By Rob Mitchum
Alan Turing is best known as the father of the modern computer, a skillful World War II codebreaker, and a pioneer in the study of artificial intelligence. But in the last years before Turing’s death at age 41, he aimed his genius at a different target: the then-stalled field of developmental biology. By the middle of the 20th century, many scientists had tried and failed to explain how a complex organism could form itself from a simple embryo originally made up of identical cells. One 19th century biologist, Hans Driesch, grew so frustrated with the problem that he gave up and wrote a text on vitalism, the doctrine that life cannot be explained by science alone.
In a 1952 paper called “The Chemical Basis of Morphogenesis,” Turing rushed headlong into this challenge, building a mathematical model of how patterned cells can be formed from non-patterned beginnings. It was Turing’s only published work on the topic; he died two years later. But in those 35 pages, he predicted elements of developmental biology that wouldn’t be discovered for 30 more years, coined a term that is central to the field today, and accidentally sparked a new sub-field of mathematical study for a bonus. In a recent Nature retrospective commemorating Turing’s 100th birthday, University of Chicago scientist John Reinitz wrote, “What Turing should receive credit for is opening the door to a new view of developmental biology…He was well ahead of his time.”
Reinitz’s own research is deeply indebted to Turing’s landmark paper. A professor with appointments in Statistics, Ecology & Evolution, and Molecular Genetics & Cell Biology, Reinitz’s laboratory studies how gene expression controls the development of the fruit fly Drosophila melanogaster. As part of those efforts, the laboratory has built several computational models of gene transcription and fly development, one of which is a specific example of a class of equations in Turing’s paper, Reinitz said in an interview about his essay.
Beyond that direct lineage, Reinitz admires the paper (“The article is just a pile of interesting ideas.”) and teaches it in his courses. But it wasn’t fully appreciated in the field of developmental biology until decades after its publication, when the role of DNA and the molecules that Turing preemptively named “morphogens” became more widely known in biology.
“When I was in grad school, this paper was circulating, and it was considered to be a sort of interesting but crazy paper,” Reinitz said. “It didn’t have anything about genes, and when I first saw it, it was really before any of these morphogens had actually been found. So it didn’t seem to have any direct bearing on actual experimental science.”
The core of the paper is a computational model — one of the first ever published, Reinitz said — that mathematically proved one could create complex patterns from a symmetrically organized cell. Early in development, the “pluripotent” cells of the embryo are each capable of developing into a wide range of cell types, from blood to skin to muscle to hair. Indeed, if an embryo is split in two early enough, it can form two entire organisms…as is the case with identical twins.
Classic linear mathematics can’t explain how one generic cell can produce so many unique descendants. So Turing’s model employed a “mathematical trick,” using the interplay of two diffusing factors (Turing’s “morphogens”) to produce the temporary instability necessary for a pattern to form. That these morphogens had never been observed in scientific experiments at the time he published was beside the point; Turing simply wanted to show that pattern-making could be done with a minimum of elements.
“I think that one of the things that’s seriously misunderstood about the paper is that a lot of people read it and think it’s making specific predictions about biological systems,” Reinitz said. “The main thing he was concerned about was just demonstrating that you could form patterns from non-patterns. He wanted to show with chemistry that you can have patterns form spontaneously.”
It was three more decades before factors similar to the morphogens Turing had predicted were actually found. In 1983, the Hox genes were found to produce factors that direct the patterning of the fly body, and then later identified to exist and perform a similar role in many other species. While these and subsequently discovered developmental factors didn’t function exactly as Turing hypothesized in his model, they supported his general idea that the future fate of a cell can be predicted from measuring the signals (or genes) active at a given moment.
“For me personally, the paper is intellectually important because it is also the first state description of development,” Reinitz said. “Whether people use equations or not, this is the way all modern developmental biologists think about the problem. They think, if I change this gene, that’s going to change what the embryo grows into, and they’re interested in how mechanical forces shape embyrogenesis. Those ideas are all thinking in a state description way.”
Unfortunately, Turing’s life came to a troubled end shortly after the publication of his paper, leaving his planned research into the natural patterns of leaves unfinished. In his absence, Turing’s ideas about development were also left open to misinterpretation and distortion, Reinitz said, until the discovery of morphogens demonstrated that his remarkable prescience. Now, as the scientific community commemorates Turing’s centennial, Reinitz hopes that the deep impact he left on developmental biology will be recognized alongside his legendary contributions to computation.
“I think that a lot of it had to do with the fact that there was such a long gap between the publication of the paper and the time when the things Turing was talking about really became amenable to investigation,” Reinitz said. “If Turing had stayed alive, things would have been very different.”
[Photo of Alan Turing sculpture by Jon Callas/Wikimedia Commons]
Reinitz, J. (2012). Turing centenary: Pattern formation Nature, 482 (7386), 464-464 DOI: 10.1038/482464a