MicroRNA signatures predict oestrogen receptor, progesterone receptor and HER2/neu receptor status in breast cancer

Abstract

Introduction: Breast cancer is a heterogeneous disease encompassing a number of phenotypically diverse tumours. Expression levels of the oestrogen, progesterone and HER2/neu receptors which characterize clinically distinct breast tumours have been shown to change during disease progression and in response to systemic therapies. Mi(cro)RNAs play critical roles in diverse biological processes and are aberrantly expressed in several human neoplasms including breast cancer, where they function as regulators of tumour behaviour and progression. The aims of this study were to identify miRNA signatures that accurately predict the oestrogen receptor (ER), progesterone receptor (PR) and HER2/neu receptor status of breast cancer patients to provide insight into the regulation of breast cancer phenotypes and progression.

Methods: Expression profiling of 453 miRNAs was performed in 29 early-stage breast cancer specimens. miRNA signatures associated with ER, PR and HER2/neu status were generated using artificial neural networks (ANN), and expression of specific miRNAs was validated using RQ-PCR.

Results: Stepwise ANN analysis identified predictive miRNA signatures corresponding with oestrogen (miR-342, miR-299, miR-217, miR-190, miR-135b, miR-218), progesterone (miR-520g, miR-377, miR-527-518a, miR-520f-520c) and HER2/neu (miR-520d, miR-181c, miR-302c, miR-376b, miR-30e) receptor status. MiR-342 and miR-520g expression was further analysed in 95 breast tumours. MiR-342 expression was highest in ER and HER2/neu-positive luminal B tumours and lowest in triple-negative tumours. MiR-520g expression was elevated in ER and PR-negative tumours.

Conclusions: This study demonstrates that ANN analysis reliably identifies biologically relevant miRNAs associated with specific breast cancer phenotypes. The association of specific miRNAs with ER, PR and HER2/neu status indicates a role for these miRNAs in disease classification of breast cancer. Decreased expression of miR-342 in the therapeutically challenging triple-negative breast tumours, increased miR-342 expression in the luminal B tumours, and downregulated miR-520g in ER and PR-positive tumours indicates that not only is dysregulated miRNA expression a marker for poorer prognosis breast cancer, but that it could also present an attractive target for therapeutic intervention.

Figure 1

Multilayer perceptron with sigmoidal activation function. Weights are adjusted at the end of each epoch by the back-propagation algorithm.

Figure 2

Performance of the models at each step of the analyses. Model performance with each input addition over the course of the analysis for (a) oestrogen receptor (ER) status – 6 optimal transcripts. After the addition of the six optimal microRNA transcripts, the accuracy of the model has reached 100% and there is no further improvement in the error. At this point the model is considered to contain the transcripts that most accurately model the data. Columns represent median model accuracy; lines represent mean squared error for the predictions at each step. (b) progesterone receptor (PR) status – four optimal transcripts, and (c) v-erb-b2 erythroblastic leukaemia viral oncogene homolog 2 receptor (HER2/neu) status – five optimal transcripts.

Figure 3

Population analysis for receptor status. Population analysis for (a) oestrogen receptor (ER) status. Using the transcript signature from the ANN model, it is possible to be able to place a patient with unknown ER status within this population structure, with 100% accuracy from an ANN prediction, (b) progesterone receptor (PR) status, and (c) HER2/neu status. White, receptor-negative patients; grey, receptor-positive patients. y axis, artificial neural network (ANN) prediction with 0 being a receptor-negative prediction and 1 a receptor-positive prediction. Error bars indicate a 95% confidence interval.

Figure 4

Response curves for miR-342, miR-520g and miR-520d. Response curves for (a) miR-342, (b) miR-520g and (c) miR-520d*. Figures show the intensity of each transcript plotted against the artificial neural network (ANN) prediction with respect to the sample being classified as either (a) oestrogen receptor (ER)-positive or ER-negative, (b) progesterone receptor (PR)-positive or PR-negative and (c) v-erb-b2 erythroblastic leukaemia viral oncogene homolog 2 receptor (HER2/neu)-positive or HER2/neu-negative. Error bars indicate 95% confidence intervals.

Figure 5

Coordinate expression of co-located microRNAs. Scatterplots of expression values for microRNAs located adjacently on the same chromosome. (a) miR-16 and miR-15a; Ch13q14.3. (b) miR-16 and miR-15b; Ch3q26.1. (c) miR-143 and miR-145; Ch5q14. (d) miR-99a and let-7c; Ch21q16. (e) miR-195 and miR-497; Ch17p13.1. (f) miR-520g and miR-520h; Ch19q13.42. (g) miR-17-5p, miR-18a, miR-19a, miR-19b, miR-20a, miR-92; Ch13q31.3.

Figure 6

Correlation between microRNA expression on microarray and RQ-PCR. For a subset of microRNAs (miRNAs) and samples we performed RQ-PCR to independently assess miRNA expression. RQ-PCR data are normalized using let-7a and miR-16. There is generally good correlation between miRNA expression using the two techniques. probe-specific differences were observed, however. R value using Pearson correlation, P < 0.05 significant.

Figure 7

Expression of miR-342 and miR-520g in breast tumours. RQ-PCR detection analysis shows that expression levels of miR-342 are increased in: (a) oestrogen receptor (ER)-positive tumours compared with ER-negative tumours (P = 0.04), (b) v-erb-b2 erythroblastic leukaemia viral oncogene homolog 2 receptor (HER2/neu)-positive compared with HER2/neu-negative tumours (P = 0.001), and (c) luminal-B subtype of breast tumours (P = 0.001). (d) miR-520g expression is increased in ER-negative tumours compared with ER-positive tumours (P = 0.005) and in progesterone receptor (PR)-negative tumours compared with PR-positive tumours (P = 0.032). MicroRNA expression presented as log₁₀ of the relative quantity. *P < 0.05, **P < 0.005.