Some Recollections and Reflections on Mutation Rates

Corresponding author: Maurice S. Fox, Department of Biology, M.I.T., Room 68-630, 77 Massachusetts Ave., Cambridge, MA 02139, msfox{at}mit.edu (E-mail).

IN thinking about JAN DRAKE's creative and persistent interest in mutation, I am prompted to reflect on the extent to which investigations, intricately interweaving the study of mutation and the use of mutants, have figured so prominently in the progress of our understanding of biological processes.

I take this opportunity to record some of my own recollections of the work on mutation in the 1940s and early 1950s that was so influential in the development of quantitative biology. In 1943, LURIA and DELBRÜCK had elegantly shown that mutation could occur prior to the selection process used to detect its products. In 1944, AVERY et al. had shown that the chemical material responsible for the transfer of genetic properties from one bacterium to another (bacterial transformation) was DNA, but the general implications of this demonstration were still regarded with suspicion. The proposal by WATSON and CRICK 1953 for the molecular structure of DNA was still in the future. In retrospect, our naiveté in those early days may seem amusing but some of the paradoxes that were encountered remain with us to this very day.

When World War II ended, many of the talented scientists who had been brought together to build the atomic bomb saw an opportunity to embark on new intellectual adventures. It was ROBERT HUTCHINS, the Chancellor of the University of Chicago, who displayed the imagination to bring many of them to the University of Chicago. Among the more prominent figures were HAROLD C. UREY, WILLARD F. LIBBY, ENRICO FERMI, JAMES FRANK, LEO SZILARD, MARIA GOEPPERT, JOSEPH E. MAYER, and EDWARD TELLER. In 1947, after a meeting of the Atomic Scientists of Chicago, SZILARD invited AARON NOVICK, a physical organic chemist, with whom he had worked the year before while lobbying in Washington for civilian control of atomic energy, to join him in "an adventure in biology." They did join forces and that summer took the newly established Phage course at Cold Spring Harbor taught by MARK ADAMS and MAX DELBRÜCK. During the following year they created a laboratory in the basement of a building, a few blocks off the main campus, that had been a synagogue before it was acquired by the university. The following summer they completed their formal education in biology with a microbiology course taught by C. B. VAN NIEL at the Hopkins Marine Station of Stanford University at Pacific Grove, California. It was during that summer that the idea of a device for growing bacteria in continuous culture, the chemostat, was developed. By 1950, the chemostat was a reality. The central feature is embodied in the paper by NOVICK and SZILARD 1950 :

In general, the growth rate of a bacterial strain may be within very wide limits independent of the concentration of a given growth factor; but since at zero concentration the growth rate is zero, there must of necessity exist, at sufficiently low concentrations of the growth factor, a region in which the growth rate falls with falling concentration of the growth factor. It therefore should be possible to maintain a bacterial population over an indefinite period of time growing at a rate considerably lower than normal simply by maintaining the concentration of one growth factor—the controlling growth factor—at a sufficiently low value, while the concentrations of all other growth factors may at the same time be maintained at high values. (p. 708)

At this time, I too began searching for a pathway into biology much to the distress of my Ph.D. thesis advisor, the physical chemist, WILLARD F. LIBBY. LIBBY felt that the way to switch fields was to do as he did: invent an instrument or a technique that would be useful in that new field and become a participant in the new field overnight. He believed that he had become an archaeologist by inventing the ¹⁴C dating approach to determine the age of organic carbon containing samples from archaeological sites. Not sharing this view, when I learned about the exciting work of NOVICK and SZILARD from a seminar NOVICK gave on the chemostat, I went to talk to him.

Our conversation led to an invitation to join their effort and I became part of the league of physical scientists enthralled by the promise of a new approach to biology. I was to participate in the investigation of the physiology of mutability using the chemostat.

By this time, work with the chemostat was progressing vigorously in a laboratory in the newly constructed Institute of Radiobiology and Biophysics. NOVICK and SZILARD 1950 , NOVICK and SZILARD 1951 had already shown that in continuous culture, for bacteria growing in steady state at rates determined by limiting concentrations of an essential growth factor, mutations to bacteriophage resistance occurred at a constant rate. The chemostat thus had provided a direct method of measuring mutation rates. They showed that phage-resistant mutants continued to accumulate in the population at a constant rate for about 35 generations after which the mutant frequently dropped and then returned to the pre-drop rate of accumulation. For chemostats that were maintained for very many generations, such discontinuities occurred repeatedly, about every 30–40 generations (NOVICK 1956 ). These discontinuities were interpreted as the result of evolutionary changes in the chemostat population driven by the growth rate limiting concentrations of the essential growth factor. For most of these experiments, a nonreverting tryptophan-requiring strain of Escherichia coli was used and tryptophan was the limiting growth factor. The rate at which fresh medium flows into the growth tube determines the growth rate of the bacteria. The steady state tryptophan concentration, established by the bacteria, is the concentration at which bacteria grow at a rate equal to the flow rate. The bacteria are thus maintained in steady state. The discontinuities that they observed were shown to result from population changes in which fast mutants, able to grow more rapidly at that tryptophan concentration, displaced the previous population and established a new lower tryptophan concentration in the growth tube. The phage-resistant mutants that had accumulated from the previous population were thus washed out and new phage-resistant mutants arising from the newly established fast population began to accumulate. Thus they provided a workable system for displaying evolutionary changes in a controlled environment. Indeed it was this system that was employed some years later by HORIUCHI et al. 1962 to investigate how bacteria evolve to better use of limiting concentrations of lactose in their medium. Among the kinds of changes they demonstrated was gene amplification of chromosomal genes required for lactose metabolism.

NOVICK and SZILARD had also used the chemostat to demonstrate that caffeine, other methyl xanthines, and adenine acted as mutagens when present in modest concentrations. It was somewhat later that they showed that low concentrations of adenosine or guanosine could reduce the mutagenic activity of the methyl xanthines and in fact could reduce spontaneous mutation rates (NOVICK and SZILARD 1952 ).

This was the context in which, with the help of the Cold Spring Harbor phage and bacterial genetics courses and, a year later, the VAN NIEL course, my own career in biology was launched. My first project was to construct a continuous culture device different from the chemostat, in which the bacteria would not be limited by low concentrations of a limiting growth factor but could grow at their maximum rates in continuous culture. We called this device a breeder (FOX and SZILARD 1955 ). The flow rate of fresh medium into the culture tube was determined by a periodic assessment of the turbidity of the culture. The continuous culture was maintained at a bacterial density that was about one-tenth of the maximum density to which the bacteria could grow in that medium.

We thought about many kinds of experiments that would be possible with such a system. Some, in retrospect, seem very naive. For example, during my first year (before WATSON and CRICK, but after AVERY), it occurred to me that the chemostat might allow us to identify and chemically distinguish different genes whose mutation rates we could measure. The idea was to irradiate the continuous culture with monochromatic UV light and measure the enhancement at each wavelength for each of the genes whose mutation rates we could measure. That is, we could determine an action spectrum for each kind of mutation. At the next Midwest Phage Meeting in Ann Arbor, Michigan, I got to know CYRUS LEVINTHAL, who had made the transition from physics to phage a few years earlier. I told him about this experiment. He listened patiently and when I was finished, he laughed and said, "Would you like to borrow a monochromator?" He had had a similar idea and got as far as building a monochromator. We talked some more and, thank goodness, I was talked out of that adventure.

Still, many lines of investigation were productive, and one of these may have particular relevance today. With the chemostat, it was possible to measure the mutation rate for cells growing at different growth rates, using tryptophan as the limiting growth factor. We found that the mutation rates, per hour, were constant over a range of growth rates, from generation times of 2–12 hr in minimal medium with lactate as a carbon source. In other words, the mutation rate, measured per generation, had increased sixfold when the rate of growth of the bacteria had decreased sixfold. Although more limited, a similar conclusion was evident when the limiting growth factor was phosphorous. Making use of a more complex medium, containing acid hydrolyzed casein (tryptophan free), I was able to extend these observations to growth rates faster than a 2-hr generation time. Indeed, we could measure mutation rates at generation times approaching 30 min. At those fast growth rates, the mutation rates per hour began to increase, perhaps approaching a constant mutation rate per generation (FOX 1955).

Some years later, KUBITSCHEK and BENDIGKEIT 1964 confirmed the observation that, in tryptophan-limited chemostats, mutation to bacteriophage T5 resistance occurred at a rate constant in time, independent of the growth rate. They also reported that when glucose was the limiting growth factor, the mutation rates per generation were constant. These contrasting observations suggested "the existence of two different kinds of mutational response" (KUBITSCHEK and BENDIGKEIT 1964 ).

It is interesting to reflect on the fact that the possibility of maintaining bacteria in continuous culture provided the first and perhaps only opportunity till now, to make direct measurements of mutation rates. To be sure, LURIA and DELBRUCK 1943 had calculated mutation rates in their analysis of mutation events. But their calculations were based on a crucial assumption, namely that the probability of mutation was constant per cell per generation. This was certainly a plausible assumption, but the chemostat experiments suggested that it might lack generality. To the best of my knowledge, steady state devices have been the only systems that provide an opportunity to measure mutation rates directly. In lieu of direct measurement, the LURIA and DELBRUCK 1943 approach, made readily applicable by LEA and COULSON 1949 , has been and continues to be the method of choice for inferring mutation rates. Their calculation yields a mutation rate whose units are in mutations per cell per generation and is probably useful for mutations in which the selection is lethal (such as phage resistance or resistance to bacteriocidal drugs).

For mutations detected by nonlethal selection, such as reversion of auxotrophic mutants or mutants that have retrieved the capacity to metabolize lactose from a defective lac gene, their calculation could raise some difficulties. Certainly mutations arise during the growth of the culture in preparation for selection, and those that occur early give rise to "jackpots." The question that is more difficult to appraise is how many mutations arise after depositing the cells on the selective plate, even though the bacteria do not increase in numbers. FOSTER and CAIRNS 1992 showed that the distribution of frequencies of mutants from lac^- to lac⁺, in independent cultures, does not fit the expectation from the LURIA and DELBRUCK 1943 picture. We now know that new mutations appear after the selected population has been deposited on the semisolid selection medium. And this fact raises questions about the utility of LURIA and DELBRÜCK's original calculation of the mutation rate during the growth of the population. It is impossible to directly compare mutation rates in units of mutations per cell per generation (as calculated from mutation frequencies in independent culture) with rates at which mutations are observed to accumulate in a stationary population undergoing selection, measured in units of mutations per cell per hour.

The chemostat observation that, when bacteria are maintained under tryptophan limitation, the mutation rates to bacteriophage resistance remain constant per hour over a wide range of generation times, remains a paradox, and it is one of possible relevance to a number of phenomena under study today. It may be that under at least some nonlethal selection or starvation conditions, mutation rates per hour remain roughly constant even when no growth is evident. Could this be one of the two different kinds of mutation response suggested by KUBITSCHEK and BENDIGKEIT 1964 ?

New evidence of mutability, at many loci, of cells under conditions of selection for lac⁺ revertants has been reported. TORKELSON et al. 1997 and FOSTER 1997 suggest the possibility of transient hypermutable states that can occur in some cells during selection and which do not require cell division.

Recent work on human cells defective in the human homologue of the E. coli mismatch repair gene Mut S (hMSH2), describes evidence of a mutator phenotype for cells maintained under conditions of contact inhibition (RICHARDS et al. 1997 ). This elevated mutability is not evident in cells that were not maintained at high density in a growth limited state for two weeks. Since mutations in the hMSH2 gene are often found associated with hereditary nonpolyposis colon cancer, the relevance of mutability under growth limiting conditions could have wide relevance. The observation many years ago of mouse cell transformation occurring in the progeny of cells that had been exposed to X-rays 20–30 generations earlier may also be the result of a high level mutability that is only evident in cells that are maintained under conditions of contact inhibition (KENNEDY et al. 1980 ). In this case it would have to be argued that the X-ray-induced change giving rise to this property was present in all or nearly all of the X-rayed cells and that this property was transmitted to progeny cells for many generations, i.e., that it was an epigenetic change.

After 50 years of dazzling progress, we find ourselves still dependent on the use of mutants for probing the intricacies of biological processes, and on an understanding of the regulation and physiology of mutation for probing the subtle biological mechanisms responsible for balancing genomic stability with plasticity.

	LITERATURE CITED

TOP LITERATURE CITED

AVERY, O. T., C. M. MACLEOD, and M. MCCARTY, 1994  Studies on the chemical nature of the substance inducing transformation of pneumococcal types: induction of transformation by a desoxyribonucleic acid fraction isolated from pneumococcus type III. J. Exp. Med. 79:137-158.

FOSTER, P. L. and J. CAIRNS, 1992  Mechanisms of directed mutation. Genetics 131:783-789[Abstract].

FOSTER, P. L., 1997  Nonadaptive mutations occur on the F' episome during adaptive mutation conditions in Escherichia coli. J. Bacteriol. 179:1550-1554[Abstract/Free Full Text].

FOX, M. S. and L. SZILARD, 1955  A device for growing bacterial populations under steady state conditions. J. Gen. Physiol. 39:261-266[Abstract/Free Full Text].

FOX, M. S., 1995  Mutation rates of bacteria in steady state populations. J. Gen. Physiol. 39:267-278.

HORIUCHI, T., J.-I. TOMIZAWA, and A. NOVICK, 1962  Isolation and properties of bacteria capable of high rates of ß-galactosidase synthesis. Biochim. Biophys. Acta 55:152-163[Medline].

KENNEDY, A. R., M. FOX, G. MURPHY, and J. LITTLE, 1980  Relationship between x-ray exposure and malignant transformation in C3H 10T1/2 cells. Proc. Natl. Acad. Sci. USA 77:7262-7266[Abstract/Free Full Text].

KUBITSCHEK, H. E. and H. E. BENDIGKEIT, 1964  Mutation in continuous cultures. I. Dependence of mutational response upon growth-limiting factors. Mutat. Res. 1:113-120.

LEA, D. E. and C. A. COULSON, 1949  The distribution of the numbers of mutants in bacterial populations. J. Genet. 49:264-285.

LURIA, S. E. and M. DELBRÜCK, 1943  Mutations of bacteria from virus sensitivity to virus resistance. Genetics 28:491-511[Free Full Text].

NOVICK, A. and L. SZILARD, 1950  Experiments with the chemostat on spontaneous mutations of bacteria. Proc. Natl. Acad. Sci. USA 36:708-719[Free Full Text].

NOVICK, A. and L. SZILARD, 1951  Genetic mechanisms in bacteria and bacterial viruses. I. Experiments on spontaneous and chemically induced mutations of bacteria growing in the chemostat. Cold Spring Harbor Symposia on Quantitative Biology 16:337-343.

NOVICK, A. and L. SZILARD, 1952  Anti-mutagens. Nature 170:926-927[Medline].

NOVICK, A., 1956 Some chemical bases for evolution of microorganisms, pp. 533–546 in Perspectives in Marine Biology, edited by A. A. BUZZATI-TRAVERSO. University of California Press, Berkeley and Los Angeles.

RICHARDS, B., H. ZHANG, G. PHEAR, and M. MEUTH, 1997  Conditional mutator phenotypes in hMSH2-deficient tumor cell lines. Science 277:1523-1526[Abstract/Free Full Text].

TORKELSON, J., R. S. HARRIS, M.-J. LOMBARDO, J. NAGENDRAN, C. THULIN, and S. M. ROSENBERG, 1997  Genome-wide hypermutation in a subpopulation of stationary-phase cells underlies recombination-dependent adaptive mutation. EMBO J. 16:3303-3311[Medline].

WATSON, J. D. and F. H. C. CRICK, 1953  A structure for deoxyribose nucleic acid. Nature 171:737-738[Medline].

This article has been cited by other articles:

				T. L. Ferea, D. Botstein, P. O. Brown, and R. F. Rosenzweig Systematic changes in gene expression patterns following adaptive evolution in yeast PNAS, August 17, 1999; 96(17): 9721 - 9726. [Abstract] [Full Text] [PDF]