Do intestinal tumors arise from mutation(s) in a single somatic progenitor cell and its descendants? This fundamental question in cancer biology has actually been addressed by only a few studies over the past four decades (1-3). The consensus was that human colonic tumors are monoclonal until Novelli and colleagues demonstrated that the majority of early adenomas isolated from an XO/XY mosaic individual with familial adenomatous polyposis (FAP) have a polyclonal structure (4). A major criticism of this study was the inherent instability of the XO/XY karyotype; this concern was overcome in a study by Merritt and colleagues in which early adenomas were analyzed from C57BL/6J (B6) mice carrying the Min allele of Apc and chimeric for ROSA26 (R26), a ubiquitously expressed cell lineage marker (5). The majority of tumors in this mouse model of FAP were polyclonal. Thus, tumors that develop in the mammalian intestine may be derived from two or more populations of neoplastic cells, but this structure was largely concealed in earlier studies because it was undetectable with the technology of the time.
The high tumor multiplicity of the mice in the study of Merritt and colleagues made it difficult to eliminate the possibility that polyclonality arose by random collision of distinct initiated clones as opposed to some form of clonal interaction. My colleagues and I tested the random collision hypothesis by using Min mice that were also homozygous for the tumor suppression allele of Mom1. Despite the markedly reduced tumor multiplicity in these mice, the observed fraction of overtly polyclonal heterotypic adenomas was significantly higher than predicted by the random collision hypothesis (6).
- My central hypothesis is that interactions among initiated clones within a single tumor in the mammalian intestine provide a selective advantage during formation, growth, and progression. Members of my laboratory will test this hypothesis by addressing three fundamental questions:
- Does the polyclonality of early intestinal tumors depend on the pathway of tumorigenesis? We will analyze tumors from mice treated with either ethylnitrosourea or azoxymethane, mice in which tumorigenesis is initiated because of a mutation in the TGF-beta signaling pathway, and mice in which Apcis inactivated somatically by silencing. If interactions among transformed clones within a single tumor provide a selective advantage, polyclonality should be common in all of these distinct mouse models.
- How do polyclonal tumors form in the mammalian intestine? In models in which polyclonality is evident, we will test several different hypotheses that could explain the formation of heterotypic tumors. Our initial study of Min mice chimeric for R26 expression indicated that the most likely explanation involved clonal interactions occurring over one to two crypt diameters. We will extend this work by attempting to determine the nature of clonal interactions using statistical approaches. An initiated clone might transform one or more neighboring clones (recruitment). Alternatively, polyclonal tumors might form because two or more independently initiated clones are simply juxtaposed (cooperation).
- Does polyclonality persist as adenomas progress to invasive adenocarcinomas in the intestine? We will monitor longitudinally tumors in three mouse models using two complementary state-of-the-art imaging platforms, microCT and optical colonoscopy. Preliminary data indicate colonic tumors in Min mice have different fates: some grow and progress, some remain static, and some regress. Our colleagues have demonstrated that this is also true of colonic tumors in humans (7). The different fates may simply reflect different clonal structures.
These studies will rigorously test the long-held view that colonic tumors are derived from a single progenitor and progress to malignancy as mutations drive successive rounds of clonal expansion. A combination of newly developed mouse models, statistical analyses, and technologies will be employed. My R01 proposal focused on these studies of tumor polyclonality recently received an outstanding score (Priority Score 121; Percentile 0.9%). The results could profoundly change our understanding of tumorigenesis in the mammalian intestine. In 2000, Hanahan and Weinberg stated, "those researching the cancer problem will be practicing a dramatically different type of science than we have experienced over the past 25 years. Surely much of this change will be apparent at the technical level. But ultimately, the more fundamental change will be conceptual" (8). A deep understanding of this issue, the importance of clonal interactions for intestinal neoplasia, can impact the development of management strategies for this disease.