Brain-derived neurotrophic factor expression predicts adverse pathological & clinical outcomes in human breast cancer

Brain-derived neurotrophic factor (BDNF) belongs to the neurotrophin (NT) superfamily of polypeptide growth factors, which includes nerve growth factor (NGF) and NTs 3-6 [1, 2]. NTs and their receptors have key physiological roles in the development and function of the central and peripheral nervous systems in vertebrates [3–5]. However, they are also widely expressed in non-neuronal tissues [6]. Expression of the BDNF gene (BDNF) is regulated by the presence of multiple activity dependent and tissue-specific promoters [7]. BDNF signals preferentially via its high affinity tyrosine kinase receptor, tropomyosin receptor kinase B (TrkB). Ligand-induced receptor dimerisation results in auto-phosphorylation and initiates multiple signalling cascades, including the mitogen-activated protein kinase (MAPK), phosphatidyl-inositide 3-kinase (PI3K) and phospholipase C-gamma (PLC- γ) pathways, that promote cellular survival [8–11]. However, BDNF also shares a common low affinity receptor (p75^NTR) with the other NTs, which is a member of the tumor necrosis factor (TNF) receptor superfamily, implicated in the modulation of cell survival, cell cycle regulation and cytoskeletal rearrangement [8]. The overall cellular response to BDNF exposure is therefore likely to reflect an equilibrium between TrkB and p75^NTR activity.

BDNF has been implicated in various human pathologies, including: depression, epilepsy, Alzheimer's, Parkinson's and Huntington's disease [7, 12]. BDNF has also been associated with several human cancers, both neuronal and non-neuronal, including: neuroblastoma [13], myeloma [14], ovarian [15, 16], lung [17], prostate [18], hepato-cellular [19], pancreatic [20][21, 22], head and neck squamous cell carcinomas [23] and pulmonary carcinoid tumours [24]. Interestingly, the archetypal neurotrophic factor NGF has been demonstrated to stimulate proliferation, angiogenesis and behave as an anti-apoptotic factor in human breast cancer (BC) [1, 25–27] with potential for therapeutic targeting [28]. In keeping with this, BDNF has been associated with cell survival in human BC cell lines [1]. BDNF has also been demonstrated to be significantly up regulated in oestrogen receptor alpha (ER-α) positive BCs [29]. Although increased NT and cognate receptor expression have been demonstrated in BC [2, 30], the precise biological role of BDNF and its utility as a novel biomarker have yet to be determined. The objective of this study was to determine the mRNA and protein expression of BDNF in a cohort of women with BC. Expression levels were compared with normal background tissues and evaluated against established pathological parameters and clinical outcome over a 10 year follow-up period.

NTs and their receptors are increasingly being implicated as novel mediators of carcinogenesis in neuronal and non-neuronal tissues. Whilst the literature regarding BC remains sparse, altered expression and function of these factors are likely to contribute to tumourigenesis and progression. The present study adds to the literature in support of the oncogenic function of BDNF in BC. Furthermore, this study is the first to quantitatively evaluate BDNF mRNA expression in a large cohort of BC patients and provide correlation with conventional pathological parameters and clinical outcomes over an extended follow-up period. Significantly higher mRNA transcript levels were found in BC specimens compared to normal tissue, corroborated at the protein level by immuno-histochemical staining. BDNF expression was found to increase with increasing NPI, nodal positivity, LR, death from BC and poor prognosis overall. After a median follow up of 10 years, higher BDNF expression levels were significantly associated with reduced OS. Furthermore, BDNF remained an independent prognostic variable in multivariate analysis. Our results differ from those of Blasco-Gutierrez et al. [2] who did not identify differential staining between tumour and normal breast tissue and reported no association between BDNF expression and pathological parameters or clinical outcomes. In their study, Tozlu et al. [29] employed real-time quantitative reverse transcription (RT)-PCR to compare the mRNA expression of 560 selected genes in BCs excised from 48 women. BDNF emerged as a growth factor significantly up regulated amongst ER-α positive tumours in both the screening and validation sets. However, the relationship was found to be reversed in vitro, where BDNF was found to be up regulated in ER-α negative BC cell lines. Interestingly cross-talk between steroid and growth factor pathways has been identified in the brain, with oestrogen implicated in the regulation of BDNF and demonstrating some degree of overlap in activity with NTs [34]. BDNF has also been associated with increased cell survival, although not proliferation, in human BC cell lines. In accordance with the fact that BDNF did not demonstrate a mitogenic effect, no expression of TrkB was found in any of the BC cells tested. However, the interaction between BDNF and P75^NTR, which specifically induces NF-ϰB, has been implicated in the protection of BC cells from apoptosis [1]. Further support for the oncogenic function of BDNF in vivo comes from a transgenic mouse model of spontaneous mammary tumor formation, following exposure to the organochlorine pesticide Dieldrin, which is a persistent environmental toxin thought to increase the risk of BC [35]. Offspring of these mice show enhanced development of mammary tumours with increased mRNA and protein expression of BDNF and TrkB [36].

The factors regulating NT and NT receptor expression in normal and malignant breast tissues remain unknown. Whilst the precise biological role of the BDNF/TrkB/P75^NTR axis in human BC has yet to be elucidated, supporting evidence of oncogenic function may be inferred from other malignancies. TrkB has been implicated in the pathogenesis of neuroblastoma where expression is correlated with poor outcome and a chemo-resistant phenotype [37, 38], increased mortality in Wilm's tumor [39], lymph node metastasis and advanced stage in non-small cell lung cancer [40], shorter survival in ovarian cancer [16][15] and distant metastases and poor prognosis in gastric cancer [41]. TrkB has also been found to regulate migration, invasion and epithelial-mesenchymal transition in head and neck squamous cell carcinoma [23]. Signalling via the BDNF/TrkB pathway stimulates pro-survival signals, resistance to anoikis and altered cellular aggregation, all features of cancer cells and prerequisites of metastases formation [42]. Emerging roles for NTs in angiogenesis provide further insight into their potential relevance to cancer development and progression [43]. Studies have identified BDNF and TrkB as key mediators of vascular development [44]. BDNF is an endothelial survival factor, deficiency of which results in reduced endothelial cell-cell contacts and apoptosis [45]. BDNF has been implicated as a novel angiogenic protein in multiple myeloma [14, 46]. In support of this, BDNF activation of TrkB has been found to induce vascular endothelial growth factor (VEGF) expression via hypoxia inducible factor-1-alpha (HIF-1-α) in neuroblastoma cells [47]. It has also been suggested that increased expression of TrkB/BDNF may indicate increased neo-neurogenesis, where tumors initiate their own innervation by releasing neurotrophic factors [48]. This may further support tumour progression by the release of neurotransmitters which enhance metastasis. Indeed, the presence of nerve cell markers in tumoral tissue has been found to be a prognostic marker in several human malignancies [49]. In addition to utility as a prognostic marker, therapeutic manipulation of the TrkB signal transduction pathway appears to be an increasingly important target in cancer biology which merits further exploration [11, 50]. Despite the inferences drawn, the mechanisms through which BDNF exerts its oncogenic activity have yet to be adequately determined and this will undoubtedly be necessary to optimise any potential therapeutic applications.

Limitations of the present study include the use of background parenchyma from BC patients to provide 'normal tissue' for comparison is also contentious. Ideally, such material should be derived from patients without BC in order to avoid any 'field change' which may exist within cancer bearing tissues. Although the sample size and follow-up period were substantial, it is possible that a larger cohort, particularly with regard to subgroup analysis, may have influenced several results which approached, but failed to reach, statistical significance. In addition to the measurement of mRNA transcript levels and qualitative immuno-histochemistry, quantitative analysis of protein expression should be undertaken to ensure concordance. Correlation with associated molecules, in particular TrkB and P75^NTR, and other markers of invasiveness and metastatic competence would also be of value.

BDNF expression was found to be significantly higher in BC specimens compared to normal tissues. Higher transcript levels were significantly associated with unfavourable pathological parameters including nodal positivity and increasing NPI; and adverse clinical outcomes including LR, death from BC, poor prognosis, reduced DFS and OS. In addition to the prognostic utility of BDNF, further mechanistic studies are warranted to explore the potential for therapeutic manipulation in human BC.