This study used a unique xenogeneic breast cancer model to study the effects of tumor cells and neighboring host cells upon each other in tumor growth and metastasis. It exploited species differences between the interacting components to determine how the host influenced the tumor and vice versa. It was found that the gene expression profiles of highly and poorly metastatic clones from the same human breast carcinoma changed differentially when the cells were transferred from growth in vitro to the mammary gland. We describe novel sets of genes, validated by human-specific probes, which were induced in the 2 isogenic, but phenotypically different, tumor lineages by the mammary environment. Conversely, the tumor cells also induced changes in gene expression in the neighboring host stromal (i.e., mesenchymal) cell lineages, validated by mouse-specific probes. Reciprocal inductive interactions were also demonstrated in the tumor deposits formed preferentially in the lungs and lymph nodes by the highly metastatic tumor cells. Subtraction of the induced gene changes in the primary site from those in the metastases revealed that the number and magnitude of specific gene inductions in colonized organs were moderate. This finding indicates that the gene expression program causing metastasis has only limited flexibility and fits well with clinical observations that tumor cells form metastases preferentially in select organs, although tumor cells are scattered ubiquitously. This dependency on suitable host niches suggests new molecular therapeutic avenues that target genes in the host-support system that is manipulated by the malignant cells. To survive and prevail, the intruding cells must co-opt local conditions to provide their needs. This prompts questions about what interactions and mechanisms are activated to promote this disorderly behavior and how these are normally suppressed.The discoveries of Spemann and colleagues, 1 that cells of different embryonic lineages actively induce each other to cooperate in the formation of tissues, organs and body regions have far reaching applications in solving such fundamental but clinically relevant questions. Their findings provided a rational framework for explaining how specialized cell lineages interact to produce and maintain the anatomic plan of the whole animal and the microscopic architecture of its individual organs throughout life and healing processes. Conversely, their observations also have important implications for understanding mechanisms underlying the progressive histological disorganization which characterizes cancer pathogenesis and progression to metastatic malignancy. Important steps in recognizing the importance of cell and tissue interactions in neoplasia include the work of Orr that showed that morphological 2 and functional 3 changes occur in the dermis of carcinogen-treated skin long before a carcinoma develops from the overlying epidermis. Later, Grobstein 4 and Sengel et al. 5 showed that the formation of the microscopic anatomy and histology of org...