Mini- and MicrosatellitesClaes RamelDepartment of Genetic and Cellular Toxicology, Stockholm University,Stockholm, Sweden

While the faithful transmission of genetic information requires a fidelity and stability of DNA thatis involved in translation into proteins, it has become evident that a large part of noncoding DNAis organized in repeated sequences, which often exhibit a pronounced instability and dynamics.This applies both to longer repeated sequences, minisatellites (about 10-100 base pairs), andmicrosatellites (mostly 2-4 base pairs). Although these satellite DNAs are abundantly distributedin all kinds of organisms, no clear function has been discerned for them. However, extension oftrinucleotide microsatellite sequences has been associated with several severe human disorders,such as Fragile X syndrome and Huntington's disease. Rare alleles of a minisatellite sequence

have been reported to be associated with the ras oncogene leading to an increased risk forseveral human cancers. A dynamic behavior of repeated DNA sequences also applies totelomeres, constituting the ends of the chromosomes. Repeated DNA sequences protect thechromosome ends from losing coding sequences at cell divisions. The telomeres are maintainedby the enzyme telomerase. Somatic cells, however, lose telomerase function and gradually die.Cancer cells have activated telomerase and therefore they acquire immortality. Environ HealthPerspect 105(Suppl 4):781-789 (1997)

Key words; minisatellites, microsatellites, amplification, genetic instability, fingerprinting,transpositions, recombination, mismatch repair, telomere

Introduction

Since around 1970 our concept of DNAhas undergone a pronounced modificationin two respects: the stability of DNA andthe organization of the genetic material inliving organisms. Previously DNA wasconsidered a highly stable entity, whichwas subjected only to alterations throughrare mutational events. This notion wassupported by experimental evidence in the1950s, which showed that DNA replicationproceeds with an extremely low frequencyof errors. The stability and precision ofthe genetic system were manifested by the

This paper was prepared as background for theWorkshop on Susceptibility to Environmental Hazardsconvened by the Scientific Group on Methodologiesfor the Safety Evaluation of Chemicals (SGOMSEC)held 17-22 March 1996 in Espoo, Finland. Manuscriptreceived at EHP 5 November 1996; accepted18 November 1996.

Address correspondence to Dr. C. Ramel,Department of Genetic and Cellular Toxicology,University of Stockholm, S-106 91 Stockholm,Sweden. Telephone: 46 8 16 20 51. Fax: 46 8 612 4004. E-mail: claes.ramel@genetics.su.se

Abbreviations used: AR, androgen receptor;HNPCC, hereditary nonpolyposis colon cancer;IDDM, type diabetes mellitus; IGM, immunoglobulinheavy-chain gene; INS, insulin gene; LINE, long inter-spersed element; MVR, minisatellite variant repeat;PCR, polymerase chain reaction; RAP, repeat associ-ated point mutations; RIP, repeat-induced point muta-tions; VNTR, variable number of tandem repeats.

formulation of the "Central Dogma,"which holds that the flow of informationin the cells proceeds from DNA to RNAand from RNA to proteins.

In the 1970s it was found that DNAwas far more dynamic than anticipated andthat the central dogma was not invariablycorrect. The detection of reverse transcrip-tase showed that RNA could be transcribedto DNA; this had important consequencesfor many cellular processes, such as theinsertion of mobile elements, the founda-tion of pseudogenes from mRNA, andretroviral replication.

The other area of DNA research, forwhich the last two decades have provided afundamentally new concept of DNA, con-cerns the organization of the genetic mater-ial in higher organisms. The vast majorityof DNA-about 97% in human DNA-does not give rise to any proteins. The func-tioning of all this DNA has been obscurefrom the beginning and it was named "self-ish DNA" by Orgel and Crick (1). Othernames, such as junk DNA, parasite DNA,and extra DNA illustrate the confusionabout the functioning of this DNA. Someorganisms have a remarkably high amountof DNA. Thus amphibians can have upto 20 times more DNA than man (1).The reason for this spectacular variation

in DNA content between organisms isstill obscure.

Some of the noncoding DNA occurs asintrones, which are spliced away beforetranslation, but most of it is organized asrepeated sequences, which somewhat consti-tute "biological dynamite," in the sense thatthey are apt to exhibit a high degree of insta-bility and thus are responsible for much ofthe instability ofDNA mentioned above.

Research in more recent years onrepeated sequences of DNA has given aninsight into the behavior and biologicalconsequences of changes in these units.Alterations, particularly of microsatellites,have been shown to be connected with sev-eral severe human disorders. Although itappears that these repeated sequences donot provide any obvious benefit to theorganism, the fact that their alteration canhave severe effects nevertheless indicatessome kind of a function behind the occur-rence of these seemingly nonsense DNAsequences. In the present report an attemptwill be made to summarize current knowl-edge of mini- and microsatellites as well assome other repeated sequences of DNA.

Major Types of RepeatedDNA SequencesAmplification of DNA is not restricted tononcoding sequences; many coding genesare also amplified or can go through such aprocess. Amplification of nuclear onco-genes, leading to an increase of theirexpression, is connected with the inductionof cancer. In many cases an amplificationof genes constitutes a defense mechanismfor detoxication of exogenous agents, i.e.,amplification of metallothioneins againstheavy metals and the amplification up to1000 times of the dehydrofolate gene giv-ing resistance to methotrexate (2). InDrosophila, a specific amplification control-ling element is responsible for the amplifi-cation of corion genes during the earlydevelopment stages of the embryo.

The ribosomal DNA in the proximalheterochromatin of Drosophila occurs as ahighly amplified gene, and the optimal num-ber of genes is gradually restored in case of adeletion of a part of the heterochromatin.

Concerning shorter and mostly noncod-ing repeats of DNA, which are the mainsubject of this presentation, we can recog-nize two classes-minisatellites: up to100 bp, but mostly about 9 to 30 bp; andmicrosatellites: 2 to 4 bp, telomeres andtelomerlike sequences, and centromeres.

Environmental Health Perspectives * Vol 105, Supplement 4 * June 1997

--------

m

781

C. RAMEL

MinisatelJitesOccurrence. Minisatellites are regions ofthe genome with noncoding, tandemlyrepeated sequences of up to about 100 bp(3,4). The number of minisatellite loci inthe human genome has been estimated tobe 1500 per haploid genome (5,6). Manyof these loci exhibit an extreme polymor-phism due to variation in the number ofrepeats, called variable number of tandemrepeats (VNTR). The background of thisgenetic variability is a mutation rate thatcan exceed 10% per gamete (7). Jeffreys etal. (3) detected and developed DNAprobes that are able to simultaneouslydetect large numbers of hypervariable mini-satellite loci. Hybridization to digested andelectrophoresed DNA with these coresequences at low stringency detects a pat-tern of fragments that is unique for unre-lated individuals. These properties ofminisatellites provide the background forthe "fingerprint analyses" (8), which havefound several important applications,including as a powerful tool in forensicmedicine, as markers for linkage studiesin genetic analyses, and as a means forestablishing kinship between individuals,including paternity determination. Throughthe development of a system of polymerasechain reaction (PCR) (minisatellite variantrepeat [MVR], below) by Jeffreys and co-workers, it is possible to analyze singlepairs of minisatellite alleles (9). This hasenabled a measurement of changes in thenumber of repeats and also the occurrence,frequency and location of point mutationsalong the sequence of repeats in a singleallele. This development has been impor-tant to the study of the mechanism forgenetic changes of these repeated sequencesand the genetic instability involved.

The minisatellites of the humangenome are not evenly distributed but areprimarily localized at the ends of the chro-mosomes, which implies a limitation in theuse of these sequences in linkage analyses(10). This subtelomeric localization of thehuman minisatellites is correlated with ahigh density of chiasmata during meiosis,indicating an association with meioticcrossing over (11,12). The human X chro-mosome has few minisatellites, but there isa cluster of minisatellites in the X-Y pair-ing region (13). Clusters of minisatellitesat the ends of the chromosomes do notapply for the mouse genome (14).

Techniques of Typing MinisateUlites.As mentioned above, variation in theminisatellite pattern of length mutationscan be studied by means of restriction

analyses and the use of probes, which canhybridize with a large number of minisatel-lite loci. The analytical technique has beenfurther developed by the use ofPCR ampli-fication, giving an additional sensitivity.However, the disadvantage of PCR analysisof length variation is that many minisatel-lite alleles are too long for efficient amplifi-cation. Jeffreys et al. (9) have introduced anew PCR system, MVR, which has implieda solution of that problem and providedincreased sensitivity by enabling analyses ofinternal and often subtle variations of inter-nal repeats. The method is based on the useof primers, which are specific for repeatvariants and which enable a successive PCRanalysis of long stretches of repeatedsequences with occasional variants. Suchinternal variation is present in almost allminisatellites.

Mutational Changes. The mutationfrequency of minisatellites does not seem tobe dependent on the length of the allele inthe same way as in microsatellites (15).Short arrays of repeats can be stable overmillions of years (16), while long alleles canhave an extremely high frequency of muta-tional changes (up to 15%). The high insta-bility of some human minisatellites seemsto be a property of the repeated sequenceitself, as indicated by the fact that theunstable human minisatellite MS1 retainedits instability also after being inserted intothe genome of yeast, Saccharomyces (17). Infive highly unstable loci the rate of length-change mutations was related to theirobserved heterozygosity, indicating that thechanges were selectively neutral.

Mutational changes of minisatellites arenot randomly distributed, but occur pre-dominantly at one end of the locus. Thispeculiar polarity was revealed by MVR-PCR analysis of three minisatellite loci (9).The occurrence of such polar hot spots wasalso found in pedigree analysis of germlinemutations (18).

Different mechanisms of germlinelength changes of minisatellites can be visu-alized-replication slippage, intramolecularrecombination, unequal sister chromatidexchange, and unequal interallelic recombi-nation or gene conversion (10). The factthat no length change has been recordedinvolving an exchange of flanking markerseliminates a simple crossing-over model.About half of the length mutations recordedfor three alleles studied by Jeffreys' group(18) were formed through small patchexchange between alleles, presumablyinvolving gene conversionlike events (aprocess through which an allele in one of

the chromosomes is replaced by an allele inthe other chromosome). Some mutationsare of intraallelic origin. Anomalous repeats,not corresponding to either allele may havebeen brought about by mismatch repair.

Dubrova et al. (19) studied minisatel-lite length germline mutations in malemice induced by 0.5 or 1.0 Gy y-radiation.The frequency of mutation was consider-ably higher than other end points, but thedoubling dose effect was approximately thesame. The data indicated that the selectionagainst the mutations was insignificant.

Practical Application ofMinisatelliteFingerprinting. The extreme individualvariation of minisatellite and microsatellitepattern (below) has provided a new andexceedingly efficient tool for the recognitionof individuals by their electrophoretic pat-tern. Already in the original fingerprintanalysis the chance of two unrelated personsexhibiting the same pattern was extremelylow-theoretically somewhere around10-12. Later methodological improvementshave increased the sensitivity.

The possibility of amplifying DNA byPCR has made it possible to use extremelysmall material for minisatellite typing-single hairs or tiny blood stains. The finger-printing of satellite DNA -therefore has lentitself to analyses in forensic medicine andalso in historical and archeological samples.The use of minisatellite fingerprinting inlegal contexts has, however, caused muchdebate. Critical comments have emphasizedthe risk for deficient laboratory control,lack of clear definition of match of elec-trophoretic bands, dependence on the gelsystem used, and the question of statisticalweight of apparent match between samples.Furthermore, it has been pointed out bythe dominant critics Lewontin and Hartl(20) that error may be brought up by varia-tion in allele frequency between subpopula-tions. A point of particular relevance inpaternity establishment are germline muta-tions. Some prudence has been justifiedwhen introducing minisatellite fingerprint-ing, i.e., for forensic and legal purposes, butit seems that these possible sources of errorscan be overcome and they are largely miti-gated by the MVR-PCR typing system ofboth alleles for both length variation andinternal variation (above). Although thereliability of the fingerprinting method hasbeen questioned in some conspicuous legalcases, the use of this tool nevertheless hasbecome more and more a routine procedurein forensic medicine.

The occurrence of minisatellites andother repetitive DNA sequences is not

Environmental Health Perspectives - Vol 105, Supplement 4 * June 1997782

MINI-AND MICROSATELUTES

restricted to humans and other mammals,but they have a wide distribution throughoutthe organism world. The use of minisatelliteand microsatellite typing has become animportant and highly valuable new tool inpopulation ecology (21). Soon after the dis-closure of the highly variable minisatellitesby Jeffreys and co-workers in humans, inves-tigations by Burke and Bruford (22) showeda pronounced variation in fingerprintswithin and between bird species. A popula-tion analysis of house sparrows throughanalyses of blood samples from each individ-ual by fingerprinting demonstrated the use-fulness of fingerprint mapping for analyses ofpopulation structure, mating selection, andvarious polygamous pairing strategies.Repetitive DNA sequences in sufficientlystable minisatellites and microsatellites arealso of use in phylogenetic investigations ofevolutionary processes by comparisons ofspecies, subspecies, and populations.

Association ofMinisatellites withHuman Diseases. While no obvious evolu-tionary advantage of minisatellites can bediscerned at the present state of our knowl-edge, there are some cases of pathogenicminisatellites. The best known case concernsthe minisatellite connected with the Ha-rasprotooncogene locus, HRASI VNTR Thisminisatellite is located 1000 bp downstreamof the polyadenylation signal (23). It con-tains repeat units of 28 bp, forming about30 alleles. Four of these, comprising 94% ofthe alleles, have given rise to the other alleles(24). The rarer alleles are three times morecommon in cancer patients than in controlsand these alleles are associated with multipleforms of cancer. The data indicate that theycontribute to 1 of 11 cases of cancer (25).The odds ratio for the association betweenthe rare HRAS1 minisatellite alleles andcancer were, according to Krontiris et al.(26), as shown in Table 1.

Concerning the mechanism behind theassociation between the HRASI minisatel-lites and cancer two possibilities have beendiscussed (25,26). The rare alleles mayexhibit a linkage with a potential disease

Table 1. Odds ratio for the association between rareHRAS1 minisatellite alleles and cancer.a

Type of cancer Odds ratio pValue

Bladder 2.30 >0.001Breast 1.68 0.001Colorectal 2.21 0.002Leukemia 2.29 0.001Lung 1.55 0.051Melanoma 1.56 0.091

"Data from Krontiris et al. (26).

locus and these alleles would then just bemarkers for the risk of cancer. Consideringthe fact that the high-risk alleles derivefrom all the four common alleles and pre-sumably from many ancestral chromo-somes, this hypothesis is not likely. Analternative hypothesis is based on the find-ing that the HRASI minisatellite binds tothe reINF-icB family of transcriptionalregulatory factors (27,28). It is suggestedthat pathogenic minisatellite mutationsmay disrupt nonpathogenic interactionswith rel proteins.A somewhat similar pathogenic situation

is indicated for minisatellite mutationslinked to the insulin gene (INS). The mini-satellite INS VNTR is located 600 bpupstream of the transcriptional start site(29). The minisatellite is composed of 14bp repeat units arranged in three allelicdasses with modal lengths of 600 (Class I),1200 (Class II) and 2200 (Class III). Thepresence of Class I minisatellite is associatedwith a doubling of the relative risk for type Idiabetes mellitus (IDDM) (25). At least sixgenes, IDDM 1 to 6, contribute to the riskfor diabetes, and IDDM2 has been mappedto the INS VNTR minisatellite. The INSminisatellite, furthermore, binds to a specifictranscription factor, Pur-1. However, in thiscase the high-risk allele exhibits a weakertranscriptional effect than the low-risk alle-les. Nevertheless, the sequence compositionof the individual repeat units, in addition tothe total length of the minisatellite, govemsthe transcriptional response (30).

It is likely that other pathogenic effectsof minisatellites will be revealed in thefuture. It can be mentioned now that aminisatellite upstream of the humanimmunoglobulin heavy-chain gene IGHenhancer may have a suppression effect onimmunoglobulin gene expression by tran-scriptional control in the same way asHRASI and INS minisatellites (31).

The minisatellites of the HRASI, INSand IGH genes do not have any homolo-gous counterpart in nonprimate genes andit is therefore unlikely that they constitutetrue transcriptional elements; rather, theyare recent acquisitions (25). It is morelikely that the variation of minisatellitessometimes produces products that interactwith transcriptional factors and that, aslong as the effect on transcription keepswithin a narrow range, it will not bestrongly selected against (25).Conclusions. The wide occurrence ofhighly variable minisatellite sequences hasprovided indispensable tools in geneticlinkage analyses, forensic medicine, pater-

nity determination and population ecology.Also, it can be foreseen that the use of min-isatellites and other repeated DNAsequences will play an even more essentialrole in the future, both in research and forvarious practical applications.The reliabilityof the fingerprinting of minisatellites forlegal purposes has been the subject of dis-cussion and some controversy. However,the application of new PCR techniques,improved control of the laboratory proce-dures, and more experience with minisatel-lite patterns in subpopulations can beexpected to remove many of the problemsunder discussion. The occurrence of patho-genic minisatellites has given anotherdimension to this field of research. TheHRASI minisatellite seems to be of majorimportance in the cancer panorama-atleast 50,000 cases of cancer a year can beexpected to depend on the rare alleles ofthis minisatellite (26). Another importantfinding is the connection between a mini-satellite linked with the insulin locus INSand type I diabetes. A minisatellite linkedto the enhancer of the immunoglobulingene IGH is a third potentially importantcase. In all these cases, the effect of the mini-satellites seems to occur through bindingto transcription factors.

MirsatdlitesOcurrence. Microsatellites are repetitivesequences of mostly 2 to 4 nucleotides witha widespread occurrence particularly inmulticellular organisms. In the humangenome, dinucleotide repeats occur onaverage every 30,000 bp and somewhatless frequently for the more complexunits (32). These repeats therefore consti-tute a significant part of human DNA.Concerning the evolutionary significanceof microsatellites, hardly anything but dis-advantages can be discerned. Severalhuman disorders have been attached toamplification of microsatellite sequencesand other evidence of negative effects ofmicrosatellites can be traced. Formation oftandem duplications of the short sequencesthat build up microsatellites can easilyoccur as an error during DNA replication,and further amplification through strandslippage can occur in successive DNAreplication, giving rise to longer stretchesof minisatellite repeats. An accumulationof dispersed repeated sequences of simplenucleotide units can be expected to implyan increased risk of homologous recombi-nation between chromosomal segmentsand resulting in translocations, deletions,and inversions. Filamentous ascomycetes

Environmental Health Perspectives * Vol 105, Supplement 4 * June 1997 783

C. RAMEL

such as Neurospora crassa do not seem to tol-erate the burden that repetitive and appar-ently useless DNA implies. Presumably as aconsequence, Neurospora has only 10%repetitive DNA as compared to 50% inhigher organisms (33). To counteract theaccumulation of dispersed homologousmicrosatellite sequences, these sequences aresubjected to a high mutation rate through"repeat-induced point mutations" (RIP).All repeated sequences above 1 kilobase inNeurospora show signs of "RIPping." Theprimary functf.n of RIP is to protect theorganism not 1y against "parasite DNA"but also aga viruses and transposons(33). Althou ;- seems that higher organ-isms are less .sitive to these repeatedsequences, the1s are reasons to believe that asimilar protective device is operating also(below). At our present state of knowledgeit is difficult to visualize any positive biolog-ical function at least for most microsatellitesand it thus seems that they constitute true"parasite" or "selfish" DNA in the senseoutlined by Orgel and Crick (1).

Although there are principal differencesbetween micro- and minisatellites, the bor-derline between them is arbitrarily set onthe bases of the length of the repeat units.From an evolutionary point of view, it islikely that minisatellites can be generatedfrom microsatellites. Two hypotheses havebeen presented to account for the commoncore of minisatellites (34). According to atransposition model proposed by Jeffreys etal. (3), related core sequences betweenminisatellites are the result of transposi-tions mediated by sequences flanking theminisatellites VNTR. In support of thishypothesis, there are observations indicat-ing an association of minisatellite VNTRswith dispersed repetitive elements such ashuman Alu and transposonlike sequences.Sequence divergence is brought about bysubsequent mutational changes, which arecarried to other repeats of the tandem arrayby unequal exchange. However, manyminisatellites with related core sequencesdo not exhibit such an association with dis-persed repeats flanking the tandem array(3,34), making it unlikely that theyemanate from this kind of a transpositionprocess. Another model, the expansionhypothesis, is based on the concept of coresequences containing motifs that enhancethe expansion of tandem repeats indepen-dently at different loci (3). Short tracts ofsimple repeats would serve as the raw mate-rial for expansion by slipped strand mis-pairing into more complex minisatellites.This model would predict that one could

trace the development from microsatellitesto minisatellites by "fossils" of microsatel-lites in close association or interdispersedwith minisatellite VNTRs (34). Severalexamples of such an association havebeen recorded, indicating the generation ofminisatellites from microsatellites.

Analytial Methodsfor Microsatelites.Simple tandem repeat loci have beenisolated from genomic libraries by hybrid-ization screening, using relatively shortoligonucleotide repeat sequences. However,the experience from isolation of minisatelliteloci suggests that the use of long (>200 bp),tandemly repeated probes is more efficientthan short probes to isolate longer tandemarrays; and longer probes would also bettertolerate interspersed variant repeats. Armouret al. (35) therefore developed a more effi-cient system for the isolation of shortrepeats. Their system is based on a priorenrichment for tandemly repeated DNAfragments by hybridization to long tan-demly repeated targets. A library of restric-tion fragments with appropriate linkersfor PCR amplification is constructed.From amplified fragments of 400 to 1000bp, tandem repeat-containing fragmentsare selected by hybridization to long arraysof either mixed trimeric or mixedtetrameric repeat sequences. Both naturaland synthetic sequences were used as tar-gets in the hybridization selection. Thisenrichment procedure enables a rapid iso-lation of a large number of microsatelliteclones. In this way, Armour et al. isolated46 tandem repeat arrays (27 tetrameric,19 trimeric), which were sequenced andcharacterized (35).

Instability and Mutational Changes.Many microsatellites are unstable-in somecases exceedingly so. In particular, CG-richtrinucleotide and CA dinucleotide repeatsexhibit high instability and they are ordersof magnitude more variable than other tan-dem repeats. The reason for this specificityin instability is not known. In extreme casesall cells in the organism have differentlengths of the microsatellite (32). Theinstability is highly influenced by the lengthof the microsatellite with an increased insta-bility with increasing length. CG-rich trin-ucleotides and CA dinucleotides form fourgroups (32):* Short repeat length-stable* Repeats of middle length-polymor-

phic, but stable between generations* Long alleles-instability increased by

several orders of magnitude, constitut-ing "premutational alleles," which arenot stable between generations

* Long and extremely unstable alleles,also exhibiting mitotic instability. Forfragile X the likelihood of instabilitywas as follows: below 50 repeats, norisk; around 60 repeats, low risk; 70 to86 repeats, high risk; above 86 repeats,absolute likelihood (36).The most common repeat length muta-

tions involve only relatively small changes.In vitro studies have indicated that strandslippage during DNA replication constitutesthe major cause of these length mutations(37). Furthermore, there is a connectionbetween replication slippage and DNArepair, as is indicated by mutations in DNArepair, giving rise to increased instab ilty. InSaccharomyces, mutations in mismatchrepair genes increased replication slippage100 to 700 times in poly-(GT) repeats (38).Infrequently, huge length increases occur,resulting in the extreme instability of thelong alleles group above. This suddenincrease in trinucleotide repeat length,which causes several human diseases, mustinvolve some mechanism other than replica-tion slippage. In cell culture 1000-foldamplification has been observed for thedihydrofolate reductase gene (39). Thisdrastic amplification involves an episomalmechanism. The gene is excised and copied,presumably by a rolling circle process, andreintegrated into nonhomologous chromo-somal sites. This mechanism, however, isnot the likely one to explain the amplifica-tion of trinucleotide repeats. Unlike the casewith the episomal mechanism, the amplifi-cation of trinucleotides never involves sur-rounding DNA and it always occurs in situ.A model to explain the expansion of trinu-cleotides (32) takes into consideration thedifficulty in replicating CG-rich sequencesby polymerases (36). It is possible thatreplication of these repeats gives rise to pre-mature termination and reinitiation events,generating multiple incomplete strands.Extensive increases in length can then beinduced by a strand switching between theincomplete strands. This model predictsthat an increased rate and an increasedlength of the expansion will occur withincreasing initial length of the trinucleotidesequence. These predictions have beenexperimentally observed (36).

Repeat-induced Point Mutations. Theaccumulation of repeated sequences, par-ticularly of microsatellites, implies a riskfor homologous recombination betweendispersed repeats, causing translocation andother chromosomal aberrations. Fungilike Neurospora (above) are less toleranttowards such repeated sequences than

Environmental Health Perspectives * Vol 105, Supplement 4 * June 1997784

MINI- AND MICROSATELLITES

higher multicellular organisms, and theyhave developed defense mechanismsagainst homologous repeated deletions byRIP (33). RIP recognizes duplicatedsequences and induces G.C to A.T muta-tion. This mutation is associated withmethylation of cytosin and a high fre-quency of recombination between tandemrepeats. Both alleles are mutated in thisprocess, and the genetic mechanism seemsto be comparable to the recognition ofhomologous sequences by recombinationprocesses. At the molecular level, the highfrequency of G.C to A.T transitions isprobably caused by enzymatic deaminationof cytosin or 5-methylcytosine. The nor-mal repair of these lesions may be turnedoff in ascogenous tissue or overwhelmed byRIP. This mutational process may be anintegral part of "genome cleaning" duringthe period between fertilization and karyo-gamy in fungi, which also includes a highfrequency of intrachromosomal recombina-tion, deleting tandemly repeated genes. Thesequence divergence by RIP can be suffi-cient to prevent recognition of homologyand subsequent recombination between dis-persed DNA regions. This form of geneticinstability potentially stabilizes the grossorganization of the genome (33).

Another form of RIPping has beendescribed by Rand (40) in the mitochondr-ial DNA in crickets (Gryllus). A repeatedsequence of 220 bp in length was found tobe a hot spot for point mutations, deletions,and insertions. The mutational changeswere localized in and around a 14 bp G.C-rich sequence. This mutation process appar-ently involves a mechanism other than theRIPing in Neurospora, as neither methyla-tion nor a bias towards G.C to A.T muta-tions was observed in the cricket mtDNA.As it is not clear if these mutations areinduced by the repeats or associated withthe repeats, this process has been namedrepeat associated point mutation (RAP).

The potential genetic and biologicaldisadvantages of repeated sequences, suchas microsatellites, can be expected to be ofgeneral relevance, although the thresholdfor negative effects presumably is higher inhigher organisms. Kricker et al. (41) have

Table 2. Mismatch repair genes.a

E coli S. cerevisiae Homo % HWPCCMutS MSH2 hMSHW < 50MutL MLH1 hMLH1 <30MutL PMS1 hPMS1 Few casesMutL hPMS2 Few cases

aData from Fishel and Kolodner (46).

pointed out that vertebrate chromosomeswould be threatened by illegitimate recom-bination between repeated sequences, suchas mobile elements and pseudogenes. Tocounteract this "genetic time bomb," astrategy based on methylation and associ-ated mutations through methylation anddeamination of 5-methylcytosine in CpGhas been developed.

Instability andMismatch Repair. Thestability of microsatellites is dependent onan intact mismatch DNA repair. The lossof this repair function in Saccharomycesincreased the instability of microsatellitesdrastically (above). The data on yeast indi-cated that the strong effect on the stabilityof poly (GT) recorded depended on errorsin the excision of mismatch bases afterDNA slippage, most ofwhich are correctedby mismatch repair in wild-type cells (38).The discovery of a similar case with coloncancer has attracted much attention.Fifteen percent of colorectal cancers have ahereditary background, hereditary non-polyposis colon cancer (HNPCC). Onegene involved in this cancer was localizedto chromosome 2 and linked to this locuswas a microsatellite with an array of ACrepeats. In the tumors ofHNPCC patients,mutations in this gene caused an extensiveinstability, not only of the AC repeatslinked to the gene, but also of microsatel-lites elsewhere in the genome, which weresubjected to thousands of changes (42,43).The gene in chromosome 2, responsible forthe genetic instability of HNPCC tumors,was homologous to the mismatch repairgene MutS in E. coli and MSH2 in yeast(44,45). Subsequently, three more humangenes, homologous with the mismatchrepair genes in E. coli and yeast, have beenlinked to HNPCC (46) (Table 2). Parsonset al. (47) recently reported a subset ofHNPCC patients with a high frequency ofmicrosatellite mutations not only in theirtumors but also in nonneoplastic cells.These patients furthermore had very fewtumors, showing that deficient mismatchrepair and succeeding mutations can be

Table 3. Diseases associated with trinucleotide reiteratior

compatible with normal development andnot sufficient for tumor development. Onthe other hand, instability of microsatellitesis also generated by mechanisms other thandeficient mismatch repair genes-for exam-ple, deficiency in exonuclease (48). Severalcancer forms have been found to be associ-ated with microsatellite instability; theseinclude gastric, pancreatic endometrial,Barrett's esophageal, and lung cancer (49).

Association ofMicrosatellites withHuman Diseases. The previous sectiondealt with microsatellite instability in con-nection with DNA repair deficiency and theassociation of this instability with cancer.This association between microsatelliteinstability and the disease is not a causalone, but presumably a common result of thelack of mismatch repair of DNA. Howevermicrosatellites have attracted a great deal ofattention in recent years because of a directconnection between expanded arrays ofCG-rich trinucleotides and several humanneurological diseases. Table 3 shows fivediseases of trinucleotide reiteration that havebeen characterized (50).

The microsatellite sequence in thesediseases are linked to a coding gene,which is affected by the expansion of thetrinucleotide sequence. These cases ofmicrosatellite-dependent disease representtwo classes. Fragile X and myotonic dystro-phy have their trinucleotide sequencelinked to the noncoding ends of the gene,while in the three diseases with CAGexpansion, coding for polyglutamine, themicrosatellite is located within the codingpart of the gene. An initial increase of thetrinucleotide sequence functions as a pre-mutational event. Above a critical numberof repeats, the system becomes unstable andusually more sequences are added, eventu-ally resulting in symptoms. Another charac-teristic of these diseases is the fact that thesymptoms tend to be more severe in subse-quent generations because of amplificationduring gametogenesis or in the zygote, aprocess named genetic anticipation. Thisanticipation is sex linked and inter alia

Reiterated Normal range Disease rangeDisease range trinucleotide of reiteration of reiteration

Spinal and bulbar muscular CAG 11-33 40-62atrophy, Kennedy's disease

Huntington's disease CAG 11-34 42-100Spinocerebellar ataxia type 1 CAG <29-36 43->60FragileX CGG 6-54 250-4000Myotonic dystrophy CTG 5-30 > 50

aData from Green (50).

Environmental Health Perspectives * Vol 105, Supplement 4 * June 1997 785

C. RAMEL

occurs through the mother in fragile X andmyotonic dystrophy but through the fatherin Huntington's disease. This process ofanticipation is connected with methylationof cytosine and genetic imprinting.

Fragile X was the first recognized caseof a genetic disease with an instability oftrinucleotide repeats. It is a common neu-rological disease that causes mental retarda-tion, which is inherited as an X-linkeddominant trait. It is manifested by chro-mosome breakages at specific sites. Thedisease is associated with an expansion of amicrosatellite repeat ofCGG trinucleotidesin the 5' untranslated end of the geneFMRJ. This leads to a hypermethylation ofthe promoter region and a down regulationof the gene expression. The mechanism ofinactivation of neighboring loci by non-coding repeats has a counterpart in the het-erochromatization of euchromatin bytandem repeats as observed in Arabidopsisand in Drosophila (51). Although the insta-bility of the trinocleotide repeats dependson the length of the sequence, the length atwhich an instability of the microsatellitebegins to occur varies between 35 and 55repeats. This long "gray zone" was shownby Eichler et al. (52) to depend on theinterspersion by AGG trinucleotides. Mostalleles of CGG repeats contain two AGGsand the instability depends on the numberof uninterrupted CGG repeats. The unin-terrupted length under which the mini-satellite is stable turned out to be 34 to 37CGGs, which is in agreement with the cor-responding number in other triplet repeatdiseases including myotonic dystrophy,Kennedy's disease, Huntington's disease,spinocerebellar ataxia, and dentatorubralpallidoluyisian atrophy. The loss ofAGGcauses an increased uninterrupted length ofCGG sequences and is therefore probablyan important mutational event for the pre-disposition to fragile X. The mechanism ofexpansion of the trinucleotide sequencepresumably rests on slippage during DNAreplication. To explain the rapid expansionof a large sequence of triplets, Eichler et al.(52) argue in favor of a slippage mecha-nism dependent on the lagging and leadingstrand. Based on the observed polarity ofexpansions at the 3' end, they propose aslippage process involving a whole Okasakifragment, spanning 150 to 200 bp withintrinucleotide repeat alleles of about 70CGGs (210 bp). Concerning the non-mendelian increase of the effect of mutatedgenes from one generation to the next, this"anticipation" has been reported as apostzygotic process in fragile X syndrome,

while premutational increase takes placeduring meiosis (51).

Myotonic dystrophy, another neurologi-cal disease, depends on an expansion of asequence ofCTG at the 3' untranslated endof the gene myotonic dystrophy proteinkinase (MDPK). The expanded trinu-deotide sequence eliminates transcription ofMDPK Above a threshold of about 146 bpno mRNA for MDPK could be observed(53). An increased nucleosome binding ofsuch expanding repeats, leading to a tran-scriptional repression, has been proposed asa mechanism (53). It has further beenshown that the increased nucleosome bind-ing exerts an effect on the post-transcrip-tional processing of the transcript fromexpanded alleles but not on the initiation ofthe transcription (54). The threshold forthe symptoms of myotonic dystrophy of146 bp corresponds with the DNA lengthfor a nucleosome (53).

Huntington's disease is an autosomalneurodegenerative disorder that dependson an expansion of CAG tandem repeats,giving rise to polyglutamin. This micro-satellite is located within the coding regionof the Huntington's disease gene, HD orIT15, which codes for the protein hunt-ingtin. Although the disease is dependenton the stretch ofCAG repeats, the instabil-ity giving rise to an expansion of the polyg-lutamine array seems to be influenced byanother trinucleotide repeat of CCGdownstream of the CAG repeat, coding forproline (55). Huntington's disease usuallyhas a late onset, but with increasing expan-sion of the trinucleotide repeats, due to"anticipation" through male gametes, thesymptoms become more severe and onsetearlier in subsequent generations. Thefunction of huntingtin is not known andits expression is similar in patients and con-trols. The length expansion makes the pro-tein not merely useless, but activelyharmful by a gain of function. Zeitlin et al.(56) showed that huntingtin is indispens-able, as null mutation of the huntingtingene in mice caused death of the embryo.The data on mice further suggested thathuntingtin is involved in counterbalancingprogrammed cell death, apoptosis (56).Reports by Li et al. (57) indicate that thepathological effects by the expansion of theCAG repeats depend on interactionbetween huntingtin and other cellular pro-teins. They identified a protein, hunt-ingtin-associated protein, HAP-1, thatbinds to huntingtin.This binding isenhanced by an expanded polyglutamine.The HAP-1 protein is enriched in the

brain, which may explain the localizedeffect of the disease to brain tissue.

Spinal and bulbar muscular atrophy,Kennedy's disease, is an X-linked diseaseand the only polyglutamine-dependentneurological disorder for which the func-tion of the protein involved is known. Itconstitutes the androgen receptor (AR),which is a ligand-activated transcriptionfactor. The AR contains, in the codingregion, the polyglutamine tract by CAGrepeats. As in the other microsatellite-dependent diseases, the severity of the dis-ease is correlated with the expansion of themicrosatellite. Chamberlain et al. (58)showed that progressive expansion of thepolyglutamine tract in human AR caused alinear decrease in the binding of the recep-tor to androgen and a decrease in activat-ing transcription of AR-responsive genes.However, the data indicated that there wasa threshold, as the expansion of the trinu-cleotides did not completely eliminate ARactivity, and that the residual activitywas sufficient to develop male primary andsecondary sex characteristics.

Inactivation Mechanisms by Poly-glutamine. Many data on the relationshipbetween expansion of trinucleotides andneurological disorders are now available,but the mechanistic cause of the effects ofthe expanded repeats at a molecular level isnot clear. An attractive possibility is thatlong trinucleotide repeats confer structuralchanges of DNA, and that these changesconstitute the ultimate reason for thepathological behavior. Yano-Yanagisawa etal. (59) found in the mouse brain two trin-ucleotide repeat-binding proteins-TRIP-1and TRIP-2-which bind specifically torepeats of AGC, AGT, GGC, and GGT,but no other trinucleotides. The AGC-repeat binding activity is of interest con-cerning polyglutamine. (CAG)-repeats werefound to contain clusters of non-B DNAstructural units, formed by each AGC trin-ucleotide repeating unit: 5'......(C AG)(CAG)(C AG)(C AG)........3'. In non-B DNA,cytosines are specifically base unpaired. Theproperty of trinucleotides to adopt anunusual DNA structure may contribute totheir abnormal behavior. Recently Gacy etal. (60) presented data suggesting that hair-pin formation in microsatellite repeatedsequences may provide a common explana-tion for several characteristics of simplenucleotide repeat expansion and pathologi-cal effects. Hairpin formation and stabilityare correlated with the length of the repeatsequence. They would, for example, explainthe stabilizing effect ofAGG punctuation

Environmental Health Perspectives * Vol 105, Supplement 4 * June 1997786

MINI- AND MICROSATELLITES

on FMRI in Fragile X (above) on the basisof its interruption of hairpin stability. Therepeats that form hairpin structures woulddisrupt normal DNA replication. Above acritical threshold length, stable hairpinstructures are formed, leading to replicationerrors and further expansion.

There are other neurological diseaseswith characteristics that resemble the onesestablished as dependent on polyglutamineexpansion, such as the phenomenon ofanticipation. It is therefore likely that moresuch neurological disorders will be added tothis class of polyglutamine degenerative dis-eases. The discovery of proteins that inter-act with polyglutamine stretches with anintensity dependent on the length of CAGrepeats opens new possibilities for identify-ing other disorders of this type. Trottier etal. (61) have characterized a monoclonalantibody that selectively recognizes polyglu-tamine expansions in Huntington's disease,spinocerebellar ataxia SCAI and Machado-Joseph disease SCA3-all known gluta-mine-repeat disorders. An expansion ofpolyglutamine was detected in this waywith spinocerebellar ataxia SCA2 and thedominant cerebellar ataxia with retinaldegeneration. There are indications thatschizophrenia and bipolar disorders maybelong to the same group of diseases.O'Donovan et al. (62) found that schizo-phrenia patients and patients suffering frombipolar disorders had expanded trinu-cleotide sequences of CAG and its comple-ment CTG as compared to controls. Theconnection between expanded trinucleotideCAG repeats and degenerative disorders mayopen the possibility in the future to designtherapeutic molecules that would interferewith abnormal polyglutamine stretches.

TelomeresThe ends of the chromosomes in mostorganisms consist of a tandem array of

simple DNA sequences, constituting thetelomeres [Zakian (63) presents a recentreview]. This arrangement of the chromo-some ends is dictated by the fact that theDNA polymerase cannot reproduce bothDNA strands to the ends without losingthe tip sequence of one of the strands.Therefore the tip of the chromosomes isorganized with noncoding repeatedsequences, which can be lost without losingcoding DNA [Ligner et al. (64) present arecent overview]. The telomeres can bereplaced by a protein-RNA enzyme, telom-erase. Most telomeric repeat sequences areshort, usually 5 to 8 bp, in mammalsTTAGGG. Drosophila has an exceptionaltelomere structure without the short con-servative repeats of the telomeres of mostother organisms (65). Instead, Drosophilahas one or more elements like long inter-spersed elements (LINE) mobile elements,and the replacement of the telomeresoccurs by transposition of the telomeresequence to the chromosome ends.Proximal to the LINE sequences inDrosophila there is a sequence of tandemrepeats, which probably are analogous tothe subterminal middle repetitive regions,telomere-associated (TA) DNA, in othereukaryotes. The array ofTA can expand orcontract by means of a recombinationmechanism. The composition of telomere-repeated sequences varies considerablyamong organisms, and evidently the telom-ere function does not require a specificDNA sequence. The DNA strand, whichruns from 5' to 3' towards the end, has reg-ularly more G residues, arranged in clus-ters, than the other strand. At least inciliated protozoans, such as Tetrahymenaand Oxytrichia, and in yeast the G strand isextended to form a single-strand G tail.The G strand can form non-Watson-Crickbase pairing structures, such as four-stranded helices and multiple G-G base

pairs. It is possible that this property isessential for the bouquet stage, formed bythe telomeres during meiosis.

The function of the telomeres is toprotect the ends of the chromosomes, notonly from losing genetic material at eachcell division, but also to prevent the endsfrom fusing with each other. As was shownby McClintock (66), broken chromosomeends fuse with each other, forming dicen-tric bridges and a breakage-fusion-bridgecycle. Because of the loss of DNA at thechromosome ends the telomeres have to bereplicated in another way than the rest ofthe chromosomes. This replication isacquired by telomerase, which has a uniquecomposition of protein and an RNA com-ponent. The protein part consists of twosubunits in Tetrahymena (67). The replica-tion of the telomere occurs by means ofreverse transcriptase from the RNA com-ponent. In Drosophila this replication isperformed through transposition of thetelomere sequence (above). In humansmost somatic tissues lose their telomeraseactivity and consequently the chromosomeends will shorten at each cell division, lead-ing to the eventual death of the cell. Thishas led to the hypothesis that the telomerelength functions as a biological clock,resulting in a programmed cell death aftera certain number of cell divisions (68). Inactual measurement of the telomeraseactivity in normal and immortal cancercells telomerase activity was invariablyrepressed in normal somatic cells but wasreactivated in various cancer cells (69).The important observation that theimmortality of malignant cells is associatedwith telomerase activity has led to specula-tion that the telomere might constitute atarget for cancer therapy. Similar specu-lations can also be applied concerningprevention of aging by reactivation oftelomerase activity.

REFERENCES

1. Orgal LE, Crick FHC. Selfish DNA: the ultimate parasite.Nature 284:604-607 (1980).

2. Schimke RT. Gene amplification, drug resistance and cancer.Cancer Res 44:1735-1742 (1984).

3. Jeffreys AJ, Wilson V, Thein SL. Hypervariable minisatelliteregion in human DNA. Nature 314:67-73 (1985).

4. Nakamura Y, Leppert M, O'Connell P, Wolff R, Holm T,Culver M, Martin C, Fujimoto E, Hoff M, Kumlin E et al.Variable number of tandem repeat (VNTR) markers for humangene mapping. Science 235:1616-1622 (1987).

5. Braman J, Barker D, Schumm J, Knowlton R, Donis-Keller H.Characterization of very highly polymorphic RFLP probes.Cytogenet Cell Genet 40:589 (1985).

6. Jeffreys AJ. 23rd Colworth Medal Lecture. Highly variableminisatellites and DNA fingerprint. Biochem Soc Trans15:309-317 (1987).

7. Jeffreys AJ, Royle NJ, Wilson V, Wong Z. Spontaneous muta-tion rates to new length alleles at tandem repetitive hypervari-able loci in human DNA. Nature 332:278-281 (1988).

8. Jeffreys AJ, Wilson V, Thein SL. Individual-specific 'finger-prints' of human DNA. Nature 316:76-79 (1985).

9. Jeffreys AJ, MacLeod A, Tamaki K, Neil DL, Monckton DG.Minisatellite repeat coding as a digital approach to DNA typ-ing. Nature 354:204-209 (1991).

10. Armour JAL, Jeffreys AJ. Biology and application of humanminisatellite loci. Curr Opin Genet Dev 2:850-856 (1992).

Environmental Health Perspectives * Vol 105, Supplement 4 * June 1997 787

C. RAMEL

11. Hulten M. Chismata distribution in the normal human male.Hereditas 76:55-78 (1974).

12. Laurie DA, Hulten M. Further studies on chiasmata distribu-tion and interference in the human male. Ann Hum Genet49:203-214 (1985).

13. Cooke HJ, Brown WR, Rappold GA. Hypervariable telomericsequences from the human sex chromosomes are pseudoautoso-mal. Nature 317: 687-692 (1985)

14. Julier C, DeGouyon B, Georges M, Guenet J-L, Nakamura Y,Avner P, Lathrop M. Minisatellite linkage maps in the mouseby cross-hybridization with human probes containing tandemrepeats. Proc Natl Acad Sci USA 87:4585-4589 (1990).

15. Jeifreys AJ, Tamaki K, MacLeod A, Monckton DG, Neil DL,Armour JAL. Complex gene conversion events in germlinemutation at human minisatellites. Nat Genet 6:136-145(1994).

16. Gray IC, Jeffreys AJ. Evolutionary transience of hypervariableminisatellites in man and the primates. Proc R Soc London BBiol Sci 243:241-253 (1991).

17. Cederberg H, Agurell E, Hedenskog M, Rannug U.Amplification and loss of repeat units of the human minisatel-lite MS 1 integrated into chromosome III of haploid yeaststrain. Mol Gen Genet 238:38-42 (1993).

18. Armour JAL, Monckton DG, Neil DL, Tamaki K, MacLeodA, Allen M, Crosier M, Jeffreys AJ. Mechanisms of mutation athuman minisatellite loci. In: Genome Rearrangement andStability (Davies KE, Warren ST, eds). Genome Analysis7:43-57 (1993).

19. Dubrova YE, Jeffreys AJ, Malashenko AM. Mouse minisatellitemutations induced by ionizing radiation. Nat Genet 5:92-94(1993).

20. Lewontin, RC, Hartl DL. Population genetics in forensic DNAtyping. Science 254:1745-1750 (1991).

21. Bruford MW, Wayne RK. Microsatellites and their applicationto population genetic studies. Curr Opin Genet Dev3:939-943 (1993).

22. Burke T, Bruford MW. DNA fingerprinting in birds. Nature327:149-152 (1987).

23. Capon DJ, Chen EY, Levinson AD, Seeburg PH, GoedderDV. Complete nucleotide sequence of the T24fhuman bladdercarcinoma oncogene and its normal homologue. Nature302:33-37 (1983).

24. Kasperczyk A, DiMartino NA, Krontiris TG. Minisatelliteallefe diversification: the origin of rare alleles of the HRASJlocus. Am J Hum Genet 47:854-859 (1990).

25. Krontiris TG. Minisatellites and human disease.Science269:1682-1683 (1995).

26. Krontiris TG, Derlin B, Karp DD, Robert NJ, Risch N. Anassociation between the risk of cancer and mutations in theHRASJ minisatellite locus. N Engl J Med 329:517-523(1993).

27. Trepicchio WL, Krontiris TG. Members of the rel/NF-kB fam-ily of transcriptional regulatory factors bind the HRASI min-isatellite DNA sequence. Nucleic Acids Res 20:247-2434(1992).

28. Green M, Krontiris TG. Allelic variation of reporter gene acti-vation by the HRASI minisatellite. Genomics 17:429-434(1993).

29. Bell GI, Selby M, Rutter WJ. The highly polymorphic regionnear the human insulin gene is composed of simple tandemlyrepeated sequences. Nature 295:31-35 (1982).

30. Kennedy GC, German MS, Rutter WJ. The minisatellite in thediabetes susceptibility locus IDDM2 regulates insulin transcrip-tion. Nat Genet 9:293-258 (1995).

31. Trepicchio WL, Krontiris TG. IGH minisatellite suppressionof USF-binding-site and Ep-mediated transcriptional activationof the adenovirus major late promoter. Nucleic Acids Res21:977-985 (1993).

32. Kuhl DPA, Caskey CT. Trinucleotide repeats and genomevariation. Curr Opin Genet Dev 3:404-407 (1993).

33. Selker EU. Premeiotic instability of repeated sequences inNeurospora crassa. Annu Rev Genet 24:579-613 (1990).

34. Wright JM. Are minisatellites the evolutionary progeny ofmicrosatellites? Genome 37:34335-34346 (1994).

35. Armour JAL, Neumann R, Gobert S, Jeffreys AJ. Isolation ofhuman simple repeat loci by hybridization selection. Hum MolGenet 3:599-605 (1994).

36. Fu Y-H, Kuhl DP, Pizzuti A, Pieretti M, Sutcliffe JS, RichardsS, Verkerk AJ, Holden JJ, Fenwick RG Jr, Warren ST et al.Variation of the CGG repeat at the fragile X site resulting ingenetic instability: resolution of the Sherman paradox. Cell67:1047-1058 (1991).

37. Sclotterer C, Tautz D. Slippage synthesis of simple sequenceDNA. Nucleic Acids Res 20:211-215 (1992).

38. Strand M, Prolla TA, Liskay RM, Petes TD. Destabilization oftracts of simple repetitive DNA in yeast by mutations affectingDNA mismatch repair. Nature 365:274-276 (1993).

39. Trask BJ, Hamlin JL. Early dihydrofolate reductase gene ampli-fication events in CHO cells usually occur on the same chro-mosome arm as the original locus. Genes Dev 3:1913 (1989).

40. Rand DM. RIPping and RAPping at Berkeley. Genetics132:1223-1224 (1992).

41. Kricker MC, Drake JW, Radman M. Duplication-targetedDNA methylation and mutagenesis in the evolution of eukary-otic chromosomes. Proc Natl Acad Sci USA 89:1075-1079(1992).

42. Peltomiki P, Aaltonen LA, Sistonen P, Pylkkanen L, MecklinJP, Jarvinen H, Green JS, Jass JR, Weber JL, Leach FS et al.Genetic mapping of a locus predisposing to human colorectalcancer. Science 260:810-812 (1993).

43. Aaltonen LA, Peltomaki P, Leach FS, Leach FS, Sistonen P,Pylkkanen L, Mecklin JP, Jarvinen H, Powell SM, Jen J et al.Clues to the pathogenesis of familial colorectal cancer. Science260:812-816 (1993).

44. Fishel R, Lescoe MK, Rao MR, Copeland NG, Jenkins NA,Garber J, Kane M, Kolodner R. The human mutator genehomolog MSH2 and its association with hereditary nonpolypo-sis colon cancer. Cell 75:1027-1038 (1993).

45. Leach FS, Nicolaides NC, Papadopoulos N, Liu B, Jen J,Parsons R, Peltomaki P, Sistonen P, Aaltonen LA, Lystrom-Lahti M et al. Mutations of a MutS homolog in hereditay non-polyposis colorectal cancer. Cell 75:1215-1225 (1993).

46. Fishel R, Kolodner RD. Identification of mismatch repair genesand their role in the development of cancer. Curr Opin GenetDev 5:382-395 (1995).

47. Parsons R, Li G-M, Longley M, Modrich P, Liu B, Berk T,Hamilton SR, Kinzler KW, Vogelstein B. Mismatch repairdeficiency in phenotypically normal human cells. Science268:738-740 (1995).

48. Johnson RE, Kowali GK, Prakash L, Prakash S. Requirementof the yeast RTH 1 5' to 3' exonuclease for the stability of sim-ple repetitive DNA. Science 269:238-240 (1995).

49. Meltzer SJ, Yin J, Manin B, Rhyu MG, Cottrell J, Hudson E,Redd JL, Krasna Mj, Abraham JM, Reid BJ. Microsatelliteinstability occurs frequently and in both diploid and aneuploidcell populations of Barrett's- associated esophageal adenocarci-noma. Cancer Res 54:3379-3382 (1994).

50. Green H. Human genetic diseases due to codon reiteration:relationship to an evolutionary mechanism. Cell 74:955-956(1993).

51. Tautz D, Schl6tterer C. Simple sequences. Curr Opin GenetDev 4:832-837 (1994).

52. Eichler EE, Holden JJ, Popovich BW, Reiss AL, Snow K,Thibodeau SN, Richards CS, Ward PA, Nelson DL. Length ofuninterrupted CGG repeats determines instability in theFMRJ gene. Nat Genet 8:88-94 (1994).

53. Wang Y-H, Amirhaeri S, Kang S, Wells RD, Griffith JD.Preferential nucleosome assembly at DNA triplet repeats frommyotonic dystrophy gene. Science 265:669-671 (1994).

54. Krahe R, Ashizawa T, Abbruzzese C, Roeder E, Carango P,Giacanelli M, Funanage VL, Siciliano MJ. Effect of myotonicdystrophy trinucleotide repeat expansion on DMPK transcrip-tion and processing. Genomics 28:1-14 (1995).

55. Andrew SE, Goldberg YP, Theilmann J, Zeisler JR, Hayden

788 Environmental Health Perspectives * Vol 105, Supplement 4 * June 1997

MINI- AND MICROSATEWTES

MR. ACCG repeat polymorphism adjacent to the CAG repeatin the Huntington disease gene: implication for diagnostic accu-racy and predictive testing. Hum Mol Genet 3:65-67 (1994).

56. Zeitlin S, Liu J-P, Chapman DL, Papaioannou VE,Efstradiadis A. Increased apoptosis and early embryonic lethal-ity in mice nullizygous or the Huntington's disease genehomologue. Nat Genet 11:155-163 (1995).

57. Li X-J, Li S-H, Sharp AH, Nucifora FC Jr, Schilling G,Lanahan A, Worley P, Snyder SH, Ross CA. A Huntington-associated protein enriched in brain with implication forpathology. Nature 378:398-402 (1995).

58. Chamberlain NL, Driver ED, Mjesfeld RL. The length andlocation of CAG trinucleotide repeat in the androgen receptorN-terminal domain affect transactivation function. NucleicAcids Res 22:3181-3186 (1994).

59. Yano-Yanagisawa H, Li Y, Wang H, Kohwi Y. Single-strandedDNA binding proteins isolatedc from mouse brain recognizespecific trinucleotide repeat sequences in vitro. Nucleic AcidsRes 23:2654-2660 (1995).

60. Gacy AM, Goellner G, Juranic N, Macura S, McMurray CT.Trinucleotide repeats that expand in human disease form hair-pin structures in vitro. Cell 81 533-540 (1995).

61. Trottier Y, Lutz Y, Stevanin G, Imbert G, Devys D, Cancel G,Saudou F, Weber C, David G, Tora L et al. Polyglutamine

expansion as a pathological epitope in Huntington's disease andfour dominant cerebellar ataxias. Nature 378:403-406 (1995).

62. O'Donovan MC, Guy C, Craddock N, Murphy KC, CardnoAG, Jones LA, Owen MJ, McGuffin P. Expanded CAG repeatsin schizophrenia and bipolar disorders. Nature Genet10:380-381(1995).

63. Zakian VA. Telomeres: beginning to understand the end.Science 270:1601-1607 (1995).

64. Lingner J, Cooper JP, Cech TR. Telomerase and DNA endreplication: no longer a lagging strand problem? Science269:1533-1534 (1995).

65. Mason JM, Biessmann H. The unusual telomeres ofDrosophila. Trends Genet 11:58-62 (1995).

66. McC[intock B. The stability of broken ends of chromosomes inZea mays. Genetics 26:234-282 (1941).

67. Collins K, Kobayashi R, Greider CW. Purification of tetrahy-mena telomerase and cloning of genes encoding the two proteincomponents of the enzyme. Cell 81:677-686 (1995).

68. Harley CB, Futcher, Greider CW. Telomeres shorten duringageing of human fibroblasts. Nature 345:458-460 (1990).

69. Kim NW, Piatyszek MA, Prowse KR, Harley CB, West MD,Ho PL, Coviello GM, Wright WE, Weinrich SL, Shay JW.Specific association of human telomerase activity with immor-tal cells and cancer. Science 266:2011-2015 (1994).

Environmental Health Perspectives * Vol 105, Supplement 4 * June 1997 789