Photochemistry and Photobiology Vol. 52, No. 4, pp. 897-901, 1990 Printed in Great Britain. All rights reserved
0031-8655190 $03.00+0.00 Copyright 0 1990 Pergamon Press plc
RESEARCH NOTE GROWING Escherichia coli MUTANTS DEFICIENT IN RIBOFLAVIN BIOSYNTHESIS WITH NON-LIMITING RIBOFLAVIN RESULTS IN SENSITIZATION TO INACTIVATION BY BROAD-SPECTRUM NEARULTRAVIOLET LIGHT (320-400 nm) R. E. LLOYD,J. L. RINKENBERGER, B. A. HUG and R. W. TUVESON* Department of Microbiology, University of Illinois, 131 Burrill Hall, 407 S. Goodwin, Urbana, IL 61801, USA (Received 12 Junuury 1990; accepted 9 April 1990)
Abstract-Two mutants of Escherichia coli unable to synthesize riboflavin were grown with limiting (2 pg me-’) and non-limiting (10 pg me-]) concentrations of riboflavin. These riboflavin auxotrophs when grown to exponential phase with non-limiting riboflavin are more sensitive to broad spectrum near-ultraviolet light (NUV, 32C-400 nm) inactivation than when they are grown with limiting riboflavin. Exponential phase cells of the riboflavin auxotrophs grown with limiting riboflavin are sensitized when irradiated in saline supplemented with riboflavin. This suggests that extracellular riboflavin is important as a N U V sensitizer when intracellular levels of riboflavin are reduced. The concentration of riboflavin in crude extracts from exponentially growing cells correlates well with the sensitivity of these mutants to N U V inactivation. The level of riboflavin supplementation has little effect on the NUV sensitivity of the parental strain.
tory chain (e.g. porphyrins, flavins and quinones) have the NUV absorption characteristics that would allow them to act as endogenous photosensitizers (Jagger, 1983). The photosensitizing properties of endogenous porphyrins have previously been examined in E. coli using strains carrying a mutant allele of the HemA gene. These strains are incapable of synthesizing porphyrins, precursors of cytochromes, unless supplemented with A-aminolevulinic acid (A-ALA) (Tuveson and Sammartano, 1986; Peak et al., 1987). Mutants deficient in HemA and grown without AALA lack porphyrins and are resistant to NUV inactivation. Cells grown with A-ALA are NUVsensitive, suggesting that porphyrins act as endogenous chrornophores for the absorption of NUV wavelengths. Additional evidence that cytochromes can serve as endogenous NUV photosensitizers comes from the observation that E. coli strains carrying cloned genes of the cyd complex and overproducing specific cytochromes are sensitive to inactivation by NUV (Sammartano and Tuveson, 1987). Testing the photosensitizing properties of endogenous flavin compounds has been more difficult due to the absence of riboflavin auxotrophs in E . coli. In their analysis of the riboflavin biosynthetic pathway, Wilson and Pardee (1962) were unable to isolate riboflavin auxotrophs. They suggested that the E . coli strain used was incapable of taking up riboflavin from the medium and, consequently, auxotrophs could not be recovered. Recently, Bandrin et a / . (1983) have reported the isolation of riboflavin auxotrophs from a strain of
Many investigations of near-ultraviolet (320-400 nm, NUV)t radiation effects have been reviewed (Jagger, 1981, 1983; Eisenstark, 1987). It has become clear that mechanisms of NUV inactivation and mutagenesis are distinctly different from those of far-ultraviolet radiation (FUV). FUV mediated damage results from direct absorption of light energy by DNA (Webb and Lorenz, 1970; Webb, 1978; Webb and Brown, 1979; Peak et al., 1984). Damage generated by FUV is not oxygen dependent, while NUV induced inactivation requires oxygen. Apparently, NUV damage results from the absorption of light by an endogenous photosensitizer or photosensitizers leading to the formation of reactive oxygen species. These oxygen species then react with and damage crucial biomolecules (Larson, 1986; Foote, 1987). Identification of endogenous chrornophores that absorb NUV radiation is a crucial step in understanding how NUV light damages biological tissues. The chromophore that absorbs NUV within the cells is probably not DNA because of its limited absorption in the NUV region of the spectrum (Peak et al., 1984). However, many of the molecules associated with the respira-
*To whom correspondence should be addressed. L4bbreviation.s: A-ALA, A-aminolevulinic acid; FUV, farultraviolet radiation, 200-290 nm; Kn, kanamycin sulfate; LB, Luria-Berlani medium; NUV, near-ultraviolet radiation.
R. E . LLOYDet al.
E. coli apparently capable of taking up exogenously supplied riboflavin. In this communication, we report results with two of these riboflavin auxotrophs consistent with the suggestion (Jagger, 1983) that riboflavin serves as an endogenous NUV photosensitizer. MATERIALS AND METHODS
Bacterial strains. The riboflavin auxotrophs described by Bandrin et al. (1983) represent Tn5 (Kn') insertions into three unlinked regions of the E. coli chromosome. The two mutants used in these investigations represent Tn5 insertions in the ribA (BSV18) and ribB (BSV11) genes. The parental strain (1100-2) has the following genotype: F- thi hsdR hsdM. The parental strain and the two mutants were kindly provided to us by Professor Aziz Sancar of the University of North Carolina at Chapel Hill. Media. The complex medium was Luria-Bertani (LB), consisting of 10 g tryptone (Difco, Detroit, MI), 5 g yeast extract (Difco), and 10 g NaCl (Mallinckrodt, Paris, KY) per liter of H 2 0 . For agar plates, LB was solidified with 1.2% Bacto-agar (Difco). The medium was supplemented, when required, with either 2 or 10 pg me-1 of riboflavin (Nutritional Biochemicals Corporation, Cleveland, OH). The riboflavin auxotrophs were grown in the presence of filter sterilized (Millex-HA 0.45 micron filter unit; Millipore Corp., Bedford, MA) kanamycin sulfate (Kn; 25 pg me-', Sigma, Chemical Co., St. Louis, MO). The Kn was added to liquid medium at room temperature or to 55°C molten agar medium just before pouring. Media and growth conditions. The mutant strains, BSVll and BSV18, were grown at 37°C in LB supplemented with Kn and non-limiting (10 pg me-l) or limiting (2 pg me-l) levels of riboflavin. One milliliter of the overnight culture was used to inoculate 50 me of the same medium and grown to exponential or stationary phase. For experiments involving the parental strain (1 1002), cells were grown overnight to stationary phase and 1.0 me of the culture was used to inoculate fresh medium. The cells were grown to either exponential or stationary phase. For specific experiments, the parental strain was grown in LB supplemented with either 2 or 10 pg me-l riboflavin. Growth was monitored using a KlettSummerson colorimeter equipped with a green filter. Broad spectrum N W puence-response curves. For experiments involving stationary phase cells, about 20 me of the culture were washed three times with 0.85% saline solution (NaC1) to remove residual medium and riboflavin. For experiments involving exponentially grown cells, about 40 me of cells were washed in the same manner. Washed cells were divided into two equal aliquots in 16 x 160 mm Pyrex test tubes and diluted to approx. 5 x loR cells me-l. Cells in one tube were diluted in 0.85% saline, while those in the other were diluted with an equal volume of saline supplemented with riboflavin to give a final concentration of 10 pg me-'. Cells were irradiated and samples taken at predetermined intervals. The samples were diluted appropriately and plated to assess survival. Colonies were counted after 2 4 t o 48 h of incubation at 37°C. Broad-spectrum N W source. The NUV source for all irradiation experiments is kept in a cold room at 10°C. The four 50 cm lamps (GE 40 BLB, integral filter, Westinghouse Electric Corp., Bloomfield, NJ), arranged in a 10 x 10 x 120 cm array, have an emission range of 300425 nm with a maximum energy emission at 350 nm, and emit 97% of their radiant energy between 300 and 400 nm. The fluence rate of the lamps was measured with a DRCl00X Digital Radiometer equipped with a DIX-365 sensor (Spectroline) and found to be 3.12 kJ m-2 min-I. Quantitation of riboflavin levels. Cells of the two mutants and the parental strain were grown to exponential
phase with riboflavin (10 or 2 pg me-'). The cells were washed, resuspended in 0.1 M potassium phosphate buffer (pH 7.0), and broken in a French pressure cell. The extracts were kept in the dark and assayed for riboflavin content. Previous experiments to determine the biological stability of riboflavin in various solvents had also indicated that 2 X lo9 cells contain approx. 1.0 pg riboflavin. This information was used in estimating the volume of cell extract that would contain 3 pg riboflavin. This volume was added to 6 m e of LB broth and sterilized (either by autoclaving or filter sterilization). The growth of the riboflavin auxotroph BSV18 is directly proportional to the amount of riboflavin available. Therefore, the medium was inoculated with BS\;'18 (1 x lo3 to 1 x lo4 cells) and Kn was added. After 48 h of growth at 37°C in the dark, the turbidity was measured in a Klett-Summerson colorimeter and the amount of riboflavin present was determined from a standard curve.
RESULTS AND DISCUSSION
The parental strain and two riboflavin auxotrophs (BSV11 and BSV18) were grown to exponential phase in the presence of non-limiting (10 pg me-l) and limiting (2 pg me-') concentrations of riboflavin. Preliminary experiments had shown that the two mutants cannot grow in LB without riboflavin supplementation. When LB is supplemented with non-limiting riboflavin, the two riboflavin auxotrophs grow at rates nearly equal to that of the parental strain (Fig. 1). When the medium is supplemented with limiting riboflavin, the doubling time during exponential growth for BSVll is increased by a factor of about two, while the maximum cell density attained, as measured in Klett units, is reduced by about a factor of five. For BSV18, growth in limiting riboflavin increased the J 10
Figure 1. Growth curves for the strains used in NUV irradiation experiments. For the purpose of this paper exponential phase was assumed to be the region between the first two sets of points. Additional data points within this region have been omitted for clarity. Mutant cells were grown in LB supplemented with 25 pg me-' kanamycin sulfate. The number after the mutant (BSVl1 or BSVl8) or parental (1100-2) strain indicates the supplementation level of riboflavin in the growth medium (2 = 2 pg me-l, 10 = 10 pg me-I).
Research Note doubling time during exponential growth by about a factor of 1.2, while the maximum cell density attained, as measured in Klett units was reduced by a factor of two. Growing the parental strain (11002) with o r without riboflavin supplementation does not influence doubling time o r maximum cell density attained at stationary phase. Previous work has shown that stationary phase cells are more resistant than exponential phase cells to NUV inactivation (Tuveson and Jonas, 1979). In agreement with this earlier work, when grown to stationary phase the mutants and the parental strains are much more resistant to inactivation by broad-spectrum NUV than exponential phase cells, even when the cells are grown with non-limiting riboflavin. The NUV inactivation kinetics of exponentially growing cells of the parental strain (1100-2) when grown in LB supplemented with either 2 or 10 pg me-' and irradiated in saline were indistinguishable (Fig. 2). When these same cells were irradiated in saline containing riboflavin (10 pg me-]), sensitivity to inactivation by NUV was enhanced by about a factor of two (calculated at the survival level). The inactivation kinetics for cells of the parental strain grown to stationary phase (2.5 h following the transition from exponential to stationary phase) were all similar, with the exception of cells grown in LB supplemented with 10 pg me-' of riboflavin. These cells were less sensitive to NUV inactivation when irradiated in saline without riboflavin. The results of NUV inactivation experiments involving the two riboflavin auxotrophs (BSV11 and 18) are presented in Figs. 3 and 4. Exponential phase cells grown in non-limiting riboflavin (10 pg
BSyl1 exponential cells
stationary B S V l l cells
Fluence (kJ m--2)
Figure 3 . NUV inactivation curves for the riboflavin auxotroph BSVI1. The cells were grown in 2 bg me riboflavin (circles) or 10 pg me-l riboflavin (triangles). They were irradiated in saline (open symbols) and riboflavinsupplemented saline (filled symbols). BSVl8 stationary cells
BSVl8 exponential cells
Fluence (W rn--2)
Figure 4. NUV inactivation curves for the riboflavin auxotroph BSV18. The cells were grown in 2 pg m e - l riboflavin (circles) or 10 pg me-' riboflavin (triangles). Cells were irradiated in saline (open symbols) and riboflavinsupplemented saline (filled symbols).
- 7 . 01 0
Fluence (kl rn^-2)
Figure 2. NUV inactivation curves for the parent strain 1100-2. The cells were grown in 2 pg me-' riboflavin (circles) or 10 pg me-l riboflavin (triangles). Cells were irradiated in saline (open symbols) and riboflavin-supplemented saline (filled symbols).
mew1)are sensitized to inactivation by NUV when compared to the same cells grown in limiting riboflavin. Calculated at the lo-* survival level, exponential phase cells of BSVll grown in non-limiting riboflavin are about three times more sensitive to NUV inactivation, while the same comparison for BSV18 shows that these mutant cells are about 2.5 times more sensitive to NUV inactivation. Irradiation of the exponential phase cells grown in
E. LLOYDet al.
non-limiting riboflavin in the presence of riboflavin (10 pg me-l) did not enhance their sensitivity to NUV inactivation. Cells grown in limiting riboflavin and irradiated in the presence of riboflavin showed an increased sensitivity to NUV irradiation. This may mean that when the mutants are grown to exponential phase in nonlimiting riboflavin the endogenous level of sensitizer is sufficient to result in maximal inactivation, while growth in limiting riboflavin reduces the level of endogenous sensitizer such that exogenous riboflavin can add to the magnitude of inactivation. Sensitivity to NUV was observed for stationary phase cells grown in nonlimiting riboflavin, although the effect was not as dramatic as that observed for exponential phase cells (Figs. 3 and 4). When estimated at the 10-I survival level, the mutants grown to stationary phase in non-limiting riboflavin were between 1.1 and 1.3 times more sensitive to inactivation by NUV. Irradiating stationary phase cells in the presence of riboflavin (10 pg me-]) leads to sensitization, independent of the level of supplementation used to grow the cells for the inactivation experiments. It has been suggested that the membrane is one of the more important targets for NUV in exponentially growing cells as compared to stationary cells (Klamen and Tuveson, 1982). The results with the riboflavin mutants (Figs. 3 and 4) are consistent with this notion; stationary phase cells irradiated in the presence of riboflavin responded as though the lethal membrane target were present, but riboflavin (or a derivative) was not present in the vicinity of the target unless supplied exogenously. The initial assumption was that growing the mutant strains in limiting riboflavin would decrease endogenous levels of riboflavin. In order to validate this initial assumption, measurements were made of endogenous riboflavin present in cells grown in high and low concentrations of riboflavin. Mutant cells grown with limiting riboflavin contained half the endogenous riboflavin of cells grown with non-limiting riboflavin (Table 1). This result suggests a direct
Table 1. Endogenous riboflavin concentrations found in exponentially growing cells Strain
Riboflavin concentration in growth medium
In vivo concentration of riboflavin*
BSVll BSVll BSV18 BSV18 1100-2t 1100-2 ~
2 10 2
*pg riboflavin mggl protein.
tStrain from which the auxotrophs were derived
correlation between a decrease in NUV sensitivity and a decrease in the endogenous riboflavin concentration. The identification of specific NUV chromophores is one of the first steps in understanding the molecular mechanisms by which NUV light damages biological tissues. The results of this investigation indicate that riboflavin (or a derivative) functions as such a chromophore and can be an important endogenous NUV photosensitizer. The location of the chromophore within the cell gives important information about the mode(s) of action. The cytoplasmic membrane constitutes roughly 6% of the volume of an E. coli cell. Preliminary fractionation studies indicate that the concentration of riboflavin in the membrane is 10-fold higher than that in the cytoplasm. It has been suggested that the membrane is a target for NUV inactivation events (Moss and Smith, 1981; Klamen and Tuveson, 1982). A high concentration of riboflavin in the membrane would help explain the importance of the membrane as a NUV target. In addition, the relatively lower mutagenicity of NUV when compared to FUV (Tyrrell, 1980; Turner and Webb, 1980) could be explained if NUV mediated damage occurred primarily at the membrane. We are continuing experiments to further localize and identify NUV photosensitizers. We are also investigating how the riboflavin biosynthetic system is regulated. It may be possible to control endogenous flavin levels experimentally in order to quantitate the amount of endogenous riboflavin necessary for NUV inactivation. Acknowledgements-This research was supported in part by a grant from the USDA (Competitive Research Grant No. 87-CRCR-1-2374) and by a Public Health service grant (1 R 0 1 ES 4397-02). REFERENCES
Bandrin, S . V., P. M. Rabinovich and A. I. Stepanov (1983) Three linkage groups of genes involved in riboflavin biosynthesis in Escherichia coli (in Russian). Genetica 19, 1419-1425. Eisenstark, A. (1987) Mutagenic and lethal effects of nearultraviolet radiation (290-400 nm) on bacteria and phage. Env. Mol. Mutagen. 10, 317-337. Foote, C. S. (1987) Type I and type I1 mechanisms of photodynamic action. In Light Activated Pesticides (Edited by J. R. Heitz and K. R. Downum), Chap. 2, pp. 22-38. ACS symposium series (339) Anaheim, CA. Jagger, J. (1981) Near-UV radiation effects on microorganisms. Photochem. Photobiol. 34, 761-768. Jagger, J . (1983) Physiological effects of near ultraviolet radiation on bacteria. In Photochemical and Photobiological Reviews (Edited by K. C. Smith), Vol. 7, pp. 1-75. Plenum Press, New York. Klamen, D. L. and R. W. Tuveson (1982) The effect of membrane fatty acid composition on the near-UV (300-400 nm) sensitivity of Escherichia coli K l W . Photochem. Photobiol. 35, 167-173. Larson, R. A. (1986) Insect defenses against phototoxic plant chemicals. J . Chem. Ecol. 12, 859-870. Moss, S . H. and K. C. Smith (1981) Membrane damage can be a significant factor in the inactivation of Escher-
ichia coli by near-ultraviolet radiation. Photochem. Photobiol. 33, 203-210. Peak, M. J., J. S . Johnson, R. W. Tuveson and J. G . Peak (1987) Inactivation by monochromatic near-UV radiation of an Escherichia coli hemA mutant growth with and without delta-aminolevulinic acid: the role of DNA vs membrane damage. Photochem. Photobiol. 45, 473-478. Peak, M. J., J. G. Peak, M. P. Moehring and R. B. Webb (1984) Ultraviolet action spectra for DNA dimer induction, lethality and mutagenesis in Escherichia coli with emphasis on the UV B region. Photochem. Photobiol. 40, 613-620. Sammartano, L. J. and R. W. Tuveson (1987) Escherichia coli strains carrying the cloned cytochrome d terminal oxidase complex are sensitive to near-UV inactivation. J . Bacteriol. 169, 5304-5307. Turner, M. A. and R. B . Webb (1980) Comparative mutagenesis and interaction between near-ultraviolet (313-405 nm) and far-ultraviolet (254 nm) radiation in Escherichia coli strains with differing repair capabilities. J . Bacteriol. 147, 410-417. Tuveson, R. W. and R. B. Jonas (1979) Genetic control of near-UV (300-400 nm) sensitivity independent of the recA gene in strains of Escherichia coli K12. Photo-
chem. Photobiol. 30, 667-676. Tuveson, R. W. and L. J. Sammartano (1986) Sensitivity of hemA mutant Escherichia coli cells to inactivation by near-UV light depends on the level of supplementation with A-aminolevulinic acid. Photochem. Photobiol. 43, 621-626. Tyrell, R. M. (1980) Mutation induction by and mutational interaction between monochromatic wavelength radiations in the near-ultraviolet and visible ranges. Photochem. Photobiol. 31, 37-46. Webb, R. B. (1978) Near-UV mutagenesis: Photoreaction of 365-nm-induced mutational lesions in Escherichia coli WP2s. J . Bacteriol. 133, 860-866. Webb, R. B. and M. S. Brown (1979) Oxygen dependence of sensitization to 254-nm radiation by prior exposure to 365 nm radiation in strains of Escherichia coli K12 differing in DNA repair capability. Radiation Rex. 80, 82-91. Webb, R. B . and J. R. Lorenz (1970) Oxygen dependence and repair of lethal effects of near ultraviolet and visible light. Photochem. Photobiol. 12, 383-289. Wilson, A. C . and A. B. Pardee (1962) Regulation of flavin synthesis by E. coli. J . Gen. Microbiol. 28, 283-303.