Somatic clonal evolution: A selection-centric perspective

Abstract

It is generally accepted that the initiation and progression of cancers is the result of somatic clonal evolution. Despite many peculiarities, evolution within populations of somatic cells should obey the same Darwinian principles as evolution within natural populations, i.e. variability of heritable phenotypes provides the substrate for context-specific selection forces leading to increased population frequencies of phenotypes, which are better adapted to their environment. Yet, within cancer biology, the more prevalent way to view evolution is as being entirely driven by the accumulation of “driver” mutations. Context-specific selection forces are either ignored, or viewed as constraints from which tumor cells liberate themselves during the course of malignant progression. In this review, we will argue that explicitly focusing on selection forces acting on the populations of neoplastic cells as the driving force of somatic clonal evolution might provide for a more accurate conceptual framework compared to the mutation-centric driver gene paradigm. Whereas little can be done to counteract the “bad luck” of stochastic occurrences of cancer-related mutations, changes in selective pressures and the phenotypic adaptations they induce can, in principle, be exploited to limit the incidence of cancers and to increase the efficiency of existing and future therapies. This article is part of a Special Issue entitled: Evolutionary principles - heterogeneity in cancer?, edited by Dr. Robert A. Gatenby.

Introduction

According to Ernst Mayr, biological causation can be separated into proximal causes that answer the “how” questions and ultimate causes that answer the “why” questions. The latter category is equated with evolutionary causation [1], epitomized in the famous saying of Theodosius Dobzhansky: “Nothing makes sense in biology except in the light of evolution” [2]. Indeed, Darwinian principles provide a unifying explanation for the astounding biological diversity and complexity. The main idea is elegantly simple: competition for limited resources within a population of individuals with heritably distinct phenotypes gives rise to increased population frequencies of individuals with phenotypes that are better adapted to a given environment. Thus, the process is shaped by the interplay between stochastic mutational processes and the deterministic context-specific natural selection.

A landmark paper by Peter Nowell in 1976 applied the concept of evolutionary causation to explain the initiation and progression of cancers. According to Nowell's argument, cancers occur and progress because of the underlying process of somatic clonal evolution. Genetic mutations within somatic cells generate heritable phenotypic variability, allowing for the outgrowth of sub-clones with higher fitness [3]. Whereas there is little disagreement about the Darwinian nature of cancer causation, the prevailing conceptual framework of somatic cancer evolution has been shaped by a mutation-centric argument articulated by Eric Fearon and Bert Vogelstein. They argued that the multistep cancer progression is the direct result of the mutational activation of oncogenes and inactivation of tumor suppressor genes, as these genomic changes “drive” tumor progression [4]. More generally, the idea of “driver” mutations is also applicable to clonally heritable changes in gene expression, without changes in sequence of the gene/protein, referred to as epimutations [5]. For the sake of simplicity, unless otherwise specified, we will use the term “mutations” to refer to both genetic mutations and epimutations.

Recent advances in DNA sequencing techniques have enabled the discovery of remarkable genetic heterogeneity within tumors, including differences in the mutational status of presumed drivers [6] suggesting a picture that is more complex than that of a series of clonal succession driven by acquisition of powerful driver mutations. Furthermore, research within the last two decades has also brought about the realization that alterations in tissue microenvironments play key roles in cancer initiation and progression. In spite of these developments, the mutation centric view of somatic evolution remains dominant, and the famous statement “The revolution in cancer research can be summed up in a single sentence: cancer, is, in essence, a genetic disease” [7] reflects a wide consensus within the cancer research community.

Whereas consideration of genes and altered gene activity provide an appropriate framework for the elucidation of proximal mechanisms of cancer causation, it might be inaccurate when applied to evolutionary causation. The evolution results from the interplay between mutational diversification and outgrowth of populations with phenotypes that are most fit within the dynamic and context-specific selection forces. Therefore, context-specific selection forces need to be taken into account to understand evolutionary changes. This distinction between proximal and evolutionary causes not only provides a more relevant framework for understanding origin and progression of cancers, but also offers new approaches for the prevention and treatment of the disease.