(b. Okazaki, Japan, 13 November 1924; d. Mishima, Japan, 13 November 1994),
molecular evolution, neutral theory, population genetics, evolutionary genetics, diffusion equations.
Kimura was the chief proponent of the neutral theory of molecular evolution. In addition, he was a noted and influential mathematical population geneticist who developed the use of diffusion equation approximations for problems in population biology.
Early Years and Education . Kimura was born in 1924 in Okazaki, Japan. His father was a businessman and Motoo was the first son. His father’s interests in flowers and ornamental plants, along with his middle school teacher’s encouragement, led Kimura to an early conviction that he would become a botanist. During school, Kimura also developed an interest in mathematics, but could see no connection to botany and did not pursue it with the same interest. In 1942 Kimura entered the National High School in Nagoya, where he eagerly studied plant cytogenetics under M. Kumazawa. Of special significance for Kimura’s future, Kumazawa also taught a course in biometry. For the first time, Kimura realized that his mathematical skills could find a place in biology.
The urgent circumstances of World War II shortened Kimura’s time spent in high school from three years to two-and-one-half, thus allowing him to enter the Kyoto Imperial University in 1944. Kimura entered as a student of botany under the Faculty of Sciences, but his main influence was Hitoshi Kihara, a geneticist in the Faculty of Agriculture. By the end of his first year at Kyoto, the atomic bombs had been dropped on Hiroshima and Nagasaki and Japan had surrendered. Wartime shortages worsened after the surrender, however, and Kimura— being a student away from home—was hit especially hard. Fortunately, Kimura had a cousin in Kyoto that he could call on occasionally for better food.
Kimura’s cousin, Matsuhei Tamura, was an associate professor under Hideki Yukawa, the theoretical physicist who had predicted the existence of the meson and was considered by Kimura to be Japan’s scientific hero. Tamura was a mathematical physicist and surely had an influence on Kimura, who had developed an ambition “to do something in genetics like what theoretical physicists were doing in physics” (Kimura, 1985, pp. 463–464). Although Tamura was not impressed with the idea of theoretical biology, he had a better understanding of what mathematical biology entailed than most biologists in Japan at the time.
Kimura’s growing interest in mathematical treatments of genetics and biology flourished after he graduated and moved into Kihara’s laboratory at Kyoto. Kihara’s attitude was remarkable; it was just right for Kimura, and for the future of population genetics. Recognizing Kimura’s talent, Kihara assigned him no specific duties, leaving him free to study. Kimura threw himself into the technical literature of mathematical genetics, then dominated by Sewall Wright, John B. S. Haldane, and Ronald Aylmer Fisher. Kimura took mathematics courses where he could, but he was largely self-taught with occasional help from Tamura. While a student, Kimura had read voraciously whatever genetics literature he could get. Pirated editions of Conrad H. Waddington’s An Introduction to Modern Genetics (1939) and Theodosius Dobzhansky’s Genetics and the Origin of Species (1937) led him to the work of Wright in particular. By graduation, he had begun to devote more and more of his time to studying Wright’s mathematical papers. Indeed, Kimura spent a full year on Wright’s 1931 paper, “Evolution in Mendelian Populations,” alone, learning the math as he went.
When the National Institute of Genetics was founded in 1949 in Mishima, Kimura was hired as a research associate with Kihara’s recommendation. He remained associated with the institute for the rest of his life. The institute was located in a wooden building that had been a wartime aircraft factory. It was hot in summer and cold in winter. Furthermore, at that time Mishima was a small, provincial city, lacking the cultural and intellectual attractions of Kyoto and Tokyo. No one in Mishima understood or cared about Kimura’s work, which increased his sense of isolation. He made frequent trips to Kyoto and Tokyo for library facilities, and undoubtedly for intellectual refreshment. Undeterred, he began writing papers and the first annual report of the Genetics Institute contained five of his reports, some startlingly original. It is interesting to read these early reports as they foreshadowed some of the later work for which Kimura was to become famous.
McDonald and Newton Morton, two American geneticists at the Atomic Bomb Casualty Commission, also recognized Kimura’s work. Together with Komai, they were able to find enough funding for Kimura to come to the United States. Kimura wanted to work with Wright, but by this time Wright was getting ready to retire from the University of Chicago and was not taking students. Instead Kimura went to Iowa State College (now Iowa State University) in 1953, where worked with America’s best-known animal breeder and Wright acolyte, Jay L. Lush.
After entering Iowa State College, Kimura became dissatisfied with the research program, which was concerned with quantitative traits and emphasized subdivision of epistatic variance (that is, the variance component caused by gene interaction). Kimura understood this, but he really wanted to work on stochastic processes. Furthermore, he developed a strong dislike of Lush. When Kimura had first arrived in the United States, he had attended the Genetics Society of America meeting in Madison, Wisconsin, where he met James F. Crow. Crow was a population geneticist and one of the few who were acquainted with Kimura’s work. Indeed, on the voyage from Japan Kimura had written a paper demonstrating how fluctuating selection could mimic the stochastic effects of random genetic drift. Kimura had cleverly found a transformation that converted a complicated partial differential equation into a simple heat-diffusion formula, known to every physics student. He gave the paper to Crow for comments, who suggested its publication in Genetics. Wright reviewed the paper with unusual enthusiasm, and it was soon published (Kimura, 1954). As Kimura’s dissatisfaction grew at Iowa State, he decided to transfer to the University of Wisconsin and study with Crow. Crow was reasonably sure that Wright would soon be moving to Wisconsin, and so accepted Kimura as a student.
Kimura spent two years, 1954–1956, getting his PhD in Wisconsin. Before coming to the United States, Kimura had discovered the two Kolmogorov equations. These are partial differential equations, one known as “forward” and other “backward,” used to describe random processes, such as Brownian motion and more general diffusion processes. Wright had used the forward equation— in fact he rediscovered it himself—but Kimura was the first geneticist to employ the backward equation. He realized while still in Wisconsin that this equation was especially useful for some previously unsolved problems. Later, for example, he used this to study the age of a mutant allele in a population.
Soon after arriving in Wisconsin, Kimura obtained the complete distribution of allele frequencies under neutral random drift, at any time from any arbitrary starting frequency. He soon extended this to three alleles, then to an indefinite number. He then included the effects of mutation, migration, and selection. These results were published in the Cold Spring Symposium (Kimura, 1955). By the time Kimura received his PhD, he was already a recognized leader in theoretical population genetics. Kimura then returned to Japan. Except for occasional stays abroad, usually a year or less, he spent the rest of his life in Mishima.
In Japan, Kimura continued to develop equations for stochastic genetic models of greater generality. He introduced the “infinite allele” and “infinite site” models, widely used for evolutionary studies many years later after the coming of molecular techniques. With his colleague Takeo Maruyama, he found a method for investigating several problems, such as the number of individuals in the path to fixation or loss, or the number of heterozygotes. Kimura also developed the “stepping-stone” model of population structure, which has served as a foundation for the study of migration by many other scientists. In addition, Kimura also did a number of studies of genetic load and on inbreeding theory. His interest in mathematical models also led him to pioneering uses of computer simulations in population genetics (Crow, 1995).
The Neutral Theory of Molecular Evolution . In the 1960s population genetics had the most beautiful theory in biology, but there were few opportunities to apply it. Molecular biology changed everything. Data on the rates of molecular evolution were appearing and were awaiting analysis. In 1968 Kimura proposed what would become known as the neutral theory of molecular evolution. Using protein sequence data generated by biochemists such as Emile Zuckerkandl and Emmanuel Margoliash, Kimura and his colleague Tomoko Ohta compared mammalian protein sequences and used the number of detected differences across species to calculate a rate of molecular evolution. Kimura then reasoned that if most mutations were in fact selected, then the rate of evolution calculated for mammals would create an intolerable genetic load (the amount of differential mortality and fertility required for such a rate was more than the population could sustain). Because mammals were not extinct or staggering under an enormous genetic load, Kimura concluded that most detected molecular variants were in fact selectively neutral, meaning that they produced no change in survival or fertility for their possessors (Kimura, 1968; also see Dietrich, 1994, and Suarez & Barahona, 1996).
Kimura’s conclusion and argument were controversial, but the dispute between neutralists and selectionists was guaranteed in 1969 when Tom Jukes and Jack King wrote their neutralist manifesto under the provocative title of “Non-Darwinian Evolution.” King and Jukes brought a large variety of evidence to bear in favor of large numbers of neutral mutations (1969). By using evidence from the growing field of molecular evolution to support the idea of neutral mutations and the importance of random drift, they spelled out the molecular consequences of the neutral hypothesis more clearly than Kimura had. King and Jukes built their case using phenomena such as synonymous mutations, the Treffers mutator, the relation between amino acid frequencies and the genetic code, and the growing body of data on specific proteins such as cytochrome c.
The neutral theory directly challenged the power of natural selection in evolutionary biology. Because the neutral theory claimed that random drift was more significant than natural selection at the molecular level, it helped drive a wedge between the way evolution was understood at the organismal and molecular levels—at the organismal level, natural selection predominated, while at the molecular level, random genetic drift was an important factor. By articulating a different set of evolutionary mechanisms for the molecular level, the neutral theory provided a theoretical foundation for the development of molecular evolution as a new field of biological inquiry.
Many biologists were extremely skeptical of the neutral theory. Classical geneticists believed, reasonably, that hardly any observable changes were completely neutral, in part, because they were thinking about morphological changes, not changes in nucleotides or amino acids. Kimura and Ohta pursued the neutral theory vigorously. One of the most attractive features of the approach was that it provided a basis for the molecular clock, described in 1964 by Zuckerkandl and Linus Pauling. Zuckerkandl and Pauling had observed that molecules collected substitutions at a remarkably constant rate. Hence, the number of changes between two species could be used to infer the time since they split from a common ancestor—a great boon to systematic biologists. The neutral theory proposed a mechanism for this constancy, because neutral evolution is driven by the mutation rate; meaning that when random drift is taken into account, the longtime average rate of nucleotide substitution becomes equivalent to the mutation rate, which was believed to be roughly constant.
In 1969 Kimura used the constancy of the rate of amino acid substitutions to argue powerfully for the importance of neutral mutations and random drift in molecular evolution. At the same time, Kimura was also calling on his earlier work on stochastic processes in population genetics to forge a solid theoretical foundation for the neutral theory. Kimura’s diffusion equation method provided the theoretical techniques he needed to formulate specific models, which in turn enabled him to address issues such as the probability and time to fixation of a mutant substitution as well as the rate of mutant substitutions in evolution. Working in collaboration with Tomoko Ohta, Kimura also extended the neutral theory to encompass the problem of explaining protein polymorphisms. This was a central concern of population genetics, and Kimura and Ohta were able to argue that protein polymorphisms were a phase in mutations’ long journey to fixation (Kimura & Ohta, 1971).
Kimura found many other arguments in favor of the neutral theory over the next few years. For instance, amino acids in regions of less importance for the function of a polypeptide evolved faster than those important for the function. Particularly revealing was the insulin molecule, which has three regions, one of which is discarded and not used. The unused part evolved fastest. Within codons, synonymous changes were faster than nonsynonymous. To Kimura, slow evolution of some nucleotides was caused by “selective constraint”: these regions already functioned well, and therefore most mutations were harmful. One of Kimura’s most striking arguments came from the fact that the number of amino acid differences between the alpha and beta hemoglobins in humans was about the same as that between human beta and carp alpha. The first two have been in the same cell for some 400 million years, while the latter two have been in fish and the line leading to humans. The difference in selective forces could hardly be greater. If the amino acid changes were due to selection, the two sequences should be enormously different from each other; but they were not. Kimura summarized his views in a widely quoted book, The Neutral Theory of Molecular Evolution, published in 1983. He devoted much of his energy for the rest of his active life to finding more evidence and arguing his case.
DNA Enters the Debate . The availability of DNA sequence data in the mid-1980s transformed the debate over the neutral theory of molecular evolution. While earlier techniques, such as electrophoresis, allowed evolutionary biologists to estimate variability at the molecular level, DNA sequencing promised more direct measurements of genetic variability. More importantly, DNA sequence data made it possible to better distinguish drift from selection.
In the 1960s Kimura and King and Jukes proposed that synonymous changes, changes in DNA that do not produce a corresponding change in the amino acid of the protein coded for, should be neutral because they have no observable effect. Because these changes presumably have no selective effect, they should evolve more quickly than most of the nonsynonymous changes because most of these are harmful and are eliminated rather than contributing to evolutionary change. The rare advantageous change would evolve more quickly than the neutral change as positive selection pushed it to fixation in the population. Kimura proposed that the differences in synonymous and nonsynonymous substitution rates could form the basis for detecting positive selection (Kimura, 1983). In 1984 Martin Kreitman introduced DNA sequencing to evolutionary genetics and extended Kimura’s idea of comparing synonymous and nonsynonymous substitutions. The McDonald-Kreitman test compares the ratio of nonsynonymous to synonymous changes within a species and between two species. If the sequences are neutral, the ratios should remain the same. If there is positive selection, then nonsynonymous changes should have accumulated over time, so there would be more nonsynonymous changes between species than within a species. This test and many other statistical tests that followed enabled evolutionary biologists to detect balancing selection, adaptive protein evolution, and population subdivision (McDonald & Kreitman, 1991; Kreitman, 2000).
During the last decade of Kimura’s life, the debate surrounding the neutral theory died down. While the neutral theory may not have been accepted in full, it became a standard part of evolutionary theory. As sequence data accumulated, biologists realized that many organisms possess a great deal of noncoding DNA, which would then have been subject to neutral evolution. At the same time, the neutral theory forms a natural null hypothesis for studies of selection and for statistical tests of selection.
Despite the controversy surrounding the neutral theory, Kimura received numerous honors, including honorary degrees from the University of Chicago and the University of Wisconsin, the Japan Academy Prize in 1968, the Japanese Order of Culture (Emperor’s Medal) in 1976, the Chevalier de L’Ordre National du MØrite in 1986, the Asahi Shimbun Prize in 1987, the John J. Carty Award from the (U.S.) National Academy of Sciences in 1987, and the Darwin Medal from the Royal Society in 1992. He is particularly honored in his hometown of Okazaki, thanks largely to efforts of his brother. In addition to a museum, Kimura is honored with a statue in the city (Crow, 1995).
Soon after his return to Japan from Wisconsin, Kimura married. He and Hiroko Kimura had one child, a son, Akio. He had one important hobby, orchid breeding. Every Sunday was devoted to this, and he produced several prize-winning clones. Throughout his life he also enjoyed philosophy, especially the writings of Bertrand Russell, and science fiction, where he was particularly fond of the writing of Arthur Clarke. Kimura’s main interest in life was his work, especially after the neutral theory, for which he became a passionate advocate. Kimura’s advocacy continued up to a short time before his death. In his late sixties, Kimura developed amyotrophic lateral sclerosis, and deteriorated very rapidly. His death came on his seventieth birthday, the result of an accidental fall.
WORKS BY KIMURA
“Process Leading to Quasi-Fixation of Genes in Natural Populations Due to Random Fluctuation of Selection Intensities.” Genetics 39 (1954): 280–295.
“Stochastic Processes and Distribution of Gene Frequencies under Natural Selection.” Cold Spring Harbor Symposium on Quantitative Biology 20 (1955): 33–53.
Diffusion Models in Population Genetics. London: Methuen, 1964.
With James Crow. “The Number of Alleles that Can Be Maintained in a Finite Population.” Genetics 49 (1964): 725–738
“Evolutionary Rate at the Molecular Level.” Nature 217 (1968): 624–626. Kimura’s initial argument for neutral molecular evolution
. “The Rate of Molecular Evolution Considered from the Standpoint of Population Genetics.” Proceedings of the National Academy of Sciences of the United States of America. 63, no. 4 (1969): 1181–1188.
With Tomoko Ohta. “Protein Polymorphism as a Phase in Molecular Evolution.” Nature 229 (1971): 467–469.
The Neutral Theory of Molecular Evolution. Cambridge, U.K.: Cambridge University Press, 1983. Kimura’s most extensive treatment of the neutral theory.
“Genes, Populations, and Molecules: A Memoir.” In Population Genetics and Molecular Evolution. Edited by Tomoko Ohta and Kenichi Aoki. Tokyo: Japan Scientific Society Press, 1985
. “Molecular Evolutionary Clock and the Neutral Theory.” Journal of Molecular Evolution 26 (1987): 24–33.
Population Genetics, Molecular Evolution, and the Neutral Theory: Selected Papers. Edited by Naoyuki Takahata. Chicago: University of Chicago Press, 1994. A collection of Kimura’s most influential papers and a complete bibliography of his publications.
Crow, James. “Motoo Kimura (1924–1994).” Genetics 140 (1995): 1–5.
Dietrich, Michael R. “The Origins of the Neutral Theory of Molecular Evolution.” Journal of the History of Biology 27 (1994): 21–59.
King, Jack L., and Thomas H. Jukes. “Non-Darwinian Evolution.” Science 164 (1969): 788–798.
Kreitman, Martin. “Methods to Detect Selection in Populations with Application to the Human.” Annual Review of Genomics and Human Genetics 1 (2000): 539–559.
McDonald, John H., and Martin Kreitman. “Adaptive Protein Evolution at the Adh Locus in Drosophila.” Nature 351 (1991): 652–654.
Ohta, Tomoko, and John Gillespie. “Development of Neutral and Nearly Neutral Theories.” Theoretical Population Biology 49 (1996): 128–142.
Provine, William. “The Neutral Theory of Molecular Evolution in Historical Perspective.” In Population Biology of Genes and Molecules, edited by Naoyuki Takahata and James Crow. Tokyo: Baifukan, 1990.
Suarez, Edna, and Anna Barahona. “The Experimental Roots of the Neutral Theory of Molecular Evolution.” History and Philosophy of the Life Sciences 18 (1996): 55–81.
Zuckerkandl, Emile, and Linus Pauling. “Evolutionary Divergence and Convergence in Proteins.” In Evolving Genes and Proteins, edited by Vernon Bryson and Henry J. Vogel. New York: Academic Press, 1965.
Michael R. Dietrich
James F. Crow