underlined by the 80 kb
`screened around gene targeted sites (Zheng et
`al., 1991). A novel application was that of muta(cid:173)
`tion profiling of selectable elements. Here a se(cid:173)
`lectable element containing randomly distributed
`point mutations was subjected to selection. Muta(cid:173)
`tions were then assayed by CCM and the cleavage
`fragments were compared with the unselected
`controls (Wurst and Pohl, I 991).
`The future may see increased use of the unla(cid:173)
`belled form of the method as well as its improve(cid:173)
`tens of kilobases need
`to be
`ment. Where
`screened it may well be the method of choice
`even over sequencing as currently performed. De(cid:173)
`spite numerous trials, alternative chemicals have
`not been encountered (R. Cotton, unpublished).
`
`Single-strand conformation polymorphism (SSCP)
`The SSCP method relics on the fact that sin(cid:173)
`gle-strand DNA in solution under certain condi(cid:173)
`tions has a cfefined secondary structure. This sec(cid:173)
`ondary structure can be altered when one of the
`bases is changed and the alteration in secondary
`structure is detected by electrophoresis in non(cid:173)
`denaturing gels. Thus normal and mutant strands
`will have a different mobility (Orita et al.,
`1989a,b). The method is analogous to DGGE in
`analytical terms, but is far simpler.
`Quite a few improvements and adaptions have
`been described in the last two years. The adap(cid:173)
`tion of precast gels and the application of silver
`staining has eliminated the use of radioactivity,
`and has improved resolution (Dockhorn-Dwor-
`
`nickzak et al., 1991; Mohabeer et al., 1991) but
`ethidium bromide staining is a more rapid proce(cid:173)
`dure to perform (Yap and McGee, 1992). Be(cid:173)
`cause mutations can be most effectively screened
`for in fragments several hundred base-pairs long,
`two strategies have been introduced to increase
`the length of the screen. In one case DNA (900
`bp) is cut with several frequent-cutting restriction
`enzymes and subject to SSCP analysis (Iwahawa
`et al., 1992). This both
`increases the length
`screened and improves the chances of mutations
`being detected due to different fragment contexts
`for the mutation. In the other case, 2.7-kb DNA
`was cut with frequent-culling restriction enzymes,
`separated according to size in a denaturing gel in
`a first dimension, and then analysed on a non
`denaturing gel in the second dimension by SSCP
`(Kovar et al., 1991). As in DGGE, alleles and
`minor components can be separated for direct
`sequencing (Suzuki et al., 1991; Hata et al., 1990).
`More recently the ARMS technique (see below)
`has been combined with SSCP for HLA genotyp(cid:173)
`ing (Lo et al., 1992). This involves selection of
`specific groupings of polymorphisms by allele
`specific amplification (ARMS) followed by analy(cid:173)
`sis of variation within these groups by SSCP with
`silver staining or fluorescent primers. This should
`allow rapid and complete analysis for matching
`purposes and for analysis of other complex sys(cid:173)
`tems. Recently the most marked modification has
`been to transfer the analysis to RNA copies of
`DNA (Danenberg et al., J 992; Sarkar ct al., 1992).
`The rationale is that RNA can assume more
`elaborate and greater numbers of conformational
`forms, and these conformational forms appear to
`be sensitive to single-base substitutions. Both of
`the groups demonstrated the increased sensitivity
`of RNA SSCP over SSCP, but the latter group
`performed a detailed comparison (Table 1).
`It is thus clear that RNA SSCP leads to detec(cid:173)
`tion of an increased percentage of mutations.
`The most clear cut advantage of SSCP as used
`on DNA is the apparent simplicity of the method.
`Use without label is also an important advantage.
`However, this is offset by the most serious draw(cid:173)
`back which is the lack of 100% detection. In one
`of the so far rare analyses of the proportion of
`mutations detected, the number of mutations
`missed in the 183- and 307-bp pieces is cause for
`
`GeneDX 1020, pg. 8
`
`

`

`TABLE 1
`MUTATIONS DETEC.."TED BY DNA SSCP AND RNA
`ssCP ( %)
`
`Number of
`mutations studied
`
`DNA SSCP RNASSCP
`
`!83 bp
`307bp
`520 bp
`rx mula-
`lions
`
`12
`22
`
`20
`
`83
`58
`Minority
`
`93
`77
`Minority
`
`35
`
`70
`
`Data from Sarkar ct al. ( 1992).
`
`131
`
`concern, particularly in the blinded study of the
`factor IX mutations (Sarkar et al., 1992) (Table
`1). It was stated that if shorter pieces were anal(cid:173)
`ysed a better percentage of these mutations would
`be detected but it would take "substantially more
`effort". 1% false positives were reported in this
`study. When compared with OGGE, SSCP de(cid:173)
`tected 5 of 7 polymorphisms found with the for(cid:173)
`mer (Sheffield et al., 1992b).
`A definite improvement in the proportion of
`mutations detected was shown with the use of
`RNA SSCP (Table 1), but this is at a cost of an
`
`NORMAL
`
`MUTANT
`
`CDI
`
`DGGE
`
`SSCP
`
`HET
`
`RNAse ,CCM
`
`ASO
`
`LIG
`
`ASA
`
`PEX
`
`AIRS
`
`COl conjugate s lower in gel
`or blocks primer extension
`
`mutant slower in gel
`
`mutation changes shape
`and mobility in gel
`
`bubble makes mutant duplex
`slo