10636
2007, 111, 10636-10638 Published on Web 08/21/2007
Bringing Electrons and Microarray Technology Together T. Solomun*,† and H. Sturm‡ Free UniVersity Berlin, Institute of Chemistry and Biochemistry, Physical and Theoretical Chemistry, Takustrasse 3, D-14195 Berlin, Germany, and Federal Institute for Materials Research and Testing (BAM), Unter den Eichen 87, D-12205 Berlin, Germany ReceiVed: July 9, 2007; In Final Form: July 29, 2007
Low-energy secondary electrons are the most abundant radiolysis species which are thought to be able to attach to and damage DNA via formation and decay of localized molecular resonances involving DNA components. In this study, we analyze the consequences of low-energy electron impact on the ability of DNA to hybridize (i.e., to form the duplex). Specifically, single-stranded thymine DNA oligomers tethered to a gold surface are irradiated with very low-energy electrons (E ) 3 eV, which is below the 7.5 eV ionization threshold of DNA) and subsequently exposed to a dye-marked complementary strand to quantify by a fluorescence method the electron induced damage. The damage to (dT)25 oligomers is detected at quite low electron doses with only about 300 electrons per oligomer being sufficient to completely preclude its hybridization. In the microarray format, the method can be used for a rapid screening of the sequence dependence of the DNA-electron interaction. We also show for the first time that the DNA reactions at surfaces can be imaged by secondary electron (SE) emission with both high analytical and spatial sensitivity. The SE micrographs indicate that strand breaks induced by the electrons play a significant role in the reaction mechanism.
In irradiated living cells, secondary electrons with an initial energy below 10 eV greatly outnumber those with higher energies.1 Experimental evidence is emerging that shows these electrons can attach resonantly to DNA and initiate its fragmentation. For example, Sanche and co-workers studied by the method of electrophoreses electron induced strand breaks in plasmid DNA as a function of electron energy.2,3 Below about 3 eV, only single strand breaks are observed, but, at higher electron energies, double strand breaks also occur. A support for this comes from mass spectrometry data4 which show that the basic DNA components in the gas phase decompose in a cross-beam of near-zero-energy electrons. A full understanding of the biological effects of ionizing radiation must therefore incorporate detailed knowledge of the action of low-energy electrons, and this is presently a field of vigorous experimental and theoretical activities.5-8 The goal of understanding the mechanism by which low-energy electrons damage DNA remains, however, elusive. This is in part because only a few experimental methods have been applied so far, and these only at extremes of DNA complexity (plasmid DNA vs the very basic DNA components). Thus, experiments using DNA oligomers to assess the electron damage mechanism are indeed scarce.9-11 We have chosen in this study thymine (dT)25 oligomers because their immobilization on gold has been investigated in detail with a number of techniques.12-14 To assess the electron damage, we analyze the ability of single-stranded DNA (ssDNA) to get involved in the interaction with the complementary strand * Corresponding author. E-mail:
[email protected]. Phone: +49-30-83855307. Fax: +49-30-83855612. † Free University Berlin. ‡ Federal Institute for Materials Research and Testing (BAM).
10.1021/jp075338v CCC: $37.00
after it has been bombarded with low-energy electrons. It is important to note here that in this study the changes in the hybridization ability obserVed by fluorescence are always normalized relatiVe to control probes on the same chip. The control DNA probes, which are shielded in an ultrahigh Vacuum (UHV) by a mask, experience exactly the same processing treatment except for the lack of electron irradiation (schematics in Supporting Information). In a previous study,15 we were concerned with the feasibility of such measurements, and we provide here more accurate data because of various system improvements. The single-stranded oligomers were obtained from Thermo Electron (Germany). Their structure 5′-SH(CH2)6-(dT)25-3′ comprises a thiol link for immobilization on gold. We note briefly that the degradation of alkanethiols such as HS-(CH2)5CH3 adsorbed on gold does not occur at electron energies used in this work.16,17 The oligomers are dissolved in SSC (sodium chloride/sodium citrate) buffer and deposited onto 12 mm × 12 mm gold-on-glass chips (Arrandee, Germany) either with a micropipette as large islands (∼1 mm diameter) or in the form of microarrays (250 µm pitch, 100 µm spot diameter) using a Genetix QArray Microarrayer and a split pin (Telechem). Essentially the same DNA damage data were obtained in both cases. Development and testing of the latter format is important for the future studies of the sequence dependence of electron processes using a large number of different probes on a single chip. For this, we have spotted five different microarrays (Figure 1) on a chip. Only the central (8 × 8) microarray is exposed to the electrons, while the four others in corners serve as control arrays and are shielded by a mask during the irradiation. The immobilized probes are then first © 2007 American Chemical Society
Letters
Figure 1. Normalized hybridization efficiency (fluorescence intensity) of 25-mer thymine oligomers on gold as a function of exposure to 3 eV electrons. Inset: Fluorescence images of a gold chip with different (dT)25 microarrays before (left) and after (right) exposure of only the central (8 × 8) array to the low-energy electrons.
hybridized with a dye-marked complementary strand (Cy5dA25). An Affymetrix 418 Array Scanner and GenePixPro6.0 software are used to obtain the initial median area intensity values (I°). After denaturizing by a brief (30 s) immersion into hot (∼95 °C) water, the chips are transferred into an UHV chamber for electron irradiation using a home-built electron gun/ control system. With the electron lens arrangement of the gun, there is no direct filament-to-probe sight and, therefore, possible light or thermal processes are excluded. The electron energy distribution is estimated to have a full width at half-maximum (fwhm) of about 0.5 eV. The incident electron-beam energy is defined by the voltage drop from the filament to the surface. The beam current at the surface was measured through an aperture in a mask in front of it by using a Keithley 6485 instrument and a commercial input/output card (NI 6014). The mask shields the control microarrays from the electrons, while permitting the irradiation of the central microarray. Negative tests are carried out, deflecting the electron beam from the Au surface to confirm that the observed effects are due to electrons. After the irradiation, the immobilized DNA probes were washed and again hybridized to obtained the fluorescence intensities (I) which are normalized as J ) (I/I°). The standard deviation of J values was estimated to be about 14%. The DNA damage parameter (percentage decrease in the fluorescence) is defined as (1 - J/Jcontrol) × 100. Secondary electron microscopy (SEM) images were obtained with a FEI-XL30 ESEM instrument at 0.7 kV primary beam energy and analyzed with GenePixPro6.0 software. In Figure 1, the dependence of fluorescence intensity as a function of electron doses is shown. Together with the images presented in the figure, the data clearly demonstrate decomposition of the oligomers by the impact of low-energy electrons. Various reaction channels (sugar-phosphate backbone fragmentation, base fragmentation or excision, reaction of radical sites created on oligomers, etc.) can contribute to oligomer
J. Phys. Chem. B, Vol. 111, No. 36, 2007 10637
Figure 2. Normalized secondary electron emission contrast for T25 oligomers on gold as a function of immobilization time (symbols). The full line refers to independent surface plasmon resonance measurements for the same system.12 Inset: SE micrographs of the pristine (below) and with 3 eV electrons irradiated (above) T25 spots on gold (100 µm spot diameter; adsorption time 120 min; intensity profiles are through the two lower spots in each micrograph).
damage, reduce its ability to hybridize, and lower the fluorescence. These different contributions are difficult to disentangle on the basis of the fluorescence data with a single oligomer. Nevertheless, in the initial linear low-electron-dose part of the curve in Figure 1, a linear fit would be proportional to a total cross section of degradation. From it, we estimate (assuming 1013 oligomers/cm2)14 a value for the total cross section of about 4 × 10-16 cm2. This value is significantly higher than the cross section (5 × 10-17 cm2) obtained by the electron stimulated, mass-selected neutral desorption yields for ssDNA on gold consisting mainly of fragments of the DNA bases.9 This might indicate a more extensive type of damage such as strand breaks that are detected by the fluorescence method. In addition, the curve in Figure 1 shows saturation at about 300 electrons per oligomer. It is worth recalling here that the formation of a single strand break in plasmid DNA at 3 eV requires about 200 electrons.2,3 This is in good agreement with this work considering the different types of DNA samples as well as different methods used. Possible scenarios for the formation of strand breaks are electron capture by the nucleotide bases and subsequent transfer of the excess charge to the backbone,7 electron attachment directly to the backbone,6 or electron induced proton transfer within a nucleotide leading to strand break.18 On the other hand, it has been reported that a mismatch within a short nucleotide strand can significantly reduce its hybridization efficiency.19 To a first approximation, a mismatch mimics the electron damage to a single DNA base, and therefore, to assess more properly the nature of the damage, application of additional methods is needed. Secondary electron (SE) imaging is seemingly a well suited technique for it. The escape depth of secondary electrons from a metal surface is relatively short (of the order of 1 nm), and it is known that adsorbed thin films can strongly modulate the SE yield of a metal surface.20 This is confirmed by the images in Figure 2. It is a useful property in this regard that the Au samples used in this study are extremely smooth, and thus minimize image contributions due to sample topography. This
10638 J. Phys. Chem. B, Vol. 111, No. 36, 2007 low background is essential to allow the quantification of SEM image contrasts in terms of adsorbate coverage. The surface sensitivity of SE microscopy relies on a variety of factors which can affect the image contrasts. These include the ad-layer thickness, its composition, electronic structure, etc. The electron scattering cross sections, however, are not expected to be severely impacted by the varying DNA composition induced by the electron impact. The observed intensity is believed to be dictated by film thickness to a large extent. We have carried out Monte Carlo calculations21 on model systems confirming this (Supporting Information). Therefore, the decrease in the SE contrast in Figure 2 upon exposure of DNA to the lowenergy electrons implies the formation of strand breaks. In conclusion, we have shown that the DNA damage (and the concomitant loss of genetic information) occurs at surprisingly low electron energies and doses. Combination of microarray and SE emission data indicates that a significant part of the damage concerns strand breaks. The fluorescence-based microarray technique developed here is a very sensitive method which can be applied to study sequence dependence of various other physical processes involving DNA. Acknowledgment. Financial support by DFG (Deutsche Forschungsgemeinschaft), help in printing of microarrays by Dr. C. Hultschig (Max-Planck-Gesellschaft, Berlin), and useful discussion with Prof. E. Illenberger (Free University Berlin) are gratefully acknowledged by one of us (T.S.). Supporting Information Available: Schematic of the experimental procedure involved in the microarray experiments and calculated secondary electron emission for DNA ad-layer on gold using CASINO software. This material is available free of charge via the Internet at http://pubs.acs.org.
Letters References and Notes (1) Radiation Damage in DNA: Structure/Function Relationships at Early Times; Fuciarelli, A. F., Zimbrick, J. D., Eds.; Battelle: Columbus, OH, 1995. (2) Boudaiffa, B.; Cloutier, P.; Hunting, D.; Huels, M. A.; Sanche, L. Science 2000, 287, 1658. (3) Martin, F.; Burrow, P. D.; Cai, Z.; Cloutier, P.; Hunting, D.; Sanche, L. Phys. ReV. Lett. 2004, 93, 068101. (4) Bald, I.; Kopyra, J.; Dabkowska, I.; Antonsson, E.; Illenberger, E. J. Chem. Phys. 2007, 126, 074308-1 and references therein. (5) Hotop, H.; Ruf, M. W.; Allan, M.; Fabrikant, I. I. AdV. At. Mol. Opt. Phys. 2003, 49, 85. (6) Sanche, L. Eur. Phys. J. D 2005, 35, 367. (7) Simons, J. Acc. Chem. Res. 2006, 39, 772. (8) Eur. Phys. J. D 2005, 35 (2) (Special Issue: Electron-Driven Molecular Processes). (9) Dugal, P.-C.; Huels, M. A.; Sanche, L. Radiat. Res. 1999, 151, 325. (10) Ray, S. G.; Daube, S. S.; Naaman, R. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 15. (11) Pan, X.; Sanche, L. Phys. ReV. Lett. 2005, 94, 198104. (12) Herne, T. M.; Tarlov, M. J. J. Am. Chem. Soc. 1997, 119, 8916. (13) Petrovykh, D. Y.; Kimura-Suda, H.; Whitman, L. J.; Tarlov, M. J. J. Am. Chem. Soc. 2003, 125, 5219. (14) Levicky, R.; Herne, T. M.; Tarlov, M. J.; Satija, S. K. J. Am. Chem. Soc. 1998, 120, 9787. (15) Solomun, T.; Hultschig, C.; Illenberger, E. Eur. Phys. J. D 2005, 35, 437. (16) Olsen, C.; Rowntree, P. A. J. Chem. Phys. 1998, 108, 3750. (17) Huels, M. A.; Dugal, P.-C.; Sanche, L. J. Chem. Phys. 2003, 118, 11168. (18) Dabkowska, I.; Rak, J.; Gutowski, M. Eur. Phys. J. D 2005, 35, 429. (19) Peterson, A. W.; Wolf, L. K.; Georgiadis, R. M. J. Am. Chem. Soc. 2002, 124, 14601. (20) Mack, N. H.; Dong, R.; Nuzzo, R. G. J. Am. Chem. Soc. 2006, 128, 7871. (21) Drouin, D.; Couture, A. R.; Gauvin, R.; Hovington, P.; Horny, P.; Demers, H. Monte Carlo simulation of electron trajectories in solids (CASINO), version 2.42; University of Sherbrooke: Quebec, Canada (http:// www.gel.usherbrooke.ca/casino/index.html).
