Gene expression in local stroma reflects breast tumor states and predicts patient outcome

Abstract

The surrounding microenvironment has been implicated in the progression of breast tumors to metastasis. However, the degree to which metastatic breast tumors locally reprogram stromal cells as they disrupt tissue boundaries is not well understood. We used species-specific RNA sequencing in a mouse xenograft model to determine how the metastasis suppressor RKIP influences transcription in a panel of paired tumor and stroma tissues. We find that gene expression in metastatic breast tumors is pervasively correlated with gene expression in local stroma of both mouse xenografts and human patients. Changes in stromal gene expression elicited by tumors better predicts subtype and patient survival than tumor gene expression, and genes with coordinated expression in both tissues predict metastasis-free survival. These observations support the use of stroma-based strategies for the diagnosis and prognosis of breast cancer.

RKIP as an Inflammatory and Immune System Modulator: Implications in Cancer

“…Furthermore, in a broader study, Bainer et al demonstrated, using species-specific RNA sequencing in a xenograft TNBC (Triple-negative breast cancer) mouse model, that gene expression in metastatic breast tumours is largely correlated with gene expression in local stroma of both mouse xenografts and human patients. In addition, changes in stromal gene expression elicited by tumours with or without RKIP expression is a better predictor of breast cancer subtype and patient survival than tumour gene expression [73]. In a recent study, Buschow et al revealed that RKIP expression, during immunotherapy of metastatic melanoma with dendritic cell (DC) vaccination, positively correlated with gene signatures involved in effective T cell responses, but inversely correlated with genes associated with myeloid cell infiltration and inflammation, such as STAT3, Notch1, and MAPK1 signalling members [74].…”

Targeting Raf Kinase Inhibitory Protein Regulation and Function

“…More broadly, using species-specific RNA sequencing in a xenograft TNBC mouse model, Bainer et al demonstrated that gene expression in metastatic breast tumors is widely correlated with gene expression in local stroma of both mouse xenografts and human patients [ 6 ]. Moreover, changes in stromal gene expression elicited by tumors that do or do not express RKIP is a better predictor of breast cancer subtype and patient metastasis-free survival than tumor gene expression.…”

“…Exploring and expanding RKIP-regulated gene networks using patient-derived expression data also allow for developing predictive and prognostic gene signatures that have mechanistic relevance to specific patient cohorts. After identifying RKIP-regulated signaling networks in tumors and their microenvironment, the Rosner group generated several RKIP pathway-based gene signatures for predicting metastasis-free survival of breast cancer patients [ 6 , 7 , 8 , 9 , 10 ]. The biomarkers used in these gene signatures span a wide range of metastatic processes such as extracellular matrix remodeling, epigenetic regulation, tumor-associated macrophage recruitment, and reprogramming of stromal gene expression by tumors.…”

“…Experimentally, over-expressing RKIP blocks invasion in vitro and metastatic progression in vivo [ 4 , 5 ] without affecting growth properties of the primary tumor. RKIP-related gene signatures can be utilized to identify cancer patients at highest risk [ 6 , 7 , 8 , 9 , 10 ]. These clinical and experimental findings establish RKIP as a suppressor of invasion, intravasation, extravasation and metastasis.…”

An antibody-based approach for the in situ evaluation of mouse contribution in human stem cell-derived xenografts

“…Currently, many different approaches exist, which do not fullfill this set of criteria. The vast majority of them, including DNA 8,9 -and RNA [10][11][12][13] -based PCR and reverse transcription quantitative real-time PCR (RT-qPCR), species-specific transcriptome [14][15][16][17][18][19][20] and proteome 21,22 analyses as well as flow cytometry [23][24][25] , are molecular profiling approaches which require tissue dissociation or lysis and do not address host contribution in situ on a single cell level. To investigate the unperturbed interaction between the tumor and its microenvironment, the use of genetically modified cells or mouse models expressing fluorescent reporter proteins [26][27][28][29][30] might want to be avoided.…”