Analysis of microsatellite instability and loss of heterozygosityin breast cancer with the use of a well characterized multi-plex system

Jolanta Powierska-Czarny1�, Danuta Miścicka-Śliwka2, Jakub Czarny2, TomaszGrzybowski2, Marcin Woźniak2, Gerard Drewa1, Włodzimierz Czechowicz3

and Jan Sir3

1Department of Human Biology and 2Forensic Medicine Institute, The Ludwik RydygierMedical University of Bydgoszcz, Bydgoszcz, Poland; 3Regional Center of Oncology,Bydgoszcz, Poland

Received: 10 December, 2003; revised: 09 July, 2003; accepted: 18 August, 2003

Key words: short tandem repeats, microsatellite instability, breast cancer, multiplex DNA typing

Analysis of microsatellite instability (MI) and loss of heterozygosity (LOH) is rec-ommended for screening patients with sporadic and hereditary malignancies. Thisstudy shows an application of a fluorescent hexaplex PCR system for microsatellitetyping on A.L.F. DNA Sequencer (Pharmacia Biotech). This technique detectschanges in microsatellites providing a time-efficient, reliable and accurate methodfor MI and LOH analyses. The Fragment Manager software was used for automatedsize calculation and quantitation of DNA fragments, enabling rapid and precise mea-surement of allelic ratios. We examined 70 breast cancer and 70 control DNA speci-mens, classified all the patterns of microsatellite alterations, and set up MI and LOHassessment criteria for the automated multiplex fluorescent method.

The detection of genomic instability is animportant step in molecular analysis of tu-morogenesis. Somatic alterations in micro-satellite sequences due to a deletion or an in-sertion of one or more repeat units have beentermed “microsatellite instabilities” (MI or

MIN). Microsatellite instability, referred toas replication error (RER) phenotype, is asso-ciated with defects in the DNA mismatch-re-pair machinery (Fishel et al., 1993; Bronner etal., 1994). In hereditary nonpolyposis colo-rectal carcinoma (HNPCC), this instability

Vol. 50 No. 4/2003

1195–1203

QUARTERLY

�Corresponding author: Jolanta Powierska-Czarny, The Ludwik Rydygier Medical University ofBydgoszcz, Powstańców Wielkopolskich 44/39, 85-090 Bydgoszcz, Poland; tel: (48 525) 853 658;e-mail: jolanta_powierska@o2.pl

Abbreviations: AMG, amelogenin; IDC, invasive ductal carcinoma; LOH, loss of heterozygosity; MI,microsatellite instability; STR, short tandem repeat.

was shown to be caused by a germline muta-tion in mismatch-repair genes (Aaltonen etal., 1993; Papadopoulos et al., 1994). MI hasbeen detected in a wide variety of humanneoplasmic tumors, both hereditary and spo-radic, and is used as a diagnostic criterion formismatch repair deficiency (Aaltonen et al.,1993; Papadopoulos et al., 1994; Wooster etal., 1994a; Yee et al., 1994).Analysis of the highly polymorphic micro-

satellite loci not only provides informationabout MI, but also allows the detection ofallelic deletion in tumor cells. Loss ofheterozygosity can only be assayed in hetero-zygous loci. Tumor suppressor genes are gen-erally inactivated by an intragenic mutationwithin one allele and the subsequent loss ofthe corresponding (wild) allele, resulting in acondition termed “loss of heterozygosity”(LOH) (Knudson, 1993). Somatic DNA alter-ations leading to LOH are generally morecomplex and may include multifocal deletion,mitotic recombination or nondisjunctionalchromosomal loss, as well as gene conversionand point mutation (Meuth, 1990). The fre-quent allelic loss at particular chromosomalregions in tumors indicates the presence of atumor suppressor gene.Loss of heterozygosity and microsatellite in-

stability are integral parts of neoplasmic pro-gression. Analysis of this abnormality is ofgreat significance for clinical diagnostics andis recommended in molecular screening forboth inherited and sporadic neoplasms (Jasset al., 1995; Aaltonen et al. 1998).This study reports a precise and efficient

method for detecting microsatellite instabil-ity and LOH with the use of an automated flu-orescent DNA sequencer. We used awell-characterized multiplex system enablingcoamplification and typing of five short tandemrepeats (STRs): HUMD1S103, HUMTH01,HUMD21S11, HUMD18S51, HUMFIBRA(FGA), and the amelogenin locus (Miścicka-Śliwka et al., 1997). This system has beenused in our laboratory in routine forensic andpaternity investigations. This study presents

its application in genetic analysis of 70 breastcancer specimens.

MATERIALS AND METHODS

DNA was extracted from blood and tumortissue specimens collected from patients whowere treated at the Regional Cancer Center inBydgoszcz during the years 1997–1998. All70 tumor samples were invasive ductal carci-nomas (IDC). Tumor tissues were surgicallyresected and whole blood samples were ob-tained from corresponding patients throughvenipuncture. Biological samples was col-lected according to Approval of IndependentEthics Committee of the Ludwik RydygierMedical University in Bydgoszcz. The tumortissue was then snap-frozen in liquid nitrogenand stored at –70°C. Genomic DNA was ex-tracted from tumor tissue and each corre-sponding whole blood sample using standardorganic protocol (Sambrook et al., 1998).DNA concentration was measured usingGeneQuant DNA Calculator (Pharmacia-Bio-tech Uppsala, Sweden).Amplification of a hexaplex consisting of five

STR loci (HUMD1S103, HUMTH01,HUMD21S11, HUMD18S51, and HUM-FIBRA) and the amelogenin locus, HUM-AMGXY, was performed according to previ-ously described procedure (Miścicka- Śliwka etal., 1997). The amplified fragments were sepa-rated in 6% denaturing polyacrylamide gels onA.L.F Automated Fluorescent DNA Sequencer(Pharmacia Biotech, Uppsala, Sweden). Exter-nal Size Standard 50–500 bp (PharmaciaBiotech) was used. An internal size standardwas the shorter product of the amelogenin lo-cus (106 bp). Fragment Manager software wasused to measure size and signal intensity ofthe fluorescent PCR products. This softwareallows for automated linear quantitation andcalculation of allelic ratios.Statistical analysis was performed using �

2

tests and Fisher exact test with the "SPSS"statistical package.

1196 Microsatellite instability and loss of heterozygosity in breast cancer 2003

RESULTS

A multiplex PCR system was used to iden-tify the expansion and/or deletion of fivemicrosatellite loci: D1S103, TH01, D21S11,D18S51, and FGA, in tumor specimens.These loci were amplified from genomic DNAthat had been extracted from tumor speci-mens and corresponding whole blood samplesobtained from each patient. One fluorescentdye was used for simultaneous amplificationof five STR sequences in a single reaction.The PCR products from the tumor and nor-mal control tissue were then separated in twodifferent lanes in the gel. We present a com-parative analysis of many genetic profilesfrom corresponding normal and neoplasmictissue samples and set up criteria for theassessment of MI and LOH.The electropherograms in Fig. 1 represent a

genetic profile at five microsatellite loci(D1S103, THO1, D21S11, D18S51, FGA) andthe AMG locus of normal lymphocyte DNAand tumor neoplasmic DNA collected fromthe same patient. Loss of heterozygosity wasidentified when either no peak or a very weakpeak (signal for one of the alleles reduced byat least 35%) was detected in the neoplasmicDNA as compared to the control lymphocyteDNA. Strength of the signal corresponding toan individual allele was defined by the soft-ware as peak height and peak area. To esti-mate the degree of LOH, peak areas of the al-leles in a heterozygote from normal (N) andtumor sample (T) were calculated. However,one must take into account measurement er-rors or slight differences in the efficiency ofPCR amplification. In this experiment, therelative difference in peak areas of normalheterozygotes was calculated to average 10%for all tested loci. For a few heterozygous STRloci, a higher imbalance of 25% was observed.We assumed LOH when �35% (25% + 10%)differences between peak areas were encoun-tered. All the results were verified by two in-dependent PCR reactions.

The electropherograms in Figs. 1 and 2show the relative amounts of the two corre-sponding alleles, indicating loss of hetero-zygosity. For example, in tumor samples themicrosatellites D21S11 in lane 2 (Fig. 1 A, B)and D1S103, D21S11 and FGA in lane 2 (Fig.2), had very weak peaks at positions corre-sponding to the peaks of the normal allele.Therefore, LOH could be assessed and quanti-tated very effectively using the FragmentManager software. Table D in Fig. 1, and E inFig. 2 demonstrate software quantitation andcalculation of allelic ratios at the analyzedheterozygous loci.Microsatellite instability was identified

when a novel allele in tumor DNA was ob-served as compared to the control lymphocyteDNA.To measure the DNA fragment size precisely,

fluorescently labeled external size markers(50–500 bp) were loaded into every fifth gellane. An internal size marker was the shorterPCR product of the amelogenin locus (106 bp),coamplified with the five microsatellite lociand coelectrophoresed. The Fragment Man-ager software used the internal size standardto normalize electrophoretic mobility differ-ences between lanes. In one-color analyses,cancer and normal DNA amplicons were la-beled with the same dye and separated in twodifferent lanes of the gel. In this system, theamplified DNA fragments were measured withan accuracy of one base pair. The case shownin lane 2 of Fig. 1 was classified as MI positivebecause a novel allele, which was 4 bp longerthan the longest allele in the normal DNA, wasidentified at the FGA locus in the cancer DNA.The results of fragment sizing, as well as peakheight and area measurements, are tabulatedin Fig. 1D. Additionally, in MI positive micro-satellites, the comparison of signal detectionbetween normal and cancer DNA revealed adecrease in signal strength for one allele incancer DNA. Thus, the presented system effi-ciently detected diverse types of microsatellitechanges in cancer specimens.

Vol. 50 Jolanta Powierska-Czarny and others 1197

1198 Microsatellite instability and loss of heterozygosity in breast cancer 2003

Figure 1. An example of LOH and MI identifiedin carcinoma ductale invasivum mammae.

A. An electropherogram of hexaplex PCR reaction.Top lines show profile for STR D1S103, THO1,D21S11, D18S51, and FGA amplicons in control nor-mal DNA (N, black). Bottom profile showsmicrosatellite instability in locus FGA detected in tu-mor DNA (T, red). Arrows indicate positions of allelealterations. B. Loss of heterozygosity in D21S11 locusis interpreted as reduced area of one allelic peak ascompared to corresponding normal DNA. C. MI posi-tive case in FGA locus, identified as additional peaks intumor profiles. D. Results of Fragment Manager analy-sis, yellow lines indicate peaks representing changedalleles.

Vol. 50 Jolanta Powierska-Czarny and others 1199

Figure 2. An electropherogram of hexaplex PCRfluorescent analysis with multilocus LOH.

A. Six microsatellite loci, D1S103, THO1, D21S11,D18S51, FGA, and AMG were simultaneously electro-phoresed in a single lane, N, normal (black); T, tumor(red). B. A single locus with LOH. Faint peaks in theposition of deleted alleles are interpreted as represent-ing PCR products from normal tissue present in the tu-mor sample. E. Results of Fragment Manager analysis,red arrows indicate peaks representing changed al-leles.

Differences between tumor and normalDNA were detected in 34 out of 70 invasiveductal carcinomas (IDC) (48%). In all, 78 al-terations were detected in tumors, includingmicrosatellite instability and loss ofheterozygosity. In 26 cancer samples,changes in more than one locus were ob-served. In 70 IDC tumor cases, 6 (8%) wereidentified with somatic microsatellite instabil-ity, and 31 tumors (44%) with LOH. The re-sults obtained are summarized in Table 1.The potential impact of MI and LOH on tu-

mor growth and progression of invasiveductal carcinoma was determined by simulta-neous assessment of various parameters, in-cluding: patient’s age, family history of breastcancer, condition of lymph nodes, and tumorgrade and size (unpublished data). A statisti-cally significant correlation was found be-tween the occurrence of genetic alterations inmicrosatellites and the family history ofbreast cancer, patient age, and the presenceof pathological lymph nodes (P < 0.05). Nocorrelation was found with tumor size andgrade (P > 0.05).

DISCUSSION

Loss of heterozygosity and microsatellite in-stability are integral parts of neoplastic pro-gression, observed in hereditary nonpoly-posis colorectal cancer (HNPCC) and othersporadic malignancies (Aaltonen et al., 1993;Papadopoulos et al., 1994; Wooster et al.,

1994a; Yee et al., 1994). A rapid, reliable, andeasy screening method is presented for thedetection of MI and LOH, based on a hexaplexPCR reaction routinely used in forensic andpaternity studies. The fluorescence-based pro-cedure does not require radioactive materialsor special facilities and can be easily per-formed in any clinical laboratory. Six lociwere amplified in one PCR reaction, using a40 lane sequencing gel (with an external sizemarker loaded in every fifth lane) so that amaximum of 18 patients could be examined inone assay.The Fragment Manager software was used

to calculate the size, and peak height and areafor each fluorescent product. This method en-ables detection of differences between neo-plastic and normal tissue, providing an effec-tive, sensitive, and accurate analysis of bothMI and LOH. Previous reports point out tohigher reliability of automatic fluorescent de-tection changes in STRs over the conven-tional radiolabeling methods (Toh et al., 1996;Wang et al., 1997).The software enabled automated linear

quantitation and calculation of allelic ratios,allowing rapid identification of LOH. Molecu-lar analyses of allelic deletion remain difficultto interpret, as samples usually contain sev-eral subpopulations of tumor cells. Frequentmixtures of tumor samples and normal cellscreate further difficulties for accurate inter-pretation (Bi�che & Lidereau, 1995). Thetechnique utilizing PCR reaction and electro-phoretic separation provides a highly sensi-

1200 Microsatellite instability and loss of heterozygosity in breast cancer 2003

Locus Chromosomal location MI LOH

HUMD1S103 lq32-44 3/70 (4%) 10/53 (18%)

HUMTH01 llpl5-15.5 0/70 (0%) 14/55 (25%)

HUMD21S11 21q21 0/70 (0%) 15/63 (24%)

HUMD18S51 18q21.33 1/70 (1%) 15/56 (27%)

HUMFIBRA 4q28-31 2/70 (2%) 17/62 (27%)

Table 1. Results of the analysis of microsatellite instability and loss of heterozygosity in breast can-cers

tive method for the detection of changes atmicrosatellite loci — a single neoplastic cellcan be identified among 500 normal ones, asdescribed by Sidransky (1997).It was observed that the detection of LOH

with the use of multiplex fluorescent PCR sys-tems was effective in tumor tissue sections.For the system used in this study, optimaltemplate DNA concentration is 5–10 ng per25 �l of PCR reaction volume (Miścicka-Śliwka et al., 1997). It was found that tem-plate concentration outside the optimal rangedisturbed the accuracy of peak measurementsfor each fluorescent product and thus poseddifficulties for the observation of micro-satellite changes.This highly discriminating multiplex system

is sensitive enough for both personal identifi-cation and cancer cell detection. Herein de-scribed is a method for single-color detectionon a single-wavelength automated DNA se-quencer, which is much simpler in use thanfour-dye sequencers. The single-color detec-tion system eliminates the problem of differ-ent detection sensitivities displayed by differ-ent dyes when using multi-wavelength se-quencers such as ABI 373, 377 or 310. This isvery important for the quantitation and calcu-lation of allelic ratios, and facilitatesmeasurement of allelic deletions in tumorDNA.This study analyzed the cases of MI and

LOH in microsatellites localized in five chro-mosomes. Alteration in the HUMFIBRA lo-cus, located at 4q28, occurs frequently in in-vasive ductal carcinomas, which may suggestthe importance of this chromosomal regionfor tumor growth (unpublished data). A highfrequency of LOH detected in loci D18S51and THO1 confirms the localization of theknown suppressor genes DCC (18q) andTSG101 (11p15) in the vicinity of these loci(Fearon et al., 1990; Li et al., 1997).Our recent study demonstrated also that ge-

netic alterations in microsatellites are corre-lated with patient age and family history ofbreast cancer, as well as with the presence of

pathological lymph nodes (unpublished data).Germinal mutations in the BRCA1 and/orBRCA2 gene are responsible for the majorityof hereditary breast cancer cases (Feunteun& Lenoir, 1996; Sobczak et al., 1997; Woosteret al., 1994). Recent studies on their biologicalfunctions suggest a role in monitoring and/orrepair of DNA lesions (Feunteun, 1998). Inac-tivation of these genes leads to genetic insta-bility and increased mutation rate in somegenes, such as the tumor suppressor genescontrolling progression of the cell cycle. Ourdata and other studies (Aldaz et al., 1995; DeMarchis et al., 1997; Sourvinos et al., 1997;Arzimanoglou et al., 1998) suggest thatchanges in microsatellites may predate tolater mammary tumorogenesis.The automated multiplex analysis of LOH

and MI presented here can be a usefulmethod for cancer screening (both inheritedand sporadic) and has a potential prognosticvalue. In summary, the fluorescent multiplextechnique of MI and LOH detection enablessimultaneous assessment of changes in multi-ple loci in human tumors and has a widerange of potential applications in manyclinical fields.

R E F E R E N C E S

Aaltonen LA, Peltomaki P, Leach FS, SistoneneP, Pylkkanen L, Mecklin JP, Jarvinen H,Powell SM, Jen J, Hamilton SR, PetersonGM, Kinzler KW, Vogelstein B, de laChapelle A. (1993) Clues to the pathogenesisof familial colorectal cancer. Science.; 260:812–6.

Aaltonen LA, Aslovaara MD, Kristo P, CanzianF, Hemminki A, Peltomaki P, Chadwick RB,Kaariainen H, Eskelinen MD, Jarvinen H,Mecklin JP, Chapelle A. (1998) Incidence ofhereditary nonpolyposis colorectal cancerand the feasibility of molecular screening forthe disease. N Engl J Med.; 338: 1481–7.

Aldaz CM, Chen T, Sahin A, Cunningham J,Bondy M. (1995) Comparative allelotype ofin situ and invasive human breast cancer:

Vol. 50 Jolanta Powierska-Czarny and others 1201

high frequency of microsatellite instability inlobular breast carcinomas. Cancer Res.; 55:3976–81.

Arzimanoglou II, Gilbert F, Barber HRK. (1998)Microsatellite instability in human solid tu-mors. Cancer.; 82: 1808–20.

Bi�che I, Lidereau R. (1995) Genetic alterationin breast cancer. Genes Chromosomes Can-cer.; 14: 227–51.

Bronner CE, Baker SM, Morrison PT, WarrenG, Smith LG, Lescoe MK, Kane M, EarabinoC, Lipford J, Lindblom A, Tannergard P,Bollag RJ, Godwin AR, Ward DC,Nodenskjold M, Fishel R, Kolodner R, LiskayRM. (1994) Mutation in the DNA mismatchrepair gene homologue hMLH1 is associatedwith hereditary nonpolyposis colon cancer.Nature.; 368: 258–61.

De Marchis L, Contegiacomo A, D’Amico C.(1997) Microsatellite instability is correlatedwith lymph node-positive breast cancer. ClinCancer Res.; 3: 241–8.

Fearon ER, Cho KR, Nigro JM, Kern SE,Simons JW, Ruppert JM, Hamilton SR,Preisinger AC, Thomas G, Kinzler KW.(1990) Identification of a chromosome 18qgene that is altered in colorectal cancer. Sci-ence.; 247: 49–56.

Feunteun J. (1998) Breast cancer and geneticinstability: the molecules behind the scenes.Mol Med Today.; 4: 263–7.

Feunteun J, Lenoir GM. (1996) BRCA1, a geneinvolved in inherited predisposition to breastand ovarian cancer. Biochim Biophys Acta.;1242: 177–80.

Fishel R, Lescoe MK, Rao MR, Copeland NG,Jenkins NA, Garber J, Kane M, Kolodner R.(1993) The human mutator gene homologMSH2 and association with hereditarynonpolyposis colon cancer. Cell.; 75:1027–38.

Jass JR, Cottier DS, Jeevaratnam P, Pokos V,Holdaway KM, Bowden ML, Van de WaterNS, Browett PJ. (1995) Diagnostic use ofmicrosatellite instability in hereditarynon-polyposis colorectal cancer. Lancet.; 346:1200–1.

Knudson AG. (1993) Antioncogenes and humancancer. Proc Nat Acad Sci U S A.; 90:10914–20.

Li L, Li X, Franke U, Cohen SN. (1997) TheTSG101 tumor susceptibility gene is locatedin chromosome 11 band p15 and mutated inhuman breast cancer. Cell.; 88: 143–54.

Miścicka-Śliwka D, Grzybowski T, Woźniak M.(1997) Optimization of a hexaplex PCR am-plification from Short Tandem Repeats(STR) and amelogenin (AMG) loci. Electro-phoresis.; 18: 1627–32.

Meuth M. (1990) The structure of mutation inmammalian cells. Biochim Biophys Acta.;1032: 1–17.

Papadopoulos N, Nicolaides NC, Wei YF, RubenSM, Carter KC, Rosen CA, Haseltine WA,Fleischmann RD, Fraser CM, Adams MD,Venter JC, Hamilton SR, Petersen GM, Wat-son P, Lynch HT, Peltomaki P, Mecklin JP,de la Chapelle A, Kinzler KW, Volgelstein B.(1994) Mutaton of a mutL homolog in hered-itary colon cancer. Science.; 263: 1625–9.

Risinger, JI, Barrett JC, Watson P. Lynch HT,Boyd J. (1996) Molecular genetic evidence ofthe occurrence of breast cancer as an inte-gral tumor in patients with the hereditarynonpolyposis colorectal carcinoma syndrome.Cancer.; 77: 1836–43.

Sambrook J, Fritsch EF, Maniatis T. (1989)Analysis and cloning of eucaryotic genomicDNA. In Molecular cloning, a labolatory man-ual. pp 16–22. Cold Spring Harbor Labora-tory Press.

Sidransky D. (1997) Nucleid acid-based methodfor the detection of cancer. Science.; 278:1054–8.

Sobczak K, Kozłowski P, Napierała M, Czarny J,Woźniak M, Kapuścińska K, Łośko M,Koziczak M, Jasińska A, Powierska J,Braczkowski R, Bręborowicz J, Godlewski D,Mackiewicz A, Krzyżosiak W. (1997) NovelBRCA1 mutations and more frequentintron-20 alteration found among 236women from Western Poland. Oncogene.; 15:1773–9.

1202 Microsatellite instability and loss of heterozygosity in breast cancer 2003

Sourvinos G, Kiaris H, Tsikkinis A, VassilarosS, Spandidos DA. (1997) Microsatellite insta-bility and loss of heterozygosity in primarybreast cancer. Tumor Biol.; 18: 157–66.

Toh Y, Oki E, Oda S, Tomoda M, Tomisaki S,Ichiyoshi Y, Ohno S, Sugimachi K. (1996) Anintegrated microsatellite length analysis us-ing automated fluorescent DNA sequencer.Cancer Res.; 56: 2688–91.

Wang Y, Hung S, Linn JF, Steiner G, GlazerAN, Sidransky D, Mathies RA. (1997)Microsatellite-based cancer detection usingcapillary array electrophoresis and en-ergy-transfer fluorescent primers. Electropho-resis.; 18: 1742–9.

Wooster R, Cleton-Jansen AM, Collins N,Mangion J, Cornelis RS, Cooper CS,Gusterson BA, Ponder BA, von Deimling A,

Wiestler OD. (1994) Instability of short tan-dem repeats in human cancer. Nat Genet.; 6:152–6.

Wooster R, Neuhausen SL, Mangion J, QuirkY, Ford D, Collins N, Nguyen K, Seal S,Tran T, Averill D, Fields P, Marshall G,Narod S, Lenoir GM, Lynch H, Feunteun J,Odevilee P, Cornelisse CJ, Menko FH, DalyPA, Ormiston W, McManus R, Pye C, LewisCM, Cannon-Albright LA, Petro J, PonderBAJ, Skolnik MH, Easton DF, Goldgar DE,Stratton MR. (1994) Localization of a breastcancer susceptibility gene, BRCA2, to chro-mosome 13q12-13. Science.; 265: 2088–90.

Yee CJ, Roodi N, Verrier CS, Parl FF. (1994)Microsatellite instability and loss ofheterozygosity in breast cancer. Cancer Res.;54: 1641–4.

Vol. 50 Jolanta Powierska-Czarny and others 1203