INSTABILITY AND LOSS OF HETEROZYGOSITY OF TUMOR SUPPRESSOR &GENES IN BOSNIAN PATIENTS WITH SPORADIC COLORECTAL CANCER

Vesna Hadžiavdić¹,², Nada Pavlović-Čalić³, Izet Eminović*¹,²

¹ Medical Centre for Molecular Biology, Institute of Biochemistry, Faculty of Medicine, University of Ljubljana, Vrazov trg ,  Ljubljana, Slovenia ² Department of Biology and Human Genetics, Faculty of Medicine, University of Tuzla, Univerzitetska ,  Tuzla, Bosnia and Herzegovina ³ Gastroenterology Clinic, University Clinics Centre, Trnovac bb,  Tuzla, Bosnia and Herzegovina

* Corresponding author

Abstract

Considering its frequency, high mortality rate as well as many etiological mysteries colorectal cancer is a challenge to contemporary science. In our study we analyzed RER + and RER – phenotypes and their re- lations with clinical-pathological characteristics of sporadic colorectal cancers. We also analyzed genetic alterations of tumor suppressor genes as well as their relation with microsatellite instability. Th e study was based on  tumor samples and  samples of the surrounding healthy tissue of patients with colorectal cancer. According to Amsterdam Criteria and Bethesda Criteria / or , belonged in the group of sporadic colorectal cancer. Mononucleotide marker Bat  showed instability in ,; Bat  in , and Bat  in / , of tumor samples. Considering dinucleotide markers, TP  showed instabil- ity in , and DS in , of tumor samples. Genetic alterations in tumor suppressor genes were found in tumor tissue: NM  in , samples, p in ,, APC in ,, DCC in ,, RB in ,  and DCC  in ,. Our studies confi rmed that genetic instability had an important role in the development of tumor type. Our results showed that mononucleotide marker Bat  might be used for an easy and fast screening procedure in Bosnian population, because it exhibited high percent of microsatellite instability and was in relation with RER+ phenotype. Th is investigation showed that dif- ferent genetic alterations may occur during cancer development in each individual patient’s tumor. Th ese changes result in MMR inactivation, which causes RER+ phenotype. Our results suggest a connection between alteration in some tumor suppressor genes and MSI phenotype of sporadic colorectal cancer in Bosnian population.

KEY WORDS: microsatellite instability, RER phenotype, sporadic colorectal cancer, loss of heterozygosity

BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2008; 8 (4): 313-321  VESNA HADŽIAVDIĆ ET AL.: MICROSATELLITE INSTABILITY AND LOSS OF HETEROZYGOSITY OF TUMOR SUPPRESSOR GENES IN BOSNIAN PATIENTS WITH SPORADIC COLORECTAL CANCER

Introduction matic cell, it is regarded as loss of heterozygosity (LOH), which presents a marker for localization and identifi ca- Through its frequency, high mortality rate as well as tion of (). Authors established with its many etiological unknowns, colorectal cancer around   loss of heterozigosity in DCC genes in is a challenge to contemporary science. In frequency, it colorectal cancers, while other reports state  LOH follows right after lung cancer in men and breast cancer (). Loss of DCC gene in gastric cancers is about  in women. Th e latest genetic information could be used (). Loss of heterozygosity in p in colorectal cancer in the near future for the prevention of colorectal cancer, is about  (), and LOH in APC gene is about  its early diagnosis and for the selection of patients for (). In our study we analyzed RER + and RER – phe- the most suitable treatment (). in genes may notype and their relation with clinicopathological stimulate early tumor development and cause multiple characteristics in sporadic colorectal cancers. Also, we mutations in tumor tissues. Mutations in genes of DNA analyzed genetic alterations of tumor suppressor genes repair system discovered thus far, as well as the expan- as well as their relation with microsatellite instability. sion of DNA repeating sequences may serve as early in- dicators for most of mutations that characterize tumor Materials and Methods phenotype (). Cloning and molecular analysis provided proofs that the inactivation of one of MMR genes causes Our study was based on  samples of tumor and microsatellite instability (MSI) as well as positive replica- surrounding healthy tissue of patients with colorec- tion error (RER +). Most authors set microsatellite insta- tal cancer. The samples were collected at Gastroen- bility in at least   of analyzed loci as criteria for RER+ terological and Surgical Clinic of University Clin- phenotype (). RER+ phenotype is present in about  ics Center in Tuzla (Bosnia and Herzegovina). Both at hereditary non-polyposis colorectal cancer and in tumor and healthy surrounding tissue was forma- about - of sporadic colorectal cancer patients. In lin fi xed, and thereafter embedded in paraffi n blocks. HNPCC, mutations are usually found in hMLH and Method of genomic DNA isolation is based on de-par- hMSH genes, which result in loss of heterozygos- affi nisation of tissue sections and cell proteolyses with ity that inactivates MMR genes. Th ey were also found proteinase K (). Fluorescent chain synthesis of DNA is in sporadic colorectal cancers with RER+ phenotype. a method with very broad application in tumor detec- Mutations are present in hMSH, hMLH, hMSH tion, and it is especially important in determination of and GTBP genes in majority of HNPCC patients, microsatellite instability (MSI) and loss of heterozygos- and in hPMS and hPMS genes in minority (,,). ity (LOH) of tumor suppressor genes. We used mono- International Collaborative Group (ICG) published nucleotide and dinucleotide microsatellite markers in criteria for determination of microsatellite instability in the detection of microsatellite instability. In the group of colorectal cancer (). When the panel of  microsatellite mononucleotide markers we used BAT, BAT and loci is used in analysis, the tumor has microsatellite insta- BAT, and in the group of dinucleotide markers: DS bility if two or more markers are unstable. Microsatellite and TP . For LOH detection, we used intragene mark- panel must include mononucleotide markers: Bat  and ers for the following tumor suppressor genes: NM, Bat , as well as DS from the group of dinucleotide p, APC, RB, DCC and DCC (). Amplifi cation markers. These are Bethesda Criteria that determine reactions contained: , μl  x buff er; μl DNA (- which colorectal cancers may be tested for microsatel- ng);  μl MgCl (mM); , μl of each dNTP (, mM); lite instability (). According to these criteria, tumors ,μl AmpliTaq Gold polymerase (, U; Perkin-Elmer); are classified with regard to microsatellite instability: μl forward (, pmol) and μl reverse primer (, • MSI – L low rank tumors that show MSI in less than pmol) and sterile water until fi nal volume of μl. PCR - loci; was performed in a PCR Thermocycler  (Perkin- • MSI-H high rank tumors that show MSI in over  Elmer). PCR conditions included an initial denatur- loci and have positive replication error - RER+, and ation step at °C for  min,  cycles of denaturation at • MSS- tumors with stable microsatellite loci. °C for  s, annealing at °C for  s and elongation In  in Amsterdam, International Collaborative at °C for  s, and fi nal extension at °C for  min. Group established Amsterdam Criteria that are applied Amplified PCR products were separated in an auto- in diagnosing HNPCC families (). Th ese criteria were mated sequencer  ABI PRISM, Genetic Analyser  also used in our study. Generally, when the other allele (Perkin Elmer), which enabled separation and quantifi - of tumor suppressor gene is lost or inactivated in a so- cation of DNA fragments according to the principles of

BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2008; 8 (4): 314-321 VESNA HADŽIAVDIĆ ET AL.: MICROSATELLITE INSTABILITY AND LOSS OF HETEROZYGOSITY OF TUMOR SUPPRESSOR GENES IN BOSNIAN PATIENTS WITH SPORADIC COLORECTAL CANCER capillary electrophoresis. Microsatellite analysis com- In the group of dinucleotide markers, TP  showed in- prises of comparison between healthy and tumor tissue stability in / (,) and DS in / (,) of the same patient using Genescan program package of tumor samples. With regard to instability among for analysis. Software detects fl uorescence peaks and mononucleotide markers the diff erence was found sig- shows them in the form of electropherograme. Each fl u- nifi cant (p<,), however, the diff erence between di- orescent peak is automatically quantifi ed for size in bp, nucleotide markers was not (p>,). Signifi cance test height and peak fi eld. All samples were tested twice for showed significant differences between mononucle- the confi rmation of results. Loss of heterozygosity was otide and dinucleotide markers with regard to micro- calculated mathematically () In heterozygous cases, satellite status (p<,). Number of unstable microsat- allele ratio was calculated for each pair of normal and tu- ellite loci in certain sample defi nes its RER phenotype mor tissue according to formula T:T/N:N , where as either RER+ or RER-. Th e study showed that / T and N are fi eld values of shorter allele; T and N (,) of tumor samples belong to RER+ phenotype, are fi eld values of longer allele in tumor (T) and healthy and / (,) belong to RER- phenotype. Th ere (N) samples. Th e results were assigned rank from , is a signifi cant diff erence between RER+ and RER- tu- to ,. If the result was lower or equal to , the loss of mor phenotype in frequency of unstable microsatel- heterozygosity of longer allele was considered signifi cant. lite loci (p<,). In the group of tumors with RER+ Homozygous cases cannot be included in calculation. phenotype microsatellite instability was established at We used χ test and Fisher’s exact test (Arcus Quickstat mononucleotide marker Bat  in / (,), and biomedical for Windows) in statistical data processing. dinucleotide marker TP in / (, ) of samples. In the group of tumors with RER- phenotype, mono- Results nucleotide marker Bat  showed instability in / (,) cases, and dinucleotide marker TP  as well as Analysis of genetic instability at microsatellite loci and Bat  showed instability in / (,) tumor tissues. RER phenotype of sporadic colorectal cancers There is a significant difference between mononucle- In the total sample of  patients  or , otide and dinucleotide markers with regard to RER belong to sporadic colorectal cancer, accord- status (p<,). Statistically signifi cant diff erences were ing to Amsterdam Criteria and Bethesda Criteria. also found among dinucleotide markers with regard Mononucleotide marker Bat  showed instabil- to RER phenotype (p=,). Analysis of some clinico- ity in / (,); Bat  in / (,) and pathological parameters (sex, ages, tumor localization, Bat  in / (,) of tumor samples (Figure ). hystopathological classifi cation) showed that colorectal

BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2008; 8 (4): 315-321  VESNA HADŽIAVDIĆ ET AL.: MICROSATELLITE INSTABILITY AND LOSS OF HETEROZYGOSITY OF TUMOR SUPPRESSOR GENES IN BOSNIAN PATIENTS WITH SPORADIC COLORECTAL CANCER

Clinicopathological RER + tumor phenotype RER – tumor phenotype Total characteristics No. % No. % Sex male 26 14 53,85 12 46,15 female 9 5 55,56 4 44,44 Age category > 50 24 12 50,00 12 50,00 < 50 11 7 63,64 4 36,36 Tumor localization Left-side 24 15 62,50 9 37,50 Right-side 11 4 36,36 7 63,64 Histopathological classifi cation Mucion 5 4 80,00 1 20,00 adenocarcinoma 30 15 50,00 15 50,00

TABLE 1. Relations between RER tumor phenotype and clinicopathological characteristics cancer was found mostly in males / (, ) and in tumor tissues with loss of heterozygosity. Microsatelite the age group over  / (,), tumors were lo- instability of dinucleotide marker TP was found in cated on the left-side (rectum and sigma region) in / tumor tissues with loss of heterozygosity at RB in / (,), and / (,) were adenocarcinomas. (,), and marker DS was present in tumors with Analysis of microsatellite instability of individual mark- loss of heterozygosity at APC tumor suppressor gene in ers and clinicopathological characteristics showed no / (,) samples. Th ere is no signifi cant diff erence statistically signifi cant diff erence (p>,). Analysis of (p>,) in the occurrence of microsatellite instability RER tumor phenotype and clinicopathological charac- of dinucleotide markers in tumor tissues with loss of teristics showed that RER+ phenotype was more fre- heterozygosity. Comparison between RER phenotype quent in females (/ or ,). In the age group un- and genetic alterations showed that loss of heterozy- der  it was present in / (,) of samples, and gosity at tumor suppressor gene APC was pronounced according to the tumor localization RER+ phenotype in tumors belonging to RER+ phenotype (in / or was more frequent in the left-side (/ or ,). ,), while the smallest percent of LOH was found at RER- phenotype was more frequent in males (/ RB and DCC (in / or ,). Loss of heterozygos- or ,); in the age group over  in / () and ity at tumor suppressor gene p was found in tumors they belong to the group of right-side tumors in / with RER- phenotype in / or ,, while the small- or , (Table ). Statistic analysis showed no sig- est frequency was found at RB in / ()(Table ). nifi cant diff erences between RER+ and RER- phenotype with regards to clinicopathological parameters (p>,). Th ere is no signifi cant diff erence between RER+ pheno- type and RER- phenotype of tumors with regards to the Genetic alterations in tumor suppressor genes were frequency of LOH at tumor suppressor genes (p>,). found in NM  in / (,) samples, in p in In RER+ tumor phenotype, allele loss most frequently / (,), in APC in / (,), in DCC in occurred at two loci simultaneously (/ or ,); / (,), in RB in / (, ) and DCC  in which is also the case in tumors with RER- phenotype / (,) of tumor tissues. Th e highest frequency of homozygous samples were found for DCC  tumor sup- RER + tumors RER – tumors Genetic alteration N=19 N=16 pressor gene in / (,). DCC  was homozygous No. % No. % in / (,) and p locus in / or ,. Th ere NM23 is a signifi cant diff erence between tumors with allele AI 10 52,63 9 56,25 P53 loss and tumors without allele loss (p<,). Microsatel- AI 8 42,11 10 62,50 lite instability of marker Bat  was found in tumor tis- APC sues with loss of heterozygosity at locus NM in / AI 11 57,89 7 43,75 RB1 (,). Microsatellite instability in Bat  was found AI 4 21,05 4 25,00 in tumor tissues with LOH in RB  in / (,) and DCC1 marker Bat  was unstable in tumors with LOH in APC AI 4 21,05 6 37,50 DCC2 tumor suppressor gene in / (,) samples. Th ere AI 6 31,58 6 37,50 is no signifi cant diff erence (p>,) in the occurrence of microsatellite instability of mononucleotide markers in TABLE 2. Relations between RER phenotype and genetic alterations

 BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2008; 8 (4): 316-321 VESNA HADŽIAVDIĆ ET AL.: MICROSATELLITE INSTABILITY AND LOSS OF HETEROZYGOSITY OF TUMOR SUPPRESSOR GENES IN BOSNIAN PATIENTS WITH SPORADIC COLORECTAL CANCER

(/ or ,). Th ere is no signifi cant diff erence between in certain markers, we also selected fi ve microsatellite RER+ and RER- phenotypes with regards to the number markers and divided them into MSI-H and MSI-L group. of loci with loss of heterozygosity (p>,). Analysis of Our results suggested , frequency of MSI-H status genetic alteration and clinicopathological characteris- markers while the frequency of MSI-L was ,. Th e tics showed that loss of heterozygosity at gene p was diff erences between mononucleotide and dinucleotide more frequent in females (/ or ,), while in males markers with regards to their microsatellite status are allele loss was found at NM gene in / (,). In signifi cant (p<.). Microsatellite instability of mono- the age group under , loss of heterozygosity was found nucleotide markers Bat  and Bat  are highly sensi- at p gene in / (,) samples, while in the age tive and they represent specific instability indicator group over  loss of heterozygosity was found at NM (). Microsatellite instability of Bat  marker in iden-  gene as well as in APC in / (,). Regarding tifi cation of sporadic forms of collorectal cancers may tumor localization, in left-side tumors loss of heterozy- become useful for identifi cation of new HNPCC cases. gosity was more frequent in APC gene with / Microsatellite instability was found in  samples. Cor- (,); while in the right-side tumors, allele loss was relation between MSI status and clinicopathological more frequent in p gene in / (,) (Figure ). parameters is signifi cant. Tumors appear in the younger In mucionus adenocarcinoma loss of heterozygosity was group of patients with predominant location in the more frequent at locus TP in / (), in adenocar- right-side, and tissues had more mucinous components. cinoma allele loss was present at NM in / (,) Analysis shows that  tumor tissues with MSI positive (Table ). No significant difference between loss of status at Bat  were new germline mutations in MLH heterozygosity in tumor suppressor gene and clinico- gene. Authors consider that mononucleotide marker patholoical characteristics was established (p>,). Bat  may serve as an easy and fast screening procedure preceding analysis (). Previous studies con- Discussion fi rmed Bat  and Bat  to be highly sensitive (in  and  cases respectively) for the detection of micro- There is a belief that changes in the system of MMR satellite instability in adenomas and cancers (). In our genes cause not only HNPCC development, but sporad- study mononucleotide marker Bat  showed instability ic collorectal cancer as well ().Analysis of MSI status in , cases; Bat  in , and Bat  was highly in Japanese patients () showed microsatellite instabil- unstable in as many as , of tumor samples (Figure ity in  tumors while  showed that they belong to ). Among dinucleotide markers, TP showed insta- MSI-H tumors. Other authors () confi rmed presence bility in , and DS in , of tumor samples. of microsatellite instability in  patients with multiple Study of Hong Kong patients showed  frequency of primary cancers in diff erent organs. At the same time, RER+ phenotype in sporadic collorectal cancers (). there are reports of , tumor tissues with unstable Some studies found that , tumors belong to RER + microsatellite loci (). According to the results of au- phenotype () or that RER + phenotype ranked between thors (,-) who analyzed microsatellite alterations  and  (). Also, replication errors were found in

BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2008; 8 (4): 317-321  VESNA HADŽIAVDIĆ ET AL.: MICROSATELLITE INSTABILITY AND LOSS OF HETEROZYGOSITY OF TUMOR SUPPRESSOR GENES IN BOSNIAN PATIENTS WITH SPORADIC COLORECTAL CANCER

 of collorectal cancers (). Our results suggest rath- males about  years old. Th ese studies showed a signifi - er high frequency of RER+ phenotype of ,, which cant relation between RER+ tumors and tumor localiza- is higher than previously reported. Also, there is signifi - tion in proximal colon. Th ere is no signifi cant diff erence cant diff erence between mononucleotide and dinucle- between these two groups of tumors and RER- tumor otide markers with regard to RER+ phenotype (p<,). localization (). Results of our studies (Table ) concur Our studies confi rmed that genetic instability plays an with previous reports that RER+ phenotype was more important role in the development of this type of tumor. frequent in age group under  with ,. Majority According to previous reports, many patients with spo- of RER+ tumors were left-side tumors ,, and these radic collorectal cancers may develop familial history appear in females in ,. RER- phenotype was more of collorectal cancer. Th e percentage of CRC was the frequent in males in ,; in the age group over  in highest in families with MSS-phenotype tumors (),  and in , those tumors were located in the right followed by MSI-L phenotype () and MSI-H () side. Statistical analysis showed no signifi cant diff erence phenotype tumors. MSS tumors were preferentially lo- between RER+ and RER- tumor phenotypes with regard calized in the distal colon (). Studies of MSI+ pheno- to clinicopathological parameters (p>,). Previous type of colorectal cancers as determined by the analysis studies have shown an inverse relation between RER of Bat  mononucleotide marker established signifi cant phenotype and loss of heterozygosity at chromosomes diff erence between males and females with regard to q, p and q (). In Hong Kong patients, RER+ tu- tumor location: MSI+ tumors were more frequent in mor phenotype, which was compared with RER- tumor females (,) in the left segment of colon. Age diff er- phenotype, showed low level of mutations in the form of ences were not signifi cant (age of tumor appearing was LOH at tumor suppressors as well as p in , MCC between  and ) (). Further studies showed that in , DCC in  and APC in . High frequencies of in   cases microsatellite instability was registered at allele loss of over   () were found at DCC and APC all of the fi ve tested dinucleotide markers (DS,etc.) locus, but at RB  locus the frequency was exceptionally Frequency of MSI positive status in proximal colon was low, less than . In Caucasian and Japanese patients, ,, in distal , and in rectum ,. High MSI-H loss of heterozygosity at q chromosome was over showed a signifi cant correlation with proximal tumor , and loss of heterozygosity at APC was between localization. All the tumors found in rectum had MSI-L  and  (). Our studies showed the highest fre- status. Also, the analysis showed no correlation between quency of genetic alterations in tumor suppressor gene clinicopathological characteristics and MSI tumor status NM  which were identifi ed in , samples. Th en, (). Results of our studies are in agreement with previ- tumor suppressor p was identifi ed in ,, APC in ous studies. Analysis of some clinicopathological param- ,, DCC tumor suppressor gene in ,, DCC  eters (sex, ages, tumor localization, histo-pathological in , samples of tumor tissue. Th e lowest percent of classification) showed the dominance of males with LOH in a tumor suppressor gene was found at RB in , and persons over  years of age with ,. Tu- ,  samples. Loci homozygosity was found at DCC mors are located in the left-side (rectum and sigma re-  tumor suppressor gene in ,, at DCC  in , gion) ,, and , of samples were adenocarcino- and at p locus in , samples. Th ere is a signifi cant mas. Analysis showed no signifi cant diff erence between diff erence between tumors with allele loss and tumors RER+ and RER- tumors with regard to age. In RER- without allele loss (p<,). Results of a study () con- group the average age was about , years and , fi rmed that replication error and loss of heterozygosity at years in RER+ group (). Further studies established a q, p, p, q and q chromosomes were exceptional relation between RER phenotype and histo-pathological pre-metastatic events in tumor genesis of colorectal variables in patients from Hong Kong. Th ere is no sig- cancer. Studies of genetic instability at tumor suppres- nifi cant diff erence between RER+ phenotype and the lo- sor genes p and NM  as well as clinicopathological calization of tumor which is usually right-sided. Th ere is forms at rectal cancers established alterations in NM a signifi cant diff erence with regards to sex and age. Th e  in all RER+ tumors, while p tumor suppressor was data suggest that RER+ phenotype is more frequent in altered in one case. Th ere is no statistically signifi cant males. Also, the data show that younger patients under diff erence between clinical parameters and RER status  years of age had RER+ tumor phenotype and that the (). Loss of heterozygosity was confi rmed at q chro- diff erence between RER+ and RER- tumor phenotype mosome in , samples and RER+ phenotype was is signifi cant (). RER+ tumors were found in patients found in , cases. Th ere is no signifi cant diff erence  years old, while RER- tumors were found in older fe- between LOH at q chromosome and clinicopatho-

 BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2008; 8 (4): 318-321 VESNA HADŽIAVDIĆ ET AL.: MICROSATELLITE INSTABILITY AND LOSS OF HETEROZYGOSITY OF TUMOR SUPPRESSOR GENES IN BOSNIAN PATIENTS WITH SPORADIC COLORECTAL CANCER

Genetic alterations in tumor suppressor Clinicopathological NM 23 p53 APC RB 1 DCC1 DCC2 N characteristics No. % No. % No. % No. % No. % No. % Sex male 26 15 57,69 11 42,3 13 50,00 5 19,23 10 38,46 12 46,15 female 9 4 44,44 7 77,78 5 55,56 3 33,33 0 0 0 0 Age category 13 54,17 > 50 24 13 54,17 10 41,67 6 25,00 7 29,17 7 29,17 5 45,45 < 50 11 6 54,55 8 72,73 2 18,18 3 27,27 5 45,45 Tumor localization 9 37,5 Left-side 24 12 50,00 14 58,33 4 16,67 7 29,17 10 41,67 9 81,82 Right-side 11 7 63,64 4 36,36 4 36,36 3 27,27 2 18,18 Histopathological classifi cation Mucinous 5 3 60,00 4 80,00 3 60,00 1 20,00 2 40,008 3 60,00 adenocarcinomas 30 16 53,33 14 46,67 15 50,00 7 23,33 26,67 9 30,00

TABLE 3. Relations between genetic alterations in tumor suppressor gene and clinicopathological characteristics logical characteristics with regard to sex, tumor state With regards to the tumor localization, loss of and diff erentiation level. Allele loss was more frequently heterozygosity was more frequent in left-sided tu- detected in adenocarcinoma located in distal part of co- mors at APC gene in , samples. In right-sided lon than in tumors located in proximal part (). Loss tumors allele loss at p was ,. In mucionous of heterozygosity was detected in malignant biopsies of adenocarcinoma, loss of heterozygosity was found at gastro-esophagus in   cases. LOH in APC, DCC tu- locus TP  in  samples. In adenocarcinoma al- mor suppressors as well as in MSH loci can be detected lele loss was found at NM  in , cases (Table in histologically normal tissues and in adenocarcinomas ). There is no significant difference between loss of (). Th ose are potential markers for early neoplastic heterozygosity in tumor suppressor genes with re- progression. Microsatellite instability in three loci simul- gard to clinicopathological characteristics (p>,). taneously was found in , cases, specifi cally, in the In a previous study () RER status and loss of heterozy- group of patients with average age of  years. Th is fi nd- gosity at p, p, q, p, p and q chromosomes were ing suggests a very important role of MMR gene in can- analyzed. Th e authors used CA repeat motive sequence cer pathogenesis at this age (). Th eir research showed of dinucleotide markers. Th ey found higher frequency of loss of heterozygosity in the region of DCC tumor sup- LOH in RER- tumor phenotype (,) than in RER+ tu- pressor in , patients at the age of . Th e analysis of mor phenotype (,). Analysis of microsatellite insta- genetic alterations and microsatellite instability revealed bility and loss of heterozygosity of p and APC tumor no correlation with clinicopathological characteristics. suppressor genes in  colon and rectum tumors con- No patient had family history of colorectal cancer and fi rmed that  tumors belong to RER+ phenotype (). all tumors were located in the left-side. Th ere are two Loss of heterozygosity at p was found in  samples large groups of colorectal cancers which can be individ- and at APC in  samples. Authors established lower ualized and which suggest correlation among molecular, frequency of RER+ phenotype in rectal tumors with ear- clinical and pathological forms of tumors (). Th e fi rst lier loss of heterozygosity. In colorectal tumors with the group consists of tumors with LOH in  cases, which same phenotype, mutations in APC locus were less fre- are located in distal part of colon. The second group quent than mutations in p. Tumors with RER- pheno- includes tumors with pronounced microsatellite insta- type and no loss of heterozygosity can still develop as a bility without allele loss. Th ese tumors make  of all result of carcinogenesis model. Some authors found loss colorectal cancers and in   belong to the group of of heterozygosity at p chromosome in  of right– right-side tumors. Authors conclude that these tumors sided and  of left-side tumors. RER+ phenotype was had better prognosis. Our analysis of genetic alterations found in  of right-sided tumors and   of left-sided and clinicopathological characteristics showed loss of tumors (, ). Th e heterogeneous pattern of tumor heterozygosity at p in , cases, and that allele mutations suggests that multiple alternative genetic loss at NM  was found in males in , samples. In pathways to colorectal cancer exist and accepted ge- the age group under , loss of heterozygosity was the netic model of cancer development is not representative most prominent at p gene in , samples, and in of major tumors (). Our results show that the LOH in the age group over  loss of heterozygosity appeared APC is specially frequent in tumors that belong to RER+ at NM  and APC genes in , samples (Figure ). phenotype (,), and that the least frequent were

BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2008; 8 (4): 319-321  VESNA HADŽIAVDIĆ ET AL.: MICROSATELLITE INSTABILITY AND LOSS OF HETEROZYGOSITY OF TUMOR SUPPRESSOR GENES IN BOSNIAN PATIENTS WITH SPORADIC COLORECTAL CANCER

LOH at RB and DCC in , cases. Tumors with in  cases (Table ). Th ere is no signifi cant diff erence RER- phenotype had presented loss of heterozygosity at between RER+ tumor phenotype in occurrence of loss p gene in , samples. Th e lowest one was at RB  of heterozygosity of tumor suppressor genes (p>,).

Conclusion

Our results suggest that mononucleotide marker Bat  can be used as an easy and fast screening procedure, because it exhibits high frequency of microsatellite instability and is associated with RER+ phenotype. Mononucleotide and di- nucleotide markers exhibit high microsatellite instability in rectum and colon. Poly-A repeat sequences (mononucle- otide markers) show higher microsatellite instability than CA repeat sequences (dinucleotide markers). Analysis of loss of heterozygosity in tumor suppressor genes established high frequency of genetic alterations in tumor suppressor gene NM  in , samples, and p and APC in , samples in the group of sporadic tumors. However, it is important to underline that diff erences in frequencies of microsatellite instability between our study and those of other authors sug- gest the importance of abundance and type of microsatellite markers analyzed in certain type of cancer. Our study has shown that genetic alterations in tumor may diff er at individual level and refl ect changes that result from the process of carcinogenesis. Th ose changes inactivate MMR genes and potentially lead to RER phenotype development. Our results suggest an association between alterations in certain tumor suppressor genes and MSI phenotype of this tumor type.

List of Abbreviations MMR - Mismatch repair genes MSI - Microsatellite instability RER +, – - Replication error positive hMSH - Human mut S homolog  hMSH - Human mutS homolog  hPMS and hPMS - Human post-meiotic segregation  and  GTBP - G/T mismatch-binding protein FAP - Familial adenomatous polyposis APC - Adenomatous polyposis coli LOH - Loss of heterozygosity DCC - Deleted colorectal carcinoma MCC - Mutated in colorectal cancer

References

() Baba S. Recent advances in molecular genetics of colorectal can- workshop on microsatellite instability for cancer detection and cer. World J. Surg. ; : -. familial predisposition: development of international criteria for () Loeb L.A. Microsatellite instability: marker of mutator phenotype the determination of microsatellite instability in colorectal can- in cancer. Cancer Res. ; : -. cer. Cancer Res. ; : -. () Kolonder R. Biochemistry and genetics of eukariotic mismatch () Rodriguez-Bigas M.A., Boland C.R., Hamilton S.R., Henson D.E., repair. Genes Dev. ; : -. Jass J.R., Khan P.M., Lynch H., Perucho M., Smyrk T., Sobin L., () Bell S.M., Kelly S.A., Hoyle J.A., Lewis F.A., Taylor G.R., Th omp- Srivastava S. A National Cancer Institute workshop on hereditary son H., Diyon F.G., Quirke. c-Ki-ras gene mutation in dysplasia nonpolyposis colorectal cancer syndrome meeting highlights and and carcinomas complicating ulceratie colitis. Br. J. Cancer ; Bethesda guidelines. J. Nat. Cancer Inst ; :-. : -. () Annonymous. Hereditery Non-Polyposis Colon Cancer Syn- () Akiyama Y., Sato H., Yamada T., Nagasaki H., Tsuchiya A., Abe drome, Mayo Press, Mayo Clinic, Mayo : -. R., Yuasa Y. Germ-line mutations of hMSH/GTBP gene in an () Cawkwell L.A., Levis F.A., Quike P. Frequency of allele loss of atypical hereditary nonpolyposis colorectal cancer kindred. Can- DCC, p, RB, WT, NF, NM  and APC/MCC in colorectal cer Res. ; :-. cancer assayed by fl uorescent multiplex polymerase chain reac- () Miyaki M., Konishi M., Tanaka K., Kikuchi-Yonoshita R., Mu- tion. Br. J. Cancer ; :-. raoka M., Yasuno M., Igari T., Koike M., Shiba M., Mori T. Germ- () Huang T.H.M., Quesenbery S.T., Martin M.B., Loy T.S., Diaz line mutations of MSH as the cause of hereditary nonpolyposis Arias A.A. Loss of heterozygosity detected in formalin fixed, colorectal cancer (letter). Nat Genet. ; :-. paraffi n embedded tissue of colorectal carcinoma using a micro- () Bolland C.R., Th ibodeau S.N., Hamilton S.R., Sydranski D., Esh- satellite located within the deleted in colorectal carcinoma gene. leman J.R., Burt R.W., Meltezer S.J., Rodriguez-Bigas M.A., Fod- Diagnosis Mol. Pathol. ; :- der R., Ranzani G.N., Srivastava S. A National Cancer Institute

 BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2008; 8 (4): 320-321 VESNA HADŽIAVDIĆ ET AL.: MICROSATELLITE INSTABILITY AND LOSS OF HETEROZYGOSITY OF TUMOR SUPPRESSOR GENES IN BOSNIAN PATIENTS WITH SPORADIC COLORECTAL CANCER

() Uchino S., Tsuda H., Noguchi H., Yokota Y., Terada M., Saito () Elsaleh H. Th e microsatellita instability phenotype in human col- T., Kabayashi M., Sugimura T., Hirohashi S. Frequent loss of orectal carcinoma: relationship to sex, age and tumor site. Gas- heterozygosity at the DCC locus in gastric cancer. Cancer Res. troenterol. ; :-. ; : - () Ikeda Y., Oda S., Abe T., Shinji O., Maehara Y., Sugimachi K. Fea- () Lothe R.A., Fossli T., Danielsen H.E., Stenwing A.E., Nesland J.M., tures of microsatellite instability in colorectal cancer: Compari- Gallie B., Borresen A.L. Molecular genetic studies of tumor sup- son between colon and rectum. Oncology ; : -. pressor gene regions on chromosomes  and  in colorectal () Feeley K.M., Fullard J.F., Heneghan M.A., Smith T., Maher M., tumors. J. Nat. Cancer Inst. ; : -. Murfy R.P., Gorman A. Microsatellite instability in sporadic col- () Neuman W.L., Wasylyshyn M.L., Jacoby R., Errol F., Angriman I., orectal carcinoma is not indicator of prognosis. J. Pathol. ; Montag A., Brasitus T., Michelasi F., West-Brook C.A. Evidence : -. for a common molecular pathogenesis in colorectal, gastric and () Iacopetta B.J., Welch J., Soong R., House A.K., Zhou X.P. Muta- pancreatic cancer. Genes Chromos. Cancer ; :-. tion of the transforming growth factor beta type II receptor gene () Brown S.R., Finana P.J., Hall N.R., Bishop D.T. Incidence of DNA in right sided colorectal cancer: relationship to clinicopathologi- replication error in patient with multiple primary cancers. Dis. cal features and genetic alterations J. Patholog. ; :-. Colon Rectum ; :-. () Tomlison I., Ilyas M., Johnoson V., Davies A., Clark G., Talbot () Yamashita K., Arimura Y., Kurokawa S., Itoh F., Endo T., Hirata I., Bodmer W. A comparasion of the genetic pathways involved K., Imamura A. Microsatellite instability in patients with multiple in the pathogenesis of three types of colorectal cancer. J. Pathol. primary cancers of gastrointestinal tract. Gut ; : -. ; : -. () Horri A., Han H.J., Shimada M., Yangisawa A., Katu Y., Ohta H., () Vogelstein B., Fearon E.R., Kern S.E., Hamilton S.R., Presinger Yasui W., Tahara E., Nakamura Y. Frequent replication errors at A.C., Leppert M., Nakamura Y., White R., Smits A.M.M., Bos J.L. microsatellite loci in tumors patients with multiple primary can- Genetic alterations during colorectal – tumor development. N. cers. Cancer Res. ; : -. Engl. J. Med. ; :-. () Fanga D.C., Jass J.R., Wang D.X., Zhou X.D., Lou Y.H., Young J. () Blacker H., Graf M., Rieker R.J., Otto H.F. Comparison of loss of Infrequent loss of heterozygosity of Apc/Mcc and Dcc, genes in heterozygosity and replication errors in primary colorectal car- gastric cancer showing DNA microsatellite instability. J. Clinic. cinomas and corresponding liver matastases. J. Pathol. ;  Pathol. ; :-. -. () Hoang J.M., Cottu P.H., Thuille B., Salmon R.J., Thomas G., () Lanza G., Matteuzzi M., Gaff a R., Irvieo E., Maestri I., Santini A., Mamellin R. Bat  an indicator of the replication error pheno- del Senno L. Chromosome q allelic loss and prognosis in stage type in colorectal cancers and cell lines. Cancer Res. ; :- II and III colon cancer. Int. J. Cancer ; : -  () Dolan K., Grade J., Walker S.J., Sutton R., Gosney J., Field J.R. () Dietmaier W., Wallinger S., Bocker T., Kulliman F., Fishel R., Histological and molecular mapping of adenocarcinoma of the Rusehoff J. Diagnostic microsatellite instability definition and oesophagus and gastroesophageal junction: loss of heterozygosity correlation with mismatch repair protein expression. Cancer Res. occurs in histologically normal epithelium in the oesophagus and ; :-. stomach. Oncol. Rep. ; : -. () Samowitz W., Slattery M.L., Potter J.D., Leppert M.F. Bat  and () Chetty R., Naidoo R., Shneider J. Allelic imbalance and microsat- Bat  instabillity in colorectal adenomas and carcinomas and ellite instability of the Dcc gene in colorectal cancer in patients germline polymorphisms. Amer. J. Pathol. ; : -. under the age of  using fl uorescent DNA technology. J. Clin. () Cravo M., Lage P., Albuguerque C., Chaves P., Claro I., Gomes T., Patholog. ; :-. Gaspar C., Fidalgo F., Soares J., Nobre-Leitao C. Bat  identifi es () Martin L, Assm M, Piard F. Are there several types of colorectal sporadic colorectal cancer with mutator phenotyp: a correlative carcinomas? Correlation with genetic data. Eur. J. Cancer Prev. study with clinoco-pathological features and mutations in mis- ; :-. match repair genes. J. Pathol. ; : -. () Massa M.J., Iniesta P., Gonzalez-Quevedo R., de Juan C., Caldes () Grady W.M., Rajput A., Mycroff L., Liu D.F., Kwon K., Willis J., T., Sanches-Pernante A., Cerdan J., Turres A.J., Balibrea J.L., Beni- Markowitz. Mutation of type II transforming growth factor beta to M. Diff erential prognosis of replication error phenotype and to receptor is concident with the transformation of human colon heterozygosity in sporadic colorectal cancer. Eur. J. Cancer. ; adenomas to malignant carcinomas. Cancer Res. ; : - : -. . () Rebishung C., Laurent–Puing P., Gerard J.P., Th omas G., Hamelin () Ko-Yee J.M., Cheung M.H.Y., Kwan M.W., Wong C.M., Lau R. Analysis of genetic disorders of cancer of rectum: diff erences W.K., Tang C.M., Lung M.L. Genomic instability and alterations in relation to cancer of the colon. Gastroenterol. Clin. Biol. ; in Apc, Mcc and Dcc in Hong Kong patients with colorectal car- : -. cinoma. Int. J. Cancer ; :-. () Watatani M., Yoshida T., Kuroda K., Ieda S., Yasutoni M. Allelic () Indinnimeo M., Cicchini C., Stazi A., Mingazzini P. Genetic insta- loss of chromosome p mutation of p gene, a microsatellite bility p and nm  mutation and clinicopathological features in instability in right and left sided colorectal cancer. Cancer ; rectal carcinoma. Antican. Res. ; : -. : -. () Speicher M.R. Microsatellite instability in human cancer. Oncol. () Narayan S., Roy D. Role of APC and DNA mismatch repair gene Res. ; .-. in the development of colorectal cancer. Mol. Cancer ; () Shiemann U., Müller-Koch., Gross M., Daum J., Lohse P., Baret- :-. ton G., Muders M., Mussack T., Kopp R., Holinski-Feder E. Ex- tended microsatellite analysis in microsatellite stable, MSH, and MLH mutation-negative HNPCC patients: genetic reclassifi ca- tion and correlation with clinical features. Digestion ; :- .

BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2008; 8 (4): 321-321 