Mv. SpaceRes.VoI.12,No.4,pp.(4)249-(4)253.l992 0273-1177192$15.00

Printed in GreatBritain. All rights reserved. Copyright© 1991 COSPAR

THE EFFECTSOFVACUUM-UV RADIATION(50-l9Onm)ON MICROORGANISMSAND DNA

Takashi Ito

Hagiyama-cho4-14-21,Higashimurayama-shi,Tokyo189, Japan

ABSTRACT

Early difficulties with the light source in the study of vacuum—UY effects on biologicalmaterials were overcome largely by the introduction of the synchrotron radiation source in1977. Highly monochromatic vacuurn—UV radiation has been used to obtain action spectra ofvarious biological targets in the range 50—190 nm. Bacillus subtilis spores wereparticularly suitable because of their inherent tolerance to vacuum, and action spectra forInactivation were measured down to 50 nm. Action spectra for DNA strand breaks were alsomeasured in the same range with isolated plasmid DNA. Studies are in progress of vacuum—UVinduced molecular changes in nucleic acids and their model compounds, especially of chainscisslon and base modification.

INTRODUCTION

The first experiments specifically on the effects of vacuurn—UV radiation were performedwith Bacillus subtilis spores by Blank and Arnold at Harvard University /1/. Theirshort paper described the irradiation of spores with 110—140 nrn UV radiation in air (airtransmits fairly well in the region 110—130 nm) in a simple apparatus using a carbon sparkand two hF crystal plates, one that transmitted radiation down to 110 nm and the other to140 nm. From their acknowledgements, It is apparent that physicist 1. Lyman assisted withthe irradiation arrangements. Their results demonstrated that radiation at 110 nrn showedgreater germicidal action than did radiation at 140 nrn.

Thirty years later, Jagger et al. /2/ did irradiation experiments with Streptomyces griseusconidia In the dry state using a series of crystal plates with different transmissions ascutoff filters. One of the emphases of their work was characterization of wavelengths forthe photoreactivation of induced damage. They found that the photoreactivable sector(known as an indication of far—WI type damage) of killing gradually decreased aswavelengths decreased Into the vacuum—UV region near 155 nm.

Our project on vacuurn—UY effects, begun 10 years later In 1977, dIffered from Jaggerss j~two important ways. First, synchrotron radiation had become available. This light sourcehas a continuous spectrum of high fluence rate well suited to biological experiments.Second, the wavelength of this light was, in principle, tunable, so any wavelength could beselected with high precision. (It became so in practice, as well as principle, in 1980.)We agreed with earlier investigators that spores or similar targets were logical firstsubjects for vacuum—UV experiments, because the irradiation must be carried out underultrahigh vacuum. We also wanted to observe the induction of sublethal genetic changes, sothat sites of action within the cell could be studied. Later, we extended the work toIsolated nucleic acids, viruses, E. coli, yeast cells, and cultured mammalian cells (forreview, see Ito /3/).

Obviously the finding by Miller in 1955 /4/ that electrical discharges produced organicmolecules from inorganic materials may be related to the effects of the vacuum—WI componentin radiation from the discharge. The evolutionary role of vacuum—UV components in solarradiation In the creation of biological substances and the possible later emergence oftolerance to or active repair of DNA damage are attractive themes to investigate,especially because repair capability is so essential a mechanism in the evolution oforganisms. Moreover, as interest grows in rnans activity in outer space, where vacuum—UVradiation and X—rays are normal components of solar radiation, the effects of thoseradiations take on practical importance as well /5/. The integrated fluence of solar UVradiation over the 115—195 nrn !~velength r~glon alone Is equivalent within minutes to thefluence lethal to bacteria (10 photons/rn ) /6/.

(4)249

(4)250 T. Ito

50 100 150 200 250 300

WAVELENGTH (nm)

FIg. 1. Compiled data of vacuum—UV action spectra for inactivation of Bacil’ussubtills spores having different repair abilities/ 12/. (0) Wild type, uvr[•) UV sensitive, uvrA ssp—1 and (A) WI and X—ray sensitive, uvrAlO ssp—TpolAl5l.

SOURCEOF VACUUM—UVRADIATION

Synchrotron radiation extracted from the electron storage ring at the Synchrotron RadiationLaboratory, the Institute for Solid State Physics, the University of Tokyo, operated at 0.4GeV (INS—SOR II), has been the light source for this project since its Inception in 1977.Monochromatic light Is obtained by use of a modified Wadsworth monochromator, which isparticularly well suited to Irradiation experiments with synchrotron radiation. Thissystem overcomes the major difficulties with vacuum—UV light sources encountered in thepast.

By 1985, we had constructed a system that enabled us to irradiate biological materials atany wavelength from the far—UV region down to 50 nm, In the vacuum—UV region /7,8/. Becauseof the high photon fluence attainable, it is now possible to produce photoproducts forchemical analyses as well as to measure their wavelength dependence (action spectra).

ACTION SPECTRA FOR BIOLOGICAL EFFECTS

The first action spectra were obtained for inactivation and mutation (hls+ reversion) ofBacillus subtilis spores In the wavelength region between 150 and 250 nm. Two mutantstrains dffferent with respect to their repair ability, a far—UV—sensitive (excisiondeficient) strain and an X—ray—sensitive (DNA polymerase I deficient) strain, wereemployed. The results Indicated that, at wavelengths longer than 190 nm, thefar—UV—sensltlve character (a repair sector) prevailed, indicating that the far—UV type ofdamage was Involved. As wavelengths decreased to 150 nm, the difference between wild typeand mutant strains decreased, suggesting that the fraction of damage these specific repairmechanisms (far—UV type) can cope with was diminishing. All these strains, including theX—ray—sensitive one, tended to converge to the same sensitivity. These characteristicswere approximately the same for both lethal action and mutation induction /9/, providingevidence that vacuum—UV radiation in this wavelength region can reach the DNA locatedInside the cell in doses high ~pou~hto cause DNA damage. The cross section for theInactivation was about 1 x 10” in’ at 150 fin. With yeast cells (4—5 urn diameter) littleInduction of genetic changes was observed at 150 nm, presumably because of the poorpenetration of photons into the nuclear area /10/. Leakage of intracellular elements fromirradiated yeast cells at various wavelengths paralleled the observed survival responses/11/, so evidence for cell—mentrane damage was rather convincing for these relatively largecells.

Effectsof Vacuuni-IJVRadiation (4)251

“. ABSORPTIONN—1016 N

E

z STRAND BREAK-18

I.-L)[ii‘I,

~ -19ttt 10 -

0CEC)

io-20 -

I I

100 150 200 250

WAVELENGTH(nm)

FIg. 2. UV action spectrum for single strand breaks Induced In pBR322 DNA.The absorption spectrum of DNA Is also shown. (Modified from /13/.)

Extension of the wavelength region toward the shorter wavelengths has been a continuedeffort. Recently 50—nm experiments have become possible. Compiled action spectra updatedwith Bacillus spores are shown in Figure 1 /12/. First, the trend seen in 150—190 nniregion continued, and the sensitivity (cross section) of the three strains convergedcompletely at 130 nm. Second, although some structure was seen in the spectrum, the crosssection at wavelengths below 150 nm was generally not higher than those above 150 nm. Thisresult is apparently due to the lower penetration of extremely short wavelength vacuum—WIradiation into the region where the DNA is located. Because. In In vitro experiments, thecross section of DNA damage Increases monotonically In this wavelength region (see Figure2), the generally flat action spectrum in this region probably does not reflect a truesensitivity change.

ACTION SPECTRUMFOR DNA STRAND BREAKS

Detailed measurements of single strand breaks in the vacuum—WI region were carried out withpBR322 plasmid ONA by S~uk~and Hieda /13/.,,~Th~ action spectrum Indicated that the crosssection ranged from 10 in at 8~nm to 10 m at 190 nm (Figure 2). At 150 tim thecross section was roughly 2 x 1o 8 in2. Although the cross section below 120 tim paralleledthe estimated DNA absorption cross section, above 120 nm It progressively deviated from theabsorption spectrum toward much lower sensitivity, suggesting that the absorption by bases,especially their w electron system, plays little role In strand breaks. This conclusion Issupported by the disproportionately smaller efficiency in the two regions, around 190 nmand 250 nm, known as the resonance absorption bands of bases.

NATURE OF VACUUM—UVINDUCED MOLECULARCHANGES

Methods better suited to investigation of the mechanisms of vacuum—WI Induced molecularchanges in nucleic acids are those that use a DNA model compound as the target. Suchexperiments employing dinucleoside monophosphates and small ollgonucleotides of knownchemical structures are in progress In my laboratory /14,15/.

With 2’—deoxyadenylyl—(3’—5’)—2—deoxyadenosine, dApdA, in which only one phosphodiesterbond Is present in the molecule, several distinct features were noted when the vacuum—WIphotoproducts were analyzed by thin—layer chromatography. (a) There were only two major

(4)252 T. Ito

d(pA)5

d(pA)4d(pA)3 d(pA)2

~I I / 4-dAMPI / /5-dAMP Adenine

~ k~ ~

SF

U NIRRADIAT ED

I I I I I0 2.0 4.0 6.0MIGRATION DISTANCE(cm)

Fig. 3. HPTLC chro9~ogram of va~uum—UV(163 nm) Irradiated d(pA)5. Thefluence was 5.4 x 10 photons /m

photoproducts, adenine and 5’—dAMP, in approximately equal quantities. (b) Thephotoproducts from dApdC followed the same rule; that Is, adenine and 5’—dCMP were themajor ones, but dCpdA was more stable and Its degradation mode more complex than those ofdApdA or dApdC. Cc) The pyrimidine dinucleoside monophosphates such as dTpdT and dTpdCwere more stable than the purine—containing ones. Cd) The 5’ phospho—derivative of dApdA,that is, adenine dinucleotide pdApdA, was different from dApdA in its degradation. Thepresence of a phosphate at the 5’ terminal appeared to render the molecule much moreresistant to vacuum—UV degradation, as demonstrated by phosphodiester bond breakage.

These results mean that strand breakage caused by vacuum—WI radiation is not a single—stepevent but is a series of events dependent on several molecular factors. Our presentconcept of these events Is that photon energy is initially absorbed in a sugar—phosphategroup and that sugar destruction follows, accompanied by breakage of the ester bond between3’C of the sugar and the phosphate, leaving a 5’ phosphate terminal at one end (5’—dAMP inthe case of dApdA). A base adenine attached to the destroyed sugar moiety may then bereleased. The other end then may have a sugar or 3’OII (inorganic phosphate may well alsobe released during these processes) residue. Different base sequence and base compositionfactors may also Influence the processes. If this concept is applied to polynucleotides,5’—dXMP would be lost at the site of the strand break.

The experiments were recently extended to include oligonucleotides of adenine, d(pA) for n= 3 5 /16/. In every case all possible degraded oligonucleotides and mononucleoti~e(predominantly 5’ type) components, namely d(pA)k for k satisfying n — k ~ 1, weredetected. Figure 3 shows a thin—layer chromatogram of vacuum—UV—irradiated d(pA)5. Therelative quantities of products Imply that adenine release occurs with every chainscission.

Base modifications such as thymine diiners and T(6—4)T have also been detected by HPLC inthe wavelength region from 150 to 300 nm using dTpdT /17/, but to a much smaller extent

Effects ofVacuum-UVRadiation (4)253

than the degradation products, thymine and TMP. With DNA these base modifications werealso measured In the same wavelength range /18,19/. The action spectrum resentled that ofDNA absorption for thymine diner formation, but not for 6—4 photoproduct below 180 fin.This result Is of interest In view of the reported similarity of action spectra for thesetwo types of photoproducts in the far—UV region /20/. These investigations show that bothtypes of base modifications occurred not only In the far—UV region but also in the vacuum—UV region, although the cross section was an order of magnitude lower at 150 tim than at 250nm.

Acknowledgements ——— Dr. N. Munakata, National Cancer Center Research Institute, kindlypermitted use of his unpublished materials. Dr. 1. Ogawa, University of Tokyo, kindlyprovided geophysical Information on solar radiation. Dr. Anne B. Thistle, Florida StateUniversity, critically edited the manuscript.

REFERENCES

1. I.H. Blank and W. Arnold, J. Bacteriol. 30, 503 (1935).

2. J. Jagger, R.S. Stafford, and R.J. Mackin, Jr., Radlat. Res. 32, 64 (1967).

3. T. Ito, in: Synchrotron Radiation in Structural Biology, ed. R. M. Sweet and

Avrll D. Woodhead, Plenum Press, 1989, p. 221.4. S.L. Miller, J. Am. Chem. Soc. 77, 2351 (1955).

5. 6. Horneck, H. Bucker, and R. Reitz, Abstract, 10th Int. Cong. Photobiol.(Jerusalem), 1988, p. 50.

6. L. Heroux and H.F. Hinteregger, J. Geophys. Res. 83, 5305 (1978).

7. T. Ito, T. Kada, 5. Okada, K. Hieda, K. Kobayashi, H. Maezawa, and A. Ito, Radlat.Res. 98, 65 (1984).

8. K. Hieda, H. Maezawa, A. Ito, K. Kobayashi, Y. Furusawa, and T. Ito, Photochem

.

Photobiol. 44, 417 (1986).

9. N. Munakata, K. Hieda, K. Kobayashi, A. Ito, and T. Ito, Photochem. Photobiol. 44,385 (1986).

10. K. Hieda and T. Ito, Photochem. Photobiol. 44, 409 (1986).

11. S. Matsumoto, A. Ito, and T. Ito, Radiat. Environ. Blophys. 23, 287 (1984).

12. N. Munakata, private coninunication (1990).

13. M. Suzuki and K. Hleda, Activity Rep. Synchr. Rad. Lab., Univ. Tokyo, 1986. p. 78.

14. T. Ito, M. Saito, and T. Taniguchi, Photochem. Photobiol. 46, 979 (1987).

15. T. Ito and M. Saito, Photochem. Photobiol. 48, 567 (1988).

16. T. ito and M. Saito, Radlat. Phys. Chem., in press (1990).

17. M. Saito and K. Hieda, Activity Rep. Synchr. Rad. Lab., Univ. Tokyo, 1989, p. 49.

18. T. Matsunaga and 0. Nikaldo, Activity Rep. Synchr. Rad. Lab., Univ. Tokyo, 1987,

p. 62.

19. T. Matsunaga, private coninunicatlon (1990).

20. G.L. Chan, M.J. Peak, 3.6. Peak, and W.A. Haseltine, mt. 3. Radiat. Blol. 50, 641(1986).