[NOTE: The letters A, B, C, D plus numbers in parentheses in the text below
refer to references in the Bibliography section.]


I grew up planning to become a physician, although my maternal grandfather, who was a physician, said that I should become a university professor because the hours were better (he was a horse and buggy doctor in a rural area of the Midwest) and the pay is better (he was a poor businessman and seldom collected his fees, if in fact he charged a fee). As it turned out I did take his advice.

Nevertheless, until my senior year in college I followed a premed curriculum, while majoring in chemistry. In my senior year I transferred from Berea College (Berea, KY) to Stanford University (Stanford, CA). At Berea I would have obtained a B.A. in chemistry, and would have had most of my senior year to become more rounded in my education by taking philosophy, logic, literature, etc. At Stanford, however, they give a B.S. in chemistry, which required about 35 additional units in chemistry, math, and physics. Therefore, I spent four quarters satisfying these extra science requirements, and never did get a chance to sample logic and philosophy.

One of the requirements for graduation at Stanford was a 6-unit senior research project. At that time biochemistry was part of the chemistry department (1946), and I took a project with a junior faculty member (Dr. Edward L. Duggan). At that time biochemistry was in large part the chemistry and physical chemistry of biological molecules. One approach to understanding the three dimensional structure of proteins was to titrate reactive groups (acid, base, sulfhydral) to see which reactive groups were on the surface and which were internal. I titrated a protein for exposed and hidden sulfhydral groups for my senior project.

When the time came to decide about medical school, I decided against it: (1) I didn't have any money, and (2) I concluded that I liked medical research (biochemistry) better than dealing with patients. I can't remember now which of these factors carried the most weight, or whether it was a combination of the two. In any case, my next decision was whether to get a job as a chemist with a B.S. degree and make some money, or to go on to graduate school and get my Ph.D. in biochemistry.

I did look for jobs as a chemist, but jobs weren't that plentiful in 1947. The one job that I remember interviewing for was as a quality control chemist for the C&H Sugar company in Benicia, CA. After seeing the "advantages" of living in Benicia and of being a quality control chemist (the previous holder of the job had opened the wrong valve and dumped millions of gallons of purified sugar water into the bay), I decided on graduate school.

Again I was faced with the problem of money. I obtained a Research Assistantship with Dr. Hubert S. Loring at Stanford to obtain a masters degree in biochemistry. The fellowship did not permit any outside work, and only covered a little more than tuition. If I was going to starve, I decided to do it at the University of California at Berkeley where starving would be cheaper. For food and shelter I became a "house mother" at the University of California dorms in the old Kaiser shipyard in Richmond, CA. For pocket money I became a technician in the laboratory of Dr. Paul L. Kirk, who was the father of microchemistry. I prepared fibrinogen and ran microphosphate analyses for his research group. Some years later I published my own method for microphosphate analysis (A29).

By the end of the first year of graduate school, I had to start thinking about what I would do for my thesis. I had no strong ideas, so I decided to interview each of the professors to see who would be the best to work with, e.g., some of the professors were known to spend very little time with their students. On the basis of this survey, I selected Dr. Frank W. Allen, and it happened that he was working on nucleic acid chemistry, which fit into my chemistry background. By this time I was awarded a teaching assistantship and could give up being a technician and a house mother, and I moved to one of the many rooming houses in Berkeley.

My first publication came before obtaining my Ph.D. It was an article in Science (1952) in collaboration with Dr. Duggan (who had taken a faculty position at Berkeley), and was related to the new field of microchemistry (A1). Microliter pipets were available, but they were calibrated to contain. There were advantages to having micropipets that would deliver an accurate amount for the new paper chromatography procedures. To get pipets of less than 100 microliter to deliver, we used the newly available silicone mixture, Desicote, to coat the inside of the pipets, and calibrated them by acid titration (i.e., to deliver), rather than with mercury (i.e., to contain). It was many years later before I published another paper in Science.

Although I started out trying to develop a method for isolating nucleic acids from wheat germ, technical difficulties prevented me from continuing this work. I believe that the problem was created by all the fat that is in wheat germ.

I then turned my attention to resolving the conflicting reports in the literature that a small portion of ribonucleic acid is not cleaved by alkaline hydrolysis at room temperature. Six fractions of greater complexity than mononucleotides were separated from such a hydrolysate using paper chromatography and the new technique of paper electrophoresis. These six fractions, about 3% of total yeast RNA, were negative for deoxyribose and were further characterized by their spectra, purine and pyrimidine content (acid hydrolysis), etc. (A2). Therefore, RNA does contain some bonds that are stable to alkali. [Dissertation: Studies on the Substances Produced by the Alkaline Hydrolysis of Ribonucleic Acid from Yeast, 1952]

As a postdoctoral fellow at Berkeley, the next project that I worked on was in collaboration with Dr. Arthur M. Crestfield and Dr. Allen. We characterized the first high molecular weight RNA that had ever been isolated from yeast (A3). The innovative procedure that we used was to rapidly inactivate the enzyme, ribonuclease, by lysing the cells in boiling detergent (sodium dodecyl sulfate), thus preventing the enzymatic degradation of the RNA during isolation. Detergents had only recently been invented. They would have been a big help with my wheat germ DNA project (see above).

[ TOP ]


1952: Ph.D. Thesis on the chemistry of ribonucleic acid (RNA)

1952 - 1955: Further work on RNA

1956 - 1957: Effects of whole-body X-irradiation on the metabolism of deoxyribonucleic acid (DNA) and RNA in rat and mouse tissues

1958 - 1960: Chromatographic fractionation of DNA and RNA, especially, amino acid transfer RNA

1961 - 1962: Sensitization of bacteria to ultraviolet (UV) and X-irradiation by incorporating 5-bromouracil into DNA instead of thymine: the biology and the chemistry

1962 - 1969: DNA-protein crosslinking after the UV irradiation of bacteria: its discovery, chemistry, and biological importance

1962 - : Photochemistry and radiation chemistry of DNA, pyrimidines, and amino acids

1968 - : DNA repair processes after UV and X-irradiation; especially those that require genetic recombination: identification and genetic control

1976 - : Radiation and spontaneous mugtagenesis: mechanisms and genetic control

[ TOP ]


In 1954, I took a job as a Research Associate with Dr. Bertram V.A. Low-Beer in the Department of Radiology at the University of California at San Francisco. I was to work on the biological effects of whole-body X-irradiation on rats, with special emphasis on the effects of radiation on the activity of enzymes that metabolize the nucleic acids. There was some support for the premise that alterations in enzyme activity may occur before the morphological effects of X-irradiation can be detected.

Although some work had appeared on the catabolism of the purines after X-irradiation, ours was the first paper dealing with pyrimidines. While the catabolic activity toward uridine and cytidylic acid remained normal, that for uridylic acid increased with the dose of radiation in both liver and pancreas. In the spleen, however, the activity toward both uridine and cytidylic acid increased above normal, while that toward uridylic acid decreased markedly. These changes in enzyme activity in the spleen correlated in part with the radiation-induced involution process in the spleen. Thus, for any one enzymatic activity determination, the data varied from the control values by different amounts depending on whether they were calculated per mg of wet tissue, per mg dry tissue, or per total organ. There was no change in the enzymatic activities of the brain following X-irradiation. Of the tissues studied, only the brain displayed cytidine deaminase activity (A4, A12).

Because of the problems of tissue involution after X-irradiation, and the presence of different enzymes in different cell types, the level of enzyme activity in tissues after whole-body X-irradiation does not appear to be a useful measure of early radiation effects.

[ TOP ]


In 1956, I took a position as a Research Associate with Dr. Henry S. Kaplan, in the Department of Radiology at Stanford University School of Medicine, then in San Francisco. My first task was to develop a method for fractionating DNA and RNA from normal and tumor cells to see if differences could be observed. A method for the quantitative isolation of total tissue nucleic acids essentially free of protein was developed, as well as a column chromatography procedure (A7) for the comparison of the nucleic acids (RNA and DNA) from isologous newborn, adult and neoplastic thymus (A9).

The major differences seen between adult normal and tumor tissue was in the chromatographic profiles for the RNA. However, when RNA was isolated from newborn thymus tissue having the same growth rate as the tumor tissue, these differences disappeared. Thus, the altered chromatographic pattern of the RNA from thymic tumors appeared to be a manifestation of rapid growth and/or cell immaturity rather than being an intrinsic property of the neoplastic state (A9).

[ TOP ]


While I was developing the chromatographic procedure for the nucleic acids, it became obvious that I needed an independent assay to show that the different fractions eluted from the column were somehow unique so that our data showing differences between normal and tumor tissue could be validated. It had just been reported that specific enzymes were needed for the incorporation of specific amino acids into small molecular weight RNA, but it wasn't known if separate RNA molecules were also required for the binding of each amino acid.

If I could show that RNAs specific for different amino acids could be fractionated on my columns, I would make a major breakthrough in protein synthesis research, and also validate our tumor work. I arranged with Dr. Richard S. Schweet (City of Hope, Duarte, CA) to spend two weeks with him fractionating guinea-pig liver RNA. He took my fractions and assayed them for their ability to accept specific amino acids. Although the data showed that I was able to separate specific RNA molecules that accepted specific amino acids, and Dr. Schweet presented our data at a Gordon Conference, he was hesitant to publish because the current dogma said that if there were specific proteins for activating specific amino acids, who needs specific RNAs. Eventually this dogma mellowed, and Dr. Schweet did allow the paper to be published in 1959 (A5), and a few years later (1968) Dr. Robert W. Holley received the Nobel Prize for fractionating and characterizing the RNA molecules that are specific for specific amino acids. This is one of those examples of "what might have been".

As a consequence of learning about protein synthesis while running the column fractionation of different transfer RNAs (A5), I developed a rapid method for isolating calf thymus transfer RNAs without the need of an ultracentrifuge (A6), and evaluated a number of parameters involved in improving the washed microsome system for studying protein synthesis in vitro (A14).

(i) E. coli grown to logarithmic phase in, and plated on, rich medium are more resistant to X rays, UV radiation, and methyl methanesulfonate than cells grown in and plated on minimal medium. We have called this enhanced survival capability medium-dependent resistance (MDR) [see section on Recovery Phenomena]. MDR was associated with an increased ability to repair X-ray-induced DNA single-strand breaks, a reduction in X-ray-induced DNA degradation, and less X-ray-induced inhibition of protein synthesis. Postirradiation protein synthesis was concluded to be critical in allowing the high X-ray survival associated with MDR (A97). (ii) UV-irradiated (4 J/m2) uvrA cells showed a similar rate of protein synthesis, whether incubated in minimal medium or rich growth medium (A108).

In many of the papers described below, we studied the radiation induction of DNA repair enzymes, and the inhibition of their induction by drugs that inhibit RNA and protein synthesis (see section on Radiation Sensitizers).

[ TOP ]


In 1960, I became involved in the biochemical aspects of one of Dr. Kaplan's projects. It was concerned with the sensitization of cells to X-rays by the incorporation of purine and pyrimidine analogues into their DNA (A8, A11). These studies eventually led to clinical trials using intra-arterial 5-bromodeoxyuridine and X-ray therapy (A22, B5).

To understand the molecular basis for the radiation sensitization of cells by pyrimidine analogues, I began studying the photochemical and radiation chemical alteration of the nucleic acids in vitro and in vivo, and a series of papers were published.

I demonstrated that 5-bromouracil is more photochemically (254 nm) labile in vivo and in vitro than is thymine (A10, A16), studied the photochemical behavior, i.e., the relative rates of photochemical alteration, and the types of photoproducts produced with thymine, uracil, uridine, cytosine and bromouracil, whether irradiated singly or in mixtures, or in solution, or while frozen, or when in dry films (A15). I showed that the UV irradiation of thymine in the presence of adenine produces a different type of thymine dimer than when thymine is irradiated alone (A20). The chemical structure of the uridine photohydrate was determined by proton magnetic resonance spectroscopy (A26).

The X-radiation chemistry of uracil was studied in the presence and absence of oxygen (A23). A large number of products were identified, both oxidized and reduced. Similar products were produced when uracil was treated with ascorbic acid and ferrous sulfate, as well as by the autoirradiation of [3H]-uracil.

Bacillus subtilis is much more resistant to UV irradiation in the spore stage, and different thymine photoproducts are produced in the spore and vegetative stages, e.g., very few cyclobutane-type thymine dimers are formed in the spore stage, while it is the major photoproduct formed in the vegetative stage. A new type of thymine photoproduct is formed in the spore stage. These differences in photochemistry appear to be related in part to the differences in hydration of the DNA in the spore and vegetative states, since similar results were obtained for DNA in vitro when UV irradiated dry versus wet (A19).

A number of reviews and symposia talks were published on the photochemistry of DNA (B1, B3, B8, B17, B25, B26, B50).

My discovery of the photochemical cross-linking of DNA and proteins in bacteria was reported in 1962 (A13), and its biological importance (A17, A21) and its chemical basis (A18, A24, A31, A37) were established in a series of papers. Although many amino acids readily cross-link with uracil, they fell into four different reactivity groups. Cysteine and cystine were the most reactive, then came phenylalanine and tyrosine, followed by histidine, arginine and lysine, followed by glycine and serine (A31). The chemical structures of a mixed photoproduct of cysteine and uracil (A18), and of thymine and cysteine (A37) were established. The kinetics for the photochemical addition of [35S]-cysteine to various polynucleotides and nucleic acids, both single- and double-stranded, were determined (A24).

The biological importance of DNA protein cross-links was determined under conditions where the growth conditions of E. coli were changed such that cells became either more sensitivie or resistant to UV irradiation. These changes in survival were correlated with similar changes in the amount of DNA cross-linked to protein (A17). Similarly, freezing cells at different temperatures produces different degrees of sensitivity to killing by UV irradiation. These changes again correlated with the production of DNA cross-links, and did not correlate with the production of thymine dimers (A21).

DNA-protein cross-links appear to be amenable to postreplication repair, but they cause a longer local delay in DNA synthesis than do pyrimidine dimers (A73).

DNA-protein cross-links were the subject of a number of reviews (B22, B36), symposia talks (B2, B4, B6, B9, B16, B23, B24, B28), and a book entitled "Aging, Carcinogenesis and Radiation Biology: The Role of Nucleic Acid Addition Reactions" (D3).

[ TOP ]


When analogs of purines (e.g., thioguanine) and pyrimidines (e.g., 5-bromouracil) are incorporated into cellular DNA in place of the normal bases, cells become much more sensitive to the lethal effects of both UV and X radiation (A8). About a 17% replacement of thymine by 5-bromouracil was required before a significant radiation sensitization was observed (A11). This sensitization is due to the fact that 5-bromouracil is much more radiation sensitive than is thymine (see section on Photochemistry and Radiation Chemistry of DNA). These bacterial studies with 5-bromouracil led to animal studies, and ultimately to clinical trials using intra-arterial 5-bromodeoxyuridine and X-ray therapy (see section on Mammalian Cells).

A number of studies were performed using different drugs to see if they would sensitize cells to UV or ionizing radiation with the view of finding drugs that might be useful in the clinic as radiation sensitizers, or when the mechanism of action of a drug was well known, e.g., an inhibitor of protein or RNA synthesis, the drug was used to determine if a particular repair system was radiation inducible or not.

While these sensitizers worked well in E. coli, they did not work well in mammalian cells. This was later explained by the fact that much of DNA repair in E. coli is inducible. Therefore, drugs that inhibit RNA and protein synthesis would be expected to have a big effect on DNA repair in bacteria. However, DNA repair systems in mammalian cells show very little induction. In a symposium paper presented at the XI International Cancer Congress (1974), I suggested that the best approach for inhibiting DNA repair in mammalian cells was to isolate these enzymes, and then to develop specific inhibitors for them (B19).

The following sensitizers were studied:
acriflavine, B15
chloramphenicol A43, A55, A65, A71,A76, A80, A100
2,4-dinitrophenol A48, A54, A76
hydroxyurea A39, A44, B15
iodoacetic acid A44
N-ethylmaleimide A44
puromycin A43
rifampicin A97, A107, A109
sodium cyanide A44, A45
trinitrophenol A48

We discovered one sensitizer by accident. We found that most brands of agar, which are used to solidify plates for growing bacteria, contain an inhibitor (not yet identified) that selectively inhibits the recA-dependent pathway of nucleotide excision repair (A57). A more highly purified form of agar, Noble agar (Difco), does not generally contain this inhibitor, but we always tested each lot of Noble agar. Obviously, if one wants to study the effect of inhibitors of nucleotide excision repair on survival after UV irradiation, one needs to use agar plates that do not themselves inhibit the repair system under study.

In a search for drugs that might directly inhibit DNA repair processes, rather than just inhibiting their induction, we conducted several studies on quinacrine (Atabrin). Quinacrine belongs to a group of DNA intercalating agents known as amino acridines, which readily bind to DNA. Since quinacrine has long been used as an anti-malaria drug, there should be little problem in obtaining permission to use it in radiation therapy if it proves to be an effective sensitizer. When added to E. coli after X-irradiation, there was a marked sensitization to killing and a marked inhibition of the repair of DNA single-strand breaks (A41). When added before X-irradiation, quinacrine enhanced the yield of DNA single-strand breaks for cells irradiated in the absence of air (presumably by selectively inhibiting their repair), but not in the presence of air (A44). However, for Chinese hamster ovary cells, quinacrine had only a small effect on survival, and on the inhibition of the repair of DNA single-strand breaks (A56). Quinacrine had only a small sensitizing effect upon UV irradiated cells, both wild-type and uvrB (A74). Quinacrine does not appear to be the ideal radiation sensitizer.

[ TOP ]


Glycerol is an efficient scavenger of OH radicals. We used glycerol to determine the role of OH radicals in cell killing and the production of DNA double-strand breaks. We irradiated log phase, minimal medium-grown cells in the presence and absence of 1 M glycerol. The dose reduction factor for survival was 2.47, and that for the production of DNA double-strand breaks was 2.42. Although the production of other DNA lesions is also inhibited by the presence of glycerol (literature), it seems reasonable to ascribe the protection by glycerol in terms of survival to protection against the formation of DNA double-strand breaks, since these lesions appear to be less repairable than are base damage and single-strand breaks (A66).

The effects of these drugs are discussed in the section on Mammalian Cells.

[ TOP ]


Recovery phenomena have played an important part in radiation biology. In fact, the recovery phenomenon now known as Liquid-Holding Recovery (LHR) provided the first evidence that cells could recover from radiation-induced DNA damage. In 1949, Roberts and Aldous observed that when E. coli cells were held in the absence of nutrients for a period of time after UV irradiation, not only did the shapes of the survival curves change with time, but the cells also exhibited a much higher survival than if they were plated on growth medium immediately after UV irradiation.

Our genetic studies have shown that there are two major requirements for observing LHR. First, the cell must be proficient in nucleotide excision repair and, second, it must be at least partially deficient in postreplication repair. In fact, cells completely deficient in postreplication repair (i.e., recA cells) show the greatest amount of LHR (A25, A30, A86, A87, B7).

In a fully repair-proficient cell (i.e., wild-type), if a DNA lesion is not repaired by the excision repair process it can usually be repaired by the postreplication repair process. However, nucleotide excision repair can occur in buffer that is devoid of nutrients, but postreplication repair requires the presence of complete growth medium (i.e., macromolecular synthesis is required). Therefore, by holding cells in buffer without essential nutrients, both DNA replication and postreplication repair are prevented, and only excision repair can proceed. Such liquid holding is beneficial to the survival of cells deficient in postreplication repair because the more lesions that are repaired by the excision repair process before the cells are placed in growth medium where DNA replication can proceed, the fewer lesions that will be left to cause the formation of DNA daughter-strand gaps and DNA double-strand breaks, which are the substrates for repair processes in which the cells that exhibit LHR are deficient. [NOTE: In mammalian cells, a similar phenomenon after X irradiation is called "plateau phase recovery", or the "repair of potentially lethal damage".]

UV-irradiated, excision repair-deficient cells of E. coli exhibit minimal medium recovery, although a better name for this process would be rich medium lethality (see below). The observation was that excision repair-deficient cells show a higher survival after UV irradiation if they are plated on minimal medium rather than on rich medium, regardless of the type of medium in which they were grown before irradiation (A25, A34, A42, A94, B7, B10). Since this process was described in excision repair-deficient cells, the presumption was that some aspect of postreplication repair was inhibited by components of the rich growth medium. This has proven to be the case. MMR is blocked by recA mutations (A34), and the major processes of postreplication repair (i.e., the repair of DNA daughter-strand gaps and the repair of DNA double-strand breaks) are partially inhibited by rich growth medium (A94, A99, A100). This inhibitory effect of rich medium can be reproduced by just adding the common amino acids to minimal growth medium (A94). A small amount of amino acids added to minimal medium also enhances UV radiation mutagenesis (Leu+ reversion) in excision repair deficient strains of E. coli. This suggests that there may be a relationship between MMR and error-free postreplication repair (A78).

MMR is an inducible process, and amino acids appear to inhibit the repair process itself, rather than its induction (A100). MMR occurs in the uvrB recB recC recF strain, and we have correlated this phenomenon with the inhibition by rich growth medium of the repair of DNA daughter-strand gaps (A107).

A mutation, mmrA, has been isolated that blocks this effect of rich medium on survival and repair, but the product of the mmrA gene is not known (A99), nor is the mechanism known by which excess amino acids inhibit postreplication repair. The rep-38 and mmrA1 mutations are located very close to each other (~85 min on the E. coli linkage map) and have been suggested to be allelic, however, we showed that, although certain phenotypes are similar, other phenotypes are quite different, and suggest that these two mutations are not allelic (A123).

UV-irradiated (4 J/m2) uvrA cells showed a similar rate of protein synthesis whether incubated in minimal medium (MM) or rich medium (RM), however, they showed a severe depression in DNA synthesis when incubated in MM that lasted for about 30 min, and the normal rate of DNA synthesis was not re-established until about 60 min after irradiation. When a sample of these cells was incubated in RM after UV-irradiation, there was only a slight slowing of DNA synthesis, and the normal rate of synthesis was re-established by 60 min. An additional mmrA mutation, or growth retardation by valine, blocked both this extra DNA synthesis in RM and the inhibitory effect of RM on survival. These findings suggest that the absence of a marked delay in DNA synthesis observed in RM may be responsible for the inhibitory effect of RM on postreplication repair, and on the survival of UV-irradiated excision-deficient cells (A108).

Cells deficient in DNA single-strand binding protein (ssb) exhibit a much greater sensitivity to UV irradiation when plated on rich medium than when plated on minimal medium, regardless of the type of medium they were grown in before irradiation. Although UV irradiated ssb-113 cells incubated in rich medium did resume DNA synthesis slightly sooner than cells incubated in minimal medium, and there was a slight increase in the formation of DNA double-strand breaks for cells incubated in rich medium, the most dramatic effect was a much enhanced filamentation. Therefore, the rich medium-enhanced UV radiation sensitivity in ssb-113 cells appears to be due to an inhibitory effect of rich medium on cell division rather than on DNA repair (A104). MMR occurs in the uvrB recB recC recF strain, and we have correlated this phenomenon with the inhibition by rich growth medium of the recF-independent pathway for the repair of DNA daughter-strand gaps (A107). Rich growth medium also inhibits the repair of DNA double-strand breaks that are formed after the UV irradiation of uvrB cells (A118).

E. coli grown to logarithmic phase in, and plated on, rich medium (yeast extract-nutrient broth) were more resistant to X rays, UV radiation, and methyl methanesulfonate (MMS) than were cells grown in and plated on minimal medium. We have called this enhanced survival capability medium-dependent resistance. The magnitude of MDR observed after oxic X irradiation was greater than that observed after anoxic X irradiation, UV irradiation, or MMS treatment. MDR was not observed in stationary-phase cells with X or UV radiation. MDR was associated with an increased ability to repair X-ray-induced DNA single-strand breaks, with a reduction in X-ray-induced DNA degradation, and a reduction in the X-ray-induced inhibition of protein synthesis. The slow, recA-dependent repair of DNA single-strand breaks has been shown to be due to the repair of DNA double-strand breaks (A113). Therefore, the major benefit of MDR is to enhance the ability of cells to repair DNA double-strand breaks (A109, A119).

Postirradiation protein synthesis was concluded to be critical in allowing the high X-ray survival associated with MDR, because of the large radiosensitization caused by a postirradiation growth medium shift-down (which inhibits protein synthesis for a time), or by treatment with rifampicin (which inhibits RNA synthesis). RecA protein must be at least one of the proteins whose synthesis is critical to MDR, as judged by the absence of MDR or of a rifampicin effect in X-irradiated recA and lexA mutants. The results with X-irradiated temperature-conditional recA cells suggest that it is only after cells have been damaged that the recA gene plays a role in MDR (A97). MDR is an inducible process that is absent in recA, lexA, and recBC cells, and is partially inhibited by other mutations (A97, A98). Therefore, the best condition for studying DNA repair processes is when these processes are fully functional, i.e., under conditions of maximal MDR.

[ TOP ]


One of the more common experiments in the literature in the 1960's was the study of the effects of radiation and of drugs on the incorporation of radioactive precursors into the DNA of cells, yet the pitfalls inherent in the methods used were not widely appreciated. Generally the radioactive precursor was added immediately after the treatment, and the data for the incorporation of, e.g., thymine, into acid-insoluble material per aliquot of culture versus time of growth after treatment were plotted as a linear function. (The growth of cells of course is logarithmic, and a finite time is required for the radioactive precursor to be taken up by the cells and into the DNA; just to mention a few of the problems.) Since by this type of graph the treated cultures appeared to show a slower rate of uptake of thymine relative to the untreated control, the conclusion was offered that the treatment inhibited DNA synthesis, and the time of inhibition was dependent upon dose (radiation or drugs). Although the conclusions were generally correct, the data were incorrect.

We showed that the same results could be obtained by simply diluting unirradiated cultures to the same colony forming units per ml present in the irradiated cultures. Therefore, the shapes of the published curves were largely an artifact of the number of surviving cells. We developed a proper protocol for studying the effects of radiation (and other agents) on macromolecular synthesis that circumvents the pitfalls of the previous protocols (A27).

This new protocol, which is quantitatively equivalent to assaying directly the amount of DNA per cell, was then applied to the study of the effects of UV irradiation on DNA synthesis in radiation sensitive and resistant strains of E. coli (A32). The kinetics of cell division were also determined at the same time. There was a close correlation between survival and DNA synthesis kinetics after UV irradiation in strains having widely different sensitivities to UV irradiation. This paper also refuted and/or discussed a number of current dogmas, e.g., our results showed clearly that pyrimidine dimers do not constitute a permanent block to DNA synthesis. This paper should have opened a new era for the sensible determination of DNA synthesis kinetics, however, by this time people were tired of and/or confused by this type of assay, and radiation biology shifted to new procedures.

Some years later we completed a project on DNA synthesis kinetics, cell division delay and postreplication repair after the UV irradiation of frozen cells of E. coli (A73). Although the isokilling dose for cells at room temperature was 5 times that for frozen cells, nevertheless, at equal survival the DNA synthesis kinetics were equal (in confirmation of the above results; A32), even though the frequency of lesions per unit length of DNA were not equal. This paper also demonstrated that the major DNA lesions produced in frozen cells, presumably DNA-protein cross-links (A21), are repaired by postreplication repair (A73).

The question of whether discontinuous DNA replication operates only for the lagging strand or for both strands in E. coli is still unresolved. Most in vitro data indicate that DNA replication is semi-discontinuous, but in vivo data suggest that it is discontinuous in both strands. The majority of the in vivo data have come from the use of temperature sensitive DNA ligase mutants (lig). However, since DNA ligase is also required for certain types of DNA repair as well as for DNA replication, we wondered if the discontinuous synthesis seen in a ligase mutant could be an artifact of DNA repair. Using the appropriate mutants, we showed that the discontinuous DNA synthesis observed in the lig mutant was not due to the occurrence of mismatch repair (mutS or mutL), nucleotide excision repair (uvrB5), or base excision repair (ung) (A130).

DNA single-strand binding protein plays an important role in the replication of DNA that has been damaged by UV irradiation, and also protects single-stranded parental DNA opposite daughter-strand gaps from nuclease attack, so that postreplication repair can proceed effectively (A93).

Mismatch repair is the major mechanism by which cells detect and correct errors made during DNA replication. In E. coli this process is controlled by the methylation of adenine residues in -GATC- sequences in the parental strand of DNA. The absence of methylation in the newly synthesized strand instructs the mismatch repair enzymes to repair the mismatched base in the newly synthesized strand. If there is no methylation on either strand, as in a dam strain that lacks DNA adenine methylase, the mismatch repair enzymes don't know which strand to repair and occasionally cut both, resulting in DNA double-strand breaks.

Since the dam recA and dam recB double mutants are not viable, we investigated the possibility that this inviability was due to the formation, by undirected mismatch repair, of DNA double-strand breaks, which then could not be repaired by a recA or recB strain. This hypothesis would also suggest that the viability of a dam recB strain should be restored by an additional mutation (e.g., mutL or mutS) that inactivates mismatch repair. Our data confirmed these predictions (A114).

[ TOP ]


Photoreactivation is the process by which UV irradiated cells can be brought back to life by a second irradiation with visible light; it is due to the light activated enzymatic splitting of cyclobutane-type pyrimidine dimers. Since the molecular basis of the response is known, the ability to photoreactivate a response (e.g., survival, mutagenesis) has been used as a diagnostic test for the involvement of pyrimidine dimers in the response. To prevent unwanted photoreactivation, experiments are generally conducted in laboratories with no windows that are lighted with "gold" fluorescent lamps or yellow "bug" lamps, since the light (>500 nm) from these lamps does not activate the photoreactivating enzyme of E. coli.

We have also used this diagnostic test to show that fewer pyrimidine dimers are formed in cells UV irradiated while frozen versus at room temperature (A21), that detectable levels of pyrimidine dimers are formed in the very UV radiation sensitive uvrA(BC) recA cells (A30), that most of the lesions that are normally repaired by postreplication repair can be repaired by photoreactivation (A35), to determine the kinetics for the postreplicational repair of pyrimidine dimers (A42), to decrease the yield of DNA double-strand breaks produced enzymatically in UV-irradiated cells (A63), to study an excision repair process that is independent of the uvrA(B) genes (A72), and to show that the recB gene product (exoV) interferes with photoreactivation, apparently by binding at or near pyrimidine dimers (A87).

In addition, we showed that the map location reported in the literature for the gene for photoreactivation (phr) was incorrect, and we mapped and reported the correct location (A79).

It was thought that photoreactivation does not occur after ionizing irradiation, however, Myasnik and Morozov reported in 1977 that they did observe photoreactivation in the doubly repair deficient strain E. coli uvrA recA after exposure to ionizing radiation. This paper raised two questions to us: (1) does the observed photoreactivation after ionizing irradiation utilize the same photoreactivating enzyme as after UV irradiation, and (2) are pyrimidine dimers formed in the DNA of cells after gamma irradiation?

Since we had a uvrA recA phr strain available, we were able to show that photoreactivation after gamma irradiation uses the same enzyme as after UV irradiation. In addition, by direct chemical analysis, we showed that the same isomer of the thymine dimer that is formed by the UV irradiation of cells is formed by gamma irradiation, albeit at a much reduced rate (A81, B30).

Excitations can be produced by ionizing radiation by two major mechanisms; by the direct excitation of the target molecule, and by the formation of Cerenkov radiation (UV and visible radiation in the present context) in the water surrounding the cells. The formation of Cerenkov UV radiation by gamma irradiation can be blocked, e.g., by using a very concentrated cell suspension (self shielding), or by suspending the cells in a solution containing a UV radiation absorber (e.g., DNA). Using these tricks, we showed that 137Cs-gamma radiation produces thymine dimers both by the direct excitation of DNA and via Cerenkov radiation, while 50 kVp X-rays, whose energy is below the threshold for producing Cerenkov radiation, still produces thymine dimers by direct excitation. The Cerenkov effect can be enhanced by placing a column of water around the cells contained in a quartz tube (i.e., a UV radiation transmitting tube) (A85). This work on Cerenkov radiation raises concern about the quality of the results from ionizing radiation mutagenesis experiments where the numbers of cells per ml that were irradiated was less than 109, since a significant contribution of pyrimidine dimers (i.e., a UV effect) could complicate the interpretation of the results (see section on UV/X-ray Synergism). Other authors subsequently confirmed this concern.

[ TOP ]


While my laboratory worked mainly on E. coli, when appropriate, we did work with mammalian cells.

(1) We investigated the effect of quinacrine on survival and on the inhibition of the repair of DNA single-strand breaks in X-irradiated Chinese hamster ovary (CHO) cells (A56). The effect on both end points was small.

(2) We investigated the nature of the oxygen effect on X-ray-induced DNA single-strand breaks in CHO cells. The initial yield of single-strand breaks was about 4 times higher in aerated samples than in those under extreme hypoxia. In cells heated to inactivate repair enzymes (15 min at 70oC) the number of single-strand breaks produced per rad increased several fold in air or moderate hypoxia (200 ppm O2), but under extreme hypoxia (<10 ppm O2) it was close to that for normal cells in air. When cysteamine was added to heated cells under extreme hypoxia, the DNA break yield was similar to that for normal cells irradiated under extreme hypoxia. Similarly, when sodium dithionate was added before irradiation to remove oxygen chemically, the breakage yield for heated cells was close to that for normal cells under extreme hypoxia.

Apparently heating cells does more than inactivate enzymes, it also causes important molecules that modulate the effects of X irradiation to leak out of cells. Thus, if a hydrogen donating species, cysteamine, is added or rigorous anoxia is obtained chemically, then heated cells show the same yield of DNA single-strand breaks as do normal cells under extreme hypoxia. Although most of the oxygen effect on the yield of DNA single-strand breaks appears to be the result of fast chemical reactions, our data do not exclude the existence of a very rapid repair process (e.g., one requiring DNA ligase only) for the rejoining of single-strand breaks (A58). Our results would suggest, however, that the amount of such ultrafast repair must be considerably less than hypothesized to occur in bacteria based upon somewhat similar heat inactivation studies (A44).

(3) We have investigated the extent to which DNA single-strand breaks produced in CHO cells by X-irradiation under air or nitrogen are rejoined after aerobic incubation in either phosphate or carbonate buffered salt solution or in growth medium, in order to detect any differences in the rejoining processes of aerobically and anaerobically irradiated cells. The initial rates of rejoining were about the same in buffer or in growth medium, but was somewhat slower for cells irradiated under nitrogen. After the initial rapid rate of repair, a slower rejoining rate was observed for cells incubated in growth medium. This slow process was not observed in cells incubated in buffer, due to buffer-induced DNA breakdown. Therefore, a medium-dependent DNA repair system could not be established conclusively in mammalian cells (A61), as we had done for bacteria (A46).

(4) The effect of actinomycin D on cell cycle kinetics in CHO and mouse mammary tumor cells (EMT6) was determined (A68). There was a reduced rate of progression of the cells through S phase and a G2 arrest, the degree of which was drug dose dependent. The lethal effects of the drug on the two cell lines were comparable. The drug also produced DNA single-strand breaks and/or alkali-labile bonds. After removal of the drug, the extent of repair of these breaks was drug dose dependent.

(5) Occasionally a course of radiation therapy must be abandoned or protracted because a patient demonstrates an unusually severe skin reaction at the site of therapy. We grew fibroblasts from untreated skin from cancer patients who had normal and abnormal skin reactions to radiation therapy, and determined the X-ray sensitivities of these cells in vitro. The cells from the sensitive patients were also more radiation sensitive in culture. The survival curves of the sensitive cells were biphasic, with a sensitive component and a resistant (normal) component. These results suggest that the cells from these patients may be unusually radiation sensitive in one portion of their cell cycle. (A84).

(6) Postreplication repair after UV irradiation was studied in fibroblasts from normal humans and from xeroderma pigmentosum Group A (XPA) patients (A112). We determined that DNA double-strand breaks are produced and repaired in XPA cells as they are in E. coli uvrB cells (A111) during the process of postreplication repair. Similar results were also observed for normal human cells, but only after about a 5-fold higher UV radiation dose.

(7) I also participated in a clinical trial to evaluate the effectiveness of incorporated 5-bromodeoxyuridine in tumors in enhancing the effectiveness of X-ray therapy (A22, B5). My job was to determine the percent replacement of thymine by 5-bromouracil in the DNA of tumor biopsies. Although the tumor cells that contained 5-bromouracil showed enhanced radiation sensitivity (head and neck tumors), so did the surrounding normal cells.

[ TOP ]


In certain bacterial and yeast strains, combined UV- and X-irradiation was reported to inactivate more cells than would be expected if the effects of both radiations were additive. We studied the genetic control and the molecular basis of this synergistic interaction between UV and X radiation. Synergism was observed, but to a reduced extent, in uvrA(BC) and polA mutants, but was absent in exrA(lexA), recA, recB, or recC mutants. Alkaline sucrose gradient studies demonstrated that the rec and lex gene-controlled, growth-medium-dependent (Type III) repair of X-ray-induced DNA single-strand breaks was inhibited by prior UV irradiation. [NOTE: This Type III (growth-medium-dependent; recA, recB, recC dependent) repair of DNA single-strand breaks is now known to be the repair of DNA double-strand breaks (A113).] This inhibition of DNA repair probably explains the synergistic effect of these two radiations on survival (A50, B14).

One unexplained feature of our results with the uvr mutants was that, after an exposure to UV radiation that leaves a surviving fraction of about 10-3, the subsequent X-ray survival curve has a shoulder. Thus, the prior UV irradiation serves to protect the cells from low doses of X-rays (up to about 5 krads), but sensitizes the cells to higher doses of X radiation (A50, B14). One possible mechanism for this protection might be the induction of DNA repair enzymes (see below).

When log phase cells of wild-type E. coli K-12 were maintained in growth medium after X irradiation, they became progressively more resistant to a subsequent exposure to UV or X radiation. The time to achieve maximum resistance was 60 min. This X-ray-enhanced resistance was characterized by the production of a shouldered survival curve with no change in the final slope for a second exposure to X radiation, whereas after UV irradiation both an enhancement of the shoulder and a reduction in the final slope was observed.

The time course for the decay of this X-ray-induced resistance was interesting. The X-ray-induced resistance to UV irradiation was still maximal after 360 min of incubation in growth medium, but the resistance to X radiation was almost back to normal after 240 min. "Studies on the factors that affect the decay of resistance should prove as interesting as those that affect its induction." Unfortunately, we never had the opportunity to continue these studies.

The uvrB5, uvrD3, polA1 and certain exrA strains (W3110 background) demonstrated this X-ray-induced resistance to subsequent UV or X irradiation, but recA, recB, lexA, and other exrA strains did not. This induced resistance appeared to depend upon protein synthesis between the two irradiations (A71, B14).

[ TOP ]


Contrary to popular belief, near-UV radiation (315-400 nm; also called UV-A) produces markedly different photochemical and photobiological effects than does far-UV radiation (210-315 nm). In general, since DNA does not absorb in the near-UV waveband, near-UV radiation exerts most of its effect on DNA indirectly, i.e., by producing excitations in other chromophores, which frequently require oxygen as a cofactor (in sharp contrast to the case for far-UV radiation). These excited chromophores then interact with DNA to produce damage.

In a study on the mechanism of cell killing by near-UV radiation, we were surprised to discover that a large component of this killing was due to radiation damage to cell membranes. This extra lethality was observed when cells were plated in the high-ionic-strength minimal medium normally used for bacterial growth, but was greatly reduced or eliminated by plating on low-ionic-strength medium such as nutrient broth. [NOTE: The treatment of cells with high-ionic-strength medium is a classical method for assaying for membrane damage after treating bacteria with heat.] Such membrane damage was not observed when cells were irradiated with far-UV radiation (254 nm) (A88).

This sensitivity to minimal medium (high ionic strength) was increased by increasing the salt concentration of the medium, and by increasing the pH of the medium. In addition, this sensitivity was greatly increased by adding a low concentration of commercial glassware washing detergent that had no effect on unirradiated cells, or cells that had been exposed to far-UV radiation. These results may explain the large variability often observed in the literature for near-UV radiation survival data, and demonstrate that, at least on minimal medium plates (high ionic strength), membrane damage contributes significantly towards killing by near-UV irradiation. This membrane effect is largely oxygen dependent (A88).

Several bifunctional furocoumarins (psoralens) are used in conjunction with near-UV radiation (UV-A) for the treatment of psoriasis (PUVA therapy; Psoralen plus UV-A), however, there are still many unanswered questions about the photochemistry of the furocoumarins and of the repair of the DNA damage that they produce. Therefore, we undertook to study the sensitivity of DNA repair-deficient strains of E. coli to various furocoumarins (both monofunctional and bifunctional) in combination with different wavelengths of near-UV radiation (A89).

The polA gene product (DNA polymerase I) appears to play a major role in the repair of furocoumarin monoadducts, but plays little or no role in the repair of DNA cross-links produced by bifunctional furocoumarins. However, the recA gene product plays a key role in the repair of both monoadducts and cross-links. 3-Carbethoxy psoralen and 5-7-dimethoxycoumarin have been reported to form only monoadducts, however, our results suggest that these compounds form additional types of photoproducts, since the genetic control of their repair response is quite different from that for angelicin, a "pure" monoadduct former. Furthermore, the type of biological response for certain of these compounds depends upon the wavelength used to produce the photochemistry. In summary, we developed a biological assay to detect the presence of furocoumarin monoadducts and cross-links in E. coli that is much more sensitive than any chemical assay, and we discuss the importance of the wavelength of near-UV radiation needed to achieve a given result, i.e., to produce selectively monoadducts or cross-links (A89).

[ TOP ]


Spontaneous mutations are alterations in chromosomal DNA that arise by unknown mechanisms. However, such mutagenic mechanisms can be grouped into three categories that involve errors in DNA replication, recombination, and repair. We undertook to evaluate the roles of genetic recombination and of error-prone DNA repair in spontaneous mutagenesis by studying the roles of DNA-repair genes (uvrA, uvrB, uvrD, recA, recB, lexA, and umuC) on reversion, frameshift and forward spontaneous mutagenesis.

Rich growth medium enhanced µ (the spontaneous mutation rate per bacterium per cell division) in a wild-type strain, but not in a uvrB5 strain. In minimal growth medium, the uvrA and uvrB strains had the largest µ (1.9 - 6.2-fold greater than that for isogenic wild-type strains, depending on the mutation assay). The uvrB strains carrying lexA, recA, umuC, or both the uvrD and recB mutations (in combination), i.e., mutations that inhibit error-prone DNA repair, had the lowest µ values (~10-fold less than the uvrB strain). The recA and lexA mutations also reduced µ (by ~2-fold) in uvr+ strains. The genetic control of the error prone repair-dependent sector of spontaneous mutagenesis was shown to be qualitatively similar to the genetic control for UV-radiation mutagenesis. The umuC mutation, which drastically reduced spontaneous mutagenesis, had no effect on genetic recombination. It is proposed that the low level of spontaneous mutagenesis observed in the recA, lexA, umuC, and the uvrD recB strains is due to errors made during DNA replication, while the enhanced level of spontaneous mutagenesis observed in the wild type, and especially in the uvrA and uvrB strains, is due to excisable lesions that are produced in the DNA by normal metabolic reactions, and that such unexcised lesions induce mutations via error-prone DNA repair (A90, B43, B44).

The nature of the putative spontaneous mutagenic lesions in E. coli can be inferred from the specificity of the UV endonuclease. The uvrA and uvrB strains (i.e., strains deficient in the UV endonuclease) are very sensitive to UV radiation, but show little sensitivity to ionizing radiation. Similarly, these strains are hypermutable by UV radiation, but not by ionizing radiation. These facts suggest that the lesions in DNA leading to "apparent" spontaneous mutagenesis are not produced by free radicals (i.e., characteristic of ionizing radiation), but are produced by excited state reactions (i.e., characteristic of UV radiation).

To validate this discovery that much of spontaneous mutagenesis is due to the error-prone repair of "UV-like" damage produced in DNA by normal metabolic reactions, we tested the mutagenicity [lacZ53 (am) -> Lac+] of 19 amino acids in E. coli K-12 uvrB. Cystine, and to a lesser extent, arginine and threonine were found to be antimutagenic; only phenylalanine was found to be mutagenic (regardless of its source or method of manufacture). At 2 mM, phenylalanine induced mutants at ~2-fold above background (a value that is "easy" to quantitate with bacteria). Tyrosine, and to a lesser extent, tryptophan inhibited the mutagenicity of phenylalanine. Phenylalanine mutagenesis was not detected in a wild-type strain (i.e., excision repair proficient). Thus, phenylalanine metabolism appears to produce "UV-like" lesions in DNA, which, if not excised, can produce mutations by error-prone DNA repair. Phenylalanine metabolism is offered as just one of the presumably numerous metabolic activities that contribute to spontaneous mutagenesis (A116).

An exhaustive review of the world literature has supported our conclusions (A90) that much of spontaneous mutagenesis in many different types of organisms (from bacteriophage to eukaryotes) is due to the error-prone repair of normal metabolic damage to DNA. This was not an easy review to prepare, since there are very few papers published that focus exclusively on spontaneous mutagenesis. Rather, what we had to do was to collect papers on radiation and chemical mutagenesis where DNA repair mutants had been studied, and to calculate the spontaneous mutation frequency for DNA repair deficient cells versus their parental cells from the data presented in these papers for untreated cells. These data suggest that 50-90% of spontaneous base substitutions are due to error-prone DNA repair. On the other hand, spontaneous frameshifts and deletions seem to result from mechanisms involving recombination and replication (B42).

In 1992, I published another review on spontaneous mutagenesis (B52) to bring the data on its genetic control up-to-date, and also to discuss the many subtle experimental variables that can have a marked effect on the amount and types of spontaneous mutations produced. Since spontaneous mutations are the resultant of all of those cellular processes that are mutagenic and all of those cellular processes that are antimutagenic, any experimental condition that upsets this balance will have an effect on the results. A thorough understanding of these variables that affect spontaneous mutagenesis eliminates the need for a theory of "directed" mutagenesis in starved cells.

(1) In 1975, we began our studies on UV radiation mutagenesis, since we wished to know which of the multiple pathways of postreplication repair were error-free, and which were error-prone and thereby produced mutations. We were particularly interested in the uvrD gene, now known to code for DNA helicase II. We measured Leu+ and Trp+ reversion after UV irradiation. By comparing the uvrD strain with its parental strain (either uvr+ or uvr-), we should be able to conclude whether a uvrD mutation has an effect on UV radiation mutagenesis or not. There were two accepted ways of plotting mutagenesis data on log-log plots, i.e., mutants/survivors versus survival or versus dose. However, these two plots led to two different conclusions about uvrD. In plots against survival, uvrD strains were clearly less mutable than their parental strain, but when plotted against dose there was no difference. Which is the correct interpretation? I discuss the molecular relevance of these two types of plots (e.g., equal dose means an equal number of initial lesions for the cell to deal with, while equal survival means an equal number of lethal events regardless of the dose) in an attempt to choose which type of plot is best for comparing two strains having greatly different sensitivities to killing. When survival is less than 100% and the two strains being compared show different sensitivities to killing, it seems more appropriate to compare mutation frequency as a function of survival. What would be ideal, of course, is to compare the number of mutations produced per number of repair events in wild-type versus repair deficient strains, but such information is not presently available, and probably will never be (A70).

(2) In order to determine if the differences that I observed in the UV-radiation mutability of the uvrA and uvrB strains (A70) was due to differences in these two genes or to their backgrounds (since one was K-12 and one was B), we constructed isogenic strains in the K-12 background and tested their UV-radiation mutability for Leu+ reversion. The medium supplementation to allow mutation expression was either DNB (0.02% nutrient broth) or Dleu (1 mg/ml leucine). The UV radiation survival on these two media were markedly different. The survival curve on DNB was exponential, while that on Dleu was shouldered, as had previously been observed for minimal medium recovery (see section on Recovery Phenomena). The mutation yield on these two media also were markedly different, being greatly enhanced on DNB. Since DNB inhibits minimal medium recovery and enhances mutagenesis, it suggests that minimal medium recovery is an error-free process.

When the data for mutagenesis on the two media were plotted against log of percent survival, there was no difference between the two media, suggesting that, in this case, lethality and mutagenesis were enhanced to the same extent by DNB. It was curious, however, that these data showed two distinct slopes; a steep induction of mutations from 100-35% survival, and a more shallow slope at lower survival. This suggests that there may be two different molecular mechanisms for UV-radiation mutagenesis over these two ranges of cell survival (A78).

(3) The nonlinear nature of the mutant frequency response to UV radiation was noted by several authors in early studies on mutagenesis in bacteria. When these mutant frequency data were plotted against dose on a log-log plot, a slope of 2 was obtained, suggesting that two radiation "hits" were required to produce mutations. The nature of these two hits was the subject of much discussion in the literature (e.g., Witkin). Since one hit must be in the gene that is mutated, the other hit could be responsible for the induction of error-prone, mutagenic processes. Alternatively, the production of the mutagenic lesion itself may require two hits (e.g., overlapping DNA daughter-strand gaps). These early experiments were performed at high doses of UV radiation. However, when low doses of UV radiation were used and the data were plotted with linear coordinates (Doudney), a distinct 1-hit (linear) region was observed at the lower doses, and 2-hit kinetics (nonlinear) were observed at the higher doses.

By determining many experimental points over a wide dose range, we observed the presence of three different mechanisms for UV radiation mutagenesis in E. coli uvrB5. We also saw the 1-hit response at low doses and the 2-hit response at high doses, but we observed another response with different kinetics in the 1-3 J/m2 range for Lac+ reversion. This new and unusual kinetic response is specific for nonsense reversion. The "hit" number could not be determined for this third kinetic element because it occurred over too short of a dose range (A82).

(4) We wished to determine if the putative three different mechanisms of UV-radiation mutagenesis that we deduced from our kinetic analyses had biological relevance. To do this we fractionated our results for the total yield of reversion mutations into those due to true back mutations and those due to suppressor mutations. When we did this we found that the 1-hit mutations were due to true back mutants, and the 2-hit mutations were due to suppressor mutations. The third process, which we termed KC for "kinetically complex", since we could not determine the "hit number" accurately because of the short range of doses over which it was expressed, is also associated with suppressor mutations, and appears to be an inducible process (A96).

(5) In the uvrB background, the uvrD and recB genes are involved in separate mutagenic DNA repair pathways after UV irradiation, but there remains at least one other error-prone DNA repair pathway that is uvrD recB independent (A83).

(6) In a study to resolve the conflicting data in the literature concerning the role of the umuC gene in ionizing radiation mutagenesis, we also studied the role of umuC in UV radiation mutagenesis for reversion, frameshift and forward mutations in the same cell line. Not surprisingly, we found that the umuC mutation blocked all types of UV radiation mutagenesis, but it did not block all types of gamma-radiation mutagenesis (A105). This latter point is discussed more fully under the section on Ionizing Radiation Mutagenesis.

(7) We studied the roles of the umuDC genes in the UV and gamma-radiation reversion of sequenced trp frameshift mutations, since it was reported that a trpE9777 mutation is not reverted by ionizing radiation. Unlike the UV radiation reversion of base-substitution mutations, the reversion of trp frameshift mutations was not enhanced in a uvrA umuC strain by photoreactivation after a post-UV-irradiation incubation. Therefore, the UmuDC proteins are suggested to have functions in the radiation induction of frameshifts that are more complex than are their functions in the induction of base substitutions (A122). The results for gamma radiation are discussed in the section on Ionizing Radiation Mutagenesis.

(8) When E. coli umuC122::Tn5 was chemically mutagenized to block the residual gamma-radiation mutagenesis observed in umuC strains, two mutants were isolated. These mutants turned out to have mutations in two genes that had previously been isolated and sequenced (i.e., ruvA and ruvB), but for which there were no mutation data. The ruvA and ruvB mutations had a partial effect on both UV- and gamma-radiation mutagenesis, but the effect was larger for UV radiation. The reduction of mutagenesis by a ruvA mutation was at all sites, in contrast with the case for umuC where only specific sites were affected. This effect by ruvA was not due to an effect on the induction of the recA or umuC genes (A129). The results for gamma-radiation are discussed more extensively in the section on Ionizing Radiation Mutagenesis.

(9) An R-plasmid (pEB017) was isolated that restored recombination ability to recA56 cells, and conferred enhanced resistance to UV-radiation, and enhanced UV-radiation mutability to the wild-type, recA56, and umuC36 strains (A132). A recA-like fragment was cloned from this plasmid that conferred both UV resistance and UV mutability to a recA56 strain. The restriction map of this fragment differed slightly from that for the recA gene from wild-type E. coli.

(10) Several symposia talks were published on UV radiation mutagenesis (B29, B32, B51, B54).

(1) In order to resolve the conflicting reports in the literature concerning the role of the umuC gene in ionizing-radiation mutagenesis, we studied the effect of a umuC mutation on gamma- and UV-radiation mutagenesis (nonsense, missense, and frameshift mutation assays) in E. coli. Although UV-radiation mutagenesis seems to depend almost exclusively on the umuC gene, for gamma-radiation mutagenesis the deficiency of a umuC strain varied from none to a 50-fold deficiency, depending upon the assay. The degree of mutability of the umuC strain for a given assay also depended upon the radiation dose. Thus, the apparent conflict in the literature on the role of the umuC gene in ionizing radiation mutagenesis was due to the use by the two laboratories of two different assays at two different loci having two different requirements for the umuC gene for the expression of mutagenesis. Therefore, while UV-radiation mutagenesis seems to be totally dependent upon the umuC gene, ionizing radiation mutagenesis shows both umuC-dependent and umuC-independent modes (A105).

(2) We have used several experimental approaches to learn why gamma-radiation mutability is only partially dependent upon the umuC gene, while UV-radiation mutagenesis seems to be totally dependent upon the umuC gene. First, we established that a recA mutation blocked Arg+ reversion mutagenesis at the same doses that induced mutations in the umuC122::Tn5 and wild-type strains, indicating that both the umuC-dependent and umuC-independent mechanisms of gamma-radiation mutagenesis function within recA-dependent misrepair.

Next, we showed that the umuC mutation blocked all oxygen-dependent base substitution mutagenesis, but did not block all of oxygen-dependent frameshift mutagenesis. For anoxically irradiated cells, the yields of GC -> AT and AT -> GC transitions were essentially umuC independent, while the yields of (AT or GC) -> TA transversions were heavily umuC dependent.

Therefore, our data for anoxically irradiated cells support the hypothesis that gamma-radiation produces two kinds of DNA lesions that require recA-dependent misrepair to induce mutations. For base-substitution mutagenesis, one kind of lesion requires the umuC gene and produces transversion mutations, while a second kind of lesion produces transition mutations and does not require the umuC gene. For cells irradiated in the presence of oxygen, there seems to be additional kinds of lesions whose mutagenic potential for base substitutions (but not frameshifts) is completely dependent on the umuC gene (A127).

(3) To gain additional insights into the portion of gamma-radiation mutagenesis that is not under the control of the umuC gene, we chemically mutagenized a umuC strain and isolated mutants that were blocked in gamma-radiation mutagenesis. Two such mutants were subsequently shown to be ruvA and ruvB mutants. The ruvA and ruvB genes had been previously cloned and sequenced, but their roles in mutagenesis had not been studied.

The RuvA and RuvB proteins appear to act in a complex, since a mutation in either gene has the same effect on mutagenesis and survival. However, in combination with a umuC mutation, a ruvA(or B) mutation had a synergistic effect on survival, suggesting that the umuC and ruvAB genes function in two different DNA repair pathways that compete for the same class of DNA damage.

A mutational spectrum analysis in the ruvA strain demonstrated a general depression of mutagenesis at all sites of both umuC-dependent and umuC-independent gamma-radiation mutagenesis, which is in marked contrast with the site-specific reduction in gamma-radiation mutagenesis that is observed in the umuC mutant. This general depression in radiation mutagenesis seen in the ruvAB mutants is not due to a depression in the induction of the recA and umuC genes.

The ruvA(B) strains were more deficient in UV-radiation mutagenesis than in gamma-radiation mutagenesis. Since the umuC mutation blocks all UV-radiation mutagenesis, this suggests that there is ruvA-dependent and ruvA-independent UV-radiation mutagenesis. For gamma-radiation mutagenesis, the ruvA(B) genes appear to function both in the umuC-dependent and umuC-independent mechanisms.

These results indicate that gamma-radiation mutagenesis is much more complex than is UV-radiation mutagenesis (A129), which is consistent with the greater complexity of the types of damage produced in DNA by gamma radiation compared with UV radiation.

(4) We studied the role of the umuDC genes in the UV-radiation (see above) and the gamma-radiation reversion of sequenced trp frameshift mutations, since it had been reported that a trpE9777 mutation was not reverted by ionizing radiation. We found that gamma-irradiation caused the reversion of all of the trp frameshifts studied, and the plasmid pKM101 enhanced this reversion 10- to 50-fold, which underscores the requirement for the UmuDC proteins. None of these reversions were frameshift suppressor mutants (A122).

(5) Ionizing radiation produces more long deletion mutations than does UV radiation and, similarly, ionizing radiation produces more DNA double-strand breaks than does UV radiation. We designed experiments to determine if long deletion mutations were the consequence of the repair (misrepair) of DNA double-strand breaks. We had shown previously (A119) that a recB mutation blocks the repair of DNA double-strand breaks after ionizing irradiation, but does not block the repair of DNA single-strand breaks (A46, A113). In our mutation studies (A131), we found that a recB mutation completely blocked the formation of long deletions, while base substitution and frameshift mutations were little affected. In addition, physiological conditions that enhance the ability of cells to repair DNA double-strand breaks (i.e., conditions of MDR) also enhanced the yield of long deletions. The formation of long deletions is linear with dose, as is the formation of DNA double-strand breaks.

The RecF pathway for the repair of DNA double-strand breaks does not appear to contribute to the formation of long deletions, but we discuss this in the context that the RecF pathway repairs DNA double-strand breaks where they are formed, whereas the RecB process unwinds the broken DNA until it reaches a Chi site (GCTGGTGG 3'), and then cuts the DNA and initiates recombination at the site of this enzymatic break. Therefore, numerous primary DNA double-strand breaks would be concentrated at a nearby Chi site, thereby enhancing or diminishing the apparent yield of long deletions per double-strand break, depending upon the location of the Chi site relative to the position of a break needed for the initiation of the recombination event required to produce a long deletion mutation (A131).

(6) Gamma radiation (anoxic)-induced and spontaneous episomal lacI(d) mutations in wild-type Escherichia coli strain NR9102 were sequenced. The most commonly found radiation-induced mutations were base substitutions (44% transversions and 41% transitions). The radiation-induced spectrum consisted of: 23% G:C->A:T, 18% A:T->G:C, 17% G:C->T:A, 14% G:C->C:G, 8% A:T->T:A, 6% A:T->C:G, 8% single-base deletions, 5% multiple mutations, 3% multi-base deletions, and essentially no single- or multi-base additions. The spontaneous spectrum differed substantially from that of the radiation-induced spectrum for virtually every class of mutation. However, like the radiation-induced spectrum, the spontaneous spectrum showed enhanced mutagenesis at G:C sites, strand asymmetry, and enhanced mutagenesis when G or C were the nearest neighbors (A133).

DNA sequence analysis of spontaneous and gamma-radiation (anoxic)- induced lacI(d) mutations in Escherichia coli umuC122::Tn5 showed a differential requirement for umuC at G:C vs. A:T sites, and for the production of transversions vs. transitions. The umuC strain was not deficient in spontaneous mutagenesis, and the mutational spectrum was very similar to that for the wild-type strain (A133). The yields of radiation-induced mutation classes in the umuC strain (as a percentage of the wild-type yield) were: 80% for A:T->G:C transitions, 70% for multi-base additions, 60% for single-base deletions, 53% for A:T->C:G transversions, 36% for G:C->A:T transitions, 25% for multi-base deletions, 21% for A:T->T:A transversions, 11% for G:C->C:G transversions, 9% for G:C->T:A transversions, and 0% for multiple mutations. Based upon these deficiencies and other factors, it is concluded that the umuC strain is near-normal for A:T->G:C transitions, single-base deletions and possibly A:T->C:G transversions; is generally deficient for mutagenesis at G:C sites and for transversions, and is grossly deficient in multiple mutations (A134).

[ TOP ]


(1) A polA1 strain, deficient in DNA polymerase I, is partially deficient in nucleotide excision repair after UV irradiation. In 1973 we reported evidence that DNA polymerase III, the replication polymerase, also participates in excision repair (A51), probably in long-patch excision repair (Hanawalt), which is not inhibited by a polA1 mutation, but is inhibited by a recA mutation.

(2) We next reported evidence that the exrA(lexA) gene functions in a different pathway of excision repair than does polA. The exrA gene also functions in postreplication repair (A52).

(3) UV irradiated E. coli K-12 wild-type cells were sensitized by a postirradiation treatment with 10-2 M 2,4-dinitrophenol (DNP). This effect was not seen in strains carrying a uvr mutation, suggesting that DNP interferes with the excision repair process. The polA strain was sensitized to the same extent as the wild-type strain, while the exrA(lexA) strain was not affected by DNP treatment. Recombination deficient strains, recA, recB, and recA recB, were protected by DNP treatment after UV irradiation. This protection was abolished by the addition of a uvr mutation.

Alkaline sucrose gradient sedimentation studies showed that DNP treatment interfered with the rejoining of DNA single-strand breaks induced by the excision repair process. This interference was apparently specific for the exr gene-dependent branch of the uvr gene-dependent excision repair process, since the uvr and exr strains were not sensitized, while the wild-type and polA strains were sensitized (A54).

(4) By comparing the extent of repair of incision breaks as a function of dose of UV radiation in wild-type, recA56, recB21, exrA(lexA), and polA1 cells incubated in buffer or in growth medium with or without chloramphenicol, we deduced that excision repair can be divided into two branches. One branch, can proceed in buffer, and is mainly dependent upon polA; the other branch, a minor branch, requires complete growth medium and functional recA, recB , and exrA(lexA) genes (A55). Analogies are drawn between our results for excision repair and those for the two pathways for the repair of X-ray-induced DNA single-strand breaks (see section on DNA Repair: Ionizing Radiation).

(5) Further support for the existence of two pathways of excision repair came from our observation that certain brands of agar used in the plates for growing bacteria selectively inhibit the pathway of excision repair controlled by the recA, recB, and exrA(lexA) genes, but does not affect the pathway controlled by polA. Most lots of Noble agar (Difco) showed no inhibition (A57).

(6) Studies on the formation and repair of excision gaps following UV irradiation have demonstrated the presence of two repair pathways. The major pathway is a growth-medium independent, recA independent, polA dependent, short-patch repair process, and the minor pathway is growth-medium dependent, recA dependent, polA independent, and produces long repair patches. Although a well established model exists for the major pathway of excision repair, a model was not even proposed for the second pathway until 1987.

Because the second pathway is under the same genetic and physiological control as postreplication repair, we wondered if the two processes might be similar. To this end we predicted that the recA-independent pathway functions in the part of the chromosome where no sister duplexes exist (i.e., recombination cannot occur), but that the recA-dependent process can only occur in the portions of the chromosome that were replicated prior to UV irradiation (i.e., where sister duplexes exist).

We used a dnaA(Ts) mutation to align the chromosomes (i.e., all rounds of DNA replication were completed, and new rounds could not be initiated), and studied the genetic control of the repair of excision gaps in cells with unreplicated chromosomes, and in cells with partially replicated chromosomes.

In support of our theory, the excision repair that occurred in cells with unreplicated chromosomes was recA independent, but the repair that occurred in cells with partially replicated chromosomes was partially recA dependent. We found no evidence of interchromosomal recombination in recA-dependent excision repair.

The majority of the recA-dependent excision repair was recF dependent, and a small portion was recB dependent. The recF and recB genes are suggested to function in excision repair in a manner similar to their function in postreplication repair, i.e., in the replicated portion of the chromosome, the RecF pathway repairs excision gaps, and the RecB pathway repairs the DNA double-strand breaks that arise at unrepaired gaps (A121).

(7) In 1975, we described the enzymatic production of DNA double-strand breaks after UV irradiation. These breaks appeared in wild-type, polA, recA, recB, and exrA cells after incubation in growth medium, but not in uvr cells. These results raise the possibility that a significant fraction of the lethal events in UV irradiated wild-type cells may be enzymatically induced DNA double-strand breaks (A59), which are not well repaired when cells are grown in minimal medium (see below).

(8) UV radiation survival curves for E. coli wild-type and polA strains have large shoulders. The regions of these survival curves at which killing approaches exponential correspond to the fluences at which DNA double-strand breaks (assumed to be lethal events for cells grown in minimal medium) accumulate linearly. Reducing the number of UV photoproducts either by photoreactivation or by fluence fractionation resulted in an increase in survival, and a decrease in the yield of DNA double-strand breaks in both strains. These results support the hypothesis that enzymatically-induced DNA double-strand breaks may be the lesion ultimately responsible for the UV radiation-induced killing of wild-type E. coli (A63). [NOTE: Cells grown in minimal medium are very deficient in their ability to repair X-ray-induced DNA double-strand breaks, but those grown in rich growth medium are very efficient in this type of repair (A95, A97, A98).]

(9) DNA single-strand breaks were produced in uvrA and uvrB strains of E. coli K-12 after UV (254 nm) irradiation. These breaks appear to be produced both directly by photochemical events, and by a temperature-dependent process. Cyclobutane-type pyrimidine dimers are probably not the photoproducts that lead to the temperature-dependent breaks, since photoreactivation had no detectable effect on the final yield of breaks. The DNA strand breaks appear to be repairable by a process that requires DNA polymerase I and DNA ligase, but not the recA, recB, recF, lexA101, or uvrD gene products. We hypothesize that these temperature-dependent breaks occur either as a result of the breakdown of a thermolabile photoproduct, or as the initial endonucleolytic event of a uvrA uvrB-independent excision repair process that acts on a UV photoproduct other than the cyclobutane-type pyrimidine dimer (A72). [NOTE: We now know that there is an additional type of excision repair called "base-excision" repair to distinguish it from the then known "nucleotide excision" repair.]

(10) Using a mutant deficient in DNA ligase, we showed that ligase is required for both excision repair (and postreplication repair) (A75). [NOTE: These results were predicted by current theory, but real experiments to prove the theory had not been performed.]

(11) The uvrD3 (DNA helicase II) strain is deficient in host-cell reactivation of UV irradiated bacteriophage, suggesting that it is deficient in some aspect of excision repair, probably in the excision and/or resynthesis steps, since it is proficient in the initial incision steps. We tested this by measuring repair replication after UV irradiation in wild-type, polA, uvrD, and polA uvrD strains. A large stimulation of repair replication was observed in the uvrD3 strain, compared to the wild-type and polA1 strains, but this was reduced in the polA1 uvrD3 strain. The enhanced repair replication observed for uvrD3 strains appears to be due to the enhanced degradation observed for these strains (A92).

(1) In 1970, we demonstrated that postreplication repair (i.e., the repair of DNA daughter-strand gaps) is completely dependent upon the recA gene, but not on the recB and recC genes. By comparing the results for wild-type and excision-deficient cell lines, it was clear that excision repair can eliminate most of the lesions that can be repaired by postreplication repair. Furthermore, photoreactivation also eliminated most of the DNA lesions (i.e., cyclobutane-type pyrimidine dimers) that are repaired by postreplication repair (A35).

(2) We determined the time course for the recovery of viability and for postreplication repair in UV-irradiated excision-repair deficient cells of E. coli. We measured (i) the time course for minimal medium recovery by measuring the kinetics for the disappearance of sensitivity to rich growth medium, (ii) the disappearance of cyclobutane-type pyrimidine dimers by the loss of photoreactivability, and (iii) the time taken for cells to be free of lesions that produce DNA daughter-strand gaps. After a UV-radiation dose of 6.3 J/m2 (about 65% survival) at 254 nm, it takes 5-6 hours for the above processes to occur in an excision-repair deficient strain of E. coli. These processes do not occur in the absence of macromolecular synthesis (A42).

(3) An exrA(lexA) mutation produces a partial defect in postreplication repair (A52).

(4) In 1976, we examined the effect of the recA, recB, uvrD, and exrA(lexA) mutations and of postirradiation treatment with chloramphenicol (CAP) on survival and postreplication repair after the UV irradiation of uvrB strains of E. coli. The survival data suggested that there are five branches of postreplication repair; three of these branches are blocked by an exrA(lexA), recB, or uvrD mutation, a fourth branch is blocked by any one of these mutations, and is also sensitive to CAP. At least one other branch is not sensitive to any of these mutations or to CAP. The extent of postreplication repair observed with each of these strains is in general agreement with the pathways postulated on the basis of the survival data, although there are several apparent exceptions to this correlation (A65). [NOTE: Our evaluation of synergism used the method of Brendel and Haynes (1973), which we showed later (A91) to yield incorrect conclusions. However, the concept of multiple pathways of postreplication repair is still valid (see below).]

(5) In 1981, we reinvestigated the concept of multiple pathways of postreplication repair by looking at the interaction of recB, uvrD, lexA, and recF mutations in the delta-uvrB background for survival after UV irradiation and for recombination ability. We developed a better test for synergism than the method of Brendel and Haynes (1973), which we had used before (A65). From our new data and methodology for evaluating synergism and additivity we concluded that there are (at least) two major pathways of postreplication repair, one is dependent on the recF gene, and the other is dependent on the recB, uvrD, and lexA genes. We also observed that there was no correlation between the degree of radiation sensitivity and the degree of deficiency in genetic recombination. This suggests that the repair of DNA daughter-strand gaps may not require all of the functions required for genetic recombination (A91).

(6) Using a mutant deficient in DNA ligase, we showed that ligase is required for postreplication repair (and excision repair) (A75). [NOTE: These results were predicted by current theory, but real experiments to prove this theory had not been performed.]

(7) It had been reported that a deficiency in DNA polymerase I (polA) sensitized a uvrA strain of E. coli to UV irradiation. This suggested to us that DNA polymerase I plays a role in postreplication repair. We tested this hypothesis by the direct measurement of the repair of DNA daughter-strand gaps. We confirmed that the polA gene plays a small role in postreplication repair, but does not function in the chloramphenicol inhibitable (mutagenic) pathway of postreplication repair (A80).

(8) We reinvestigated the role of DNA polymerase I in postreplication repair when a deletion mutation (Δ) for DNA polymerase I became available. In marked contrast with the polA1 mutation, which only produced a small sensitization of a uvrA strain, the (ΔpolA mutation produced a large sensitization, about the same magnitude as a recF mutation. The (ΔpolA mutation acted synergistically with the recF mutation, suggesting that the polA gene functions in pathways of postreplication repair that are largely independent of the recF gene. When compared to a uvrA strain, the uvrA (ΔpolA strain was deficient in the repair of DNA daughter-strand gaps, but was not as deficient as a uvrA recF strain. A uvrA recF (ΔpolA strain was also deficient in the repair of DNA double-strand breaks.

The UV-radiation sensitivity of a uvrA polA546 (Ts) strain (deficient in the 5' -> 3' exonuclease function of DNA polymerase I), determined at the restrictive temperature, was very close to that for the uvrA (ΔpolA strain, while the polA1 mutation (deficient in the polymerizing function but not the 5' -> 3' exonuclease function) had only a small effect on the survival of a uvrA strain. This suggests that it is the 5' -> 3' exonuclease function of DNA polymerase I that is important in the recF-independent repair of daughter-strand gaps, and also in the recB-dependent repair of double-strand breaks (A124).

(9) The DNA single-strand binding protein is coded by the ssb gene, and is essential for DNA replication. We examined the effect of ssb mutations on DNA synthesis and the formation of DNA double-strand breaks in UV irradiated uvrB cells. Our results suggest that the major role of the Ssb protein in the repair of UV radiation-damaged DNA is to protect single-stranded parental DNA opposite daughter-strand gaps from nuclease attack (A93).

(10) Postreplication repair in uvrA and uvrB strains of E. coli is partially inhibited by the presence of rich growth medium after UV irradiation. This appears to be the molecular basis of minimal medium recovery (see section on Recovery Phenomena). The rich growth medium can be just minimal medium plus Casamino Acids or the 13 pure amino acids therein. It does not matter what type of medium the cells were grown in prior to UV irradiation (A94). The postreplication repair process that is inhibited by rich growth medium appears to be error free (A78).

(11) Based upon survival analysis we proposed the presence of two independent pathways of postreplication repair, one controlled by recF and one by recB. We undertook to determine the molecular basis of these two pathways. We confirmed that recF is required for the repair of DNA daughter-strand gaps, but about 50% of the gaps are repaired even in a uvrB recF recB strain. This suggests that there is a pathway for the repair of DNA daughter-strand gaps that is independent of recF, but is dependent upon recA. The recB gene is not required for the repair of DNA daughter-strand gaps, rather it is required for the repair of DNA double-strand breaks that arise at daughter-strand gaps. The recF and recB mutations have about the same effect on survival after UV irradiation, suggesting that these two different pathways of postreplication repair are of about equal importance to the survival of UV irradiated uvrB cells (A101).

(12) The processes for repairing DNA daughter-strand gaps were studied in UV irradiated uvrB, uvrB recB, uvrB recF, and uvrB recB recF cells of E. coli K-12. The thymine dimer-containing parental DNA was found to be joined to daughter strands during postreplication repair in all four strains examined. Therefore, both the recF-dependent and the recF recB-independent gap-filling processes repair DNA daughter-strand gaps by transferring DNA from the parental strands into the daughter strands (A103).

(13) Since postreplication repair is mutagenic, and since the umuC gene controls all of UV radiation mutagenesis, it is important to know in which pathway of postreplication repair that the umuC gene functions. While a umuC mutation increased the UV radiation sensitivities of uvrB, uvrB recF, uvrB recB, and uvrB recF recB cells, and increased the deficiencies in the repair of DNA daughter-strand gaps in these strains, it did not affect the repair of DNA double-strand breaks that arose in these cells from unrepaired DNA daughter-strand gaps. We suggest that umuC functions in a minor system for the error-prone (mutagenic) repair of DNA daughter-strand gaps, possibly the repair of overlapping DNA daughter-strand gaps (A106).

(14) Although we had suggested that there was a third pathway of postreplication repair that was independent of recF143 and recB21, this conclusion could only be valid if these mutations were not leaky. Therefore, we reinvestigated this problem using insertion and deletion mutants (D). Our results with the uvrB5 recF322::Tn3 DrecBC were similar to those we found using point mutants. Therefore, the recF143 and recB21 mutations are not leaky. Furthermore, treatment of the UV irradiated triple mutant with rifampicin had no effect on survival or the repair of DNA daughter-strand gaps. Therefore, a third pathway of postreplication repair has been demonstrated that does not require the recF recBC genes, and is constitutive. Minimal medium recovery occurs in the uvrB recB recC recF strain, and we have correlated this phenomenon with the inhibition by rich growth medium of the repair of DNA daughter-strand gaps (A107).

(15) UV-irradiated E. coli K12 uvrA(BC) cells show higher survival if plated on minimal growth medium (MM) rather than on rich growth medium (RM). This phenomenon has been referred to as minimal medium recovery (see section on Recovery Phenomena). UV-irradiated (4 J/m2) uvrA cells showed a similar rate of protein synthesis, whether incubated in MM or RM, however, they showed a severe depression in DNA synthesis when incubated in MM that lasted for about 30 min, and the normal rate of DNA synthesis was not reestablished until about 60 min after irradiation. When a sample of these same cells was switched to RM immediately after UV-irradiation, there was only a slight slowing of DNA synthesis, and the normal rate of synthesis was reestablished by 60 min. An additional mmrA mutation, or growth retardation by the addition of valine, blocked both this extra DNA synthesis in RM, and the inhibitory effect of RM on survival. These findings suggest that the absence of a marked delay in DNA synthesis observed in RM may be responsible for the inhibitory effect of RM on the survival of UV-irradiated excision-deficient cells. Two hypotheses, which are not mutually exclusive, were proposed and supported by data to explain why a fast rate of DNA synthesis after UV-irradiation partially inhibits postreplication repair and enhances cell lethality (A108).

(16) The mechanism by which an sbcB mutation suppresses the deficiency in postreplication repair shown by recB recC mutants of E. coli was studied. The presence of an sbcB mutation in uvrA recB recC cells increased their resistance to UV radiation. This enhanced resistance was not due to a suppression of the minor deficiency in the repair of DNA daughter-strand gaps or to an inhibition of the production of DNA double-strand breaks in UV-irradiated uvrA recB recC cells; rather, the presence of an sbcB mutation enabled uvrA recB recC cells to carry out the repair of DNA double-strand breaks. In the uvrA recB recC sbcB background, a mutation in recF produced a huge sensitization to UV radiation, and it rendered cells deficient in the repair of both DNA daughter-strand gaps and DNA double-strand breaks. Thus, an additional sbcB mutation in uvrA recB recC cells restored their ability to repair DNA double-strand breaks, but the further addition of a recF mutation blocked this repair capacity (A110).

(17) The number of DNA double-strand breaks formed in UV-irradiated uvrB recF recB cells correlates with the number of unrepaired DNA daughter-strand gaps, and is dependent on DNA synthesis after UV-irradiation. These results are consistent with the model that the DNA double-strand breaks that are produced in UV-irradiated excision-deficient cells occur as the result of breaks in the parental DNA opposite unrepaired DNA daughter-strand gaps. By employing a temperature-sensitive recA200 mutation, we devised an improved assay for studying the formation and repair of these DNA double-strand breaks. Possible mechanisms for the postreplication repair of DNA double-strand breaks were discussed (A111).

(18) We isolated a radiation sensitive mutant of E. coli, radB101, which is similar to the recB21 mutation in that it sensitizes cells to gamma-radiation more than to UV radiation. All other mutations tested sensitize cells much more to UV radiation (A98). The radB101 mutation maps in the recN locus, so we have suggested that the radB101 mutation now be referred to as recN2001 (A125).

Since the recN2001(radB) mutation sensitizes a uvrB strain to UV irradiation, this suggests strongly that it must play a role in postreplication repair. We examined this possibility, and demonstrated that the recN(radB) gene plays little or no role in the repair of DNA daughter-strand gaps, but is involved in the recB-dependent postreplication repair of DNA double-strand breaks (A117).

(19) The recB-dependent repair of DNA double-strand breaks in UV irradiated E. coli uvrB recF cells is inhibited by rich growth medium. Furthermore, the DNA double-strand breaks that were formed in UV irradiated uvrB recA200 (Ts) cells that were incubated at 42oC in rich growth medium were not repaired during subsequent repair incubation at 30oC in minimal medium or in rich growth medium, while the breaks formed at 42oC in minimal medium could be repaired at 30oC regardless of the medium used. The same number of breaks was formed in each case. The absence of an abrupt cessation in DNA synthesis following UV irradiation, when cells are held in rich growth medium (A108), may explain why these double-strand breaks are not repairable (A118).

(20) The recF gene plays a major role in the repair of DNA daughter-strand gaps after UV irradiation, but the function of the recF gene product is not known. One approach to understanding the function of the recF gene product is to isolate mutants that suppress the RecF phenotype, and to characterize the mechanism by which these suppressor mutations function. To this end we have isolated a suppressor of recF(srf). The srf mutation maps in the recA gene, but does not appear to affect the synthesis of UV radiation-induced proteins, nor does it result in an altered RecA protein, as determined by two-dimensional gel electrophoresis. Rather, the recA(srf) mutations allow the RecA protein to participate in the recF-dependent postreplication repair process without the need for the RecF protein (A120). [NOTE: Volkert and Hartke (1984) reached a similar conclusion with their srf mutants.]

(21) Two mutations known to affect genetic recombination in a recBC sbcBC strain, recJ and recN, were examined for their effects on postreplication repair. The recN mutation produced a partial deficiency in the repair of double-strand breaks, while the recJ mutation produced a deficiency in the repair of daughter-strand gaps in uvrB recB cells (but not in uvrB cells), and a deficiency in the repair of both daughter-strand gaps and double-strand breaks in uvrA recBC sbcBC cells. Thus, the recJ gene is involved in recF-dependent repair processes, while the recN gene is involved largely in recBC-dependent repair processes (A126).

(22) A number of reviews and symposia talks were published on DNA repair (B10, B11, B12, B18, B19, B20, B21, B26, B29, B31, B32, B33, B35, B40, B41, B45, B46, B47, B48, B49, B51).

[ TOP ]


(1) DNA ligase, which rejoins polynucleotides having a 3'-OH terminus on one polynucleotide and a 5'-PO4 terminus on the other, had just become available through Drs. Olivera and Lehman at Stanford (1967). My first graduate student, Daniel S. Kapp, X-irradiated (19.6 krads) 14C-DNA that had been isolated from E. coli, and then subjected it to treatment with ligase. While there was no joining of the 14C-DNA by the action of ligase, 32P-poly dT in the same reaction mixture was joined together. Therefore, although DNA ligase may well be required for the final step in the repair of DNA chain breaks, it appears that both nuclease action (to clean the termini) and polymerase action are first required before ligase can act on DNA that has been X irradiated in vitro in air (A28).

(2) We next investigated the locations and end products of chain breakage produced by the X irradiation of aqueous solutions of calf thymus DNA. We determined the quantitative relationships between chain breakage, inorganic phosphate yields, phosphomonoester group formation, and deoxyribose damage. The effects of histidine (10-3 M) on the above parameters of X-ray damage were also determined.

For each single-chain break produced in DNA by X irradiation in 0.15 M NaCl, 0.60 molecules of malonic aldehyde-like product were formed (i.e., damage to deoxyribose), 0.33 molecules of inorganic phosphate were liberated, and 1.35 molecules of phosphomonoester groups were produced, of which 29% were 5'-phosphoryl groups. The ratio of newly formed 5'-phosphoryl termini to 5'-hydroxyl termini was 32:1. The presence of histidine (a radical scavenger) reduced the number of single-chain breaks by a factor of 0.39, the yield of inorganic phosphate by 0.24, the formation of phosphomonoester groups by 0.42, and deoxyribose damage by 0.41 (A33).

(3) Strains of E. coli K-12 deficient in excision repair (uvr) and genetic recombination (rec) were studied with reference to their radiosensitivity and their ability to repair X-ray-induced DNA single-strand breaks. Mutations in the recA, recB, and recC genes appreciably increased the radiosensitivity of E. coli, whereas a uvrB5 mutation produced little if any effect on radiosensitivity. For a given dose of X-rays, the yield of single-strand breaks was largely independent of the presence of a uvr or rec mutation. The rec+ cells (± uvrB5) could efficiently rejoin the strand breaks, but the recA56 cells could not. The recB21 and recC22 cells showed some indication of repair capacity. We conclude that there is a correlation between the inability to repair DNA single-strand breaks and the radiosensitivity of the rec mutants of E. coli. This suggests that unrepaired DNA single-strand breaks may be lethal in E. coli (A36).

(4) The above studies correlated changes in the repair capacity of cells differing from one another by one or a few mutations that change their X-ray sensitivity. It seemed desirable also to study the basis of variations in radiosensitivity within a single strain. The X-ray sensitivity of E. coli B/r can be varied widely by different preirradiation culture conditions. We examined both the yield and repair of X-ray-induced DNA single-strand breaks and postirradiation DNA degradation in E. coli B/r under different growth conditions.

Late log phase cells were 1.6 times more sensitive to killing by X-rays than were stationary-phase cells, when grown in Brain Heart Infusion plus glucose. The number of strand breaks/krad was the same for the two growth conditions, but the stationary-phase cells showed a somewhat greater ability to repair these breaks. The rapidity and extent of DNA degradation was greater in log-phase cells. The enhanced survival in stationary-phase cells correlates both with the enhanced repair capacity and the reduced degradation of DNA.

Cells grown to stationary phase in peptone medium (PO) were 3.4 times more sensitive to killing by X-rays than when the peptone medium was supplemented with glucose and phosphate buffer (PG). The yield of DNA single-strand breaks was the same under these two conditions, but was about 2-fold higher than for cells grown in Brain Heart Infusion plus glucose. The kinetics for the repair of these breaks was about the same for both types of cells for the first 30 min. After that time further repair ceased in the PO cells, but continued in the PG cells. Postirradiation DNA degradation was more rapid and more extensive in the PO cells. For both the log-stationary effect and the glucose effect, the more resistant cells showed more strand rejoining and less DNA degradation than did the sensitive cells (A38).

(5) DNA polymerase I is required for the rapid repair of X-ray-induced DNA single-strand breaks in E. coli. A much higher yield of DNA single-strand breaks/krad was observed in a polA1 strain than was reported before for wild-type E. coli. This increased yield of breaks is due to the absence of a rapid repair system, which had not been described previously in E. coli. This absence of repair probably accounts for the X-ray sensitivity of the polA1 mutant. The rapid repair of strand breaks could be reversibly inhibited in pol+ cells by temperature, pH, and EDTA (A40).

(6) Evidence for an ultrafast process for the repair of DNA single-strand breaks that is independent of DNA polymerase I was reported. For log phase cells of E. coli polA1 X-irradiated in phosphate buffered saline, pH 7.3, the rate of production of breaks per single-strand genome per krad was 2.13 in air, and 0.66 in its absence (an OER for breaks of 3.2). To determine if this oxygen effect was due to a difference in absolute yield of breaks or to a differential ability for repair, the yield of single-strand breaks was determined in cells that had been inactivated either by heat treatment (52oC) or cold shock (0oC). In both cases there was a large increase in the yield of radiation induced anoxic breaks (2.8 fold), and a smaller increase (1.25 fold) in aerobic breaks. Sodium cyanide (1 mM) at 0oC had no effect on the anoxic yield of breaks. However, in the presence of quinacrine (0.2 mM) the level of anoxic breaks increased above that seen in untreated polA cells, while the aerobic level was the same. N-Ethylmaleimide (NEM) (0.5 mM) had a similar effect, but another SH-enzyme inhibitor, iodoacetic acid (1 mM), had no effect. Hydroxyurea (10 mM) had an effect qualitatively similar to NEM. These results suggest that E. coli possess an ultrafast repair system, which operates mainly on anoxic breaks, and can be inhibited by physical or chemical pretreatment of the cells. Such inhibition demonstrates that the initial yield of X-ray-induced DNA single-strand breaks in vivo is largely independent of the presence of oxygen, but that the yield of breaks in anoxic cells is very rapidly modified by repair (A44).

(7) Starvation for niacin reduces the pools of nicotinamide adenine dinucleotide (NAD) in niacin-requiring cells, and causes an inhibition of DNA ligase, for which NAD is a cofactor. Since it was thought that DNA ligase is required in the final rejoining step in the repair of X-ray-induced DNA single-strand breaks, we examined the yield of such breaks in cells that had been starved for niacin. Such starvation increased the radiation sensitivity of cells, and caused a 1.9-fold increase in the number of X-ray-induced DNA single-strand breaks observed immediately after irradiation in repair proficient cells after anoxic irradiation, but a much smaller effect (1.2-fold) under aerobic conditions, when compared with polA1 cells irradiated in air and nitrogen. No rejoining of breaks was seen when niacin or NAD was added back to anoxically irradiated cells. These results suggest that DNA ligase is involved in the ultrafast repair system, which mainly rejoins DNA single-strand breaks produced under anoxia (A45).

(8) For cells grown to logarithmic phase in minimal medium, we measured the repair of DNA single-strand breaks in E. coli K-12 (W3110) X-irradiated in the presence or absence of oxygen, and determined the influence of repair on survival. After the completion of a very fast repair process in buffer (Type I), there were 3 times as many breaks remaining after irradiation in the presence of oxygen as there were after irradiation in the absence of oxygen. These breaks were then rapidly repaired in buffer by the Type II system, which requires DNA polymerase I. After low doses of X-rays, the Type II system can repair all of the breaks, but beyond a threshold dose it repairs a constant ~90% of the breaks presented to it (aerobic or anerobic breaks). After Type II repair is complete, the remaining breaks can be repaired by the growth-medium-dependent Type III system, which requires functional recA and recB genes. This system repairs only about two additional breaks, whether produced aerobically or anerobically. [NOTE: Cells grown in minimal medium are very deficient in their ability to repair X-ray-induced DNA double-strand breaks, but those grown in rich growth medium are very efficient in this type of repair (A95, A97, A98).]

While the recB strain showed only a slight reduction in oxygen enhancement ratio (OER), the recA strain had an OER of about 2.0, compared to 3.4 for the rec+ parental strain. This suggests that the recA gene must participate in the repair of lesions other than strand breaks, whose formation is affected by the presence of oxygen (A46).

(9) A polA1 exrA(lexA) strain of E. coli was constructed, and found to be more sensitive to aerobic or anoxic X-irradiation than were either of the single mutants. The polA strain was deficient in Type II repair, but not in Type III repair. The exrA(lexA) strain was not deficient in Type II repair, but was deficient in Type III repair. The double mutant was deficient in both types of repair. Thus, the increased X-ray sensitivity of the double-mutant correlates with its increased deficiency in the repair of DNA strand breaks. The OER values for the different strains are reported (A47).

(10) Strains deficient in DNA polymerase II (polB) showed no change in X-ray sensitivity under air or nitrogen when compared to their wild-type or polA1 parent strains (A49).

(11) E. coli wild-type cells suspended in growth medium during X-irradiation had a higher survival than cells irradiated in buffer, whether irradiated in air or nitrogen. This does not appear to be a radiation chemistry problem, since mutations in recA, recB, or exrA(lexA) abolished this higher survival response. Since there did not appear to be any difference in the final amount of repair of DNA single-strand breaks after irradiation in buffer or in growth medium, the increased X-ray sensitivity resulting from irradiation in buffer must be due to the inhibition of some rec and exr(lexA) gene-dependent repair process other than for the repair of DNA single-strand breaks (A53).

(12) We undertook a very careful study of the yield and repair of DNA single-strand breaks in wild-type and polA1 cells, and compared these results with earlier work. Particular attention was given to overcoming the two major technical problems of sucrose gradient sedimentation, i.e., speed dependence and the nonrandomness of the molecular weights of DNA samples. To this end we reported that some of our earlier data are in error by factors of 3.4 - 7.6. Technical and interpretive problems using alkaline sucrose gradients are discussed. From the present work, the initial yield of DNA single-strand breaks was found to be 32.4/genome krad-1 (9.0 eV/break) (A67).

(13) Cell survival, DNA degradation and the repair of DNA single-strand breaks were measured for E. coli pol+, polA1, polC1026 (Ts), and polA1 polC1026 (Ts) cells after 137Cs gamma irradiation. The results indicate that DNA polymerase III (polC ) is required for the growth medium dependent repair (Type III) in polA+ or polA cells, and is necessary for growth medium independent repair (Type II) in polA cells deficient in DNA polymerase I. The relative deficiencies of each of these strains generally correlate with their relative sensitivities to cell killing, and with the extent of DNA degradation observed (A69).

(14) A uvrD mutation (DNA helicase II) increased the X-ray sensitivities of E. coli wild-type and polA strains, but had no effect on the sensitivities of recA and recB strains, and little effect on a lexA strain. Incubation of irradiated cells in medium containing 2,4-dinitrophenol or chloramphenicol decreased the survival of wild-type and uvrD cells, but had no effect on recA, recB, and lexA strains. The uvrD strain is deficient in the Type III repair of DNA single-strand breaks. The uvrD mutation inhibits certain rec+ lex+ dependent repair processes, e.g., Type III repair, but does not inhibit other rec+ lex+ dependent processes that are sensitive to 2,4-dinitrophenol and chloramphenicol (A76).

(15) A major factor in the rapid accumulation of knowledge on the repair of radiation damaged DNA has been the availability of DNA repair-deficient mutants of E. coli. In order to further our studies on the repair of X-ray-induced DNA damage, we initiated a program of isolating new X-ray sensitive mutants of E. coli. Our first mutant was called radA100, and it maps at 99.6 min between serB and deoC. The radA100 mutation sensitized stationary phase cells to X-rays if they had been grown in glucose-supplemented rich medium, but not if they had been grown in nonsupplemented rich medium (indicating a defect in the phenomenon called glucose-induced resistance). Similarly, log phase cells were sensitized to X-rays, UV radiation and methyl methanesulfonate if they were grown in rich medium, but not if grown in minimal medium (indicating a defect in medium-dependent resistance). Relative to the wild-type strain, the radA strain was deficient in the repair of X-ray-induced DNA single-strand breaks when grown to logarithmic phase in rich medium, but not when grown in minimal medium. This is a novel mutation in that it does not sensitize log phase cells grown in minimal medium to either X- or UV-radiation (A95).

A second X-ray sensitive mutant, radB101, was isolated that maps at 56.5 min on the E. coli linkage map. The radB mutation sensitized wild-type cells to gamma and UV radiation, and to methyl methanesulfonate. The radB mutant was normal for gamma- and UV-radiation mutagenesis, it showed only a slight enhancement of gamma- and UV-radiation-induced DNA degradation, and it was ~60% deficient in recombination ability. The radB gene is suggested to play a role in the recA gene-dependent (Type III) repair of DNA single-strand breaks after gamma irradiation, and in postreplication repair after UV irradiation for the following reasons: the radB strain was normal for the host-cell reactivation of gamma- and UV-irradiated bacteriophage lambda; the radB mutation did not sensitize a recA strain, but did sensitize a polA strain to gamma and UV radiation; the radB mutation sensitized a uvrB strain to UV radiation (A98).

A third radiation-sensitive mutant, radC102, was isolated. The radC gene is located at 81.0 min on the E. coli K-12 linkage map. The radC mutation sensitized cells to UV radiation, but unlike most DNA repair mutations, sensitization to X rays was observed only for rich medium-grown cells. For cells grown in rich medium, the radC mutant was normal for gamma-radiation mutagenesis, but showed less UV-radiation mutagenesis than the wild-type strain; it showed normal amounts of X- and UV-radiation-induced DNA degradation, and it was ~60% deficient in recombination ability. The radC strain was normal for host cell reactivation of gamma- and UV-irradiated bacteriophage lambda; the radC mutation did not sensitize a recA strain, but did sensitize radA and polA strains to X and UV radiation, and a uvrA strain to UV radiation. Therefore, we suggest that the radC gene product plays a role in the growth medium-dependent, recA gene-dependent repair of DNA single-strand breaks after X irradiation, and in postreplication repair after UV irradiation (A102).

E. coli K-12 cells incubated in buffer can repair most of their X-ray-induced DNA single-strand breaks, but additional single-strand breaks are repaired when the cells are incubated in growth medium. While the radC mutant was proficient at repairing DNA single-strand breaks in buffer (polA-dependent repair), it was partially deficient in repairing the additional single-strand breaks (or alkali-labile lesions) that the wild-type strain can repair in growth medium (recA-dependent repair), and this repair deficiency correlated with the X-ray survival deficiency of the radC strain. In studies using neutral sucrose gradients, the radC strain consistently showed a small deficiency in rejoining X-ray-induced DNA double-strand breaks, and it was deficient in restoring the normal sedimentation characteristics of the repaired DNA (A115).

(16) A lengthy review article was published on the repair of X-ray-induced DNA single-strand breaks (B13).

(1) Using an improved technique for measuring DNA double-strand breaks, we showed that these breaks are produced linearly as a function of dose under air or nitrogen, with an OER of 2.42. The OER for survival was 2.79. From these data we calculate that 1.3-1.4 double-strand breaks per genome were produced per lethal event. [NOTE: The cells were grown to log phase in minimal medium. Under these conditions the cells are very deficient in the repair of DNA double-strand breaks (A95, A97, A98).] The energy to produce a DNA double-strand break under air was 532 eV, and 1290 eV under nitrogen (A60).

(2) The polA1 mutation (DNA polymerase I) increased the sensitivity of E. coli to killing by gamma irradiation in air by a factor of 2.9, and increased the yield of DNA double-strand breaks by a factor of 2.5. These additional breaks appear to be due to the action of nucleases in the polA1 strain rather than to the rejoining of strand breaks in the pol+ strain. For example, gamma irradiation at 3oC did not affect the yield of double-strand breaks in the pol+ strain, but decreased the yield in a polA strain by a factor of 2.2. We suggest that the increased yield of DNA double-strand breaks in the polA1 strain may be the result of the unsuccessful excision repair of ionizing radiation-induced DNA base damage on opposite strands of the DNA (A62).

(3) We determined the relative biological effectiveness (RBE) of 50 kVp X-rays and of 660 keV 137Cs gamma-rays with respect to the production of DNA damage, repair, and cell killing in E. coli. The 50 kVp X-rays were 1.93 times more effective in producing DNA double-strand breaks, but there was no significant difference in the two qualities of radiation for the production of DNA single-strand breaks. There were 1.57 times more unrepaired single-strand breaks with X-rays versus gamma-rays. This increase in unrepaired single-strand breaks may be due to the increased yield of DNA double-strand breaks, since these two types of breaks are indistinguishable on alkaline sucrose gradients. These results suggest that the greater RBE of 50 kVp X-rays may be related to an increased efficiency of producing DNA double-strand breaks compared with 137Cs gamma-rays (A64).

(4) E. coli C cells unifilarly labeled with 5-bromouracil (5-BU) in place of thymine were 2.55 times as sensitive as unsubstituted cells to killing by gamma-irradiation under air. The yield of DNA double-strand breaks was only increased by a factor of 1.55, suggesting that other lesions also contribute to cell killing. The presence of 5-BU in one strand exerts an effect on the complementary strand, since the repairability of the single-strand breaks in the non-5-BU-containing strand is diminished when 5-BU is incorporated into the complementary strand (A77).

(5) The X-ray resistance of logarithmic phase cells of E. coli K-12 is enhanced threefold by growth in rich medium versus minimal medium (A97). In this work, X-ray-induced DNA strand breaks were assayed by sedimentation in alkaline and neutral sucrose gradients to correlate the enhanced survival of rich-medium-grown cells with an enhanced capacity for DNA repair. While rich-medium-grown cells showed no enhanced capacity for repairing DNA single-strand breaks in buffer, i.e., fast, polA-dependent repair, they did show an enhanced capacity to repair both single-strand and double-strand breaks in growth medium, i.e., slow, recA-dependent repair. This enhanced capacity for DNA repair in rich-medium-grown cells was inhibited by a post-irradiation treatment with rifampicin, indicating the requirement for de novo RNA synthesis. Kinetic studies indicated that the repair of DNA double-strand breaks was a complex process. Relative to the sedimentation rate in neutral sucrose gradients of nonirradiated DNA, the sedimentation rate of X-irradiated DNA first changed from slow to very fast. Based on alkaline sucrose gradient sedimentation studies, all the strand breaks had been repaired during the formation of the very fast sedimenting DNA. With continued incubation, the sedimentation rate of the DNA on neutral sucrose gradients decreased to the normal rate (A109).

(6) The repair of X-ray-induced DNA single-strand breaks was studied after the completion of growth-medium-independent repair in E. coli K-12. A comparison of the sedimentation of DNA from the bacteriophages T2 and T7 was used to test the accuracy of our alkaline and neutral sucrose gradient procedures for determining the molecular weight of bacterial DNA. The repair of DNA single-strand breaks by cells incubated in buffer occurred by two processes. About 85% of the repairable breaks were resealed rapidly (t1/2 = < 6 min), while the remainder were resealed slowly (t1/2 = ~20 min). After the completion of the repair of DNA single-strand breaks in buffer, about 80% of the single-strand breaks that remained were found to be associated with DNA double-strand breaks. The subsequent resuspension of cells in growth medium allowed the repair of both DNA single- and double-strand breaks in wild-type, but not in recA cells. Thus, the recA-dependent, growth-medium-dependent repair of DNA single-strand breaks is essentially the repair of DNA double-strand breaks (A113).

(7) Isogenic E. coli strains carrying single DNA-repair mutations were compared for their capacity for (i) the repair of X-ray-induced DNA double-strand breaks as measured using neutral sucrose gradients, (ii) medium-dependent resistance, i.e., a recA-dependent X-ray survival phenomenon that correlates closely with the capacity for repairing DNA double-strand breaks; and (iii) the growth medium-dependent, recA-dependent repair of X-ray-induced DNA single-strand breaks, as measured using alkaline sucrose gradients (about 80% of these single-strand breaks are actually parts of double-strand breaks). These three capacities were measured to quantitate more accurately the involvement of the various genes in the repair of DNA double-strand breaks over a wide dose range. The mutations tested were grouped into five classes according to their effect on the repair of X-ray-induced DNA double-strand breaks: (I) the recA, recB, recC, and lexA mutants were completely deficient; (II) the radB and recN mutants were about 90% deficient; (III) the recF and recJ mutants were about 70% deficient; (IV) the radA and uvrD mutants were about 30% deficient; and (V) the umuC mutant resembled the wild-type strain in its capacity for the repair of DNA double-strand breaks (A119).

(8) An in vitro model was developed for the repair of DNA double-strand breaks, i.e., the formation of heteroduplex DNA from linear duplex DNA using purified RecA, Ssb, and RecBCD proteins. The RecBCD enzyme was provided to unwind the duplex DNA, and the presence of Ssb protein prevented its reassociation. When the reaction conditions were then changed to favor reassociation, heteroduplex DNA was formed slowly in the absence of RecA protein, but its presence increased the kinetics of heteroduplex formation (A128).

[ TOP ]


[NOTE: Since LEDs (light emitting diodes) are now replacing lasers as the source for light for therapy (they are less expensive, and produce specific wavelengths of light), Low Level Light Therapy is the preferred definition for LLLT, instead of Low Level Laser Therapy.]

Throughout my career as a photobiologist I have attended a number of laser meetings. My participation at these meetings was mostly to teach MD's and engineers the First Law of Photochemistry, i.e., light must be absorbed before photochemistry can occur. It follows that the biological effects of light are the result of the photochemistry.

I also had to "try" to teach them that lasers are NOT magical; lasers are just expensive flashlights. It is the light that lasers emit that produces the biological effect, NOT the laser itself. The only way to get a biological effect just from a laser is to drop it on your foot. Therefore, I had to stress that they mention the wavelength of the light that their laser produces, not just the gas used in the production of the laser (e.g., He-Ne Laser).

It has been both an interesting and frustrating experience.

In 1990, I was invited along with Dr. Arthur L. Schawlow, co-inventor of the laser, as the two major speakers at the First International Congress of the International Laser Therapy Association, to be held in Okinawa, Japan, October 26 - 28, 1990, with subsequent lectures on the mainland of Japan.

There were so many quacks and uneducated people (i.e., they don't know the First Law of Photchemistry) in the Low Level Laser Therapy (LLLT) field that I was hesitant to attend this meeting. I decided to first review the literature to see if there were any acceptable scientific papers to support the claims that near-UV and visible light from non-thermal lasers could, e.g., enhance wound healing. Fortunately, I did find a number of papers that met my scientific criteria, and demonstrated a scientific basis for the enhancement of wound healing. I decided to accept the invitation.

Again, I tried to teach the audience at the Congress about photobiology, and a paper was published entitled "The Photobiological Basis of Low Level Laser Radiation Therapy." (B53).

After the publication of this paper in 1991, I began to receive requests to review papers in the field, so I kept an eye on the field. I was most distressed when I found a paper from Stanford on low level laser effects that would never have been published if the authors or the reviewers had known the First Law of Photochemistry. I was compelled to write a Letter to the Editor of the journal Pain. The title of my letter was "Ignorance of Photobiology: A Major Pitfall in Using Lasers in Medicine" (C9). What was even worse than having the original paper come from Stanford, was that the authors published reply to my letter at the same time my letter was published, indicating even more clearly that they did not know what they were talking about.

After reading and reviewing several papers for the journal Laser Therapy, I felt again compelled to write another paper (1999) explaining "The First Law of Photochemistry and Lasers" (C10).

I know that low levels of light (both non-thermal laser and non-laser light) at certain wavelengths can have beneficial biological effects, and in certain cases with a proven scientific basis. However, there still remain in the field such a high percentage of quacks, and well meaning people uneducated in the principles of photobiology, that the field of LLLT is shunned by the core scientific community. It is too bad that the general public continues to be deprived of a potentially good therapeutic modality because of bad science, or no science.

Then I found a paper by the Co-Editor of the journal "Photomedicine and Laser Surgery" that was really bad (2004). In fact, I had rejected this paper 4 years earlier for another journal. The author knows nothing about photobiology. In our conversations about his published paper, which I copied to the Chief Editor (Lanzafame), the author sent a section of a paper that he had published to prove to me that he was an expert on photobiology. The paper said that both X-radiation and UV-radiation produce ionizations. Obviously this statement is totally wrong. I spent my career working on X-radiation and UV-radiation. The Chief Editor asked me to submit a paper to his journal. Rather than just rant about his Co-Editor's terrible paper, I decided to write (2005) a "lecture" on the basics of photobiology as related to Low Level Light Therapy, "Laser (and LED) Therapy is Phototherapy" (C11).

See also:
(B52) Light and Life: The Photobiological Basis of the Therapeutic Use of Radiation From Lasers, K.C. Smith, in Progress in Laser Therapy (Selected papers from the October 1990 ILTA Congress), pp. 11-18 (T. Ohshiro and R.G. Calderhead, ed.), John Wiley & Sons, Chichester, England, (1991).

(B53) The Photobiological Basis of Low Level Laser Radiation Therapy. K.C. Smith, Laser Therapy 3, 19-24 (1991).

(B57) Laser and LED Photobiology, K.C. Smith, Laser Therapy 19: 72-78 (2010).

(B58) Molecular Targets for Low Level Light Therapy, K.C. Smith, Laser Therapy 19: 135-142 (2010).

[ TOP ]


Allen, Frank W. [A2, A3]

Aplin, Robin T. [A18]

Bagshaw, Malcolm A. [A22, B5]

Barfknecht, Thomas R. [A78, A80, A94]

Bockrath, Richard C. [A96]

Bonura, Thomas [A59, A60, A62, A63, A64, A66, A77]

Carlson, Kenneth M. [A92, B31]

Cordes, Eugene [A5]

Crestfield, Arthur M. [A3]

Diver, William P. [A95, A97]

Doggett, R. L. Scotte [A22, B5]

Duggan, Edward L. [A1]

Earle, John D. [A84]

Felzenszwalb, Israel [A102, A115, B41]

Fuks, Zvi [A41, A74]

Ganesan, Ann K. [A25, A30, A34, A42, A43, B7]

Grossweiner, Leonard I. [A89, B50]

Hahn, George M. [A84]

Hamelin, Claude [A69, A73]

Hays, James E. [A23]

Hodgkins, Brenda [A17]

Hoppe, Richard T. [A84]

Kaplan, Henry S. [A7, A8, A9, A11, A22, A38, A40, A44, A45, A46, A48, A56, A60, A62, B5, B13]

Kapp, Daniel S. [A28, A33, A36, A39, B15]

Low-Beer, Bertram V. A. [A4]

Martignoni, Klaus D. [A50, A71, B14]

Meun, Dieter H. C. [A24, A29, A35]

Moss, Stephen H. [A85, A88]

Nelsen, Thomas S. [A22, B5]

Obaseiki-Ebor, Emmanuel E. [A132]

O'Leary, Mary E. [A17, A21, A27]

Packer, Lester [D3]

Rebhun, Stella [A7]

Roots, Ruth [A58, A61, A68]

Sargentini, Neil J. [A82, A83, A90, A95, A96, A97, A98, A99, A102, A105, A109, A113, A115, A116, A117, A119, A122, A125, A127, A129, A131, A133, A134, B31, B41, B42, B43, B44, B45, B51]

Schweet, Richard S. [A5]

Seear, Joan [D3]

Sharma, Rakesh C. [A94, A99, A100, A107, A108, A118, A121, A123, A124, B45, B46, B47]

Tang, Moon-shong, [A86, A87]

Tomlin, Patricia A. [A8, A11]

Town, Christopher D. [A38, A40, A44, A45, A46, A60, B13]

van der Schueren, Emmanuel [A48, A53, A54, A55, A57, A76, B20, B21, B31]

Voiculetz, Nicolae [A56]

Wang, Tzu-chien V. [A81, A91, A93, A101, A103, A104, A106, A110, A111, A112, A114, A120, A126, A128, A130, B30, B45, B46, B47, B48, B49]

Wechter, William J. [A26]

Yoshikawa, Hiroshi [A19]

Youngs, David A. [A47, A49, A51, A52, A53, A55, A57, A64, A65, A67, A69, A72, A75, A76, A79, B20, B21, B31]

[ TOP ]