Purpose
Primary breast and prostate cancers can be cured, but metastatic disease cannot. Identifying cell factors that predict metastatic potential could guide both prognosis and treatment.
Methods
We used Cell-SELEX to screen an RNA aptamer library for differential binding to prostate cancer cell lines with high vs. low metastatic potential. Mass spectroscopy, immunoblot, and immunohistochemistry were used to identify and validate aptamer targets. Aptamer properties were tested in vitro, in xenograft models, and in clinical biopsies. Gene expression datasets were queried for target associations in cancer.
Results
We identified a novel aptamer (Apt63) that binds to the beta subunit of F
1
F
o
ATP synthase (ATP5B), present on the plasma membrane of certain normal and cancer cells. Apt63 bound to plasma membranes of multiple aggressive breast and prostate cell lines, but not to normal breast and prostate epithelial cells, and weakly or not at all to non-metastasizing cancer cells; binding led to rapid cell death. A single intravenous injection of Apt63 induced rapid, tumor cell-selective binding and cytotoxicity in MDA-MB-231 xenograft tumors, associated with endonuclease G nuclear translocation and DNA fragmentation. Apt63 was not toxic to non-transformed epithelial cells in vitro or adjacent normal tissue in vivo. In breast cancer tissue arrays, plasma membrane staining with Apt63 correlated with tumor stage (
p
< 0.0001,
n
= 416) and was independent of other cancer markers. Across multiple datasets, ATP5B expression was significantly increased relative to normal tissue, and negatively correlated with metastasis-free (
p
= 0.0063, 0.00039, respectively) and overall (
p
= 0.050, 0.0198) survival.
Conclusion
Ecto-ATP5B binding by Apt63 may disrupt an essential survival mechanism in a subset of tumors with high metastatic potential, and defines a novel category of cancers with potential vulnerability to ATP5B-targeted therapy. Apt63 is a unique tool for elucidating the function of surface ATP synthase, and potentially for predicting and treating metastatic breast and prostate cancer.
Electronic supplementary material
The online version of this article (10.1007/s10549-019-05174-3) contains supplementary material, which is available to authorized users.