ORIGINAL ARTICLE

Assessment of microsatellite instability in colorectalcarcinoma at an Indian center

Vijay Pandey & Jyothi S. Prabhu & Payal K. &Veena Rajan & C. Deepak & Shubhada Barde &

P. Jagannath & Anita Borges & Sridhar T. S.

Accepted: 13 October 2006 / Published online: 12 December 2006# Springer-Verlag 2006

AbstractAim The purpose of the present study was to evaluate thefrequency of microsatellite instability (MSI) in colorectalcancers in an Indian cohort.Materials and methods Paraffin embedded tissue samplesof colorectal cancers from 46 patients were assessed formismatch repair protein expression (hMLH1 and hMSH2)by immunohistochemistry. Subsequently, MSI analysis wasdone after PCR amplification of five Bethesda markers.Results Amongst 46 cases studied, only 5 patients (10.8%)showed MSI. Out of these, two (4.3%) had high micro-satellite instability (MSI-H) and three (6.5%) showed lowmicrosatellite instability (MSI-L). Out of 46 cases, 41 weremicrosatellite stable (MSS). In the 46 cases tested byimmunohistochemistry, 7 (15.7%) showed the absence ofhMLH1 and 1 case showed the absence of hMSH2.Conclusion Our study indicates a similar rate of incidenceof MSI in colorectal cancers in the Indian cohort comparedto the West (10–15%) despite lower incidence of colorectalcancers and predominance of rectosigmoid tumors in theIndian population.

Keywords Microsatellite instability . Colorectal cancer .

hMLH1 . hMSH2 . Immunohistochemistry . PCR

AbbreviationsMSI microsatellite instabilityCRC colorectal cancergDNA genomic DNAMSS microsatellite stableMSI-L microsatellite instability lowMSI-H microsatellite instability highHNPCC hereditary non-polyposis colorectal cancerH & E hematoxylin and eosinIHC immunohistochemistryCAP College of American PathologistshMLH1 human mutL homolog 1hMSH2 human mutS homolog 2

Introduction

Two molecular pathways of neoplasia were described incolorectal cancers (CRC). The chromosomal instabilitypathway as described initially by Fearon and Vogelstein in1990 [1] emphasizes the role of tumor suppressor genes,DNA hypomethylation and loss of heterozygosity at keyloci. The second pathway known as the microsatelliteinstability pathway is characterized by errors in themismatch repair genes [2, 3].

Microsatellites are short tandem mononucleotide, dinu-cleotide, trinucleotide or tetranucleotide repeat sequencesdistributed throughout the genome. Expansion or contrac-tion of these sequences due to increase or decrease in therepeat units gives rise to what is referred to as microsatelliteinstability (MSI) [4, 5]. Tumors not showing this phenotypeare referred to as microsatellite stable (MSS) [6]. Numerousstudies have confirmed the better prognosis of stage IIIcolon cancer patients with MSI [7, 8]. Other studies showed

Int J Colorectal Dis (2007) 22:777–782DOI 10.1007/s00384-006-0241-3

V. Pandey (*) : J. S. Prabhu : P. K. :V. Rajan : S. T. S.Triesta Sciences, IPHCR Building,St. John’s Medical College Campus,Opposite Koramangala BDA Complex,Bangalore 560034, Indiae-mail: vijaypandey7@gmail.com

C. Deepak : S. Barde : P. Jagannath :A. BorgesAsian Institute of Oncology, S.L. Raheja Hospital,Raheja Hospital Road, Mahim,Mumbai 400016, India

that 5-FU is beneficial in stage II and III cancers with MSSor MSI-L phenotype [9].

Hereditary non-polyposis colorectal cancer syndrome(HNPCC), which accounts for 1–5% of all CRC ischaracterized by MSI [10]. Most of these patients (95%)have defects in the mismatch repair genes hMSH2 orhMLH1 [11]. Occasionally, defects in other human DNAmismatch repair genes (hPMS2 and hMSH6) were reported[12, 13]. It is interesting to note that even among sporadiccolorectal cancers, 10–15% of patients exhibit MSI [14, 15].

Colorectal cancer is uncommon in India compared to theWest with the highest reported regional incidence being3.7/100,000 among men and 3/100,000 among women.Unlike Western populations that tend to have more ofproximal tumors [16], the rectum is the commonest site oforigin amongst Indians [17]. The frequency of micro-satellite instability in colorectal cancers in Indians isunknown. In the present study, we evaluated the frequencyof MSI in colorectal cancers at a center in India.

Materials and methods

Tissue specimens

We studied 46 patients with histopathologically confirmeddiagnosis of CRC from a single center in India. The studywas approved by the institutional ethics committee.Informed consent was obtained from all patients. These46 patients had undergone tumor resection between the year2001 and 2005. Histological analysis was performed on H& E stained slides and classified according to the Dukes/Astler–Coller staging. Clinical information of age, sex, siteof the tumor, family history and histological type of thetumor was obtained from the clinical records. A tumorblock and a block of adjacent normal colonic epitheliumwere selected for analysis in each of the 46 patients. Thetumors were classified by TNM staging (according tothe CAP guidelines) using the information provided in thehistopathological reports.

Histopathological analysis

Two pathologists independently assessed each of thetumors for histological features suggestive of microsatelliteinstability such as signet ring cells, mucinous histology andpoor differentiation.

Immunohistochemistry

5 μm sections were affixed to poly-L-lysine (0.1% w/v,Sigma-Aldrich) coated slides and dried overnight at 56°C.After deparaffinization, the slides were dehydrated through

graded ethanol and finally with distilled water for rehydra-tion. Slides were incubated in 3% hydrogen peroxide inmethanol to block endogenous peroxidase activity. Sectionswere subjected to moist heat antigen retrieval by pressurecooking in EDTA buffer at pH 8.0 and then cooled to roomtemperature and washed in TBST buffer. The sections weretreated with anti-hMLH1 mouse monoclonal antibody(G168-15; Pharmingen, San Diego, USA, dilution 1:50)and anti-hMSH2 mouse monoclonal antibody (Oncogene,San Diego, USA, dilution 1:200) for 1 h. All steps wereperformed at room temperature. Development of colorreaction was performed with the Envision+R system-HRP(DAB) kit (Dakocytomation, Denmark). Sections werewashed in distilled water and then lightly counterstainedin Harris haematoxylin. Each batch of IHC was run withknown positive and negative controls. After mounting, thesections were examined under light microscope by twopathologists independently and were categorized as positive(antigen detected) when more than 10% of the malignantcells showed brown nuclear staining. Tumors were recordedas negative (antigen not detected) when nuclear stainingwas absent from all malignant cells but preserved in normalepithelial and stromal cells.

DNA extraction

Three 6 μm thick sections were cut from each tumor andadjacent normal blocks onto separate plain glass slides. Thetumor cell population was demarcated on the slides using anon-erasable marker after superimposing them on H & Estained slides of the same tissue block. After deparaffiniza-tion and dehydration through gradient alcohol, the marked

Table 1 Primer sequences and PCR conditions for the five Bethesdamarkers

Primer sequence PCR conditions

BAT25 TCG-CCT-CCA-AGA-ATG-TAA-GT TCT-GGA-TTT-TAA-CTA-TGC-CTC

35 cycles of 95°C for1 min, 55°C for 30 s,72°C for 45 s

BAT26 TGA-CTA-CTT-TTG-ACT-TCA-GCC AAC-CAT-TCA-ACA-TTT-TTA-ACC

D2S123 AAA-CAG-GAT-GCC-TGC-CTT-TA GGA-CTT-TCC-ACC-TAT-GGG-AC

D5S346 ACT-CAC-TCT-AGT-GAT-AAA-ACG-GG AGC-AGA-TAA-GAC-AAG-TAT-TAC-TAG

D17S250 GGA-AGA-ATC-AAA-TAG-ACA-AT GCT-GGC-CAT-ATA-TAT-ATT-TAA-ACC

35 cycles of 95°C for1 min, 52°C for 30 s,72°C for 45 s

778 Int J Colorectal Dis (2007) 22:777–782

areas were dissected using a sterile surgical blade into amicrofuge tube. Care was taken to avoid the inclusion ofthe non-tumorous areas for DNA extraction. The sampleswere digested with proteinase K (Qiagen, USA) overnightand DNA was extracted using the protocol described in thekit (Qiagen, DNA mini Kit USA). DNA concentration wasmeasured using the Nanodrop (Nanodrop Technologies,DE, USA).

MSI analysis

DNA from the tumor tissue and from the adjacent normaltissue was used as template in separate PCR reactions usingprimers for the Bethesda panel of markers (BAT25, BAT26,D2S123, D5S346, and D17S250) (Sigma, Germany) forexamining the instability of the microsatellites.

PCR for Bethesda primers

The sequences for the Bethesda panel of markers wereobtained from published literature [18] and are presented inTable 1. Each PCR reaction was carried out in a 20 μl

reaction volume containing 50 ng of genomic DNA,10 pmol of each primer and 8 μl of Master mix (2.5x)(Eppendorf, Germany) and were amplified for 35 cycles(Mastercycler Gradient, Eppendorf, Germany). PCR con-ditions (template amount, annealing temperature) werestandardized by performing gradient PCR. The PCRconditions are summarized in Table 1.

Polyacrylamide gel electrophoresis and determinationof stutter

The PCR products were run on 10% polyacrylamide geland stained with ethidium bromide and the bands wererecorded on a gel documentation system. The PCR productof normal pooled DNA (mixture of five adjacent normalmucosal epithelium gDNA) was taken as a control forPAGE analysis.

The gel image was analyzed by using the NIH imagesoftware (Scion Image). The midpoint of each marker bandwas visually identified and the distance to this point fromthe edge of the well was measured. These measurementswere plotted on a semi-log graph and a straight line was

Table 2 Reported and observed PCR amplicon size of Bethesdamarkers

Primer Reported PCRproduct size* (bp)

Observed PCRproduct size** (bp)

BAT25 103–115 113–127BAT26 110–124 108–126D17S250 142–155 146–160D5S346 101–114 114–120D2S123 197–227 204–228

*Reference [19]**PCR products’ size observed on 10% PAGE

Table 3 Clinical details—all patients

Variable N=46

Age range (years) 27–90GenderMale 26Female 20Tumor siteProximal to splenic flexure 11Distal to splenic flexure 35Family historyPositive 3Negative 43StageI 3II 18III 18IV 7

Table 4 Clinical details of patients with MSI

Pt-MSI

Age(year)

Sex Site of tumor Familyhistory

Stage

1 63 M Ascending No II2 50 M Distal to splenic

flexureNo II

3 34 F Distal to splenicflexure

No III

4 39 M Descending No III5 46 M Sigmoid colon No II

Fig. 1 MSI analysis. This is a composite of a gel image and adensitometric scan of the image using the NIH image (Scion Image)software. Using BAT26 primer, the PCR products of the DNA fromthe tumor tissue show stutter compared to the single sharp bandobtained with the DNA from adjacent normal epithelium. Normal:112-120 bp. Tumor: 97-120 bp

Int J Colorectal Dis (2007) 22:777–782 779

fitted to the points. The range of the PCR amplicon sizeobtained for each Bethesda marker compared with thereported range of PCR amplicon size obtained frompublished literature [19] is presented in Table 2. Thethickness of what were visually single bands in the normalmucosa were measured and found to represent a resolutionof 6–12 base pairs. We hence set a cut-off thickness of atleast 18 base pairs as the minimum criterion for terming aproduct as demonstrating stutter.

The presence of stutter bands/shift of bands observed inthe tumor DNA compared to the DNA from the adjacentnormal tissue was scored as the instability at that particularlocus. The samples were classified as displaying highfrequency microsatellite instability (MSI-H) if two or moreof the loci studied showed instability or as low frequencymicrosatellite instability (MSI-L) if only one of the

screened loci showed instability. Samples with no instabil-ity at these loci were scored as microsatellite stable (MSS).

Statistical analysis

Descriptive statistics were used to describe data such aspatient demographics and tumor-specific variables.

Statistical analysis was done to assess any significantcorrelation of MSI with stage, histopathological features orgrade of the disease using Student_s t test.

Results

Clinical characteristics of the CRC cohort

As shown in Table 3, 46 patients were enrolled for thisstudy out of which 26 were male and 20 were female. Theage range was between 27 and 90 years with a mean ageof 58 years. Out of 46 patients, 3 had positive familyhistory of colorectal cancer in the first/second degreerelatives. Of the 46 patients, 11 had a tumor proximal tothe splenic flexure and 35 cases were distal to the splenicflexure. A predominance of stage II–III (78.2%) diseasewas seen. Table 4 shows the clinical details of the patientswith MSI.

Table 5 Status of the microsatellite at the different loci of theBethesda panel (N=46)

Marker BAT25 BAT26 D2S123 D5S346 D17S250

MSI 1 4 1 1 1MSS 45 42 45 45 45

The table shows the number of patients who showed instability at eachof the loci.MSI: microsatellite instability.MSS: microsatellite stable.

Fig. 2 IHC staining of tumorsections for the detection ofmismatch repair proteins. Theupper panels demonstratethe dense nuclear staining ofthe glandular epithelium usinganti-MLH1 antibodies on the leftand anti-MSH2 antibodies onthe right. The lower panelsshow the haematoxylin counter-stained nuclei without the densebrown nuclear stain indicatingthe absence of the proteins

780 Int J Colorectal Dis (2007) 22:777–782

Molecular assessment of microsatellite instability

Microsatellite instability was determined using theBethesda panel of markers as shown in Table 1. Anexample of MSI is shown in Fig. 1. The shift in the bandfor BAT26 observed in the tumor DNA compared to theDNA from the adjacent normal tissue is indicative ofstutter. Of the 46 cases examined, 5 (10.8%) showedmicrosatellite instability. Of these, the instability at two ormore loci was seen in two cases and were categorized asMSI-H (4.34%) and the remaining three (6.52%) showedinstability only at one of the screened loci and wereclassified as MSI-L. Of the 46 cases, 41 (89.1%) showedno instability at any of the loci screened and were classifiedas MSS. Instability at various loci is presented in Table 5.Of the five Bethesda markers chosen, instability was seenmost frequently at BAT26 (4/5) followed by D17S250 (2/5). Our findings are consistent with other published reports[18].

Immunohistochemical detection of MMR proteins

Of the 46 cases, 8 patients had 1 of the mismatch repairproteins missing, 7 (15.2%) lacked detectable hMLH1 and1 (2.1%) hMSH2 by IHC (Fig. 2). MSI was significantlyassociated with hMLH1 (Chi square test: p<0.01) andhMSH2 defects (Chi square test: p<0.01).

Correlation with clinical and histological parameters

Statistical analysis did not reveal any significant correlationof MSI with stage, histopathological features or grade of thedisease. However, as the number of cases with familyhistory of CRC was small, no correlation was attempted.

Discussion

This study is the first report of MSI frequency in sporadiccolorectal tumors at an Indian center. Rajkumar et al. [20]have looked at the mutations in hMSH2 and hMLH1 inpatients with HNPCC of Indian cohort. It is quite possiblethat our observed frequency of MSI (10.8%) may not berepresentative of the frequency of MSI across all IndianCRC patients because all our cases were from a singleinstitution and were not consecutive referrals. Our studyindicates the preservation of the same pathogenic pathwaysin Indians as in Western cohorts despite the lower incidenceand greater proportion of rectosigmoid tumors [17, 21].

Given the lack of expression of the mismatch repairproteins hMLH1 and hMSH2 in patients with sporadicCRC who show the MSI phenotype, it might be reasonableto use IHC detection of these proteins as a screen for MSI

[22]. Though the current gold standard technique for theevaluation for MSI from tumor samples is the estimation ofthe length of PCR fragments by capillary electrophoresis offluorescent-labeled samples, we chose to use ethidiumbromide staining of fragments size fractionated by PAGEbecause it provides an easy and inexpensive methodavailable in all research labs. The primary draw back ofthis method is the lack of the resolution. While capillaryelectrophoresis of fluorescent-labeled fragments can sepa-rate fragments that differ by a single nucleotide, theslimmest band on PAGE was about six nucleotides thick.Previous comparisons of these methods have indicated thatbetween 10% and 15% of patients with instability werefalsely labeled negative [23, 24]. Hence, it is possible thatour numbers are an underestimate of the true positives.

Acknowledgement We thank Aruna Korlimarla, Shivangi Wani,Joseph Manoj Victor, and Sriranga Raju for their valuable technicalassistance.

References

1. Fearon ER, Vogelstein B (1990) A genetic model for colorectaltumorigenesis. Cell 61:759–767

2. Leob LA (1994) Microsatellite instability: marker of a mutatorphenotype in cancer. Cancer Res 54:5059–5063

3. Keller G, Rotter M, Vogelsang H, Bischoff P, Becker KF,Mueller J, Brauch H, Siewert JR, Hofler H (1995) Microsatelliteinstability in adenocarcinomas of the upper gastrointestinal tract.Relation to clinicopathological data and family history. Am JPathol 147:593–600

4. Horii A, Han H, Shimada M, Yanagisawa A, Kato Y, Ohta H,Yasui W, Tahara E, Nakamura Y (1994) Frequent replicationerrors at microsatellite loci in tumors of patients with multipleprimary cancers. Cancer Res 54:3373–3375

5. Brentnall TA (1995) Microsatellite instability. Shifting concepts intumorgenesis. Am J Pathol 147:561–563

6. Boland CR, Thibodeau SN, Hamilton SR, Sidransky D, EshlemanJR, Burt RW, Meltzer SJ, Rodriguez-Bigas MA, Fodde R,Ranzani GN, Srivastava S (1998) A National Cancer InstituteWorkshop on microsatellite instability for cancer detection andfamilial predisposition: development of international criteria forthe determination of microsatellite instability in colorectal cancer.Cancer Res 58:5248–5257

7. Gryfe R, Kim H, Hsieh ET, Aronson MD, Holowaty EJ, Bull SB,Redston M, Gallinger S (2000) Tumor microsatellite instabilityand clinical outcome in young patients with colorectal cancer. NEngl J Med 342:69–77

8. Watanabe T, Wu TT, Catalano PJ, Ueki T, Satriano R, Haller DG,Benson AB, Hamilton SR (2001) Molecular predictors of survivalafter adjuvant chemotherapy for colon cancer. N Engl J Med344:1196–1206

9. Ribic CM, Sargent DJ, Moore MJ, Thibodeau SN, French AJ,Goldberg RM, Hamilton SR, Laurent-Puig P, Gryfe R, ShepherdLE, Tu D, Redston M, Gallinger S (2003) Tumor microsatelliteinstability status as a predictor of benefit from flourouracil-basedadjuvant chemotherapy for colon cancer. N Engl J Med349:247–257

10. Aaltonen LA, Salovaara R, Kristo P, Canzian F, Hemminki A,Peltomäki P, Chadwick RB, Kääriäinen H, Eskelinen M, JärvinenH, Mecklin JP, de la Chapelle A (1998) Incidence of hereditary

Int J Colorectal Dis (2007) 22:777–782 781

nonpolyposis colorectal cancer and the feasibility of molecularscreening for the disease. N Engl J Med 338:1481–1487

11. Muc R, Naidoo R (2002) Microsatellite instability in diagnosticpathology. Curr Diagn Pathol 8:318–327

12. Chung DC, Rustgi AK (1995) DNA mismatch repair and cancer.Gastroenterology 109:1685–1699

13. Eshleman I, Markowitz S (1996) Mismatch repair defects inhuman carcinogenesis. Hum Mol Genet 5:1489–1494

14. Lothe RA, Peltomaki P, Meling GI, Aaltonen LA, Nystrom-LahtiM, Pylkkanen L, Heimdal K, Andersen TI, Moller P, Rognum TOet al (1993) Genomic instability in colorectal cancer: relationshipto clinicopathological variables and family history. Cancer Res53:5849–5852

15. Vogelstein B, Fearon ER, Hamilton SR, Kern SE, Preisinger AC,Leppert M, Nakamura Y, White R, Smits AM, Bos JL (1988)Genetic alterations during colorectal-tumor development. N Engl JMed 319:525–532

16. Wu X, Chen VW, Martin J, Roffers S, Groves FD, Correa CN,Hamilton-Byrd E, Jemal A, Comparative Analysis of IncidenceRates Subcommittee, Data Evaluation and Publication Committee,North American Association of Central Cancer Registries (2004)Subsite-specific colorectal cancer incidence rates and stage dis-tributions among Asians and Pacific Islanders in the United States,1995 to 1999. Cancer Epidemiol Biomark Prev 13:1215–1222

17. Mohandas KM, Desai DC (1999) Epidemiology of digestive tractcancers in India. V. Large and small bowel. Indian J Gastroenterol18(3):118–121

18. Loukola A, Eklin K, Laiho P, Salovaara R, Kristo P, Jarvinen HJ,Mecklin JP, Launonen V, Aaltonen LA (2001) Microsatellitemarker analysis in screening for hereditary nonpolyposis colorec-tal cancer (HNPCC). Cancer Res 61(11):4545–4549

19. Buhard O, Cattaneo F, Wong YF, Yim SF, Friedman E, Flejou JF,Duval A, Hamelin R (2006) Multipopulation analysis of poly-morphisms in five mononucleotide repeats used to determine themicrosatellite instability status of human tumors. J Clin Oncol 24(2):241–251

20. Rajkumar T, Soumittra N, Pandey D, Nancy KN, Mahajan V,Majhi U (2004) Mutation analysis of hMSH2 and hMLH1 incolorectal cancer patients in India. Genet Test 8(2):157–162

21. Deo SV, Shukla NK, Srinivas G, Mohanti BK, Raina V, Sharma ARath GK (2001) Colorectal cancers—experience at a regionalcancer centre in India. Trop Gastroenterol 22(2):83–86

22. Stone JG, Robertson D, Houlston RS (2001) Immunohisto-chemistry for MSH2 and MLH1: a method for identifyingmismatch repair deficient colorectal cancer. J Clin Pathol54:484–487

23. Cawkwell L, Li D, Lewis FA, Martin I, Dixon M, Quirke P (1995)Microsatellite instability in colorectal cancer: improved assessmentusing fluorescent polymerase chain reaction. Gastroenterology109:465–471

24. Ziegle JS, Su Y, Corcoran KP, Nie L, Mayrand PE, Hoff LB,McBride LJ, Kronick MN, Diehl SR (1992) Application ofautomated DNA sizing technology for genotyping microsatelliteloci. Genomics 14:1026–1031

782 Int J Colorectal Dis (2007) 22:777–782