Personal Medicine and Bone Metastases: Biomarkers, Micro-RNAs and Bone Metastases

Bone metastasis is a major cause of morbidity within solid tumours of the breast, prostate, lung and kidney. Metastasis to the skeleton is associated with a wide range of complications including bone fractures, spinal cord compression, hypercalcaemia and increased bone pain. Improved treatments for bone metastasis, such as the use of anti-bone resorptive bisphosphonate agents, within post-menopausal women have improved disease-free survival; however, these treatments are not without side effects. There is thus a need for biomarkers, which will predict the risk of developing the spread to bone within these cancers. The application of molecular profiling techniques, together with animal model systems and engineered cell-lines has enabled the identification of a series of potential bone-metastasis biomarker molecules predictive of bone metastasis risk. Some of these biomarker candidates have been validated within patient-derived samples providing a step towards clinical utility. Recent developments in multiplex biomarker quantification now enable the simultaneous measurement of up to 96 micro-RNA/protein molecules in a spatially defined manner with single-cell resolution, thus enabling the characterisation of the key molecules active at the sites of pre-metastatic niche formation as well as tumour-stroma signalling. These technologies have considerable potential to inform biomarker discovery. Additionally, a potential future extension of these discoveries could also be the identification of novel drug targets within cancer spread to bone. This chapter summarises recent findings in biomarker discovery within the key bone metastatic cancers (breast, prostate, lung and renal cell carcinoma). Tissue-based and circulating blood-based biomarkers are discussed from the fields of genomics, epigenetic regulation (micro-RNAs) and protein/cell-signalling together with a discussion of the potential future development of these markers towards clinical development.

Laser-capture and microdissected samples from PCa tumours, as well as analysis of the PCa celllines LNCaP and DU145.
Bisulphite sequencing and treatment of cells with demethylating agents.
ER-beta expression is high in normal prostate epithelium and bone metastases but low in grade 4/5 tumours, a phenomenon inversely correlated with ER-beta promoter methylation. [3] Prostate tissues from WT mice and mice with homozygour deletion of the HOXC6-gene.
Genome-wide localization analysis for HOXC6.
HOXC6 regulates both tumour promoters and suppressors related to bone metastasis including BMP7, FGFR2, IGFBP3 and PDGFRA. [4] DTCs from patients with organ confined diseasae and patients with advanced disease.
Katanin-p60 is expressed in bonemetastatic PCa cells, along with basal-cell type markers p63 and [9] high molecular weight cytokeratins.
Pooled serum samples from BPH patients, patients with non-progressing PCa, patients with evidence of biochemical progression and patients with bone metastatic PCa.
25 proteins identified which could distinguish progressing from non-progressing PCa. One protein eukaryotic transalation elongation factor 1 alpha (eEF1A1) validated by IHC as being elevated within osteoblasts adjacent to PCa tumour cells. [10] Conditioned media profiled from three PCa cell-lines -PC3 (bone metastasis), LNCaP (lymph-node metastasis) and 22Rv1 (prostate localized). Validation of markers performed using serum.
2D-LC and MSMS bottomup analysis of conditioned media and serum samples.
7 proteins identified which distinguish PCa from BPH. [13] Urine samples from PCa donors and healthy controls.

CE-MS.
Peptide panel which distinguishes PCa from healthy patients. [14] Urine samples from patients with benign lesions and PCa. SELDI-TOF-MS.
72-peak peptide panel which distinguishes PCa from benign conditions.

[15]
Urine samples from PCA patients Mutations within PIK3Ca and loss of PTEN correlate with shorter survival times and reduced efficiacy of treatment. [18] Breast tumour-cells lines cultured +/-DNAdemthylating agents.

cDNA-Microarray
Essential role for Src and AKT pathways within survival of BCa metastatic cells. [23] MDA-MB-231 cell-line and variants with increasing degrees of bone homing ability.
MSP and bisulphite sequencing and cDNAmicroarray.
Upregulation of miR-224 with resulting down-regulation of RKIP expression and dereprepression of MMP1, CXCR4 and OPG. [26] Publicly available geneexpression datasets from breast cancer metastasis studies.
Computational analysis of gene expression datasets.
SMAD4 and HIF1 identified as regulators of bone metastasis, BACH1 identified as a master regulator of bone-homing. [27] Breast cancer metastases to bone, brain and lung.
cDNA-microarray analysis with network-based classification analysis.
A protein-protein network identified with metastasis to bone mainly involving immuneresponse proteins. [28]
GRO and IL-8 expression levsl alter in response to HER2 overexpression. Treatment with the TKI-gefitinib reverses these alterations. [29] Breast tumour derived celllines and tissue microarrays.

SILAC-MS followed by validation by IHC of TMAs.
Protein-expression signature predictive of breast cancer stage within ER-negative BCa. Signature validated by TMA studies. [30] MDA-MB-231 cell-line and bone homing variants. LFQ-MS.
Elevated levels of MMP -2, -7 and -14 and TIMP-3 and low expression of MMP-9 are predictive of breast cancer metastasis. [32] Breast cancer cell line series from low-grade through to high-grade.
iTRAQ Warburg-effect proteins elevated as well as altered expression of adenylate-kinase-1, coppertransport-protein ATOX1 and histone H2B type 1M. [33] Breast cancer section TMAs. TMA-analysis Identification of MMP1 and ADAMTS1 as active enzymes within the release of soluble EGFlike ligands promoting breast cancer metastasis. [34] Primary breast cancer specimens. IHC within TMAs.
Altered expression of serpins, laminin chains and MMPs within subclasses of tumours correlating with altered invasiveness. elevated levels of integrins and metalloproteases as well as reduced expression of laminins correlates with poor prognosis. [35]

Lung Cancer Genomic/functional genomic studies Study samples Method used Key findings Reference
Analysis of lung cancer celllines.
The Nm23-H1 gen suppresses expression of micro-RNA. miR-660-5p regulates metastasis via regulation of SMARCA5. [36] Lung cancer samples from 8 NSCLC patients with linked data relating to bone metastasis occurrence.
PTHrP levels predict bone metastatic outcomes, hypercalcaemia and survival. Eight proteins involved in mitochondrial metabolism altered in bone homing cells. [44] Patient derived samples (tissues and serum).