Introduction
Colorectal cancer (CRC) is the second leading cause of cancer-related deaths in the United States, despite recent improvements in the diagnosis and treatment of the disease [
1]. The incidence and mortality rate of CRC are higher in African Americans than in the general population [
1,
2]. Many epidemiological and genetic investigations have focused on African Americans [
2‐
4] with the goal of deciphering the reasons for such disparities. While socio-economic factors are likely involved (e.g., African Americans tend to reach more advanced stages of disease before diagnosis), biological factors also contribute to the disparity [
5]. Dietary and environmental factors certainly play an important role, as Japanese immigrants to the US show higher colon cancer rates than do Japanese people living in Japan. The latter also started showing higher rates of colon cancer along with the westernization of their diets [
6,
7].
The recent approval by the FDA of Bidil, a drug for heart failure in African Americans, illustrates ethnical/racial genetic specificities that lead to different interactions with a given drug or molecule [
8]. Only limited genetic and epigenetic studies exist from Africa and other continents, including Asia, that have populations known to be of African descent. In Oman, some tribes have Bedouin roots and others have Asian and/or African origins [
9]. In general, the Omani population has Asian and African ethnicities [
10]. Other ethnic groups in Asia are not considered to be of African descent, such as Caucasian people in Iran. In this study, we used samples from patients treated for CRC in Oman, Iran and the US (African Americans) for genetic and epigenetic investigation. We chose these three populations based on their similarities and differences in African ethnicity [
10‐
14].
Colon cancer develops through different pathways, all involving changes at the chromosomal or genetic levels. Other modifications occur epigenetically and affect the level of expression of certain targeted genes that are essential for the normal control of cell division within the colon mucosa [
15]. It is now widely accepted that sporadic colorectal cancers frequently arise from pre-neoplastic lesions through the activation of oncogenes (K-ras, BRAF) and the inactivation of tumor suppressor genes (APC, p53, DCC) and mismatch repair (MMR) genes (MLH1 and MSH2) [
16,
17]. Also, activating mutations in BRAF, one of the RAF genes which encode kinases that are regulated by Ras and mediate cellular responses to growth signals, have been found to be associated with MSI-H cancers [
18]. The DNA MMR process can be impeded by genes that are either mutated [
19] or silenced [
4], leading to the expression of non-functional proteins or to a lack of expression, respectively. MMR pathway is primarily responsible for colon cancer development in families with Hereditary Non Polyposis Colon Cancer (HNPCC), which represents about 5% of the generally inventoried colon cancer cases [
20].
In the present study, we analyzed samples from two different populations of African descent (African American and Omani) and a Caucasian population (Iranian) for microsatellite instability and BRAF mutations. We correlated our findings with the silencing of genes involved in DNA MMR. We evaluated and compared the demographic and clinicopathological data in relation to molecular alterations in specific genes (BRAF, DNA MMR) to improve our understanding of colon tumorgenesis in people living in different environments.
Discussion
DNA MMR gene silencing and/or mutation are the main factors leading to accumulation of mutations within genes and general genome instability [
19]. The MSI phenotype resulting from such alterations is one of the early events leading to the development of certain cancers. It occurs in patients with Lynch syndrome and in those that acquire such alterations somatically [
24]. Another event occurring in adenoma polyp is mutation in oncogenes, specifically in K-ras and BRAF genes in the case of colon cancer [
24‐
26]. The linkage between BRAF mutations and MSI was previously reported [
27‐
30]. In general, Omani people have a mixed Asian and African ethnic background [
10] and Iranians are of Caucasian ethnicity. Here we demonstrate that in African Americans, Omanis and Iranians, the MSI phenotype occurs in 31%, 26% and 13% of CRC tumors, respectively. In addition, the locations of the tumors in African Americans and Iranians were primarily proximal compared to distal for Omanis; and BRAF mutations were more prevalent in Omanis than African Americans or Iranians (19%, 10% and 2%, respectively). While the MSI phenotype in this sample of African Americans is less than we previously reported (43–45%), [
3,
4] it appears to be more prevalent in African Americans than in others and still high in the general US population [
31‐
33]. The current sample contained more African Americans and more right-sided than left-sided tumors. The MSI phenotype was more common in proximal tumors than in distal ones (p = 0.47), strengthening earlier findings of such an association [
31‐
33]. The prevalence of MSI-H tumors indicates the importance of BRAF mutations, particularly in the African Americans, since they represent a higher mutation profile than Iranians and Omanis. This finding suggests that, at least within the MSI-H group of patients, the BRAF V600E mutation in exon 15 is the major event leading to tumor development. In addition, the distal location of Omani MSI-H tumors may also play an important role in the presentation of the disease compared to proximal ones in the African American and Iranian populations. The lower frequency of MSI-H tumors observed in Omani patients could be related to the higher frequency of the distal location. This needs to be verified in a larger sample size.
The absence of hMLH1 expression was more pronounced in MSI-H tumors from African American and Iranian patients (77% of tumors in these groups showed no hMLH1 expression). However, for Omanis, the moderate hMLH1expression (62%) in MSI-H may indicate some other defects in the component of base excision repair (Oxidative DNA damaged markers such as 8-hydroxyguanosine), or distinct lymphangiogenic phenotype associated with MSI other than in mismatch repair genes. The tumors that we have studied included MSI-H tumors that were negative using immunohistochemistry for hMLH1. This incidence is less than compared to that described in the literature (77% [
34]), although the numbers described are small.
These results are in concordance with the MSP analysis showing that a majority of samples displayed methylation at the hMLH1 promoter. The methylation of hMLH1 is most likely from a specific genome methylation process in progression of colon cancer. The African American group was older than both the Omani and Iranian group. It has been demonstrated that older people, particularly those in the 7th–9th decade of life, have a much higher chance of hypermethylation of hMLH1 and MSI. This is consistent with hMLH1 methylation in the African American cancers both in MSI-H and non-MSI cases. However, in Iranians, a higher profile of epigenetic silencing of hMLH1 may explain the MSI-H tumors but not in non-MSI. In addition, the distal location of Omani tumors may also play an important role in the presentation of the disease compared to the African American who has generally proximal CRC.
Our results demonstrate that methylation of hMLH1 is the major cause of MSI in sporadic CRC consistent with our previous findings [
3,
4]. The fact that the hMLH1 gene is methylated indicates that it may be inactivated by an epigenetic mechanism. The association of higher levels of CpG island methylation with more advanced histological changes suggests that CpG methylation plays a role in CRC [
30]. However, the pathophysiology of hyper-methylation (the why, when and where) has yet to be elucidated. Cancers can be classified according to their degree of methylation. Those with high degrees of methylation (the CpG island methylator phenotype, or CIMP) represent a clinically and etiologically distinct group that is characterized by 'epigenetic instability'. The MSI-H and CIMP phenotype may explain the criteria in proximal tumors in African American and Iranian but not the distal ones in Omani tumors; however, the CpG island methylation status of a broader panel of genes needs to be investigated to determine whether this is the case. Some samples for MLH1 methylation failed to display MLH1 protein by immunohistochemistry and were non-MSI (MSS and MSI-L). Possible explanations might be that the targeted CpG island for methylation in this study is located upstream of the gene, and has a minor effect on the transcription of MLH1 protein [
35] or that the detected methylation is only in a small cell population that does not reflect the overall tumor phenotype. Full methylation of the hMLH1 promoter region and subsequent gene inactivation may play a crucial role in carcinogenesis of MSI-H CRCs. Therefore, in our on-going investigation, we are studying the methylation status of the lesions by examining all CpG sites [
36] especially within non-MSI tumors where data reveals lack of protein expression. Some non-MSI tumors were methylated at the MLH1 site and this may be due to the partial or hemi-methylation at an altered MLH1 site. This is consistent with the lack of correlation of between the methylation status and level of expression of MLH1[
37]. Indeed, more markers need to be considered to measure methylation in order to establish a methylation phenotype that correlates the protein expression status and allows for the understanding of preferential carcinogenic pathways. Positive staining was confirmed in adjacent normal tissue within the same slide validating of the staining for specimens that were found to be negative for either MLH1 or MSH2. In addition, preferential microsatellite loci containing large repeat units, but not loci containing mono- or dinucleotide repeats units, may contribute to the non-MSI tumors, particularly MSI-L CRC tumors.
Many explanations could account for the differences between African Americans and the other populations. The difference in MSI frequency might be due to genetic specificities or to behavioral/dietary causes [
38]. Dietary factors such as folate, vitamins, and methionine may be associated with colon cancer because of involvement in DNA methylation and hence on the CpG island, MSI and BRAF. A recent study questioned the unique role for dietary folate, alcohol, vitamins B
6 and B
12 and methionine in the CIMP phenotype [
39]. In addition, use of alcohol and obesity were associated with an increased risk of tumors that were MSI-H and CIMP-low (less than two markers methylated) [
39]. The number of lymph nodes, distance organ metastases and additional impairment of the MMR system may associate with the more aggressive behavior of CRC in African Americans.
Studies have investigated the traditional (nonserrated) adenoma-carcinoma sequence and the serrated polyp neoplasia for BRAF and MSI. MSI-H was identified only in the adenocarcinoma component of serrated carcinomas [
40,
41]. BRAF mutation has been shown to be a specific marker for a serrated polyp pathway that has its origin in a hyperplastic polyp and a potential end point as MSI carcinoma. A recent study indicated that MSI-H in the sporadic colorectal cancers may be part of a clinically distinct subgroup with a high incidence of BRAF mutation developed from serrated polyps [
26]. In this study we categorized the histological status of the end point such as carcinomas with residual adenoma, the serrated polyp neoplasia pathway and the traditional (nonserrated) adenoma-carcinoma sequence. However, we did not find any evidence of serrated or serrated hyperplastic tissue in the MSI-H tumors. Therefore, our BRAF and MSI-H analyses indicate that tumors may not belong to the serrated pathway, which needs additional investigation. Most hyperplastic polyps occur in the rectum and sigmoid colon, but most serrated (hyperplastic) polyps occur in the right colon and are associated with proximal cancers [
40,
41]. All but one (13%) Omani, 79% Iranian, 66% of African American MSI-H tumors were proximal with no evidence of serrated hyperplastic polyps. Tumors in these populations were moderately differentiated with mixed adenomas. Therefore, at this time it is not clear whether or not mixed adenoma types have any role in the MSI-H in African Americans.
The population of the Sultanate of Oman is especially interesting because it represents a combination of African and Asian heritage in a small country with the distribution of CRC disease in young people, which may be due to the age structure of the population in Oman. The effect of environmental factors such as diet, physical activity, access to health care, frequent tribal marriage and lifestyle can not be ruled out for the occurrence of the CRC in young age [
9,
10,
42,
43]. The Iranian population is a large collection of ethnic groups and their descent were from ancient Iranian peoples[
44]. Modern Persians themselves are also a heterogeneous group of peoples descended from various Caucasian peoples [
14,
45]. The younger age distribution of Iranian CRC disease also may be due to the age structure of the population. However, the effect of environmental factors such as diet and lifestyle can not be ruled out. Our data on CRC lead us to speculate the complex interactions of genotype, environmental factors (such as diet), and other lifestyle factors in the pattern of bowel cancer. Further epidemiology data to compare the age-specific incident rates for Oman, Iran and African American is needed to confirm this finding.
The proximal location of tumors in African Americans is consistent with the tumors in Western and Asian studies [
31‐
33,
46,
47] which may reflect the impact of western diets on the African Americans. However, the high MSI level for African Americans for this limited sample size may in part be due to the increased age of these patients along with finding the majority of the tumors on the right side of the colon. The Iranian tumors closely mimic what we see in Caucasians in the U.S., most tumors are left sided, most of the MSI-H tumors are right sided and the MSI-H tumors are associated with higher tumor differentiation. The distal location of the Omani tumors and the low MSI level is an important observation compared to the moderate and high proximal MSI level in Iranian and African American tumors, respectively. The alteration of K-RAS may also contribute to the methylator phenotype in CRC, especially in the Iranian tumors, which we plan to study in the future [
30]. The rate of mutation and its level in MMR deficiency may contribute and determine the frameshift rate for the loss of these proteins in CRC, hence different behavior from these tumors.
All of these factors may contribute to the risk of colon cancer and presentation of the disease in African American, Omani and Iranian patients. We cannot rule out a genetic predisposition in these patients since we did not have access to detailed family history and, therefore, some of the subjects may have had genetic and epigenetic predispositions. In this study we did not use other groups such as white Americans, other Africans and Asian in our controls for any direct or indirect comparison analysis. We plan to have a more detailed study regarding the diversity of these populations since the scope of this study focused on the clinicopathological features.
In conclusion, this first comparative clinicopathological investigation in three different populations suggests that the MSI-H CRC phenotype in African American, Iranian and Omani patients is significantly associated with BRAF mutation and hMLH1 expression. CRCs in African Americans tend to be higher in microsatellite instability (32%) and more often located in the proximal colon, compared to Iranian and Omani CRC tumors. The high level of MSI-H in African Americans may have significant implications in treatment plans, because MSI-H lesions are often right-sided and may show a different response to chemotherapeutic agents such as 5-fluorourcil [
48].
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
HB carried out the MSI study. PM, FN, and MS recruited Iranian samples and run MSI and immunohistochemistry. MA, KAl-M, RAl-M, and AAl-S, and SR recruited Omani samples from different hospital in Oman and run methylation analysis. FG recruited samples from Johns Hopkins and performed the H&E staining. DTS help in the sample recruitment and reviewing the data. from Howard university. AV, RCB and AG performed run the BRAF mutation analysis. EL, MH performed the pathological analysis, MN run the statistics. HA designed the experiment and data analysis. All authors read and approved the final manuscript.