ERK/pERK expression and B-raf mutations in colon adenocarcinomas: correlation with clinicopathological characteristics

Colon cancer is the fourth most common malignancy in terms of both incidence and mortality worldwide [1]. The development of colon cancer is a complex multistep process dependent on both genetic and environmental factors, in which oncogenes, tumor suppressor genes as well as genes involved in DNA damage recognition and repair have been implicated. Gaining insight into the molecular pathways involved in the progression of colon cancer is imperative for the development of innovative individualised cancer treatment strategies.

Colorectal (CRC) carcinogenesis through various morphological stages has been linked to several genetic and epigenetic changes. Until recently, two main pathways of sporadic colorectal carcinogenesis were proposed: a) the chromosomal instability (CIN) pathway which affects proto-oncogenes and tumor suppressor genes and is characterized by alterations of chromosomal number and structure, and b) the microsatellite instability (MSI) pathway, responsible also for Lynch syndrome CRCs, which features size variations of repeated nucleotides (microsatellites) mainly in non coding sequences, due to defects in the mismatch repair system (hMLH1, hMSH2, mHMS6, PMS1, PMS2) [2, 3]. In addition to CIN and MSI pathways, a third pathway, the epigenetic instability, which is thought to be largely driven by hypermethylation-induced silencing of tumor suppressor-like genes, has been implicated in the progression of colorectal carcinogenesis [4]. According to this notion, contemporary literature suggests that CRC in general develops through two independent pathways that involve sequences of genetic and epigenetic alterations associated with pathological and clinical features: the adenoma pathway in 70-80% and the newly recognized, the serrated pathway in the remaining 20-30% [5]. The somatic molecular features which characterize the newly introduced serrated pathway to CRC include activating mutations in B-raf[6] and widespread hypermethylation of gene promoters (CIMP) [7] with or without MSI [6].

The kinases of mitogen-activated protein (MAP) kinase superfamily participate in signaling cascades conserved through evolution, which transduce extracellular signals into intracellular responses. MAP kinases are major components of pathways controlling embryogenesis, cell differentiation, proliferation and death. The Ras/Raf/mitogen-extracellular signal-regulated kinase 1/2 (MEK1/2)/extracellular signal regulated kinase 1/2 (ERK1/2) cascade is activated by mitogenic factors, differentiation stimuli and cytokines [8, 9]. The Raf family of protein kinases, which is one class of Ras effectors, phosphorylates the dual specific MAP kinases MEK1 and MEK2, which in turn phosphorylate and activate the effector MAP kinases ERK1 and ERK2 [9]. ERKs are multifunctional serine/threonine kinases that target a vast array of substrates localized in all cellular compartments, such as protein kinases, signaling effectors, receptors, cytoskeletal and nuclear proteins and transcription factors, that can influence cell fate [9‐11].

Importantly, MAP kinases are capable of affecting gene expression via intermediary kinases by phosphorylating proteins in the cytoplasm, but also translocate to the nucleus, a critical step for the fulfilment of many cellular functions of ERK, such as gene transcription, cell proliferation and differentiation [12]. Through phosphorylation of these various substrates, constitutively activated ERKs are able to influence many of the hallmarks of carcinogenesis, as defined by Hanahan and Weinberg [13]. Constitutive activation of this pathway has been observed in several human malignancies and cell lines such as breast, colon, thyroid carcinomas and melanomas and provides a potent promitogenic force resulting in uncontrolled proliferation and differentiation [14, 15].

The current study investigates the presence of mutations in K-ras and B-raf genes in colorectal carcinoma, in correlation with MAP kinase ERK expression and the expression of mismatch repair proteins (MMR) hMLH1 and hMSH2, attempting to elucidate the involvement of these MAP kinases in the development of colorectal cancer, as well as their correlations with standard clinicopathological parameters.

Various molecular markers have been proposed for the classification of colon cancer with regard to clinical course and outcome, including cell cycle as well as apoptotic regulators (c-FLIP, Ki67, COX-2) [22, 23]. Deregulation of the MAPK signalling pathway has often been associated with oncogenic transformation [24]. In this regard, there is accumulating evidence involving ERK activation in the tumorigenesis of various human cancers such as prostate, breast, colon and ovary [25‐29]. Furthermore, activating B-raf mutations are frequently detected in different tumor types such as melanomas, thyroid and colon carcinomas implying their significance as potential targets for anticancer treatment [30, 31]. In the present study we used immunohistochemistry to examine the expression of total and activated ERK1/2 in a panel of 94 colorectal carcinomas in parallel with the expression of two MMR proteins (hMLh1 and hMSH2), as well as K-ras and B-raf mutations, which may lead to constitutive activation of MAPK pathway.

To the best of our knowledge, this appears to be the first study examining ERK expression in mismatch repair deficient and mismatch repair proficient colorectal cancer simultaneously with upstream gene alterations such as K-ras and B-raf mutations, which contribute to MAPK activation. In our cohort, ERK immunoreactivity was observed in the cytoplasm as well as in the nucleus in the majority of the cases (93% and 91% respectively). More than 78% of the examined cases displayed concurrent ERK nuclear and cytoplasmic immunoreactivity, in concordance with previous observations [29, 30]. Increased levels of nuclear ERK positivity correlated with strong cytoplasmic ERK expression. The observed correlation between nuclear and cytoplasmic expression as well as the concurrent nuclear and cytoplasmic immunolocalization of ERK in the present series is compatible with its multiple functions targeting proteins localized in the cytoplasm as well as the nucleus of the cell. Moreover, pERK immunoreactivity was predominantly nuclear (68%) whereas cytoplasmic positivity was found in a subset of tumors (24%). This staining pattern is in line with previous investigations in colon as well as other neoplasms, namely, endometrial, head and neck tumors and melanomas [27, 29, 32, 33]. Additionally, total nuclear or cytoplasmic ERK staining was independent of nuclear or cytoplasmic pERK status, in our cohort. This observation is in agreement with previous studies in non small cell lung and endometrial cancer as well as in melanomas and could been explained by the hypothesis that pERK immunopositivity could arise due to ERK hyperactivation rather than overexpression [29, 30, 32].

Furthermore we show that high nuclear, but not cytoplasmic, ERK immunopositivity is correlated with tumor stage (Table 1). Interestingly, our results demonstrate that nuclear pERK expression parallels tumor grade and tumor stage (Table 2, Figure 2) in keeping with findings of previous studies in NSCLC and prostate cancer [30, 34]. This observation is in contrast with a previous investigation [25] in which pERK expression was not correlated with tumor stage, whereas there are other studies in various other tumors (breast, pancreatic, endometrial and ovarian cancer) that have failed to substantiate a positive correlation between pERK expression and classical clinicopathological parameters [26, 32, 35‐37]. Taking into account that in our study normal colorectal mucosa displays minimal nuclear pERK expression, our results speak in favor of the potential role of this molecule in tumor evolution as well as, in the acquisition of a more aggressive phenotype in colorectal carcinogenesis. Along this line, it has been suggested that constitutively active ERKs are capable of affecting gene expression, being able to influence many of the hallmarks of carcinogenesis [35]. It could be hypothesised that as colorectal cancer progresses to a more advanced disease, an increase in the activation of the MAP kinase signal transduction pathway occurs [34]. Unfortunately, the statistical power of the present investigation is reduced due to the small cohort of pERK positive cases. Further studies of larger cohorts are warranted to confirm our results. Along this line, pERK expression has been shown to be associated with poor prognosis in colorectal carcinomas [27]. It has been speculated that the activation of the MAPK pathway initiates cellular processes, which could result in either favorable or worse clinical outcome, a fact due to the complex signal transduction network ERKs are involved in. In fact, adverse effects can be observed, depending on the intensity and duration of the promitogenic signal [10].

The aforementioned findings are quite significant in view of the fact that protein kinases represent a group of molecular targets characterized by a cancer specific potential, allowing the development of new generation chemotherapeutic agents acting as kinase inhibitors. The first oral multi-kinase inhibitor that targets Raf kinases have already been approved for the treatment of renal cell cancer, whereas these factors display a broad spectrum antitumor activity in colon, breast and non-small-cell lung cancer in xenograft models and also hepatocellular carcinoma and sarcoma [38, 39]. In this context, the presence of B-raf mutations has been suggested as a possible surrogate marker of sensitivity to those drugs which target the ERK pathway at the level of Raf kinase [38, 39].

The frequency of K-ras mutations detected in this study (about 23% of the examined cases) was comparable to that found in previous reports (Table 4) [39‐41]. Furthermore, the presence of B-raf mutations in colorectal cancer is estimated to be about 10% of unselected colorectal cancers [42‐44]. In particular, B-raf T1799A mutation (V600E) has been reported in 4% of microsatellite stable (MSS) tumors whereas in microsatellite unstable tumors the percentage rises up to 27-52% [42, 44‐46]. By analogy to those investigations we detected V600E B-raf mutations in about 7% of MSS tumors and in 21% of MSI unstable tumors. Interestingly, our cases exhibited only either B-raf or K-ras mutations in accordance with previous observations suggesting that these are mutually exclusive defects that probably exert equivalent effects in tumorigenesis [39, 46].

Moreover, in our series the expression levels of total and activated ERK1/2 were independent of the mutation status of B-raf and K-ras genes. These results are in favor of the view that constitutive pERK activation occurs in a K-ras or B-raf -independent manner in a large subset of primary colon cancer cases. Recently, several negative regulators of the MAPK signalling pathway upstream of ERK at the level of Raf were identified, including Sprouty and Spred. Activation of these negative regulators inhibits phosphorylation of ERK1/2, even in the presence of mutation in K-ras gene [32, 47]. This finding has also previously been observed in ulcerative colitis-related carcinomas [48]. The genetic nature of constitutive activation of the RAS/RAF/MEK/ERK pathway in colorectal tumors with no B-raf or K-ras mutation remains unknown, although it may in part due to increased activity of growth factor receptor induced cell proliferation pathways. It could be speculated that in cancer constitutive activation of MAP kinase could be triggered by upstream oncogenic regulators due to the presence of paracrine/autocrine growth factor stimulation, rather than Ras or B-raf mutations or components of the various other signal transduction pathways that interact with MAPK, since the mutation of K-ras and B-raf obviously constitutes one of multiple ways to activate this pathway.

In sporadic colorectal carcinogenesis B-raf mutations like K-ras mutations appear to occur early at the transition from small to medium size adenoma and are extremely frequent in so-called serrated adenomas [49]. According to the MSI colorectal pathway, MSI in sporadic tumors has been suggested to be mostly due to hypermethylation of the promoters of MMR genes and is correlated with B-raf mutations (44). However, in our study B-raf mutations were not correlated with loss of hMLH1 or hMSH2 protein, suggesting that the B-raf mutated cases of our cohort may belong to more than one colorectal carcinogenesis pathways.

In this study, we have shown that mutations of the B-raf gene are not associated with mismatch-repair deficiency through loss of hMLH1 or hMSH2 expression. Disruption of the MAP kinase pathway-either through K-ras or B-raf mutation-was detected in 37% of all the examined cases, although the overexpression of total and activated ERK1/2 was not correlated with the mutational status of K-ras or B-Rraf genes. Interestingly, we present evidence that the expression of activated ERK expression parallels histological grade and stage in colon carcinomas, thus being implicated in tumor invasiveness as well as in the acquisition of a more aggressive phenotype. Our findings encourage prospective investigations to further elucidate and validate the potential role of pERK as a prognostic factor and therapeutic target in colon carcinogenesis.