- Email: [email protected]
Application of capillary gas chromatography-mass spectrometry to chemical characterization of radiation-induced base damage of DNA: Implications for assessing DNA repair processes
ANALYTICAL
BIOCHEMISTRY
14,593-603
(1985)
Application of Capillary Gas Chromatography-Mass Spectrometry to Chemical Characterization of Radiation-Induced Base Damage of DNA: Implications for Assessing DNA Repair Processes MIRAL DIZDAR~GLU Center for Radiation
Research,
National
Bureau
of Standards,
Gaithersburg,
Maryland
20899
Received April 17, 1984 The application of capillary gas chromatography-mass spectrometty (CC-MS) to the chemical characterization of radiation-induced base products of calf thymus DNA is presented. Samples of calf thymus DNA irradiated in N@-saturated aqueous solution were hydrolyzed with HCOOH, trimethylsilylated, and subjected to CC-MS analysis using a fused-silica capillary column. Hydrolysis conditions suitable for the simultaneous analysis of the radiation-induced products of all four DNA bases in a single run were determined. The trimethylsilyl derivatives of these products had excellent GC properties and easily interpretable mass spectra; an intense molecular ion (MC) and a characteristic (M-CH,)+ ion were observed. The complementary use of t-butyldimethylsilyl derivatives was also demonstrated. These derivatives provided an intense characteristic (M-57)+ ion, which appeared as either the base peak or the second most intense ion in the spectra. All mass spectra obtained are discussed. Because of the excellent resolving power of capillary GC and the accurate high-sensitivity identification by MS, the capillary CC-MS is suggested as a very suitable technique for identification of altered bases removed from DNA by base excision-repair enzymes such as DNA glycosylasesand, thus, as very useful for an understanding of the base excision-repair of DNA. o 1985 Academic RS.S, IW. KEY WORDS: DNA damage; ionizing radiation; DNA repair enzymes; mass spectra; trimethylsilylation; t-BDMS derivatives; high sensitivity.
Ionizing radiation creates a number of lesions in DNA, including modified bases, base-free sites, and single- and double-strand breaks (1). DNA-protein crosslinks are also formed when cells or mixtures of DNA and proteins are exposed to ionizing radiation (2,3). In living cells, lesions produced in DNA are subject to repair processes [for a review see Ref. (4)]. Unless repaired, such damages may have detrimental biological consequences, such as cell death, mutagenesis, or carcinogenesis (5). Elucidation of the chemical nature of radiation-induced DNA lesions is, therefore, necessary for an understanding of their biological consequences and enzymatic repairability. The effects of ionizing radiation on DNA and its constituents have been extensively studied under various conditions using a 593
variety of analytical techniques. For instance, thin-layer chromatography, high-performance liquid chromatography (HPLC), nuclear magnetic resonance spectroscopy, and mass spectrometry (MS)’ were used for identification of radiation-induced products of DNA bases (6- lo), and capillary gas chromatography (GC) combined with MS (GC-MS) was extensively applied to characterization of radiation-induced products of the DNA sugar moiety (11-16). In the past, GC-MS has played an important role in the characterization of natural ’ Abbreviations used: CC-MS, gas chromatographymass spectrometry; TMS, trimethylsilyl; BSTFA, bis(trimethylsilyl)trifluoroacetamide; t-BDMS, t-butyldimethylsilyl; MC, molecular ion; SIM, selected-ion monitoring; ‘H NMR, proton nuclear magnetic resonance spectroscopy. 0003-2697185 $3.00 Copyright Q 1985 by Academic Press. Inc. ,411 rigbu of reproduction in any form rmwed.
MIRAL
DIZDARGGLU
FIG. 1. GC separation of a trimethylsilylated HCOOH hydrolysate of irradiated calf thymus DNA. Dose: 330 Gy. Column: fused silica capillary (12 m, 0.2-mm i.d.) coated with crosslinked SE-54, programmed after 3 min at 100°C from 100 to 250°C with a rate of 7”C/min. For other details, see Materials and Methods. Peak identifications are given in Table 1.
and synthetic nucleic acid bases and nucleosides (17-22). However, this technique has not yet been applied to identification of radiation-induced base damage of DNA. Recently, we made extensive use of capillary GC-MS for isolation and characterization of radiation-induced products of DNA bases such as thymine and cytosine and their nucleosides and nucleotides (23-25), as well as of radiation-induced crosslinks between a DNA base, i.e., thymine, and aromatic amino acids such as phenylalanine (26) and tyrosine (27). In all instances, trimethylsilyl (TMS) derivatives of monomeric and dimeric products were shown to have excellent GC properties and easily interpretable mass spectra.* The present paper describes the application of capillary GC-MS to identification of * For details such as separation conditions, sensitivity, use of capillary columns with different coatings, interpretation of mass spectra obtained, and comparison of different derivatization methods, see Ref. (25).
radiation-induced mus DNA. MATERIALS
base products of calf thyAND METHODS3
Chemicals. Calf thymus DNA and bis(trimethylsilyl)trifluoroacetamide (BSTFA), 5methylcytosine, and 5-hydroxyuracil (isobarbituric acid) were purchased from Sigma. Acetonitrile and pyridine were obtained from Pierce. N-Methyl-N-(t-butyldimethylsilyl)trifluoroacetamide was from Regis. Isodialuric acid (5,6-dihydroxyuracil) was synthesized by oxidation of isobarbituric acid with bromine (28). Irradiations. Aqueous solutions of calf thymus DNA (250 mg/liter) were saturated 3 Certain commercial equipment, instruments, or materials are identified in this paper in order to specify adequately the experimental procedure. Such identification does not imply recommendation or endorsement by the National Bureau of Standards, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.
GAS CHROMATOGRAPHY-MASS
SPECTROMETRY TABLE
PEAK IDENTETCATION
Peak No.’ I II
2 III d 3 4 5 IIIa 6
Compoundb
OF DNA RADIATION
595
1 IN FIGURE
Peak No.“
Phosphoric acid Uracil Thymine 5,6-Dihydrothymine
7, 8 9 IV 10
Cytosine 5-Methylcytosine 5-Hydroxy-5,6dihydrothymine
IVa 11 12
5-Hydroxyuracil 5-Hydroxy-5,6dihydrouracil Cytosine’ 5-Hydroxycytosine
DAMAGE
V Va 13
1
Compound b cis- and trans-thymine glycol 5,6Dihydroxyuracil Adenine 4,6-Diamino-5-formamidopyrimidine Adenine’ I-Hydroxyadenine 2,6-Diamino-4-hydroxy-5formamidopyrimidine Guanine GuanineC 8-Hydroxyguanine
n Peaks a to g were also present in samples of nonirradiated calf thymus DNA, which were treated in the same manner as the irradiated samples. Their identities were not determined, except for peak d, which represents the TMS derivative of 5-methylcytosine. b As their TMS derivatives. c With an additional TMS group attached to the amino group.
with oxygen-free N20 and irradiated in a 6oC y-source. (dose range 1lo-440 Gy; dose rate 110 Gy/min).4 Samples were then lyophilized. Hydrolysis with HCOOH. About 2 mg of lyophilized samples was hydrolyzed with 1 ml of HCOOH (88%) at 150°C for 30 min in evacuated and sealed tubes. After hydrolysis, the samples were dried in a desiccator in vacua prior to derivatization. Derivatization. Hydrolyzed and dried samples were trimethylsilylated in PTFE-capped Hypo-vials (Pierce) with 0.2 ml of a mixture of BSTFA and acetonitrile (1: 1) by heating for 30 min at 140°C. t-Butyldimethylsilyl (tBDMS) derivatives were prepared according to Crain and McCloskey (29). Gas chromatography. A Hewlett-Packard Model 5880A microprocessor-controlled gas chromatograph equipped with a flame ionization detector was used. The injection port and detector were maintained at 250°C. Separations were achieved on a fused-silica capillary column ( 12 m, 0.2-mm i.d.) coated 4 1 Gray = 1 joule/kg = 100 rad.
with crosslinked SE-54 [5% phenylmethylsilicone; wall coated open tubular; siloxane deactivated; film thickness, 0.11 pm (Hewlett-Packard)]. The measured efficiency of the column was ca. 5400 theoretical plates per meter based on the pentadecane peak at 120°C (capacity factor, 6.42; linear velocity of the carrier gas, 32.8 cm/s). Helium was used as the carrier gas at an inlet pressure of 100 kPa. The split ratio was 20: 1. Gas chromatography-mass spectrometry.
Mass spectra were taken at 70 eV using a Hewlett-Packard Model 5970A mass selective detector interfaced to the above gas chromatograph. GC conditions were as above. The ion source temperature was ca. 200°C. RESULTS AND DISCUSSION Acid hydrolysis. When DNA is irradiated, free bases and some degradation products are released from DNA; however, the modified bases remain attached mostly to the polynucleotide chain (6,7,9). It is therefore necessary to remove them from the DNA backbone prior to their chemical analysis. In the
596
MIRAL
DIZDAROGLU
urm N/ iT a 52 A d
73
TMSo
ks
329 mMs
N
Mu 344
343
147
FIG. 2. Mass spectrum taken from peak 4 in Fig. I.
past, formic or perchloric acid hydrolysis was used to determine the extent of the radiationinduced base modification of DNA (6). Recently, Teoule et al. (9) described the optimal conditions for the removal, by acid hydrolysis, of radiation-induced products of thymine, cytosine, and adenine from irradiated DNA. Different conditions and acids, i.e., HCOOH and HF, were used for different bases. Products of guanine were not discussed. In the present paper, hydrolysis with 88% HCOOH at 150°C for 30 min was found to be suitable, as a result of the studies of concentration and temperature, for simultaneous identification of radiation-induced products of all four bases of DNA in a single run. GC-MS anafyszs. Figure 1 shows the GC separation of a trimethylsilylated HCOOH hydrolysate of irradiated calf thymus DNA. Peak identification is given in Table 1. Peaks I, II, III, IV, and V represent the TMS derivatives of phosphoric acid, thymine, cytosine, adenine, and guanine, respectively. Peaks IIIa, IVa, and Va also represent the last three bases, but with an additional TMS group attached to their amino groups as revealed by comparison with authentic material and by MS. Peaks a to g were also present in the gas chromatogram of nonirradiated HCOOH-hydrolyzed DNA samples. Their origin could not be determined, with the exception of peak d, which corresponds to the TMS derivative of 5-methylcytosine, for which authentic material was available. Calf thymus DNA is known to have a 5-
methylcytosine content of about 1.3% (30). Peaks 1 to 13 represent the TMS derivatives of the radiation-induced base products of DNA. Products represented by peaks 1 to 3, 6 to 8, and 10 to 12 (Table 1) were reported in previous studies using analytical techniques other than CC-MS (6,7). 8-Hydroxyguanine represented by peak 13 in Fig. 1 has recently been identified in y-irradiated DNA as 8hydroxy-2’deoxyguanosine alter enzymatic hydrolysis using CC-MS, HPLC, uv spectroscopy, and ‘H NMR.’ The presence of 8hydroxy-2’deoxyguanosine in y-irradiated 2’deoxyguanosine and 2’deoxyguanosine 5’monophosphate was also shown using the same techniques without the acid hydrolysis5 The acid hydrolysis of 8-hydroxy-2’deoxyguanosine yielded 8-hydroxyguanine.5 The remaining products (peaks 4, 5, and 9) have not been previously described.6 Mass spectra of the TMS derivatives of 5,6dihydrothymine (peak 2), 5-hydroxy-5,6dihydrothymine (peak 3) 5-hydroxycytosine (peak 6), and thymine glycol (peaks 7 and 8) were published in a previous paper (25) and will not be discussed here. A mass spectrum taken from peak 4 is given in Fig. 2. From this mass spectrum, peak 4 was assigned as the TMS derivative of 5-hydroxyuracil (isobarbituric acid), for ’ M. Dizdaroglu, submitted for publication. 6 SHydroxyuracil (isobarbituric acid) and 5,6dihydroxyuracil (as isodialuric acid) were observed in the presence of oxygen [see Ref. (9)].
GAS CHROMATOGRAPHY-MASS
SPECTROMETRY
OF DNA RADIATION
DAMAGE
597
331
116
346
q-T--
400
FIG. 3. Mass spectrum taken from peak 5 in Fig. 1.
which authentic material was commercially available. The molecular ion (M+*) and a characteristic (M-H)+ ion were observed at m/z 344 and 343, respectively, and the (MCH$ ion appeared as the base peak at m/z 329; loss of a hydrogen or methyl radical from M+’ of TMS derivatives of nucleic acid bases constitutes their major fragmentation pathways (17-19). Ions at m/z 73 and 147 are commonly observed with TMS derivatives and serve no diagnostic purpose. Figure 3 shows the mass spectrum taken from peak 5 in Fig. 1. M+’ and (M-CH$ ions were observed at m/z 346 and 33 1, respectively. No authentic material was available for this compound. On the other hand, M+’ and (M-CHj)+ ions clearly reveal the addition of an OH group and H atom to the uracil skeleton. Based on the well-known reaction mechanisms of OH radicals with pyrimidines (31,32), this compound can be attributed to the TMS derivative of 5- or 6hydroxy-5,6dihydrouracil. These two com-
pounds should give similar mass spectra as derived from the known mass spectra of TMS derivatives of corresponding thymine derivatives (25). 6-Hydroxy-5,6-dihydropyrimidines, however, are well known to be unstable in hot acid solutions (6). Thus, peak 5 was assigned as the TMS derivative of 5hydroxy-5,6-dihydrouracil. In addition, characteristic fragment ions in Fig. 3 clearly confirm this assignment as explained below: for instance, one of the most structurally diagnostic ions in mass spectra of TMS pyrimidines results from opening of the pyrimidine ring after formation of the (M-CHj)+ ion (17-19). As measurement of exact mass and deuterium labeling clearly showed (17) the resulting product contains carbons 4 and 5 and their substituents and is useful in distinguishing modification at C-5 versus C6. The ion at m/z 188 +
4
5
[(CH&Si=O-m--OTMS] most probably arises from such a fragmen___________________ 417
73
I44 432 432
FIG. 4. Mass spectrum taken from peak 9 in Fig. 1
598
MIRAL
DIZDAROGLU
SO 354
73
I
I
368 365
I .*t-~-;lz;k,,,=i=~~~~~~=~~~~!i~~--~-~~~i4-~~. 50 100 150 200m,= 250 FE. 5. Mass spectrum taken from peak 10 in Fig. 1.
tation and clearly indicates the presence of an OTMS group at C-5. Its intensity (-4% relative intensity) is not as high as that of the coresponding ions from TMS derivatives of uracil, thymine, or cytosine because the presence of a saturated 5,6-double bond in the pyrimidine nucleus supresses the formation of this ion (17). The prominent ion at m/z 116 (-44% relative intensity) apparently results from a cleavage across the pyrimidine ring and contains carbons 5 and 6
Peak 9 in Fig. 1 was attributed to the TMS derivative of 5,6-dihydroxyuracil (dialuric or isodialuric acid). Isodialuric acid was synthesized and used as authentic material. The TMS derivative of isodialuric acid is known to be identical to that of dialuric acid because of enolization (33). In the mass spectrum of the TMS derivative of 5,6dihydrouracil (Fig. 4), intense M+’ and (M-CH$ ions were found at m/z 432 and 417 (base peak), respectively. The three uracil derivatives discussed above, i.e., 5-hydroxyuracil, 5-hydroxy-5,6dihydrouracil, and 5,6-dihydroxyuracil, are believed to be derived from some radiationinduced products of the cytosine moiety by acid treatment. Deamination of 5-hydroxycytosine and 5,6dihydroxycytosine, which are known to be radiation-induced products of cytosine (6), might have given rise to formation of 5-hydroxyuracil and 5,6-dihy-
(TMS &=~H--;H~), confirming dihydropyrimidine structure with an OTMS group. The ion at m/z 158 (-8% relative intensity) is equal in mass to the doubly charged ion (M-30)*+, which is characteristic of TMS derivatives of pyrimidines and useful to confirm the molecular ion (17). The ion at m/z 257 corresponds to the loss of 6TMS from M+‘.
NHTMS
352
367 NW 367
73
..+-i-,li=4.-.,L.,-“=-~i=:~=:-~~~~-~~-=~~~-i=-~~-~~-~~-~--,--~-~-~--~ I
50
100
150
200
ml2
250
----,
300
350
FIG. 6. Mass spectrum taken from peak I 1 in Fig. 1.
400
GAS CHROMATOGRAPHY-MASS
SPECTROMETRY
OF DNA RADIATION
DAMAGE
599
FIG. 7. Mass spectrum taken from peak 12 in Fig. 1.
droxyuracil, respectively. It should also be pointed out that formic acid-induced formation of the three compounds represented by peaks 4, 5, and 9 in Fig. 1 cannot be excluded because these compounds were not identified in y-irradiated DNA without acid hydrolysis. The mass spectrum taken from peak 10 in Fig. 1 is shown in Fig. 5. It was assigned to the TMS derivative of 4,6-diamino-5formamidopyrimidine. M+‘, (M-H)‘, and (M-CH# ions were observed at m/z 369, 368, and 354, respectively. The base peak at m/z 280 was probably due to elimination of 6TMS (89 amu) from M+‘. Peak 11 in Fig. 1 gave the mass spectrum shown in Fig. 6. It was assigned to the TMS derivative of 8hydroxyadenine.’ Intense M” and (M-CH,)+ ions were present at m/z 367 and 352 (base peak), respectively. Peak 12 in Fig. 1 was attributed to the TMS derivatives of 2,6diamino-4-hydroxy5-formamidopyrimidine.8 Abundant M+’ and (M-CH$ ions were observed in its mass spectrum (Fig. 7). The intense ion at m/z 368 was presumably formed by elimination of GTMS (89 amu) from M+‘. The mass spectrum from peak 13 in Fig. I is shown in Fig. 8. Intense M+’ and (M’ This compound was also named 7,8-dihydro-8-oxoadenine [see Ref. (7)]. a In Ref. (6), this compound was named 2,4-diamino5-formamido&hydroxypyrimidine.
CH$ ions were observed at m/z 455 and 440 (base peak), respectively. This spectrum was assigned to the TMS derivative of 8hydroxyguanine, which was recently identified in y-irradiated DNA.5 Use of t-BDMS derivatives. t-BDMS derivatives of nucleic acid bases were shown to be very useful for confirming molecular weight and for GC-MS with selected-ion monitoring (SIM) by providing an intense (M-57)’ ion in their mass spectra (29,34). The complementary use of these derivatives, which are also prepared in a simple procedure and have excellent GC properties (29,34), may greatly contribute to identification of base products of irradiated DNA. Recently, the use of tBDMS derivatives for the GC-MS analysis of the monomeric and dimeric radiationinduced products of a DNA base, i.e., thymine, was demonstrated (25). Here, as an example, the mass spectra of the t-BDMS derivatives of 8-hydroxyadenine and 8-hydroxyguanine are shown in Figs. 9 and 10, respectively. In both cases, the characteristic (M-57)+ ion appeared as the base peak at m/z 436 and 566, respectively. M+’ and (MCH&+ ions were also observed (m/z 493 and 478 in Fig. 9, and m/z 623 and 608 in Fig. 10). Selected-ion monitoring. GC-MS with the SIM technique is a useful means for identification of compounds in a complex mixture, based on recording of an intense characteristic ion in their mass spectra. TMS derivatives
MIRAL
DIZDAROGLU
50
0 FIG. 8. Mass spectrum taken from peak 13 in Fig. 1.
of the radiation-induced base products of DNA discussed in the present paper provided intense M+’ and (M-CH,)+ ions in their mass spectra, which can be usefully employed for GC-MS/SIM analysis. Figure 11 illustrates the use of this technique for some of the products described in this paper, where (MCH$ ions of the TMS derivatives of 5hydroxycytosine (m/z 328),9 5-hydroxyuracil (m/z 329; see also Fig. 2), 5-hydroxy-5,6dihydrouracil (m/z 331; see also Fig. 3), 5hydroxy-5,6-dihydrothymine (m/z 345),9 8hydroxyadenine (m/z 352; see also Fig. 6), 5,6-dihydroxyuracil (m/z 417; see also Fig. 4) thymine glycol (m/z 433),9 and 8-hydroxyguanine (m/z 440; see also Fig. 8) were monitored. Peaks indicated by arrows show the presence of the products listed above in a sample of a trimethylsilylated HCOOH hydrolysate of irradiated DNA. In some instances, more than one peak for the ion monitored was obtained, due to the presence of that particular ion in the mass spectrum of another compound in the sample. For instance, the m/z 331 ion is also present in the mass spectrum of TMS 5hydroxyuracil (Fig. 2) as an isotope peak of the (M-CH$ ion (m/z 329). Similarly, the m/z 345 ion, which was monitored as the (M-CH$ ion of TMS 5-hydroxy-5,6dihy-
9 For the mass spectra of these compounds see Ref. (25).
drothymine, is found in the mass spectrum of TMS 5-hydroxyuracil (Fig. 2) as an isotope peak of M+ (m/z 344). On the other hand, the m/z 352 ion recorded as the (M-CH# ion of TMS 8-hydroxyadenine is also the (M-CH$ ion of TMS guanine [for the mass spectrum see Ref. (17)], to which the large peak between 15 and 16 min in Fig. 11 corresponds. In another instance, the presence of the ions m/z 329, 331, and 345 is clearly seen in the mass spectrum of TMS 5-hydroxycytosine (peaks between 9 and 10 min in Fig. 1 1), where these peaks constitute the isotope peaks of either (M-CH# or M” ions [for the mass spectrum see Ref. (25)]. These results clearly show that the SIM technique can be usefully applied to monitor the radiation-induced products of DNA in a complex mixture. The mass spectra then taken from the corresponding GC peaks would unequivocally reveal their presence. The intense characteristic (M-57)+ ion of their t-BDMS derivatives can also be usefully applied for the same purpose. Sensitivity. It is well known that the GCMS technique provides a high degree of sensitivity for detection of organic compounds. In the present work, amounts as low as 0.1 pmol per injection of standard compounds such as thymine, cytosine, adenine, and guanine were found to be detectable. This finding about the sensitivity of this method, however, provides no information about the lowest level of detection of damaged
GAS CHROMATOGRAPHY-MASS
SPECTROMETRY
OF DNA RADIATION
601
DAMAGE
___--_-_________________________________~~~~.~~~~~-~~---------~-~--~-~-~~~-~~~~~~~~~~~~~~~~~~~~~~~~~~~ NHR
436
R = t-BDMS
2 1 s c(
R 73
iw
493
a 478
50
100
150
200
250
m/z 300
350
400
450
493
500
550
FIG. 9. Mass spectrum of the t-BDMS derivative of 8-hydroxyadenine.
bases in DNA when one works with cells. In other words, the questions concerning how much DNA, how many cells, or how much radiation the lowest level of sensitivity of this method translates remain to be answered. CONCLUSIONS
Capillary GC-MS was shown to be an excellent analytical technique for identification of radiation-induced base products of DNA. It was possible to show the presence of the products reported previously. TMS and t-BDMS derivatives of the radiationinduced products provide sharp and symmetrical GC peaks and easily interpretable mass spectra. The presence of intense M+’ and (M-CH$ ions in the mass spectra of their TMS derivatives and an intense (M57)+ ion in those of their t-BDMS derivatives provides a highly useful means for structural elucidation. Moreover, because of the high
intensity of these ions in the mass spectra, they can be used to monitor the corresponding compounds in a complex mixture by GC-MS with SIM. Hydrolysis of irradiated DNA with 88% HCOOH at 150°C for 30 min permits simultaneous identification of the radiation-induced products of all four DNA bases in a single run. Because of the simple preparation of TMS and t-BDMS derivatives, irradiated DNA can be hydrolyzed and derivatized in a single tube, without any further sample manipulation. The overall assay time (hydrolysis, derivatization, GCMS) is less then 2 h. Moreover, the GC-MS technique provides a high degree of sensitivity for identification. Quantitative analysis of the products is also possible by addition of an internal standard to the irradiated DNA prior to acid hydrolysis. There are base exision-repair enzymes, such as DNA glycosylases, which participate in DNA repair by catalyzing the release of
R =t-BDMS ___
5866
MW 623 623 608 ._-- ..?.v-;-T-I 600
FIG.
10. Mass spectrum of the T-BDMS derivative of 8-hydroxyguanine.
,50
602
MIRAL
DIZDAROGLU Y
de 440 mlz 433 mlr 417 mlr 352 m,r 345 J
1
mlz 331 mlr 329
r’
m/r 328 2
3
4
I 5
I, 6
7
8
9
I 10 11 TI”E hi”)
12
13
14
1 15
’ 16
17
18
FIG. 1 I. SIM of a trimethylsilylated HCOOH hydrolysate of irradiated calf thymus DNA. GC conditions were as in Fig. I.
abnormal bases from DNA in free form (35,36). Lindahl(36) pointed out that, in the past, it had often been difficult to observe the removal of a specific altered base from DNA in the presence of several other forms of damage. The capillary GC-MS method described here can be usefully applied (without the acid hydrolysis, of course) to the separation and identification of abnormal DNA bases removed by DNA glycosylases. In this respect, this technique may contribute a great deal to an understanding of the base excision-repair of DNA. ACKNOWLEDGMENTS This work was conducted pursuant to a contract with the National Foundation for Cancer Research, Bethesda, Maryland. The author acknowledges the continuous support of Dr. Michael G. Simic (National Bureau of Standards). REFERENCES 1. Ward, J. F. (1975) Adv. Radiat. Biol. 5, 181-239. 2. Smith, K. C. (1976) in Aging, Carcinogenesis and Radiation Biology (Smith, K. C., ed.), pp. 67-81, Plenum, New York/London.
3. Mee, L. K. and Adelstein, S. J. (1981) Proc. Natl. Acad. Sci. U. S. A. 78, 2194-2198. 4. Lehmann, A. R. (1978) in Effects of Ionizing Radiation on DNA (Hiittermann, J., Kiihnlein, W., and Teoule, R., eds.), pp. 312-334, SpringerVerlag, Berlin/Heidelberg/New York. 5. Painter, R..B. (1980) in Radiation Biology in Cancer Research (Meyn, R. E. and Withers, H. R., eds.), pp. 59-68, Raven Press, New York. 6. Teoule, R. and Cadet, J. (1978) in Effects of Ionizing Radiation on DNA (Hiittermann, J., Kahnlein, W., and Teoule, R., eds.), pp. 17 I-203, SpringerVerlag, Berlin/Heidelberg/New York. 7. Teoule, R. (1979) in Proceedings of the Sixth Congress of Radiation Research (Okada, S., Imamura, M., Terasima, T., and Yamagushi, H., eds.), pp. 400-407, Tappan Printing Co., Tokyo. 8. Cadet, J., Berger, M., and Voituriez, L. (1982) J. Chromatogr. 238, 488-494. 9. Teoule, R., Bonicel, A., Mariaggi, N., and Polverelli, M. (1983) in Proceedings of the Seventh International Congress of Radiation Research (Broerse, J. J., Barendsen, G. W., Kal, H. B., and van der Kogel, A. J., eds.), pp. A3-43, Nijhoff, Amsterdam. 10. Frenkel, K., Goldstein, M. S., and Teebor, G. W. (1981) Biochemistry 20, 7566-7571. 11. Dizdaroglu, M., von Sonntag, C., and Schulte-Frohlinde, D. (1975). J. Amer. Chem. Sot. 97, 22772278. 12. Dizdaroglu, M., Schulte-Frohlinde, D., and von Sonntag, C. (1975) Z. Naturforsch. 3Oq 826-828.
GAS CHROMATOGRAPHY-MASS
SPECTROMETRY
13. Dizdaroght, M., Neuwald, K., and von Sonntag, C. (1976) Z. Naturforsch. 31b, 227-233. 14. Dizdaroglu, M., Schulte-Frohlinde, D., and von Sonntag, C. (1977) Int. J. Radiat. Biol. 32, 481483. 15. Dizdaroglu, M., Schulte-Frohlinde, D., and von Sonntag, C. (1977) Z. Naturjbrsch. 32c, 10211022. 16. Beesk, F., Dizdaroglu, M., Schulte-Frohlinde, D., and von Sonntag, C. (1979) Int. J. Radiat. Biol. 36, 565-576. 17. White, E., Krueger, P. M., and McCloskey, J. A. (1972) J. Org. Chem. 37, 430-438. 18. McCloskey, J. A. (1974) in Basic Principles in Nucleic Acid Chemistry (Ts’o, P. 0. P., ed.), pp. 209309, Academic Press, New York. 19. Hignite, C. (1980) in Biochemical Applications of Mass Spectrometry (Waller, G. R. and Dermer, 0. C., eds.), pp. 527-566, Wiley, New York. 20. Gelijkens, C. F., Smith, D. L., and McCloskey, J. A. (198 1) J. Chromatogr. 225, 291-299. 2 1. Sasaki, Y., and Hashizuma, T. (1966) Anal. B&hem. 16, l-19. 22. Pang, H., Schram, K. H., Smith, D. L., Gupta, S. P., Towsend, L. B., and McCloskey, J. A. (1982) J. Org. Chem. 47, 3923-3932. 23. Dizdaroglu, M., and Simic, M. G. (1984) Int. J. Radiat. Biol. 46, 241-246.
OF DNA RADIATION
DAMAGE
603
24. Dizdaroglu, M., and Simic, M. G. (1984) Radiat. Rex 100,41-46. 25. Dizdaroglu, M. (1984) J. Chromatogr. 295, 103121. 26. Dizdaroglu, M., and Simic, M. G. (1985) Int. J. Radiat. Biol., in press. 27. Simic, M. G., and Dizdaroglu, M. (1985) Biochemistry, in press. 28. Behrend, R., and Roosen, 0. (1889) Liebig’s Ann. Chem. 251,235-238. 29. Crain, P. R., and McCloskey, J. A. (1983) Anal. Biochem. 132, 124-131. 30. Adams, R. L. P., Burdon, R. H., Campbell, A. M., Leader, D. P., and Smellie, R. M. S. (198 1) in The Biochemistry of the Nucleic Acids, p. 33, Chapman & Hall, London/New York. 3 1. Fujita, S., and Steenken, S. (1981) J. Amer. Chem. Sot. 103, 2540-2545.
32. Hazra, D. K., and Steenken, S. (1983) J. Amer. Chem. Sot. 105,4380-4386. 33. Schuchmann, M. N., and von Sonntag, C. (1983) J. Chem. Sot. Perkin Trans. 2, 1525-153 1. 34. Quilliam, M. A., and Westmore, J. B. (1978) Anal. Chem. 50, 59-68. 35. Lindahl, T. (1979) in Nucleic Acid Research and Molecular Biology (Cohn, W. E., ed.), pp. 135192, Vol. 22, Academic Press, New York. 36. Lindahl, T. (1982) Annu. Rev. Biochem. 51, 61-87.
BIOCHEMISTRY
14,593-603
(1985)
Application of Capillary Gas Chromatography-Mass Spectrometry to Chemical Characterization of Radiation-Induced Base Damage of DNA: Implications for Assessing DNA Repair Processes MIRAL DIZDAR~GLU Center for Radiation
Research,
National
Bureau
of Standards,
Gaithersburg,
Maryland
20899
Received April 17, 1984 The application of capillary gas chromatography-mass spectrometty (CC-MS) to the chemical characterization of radiation-induced base products of calf thymus DNA is presented. Samples of calf thymus DNA irradiated in N@-saturated aqueous solution were hydrolyzed with HCOOH, trimethylsilylated, and subjected to CC-MS analysis using a fused-silica capillary column. Hydrolysis conditions suitable for the simultaneous analysis of the radiation-induced products of all four DNA bases in a single run were determined. The trimethylsilyl derivatives of these products had excellent GC properties and easily interpretable mass spectra; an intense molecular ion (MC) and a characteristic (M-CH,)+ ion were observed. The complementary use of t-butyldimethylsilyl derivatives was also demonstrated. These derivatives provided an intense characteristic (M-57)+ ion, which appeared as either the base peak or the second most intense ion in the spectra. All mass spectra obtained are discussed. Because of the excellent resolving power of capillary GC and the accurate high-sensitivity identification by MS, the capillary CC-MS is suggested as a very suitable technique for identification of altered bases removed from DNA by base excision-repair enzymes such as DNA glycosylasesand, thus, as very useful for an understanding of the base excision-repair of DNA. o 1985 Academic RS.S, IW. KEY WORDS: DNA damage; ionizing radiation; DNA repair enzymes; mass spectra; trimethylsilylation; t-BDMS derivatives; high sensitivity.
Ionizing radiation creates a number of lesions in DNA, including modified bases, base-free sites, and single- and double-strand breaks (1). DNA-protein crosslinks are also formed when cells or mixtures of DNA and proteins are exposed to ionizing radiation (2,3). In living cells, lesions produced in DNA are subject to repair processes [for a review see Ref. (4)]. Unless repaired, such damages may have detrimental biological consequences, such as cell death, mutagenesis, or carcinogenesis (5). Elucidation of the chemical nature of radiation-induced DNA lesions is, therefore, necessary for an understanding of their biological consequences and enzymatic repairability. The effects of ionizing radiation on DNA and its constituents have been extensively studied under various conditions using a 593
variety of analytical techniques. For instance, thin-layer chromatography, high-performance liquid chromatography (HPLC), nuclear magnetic resonance spectroscopy, and mass spectrometry (MS)’ were used for identification of radiation-induced products of DNA bases (6- lo), and capillary gas chromatography (GC) combined with MS (GC-MS) was extensively applied to characterization of radiation-induced products of the DNA sugar moiety (11-16). In the past, GC-MS has played an important role in the characterization of natural ’ Abbreviations used: CC-MS, gas chromatographymass spectrometry; TMS, trimethylsilyl; BSTFA, bis(trimethylsilyl)trifluoroacetamide; t-BDMS, t-butyldimethylsilyl; MC, molecular ion; SIM, selected-ion monitoring; ‘H NMR, proton nuclear magnetic resonance spectroscopy. 0003-2697185 $3.00 Copyright Q 1985 by Academic Press. Inc. ,411 rigbu of reproduction in any form rmwed.
MIRAL
DIZDARGGLU
FIG. 1. GC separation of a trimethylsilylated HCOOH hydrolysate of irradiated calf thymus DNA. Dose: 330 Gy. Column: fused silica capillary (12 m, 0.2-mm i.d.) coated with crosslinked SE-54, programmed after 3 min at 100°C from 100 to 250°C with a rate of 7”C/min. For other details, see Materials and Methods. Peak identifications are given in Table 1.
and synthetic nucleic acid bases and nucleosides (17-22). However, this technique has not yet been applied to identification of radiation-induced base damage of DNA. Recently, we made extensive use of capillary GC-MS for isolation and characterization of radiation-induced products of DNA bases such as thymine and cytosine and their nucleosides and nucleotides (23-25), as well as of radiation-induced crosslinks between a DNA base, i.e., thymine, and aromatic amino acids such as phenylalanine (26) and tyrosine (27). In all instances, trimethylsilyl (TMS) derivatives of monomeric and dimeric products were shown to have excellent GC properties and easily interpretable mass spectra.* The present paper describes the application of capillary GC-MS to identification of * For details such as separation conditions, sensitivity, use of capillary columns with different coatings, interpretation of mass spectra obtained, and comparison of different derivatization methods, see Ref. (25).
radiation-induced mus DNA. MATERIALS
base products of calf thyAND METHODS3
Chemicals. Calf thymus DNA and bis(trimethylsilyl)trifluoroacetamide (BSTFA), 5methylcytosine, and 5-hydroxyuracil (isobarbituric acid) were purchased from Sigma. Acetonitrile and pyridine were obtained from Pierce. N-Methyl-N-(t-butyldimethylsilyl)trifluoroacetamide was from Regis. Isodialuric acid (5,6-dihydroxyuracil) was synthesized by oxidation of isobarbituric acid with bromine (28). Irradiations. Aqueous solutions of calf thymus DNA (250 mg/liter) were saturated 3 Certain commercial equipment, instruments, or materials are identified in this paper in order to specify adequately the experimental procedure. Such identification does not imply recommendation or endorsement by the National Bureau of Standards, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.
GAS CHROMATOGRAPHY-MASS
SPECTROMETRY TABLE
PEAK IDENTETCATION
Peak No.’ I II
2 III d 3 4 5 IIIa 6
Compoundb
OF DNA RADIATION
595
1 IN FIGURE
Peak No.“
Phosphoric acid Uracil Thymine 5,6-Dihydrothymine
7, 8 9 IV 10
Cytosine 5-Methylcytosine 5-Hydroxy-5,6dihydrothymine
IVa 11 12
5-Hydroxyuracil 5-Hydroxy-5,6dihydrouracil Cytosine’ 5-Hydroxycytosine
DAMAGE
V Va 13
1
Compound b cis- and trans-thymine glycol 5,6Dihydroxyuracil Adenine 4,6-Diamino-5-formamidopyrimidine Adenine’ I-Hydroxyadenine 2,6-Diamino-4-hydroxy-5formamidopyrimidine Guanine GuanineC 8-Hydroxyguanine
n Peaks a to g were also present in samples of nonirradiated calf thymus DNA, which were treated in the same manner as the irradiated samples. Their identities were not determined, except for peak d, which represents the TMS derivative of 5-methylcytosine. b As their TMS derivatives. c With an additional TMS group attached to the amino group.
with oxygen-free N20 and irradiated in a 6oC y-source. (dose range 1lo-440 Gy; dose rate 110 Gy/min).4 Samples were then lyophilized. Hydrolysis with HCOOH. About 2 mg of lyophilized samples was hydrolyzed with 1 ml of HCOOH (88%) at 150°C for 30 min in evacuated and sealed tubes. After hydrolysis, the samples were dried in a desiccator in vacua prior to derivatization. Derivatization. Hydrolyzed and dried samples were trimethylsilylated in PTFE-capped Hypo-vials (Pierce) with 0.2 ml of a mixture of BSTFA and acetonitrile (1: 1) by heating for 30 min at 140°C. t-Butyldimethylsilyl (tBDMS) derivatives were prepared according to Crain and McCloskey (29). Gas chromatography. A Hewlett-Packard Model 5880A microprocessor-controlled gas chromatograph equipped with a flame ionization detector was used. The injection port and detector were maintained at 250°C. Separations were achieved on a fused-silica capillary column ( 12 m, 0.2-mm i.d.) coated 4 1 Gray = 1 joule/kg = 100 rad.
with crosslinked SE-54 [5% phenylmethylsilicone; wall coated open tubular; siloxane deactivated; film thickness, 0.11 pm (Hewlett-Packard)]. The measured efficiency of the column was ca. 5400 theoretical plates per meter based on the pentadecane peak at 120°C (capacity factor, 6.42; linear velocity of the carrier gas, 32.8 cm/s). Helium was used as the carrier gas at an inlet pressure of 100 kPa. The split ratio was 20: 1. Gas chromatography-mass spectrometry.
Mass spectra were taken at 70 eV using a Hewlett-Packard Model 5970A mass selective detector interfaced to the above gas chromatograph. GC conditions were as above. The ion source temperature was ca. 200°C. RESULTS AND DISCUSSION Acid hydrolysis. When DNA is irradiated, free bases and some degradation products are released from DNA; however, the modified bases remain attached mostly to the polynucleotide chain (6,7,9). It is therefore necessary to remove them from the DNA backbone prior to their chemical analysis. In the
596
MIRAL
DIZDAROGLU
urm N/ iT a 52 A d
73
TMSo
ks
329 mMs
N
Mu 344
343
147
FIG. 2. Mass spectrum taken from peak 4 in Fig. I.
past, formic or perchloric acid hydrolysis was used to determine the extent of the radiationinduced base modification of DNA (6). Recently, Teoule et al. (9) described the optimal conditions for the removal, by acid hydrolysis, of radiation-induced products of thymine, cytosine, and adenine from irradiated DNA. Different conditions and acids, i.e., HCOOH and HF, were used for different bases. Products of guanine were not discussed. In the present paper, hydrolysis with 88% HCOOH at 150°C for 30 min was found to be suitable, as a result of the studies of concentration and temperature, for simultaneous identification of radiation-induced products of all four bases of DNA in a single run. GC-MS anafyszs. Figure 1 shows the GC separation of a trimethylsilylated HCOOH hydrolysate of irradiated calf thymus DNA. Peak identification is given in Table 1. Peaks I, II, III, IV, and V represent the TMS derivatives of phosphoric acid, thymine, cytosine, adenine, and guanine, respectively. Peaks IIIa, IVa, and Va also represent the last three bases, but with an additional TMS group attached to their amino groups as revealed by comparison with authentic material and by MS. Peaks a to g were also present in the gas chromatogram of nonirradiated HCOOH-hydrolyzed DNA samples. Their origin could not be determined, with the exception of peak d, which corresponds to the TMS derivative of 5-methylcytosine, for which authentic material was available. Calf thymus DNA is known to have a 5-
methylcytosine content of about 1.3% (30). Peaks 1 to 13 represent the TMS derivatives of the radiation-induced base products of DNA. Products represented by peaks 1 to 3, 6 to 8, and 10 to 12 (Table 1) were reported in previous studies using analytical techniques other than CC-MS (6,7). 8-Hydroxyguanine represented by peak 13 in Fig. 1 has recently been identified in y-irradiated DNA as 8hydroxy-2’deoxyguanosine alter enzymatic hydrolysis using CC-MS, HPLC, uv spectroscopy, and ‘H NMR.’ The presence of 8hydroxy-2’deoxyguanosine in y-irradiated 2’deoxyguanosine and 2’deoxyguanosine 5’monophosphate was also shown using the same techniques without the acid hydrolysis5 The acid hydrolysis of 8-hydroxy-2’deoxyguanosine yielded 8-hydroxyguanine.5 The remaining products (peaks 4, 5, and 9) have not been previously described.6 Mass spectra of the TMS derivatives of 5,6dihydrothymine (peak 2), 5-hydroxy-5,6dihydrothymine (peak 3) 5-hydroxycytosine (peak 6), and thymine glycol (peaks 7 and 8) were published in a previous paper (25) and will not be discussed here. A mass spectrum taken from peak 4 is given in Fig. 2. From this mass spectrum, peak 4 was assigned as the TMS derivative of 5-hydroxyuracil (isobarbituric acid), for ’ M. Dizdaroglu, submitted for publication. 6 SHydroxyuracil (isobarbituric acid) and 5,6dihydroxyuracil (as isodialuric acid) were observed in the presence of oxygen [see Ref. (9)].
GAS CHROMATOGRAPHY-MASS
SPECTROMETRY
OF DNA RADIATION
DAMAGE
597
331
116
346
q-T--
400
FIG. 3. Mass spectrum taken from peak 5 in Fig. 1.
which authentic material was commercially available. The molecular ion (M+*) and a characteristic (M-H)+ ion were observed at m/z 344 and 343, respectively, and the (MCH$ ion appeared as the base peak at m/z 329; loss of a hydrogen or methyl radical from M+’ of TMS derivatives of nucleic acid bases constitutes their major fragmentation pathways (17-19). Ions at m/z 73 and 147 are commonly observed with TMS derivatives and serve no diagnostic purpose. Figure 3 shows the mass spectrum taken from peak 5 in Fig. 1. M+’ and (M-CH$ ions were observed at m/z 346 and 33 1, respectively. No authentic material was available for this compound. On the other hand, M+’ and (M-CHj)+ ions clearly reveal the addition of an OH group and H atom to the uracil skeleton. Based on the well-known reaction mechanisms of OH radicals with pyrimidines (31,32), this compound can be attributed to the TMS derivative of 5- or 6hydroxy-5,6dihydrouracil. These two com-
pounds should give similar mass spectra as derived from the known mass spectra of TMS derivatives of corresponding thymine derivatives (25). 6-Hydroxy-5,6-dihydropyrimidines, however, are well known to be unstable in hot acid solutions (6). Thus, peak 5 was assigned as the TMS derivative of 5hydroxy-5,6-dihydrouracil. In addition, characteristic fragment ions in Fig. 3 clearly confirm this assignment as explained below: for instance, one of the most structurally diagnostic ions in mass spectra of TMS pyrimidines results from opening of the pyrimidine ring after formation of the (M-CHj)+ ion (17-19). As measurement of exact mass and deuterium labeling clearly showed (17) the resulting product contains carbons 4 and 5 and their substituents and is useful in distinguishing modification at C-5 versus C6. The ion at m/z 188 +
4
5
[(CH&Si=O-m--OTMS] most probably arises from such a fragmen___________________ 417
73
I44 432 432
FIG. 4. Mass spectrum taken from peak 9 in Fig. 1
598
MIRAL
DIZDAROGLU
SO 354
73
I
I
368 365
I .*t-~-;lz;k,,,=i=~~~~~~=~~~~!i~~--~-~~~i4-~~. 50 100 150 200m,= 250 FE. 5. Mass spectrum taken from peak 10 in Fig. 1.
tation and clearly indicates the presence of an OTMS group at C-5. Its intensity (-4% relative intensity) is not as high as that of the coresponding ions from TMS derivatives of uracil, thymine, or cytosine because the presence of a saturated 5,6-double bond in the pyrimidine nucleus supresses the formation of this ion (17). The prominent ion at m/z 116 (-44% relative intensity) apparently results from a cleavage across the pyrimidine ring and contains carbons 5 and 6
Peak 9 in Fig. 1 was attributed to the TMS derivative of 5,6-dihydroxyuracil (dialuric or isodialuric acid). Isodialuric acid was synthesized and used as authentic material. The TMS derivative of isodialuric acid is known to be identical to that of dialuric acid because of enolization (33). In the mass spectrum of the TMS derivative of 5,6dihydrouracil (Fig. 4), intense M+’ and (M-CH$ ions were found at m/z 432 and 417 (base peak), respectively. The three uracil derivatives discussed above, i.e., 5-hydroxyuracil, 5-hydroxy-5,6dihydrouracil, and 5,6-dihydroxyuracil, are believed to be derived from some radiationinduced products of the cytosine moiety by acid treatment. Deamination of 5-hydroxycytosine and 5,6dihydroxycytosine, which are known to be radiation-induced products of cytosine (6), might have given rise to formation of 5-hydroxyuracil and 5,6-dihy-
(TMS &=~H--;H~), confirming dihydropyrimidine structure with an OTMS group. The ion at m/z 158 (-8% relative intensity) is equal in mass to the doubly charged ion (M-30)*+, which is characteristic of TMS derivatives of pyrimidines and useful to confirm the molecular ion (17). The ion at m/z 257 corresponds to the loss of 6TMS from M+‘.
NHTMS
352
367 NW 367
73
..+-i-,li=4.-.,L.,-“=-~i=:~=:-~~~~-~~-=~~~-i=-~~-~~-~~-~--,--~-~-~--~ I
50
100
150
200
ml2
250
----,
300
350
FIG. 6. Mass spectrum taken from peak I 1 in Fig. 1.
400
GAS CHROMATOGRAPHY-MASS
SPECTROMETRY
OF DNA RADIATION
DAMAGE
599
FIG. 7. Mass spectrum taken from peak 12 in Fig. 1.
droxyuracil, respectively. It should also be pointed out that formic acid-induced formation of the three compounds represented by peaks 4, 5, and 9 in Fig. 1 cannot be excluded because these compounds were not identified in y-irradiated DNA without acid hydrolysis. The mass spectrum taken from peak 10 in Fig. 1 is shown in Fig. 5. It was assigned to the TMS derivative of 4,6-diamino-5formamidopyrimidine. M+‘, (M-H)‘, and (M-CH# ions were observed at m/z 369, 368, and 354, respectively. The base peak at m/z 280 was probably due to elimination of 6TMS (89 amu) from M+‘. Peak 11 in Fig. 1 gave the mass spectrum shown in Fig. 6. It was assigned to the TMS derivative of 8hydroxyadenine.’ Intense M” and (M-CH,)+ ions were present at m/z 367 and 352 (base peak), respectively. Peak 12 in Fig. 1 was attributed to the TMS derivatives of 2,6diamino-4-hydroxy5-formamidopyrimidine.8 Abundant M+’ and (M-CH$ ions were observed in its mass spectrum (Fig. 7). The intense ion at m/z 368 was presumably formed by elimination of GTMS (89 amu) from M+‘. The mass spectrum from peak 13 in Fig. I is shown in Fig. 8. Intense M+’ and (M’ This compound was also named 7,8-dihydro-8-oxoadenine [see Ref. (7)]. a In Ref. (6), this compound was named 2,4-diamino5-formamido&hydroxypyrimidine.
CH$ ions were observed at m/z 455 and 440 (base peak), respectively. This spectrum was assigned to the TMS derivative of 8hydroxyguanine, which was recently identified in y-irradiated DNA.5 Use of t-BDMS derivatives. t-BDMS derivatives of nucleic acid bases were shown to be very useful for confirming molecular weight and for GC-MS with selected-ion monitoring (SIM) by providing an intense (M-57)’ ion in their mass spectra (29,34). The complementary use of these derivatives, which are also prepared in a simple procedure and have excellent GC properties (29,34), may greatly contribute to identification of base products of irradiated DNA. Recently, the use of tBDMS derivatives for the GC-MS analysis of the monomeric and dimeric radiationinduced products of a DNA base, i.e., thymine, was demonstrated (25). Here, as an example, the mass spectra of the t-BDMS derivatives of 8-hydroxyadenine and 8-hydroxyguanine are shown in Figs. 9 and 10, respectively. In both cases, the characteristic (M-57)+ ion appeared as the base peak at m/z 436 and 566, respectively. M+’ and (MCH&+ ions were also observed (m/z 493 and 478 in Fig. 9, and m/z 623 and 608 in Fig. 10). Selected-ion monitoring. GC-MS with the SIM technique is a useful means for identification of compounds in a complex mixture, based on recording of an intense characteristic ion in their mass spectra. TMS derivatives
MIRAL
DIZDAROGLU
50
0 FIG. 8. Mass spectrum taken from peak 13 in Fig. 1.
of the radiation-induced base products of DNA discussed in the present paper provided intense M+’ and (M-CH,)+ ions in their mass spectra, which can be usefully employed for GC-MS/SIM analysis. Figure 11 illustrates the use of this technique for some of the products described in this paper, where (MCH$ ions of the TMS derivatives of 5hydroxycytosine (m/z 328),9 5-hydroxyuracil (m/z 329; see also Fig. 2), 5-hydroxy-5,6dihydrouracil (m/z 331; see also Fig. 3), 5hydroxy-5,6-dihydrothymine (m/z 345),9 8hydroxyadenine (m/z 352; see also Fig. 6), 5,6-dihydroxyuracil (m/z 417; see also Fig. 4) thymine glycol (m/z 433),9 and 8-hydroxyguanine (m/z 440; see also Fig. 8) were monitored. Peaks indicated by arrows show the presence of the products listed above in a sample of a trimethylsilylated HCOOH hydrolysate of irradiated DNA. In some instances, more than one peak for the ion monitored was obtained, due to the presence of that particular ion in the mass spectrum of another compound in the sample. For instance, the m/z 331 ion is also present in the mass spectrum of TMS 5hydroxyuracil (Fig. 2) as an isotope peak of the (M-CH$ ion (m/z 329). Similarly, the m/z 345 ion, which was monitored as the (M-CH$ ion of TMS 5-hydroxy-5,6dihy-
9 For the mass spectra of these compounds see Ref. (25).
drothymine, is found in the mass spectrum of TMS 5-hydroxyuracil (Fig. 2) as an isotope peak of M+ (m/z 344). On the other hand, the m/z 352 ion recorded as the (M-CH# ion of TMS 8-hydroxyadenine is also the (M-CH$ ion of TMS guanine [for the mass spectrum see Ref. (17)], to which the large peak between 15 and 16 min in Fig. 11 corresponds. In another instance, the presence of the ions m/z 329, 331, and 345 is clearly seen in the mass spectrum of TMS 5-hydroxycytosine (peaks between 9 and 10 min in Fig. 1 1), where these peaks constitute the isotope peaks of either (M-CH# or M” ions [for the mass spectrum see Ref. (25)]. These results clearly show that the SIM technique can be usefully applied to monitor the radiation-induced products of DNA in a complex mixture. The mass spectra then taken from the corresponding GC peaks would unequivocally reveal their presence. The intense characteristic (M-57)+ ion of their t-BDMS derivatives can also be usefully applied for the same purpose. Sensitivity. It is well known that the GCMS technique provides a high degree of sensitivity for detection of organic compounds. In the present work, amounts as low as 0.1 pmol per injection of standard compounds such as thymine, cytosine, adenine, and guanine were found to be detectable. This finding about the sensitivity of this method, however, provides no information about the lowest level of detection of damaged
GAS CHROMATOGRAPHY-MASS
SPECTROMETRY
OF DNA RADIATION
601
DAMAGE
___--_-_________________________________~~~~.~~~~~-~~---------~-~--~-~-~~~-~~~~~~~~~~~~~~~~~~~~~~~~~~~ NHR
436
R = t-BDMS
2 1 s c(
R 73
iw
493
a 478
50
100
150
200
250
m/z 300
350
400
450
493
500
550
FIG. 9. Mass spectrum of the t-BDMS derivative of 8-hydroxyadenine.
bases in DNA when one works with cells. In other words, the questions concerning how much DNA, how many cells, or how much radiation the lowest level of sensitivity of this method translates remain to be answered. CONCLUSIONS
Capillary GC-MS was shown to be an excellent analytical technique for identification of radiation-induced base products of DNA. It was possible to show the presence of the products reported previously. TMS and t-BDMS derivatives of the radiationinduced products provide sharp and symmetrical GC peaks and easily interpretable mass spectra. The presence of intense M+’ and (M-CH$ ions in the mass spectra of their TMS derivatives and an intense (M57)+ ion in those of their t-BDMS derivatives provides a highly useful means for structural elucidation. Moreover, because of the high
intensity of these ions in the mass spectra, they can be used to monitor the corresponding compounds in a complex mixture by GC-MS with SIM. Hydrolysis of irradiated DNA with 88% HCOOH at 150°C for 30 min permits simultaneous identification of the radiation-induced products of all four DNA bases in a single run. Because of the simple preparation of TMS and t-BDMS derivatives, irradiated DNA can be hydrolyzed and derivatized in a single tube, without any further sample manipulation. The overall assay time (hydrolysis, derivatization, GCMS) is less then 2 h. Moreover, the GC-MS technique provides a high degree of sensitivity for identification. Quantitative analysis of the products is also possible by addition of an internal standard to the irradiated DNA prior to acid hydrolysis. There are base exision-repair enzymes, such as DNA glycosylases, which participate in DNA repair by catalyzing the release of
R =t-BDMS ___
5866
MW 623 623 608 ._-- ..?.v-;-T-I 600
FIG.
10. Mass spectrum of the T-BDMS derivative of 8-hydroxyguanine.
,50
602
MIRAL
DIZDAROGLU Y
de 440 mlz 433 mlr 417 mlr 352 m,r 345 J
1
mlz 331 mlr 329
r’
m/r 328 2
3
4
I 5
I, 6
7
8
9
I 10 11 TI”E hi”)
12
13
14
1 15
’ 16
17
18
FIG. 1 I. SIM of a trimethylsilylated HCOOH hydrolysate of irradiated calf thymus DNA. GC conditions were as in Fig. I.
abnormal bases from DNA in free form (35,36). Lindahl(36) pointed out that, in the past, it had often been difficult to observe the removal of a specific altered base from DNA in the presence of several other forms of damage. The capillary GC-MS method described here can be usefully applied (without the acid hydrolysis, of course) to the separation and identification of abnormal DNA bases removed by DNA glycosylases. In this respect, this technique may contribute a great deal to an understanding of the base excision-repair of DNA. ACKNOWLEDGMENTS This work was conducted pursuant to a contract with the National Foundation for Cancer Research, Bethesda, Maryland. The author acknowledges the continuous support of Dr. Michael G. Simic (National Bureau of Standards). REFERENCES 1. Ward, J. F. (1975) Adv. Radiat. Biol. 5, 181-239. 2. Smith, K. C. (1976) in Aging, Carcinogenesis and Radiation Biology (Smith, K. C., ed.), pp. 67-81, Plenum, New York/London.
3. Mee, L. K. and Adelstein, S. J. (1981) Proc. Natl. Acad. Sci. U. S. A. 78, 2194-2198. 4. Lehmann, A. R. (1978) in Effects of Ionizing Radiation on DNA (Hiittermann, J., Kiihnlein, W., and Teoule, R., eds.), pp. 312-334, SpringerVerlag, Berlin/Heidelberg/New York. 5. Painter, R..B. (1980) in Radiation Biology in Cancer Research (Meyn, R. E. and Withers, H. R., eds.), pp. 59-68, Raven Press, New York. 6. Teoule, R. and Cadet, J. (1978) in Effects of Ionizing Radiation on DNA (Hiittermann, J., Kahnlein, W., and Teoule, R., eds.), pp. 17 I-203, SpringerVerlag, Berlin/Heidelberg/New York. 7. Teoule, R. (1979) in Proceedings of the Sixth Congress of Radiation Research (Okada, S., Imamura, M., Terasima, T., and Yamagushi, H., eds.), pp. 400-407, Tappan Printing Co., Tokyo. 8. Cadet, J., Berger, M., and Voituriez, L. (1982) J. Chromatogr. 238, 488-494. 9. Teoule, R., Bonicel, A., Mariaggi, N., and Polverelli, M. (1983) in Proceedings of the Seventh International Congress of Radiation Research (Broerse, J. J., Barendsen, G. W., Kal, H. B., and van der Kogel, A. J., eds.), pp. A3-43, Nijhoff, Amsterdam. 10. Frenkel, K., Goldstein, M. S., and Teebor, G. W. (1981) Biochemistry 20, 7566-7571. 11. Dizdaroglu, M., von Sonntag, C., and Schulte-Frohlinde, D. (1975). J. Amer. Chem. Sot. 97, 22772278. 12. Dizdaroglu, M., Schulte-Frohlinde, D., and von Sonntag, C. (1975) Z. Naturforsch. 3Oq 826-828.
GAS CHROMATOGRAPHY-MASS
SPECTROMETRY
13. Dizdaroght, M., Neuwald, K., and von Sonntag, C. (1976) Z. Naturforsch. 31b, 227-233. 14. Dizdaroglu, M., Schulte-Frohlinde, D., and von Sonntag, C. (1977) Int. J. Radiat. Biol. 32, 481483. 15. Dizdaroglu, M., Schulte-Frohlinde, D., and von Sonntag, C. (1977) Z. Naturjbrsch. 32c, 10211022. 16. Beesk, F., Dizdaroglu, M., Schulte-Frohlinde, D., and von Sonntag, C. (1979) Int. J. Radiat. Biol. 36, 565-576. 17. White, E., Krueger, P. M., and McCloskey, J. A. (1972) J. Org. Chem. 37, 430-438. 18. McCloskey, J. A. (1974) in Basic Principles in Nucleic Acid Chemistry (Ts’o, P. 0. P., ed.), pp. 209309, Academic Press, New York. 19. Hignite, C. (1980) in Biochemical Applications of Mass Spectrometry (Waller, G. R. and Dermer, 0. C., eds.), pp. 527-566, Wiley, New York. 20. Gelijkens, C. F., Smith, D. L., and McCloskey, J. A. (198 1) J. Chromatogr. 225, 291-299. 2 1. Sasaki, Y., and Hashizuma, T. (1966) Anal. B&hem. 16, l-19. 22. Pang, H., Schram, K. H., Smith, D. L., Gupta, S. P., Towsend, L. B., and McCloskey, J. A. (1982) J. Org. Chem. 47, 3923-3932. 23. Dizdaroglu, M., and Simic, M. G. (1984) Int. J. Radiat. Biol. 46, 241-246.
OF DNA RADIATION
DAMAGE
603
24. Dizdaroglu, M., and Simic, M. G. (1984) Radiat. Rex 100,41-46. 25. Dizdaroglu, M. (1984) J. Chromatogr. 295, 103121. 26. Dizdaroglu, M., and Simic, M. G. (1985) Int. J. Radiat. Biol., in press. 27. Simic, M. G., and Dizdaroglu, M. (1985) Biochemistry, in press. 28. Behrend, R., and Roosen, 0. (1889) Liebig’s Ann. Chem. 251,235-238. 29. Crain, P. R., and McCloskey, J. A. (1983) Anal. Biochem. 132, 124-131. 30. Adams, R. L. P., Burdon, R. H., Campbell, A. M., Leader, D. P., and Smellie, R. M. S. (198 1) in The Biochemistry of the Nucleic Acids, p. 33, Chapman & Hall, London/New York. 3 1. Fujita, S., and Steenken, S. (1981) J. Amer. Chem. Sot. 103, 2540-2545.
32. Hazra, D. K., and Steenken, S. (1983) J. Amer. Chem. Sot. 105,4380-4386. 33. Schuchmann, M. N., and von Sonntag, C. (1983) J. Chem. Sot. Perkin Trans. 2, 1525-153 1. 34. Quilliam, M. A., and Westmore, J. B. (1978) Anal. Chem. 50, 59-68. 35. Lindahl, T. (1979) in Nucleic Acid Research and Molecular Biology (Cohn, W. E., ed.), pp. 135192, Vol. 22, Academic Press, New York. 36. Lindahl, T. (1982) Annu. Rev. Biochem. 51, 61-87.