Correlation of miRNA Expression Profiling in Surgical Pathology Materials,with Ki-67,HER2,ER and PR in Breast Cancer Patients

Abstract

Background

New molecular markers related to prognosis and/or clinical outcome have been extensively studied in breast cancer. In particular, microRNA (miRNA) has attracted the interest of both basic and clinical investigators as one of the promising molecular markers of breast cancer patients. MiRNAs are a class of short noncoding RNAs that regulate mRNAs at posttranscriptional level and are deregulated in various human malignancies. Previous studies have reported that miRNAs were stably conserved in 10% formalin-fixed paraffin-embedded tissues without significant degradation, in contrast to more fragile RNA.

Methods

Therefore, in this study, we examined 21 surgical breast cancer specimens using the Human Cancer microRNA PCR Array system (QIAGEN) to explore potential molecular targets of miRNAs.

Results

Profiling of miRNA expression in archival materials demonstrated that a group of deregulated miRNAs was associated with clinicopathological parameters of the patients, such as Ki-67, HER2, ER and PR. For instance, an abundant expression of multiple let-7 miRNA family, also known as tumor suppressor, was detected in low Ki-67 and HER2 groups. Elevated expression of 8 miRNAs overlapped between Ki-67+/HER2+/ER+/PR+ groups, including several known oncogenic miRNAs such as miR-148b, miR-15b, miR-200c, miR-150, miR-191, miR-96, miR-25 and miR-21.

Conclusions

These results all indicated that when analyzing miRNAs in surgical pathology specimens of breast cancer as a biomarker, they should be examined as a cluster through miRNA profiling, rather than relying on the analysis of a single miRNA.

Keywords

Breast cancer ER HER2 Ki-67 MicroRNA PR

Introduction

Gene expression analysis has become more readily available in routine clinical practice due to the development of a few pioneering diagnostic tools. One of the first DNA microarray profiling pioneers, Perou and colleagues, utilized microarray analysis to subclassify breast cancer into 5 different groups: luminal A, luminal B, human EGFR-related 2 (HER2) and basal-like, in addition to unclassified types (1). Their work has been cited repeatedly since, with numerous studies verifying a statistically significant correlation with immunohistochemical (IHC) findings in large numbers of the patients (Tab. I) (2, 3).

TABLE I

Comprehensive results for age, ERα LI, PR LI, HER2 and Ki-67 LI in 21 10% formalin-fixed and paraffin-embedded (FFPE) breast cancer tissues

Case	Age	ERα LI (%)	PR LI (%)	HER2	Ki-67 LI (%)
1	72	96	72	0	3
2	48	65	35	0	26
5	51	92	78	1	14
6	56	1	0	3	27
7	54	65	36	2	6
8	65	1	0	1	32
9	55	82	2	1	20
10	65	56	30	0	17
11	43	1	2	3	28
13	41	0	0	3	33
14	41	75	83	2	11
15	51	62	26	2	30
16	65	70	35	3	12
17	86	0	0	1	29
18	71	95	90	0	12
20	54	78	54	2	7
21	63	8	1	1	34
22	50	4	2	2	6
23	77	74	13	2	23
24	66	92	0	2	10
25	55	63	23	2	17

Correlations between molecular markers are shown in Supplementary Figures 1 and 2. Available online at www.biological-markers.com.

ER = estrogen receptor α; LI = labeling index; PR = progesterone receptor.

For clinical applications of DNA expression analysis, MammaPrint (Agenda, Amsterdam, Netherlands) and Oncotype DX (Generic Health, Red Wood City, CA, USA) are well-known commercialized tools in both the United States and Japan (4, 5). The 70-gene classifier, MammaPrint DNA profiling provides predictions of distant metastasis risk for early-stage breast cancer patients (4). Oncotype DX uses 21 genes to estimate risk of recurrence for early breast cancer patients. The accuracy of both DNA array-based disease outcome predictions has been validated across countries for node-negative or early-stage breast cancer patients (6–8). To conduct, employ, or proceed the array, MammaPrint requires fresh tissue samples, whereas Oncotype DX evaluates target genes with quantitative real-time polymerase chain reaction (qRT-PCR) in 10% formalin-fixed and paraffin-embedded (FFPE) tissue samples. Human tissue specimens are often frozen to avoid RNA degradation, but limitations of storage are considered to make this unsuitable for retrospective analysis. Thus FFPE is frequently used for storage of the tissue specimens, which enables us to perform relatively larger scale retrospective studies. However, highly degraded RNAs in FFPE specimens make it almost impossible to perform or interpret data from conventional retrospective DNA microarrays.

Accumulated evidence has indicated that correlations of microRNA (miRNA) between FFPE and fresh frozen samples exceeds those of total RNA (9, 10). MiRNAs are a class of noncoding small RNAs that were highly conserved across species. Nucleic primary miRNA transcript is processed by Drosha to shorter precursor miRNA, and exported to cytoplasm by Exportin. After cytoplasmic processing of precursor miRNA by Dicer, 20-25 nucleotide lengths of mature miRNA bind to complementary mRNAs for posttranscriptional regulation via degradation or inhibition of translation (11, 12). A number of publications in past decades have suggested diverse miRNAs as new molecular targets and explored their potential for clinical applications in oncogenic regulations. A variety of miRNAs have been proposed as new molecular targets in past decades. Our previous study identified miRNAs in association with breast cancer drug treatment, as well as miRNA expression changes under estrogen-deprived conditions using breast cancer cell lines and FFPE specimens (13). In this study, we collected a breast cancer we collected a breast cancer clinical database at Tohoku Kosai hospital clinical database at Tohoku Kosai hospital, to construct breast cancer miRNA expression profiles (miRNA profiles) based on Ki-67, HER2, estrogen receptor α (ER) and progesterone receptor (PR) status using miRNA microarrays. Based on miRNA profiles, we suggest a cluster of oncogene-related miRNAs to provide further insights into the biological and/or clinical profiles of breast cancer patients.

Materials and Methods

Preparation of breast cancer tissue samples

A total of 21 primary breast cancer specimens were retrieved from Japanese female patients who underwent surgical treatment from 2004 to 2010 at the Tohoku Kosai Hospital (Sendai, Japan) and were available for examination in this study. All of the cases were diagnosed as invasive ductal carcinoma, with age ranging from 41 to 86 years (median 58.6 years). Clinicopathological characteristics of the patients are summarized in Table I. Tissues were fixed in 10% formalin for 24-48 hours at room temperature and embedded in paraffin. The research protocol for this study was approved by the ethics committee of both the Tohoku University Graduate School of Medicine (2010-572) and Tohoku Kosai Hospital (H17.8.5).

Immunohistochemistry of molecular markers

A Histofine SAB PO kit (Nichirei Co Ltd., Tokyo, Japan) was used in combining each primary antibody. The following primary antibodies were used in this study: monoclonal antibodies for ER (ER1D5; Immunotech, Marseille, France), PR (MAB429; Chemicon International Inc., Temecula, CA, USA) and Ki-67 (MIB-1; DakoCytomation Co. Ltd., Kyoto, Japan), as well as rabbit polyclonal antibody for HER2/neu (AO485; DakoCytomation).

The Hercep test was performed to study HER2 status of breast carcinoma (0 to 3+) and classified as follows: negative for 0 and 1+, equivocal for 2+ and positive for 3+, according to the American Society of Clinical Oncology (ASCO) guidelines of 2013 (14). For ER, PR and Ki-67 immunoreactivity, at least 1,000 cells were counted to calculate the ratio of positive cells to determine the labeling Index (LI) (15).

RT² miRNA PCR array

Human cancer microRNA PCR array: MAH-102A (RT² miRNA PCR Array System; QIAGEN, Mannheim, Germany) was used to evaluate expression of 88 cancer-related miRNAs quantitatively in isolated breast carcinoma cells, as described in previous reports(13). This semicomprehensive miRNA array analysis included miRNAs of 4 internal control miRNAs, a reverse transcription control miRNA and a positive control miRNA, and has been validated by individual studies (16). To isolate carcinoma cells from 8-μm FFPE tissue sections, MMI CellCut (Molecular Machines and Industries, Glattbrug, Switzerland) was used to carry out laser capture microdissection. Approximately 5,000 carcinoma cells were obtained from individual tissue sections. These isolated carcinoma cells were deparaffinized at 60°C for 3 hours. Extraction of miRNA was performed as instructed using the Pure Link miRNA Isolation Kit (Invitrogen, Carlsbad, CA), and followed by cDNA synthesis using the RT² miRNA First Strand Kit (QIAGEN). qRT-PCR was employed with the ABI7500 real-time PCR system (Applied Biosystems, Forster city, CA, USA) to determine expression levels of miRNAs.

RT² Profiler PCR Array Data Analysis

The Web-based profiler RT² Profiler PCR Array Data Analysis (QIAGEN) was used to complete the hierarchical clustering analysis and scatterplot analysis for miRNA expression and its relationship with established molecular markers. (http://pcrdataanalysis.sabiosciences.com/pcr/arrayanalysis.php). Ki-67 LI and HER2 status were analyzed by nonsupervised hierarchical clustering to visualize groups of coregulated miRNAs with U6 miRNA internal control normalization. For ER- and PR-associated miRNA expression, a scatterplot was formulated with normalization of global housekeeping miRNAs.

Statistical Analysis

JMP Pro 11 (SAS Institute Inc., Cary, NC, USA) was used to evaluate findings of clinicopathological features in the cases examined. The significance of each value was set as a p value less than 0.05. Associations between molecular markers such as Ki-67, HER2, ER and PR were determined by ANOVA. Fold differences and p values for miRNA expression changes in relation to ER and PR were calculated by RT² Profiler PCR Array Data Analysis.

Results

Comparisons between Ki-67 and Endogenous up-/down-regulated miRNA Profiles

In this study, the Ki-67 LI results were categorized into 3 groups based on several cutoff values reported in previous studies (17–19): 0-14% as control, 15%-24% as intermediate (group 1) and >25% as high (group 2) for clustering analysis (Fig. 1 and Tab. II). There were 20 up-regulated miRNAs in the high Ki-67 group, derived from 3 subclustered groups (Fig. 1, sections A-C). Particularly, hsa-miR-191 in group A (4.062-up, p = 0.080) was identified as an oncomiR in hepatocellular and breast cancer (20, 21) (Fig. 1, section A). In group B (Fig. 1, section B), hsa-miR-7 resulted in the highest fold regulation (7.467-up) with the lowest p value (p = 0.051), in concordance with our previous report that miR-7 contributes to breast cancer malignancy via regulation of EGFR, among endocrine therapy resistant patients (13). In addition, up-regulation of hsa-miR-15b in group C (3.759-up, p = 0.020) was also consistent with previous reports in colon and hepatocellular carcinoma patients (22, 23). In contrast, there were 10 highly expressed miRNAs in the low Ki-67 group (Fig. 1, section D), including several tumor-suppressing miRNAs known from previous reports. For example, hsa-let-7a, hsa-let-7b and hsa-let-7e belong to a family of tumor-suppressing let-7 miRNA in various cancers (24, 25).

Fig. 1

Cluster analysis of oncogene-related miRNAs for Ki-67 expression status. Control group = low Ki-67 group (0-14%); Group 1 = intermediate Ki-67 group (15%-24%); Group 2 = high Ki-67 group (>24%). Magnitude of miRNA expression is represented by the heat map: red indicates high expression, and green indicates low expression. Three subclusters (sections A-C) identified 20 up-regulated miRNAs in the high Ki-67 group (groups 1 and 2), as shown in Table II, and 10 up-regulated miRNAs in the low Ki-67 group (section D).

TABLE II

The list of up- or down-regulated miRNAs accroding to Ki-67 status (Fig. 1)

		Group 1 (Intermediate)		Group 2 (High)		ANOVA
	Mature ID	Fold Regulation	p value	Fold Regulation	p value	p value
A	hsa-miR-132	-1.005	0.813	1.595	0.517	0.848
	hsa-miR-34a	1.144	0.687	2.574	0.365	0.382
	hsa-miR-150^†	1.326	0.599	4.010	0.212	0.303
	hsa-miR-191^†	1.277	0.528	4.062	0.176	0.080
	hsa-miR-27a^‡	1.179	0.783	3.184	0.294	0.370
B	hsa-miR-301a^‡	1.320	0.894	2.674	0.341	0.317
	hsa-miR-335^‡	1.067	0.545	1.565	0.662	0.745
	hsa-miR-27b^†	1.432	0.939	4.009	0.298	0.286
	hsa-miR-7	2.004	0.152	7.467	0.207	0.051
	hsa-miR-199a-3p	1.166	0.808	1.861	0.383	0.910
	hsa-miR-18b	1.142	0.873	1.939	0.649	0.792
	hsa-miR-125b^§	1.497	0.826	3.440	0.316	0.458
	hsa-miR-133b	1.083	0.577	1.520	0.988	0.657
	hsa-miR-200c^†	1.506	0.851	3.367	0.215	0.134
C	hsa-miR-92a^‡	-1.012	0.830	2.522	0.346	0.465
	hsa-miR-16^‡	-1.044	0.766	2.266	0.300	0.262
	hsa-miR-25^†	-1.023	0.627	4.279	0.184	0.090
	hsa-miR-21^{† §}	-1.174	0.816	2.755	0.250	0.277
	hsa-miR-96^{† §}	-1.257	0.462	2.622	0.514	0.175
	hsa-miR-15b^†	-1.393	0.465	3.759	0.171	0.020^*
D	hsa-let-7e	-6.799	0.185	-2.140	0.128	0.028^*
	hsa-miR-143	-1.751	0.451	-1.398	0.382	0.608
	hsa-let-7b	-1.867	0.459	-1.426	0.244	0.171
	hsa-miR-29a	-1.618	0.356	-1.307	0.659	0.660
	hsa-miR-218^‡	-2.079	0.302	-1.256	0.788	0.396
	hsa-let-7a	-4.426	0.206	-1.445	0.231	0.100
	hsa-miR-214^‡	-3.186	0.174	-1.360	0.363	0.276
	hsa-miR-181a	-1.366	0.797	-1.119	0.858	0.751
	SNORD48	-1.701	0.088	-1.212	0.683	0.364
	hsa-miR-15a^‡	-2.959	0.334	-1.097	0.925	0.587

Comparisons between HER2 Status and Endogenous up-/down-regulated miRNA Profiles

Expression of HER2 was evaluated based on IHC results from 0 to 3+, and classified into 3 groups as described in “Materials and methods,” to employ cluster analysis. We identified 5 miRNAs (Fig. 2, section A and Tab. III) that were significantly up-regulated as HER2 expression increased. Of interest, hsa-miR-21, hsa-miR-96 and hsa-miR-125b were also overexpressed in the high Ki-67 group, which is consistent with the fact that hsa-miR-21 was enlisted above as oncogenic miRNA in breast cancer (26). On the other hand, 26 miRNAs were dysregulated in the low HER2 group, including hsa-let-7c, hsa-let-7d and hsa-let-7f (Fig. 2, section B). Among these reverse-correlated miRNAs, hsa-let-214 and hsa-miR-15a were also detected in the low Ki-67 group (Tab. II), and both miRNAs have been reported as tumor suppressors in breast cancer (27, 28). However, miRNAs positively correlated with Ki-67 status, such as hsa-miR-27a, hsa-miR-92a, hsa-miR-301a, hsa-miR-335a and hsa-miR-16 were also abundant in the low HER2 group, indicating that this group contained heterogeneous oncogene-related miRNAs.

Fig. 2

Cluster analysis of oncogene-related miRNAs for HER2 expression status. Control group: low HER2 group (immunohistochemistry [IHC] score 0, 1); Group 1: intermediate HER2 group (IHC score 2); Group 2: high HER2 group (IHC score 3). Magnitude of miRNA expression is represented by the heat map; red indicates high expression, and green indicates low expression. Section A: up-regulated miRNAs in the high HER2 group; section B: up-regulated miRNAs in the low HER2 group.

TABLE III

The list of up- or down-regulated miRNAs according to HER2 status (Fig. 2)

		Group 1		Group 2		ANOVA
	Mature ID	Fold Regulation	p value	Fold Regulation	p value	p value
A	hsa-miR-146a	1.029	0.961	1.059	0.679	0.894
	hsa-miR-21^{† §}	1.093	0.434	1.244	0.891	0.709
	hsa-miR-363	1.279	0.971	2.102	0.127	0.123
	hsa-miR-96^{† §}	1.042	0.461	1.455	0.844	0.640
	hsa-miR-125b^§	1.009	0.536	1.159	0.936	0.801
B	hsa-miR-98	-3.617	0.073	-1.877	0.481	0.164
	hsa-miR-27a^§	-2.021	0.289	-1.421	0.501	0.455
	hsa-miR-203^†	-1.804	0.239	-1.369	0.445	0.375
	hsa-miR-18a	-3.926	0.001^*	-2.269	0.462	0.021^*
	hsa-let-7c	-1.386	0.101	-1.276	0.285	0.159
	hsa-miR-301a^§	-2.318	0.142	-1.785	0.317	0.254
	hsa-miR-30c	-1.943	0.411	-1.538	0.413	0.559
	hsa-miR-20b	-5.293	0.237	-2.239	0.437	0.387
	hsa-miR-181d	-5.073	0.038^*	-2.262	0.234	0.060
	hsa-let-7d	-5.588	0.244	-2.309	0.452	0.387
	hsa-miR-125a-5p	-1.686	0.478	-1.786	0.470	0.487
	hsa-miR-15a^‡	-1.798	0.225	-1.950	0.291	0.280
	hsa-miR-134	-1.300	0.436	-1.354	0.415	0.589
	hsa-miR-335^§	-2.372	0.717	-3.000	0.049^*	0.400
	hsa-miR-218^‡	-1.287	0.552	-1.351	0.644	0.832
	hsa-miR-122	-2.600	0.663	-2.088	0.528	0.797
	hsa-miR-372	-3.947	0.148	-2.510	0.337	0.212
	hsa-miR-17	-4.088	0.021^*	-2.646	0.149	0.032^*
	hsa-miR-149	-2.145	0.202	-1.780	0.286	0.294
	hsa-miR-92a^§	-2.620	0.075	-2.290	0.300	0.151
	hsa-miR-196a	-2.538	0.161	-2.190	0.696	0.336
	hsa-miR-193b	-6.265	0.016^*	-4.047	0.146	0.027^*
	hsa-let-7f	-2.791	0.109	-2.566	0.142	0.198
	hsa-miR-19a	-2.638	0.090	-2.370	0.317	0.199
	hsa-miR-214^‡	-2.323	0.810	-2.143	0.499	0.799
	hsa-miR-16^§	-1.342	0.204	-1.315	0.472	0.367

Identification of down-regulated miRNAs based on ER and PR Status

In this study, 4 cases were categorized as part of the ER- group (0-8%), whereas the rest of the 17 cases were ER+ (56%-96%). For PR status, 10 samples were PR- (0-13%), and 11 were PR+ (23%-90%). Normalization was performed using global housekeeping genes (SNORD44, SNORD47 and SNORD48). RNU6 was excluded since its expression was not necessarily consistent between the ER+ and ER- groups, as previously reported (29).

There were 15 down-regulated miRNAs significantly correlated with ER- status, while 11 miRNAs significantly underexpressed in cases of PR- status exhibited a similar trend to that of the ER- group, as shown in Figure 3 (Tabs. IV and V). Of those, 8 miRNAs were also detected in a group of up-regulated miRNAs in Ki-67 positive cases. Previous studies have indicated that hsa-miR-15b and hsa-miR-200c were dysregulated in ER+ breast cancer, using cell lines in vitro (30). Notably, hsa-miR-21 and hsa-miR-96 were identified to be positively correlated with Ki-67, HER2, ER and PR status.

Fig. 3

Scatterplot analysis of miRNA expression based on ER/PR expression in 21 human breast cancer patients. Normalization was performed with global housekeeping genes. Solid lines indicate a 2-fold increment or decrement of miRNA expression levels. A) Estrogen receptor α (ER) expression was categorized into ER- (0-8%, n = 4) and ER+ (56%-96%, n = 17) groups. B) Progesterone receptor (PR) expression was categorized into PR- (0-13%, n = 10) and ER+ (23%-90%, n = 11) groups.

TABLE IV

Fifteen miRNAs under-expressed in ER- group vs. ER+ group

Mature ID	Fold regulation	p value	Correspondence
hsa-miR-148b^*	-4.609	0.016
hsa-miR-150^*	-7.150	0.037
hsa-miR-15b^*	-4.847	0.004
hsa-miR-191^*	-5.655	0.004
hsa-miR-200c^*	-3.461	0.010	Up-regulated miRNAs in high Ki-67
hsa-miR-21^*^†	-3.587	0.041
hsa-miR-25^*	-6.280	0.003
hsa-miR-27b	-4.052	0.032
hsa-miR-96^*^†	-3.895	0.019
hsa-miR-203^*	-8.656	0.020	Up-regulated miRNAs in low HER2
hsa-miR-140-5p	-2.522	0.038	^*Down-regulated miRNAs in low PR status
hsa-miR-20a	-2.707	0.050
hsa-miR-135b	-5.489	0.034
hsa-miR-183	-4.793	0.003
hsa-miR-23b	-6.507	0.036

Mature ID	Fold regulation	p value	Correspondence
hsa-miR-148b	-2.529	0.016
hsa-miR-150	-2.777	0.037
hsa-miR-15b	-2.345	0.003
hsa-miR-191	-3.076	0.004
hsa-miR-200c	-3.839	0.010	Up-regulated miRNAs in high Ki-67
hsa-miR-21^*	-2.458	0.041
hsa-miR-25	-2.206	0.003
hsa-miR-96	-4.263	0.020
hsa-miR-203	-4.261	0.019	Up-regulated miRNAs in low HER2
hsa-miR-140-5p	-3.298	0.038	†Down-regulated miRNAs in low ER status
hsa-miR-20a	-2.757	0.050

Discussion

Possible roles of miRNAs in human cancer development have emerged in the results of numerous studies. In fact, a phase II clinical trial using miRNA as a molecular target was completed in the United States (31). Yet, clinically relevant profiling of miRNAs is still demanded to proceed to therapeutic intervention. A number of array-based analyses have shown that miRNA expression in FFPE tissues is consistent with fresh frozen samples, whereas total RNA expression in FFPE tissues tends to be changed due to degradation and modification during the fixation process (9). In this study, miRNA expression of breast cancer related biomarkers was profiled from the same FFPE tissues that were evaluated for IHC of Ki-67, HER2, ER and PR. The number of the cases examined in our study was limited, but statistical analysis of pathological parameters among the breast cancer tissues was reasonably consistent with larger scale analyses in the past. Consequently, ER LI and PR LI were positively correlated (p = 0.0002; Suppl Fig. 1A), while both ER LI and PR LI were reverse-correlated with Ki-67 LI (p = 0.0005, p = 0.0088; Suppl Fig. 1B, C).

Hierarchical cluster analyses have demonstrated that expression patterns of the let-7 family (especially, hsa-let-7a-f) were negatively correlated with Ki-67 and HER2 in the cases examined. Other miRNA profiling analyses using breast cancer cell lines and tissues have also suggested that the let-7 family were significantly under-regulated comparing with normal mammary epithelial cells (32, 33). A family of let-7 miRNAs was one of the first miRNAs discovered during terminal differentiation of Caenorhabditis elegans in 2000 (34). In human, let-7 is well known to be involved in differentiation, development, cell cycle and gene regulation (32). In particular, let-7 expression was reported to be down-regulated in various cancers, and restoration of its expression was also reported to suppress carcinoma cell growth (35). We have previously reported that let-7f may repress an estrogen-synthesizing enzyme, aromatase, and let-7f and let-7d were down-regulated in Lin28-positive breast cancer (36, 37). These results were all consistent with earlier studies in which let-7 was suggested as a therapeutic target for suppressing cancer cell proliferation (35, 38, 39).

Among the 15 down-regulated miRNAs in the ER+ group, 9 miRNAs were abundantly expressed in the high Ki-67 group (Tab. III, group A). Of those, 5 miRNAs (hsa-miR-191, hsa-miR-15b, hsa-miR-200c, hsa-miR-21, hsa-miR-25) are known to be involved in ER activation in breast cancer (40–42). In particular, accumulated reports have shown that expression of miR-21 is markedly elevated in glioblastoma and lung and breast cancer, and promotes tumor progression by inhibiting apoptosis or regulation of epithelial-mesenchymal transition (43–46). Furthermore, miR-150 was suggested to have opposite roles depending on cell types. In gastric cancer, miR-150 inhibits proapoptotic factors and promotes cancer progression (47). Conversely, decreased expression of miR-150 was significantly associated with poor survival rate and worse response to adjuvant chemotherapy in colorectal cancer (48). These opposite effects toward carcinogenesis imply that miRNAs may be influenced by other regulatory factors according to cellular characteristics.

Our results indicated that miRNAs may be involved in the development of HER2 overexpression (Fig. 2). HER2 overexpression was reported to be correlated with effectiveness of chemotherapy (49, 50). For 2+ HER2 groups, the FISH test is commonly used for further evaluation to determine sensitivity to additional chemotherapy. Figure 2, section A, summarizes commonly up-regulated miRNAs in 2+ and 3+ HER2 groups, whereas Figure 2, section B, summarizes similar expression levels between negative and 2+ HER2 groups. Therefore, the analysis of several miRNAs could further subclassify the group of patients with 2+ HER2. Results of our study indicated that the possible evaluation of miRNAs in the same archival materials in which IHC was performed could contribute to further refining the interpretation of immunostaining results. The fact that this approach could identify the abnormalities in protein and miRNAs in the same groups of carcinoma cells is considered especially pivotal in light of recent controversies regarding intratumoral heterogeneity among breast cancer patients.

It is also true that detection of single miRNA expression may not necessarily provide clinically relevant information, as our study illustrated. This could be due to the fact that miRNA expression may be reflected in multiple factors, and detection of endogenous underexpressed miRNA can often be challenging. It is highly possible that improper expression of miRNAs could be identified if miRNAs were epigenetically modulated in heterogeneous cancer cells. Thus, observation of miRNA clusters rather than single miRNAs is desirable for the application of miRNA diagnosis to achieve accuracy in the results. Here, we provide evidence that a group of distinctive miRNAs in accordance combination with established molecular markers such as ER, PR, HER2 and Ki-67 could be a novel diagnosis tool for breast cancer. In our study, there were 8 miRNAs that were significantly associated with Ki-67, HER2, ER and PR status: namely, miR-148b, miR-15b, miR-200c, miR-150, miR-191, miR-96, miR-25 and miR-21 could be novel candidates as a group of deregulated miRNA markers that occur among hormone therapy-resistant patients. Evaluation of collective miRNA expression patterns can lead to the development of a more stable molecular biology based diagnostic tool (51). There are many reports of similar experiments, but the results were based on in vitro cell lines. In our study, we successfully extracted RNAs from tissue specimens and performed miRNA PCR array to compare the results with their steroid receptor status. This is the first study to show comprehensive miRNA expression patterns in relation to clinical markers such as ER, PR, HER2 and Ki-67. MiRNA studies of cancer have investigated various miRNA targets in past decades, and therefore, we believe that this comprehensive miRNA analysis could be useful to narrow down research to focus on more reliable markers in hormone-dependent cancers. In addition, the aim of our study was partly to suggest a new method of more detailed classification for each breast cancer phenotype by profiling groups of miRNA expression patterns. It is expected that introduction of an miRNA profiling method can contribute to early noninvasive detection of breast cancer risk as well as oncomiR-targeted therapy for patients with drug treatment resistance.

Footnotes

Acknowledgement

We would like to thank Mr. Katsuaki Ono, our lab technician, for generous advice regarding tissue sample preparation.

Financial support: This research was supported by a Grant-in-Aid for Challenging Exploratory Research, Japan Society for the Promotion of Science (JSPS) KAKENHI grant number 23659183.

Conflict of interest: The authors declare they have no conflicts of interest in this study.

References

Perou

Sørlie

Eisen

Molecular portraits of human breast tumours.

Nature 2000;406(6797):747–752

Fan

Krutilina

Sun

Comprehensive analysis of microRNA (miRNA) targets in breast cancer cells.

J Biol Chem 2013;288(38):27480–27493

Nielsen

Hsu

Jensen

Immunohistochemical and clinical characterization of the basal-like subtype of invasive breast carcinoma.

Clin Cancer Res 2004;10(16):5367–5374

van't Veer

Dai

van de Vijver

Gene expression profiling predicts clinical outcome of breast cancer.

Nature 2002;415(6871):530–536

Paik

Shak

Tang

A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer.

N Engl J Med 2004;351(27):2817–2826

Bueno-de-Mesquita

van Harten

Retel

Use of 70-gene signature to predict prognosis of patients with node-negative breast cancer: a prospective community-based feasibility study (RASTER).

Lancet Oncol 2007;8(12):1079–1087

Ishitobi

Goranova

Komoike

Clinical utility of the 70-gene MammaPrint profile in a Japanese population.

Jpn J Clin Oncol 2010;40(6):508–512

Wittner

Sgroi

Ryan

Analysis of the MammaPrint breast cancer assay in a predominantly postmenopausal cohort.

Clin Cancer Res 2008;14:2988–2993

Nakajima

Gavin

Systematic analysis of microRNA expression of RNA extracted from fresh frozen and formalin-fixed paraffin-embedded samples.

RNA 2007;13(10):1668–1674

10.

Liu

Tetzlaff

Vanbelle

MicroRNA expression profiling outperforms mRNA expression profiling in formalin-fixed paraffin-embedded tissues.

Int J Clin Exper Pathol 2009;2:519–527

11.

Zamore

Haley

Ribo-gnome: the big world of small RNAs.

Science 2005;309(5740):1519–1524

12.

Nana-Sinkam

Croce

Clinical applications for microRNAs in cancer.

Clin Pharmacol Ther 2013;93(1):98–104

13.

Masuda

Miki

Hata

An induction of microRNA, miR-7 through estrogen treatment in breast carcinoma.

J Transl Med 2012;10Suppl 1:S2

14.

Wolff

Hammond

Hicks

Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American Society of Clinical Oncology/College of American Pathologists clinical practice guideline update.

J Clin Oncol 2013;31:3997–4013

15.

Pinto

André

Pereira

Nóbrega

Soares

Prognostic comparative study of S-phase fraction and Ki-67 index in breast carcinoma.

J Clin Pathol 2001;54(7):543–549

16.

Philippidou

Schmitt

Moser

Signatures of microRNAs and selected microRNA target genes in human melanoma.

Cancer Res 2010;70(10):4163–4173

17.

Tamaki

Ishida

Tamaki

Analysis of clinically relevant values of Ki-67 labeling index in Japanese breast cancer patients.

Breast Cancer 2012

18.

Urruticoechea

Smith

Dowsett

Proliferation marker Ki-67 in early breast cancer.

J Clin Oncol 2005;23(28):7212–7220

19.

Goldhirsch

Wood

Coates

Gelber

Thürlimann

Senn

Panel members. Strategies for subtypes: dealing with the diversity of breast cancer: highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2011.

Ann Oncol 2011;22(8):1736–1747

20.

Mar-Aguilar

Luna-Aguirre

Moreno-Rocha

Differential expression of miR-21, miR-125b and miR-191 in breast cancer tissue.

Asia Pac J Clin Oncol 2013;9(1):53–59

21.

Elyakim

Sitbon

Faerman

hsa-miR-191 is a candidate oncogene target for hepatocellular carcinoma therapy.

Cancer Res 2010;70(20):8077–8087

22.

Liu

Yao

Wang

Circulating miR-15b and miR-130b in serum as potential markers for detecting hepatocellular carcinoma: a retrospective cohort study.

BMJ Open 2012;2(2):e000825

23.

Frampton

Krell

Gall

Castellano

Stebbing

Jiao

miR-15b and miR-17 are tumor-derived plasma microRNAs dysregulated in colorectal neoplasia.

Ann Surg 2014 Mar 18

24.

Schubert

Spahn

Kneitz

Distinct microRNA expression profile in prostate cancer patients with early clinical failure and the impact of let-7 as prognostic marker in high-risk prostate cancer.

PLoS ONE 2013;8(6):e65064

25.

Johnson

Grosshans

Shingara

RAS is regulated by the let-7 microRNA family.

Cell 2005;120(5):635–647

26.

Iorio

Croce

MicroRNAs in cancer: small molecules with a huge impact.

J Clin Oncol 2009;27:5848–56

27.

Derfoul

Juan

Difilippantonio

Palanisamy

Ried

Sartorelli

Decreased microRNA-214 levels in breast cancer cells coincides with increased cell proliferation, invasion and accumulation of the Polycomb Ezh2 methyltransferase.

Carcinogenesis 2011;32(11):1607–1614

28.

Luo

MiR-15a is underexpressed and inhibits the cell cycle by targeting CCNE1 in breast cancer.

Int J Oncol 2013;43(4):1212–1218

29.

Appaiah

Goswami

Mina

Persistent upregulation of U6:SNORD44 small RNA ratio in the serum of breast cancer patients.

Breast Cancer Res 2011;13(5):R86

30.

Cochrane

Cittelly

Howe

MicroRNAs link estrogen receptor alpha status and Dicer levels in breast cancer.

Horm Cancer 2010;1(6):306–319

31.

Lanford

Hildebrandt-Eriksen

Petri

Therapeutic silencing of microRNA-122 in primates with chronic hepatitis C virus infection.

Science 2010;327(5962):198–201

32.

Wang

Cao

Wang

Liu

You

Regulation of let-7 and its target oncogenes [review].

Oncol Lett 2012;3:955–960

33.

Riaz

van Jaarsveld

Hollestelle

miRNA expression profiling of 51 human breast cancer cell lines reveals subtype and driver mutation-specific miRNAs.

Breast Cancer Res 2013;15(2):R33

34.

Reinhart

Slack

Basson

The 21-nucleotide let-7 RNA regulates developmental timing in Caenorhabditis elegans.

Nature 2000;403(6772):901–906

35.

Johnson

Esquela-Kerscher

Stefani

The let-7 microRNA represses cell proliferation pathways in human cells.

Cancer Res 2007;67(16):7713–7722

36.

Shibahara

Miki

Onodera

Aromatase inhibitor treatment of breast cancer cells increases the expression of let-7f, a microRNA targeting CYP19A1.

J Pathol 2012;227(3):357–366

37.

Sakurai

Miki

Masuda

LIN28: a regulator of tumor-suppressing activity of let-7 microRNA in human breast cancer.

J Steroid Biochem Mol Biol 2012;131:101–106

38.

Barh

Malhotra

Ravi

Sindhurani

MicroRNA let-7: an emerging next-generation cancer therapeutic.

Curr Oncol 2010;17(1):70–80

39.

Frost

Olson

Control of glucose homeostasis and insulin sensitivity by the Let-7 family of microRNAs.

Proc Natl Acad Sci USA 2011;108(52):21075–21080

40.

Di Leva

Piovan

Gasparini

Estrogen mediated-activation of miR-191/425 cluster modulates tumorigenicity of breast cancer cells depending on estrogen receptor status.

PLoS Genet 2013;9(3):e1003311

41.

Klinge

MiRNAs and estrogen action.

Trends Endocrinol Metab 2012;23(5):223–233

42.

Maillot

Lacroix-Triki

Pierredon

Widespread estrogen-dependent repression of microRNAs involved in breast tumor cell growth.

Cancer Res 2009;69(21):8332–8340

43.

Chan

Krichevsky

Kosik

MicroRNA-21 is an antiapoptotic factor in human glioblastoma cells.

Cancer Res 2005;65(14):6029–6033

44.

Gao

Cao

Deregulated expression of miR-21, miR-143 and miR-181a in non small cell lung cancer is related to clinicopathologic characteristics or patient prognosis.

Biomed Pharmacother 2010;64:399–408

45.

Han

Wang

Liu

MiR-21 regulates epithelial-mesenchymal transition phenotype and hypoxia-inducible factor-1α expression in third-sphere forming breast cancer stem cell-like cells.

Cancer Sci 2012;103(6):1058–1064

46.

Esquela-Kerscher

Slack

Oncomirs: microRNAs with a role in cancer.

Nat Rev Cancer 2006;6(4):259–269

47.

Jin

Yang

MiR-150 promotes gastric cancer proliferation by negatively regulating the pro-apoptotic gene EGR2.

Biochem Biophys Res Commun 2010;392(3):340–345

48.

Zhang

Wang

miR-150 as a potential biomarker associated with prognosis and therapeutic outcome in colorectal cancer.

Gut 2012;61(10):1447–1453

49.

Slamon

Leyland-Jones

Shak

Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2.

N Engl J Med 2001;344(11):783–792

50.

Piccart-Gebhart

Procter

Leyland-Jones

Herceptin Adjuvant (HERA) Trial Study Team.

Trastuzumab after adjuvant chemotherapy in HER2-positive breast cancer. N Engl J Med 2005;353(16):1659–1672

51.

Blenkiron

Goldstein

Thorne

MicroRNA expression profiling of human breast cancer identifies new markers of tumor subtype.

Genome Biol 2007;8(10):R214

Case	Age	ERα LI (%)	PR LI (%)	HER2	Ki-67 LI (%)
1	72	96	72	0	3
2	48	65	35	0	26
5	51	92	78	1	14
6	56	1	0	3	27
7	54	65	36	2	6
8	65	1	0	1	32
9	55	82	2	1	20
10	65	56	30	0	17
11	43	1	2	3	28
13	41	0	0	3	33
14	41	75	83	2	11
15	51	62	26	2	30
16	65	70	35	3	12
17	86	0	0	1	29
18	71	95	90	0	12
20	54	78	54	2	7
21	63	8	1	1	34
22	50	4	2	2	6
23	77	74	13	2	23
24	66	92	0	2	10
25	55	63	23	2	17

Case	Age	ERα LI (%)	PR LI (%)	HER2	Ki-67 LI (%)
1	72	96	72	0	3
2	48	65	35	0	26
5	51	92	78	1	14
6	56	1	0	3	27
7	54	65	36	2	6
8	65	1	0	1	32
9	55	82	2	1	20
10	65	56	30	0	17
11	43	1	2	3	28
13	41	0	0	3	33
14	41	75	83	2	11
15	51	62	26	2	30
16	65	70	35	3	12
17	86	0	0	1	29
18	71	95	90	0	12
20	54	78	54	2	7
21	63	8	1	1	34
22	50	4	2	2	6
23	77	74	13	2	23
24	66	92	0	2	10
25	55	63	23	2	17

Correlation of miRNA Expression Profiling in Surgical Pathology Materials,with Ki-67,HER2,ER and PR in Breast Cancer Patients

Abstract

Background

Methods

Results

Conclusions

Keywords

Introduction

Materials and Methods

Preparation of breast cancer tissue samples

Immunohistochemistry of molecular markers

RT2 miRNA PCR array

RT2 Profiler PCR Array Data Analysis

Statistical Analysis

Results

Comparisons between Ki-67 and Endogenous up-/down-regulated miRNA Profiles

Comparisons between HER2 Status and Endogenous up-/down-regulated miRNA Profiles

Identification of down-regulated miRNAs based on ER and PR Status

Discussion

Footnotes

Acknowledgement

References

RT² miRNA PCR array

RT² Profiler PCR Array Data Analysis

Case	Age	ERα LI (%)	PR LI (%)	HER2	Ki-67 LI (%)
1	72	96	72	0	3
2	48	65	35	0	26
5	51	92	78	1	14
6	56	1	0	3	27
7	54	65	36	2	6
8	65	1	0	1	32
9	55	82	2	1	20
10	65	56	30	0	17
11	43	1	2	3	28
13	41	0	0	3	33
14	41	75	83	2	11
15	51	62	26	2	30
16	65	70	35	3	12
17	86	0	0	1	29
18	71	95	90	0	12
20	54	78	54	2	7
21	63	8	1	1	34
22	50	4	2	2	6
23	77	74	13	2	23
24	66	92	0	2	10
25	55	63	23	2	17