Expression of the RNA Helicase DDX3 and the Hypoxia Response in Breast Cancer

Aims DDX3 is an RNA helicase that has antiapoptotic properties, and promotes proliferation and transformation. In addition, DDX3 was shown to be a direct downstream target of HIF-1α (the master regulatory of the hypoxia response) in breast cancer cell lines. However, the relation between DDX3 and hypoxia has not been addressed in human tumors. In this paper, we studied the relation between DDX3 and the hypoxic responsive proteins in human breast cancer. Methods and Results DDX3 expression was investigated by immunohistochemistry in breast cancer in comparison with hypoxia related proteins HIF-1α, GLUT1, CAIX, EGFR, HER2, Akt1, FOXO4, p53, ERα, COMMD1, FER kinase, PIN1, E-cadherin, p21, p27, Transferrin receptor, FOXO3A, c-Met and Notch1. DDX3 was overexpressed in 127 of 366 breast cancer patients, and was correlated with overexpression of HIF-1α and its downstream genes CAIX and GLUT1. Moreover, DDX3 expression correlated with hypoxia-related proteins EGFR, HER2, FOXO4, ERα and c-Met in a HIF-1α dependent fashion, and with COMMD1, FER kinase, Akt1, E-cadherin, TfR and FOXO3A independent of HIF-1α. Conclusions In invasive breast cancer, expression of DDX3 was correlated with overexpression of HIF-1α and many other hypoxia related proteins, pointing to a distinct role for DDX3 under hypoxic conditions and supporting the oncogenic role of DDX3 which could have clinical implication for current development of DDX3 inhibitors.


Introduction
In the Western world, one in eight women will develop breast cancer during their life and breast cancer is causing about 458.000 deaths worldwide per year [1,2]. Aggressive forms of breast cancer are frequently refractory to treatment [3], even to established targeted therapy, and thus have a high risk of relapse and formation of distant metastases [4]. Identification of molecular pathways involved in aggressive forms of breast cancer is therefore important to design novel targeted therapeutic agents to counteract tumor progression and metastasis.
A recent in vitro study [21] showed that DDX3 is a direct downstream target of HIF-1a, the predominant factor in the mammalian hypoxia response [22]. Hypoxia is an important event in breast carcinogenesis [23][24][25][26], causing a more aggressive phenotype with increased invasiveness and proliferation, formation of metastases, resistance to therapy [27] and poorer survival [28,29].

Patients
Representative paraffin embedded tissue blocks of 422 breast cancer patients collected between 2004 and 2007 were taken from the archive of the Department of Pathology of the University Medical Centre in Utrecht and routinely processed to four tissue microarrays (TMA) as described before [56,57].
Use of anonymous or coded left over material for scientific purposes is part of the standard treatment contract with patients in the UMCU [63].

Scoring of Immunohistochemistry
Scoring was done by a single experienced pathologist (PJvD). Intensity of cytoplasmic DDX3, FER kinase and membranous Ecadherin, TfR and c-Met was scored semi-quantitatively from 0-3 and percentages of cells with nuclear DDX3 and Notch1 expression were estimated. Out of three cores from the same patient, the maximum cytoplasmic DDX3 score was used for further analysis.
DDX3 scores 1 and 2 were grouped as low DDX3 expression and evaluated against high DDX3 expression (scores 3). For Ecadherin, TfR, c-Met and FER kinase scores 0 and 1 were defined as low expression versus score 2 and 3 as high expression. For HIF-1a, the 1% threshold was used as before [59].

Statistics
Expression levels of DDX3 and the other proteins were compared by chi-square test or t-test whenever applicable. Logistic regression or ANCOVA was used for multivariate analysis to determine dependence of these relations on HIF-1a.
Since EGFR and HER2 are upstream regulators of HIF-1a via PI-3K/AKT, we also assessed the relation of EGFR and HER2 with DDX3 independent of Akt1 and HIF-1a. In lobular breast cancer there is very little or no expression of E-cadherin, so the lobular cancers were excluded in analysis with respect to Ecadherin.
Pearson correlation coefficient was determined for correlation analysis.

Discussion
The aim of this study was to investigate the relation between DDX3 and the hypoxic response in human breast cancer in the light of in vitro results pointing to regulation of DDX3 by HIF-1a. We indeed show a positive correlation between HIF-1a and DDX3 overexpression in a large series of human breast cancer cases, as well as an association between DDX3 overexpression and various other hypoxia related proteins.
However, we have established a correlation between DDX3 overexpression and nuclear HIF-1a overexpression which supports the direct regulation of DDX3 by HIF-1a found in vitro [21], but this is obviously no more than an association at this point no proof for a causal relationship. Immunohistochemistry has some limitations like being inherently a more qualitative than quantitative method, and semiquantitative scoring and dichotomization with non-optimal reproducibility. To compensate for these issues we standardized the IHC procedure, used control tissue throughout, scored three samples per patient, studied a large cohort of breast cancer patients and results obtained from dichotomized parameters were confirmed by correlation analysis for the most important parameters with the DAKO score of DDX3 (table 5). Patient features in this study corresponded with known clinicopathological characteristics in breast cancer (table 1) [65].
Furthermore, DDX3 correlated with EGFR, HER2, FOXO4, ERa and c-Met in a HIF-1a dependent way. Also, we found a positive correlation with COMMD1, FER kinase, Akt1, Ecadherin, TfR and FOXO3A independent of HIF-1a. COMMD1 down regulates HIF-1a by competition with HSP90b [41], or  down regulates the transcriptional activity of HIF-1a [40]. However, we could not detect an association with COMMD1 and HIF-1a expression or its downstream targets: E-cadherin, TfR, p21, p27 or c-Met. Nonetheless, COMMD1 correlates with DDX3 independent of HIF-1a. FER kinases help cells to withstand stress, including hypoxia, via up regulation of HIF-1a [42]. We found a strong relation with FER kinase with both HIF-1a and DDX3. After correction for the effect FER kinase has on HIF-1a, a strong relation between FER kinase and DDX3 remained, implying a HIF-1a dependent and independent relation.
DDX3 was shown to down regulate E-cadherin [12], but in the present study we show a positive correlation, for which we have no obvious explanation. TfR is also under transcriptional control of HIF-1a [47]. TfR is overexpressed in many cancers, which could be attributed to the increased need for iron as a cofactor of the ribonucleotide reductase enzyme involved in DNA synthesis of rapidly dividing cells. Thus, the HIF-1a independent relation between DDX3 and TfR corroborates previous reports on the oncogenic properties of DDX3. Nuclear expression of FOXO3A in breast cancer is associated with anti-apoptotic signaling via Akt1, an aggressive phenotype and poor survival [66]. In response to hypoxia, FOXO3A accumulates in a HIF-1a dependent way to inhibit HIF-1a induced apoptosis [48]. Although we did not find a relation between HIF-1a and FOXO3A we did find a relation between FOXO3A and DDX3, independent of HIF-1a and Akt1. Perhaps DDX3 and FOXO3A function in a concerted survival response after stress stimuli.
In conclusion, ten of eighteen proteins analyzed by IHC showed a similar HIF-1a related effect as described in the literature. All these ten HIF-1a related proteins were associated with expression of DDX3 as well, indicating an important role for DDX3 in the hypoxia response via HIF-1a, and underlying the oncogenic role of DDX3. Since hypoxic tumor regions are typically resistant to current therapy [27], this emphasizes the potential of DDX3 inhibitors, perhaps in combination with HER2 and/or EGFR inhibitors.