MISMATCH REPAIR PROTEINS AND SURVIVIN IN ADENOMATOUS COLON POLYPS WITH LOW GRADE AND HIGH GRADE DYSPLASIA : AN IMMUNOHISTOCHEMICAL STUDY

Objective: The aim of our study was to observe the immunohistochemical expression pattern of mismatch repair proteins (MMRP) MLH1, MSH2, MSH6 and PMS2, as well as survivin, in colon polyps. Methods: We assessed above mentioned proteins in a unified group of 124 tubular adenomatous colon polyps with regard to the presence of dysplastic abnormalities in order to explore their relationship. Furthermore, we studied their relation to such clinic-morphological parameters as the age of patients, size of adenoma, degree of dysplastic changes and localization of the lesion. Results: Survivin was expressed in 97 cases (78.2%), MLH1 was found in 111 cases (89.5%), MSH2 in 115 cases (92.7%), MSH6 in 118 cases (95.2%) and PMS2 in 105 cases (84.7%). The majority of absent MMRP cases was detected where the adenoma size was less than 10 mm with LGD (lowgrade dysplasia). Survivin expression significantly correlated with the adenoma size and dysplasia grade. Subcellular survivin compartmentalization was statistically associated with the adenoma size, dysplasia grade and adenoma localization. Furthermore, we confirmed a significant relation between survivin expression and MMRP. In general, the intensity of immunoreaction was stronger in the MMRP than in survivin. Conclusions: Our recent results suggest that MMRP may suppress the antiapoptotic activity of survivin in LGD and HGD (high grade dysplasia) colon adenomas.


INTRODUCTION
The mismatch repair system (MMR) is utilized by proliferating cells to correct errors (mutations) that may develop during DNA replication (Jover et al., 2004;Vilar and Gruber, 2010;Hassen et al., 2012).DNA MMR also controls cell cycle checkpoints and thus genomic stability, and play an important role in apoptosis in response to DNA damage (Cejka et al., 2003).MMR genes are ubiquitous genes encoding mismatch repair proteins (MMRP).There is a tendency to heterodimerization of MMRP to functional doublets.MSH2 forms a heterodimer with MSH6 (or MSH3), which is involved in the diagnostic recognition of misaligned nucleotides and mispaired insertion-deletion loops.Furthermore, MSH2-MSH6 (MSH3) doublet recruits and activates proteins MLH1 and PMS2 (Peltomäki, 2003).Subsequently, the MLH1-PMS2 protein complex recruits nucleases, polymerases and other assorted proteins, which initiate downstream repair functions including the excision of the mismatched DNA strands or microsatellite instable (MSI) sequence repeats (Scherer et al., 2005;Vilar and Gruber, 2010).Defects in the MMR cause an increased spontaneous mutation rate known as the "mutator phenotype" (Charames and Bapat, 2003).In addition to elevated mutation rate, loss of MMR function may also lead to instability in simple sequence repeats (microsatellites).During mono-, di-, tri-, and tetranuclolide repeats replication, a newly formed strand may slip along the original template, thus leading to a bulged mispaired insertion-deletion loop.MMR mutations predominantly in MLH1, MSH2, MSH6, and PMS2 may cause the production of inactivated or abnormally short proteins that cannot perform their normal functions.In the light of all this, loss of protein expression suggests a defective MMR.Protein survivin is a unique member of the inhibitor of apoptosis protein (IAP) family (Amiri and Richmond, 2005).IAP proteins play a key role in the negative regulation of apoptosis (type of programmed cell death spectrum).The multifunctional survivin protein possesses a number of distinct features not shared with the other IAP family members, it is involved in the regulation of cell cycle, inhibits the apoptotic cascade and stimulates angiogenesis.Survivin is highly expressed in embryonic and fetal tissues as well as in human malignancies (Li, 2005;Li and Brattain, 2006), while being almost undetectable in most terminally differentiated normal cells.Furthermore, survivin appears to be localized in different subcellular compartments: in nucleus, in cytoplasm, or as combined immuno-histochemical positivity in both nucleus and cytoplasm (Brennan et al., 2008;Oh et al., 2009).Due to large quantitative differences in the degree of survivin expression in normal adult tissues and in corresponding malignant tumors as well as different subcellular compartmentalization, survivin appears to represent a promising tumor biomarker and prognostic factor (Piras et al., 2007;Ge et al., 2013).Basically, MMRP and survivin are known to be diametrically opposing signals able to direct the apoptotic pathways.Both MMRP and survivin represent powerful prognostic markers.Therefore, we analyze the immunohistochemical expression of these proteins in sporadic colorectal adenomas with respect to the degree of dysplastic changes in order to explore their relationship.Moreover, we correlate these proteins with clinicomorphological parameters.Both MMRP and survivin are involved in apoptotic cascades.Abnormalities in the regulation of apoptosis may influence the development of early and developed malignant tumors (Ioana et al., 2010;Yurgelun et al., 2012;Ge et al., 2013).Although main functions of survivin and MMRP are well described in numerous papers (Duffy et al., 2007;Brennan et al., 2008;Li, 2008;Sun et al., 2014), their relation is studied very rarely.To the best of our knowledge, the relation between MMRP and survivin has not yet been elucidated in dysplastic colon adenomas.

MATERIALS AND METHODS
Formalin-fixed paraffin-embedded tissue samples from 124 cases of sporadic colon adenomatous polyps were enrolled in the present study.Pathology reports from all patients were reviewed, and their age, sex, localization as well as size of polyps were recorded.The hematoxylin and eosinstained slides from each patient were then reviewed to confirm the degree of dysplasia.The cecum, ascending and transverse colon were regarded as the right or proximal colon, whereas the descending colon, sigmoid and rectum were referred to as the left or distal colon (Pino et al., 2009).Our adenoma group included 87 males (70.2%) and 37 females (29.8%) (average age of males was 60.9 ±10.1 years and the average age of females was 68.7 ±9.1 years).Of all of the male/female patients, the polyps were located in 33/13 (26.6% / 10.5%) cases in the right colon and in 54/24 (43.6% / 19.3%) cases in the left colon.Each paraffin block was cut into three-micrometer thick sections and subjected to immunohistochemical staining, three sections for each primary antibody using Thermo Scientific Microm HM430.
In order to achieve greater adherence of the tissue sections to the glass surface, we used Flex slides (Dako, Glostrup, Denmark), which had been baked for two hours in an oven at 59°C.The slides were then treated in a PT Link System (Dako).The endogenous peroxidase activity was quenched with 3% hydrogen peroxide for ten minutes.Immunohistochemical reactions for MMRP were performed using Flex monoclonal mouse anti-human MLH1 and MSH2 antibodies (Dako, Clone ES05 and Clone FE1, respectively), and Flex monoclonal rabbit anti-human PMS2 and MSH6 antibodies (Dako, Clone EP51 and Clone EP49, respectively).For survivin, the immunohistochemical staining was performed using monoclonal mouse anti-survivin antibody (Dako, Clone 12C4, dilution 1:50).For MMRP immunoreactions, the sections were incubated for 20 minutes with the primary antibody at room temperature, and sections for MSH2 reaction www.anatclinar.com.ar were treated by Linker / Mouse for 20 min.The MLH1, MSH2, PMS2 and MSH6 proteins were visualized by means of the EnVision TM Flex / HRP System (Dako) using 3.3´ -diaminobenzidine (DAB) chromogen as substrate.After a one hour-long incubation with the primary antibody and Linker / Mouse treatment for 20 minutes, survivin was visualized by means of the EnVision TM Flex / HRP System using 3-amino-9-ethylcarbazole (AEC) chromogen as substrate, according to the manufacturer´s instructions.All sections were counterstained with Mayer´s hematoxylin (Dako).Negative controls were obtained by omitting the primary antibody.
Analyzed data: The expression and staining intensity of mismatch repair proteins and antiapoptotic protein survivin in 124 cases of sporadic colorectal adenomas.

MLH1:
In evaluation of absent cases, majority of them (10/13 cases, 76.9%) were found in adenoma size <10 mm (Table 2).MLH1 expression significantly correlated with dysplasia grade (p=0.019).High-grade adenomas expressed MLH1 in 96% of cases, low grade adenomas only in 83% of cases.Intensity of immunoreactivity of MLH1 significantly increased with the increasing tumor size and higher dysplasia grade (p=0.0001,p<0.0001 respectively).Interestingly, proximally located adenomas were statistically associated with stronger immunoreaction intensity (p=0.005)(Figure 2a).The comparison between MSH2 positive and negative adenomas did not reveal any significant relation to the observed clinicopathological parameters (p>0.05).However, the intensity of immunoreactivity in MSH2 positive samples significantly increased with an increasing tumor size and higher dysplasia grade (p=0.002,p=0.001 respectively).Additionally, the intensity of immunoreactivity was significantly higher in adenomatous polyps with proximal localization (p=0.006)(Figure 2b).MSH6: MSH6 protein was absent in 6 cases, 4/6 cases (66.7%) were found in adenoma size <10 mm (Table 4).MSH6 expression showed similar characteristics as MSH2.We confirmed a relation between MSH6 expression and age (p=0.007),but the other clinicopathological parameters did not reveal any significant relation (p>0.05).Nonetheless, the intensity of immunoreactivity in MSH6 positive samples significantly increased with the increasing tumor size and higher dysplasia grade (p<0.0001,p=0.0003, respectively).PMS2: Majority of absent cases (16/19, 84.2%) were detected in adenomas <10 mm (Table 5).PMS2 expression significantly correlated with dysplasia grade (p=0.018).PMS2 positivity was increased with higher dysplasia grade.Moreover, the intensity of immunoreactivity in PMS2 positive samples also significantly increased with the increasing tumor size and higher dysplasia grade (p=0.008,p<0.0001, respectively).Also, while 66% of adenomatous polyps with proximal localization demonstrated strong intensity of immunoreactivity, it was only 43% of polyps with distal localization that demonstrated the same (p=0.017).We confirmed a significant relation between survivin expression and all mismatch proteins (p<0.05).Subsequent analysis revealed that adenomatous polyps in which survivin intensity was lower than the intensity of mismatch protein -MLH1, MSH2, MSH6 and PMS2 formed the following percentages of all samples: 64%, 72%, 71% and 52%, respectively.Adenomas with identical intensity of survivin and mismatch protein -MLH1, MSH2, MSH6 and PMS2 formed 24%, 21%, 24% and 29%, respectively.and Bapat, 2003).Loss of MMR function may accelerate the development and accumulation of mutations in those genes, which are responsible for controlling cell growth.This fact provides support to a reasonable hypothesis for a rapid enlargement of colon adenomatous polyps and their progressive structural transformation to carcinomas (Molaei et al., 2011).

Relation between survivin and mismatch proteins (Table 6).
In our series, we found a vast majority of MMRP absent cases in adenomas <10 mm.In general, size of adenoma is considered to be a valuable prognostic marker, since large adenomas >10 mm are associated with worse histomorphological features (Toll et al., 2011).On the contrary, Sheridan et al. (2006) suggest that small sized sessile serrated adenomas (SSA) may develop into carcinomas despite their relatively small size.Molecular abnormalities develop through MSI (microsatellite instability).Decreased or absent immunostaining for MMRP was described in many SSA (Lee et al., 2005;Oh et al., 2005).Recently, several studies demonstrated that immunohistochemical detection of abnormal expression of MMRP is capable of identifying defective MMR genes.Molecular testing of MSI status consists of polymerase chain reaction and gel electrophoresis to examine the DNA sequences (Khoo et al., 2013).
There is an excellent correlation between immunohistochemical results and MSI analysis.
Both of these approaches are recommended to diagnose the abnormal status of MMR (Lanza et al., 2011;Khoo et al., 2013).Based on our recent results, we point out that loss of MMRP expression is also related to small sized sporadic adenomas <10 mm, and that increased risk of malignancy should be taken into consideration in these colon lesions.Data regarding the expression absence of MMRP in large vs. small sporadic colorectal adenomas is scarce.Most papers describe and study these proteins in adenomas associated with Lynch Syndrome (Halvarsson et al., 2005;Pino et al., 2009;Walsh et al., 2012).Furthermore, in our adenoma group, absent cases were mainly associated with low grade dysplasia.The degree of dysplastic changes is an important histomorphological and prognostic parameter.High grade dysplasia is the strongest predictor for the development of malignant tumor.Interestingly enough, we found a higher prevalence of abnormal staining for four MMRP in low grade adenoma cases in comparison to high grade cases.Currently, there is a heated discussion within literature concerning the expression of MMRP and severity of dysplasia.Some authors concluded that loss of MMRP www.anatclinar.com.arexpression is also detected in the absence of high grade dysplasia (De Jong et al., 2004).Contrarily, significant correlation was found in adenomas between MSI and high grade dysplasia (Iinoa et al., 2000).Pino et al. (2009) observed significant association between absent immunohistochemical staining of MMRP and high grade dysplasia.Another study by Walsh et al. (2012) (Hawkins and Ward, 2001).We found loss of MLH1 protein in 13/124 cases (10.5%).Typically, all absent MLH1 cases were accompanied by loss of immunohistochemical positivity for PMS2.PMS2 protein is probably unstable without its heterodimer twin (Young et al., 2002).The absence of more than one MMRP demonstrates progression via MSI pathway, and this pattern may suggest progressive transformation through adenoma-dysplasia-carcinoma sequence in a portion of assessed cases (Sheridan et al., 2006).Surprisingly, Nakagawa et al. (2001) concluded that normal colonic mucosa may also represent a possible precursor lesion by the spread of MLH1 promoter methylation with subsequent development of sporadic MSI+ colorectal cancer.Methylation of MLH1 promoter was also presented by Kuniyasu et al. (2004) in hyperplastic mucosa adjacent to colon cancer in athymic mice.Thus, our findings indicate that expression abnormalities in MMRP system may play a critical role in the early stages of development of premalignant and malignant colon lesions.
In the present study, distally located sporadic adenomas displayed loss of immunostaining for four MMRP more frequently.Our results are in concordance with previous observations (Pino et al., 2009).Several research groups revealed that increased risk of malignant transformation of sporadic colorectal adenomas is higher in distal colon than in proximal colon (Nusko et al., 1997;Chaves et al., 2000;Rijcken et al., 2002).By contrast, in patients with HNPCC, abnormal MMRP immunostaining was detected frequently in adenomas of proximal colon (Samowitz et al., 2001;Rijcken et al., 2002) (Mesri et al., 2001).Further, proliferative fenotype of survivin is associated with poor prognostic histomorphological parameters, such as vascular invasion and tumor grade 3 (Adamkov et al., 2012).As expected, the presence of survivin expression significantly correlated with adenoma size and degree of dysplasia, statistically significant differences were observed between subcellular localization of survivin and adenoma size, dysplasia grade as well as its localization.Furthermore, there is a significant trend between the intensity of survivin immunoreaction and dysplasia grade and localization.These recent results suggest that survivin expression pattern in colon adenomas is also associated with worse prognostic features.In general, there is a higher incidence of focal malignant changes in larger adenomas and in HGD adenomas (Muto et al., 1975).This may prove a definitive relationship between adenomas and colorectal carcinomas (the adenoma-carcinoma sequence) (Talbot et al., 2006).Briefly, MMRP are responsible for the correction of mutations during DNA replication, and the antiapoptotic survivin is an ideal protein for the development of premalignant and malignant lesions.Our analysis revealed a significant relation between the expression of MMRP and survivin (Table 6).Likewise, the evaluation of intensity of immunoreaction also resulted in interesting findings.As described in our recent results, a majority of adenomatous polyps demonstrated lower intensity of survivin immunowww.anatclinar.com.arreaction as opposed to the intensity of MMRP.In addition, some percentage of cases revealed identical immunoreaction intensity for both MMRP and survivin.All of this may slightly uncover the relation of these proteins.MMRP MLH1, MSH2, and PMS2 are also linked with apoptotic cascade via p53 or its homologue p73 (Stojic et al., 2004), e.g.p73 protein is directly stabilized by PMS2 protein and this interaction enhances its proapoptotic function (Shimodaira et al., 2003).Other studies, such as Luo et al., (2004) and Hassen et al., (2012), described the key role of MLH1 and PMS1/PMS2 heterodimerization.Nuclear accumulation of this complex may increase the activation of p53 by ataxia-teleangiectasia mutated (ATM) protein kinase.Köster et al. (2007) reported a significant correlation between immunohistochemical expression for MSH2 and p53, and apoptosis in cervical carcinoma.Zhang et al. (1999) showed that overexpression of MSH2 and MLH1 may induce apoptosis and that MSH2-deficient cells do not display apoptotic features.The expression of functional activity of p53 may regulate the expression of MSH2, since a binding site for p53 was discovered in the promoter region of the MSH2 gene (Scherer et al., 1996a;Scherer et al., 1996b).Thus, it seems that the common denominator between MMRP and apoptosis is the p53 protein.The p53 protein can either activate the apoptotic cascade by up-regulation of several associated genes or suppress those genes with antiapoptotic functions.Survivin is known to be suppressed by wild type p53, because it interacts with the survivin promoter, proven to be the first promoter to confer p53dependent repression (Hoffman et al., 2002;Mirza et al., 2002).In addition, p53 may interfere with bcl-2 proteins in mitochondria with subsequent release of cytochrome c (Chipuk et al., 2004) and formation of proapoptotic multiprotein complex apoptosome.Köster et al. (2007) indicate that MMRP system may be more active in the early stages of cancer development.The presence of dysplastic changes in colon adenomas provides evidence for their malignant potential.In our group of colon adenomas, we detected moderate and strong intensity of immunoreaction for MMRP in a majority of positive cases.Additionally, we revealed a statistically significant relation between immunoreactivity of MMRP and survivin.Increasing intensity of MMRP immunoreactivity was accompanied by decreasing intensity of survivin immunoreactivity.Taking into account our current study results as well as the above-mentioned interactions between the proteins in question we suggest that MMRP may influence the antiapoptotic function of survivin by indirect mechanism via activation of p53 in LGD and HGD colon adenomas.

Figure 1b -
Figure 1b -Strong intensity of survivin immunoreaction in HGD distally located colon adenoma

Table 4 -
Relationship between MSH6 expression and clinicomorphological parameters in colon adenomas.+ weak intensity, ++ / +++ moderate / strong intensity Based on the Cochran-Armitage test for trend (Table1), we confirmed a statistically significant trend between survivin intensity of immunoreactivity (weak/strong) (Figures