There is great variability in the reported frequencies of triple-negative and basal-like breast cancers in the literature, depending on the criteria used. The prevalence of TNBC has ranged from 17.1% to 30.5% and that of BLBC from 8.0% to 55.7% [13
]. Here we reassessed the steroid receptor status and HER2 status of tumors; those tumors without any ER and PR immunostaining were considered as receptor negative and tumors exhibiting more than 10% of invasive tumor cells were considered as steroid receptor-positive. In addition, only grade II-III tumors with ductal histology were included in the TNBC cohort. It is also known that triple-negative and basal-like subtypes of breast carcinoma are not single cohesive entities but instead reflect a collection of different diseases. Outcome in cases of TNBC has consistently been worse compared with ER- and PR-positive tumors, which was again confirmed in the current material. TNBC and BLBC tumors are usually larger and of higher grade compared with receptor-positive breast cancer tumors and some reports suggest that cases of BLBC are more often node-negative [3
]. In the current study, triple-negative carcinomas were larger than ER+/PR+/HER2- cancers, as were basal-like carcinomas compared to non-basal-like triple-negative breast cancers, but no association with nodal status was observed. The proportion of basal-like subtype showing cancers in our material is in line with previous reports [16
One of the most important estrogen-related carcinogenic mechanisms is oxidative metabolism of estrogen and subsequent formation of ROS [17
]. Published data from several laboratories suggest that in vitro
, physiological estrogen concentrations induce significant oxidative stress and that estrogen-induced ROS formation takes place in mitochondria in particular [19
]. Other studies have provided evidence that 8-OHdG levels in the ER-positive MCF-7 breast cancer cell line are over 9-fold higher than in triple-negative MDA-MB-231 cells [9
] and inhibition of estrogen receptor alpha expression significantly reduces estrogen-induced 8-OHdG formation in MCF-7 cells [21
]. Since estrogen levels in ER-positive tumors are higher than in ER-negative ones [22
], our finding that 8-OHdG (as a marker of ROS-derived DNA damage) was highly overexpressed in the ER+/PR+/HER2- group (73.2%) compared with TNBCs (37.1%) is in line with the above data. Previous and current data taken together suggest that ROS play a major role in steroid receptor-positive breast cancer pathogenesis in particular, but not necessarily in cases of TNBC (including BLBC).
A negative prognostic value of immunohistochemical 8-OHdG expression has been reported in connection with at least colorectal carcinoma [23
], ovarian cancer [24
] and malignant melanoma [25
]. Nevertheless, there is a growing body of evidence that in breast cancer the prognostic value of 8-OHdG is different. We have recently reported that immunohistochemical 8-OHdG expression is an independent factor related to good prognosis in breast cancer, especially as regards ductal histology [26
]. In another previous study we found that 8-OHdG expression was significantly diminished in invasive breast carcinomas compared with non-invasive breast lesions [27
]. The current data confirm previous results, as 8-OHdG was a marker of good BCSS also in the current population.
When cells are exposed to oxidative stress, Keap1 undergoes a modification that allows Nrf2 to be released from a complex with it and translocate to the nucleus where it binds to antioxidant response elements of DNA [8
]. Nrf2-mediated antioxidant enzyme induction is one of the major defense mechanisms against excessive ROS production and, on the other hand, PRDX enzymes are considered to be among the most efficient cell redox state-regulating enzymes [8
]. The importance of PRDXs is derived partly from their wide subcellular distribution, in contrast to most other antioxidant enzymes. The majority of ROS are produced in mitochondria under physiological conditions and PRDX III is an especially important part of antioxidant defense, since it is located mainly in mitochondria. Peroxiredoxin IV is found in lysosomes, peroxisomes and the endoplasmic reticulum, where oxidative stress is also a potent threat [5
]. Previous studies carried out in vitro
have demonstrated induction of PRDXs via the Nrf2/Keap1 pathway [28
], but there are no reports on Nrf2 or Keap1 in clinical breast cancer material. The current data is in line with previous in vitro
results as regards steroid receptor-positive breast cancer, since there was highly significant co-expression of PRDX III and Nrf2 in ER+/PR+/HER2- cases, which may reflect Nrf2-mediated PRDX induction after estrogen-induced oxidative stress. Furthermore, PRDX III was associated with 8-OHdG expression in the non-TNBC cohort, probably representing antioxidant induction as a response to oxidative imbalance. Keap1 was highly overexpressed in the TNBC group compared with the steroid receptor-positive control group, which implies that there is no need for intensive (Nrf2-mediated) free radical scavenging in cases of TNBC as a result of a lack of estrogen-induced oxidative stress. Keap1-positive tumors were more aggressive than Keap1-negative ones and Keap1 expression associated also to poor prognosis. This probably derives from the sensitive induction of Keap1 in stressed and damaged tumors, rather than carcinogenesis promoting features of Keap1 itself. However, further mechanistic investigations are required to confirm these hypotheses.
In a previous tissue microarray study we reported on PRDX III overexpression in ER- and PR-positive breast cancers and PRDX IV in PR-positive cases [10
]. In that study, with unselected breast cancer cases and with older methods of steroid receptor assessment, PRDX III- and PRDX IV-positive cases were associated with better prognosis. In the current material neither of the studied PRDX enzymes showed significant association with survival. We observed PRDX IV overexpression in TNBC, especially in non-basal-like breast cancers. PRDX III expression was similar in the TNBC and non-TNBC groups, but PRDX III-positive tumors tended to be smaller and of lower grade. This association with lower grade was observed in a previous study [10
] and suggests a protective role of this mitochondrial enzyme in breast carcinogenesis. All in all, PRDXs III and IV could function as protective enzymes in ER- and PR-positive breast carcinomas, working against the ROS induction of estrogen metabolites.