Human 3β‐hydroxysteroid dehydrogenase type 1 in human breast cancer: clinical significance and prognostic associations

Abstract Active sex steroids including estrogens and androgens are locally produced from circulating inactive steroids by various steroid‐metabolizing enzymes, and play pivotal roles in the progression of hormone‐dependent breast cancers. Human 3β‐hydroxysteroid dehydrogenase type 1 (3β‐HSD type 1) is a critical enzyme in the formation of all classes of active steroid hormones, and is also involved in the inactivation of potent androgen dihydrotestosterone (DHT). Therefore, this enzyme is suggested to modulate active sex steroid production or inactivation, with a role in hormone‐dependent breast cancer. The purpose of this study was to investigate the clinical significance of 3β‐HSD type 1 in human breast cancer. Using immunohistochemistry (IHC), we evaluated 3β‐HSD type 1 expression in 161 human breast cancers and analyzed correlations of 3β‐HSD type 1 expression with various clinicopathological factors. Of 161 breast cancer cases, 3β‐HSD type 1 expression in cancer cells was detected in 119 cases (73.9%), and was positively correlated with estrogen receptor (ER)‐positivity but not HER‐2 status. In ER‐positive cases (n = 130), 3β‐HSD type 1 expression was inversely correlated with invasive tumor size (P = 0.0009), presence of invasive region (P = 0.0107), and lymphatic involvement (P = 0.0004). 3β‐HSD type 1 expression was significantly associated with decreased risk of recurrence or improved prognosis by both univariate (P = 0.0003 and P = 0.009, respectively) and multivariate (P = 0.027 and P = 0.023, respectively) analyses. Our findings indicate that this enzyme is a prognostic factor in hormone‐dependent breast cancer.


Introduction
Breast carcinoma is one of the most common malignancies in women worldwide, and the majority of cases are hormone-dependent [1,2]. In hormone-dependent breast cancers, estrogens play crucial roles in cancer progression through the nuclear estrogen receptor (ER) [1,3]. Other bioactive sex steroids androgens, such as dihydrotestosterone (DHT) and its precursor testosterone (TS), have been reported to have inhibitory effects in hormone-dependent breast cancer cells [4][5][6]. Consequently, it is postulated that the progression of hormone-dependent breast cancers is regulated by the overall effect of bioactive sex steroids. These active sex steroids are produced in endocrine organs including the ovary in females and the testis in males, and are secreted into the circulation. However, the serum concentration of hormones does not necessarily reflect the local hormonal activities in the target tissues. Numerous recent studies reported that active sex steroids are also locally produced from circulating inactive steroids; this ORIGINAL RESEARCH Human 3β-hydroxysteroid dehydrogenase type 1 in human breast cancer: clinical significance and prognostic associations Toru Hanamura 1 , Tokiko Ito 1 , Toshiharu Kanai 1 , Kazuma Maeno 1 , Yasuyo Shimojo 2 , Takeshi Uehara 2 , Takashi Suzuki 3 , Shin-ichi Hayashi 4 & Ken-ichi Ito 1 system of hormone action is known as the intracrine system [5,[7][8][9][10]. It is also reported that a substantial proportion of estrogens in women (approximately 75% before menopause, and almost 100% after menopause) are synthesized in peripheral hormone-target tissues from abundant circulating precursor steroids of adrenal origin [5,10]. It is thus important to evaluate the physiological and/or pathological significance of this intracrine activity [7].
Human 3β-HSD type 1 is a critical enzyme in the conversion of DHEA and A-diol to estrogen precursors A-dione and TS. This enzyme is also involved in conversion of the potent androgen DHT to its inactive form 5α-androstane-3β,17β-diol (3β-diol), which has substantial estrogenic activity [14][15][16][17][18][19][20]. Therefore, this enzyme is suggested to be a modulator of active sex steroid production or inactivation, and plays a role in hormone-dependent breast cancer. However, the expression of 3β-HSD type 1 has not been sufficiently examined in human breast cancer tissue, and the biological significance of 3β-HSD type 1 remains unknown.
The purpose of this study was to examine the expression of 3β-HSD type 1 and its biological and prognostic significance in human breast cancer. To accomplish this, we evaluated the expression of 3β-HSD type 1 in 161 human breast cancer specimens by immunohistochemistry (IHC) and analyzed for correlations with various clinicopathological factors. Our results indicate for the first time that this enzyme is a potent prognostic factor in hormonedependent breast cancer.

Tumor samples
All human breast cancer tissues were obtained during surgery at the Shinshu University Hospital (Nagano, Japan) after patient consent and with approval from the Shinshu University Hospital Ethics Committee. Tumor samples were obtained from all patients with untreated breast cancer who had undergone surgery during January 2004 and December 2005 (n = 161). Clinicopathological data including age, menopausal state, invasive tumor size, histological type, lymphatic involvement, lymph node status, histological grade, ER status, progesterone receptor (PgR) status, human epidermal growth factor receptor-2 (HER-2) status, and prognosis were collected by reviewing patient case records. At the time of surgery, patients who had not menstruated for more than 1 year were defined as postmenopausal; other patients were defined as premenopausal. The histological grade of each specimen was evaluated by one pathologist (T. U.) based on the method of Robbins et al. [21]. ER, PgR, and HER-2 statuses were evaluated by IHC staining. The cut-off value for ER and PgR positivity was set at ≥10% [22]. Tumors were considered to overexpress HER-2 if they were given a score of 3 following IHC staining, or if they showed >2.2-fold amplification of the HER-2 gene, as assessed by fluorescence in situ hybridization. FISH testing was only performed for tumors that scored 2 during IHC staining [23]. Because adjuvant therapies were administered based on criteria in 2004, it should be noted that the criteria for ER-, PgR-, and HER-2 positivity are not new. None of the patients examined in our study received irradiation, chemotherapy, or endocrine therapy prior to surgery. Sixty-four patients received adjuvant chemotherapy, while 128 patients received adjuvant endocrine therapy after the surgery. The mean follow-up time was 104 months (range 4-132 months). Disease-free and disease-specific survival data were available for all patients. All specimens were fixed with 10% formalin and embedded in paraffin wax. Snap-frozen tissues were not available for examination in this study. Patient characteristics are listed in Table 1.

Immunohistochemistry
A Vectastain ABC kit (VectorLaboratories, Burlingame, CA, USA), which uses the avidin-biotin complex (ABC) Human 3β-hydroxysteroid dehydrogenase type 1 in human breast cancer T. Hanamura et al. method, was used for the 3β-HSD type 1 immunoreactive staining. After deparaffinization, antigen retrieval was performed by heating the slides in a microwave oven twice for 6 min in a citric acid buffer. The dilution of 3β-HSD type 1 primary antibody was 1:200. After overnight incubation at 4°C with the primary antibody, incubation with the biotinylated anti-mouse IgG (Vectastain ABC kit) was performed for 1 h at RT. Then, sections were incubated with avidin-biotinperoxidase complex (Vectastain ABC Kit) at RT for 1 h. Finally, the antigen-antibody complex was visualized after a 6-min incubation with diaminobenzidine. Counterstaining was performed using hematoxylin. Tissue sections of human placenta [15], human adrenal gland (zona glomerulosa) [24], and a cell block of HSD3B1-overexpressing breast cancer cell line (E10-HSD3B1) and its control cell line (E10-control), which were established from the MCF-7 human breast cancer cell line [25], were used as positive controls for 3β-HSD type 1. As a negative control, mouse IgG1-kappa monoclonal antibody (ab18447) was used instead of the primary antibodies for isotype control. Immunoreactive staining of ER, PgR, and HER-2 were performed as previously described [26].

Scoring of immunoreactivity
For statistical analyses of 3β-HSD type 1, the cancer samples were independently and blindly classified into two groups (positive and negative cancer cells) by three of the authors (T. H., T. S., and T. U.). Because there are no reports describing the evaluation criteria of 3β-HSD type 1 in breast cancer in the past, we used the evaluation criteria by Suzuki et al., which reported other steroidmetabolizing enzymes, with some modifications [27,28]. The cases were defined as positive if more than 50% cancer cells showed cytosolic staining regardless of its intensity; others were defined as negative. Interobserver differences were 11.8% (19 cases). Cases with discordant results among the observers were simultaneously reevaluated using a multiheaded microscope.

Statistical analysis
Statistical analyses were performed using the StatFlex 6.0 software program (Artech Co., Ltd., Osaka, Japan). Values for patient age and invasive tumor size were summarized as the median (range). Other patient characteristics were expressed in absolute numbers (%). Quantitative data and categorical data were compared using unpaired t-tests and chi-square tests, respectively. Disease-free and disease-specific survival curves were generated according to the Kaplan-Meier method. The statistical significance of differences in the survival analyses were calculated using the log-rank test. In addition, univariate and multivariate analyses for prognoses were evaluated by a proportional hazard model (Cox). Significant variables evaluated by univariate analyses were only examined in the multivariate analyses. P-values of <0.05 were considered significant.

Immunohistochemistry
Strong immunoreactivity of 3β-HSD type 1 was detected in the syncytiotrophoblast of human placenta and the  (Fig. 1A, B), whereas moderate immunoreactivity was detected in the zona glomerulosa of human adrenal glands and the cytoplasm of E10-control cells (Fig. 1C, D, E). Immunoreactivity for 3β-HSD type 1 was absent in the zona fasciculata of human adrenal gland (Fig. 1C, F). No immunoreactivity was identified in specimens using isotype control antibodies by standard immunohistochemistry (Fig. 1G). 3β-HSD type 1 immunoreactivity was detected in the cytoplasm of breast cancer cells. The number of cases evaluated as 3β-HSD type 1-positive and -negative were 119 (73.9%) and 42 (26.1%), respectively. Figure 2 shows the typical cases evaluated as positive ( Fig. 2A, B) or negative (Fig. 2C, D). Of 118 available cases that had morphologically normal glands adjacent to the cancer tissue, 112 cases (94.9%) had moderate (56 cases; 47.5%) or strong (56 cases; 47.5%) immunoreactivity for 3β-HSD type 1 in morphological normal epithelial cells (data not shown). Moderate (43 cases; 26.7%) or strong (16 cases; 9.9%) immunoreactivity of 3β-HSD type 1 was also detected in stromal cells adjacent to the cancer tissue (data not shown). As mentioned above, cytosolic staining of cancer cells was analyzed in detail in this study.
Correlation between 3β-HSD type 1 expression and breast cancer subtype First, we analyzed 3β-HSD type 1 expression according to ER or HER-2 status. The number of cases expressing immunoreactive 3β-HSD type 1 according to ER and HER-2 status is summarized in Figure 3. Expression of 3β-HSD type 1 was positively correlated with ER-positivity ( Fig. 3A) but not with HER-2 status (Fig. 3B). For further Figure 1. Immunoreactivity of 3β-HSD type 1 in control specimens. Strong immunoreactivity of 3β-HSD type 1 was detected in the syncytiotrophoblast of human placenta (A) and the cytoplasm of E10-HSD3B1 cells (B). Moderate immunoreactivity was detected in the zona glomerulosa of human adrenal glands (C, D) and the cytoplasm of E10-control cells (E). Immunoreactivity for 3β-HSD type 1 was absent in the zona fasciculata of human adrenal gland (C, F). No immunoreactivity was identified in specimens by isotype control antibodies using standard immunohistochemistry (G). Original magnifications were ×200 (A, B, D, E, F, G) and ×100 (C). Bars indicate 300 μm.
Human 3β-hydroxysteroid dehydrogenase type 1 in human breast cancer T. Hanamura et al. analysis, using the ER and HER-2 status of the tumor, breast cancer subtypes were approximated as follows: "luminal" (ER-positive and HER-2-negative), "luminal HER-2" (ER-positive and HER-2-positive), "HER-2" (ERnegative and HER-2-positive), and "TNBC" (ER-negative and HER-2-negative). Cases evaluated as 3β-HSD type 1-negative were most common in TNBC samples without statistical significance (Fig. 3C). Because of these results, we hypothesized that 3β-HSD type 1 has some significant role especially in hormone-dependent breast cancers. Correlation between 3β-HSD type 1 expression and clinicopathological parameters in 130 ER-positive breast cancers Table 2 summarizes the identified associations between 3β-HSD type 1 expression and clinicopathological parameters in 130 ER-positive breast cancers. Of 130 ER-positive cases, the number of cases evaluated as 3β-HSD type 1-positive and -negative were 101 (77.7%) and 29 (22.3%), respectively. 3β-HSD type 1 expression was inversely correlated with invasive tumor size (P = 0.029), presence of invasive region (P = 0.0107),w and lymphatic involvement (P = 0.0004). There was, however, no significant relationship between 3β-HSD type 1 expression and patient age, menopausal status, lymph node metastasis histological grade, ER status, or HER-2 status.
Correlation between 3β-HSD type 1 expression and clinical outcome in 130 ER-positive breast cancers Disease-free survival curves of 130 ER-positive breast cancers are illustrated in Figure 4A. 3β-HSD type 1 expression was associated with a significant decreased risk of Human 3β-hydroxysteroid dehydrogenase type 1 in human breast cancer T. Hanamura et al. recurrence (log-rank test; P = 0.00003). After univariate analysis by Cox (Table 3), lymph node metastasis (P = 0.006), histological grade 3 (P = 0.003) and 3β-HSD type 1-negative (P = 0.0003) were demonstrated to correlate significantly with worse prognosis for disease-free survival in 130 ER-positive breast cancer patients. Multivariate analysis (Table 3) revealed that histological grade 3 (P = 0.020) and negativity of 3β-HSD type 1-negative (P = 0.027) were independent prognostic factors with relative risks of 4.23 and 3.36, respectively, whereas the other factors described above were not significant.

Discussion
In this study, expression of 3β-HSD type 1 was inversely correlated with invasive tumor size, presence of invasive region, and lymphatic involvement. Moreover, our findings indicated that 3β-HSD type 1 is a potent prognostic factor for better outcome in hormone-dependent breast cancer.
In humans, there are two 3β-HSD isoenzymes, which are designated as type 1 and type 2 and encoded by HSD3B1 and HSD3B2 genes, respectively [15]. Type 1 isozyme is predominant in the placenta and peripheral tissues, such as the skin (principally in sebaceous glands), mammary gland, and prostate [29][30][31][32]. In comparison, the type 2 isozyme, which shares 93.5% identity with type 1, is almost exclusively expressed in the adrenal glands, ovary, and testis [29,33,34]. In normal adrenal cortex, it is reported that 3β-HSD type 1 immunoreactivity was essentially confined to the zona glomerulosa. In contrast, 3β-HSD type 2 was not confined to the zona glomerulosa, but was found across the zona fasciculata [24]. Based on these data, we used tissue sections of human placenta and human adrenal gland as positive controls for 3β-HSD type 1 in this study, and our findings were consistent with previous studies. Furthermore, consistent with our expectations, immunoreactivity of 3β-HSD type 1 was stronger in E10-HSD3B1 cells compared with E10-control cells. From these data, the antibody used in this study was considered to have sufficient specificity for 3β-HSD type 1. Enzymatic activity of 3β-HSD-expressing tissue has been reported in human breast cancer tissues [35], and 3β-HSD protein was observed in 36% of breast cancer samples [36]. In mammary gland, sections immunolabeled for 3β-HSD localization, labeling was observed in the cytoplasm of epithelial cells in both the acini and terminal ducts. Immunolabeling was also found in endothelial cells as well as in fibroblasts in the stroma and blood vessels [37]. Our results do not necessarily coincide with previous reports in terms of the positive rate of 3β-HSD in breast cancer tissues because of the different sample number and antibody used for 3β-HSD detection. However, localization of 3β-HSD type 1 in present study is in good agreement with previous studies.
In this study, multivariate analyses revealed that 3β-HSD type 1-negative is an independent prognostic factor, and that relative risks for disease-free survival and diseasespecific survival were 3.36 and 12.23, respectively. These data suggest that the effectiveness of 3β-HSD type 1 as a prognostic marker for is at least equal or higher than other prognostic markers previously reported, which include invasive tumor size, lymph node status, histological grade, PgR status, and HER-2 status [38,39]. However, prospective studies are needed to clarify whether 3β-HSD type 1 can be used as a new prognostic marker of breast cancers in routine practice. The current view is that inhibition of 3β-HSD1 would decrease conversion of DHEA to estrogen precursors or DHT to 3β-diol, to slow ERpositive tumor growth [40]. In our previous report, we suggest that increased expression of HSD3B1 might reduce sensitivity to aromatase inhibitors (AIs) in human breast cancer cell lines, as demonstrated by enhanced 3β-diolinduced ER activation and growth mechanisms [25]. Another study suggested that the steroid-metabolizing pathway activated by 3β-HSD type 1 might function as an alternative estrogenic steroid-producing aromataseindependent pathway in human breast cancers [41]. Therefore, we initially focused on the steroid-metabolizing pathway of 3β-HSD type 1 as a tumor progression factor, or one candidate of the AI-resistance mechanism. However, in this study, expression of 3β-HSD type 1 was inversely correlated with invasive tumor size, presence of invasive region, and lymphatic involvement. Moreover, it was associated with a decreased risk of recurrence in cases that were treated with AI as an adjuvant therapy (n = 44; data not shown), and this result was inconsistent with our previous report. In this study, we found a significant positive correlation between ER positivity and 3β-HSD type 1 expression. We already know that ER+ breast cancers have better outcomes. Therefore, it is suggested that expression of 3β-HSD might just indirectly indicate the hormone-responsiveness of individual breast cancer cases. Further analyses including validation using other breast cancer databases or in vitro experiments are required to resolve this paradox and to understand the function of 3β-HSD type 1 in breast cancer.
As mentioned above, bioactive sex steroids are locally produced from circulating inactive steroids by the comprehensive action of the various steroid-metabolizing enzymes (Fig. 5). AIs are established as a standard treatment option for hormone-dependent breast cancer. It is suggested that STS-17β-HSD type 1 pathway could function as another estrogenic steroid-producing pathway [41]. Indeed, the clinical and prognostic significance of STS in human breast cancer has been reported [28]. Furthermore, the STS pathway has been noted as a therapeutic target, and its clinical application is already underway [12,42]. Expression of the 5α-reductase family in the breast has been reported to be associated with better prognosis in breast cancer patients [43,44]. Thus, many suggestions for therapeutic targets or biological markers of hormonedependent breast cancer have been generated through understanding the intracrine systems of breast cancer. Further understanding of 3β-HSD type 1 may provide insights for the development of novel therapeutic strategies in breast cancer treatment.
In this analysis, we have not studied 3β-HSD type 1 with respect to AR expression. However, it is necessary Human 3β-hydroxysteroid dehydrogenase type 1 in human breast cancer T. Hanamura et al. to extend the analysis to its association with AR expression in the future because this enzyme is an androgenmetabolizing enzyme. The role of 3β-HSD type 1 in ER-negative breast cancer was not sufficiently examined because of the small number of cases. Of 31 ER-negative cases, the number of cases evaluated as 3β-HSD type 1-positive and -negative were 18 (58.1%) and 13 (41.9%), respectively. Although there was no statistical significance, 3β-HSD type 1 expression tended to be inversely correlated with invasive tumor size, presence of invasive region, lymphatic involvement, and lymph node metastasis. Therefore, a possibility that the function of 3β-HSD type 1 is independent of ER must also be considered. As mentioned above, prospective studies with a larger number of cases are needed in the future to promote our further understanding of 3β-HSD type 1.
In summary, we have investigated the clinical and prognostic significance of 3β-HSD type 1, which may regulate in situ production or inactivation of active sex steroids in hormone-dependent breast cancers using IHC. This is the first study to indicate that 3β-HSD type 1 is a potent prognostic factor in hormone-dependent breast cancer.  [TS] and A-dione). Steroid sulfatase (STS) metabolizes estrone sulfate (E1S) and DHEAS to E1 and DHEA, respectively, which are further metabolized to E2 or androst-5-ene-3β,17β-diol (A-diol), respectively, by 17β-hydroxysteroid dehydrogenase type 1 (17β-HSD type 1). dihydrotestosterone (DHT) is synthesized from TS in an irreversible reaction catalyzed by the 5α-reductase family, and is a highly potent androgen. Human 3β-HSD type 1 is a critical enzyme in the conversion of DHEA and A-diol to estrogen precursors A-dione and TS. This enzyme is also involved in conversion of the potent androgen DHT to its inactive form 5α-androstane-3β,17β-diol (3β-diol), which has substantial estrogenic activity.