Kenneth Eward/BioGrafx/Photo Researchers Inc.
Nearly two decades ago, Paul H. Silverman testified before Congress to advocate the Human Genome Project. He later became frustrated when the exceptions to genetic determinism, discovered by this project and other investigations, were not sufficiently incorporated in current research and education.
In "Rethinking Genetic Determinism,"1 Silverman questioned one of the pillars of molecular genetics and documented the need for determinism's expansion into a far more valid and reliable representation of reality. He would receive correspondence from all over the world that reinforced this vision.
Silverman firmly believed that we needed a wider-angled model, with a new framework and terminology, to display what we know and to guide future discovery. He also viewed this model as being a catalyst for exploring uncertainty, the vast universe of chance differences on a cellular and molecular level that can considerably influence organismal variability. Uncertainty not only undermines molecular...
Arnold Goodman (left) is an associate director of the Center for Statistical Consulting at the University of California, Irvine. Cláudia Bellato (center) is an independent researcher at CENA, University of São Paulo, Brazil. Lily Khidr (right) is a PhD candidate at UC-Irvine. They dedicate this article to the memory of Paul Silverman and thank Nancy, his wife, for her assistance.
Various commentaries detail deviation from determinism within the cellular cycle. Here we use the term cellular cycle not in the traditional sense, but rather to describe the cyclical program that starts with gene regulation through transcription, translation, post-processing and back into regulation.
Richard Strohman at UC-Berkeley describes the program in terms of a complex regulatory paradigm, which he calls "dynamic epigenetics." The program is dynamic because regulation occurs over time, and epigenetic because it is above genetics in level of organization.2 "We thought the program was in the genes, and then in the proteins encoded by genes," he wrote, but we need to know the rules governing protein networks in a cell, as well as the individual proteins themselves.
John S. Mattick at the University of Queensland focuses upon the hidden genetic program of complex organisms.3 "RNAs and proteins may communicate regulatory information in parallel," he writes. This would resemble the advanced information systems for network control in our brains and in computers. Indeed, recent demonstrations suggest that RNA might serve as a genetic backup copy superseding Mendelian inheritance.4
Gil Ast of Tel Aviv University writes: "Alternative splicing enables a minimal number of genes to produce and maintain highly complex organisms by orchestrating when, where, and what types of proteins they manufacture."5 About 5% of alternatively spliced human exons contain retrotransposon Alu sequences. These elements represent an engine for generating alternative splicing.
Thus we see a genetic control system regulated by protein products, RNAs, and interventions from DNA itself. Yet throughout, the consideration of genetic uncertainty as a bridge to cellular behavior is conspicuously absent.
Genetic reductionism, the other pillar of molecular genetics, has many challengers. Among them is Stephen S. Rothman at UC-Berkeley, who described the limits of reductionism in great detail within his comprehensive and well-constructed book.6
A more recent publication by Marc H.V. Van Regenmortel at France's National Center for Scientific Research updated this assessment by discussing not only the deficiencies of reductionism, but also current ways of overcoming them. "Biological systems are extremely complex and have emergent properties that cannot be explained, or even predicted, by studying their individual parts."7
NEW CELL MODEL
Molecular genetics appears to be at a crossroads, since neither determinism nor reductionism is capable of accurately representing cellular behavior. In order to transition from a passive awareness of this dilemma to its active resolution, we must move from simply loosening the constraints of determinism and reductionism toward a more mature and representative combination of determinism, reductionism, and uncertainty.
To facilitate this expansion, we propose a model for the cellular cycle. Although only a framework, it provides a vehicle for broader and deeper appreciation of the cell. The figure on page 25 provides a novel structure for understanding current knowledge of the cycle's biological stages, as well as a guide for acquiring new knowledge that may include genetic uncertainty.
Organismal Regulation: The organism specifies its cellular needs (bottom red) for the cell to act upon. It converts the comparison of proteins with organismal needs into metabolic agents. The organism then defines its cellular needs (top red). It employs metabolic effects to alter the extra-cellular matrix and signal other needs.
Cellular Regulation: Within the bounds of a cell's membrane, cellular needs transmission (top blue) directs the cell in various ways, including proliferation, differentiation, and programmed cell death. It uses such factors as receptors and enzymes to yield molecular messengers. In the cell's nucleus, chromatin remodeling (bottom blue) then rearranges DNA accessibility by uncoiling supercoiled DNA and introducing transcription factors.
Transcription: Transcription (left green) DNA serves as the template for RNAs, both regulatory sequences and pre-messenger RNAs. It transcribes polymerases and binding partners into heterogeneous nuclear RNAs. Pre-messenger RNAs then undergo highly regulated splicing and processing (right green). They turn pre-messenger RNAs into mature messenger RNAs.
Translation: Within the cytoplasm, messenger RNAs and ribosomes translate 2D-unfolded proteins (left magenta). Secondary structuring and thermodynamic energy (right magenta) then enable physical formations that complete the process with folded proteins and oligonucleotides.
Postprocessing: Again within the cytoplasm, tertiary structuring and modification (top aqua) use assemblers, modifiers and protein subunits to supply regulated proteins. Then feedback regulation (bottom aqua) produces heritable gene expression from small RNAs, proteins and DNA. The proteins and gene expression, rather than being an endpoint, now begin the whole process over again by signaling other cells, altering and maintaining the genome, and editing RNA transcripts.
Helen M. Blau was a keynote speaker at the recent UC-Irvine stem-cell symposium in memory of Paul Silverman and Christopher Reeve.8 She observed: "Where we look and how we look determine what we see." Although only a brief prescription, we now propose an approach to the exploration for uncertainty that involves both where we look and how we look. We examine those cellular-cycle outputs having a relatively high likelihood of diversity and its frequent companion, uncertainty.
As an example of exploring for uncertainty in a cellular cycle, consider the following example: Suppose an organismal regulatory program for cellular differentiation might alter the signaling milieu in the extracellular matrix. The signal is internalized by a cell, which might, in turn, alter transcription, produce mature messenger RNAs, produce the 3D-folded proteins, and feed back to alter gene expression for all daughter cells.
Now suppose the ECM signaling milieu is altered with a probability p1; the signal is internalized by a cell with a probability p2; transcription will change with a probability p3; mature mRNAs are produced with a probability p4, producing the 3D-folded protein with a probability p5 and altering heritable gene expression with a probability p6. The probabilities p2, p3, p4, p5, and p6 are all conditional on results from the step preceding them, so that the resulting probability of altered heritable gene expression is the product of all of them. Although this probability may be small, is it not preferable to know its form and to later estimate it, than to simply ignore its existence?
When we consider all possible stage alterations, the diversity of outputs and complexity of our probability calculations will increase. If we also consider all possible interactions, the diversity of outputs and complexity of probability calculations will increase quite substantially.
The implications reach far beyond the regulation of a single cell or organism. Sean B. Carroll of the University of Wisconsin, Madison, summarizes evolutionary developmental biology,9 invoking Jacques Monod's landmark
Why wouldn't chance also be included in our observations of biology at the molecular level? We've proposed a brief overview of the "what" and "how" for constructing an uncertainty bridge from genetic determinism and reductionism to actual cellular behavior. We hope and believe it meets the spirit of Paul Silverman's prescient vision, as well as his final wishes.