Abstract
The growth of human tumor cells transplanted into immunodeficient mice is frequently studied to gain understanding about the way potential drug treatments interfere with growth in vivo. A wide range of methods is available for learning about specific aspects of tumor cell behavior, for example, cells may be administered to follow their ability to grow close to the site of injection which may be at a generic site or one specific to that type of tumor. Some models of metastasis follow the appearance of a tumor mass after intravascular administration of tumor cells; others score remote growth after removal of a primary tumor implanted subcutaneously. Assessing metastatic growth may increasingly rely on serial observation of tumor cell numbers as seen by whole-body imaging, but the sensitivity of these methods is poor in terms of the minimum number of cells detectable, and histological follow-up to establish tumor cell numbers can be confounded by variable expression or even silencing of reporter genes. Here we describe how fluorescence in situ hybridization (FISH) using commercially available probes can very easily be used to detect even single metastatic tumor cells in mouse models, using routinely fixed and processed tissue samples, and without the tumor cell lines needing to express engineered reporter genes. The FISH protocol can be combined with other standard histological protocols to study the behavior of tumor cells and adjacent host cells to improve our understanding of tumor-stroma interactions, and is also useful for simultaneous demonstration of the cell of origin and phenotype of cells used in regenerative medicine-based applications.