The role of ERK1/2 activation in breast cancer prognosis has been studied extensively, with conflicting results. Experimental data from breast cancer cell lines show that MAP kinases, particularly ERK, are involved in cell proliferation and motility. However, clinical studies on human tumor tissues reveal mixed findings. Some studies suggest that high ERK1/2 activity correlates with a poor prognosis, while others indicate a favorable outcome. For instance, phosphorylated ERK1/2 levels in breast tumors were associated with longer recurrence-free survival in some cases but linked to aggressive disease in others. Technical variations, such as antibody specificity, sample processing, and study design, contribute to these discrepancies. Additionally, the localization and activation state of ERK1/2 influence their prognostic significance. ERK1/2 phosphorylation typically translocates to the nucleus, affecting transcription factors like AP-1, but this process may differ in clinical contexts. Overall, while ERK1/2 activation appears to play a role in breast cancer progression, its exact prognostic relevance remains unclear, necessitating further research to clarify its role and optimize therapeutic targeting.
Experimental studies on breast cancer cell lines have demonstrated that MAP kinases play a role in normal alveolar morphogenesis, as shown by Wang et al. (2002). Additionally, high ERK activity has been linked to proliferation and motility in breast cancer cells, as observed in research by Seddighzadeh et al. (1999). However, clinical studies on human tumor tissues exploring ERK1/2 expression and activation remain limited.
Investigations into non-breast tumors have yielded mixed results regarding the prognostic significance of ERK expression. For instance, in salivary gland mucoepidermoid carcinomas, high phosphorylated ERK1/2 levels were associated with early progression, as reported by Handra-Luca et al. (2003). Conversely, ovarian cancer patients with high ERK1/2 and p-ERK1/2 expression in pleural effusions exhibited better overall survival compared to those with lower levels, according to Givant-Horwitz et al. (2003). Interestingly, ERK expression and activity in these cases increased following chemotherapy. In prostate cancer, p-ERK1/2 staining intensity decreased with disease progression, as noted by Malik et al. (2002). In small cell lung carcinomas, tumors with cytoplasmic p-ERK expression had a better prognosis than p-ERK-negative cases, while nuclear immunoreactivity showed no prognostic value, as found by Blackhall et al. (2003).
Regarding breast cancers, early studies suggest a role for activated MAP kinases in carcinogenesis and progression. Sivaraman et al. (1997) observed increased MAPK activity in cytosols from 12 mammary carcinomas compared to normal tissues or benign lesions. Salh et al. (1999) reported elevated ERK1 and ERK2 amounts in immunoprecipitation studies, though ERK2 staining intensity in immunohistochemistry was reduced in tumor cells, potentially due to the lower percentage of epithelial cells in normal tissue samples. Our IHC and Western blot data reveal significant variations between tumors, with decreased p-ERK1/2 expression in many carcinomas and overexpression in others.
In a separate study by Mueller et al. (2000), MAPK activity in cytosols from 131 mammary carcinomas was measured using an enzymatic assay and found to correlate with nodal involvement and higher relapse risk, though these associations were not statistically significant. This contrasts with our findings, where high activated ERK1/2 expression significantly predicts favorable recurrence-free survival. The discrepancy may arise from differences in sample processing: Mueller et al. used mild lysis buffers, which may not effectively measure nuclear proteins, whereas our method employed SDS-containing buffers to maximize protein extraction. Unphosphorylated MAP kinases are cytoplasmic, while phosphorylated forms translocate to the nucleus to regulate transcription, as Brunet et al. (1999) demonstrated. Thus, enzymatic assays for cytosolic extracts may only capture a portion of total p-ERK activity.
Our study is more comparable to two IHC studies concluding that activated ERK1/2 may influence tumor progression and prognosis. Adeyinka et al. (2002) used a polyclonal phospho-p44/p42 (Thr 202/Tyr 204) antibody, while Gee et al. (2001) employed a polyclonal anti-ACTIVE™ MAPK antibody targeting dually phosphorylated ERK1/2 forms. We utilized a monoclonal phospho-p44/p42 antibody from Cell Signaling Technology.
Despite differences in antibodies, variations in methodology and patient cohorts likely drive discrepancies. Adeyinka et al. (2002) found no significant endocrine sensitivity differences among 29 ER-positive, nodal-negative patients treated with tamoxifen, suggesting activated MAPK may not predict endocrine response. High p-MAPK positivity in nodal-positive tumors (90%) compared to mixed nodal cases (48%) led them to suggest p-MAPK as a poor prognosis marker, though this was not statistically confirmed. Gee et al. (2001) analyzed 90 carcinomas with 72% positivity, finding high p-ERK1/2 reactivity linked to poor antihormonal therapy response and reduced survival in ER-positive and all patients. Multivariate analysis highlighted a significant survival correlation with p-MAPK positivity in ER-positive cases. However, their cohort differed from ours, with 82% mortality versus 17.5%, complicating direct comparisons.
Our data uniquely show that high p-ERK1 expression is an independent, favorable prognostic indicator. Both IHC and Western blot methods consistently linked negative p-ERK staining to shorter recurrence-free survival, supporting our conclusion. For p-ERK2, only weak, nonsignificant correlations were observed, suggesting limited role in progression.
Western blot analysis of p-ERK1/2 allows differentiation between the two kinases but fails to distinguish tumor from non-malignant cells. Immunohistochemistry specifically measures tumor cell expression, which we confirmed by correlating results between methods. Our findings suggest p-ERK proteins detected by Western blot primarily originate from tumor cells. The correlation of p-ERK1/2 levels with recurrence-free survival appears driven by p-ERK1 expression, as p-ERK2 showed no statistical significance.
Interestingly, high p-ERK1 levels correlate with low MMP9 and high MMP1 expression, reflecting tumor and stromal cell contributions, as Brummer et al. (1999) and Dublin et al. (2000) noted. Prior studies in our cohort linked active MMP1 with ER positivity and negative nodal status, while high MMP9 levels correlated with nodal involvement (Milde-Langosch et al., 2004). Since MMP9 is a known negative prognosticator, its low levels in high p-ERK1 cases align with our favorable prognosis findings, though causality remains unproven.
No significant associations between p-ERK1/2 expression and cell-cycle regulators or Ki67 were found, suggesting minimal direct involvement in proliferation. Immunoblots show weaker expression of cell-cycle regulators in normal tissue versus carcinomas, ruling out nontumor cell influence. Experiments on MCF7 cells by Alblas et al. (1998) showed TPA-mediated proliferation inhibition accompanied by high ERK2 expression, supporting a growth-inhibiting ERK1/2 role in breast cancer.
Results for unphosphorylated ERK1 and ERK2 differ from their activated forms. No significant correlations with prognostic parameters or survival were found, but high ERK1 and ERK2 levels associated with cyclin D1, Rb, and p16, known tumor-suppressors, suggesting non-proliferative roles. ERK1’s correlation with PAI-1 expression hints at a potential role in tumor invasion regulation.
Comparing ERK1/2 proteins with downstream AP-1 transcription factors, c-Fos, Fra-1, Fra-2, and FosB levels are weak in normal tissue, reflecting tumor cell expression in our samples. Gruda et al. (1994) reported Fra-1 and Fra-2 stabilization post-MAPK phosphorylation, yet we found weak correlations between strong p-ERK1 and high FosB or low JunD and p-Fra-1 levels. Fra-1 and Fra-2 phosphorylation involves multiple kinases, not solely ERK1/2, as evidenced by differing mobility shifts, suggesting minimal ERK1/2 involvement in AP-1 regulation in vivo.
Our prior work showed FosB overexpression in differentiated, hormone receptor-positive tumors (Milde-Langosch et al., 2003). Our findings confirm that p-ERK overexpression associates with favorable prognosis. Nonphosphorylated ERK correlations with AP-1 proteins may reflect regulatory mechanisms rather than direct effects.
Despite in vitro evidence linking ERK1/2 activation to proliferation and motility, our clinical findings suggest the opposite, with high p-ERK1/2 correlating with better outcomes. This paradox may reflect tumor independence from extracellular signals during progression, as highly malignant tumors often lose hormone dependence. As Shen and Brown (2003) noted, targeting the MEK-ERK pathway may be ineffective in advanced cancers, underscoring the need for further research into alternative therapeutic strategies and larger studies to validate p-ERK proteins as prognostic indicators.
