Nuclear expression of KLF6 tumor suppressor factor is highly associated with overexpression of ERBB2 oncoprotein in ductal breast carcinomas

Abstract

Background: Krüppel-like factor 6 (KLF6) is an evolutionarily conserved and ubiquitously expressed protein that belongs to the mammalian Sp1/KLF family of transcriptional regulators. Though KLF6 is a transcription factor and harbors a nuclear localization signal it is not systematically located in the nucleus but it was detected in the cytoplasm of several tissues and cell lines. Hence, it is still not fully settled whether the tumor suppressor function of KLF6 is directly associated with its ability to regulate target genes.

Methodology/principal findings: In this study we analyzed KLF6 expression and sub-cellular distribution by immunohistochemistry in several normal and tumor tissues in a microarray format representing fifteen human organs. Results indicate that while both nuclear and cytoplasmic distribution of KLF6 is detected in normal breast tissues, breast carcinomas express KLF6 mainly detected in the cytoplasm. Expression of KLF6 was further analyzed in breast cancer tissues overexpressing ERBB2 oncoprotein, which is associated with poor disease prognosis and patient's survival. The analysis of 48 ductal carcinomas revealed a significant population expressing KLF6 predominantly in the nuclear compartment (X(2)p = 0.005; Fisher p = 0.003). Moreover, this expression pattern correlates directly with early stage and small ductal breast tumors and linked to metastatic events in lymph nodes.

Conclusions/significance: Data are consistent with a preferential localization of KLF6 in the nuclear compartment of early stage and small HER2-ERBB2 overexpressing ductal breast tumor cells, also presenting lymph node metastatic events. Thus, KLF6 tumor suppressor could represent a new molecular marker candidate for tumor prognosis and/or a potential target for therapy strategies.

Figure 1. Immunohistochemical assay for KLF6 in placenta tissue section.

Immunohistochemistry assays were performed as described in Material and Methods. The specific stain is shown as a brown precipitate and nuclei were blue counterstained with hematoxylin. Magnification is 600×. A. Stain pattern obtained for the commercial anti-KLF6 polyclonal antibody (Zf9, top) and for in house produced anti-KLF6 monoclonal antibody (clone 2C11, bottom) in term placenta tissue sections. B. Background control of the secondary detection complex EnVision® System Peroxidase Labeled Polymer (Dako, Carpinteira, USA). The placenta tissue section was incubated only with primary antibody dilution buffer, and the secondary detection complex mentioned above.

Figure 2. KLF6 and ERBB2 expression in breast cancer tissues.

Immunohistochemistry assays in two lots of Tumor tissues Microarray from DAKO (CheckBoard Tissue Tumor human Tissues, T1064, Dako, USA). Tissues were incubated with anti-KLF6 (clone 2C11, left panel) or anti-ERBB2 (right panel) antibodies. Lot 1 (#00071) is represented in the panel L1, and Lot 2 (#00111) in panels L2a, L2b and L2c. The KLF6 and ERBB2 stains were photographed in the equivalent area of the tissue section at 400× of magnification.

Figure 3. KLF6 and ERBB2 expression in ductal breast cancer tissues.

Immunohistochemistry assays in LandMark LD Breast TMA tissue microarray (Ambion, Austin, TX, USA). Tissues were incubated with the anti-KLF6 and anti-ERBB2 antibodies. The left panel of the KLF6 and ERBB2 column pictures has 400× magnification. The square into each 400× picture indicates the area photographed at 1000× and showed in the right panel of each column. Photomicrographs of the immunostain for KLF6 and ERBB2 were taken from equivalent tissue areas. The horizontal panels represents the KLF6 and ERBB2 expression in a fibroadenoma tissue (i) and four representative cases for both positive nuclear KLF6 and ERBB2 (ii), negative nuclear KLF6 and positive ERBB2 (iii), positive nuclear KLF6 and negative ERBB2 (iv), and both negative nuclear KLF6 and ERBB2 (v). Tissues in ii–v correspond to the cases 40, 18, 45, and 6, respectively, described in Tables S2 and S3.

Figure 4. Association between nuclear KLF6 expression and ERBB2-overexpression in ductal breast tumor tissues.

The immunohistochemical staining intensity for KLF6 (nuclear and cytoplasmic) and c-erbB2 (cytoplasmic membrane) is represented as scores used a 0.0–4.0 scale estimated over a characteristic cell cluster of each tissue. The scores estimation was obtained as described , (see Material and Methods). A score of 1.0 represent less than 10% of positive labeled cells and was set as the cutoff value. A. Scatter plot analysis of nuclear KLF6 versus cytoplasmic KLF6. The oval includes positive tissue for both staining patterns. B. Scatter plot analysis of nuclear KLF6 versus ERBB2 immunostains. Filled lines settled at 0.0 and 1.0 score indicate cutoff values and partite four populations determined by the KLF6 and ERBB2 staining. The percentage of each population is represented into a gray circle. Especially for the double-negative population (minus 1.0 and 0.0 scores) a point could represent two or more cases.