A Stable Isotope Dilution Nanoflow Liquid Chromatography Tandem

Subscriber access provided by READING UNIV

Article

A Stable Isotope Dilution Nanoflow Liquid Chromatography Tandem Mass Spectrometry Assay for the Simultaneous Detection and Quantification of Glyoxal-Induced DNA CrossLinked Adducts in Leukocytes from Diabetic Patients Hauh-Jyun Candy Chen, Ya-Lang Chang, Yi-Chun Teng, Chiung-Fong Hsiao, and Tsai-Shiuan Lin Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b04296 • Publication Date (Web): 25 Nov 2017 Downloaded from http://pubs.acs.org on November 26, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

1

A Stable Isotope Dilution Nanoflow Liquid Chromatography Tandem Mass

2

Spectrometry Assay for the Simultaneous Detection and Quantification of

3

Glyoxal-Induced DNA Cross-Linked Adducts in Leukocytes from Diabetic Patients

4

5

Hauh-Jyun Candy Chen,* Ya-Lang Chang, Yi-Chun Teng, Chiung-Fong Hsiao, and

6

Tsai-Shiuan Lin

7

Department of Chemistry and Biochemistry, National Chung Cheng University, 168

8

University Road, Ming-Hsiung, Chia-Yi 62102, Taiwan

9

*To

whom

correspondence

should

be

addressed.

Phone:

886-5-242-8176.

Fax:

10

886-5-272-1040. E-mail: [email protected].

11

Keywords: cross-links; diabetes; glyoxal; human leukocyte DNA; nanoLC-NSI/MS/MS;

12

Abbreviations:

13

chromatography−nanospray ionization tandem mass spectrometry; H-SRM, highly selected

14

reaction monitoring; SID, stable isotope dilution; T2DM, type 2 diabetes mellitus.

gx,

glyoxal;

nanoLC−NSI/MS/MS,

15

1

ACS Paragon Plus Environment

nanoflow

liquid

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

16

TOC graphics

17

2


Page 2 of 32

Page 3 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


18

Abstract

19

Glyoxal is a bifunctional electrophile capable of cross-linking DNA. Although it is

20

present in foods and from the environment, endogenous formation of glyoxal occurs through

21

metabolism of carbohydrates and oxidation of lipids and nucleic acids. Plasma concentrations

22

of glyoxal are elevated in in diabetes mellitus patients compared to non-diabetics. The most

23

abundant 2′-deoxyribonucleoside adducts cross-linked by glyoxal are dG-gx-dC, dG-gx-dA,

24

and dG-gx-dG. These DNA cross-links can be mutagenic by damaging the integrity of the

25

DNA structure. Herein, we developed a highly sensitive and specific assay for the

26

simultaneous detection and quantification of the dG-gx-dC and dG-gx-dA cross-links based

27

on stable isotope dilution nanoflow liquid chromatography nanospray ionization tandem mass

28

spectrometry (nanoLC-NSI/MS/MS) under the highly selected reaction monitoring mode and

29

using a triple quadrupole mass spectrometer. The entire assay procedures involved addition of

30

the stable isotope standards [15N5]dG-gx-dC and [15N5]dG-gx-dA as internal standards,

31

enzyme hydrolysis to release the cross-links as nucleosides, enrichment by a reversed phase

32

solid-phase extraction column, and nanoLC-NSI/MS/MS analysis. The detection limit is 0.19

33

amol for dG-gx-dC and 0.89 amol for dG-gx-dA, which is 400 and 80 times more sensitive,

34

respectively, than capillary LC−NSI/MS/MS assay of these adducts. The lower limit of

35

quantification was 94 and 90 amol for dG-gx-dC and dG-gx-dA, respectively, which is

36

equivalent to 0.056 and 0.065 adducts in 108 normal nucleotides in 50 µg of DNA. In type 2 3



37

diabetes mellitus (T2DM) patients (n = 38), the levels of dG-gx-dC and dG-gx-dA in

38

leukocyte DNA were 1.94 ± 1.20 and 2.10 ± 1.77 in 108 normal nucleotides, respectively,

39

which were significantly higher than those in non-diabetics (n = 39: 0.83 ± 0.92 and 1.05 ±

40

0.99 in 108 normal nucleotides, respectively). Excluding the factor of smoking, an exogenous

41

source of glyoxal, levels of these two cross-linked adducts were found to be significantly

42

higher in nonsmoking T2DM patients than in nonsmoking control subjects. Furthermore, the

43

levels of dG-gx-dC and dG-gx-dA correlated with HbA1c with statistical significance. To our

44

best knowledge, this is the first report of the identification and quantification of

45

glyoxal-derived cross-linked DNA adducts in human leukocyte DNA and their association

46

with T2DM. This stable isotope dilution nanoLC-NSI/MS/MS assay is highly sensitive and

47

specific and it requires only 50 µg of leukocyte DNA isolated from 2−3 mL of blood to

48

accurately quantify these two cross-linked adducts simultaneously. Our assay thus provides a

49

useful biomarker for the evaluation of glyoxal-derived DNA damage.

4


Page 4 of 32

Page 5 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


50

Introduction

51

Glyoxal (gx) is the most reactive α-dicarbonyl compound generated from glycated

52

proteins in the Maillard reaction.1 It can also be derived from the metabolism of

53

N-nitrosamines, amino acids,2,3 and carbohydrates, and from the oxidation of lipids and

54

nucleic acids.3-5 Being mutagenic in bacteria and mammalian cells, glyoxal is used in the

55

industry, is widely distributed in the environment (including cigarette smoke),6 and is found

56

in foods and beverages.7,8 Serum concentrations of glyoxal are elevated in patients with

57

diabetes, uremia, and peritoneal dialysis.9-12 Coincidentally, diabetic patients have been

58

shown to have an increased risk of various types of cancers.13

59

Glyoxal can react with DNA and proteins forming single adducts as well as cross-links

60

in DNA, in proteins, and between DNA and proteins.14-17 We recently characterized the sites

61

and types of hemoglobin adducts of glyoxal and identified certain glyoxal-modified peptides

62

as a biomarker for diabetes mellitus using the shot-gun proteomic quantitation approach.18

63

The

reactions

of

glyoxal

with

the

2′-deoxyribonucleosides

give

rise

to

64

3-(2′-deoxy-D-erythro-pentofuranosyl)-5,6,7-trihydro-6,7-dihydroxyimidazo[1,2-a]purine-

65

-9-one (dG-gx),15 which can be converted to the stable product N2-carboxymethyl-2′-deoxy-

66

guanosine (N2-CMdG).14 In DNA, dG-gx reacts further with the nucleosides to form the

67

cross-linked products dG-gx-dC, dG-gx-dA, and dG-gx-dG.19,20 Among the various types of

68

DNA damage, interstrand cross-links are the most harmful because they completely block 5



69

DNA transcription and replication.21-23 In this laboratory, we previously reported

70

quantification of these three cross-linked adducts in glyoxal-treated calf thymus DNA and in

71

human placental DNA by capillary liquid chromatography tandem mass spectrometry

72

(LC−MS/MS).24 However, stable isotope dilution method was not employed owing to the low

73

yield in synthesis of these isotope standards. Alternatively, a calibration curve was

74

constructed along with each set of samples. In this current study, the stable isotopomers of

75

dG-gx-dC and dG-gx-dA were synthesized and used as internal standards. The isotope of

76

dG-gx-dG was omitted because its yield was the lowest and the abundance of dG-gx-dG in

77

DNA samples is also the lowest among the three cross-linked adducts. With the nanoflow LC

78

system coupled to nanospray ionization tandem mass spectrometry (nanoLC−NSI/MS/MS),

79

the sensitivity and specificity of the assay allowed accurate quantification of these two

80

cross-linked adducts in leukocyte DNA from type 2 diabetes mellitus (T2DM) patients and

81

nondiabetic subjects.

82

6


Page 6 of 32

Page 7 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


83

Material and Methods

84

Materials. Glyoxal (40% aqueous solution), adenosine deaminase, micrococal nuclease,

85

snake venom phosphodiesterase I, calf spleen phosphodiesterase II, nuclease P1, calf thymus

86

DNA, and human placental DNA were obtained from Sigma Chemical Co. (St. Louis, MO).

87

[15N5]2′-Deoxyguanosine ([15N5]dGuo) was from Cambridge Isotope Laboratories (Andover,

88

MA). All reagents are of reagent grade or above.

89

Synthesis of [15N5]dG-gx-dC and [15N5]dG-gx-dA. Standard isotopomers of dG-gx-dC

90

and dG-gx-dA were individually synthesized from [15N5]gx-dG with dC or dA, respectively,

91

and purified as described previously.24 Standard [15N5]gx-dG was synthesized from reaction

92

of [15N5]dGuo (0.5 mM) with glyoxal (5.0 mM) in 0.4 M potassium phosphate buffer (pH 7.4)

93

at room temperature with stirring for 15 hr. The reaction mixture was purified by a disposable

94

C18 solid-phase extraction (SPE) column [Bond Elut Bonded phase C18, 500 mg, 3 mL,

95

Varian (Harbor City, CA)] conditioned with 15 mL of methanol, followed by 15 mL of water.

96

After the sample was loaded, the SPE column was washed with 9 mL of water and eluted

97

with 15% MeOH.

98

[15N5]dG-gx-dC and [15N5]dG-gx-dA were synthesized by the same manner as reported for

99

dG-gx-dC and dG-gx-dA24 except that they were purified separately by a disposable C18 SPE

100

column to avoid contamination from the HPLC system. The C18 SPE column was

101

conditioned with 15 mL of methanol, and followed by 15 mL of water. After the sample was 7



102

loaded, the SPE column was washed with 12 mL of water and 3 mL of 5% aqueous methanol,

103

and [15N5]dG-gx-dC was eluted with 3 mL of 10% aqueous MeOH. For [15N5]dG-gx-dA, the

104

SPE column was washed with an additional 3 mL of 15% aqueous MeOH, and

105

[15N5]dG-gx-dA was eluted with 3 mL of 25% aqueous MeOH. The purity of [15N5]dG-gx-dC

106

and [15N5]dG-gx-dA obtained thereof was confirmed and quantified by comparing with

107

dG-gx-dC and dG-gx-dA using nanoLC-NSI/MS/MS described below (Figure S1, Supporting

108

Information). The yield was 15% for both isotope-labeled standards. The daughter ion scan

109

spectra were provided in Figure S2, Supporting Information. [15N5]dG-gx-dC: ESI+ MS: m/z

110

540 ([M + H]+), 424 ([M + H − dR]+), 308 ([M + H − 2 dR]+), 228 ([dC + H]+).

111

[15N5]dG-gx-dA: m/z 564 ([M + H]+), 448 ([M + H − dR]+), 332 ([M + H − 2 dR]+), 252 ([dA

112

+ H]+).

113

DNA Extraction from Blood. Human blood freshly drawn from the vein was added to a

114

solution of of 10% (v/v) citrate-dextrose as an anticoagulant and stored at 4 °C. Within a

115

week, the blood/anticoagulant solution was subjected to the Blood Genomic Midiprep System

116

(Viogen, Sunnyvale, CA) or the Blood & Cell Culture DNA Maxi Kit (QIAGEN, Duesseldorf,

117

Germany), following the protocols provided by the manufacturers. The mount of DNA was

118

quantified by a NanoPhotometer (Implen, Inc. Westlake Village, CA). The purity of DNA was

119

confirmed by the absorbance ratio at 260 and 280 nm being between 1.7 and 1.9.

120

DNA Hydrolysis and Quantitation of dG. Typically, a solution containing DNA (50 µg) 8


Page 8 of 32

Page 9 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


121

and 100 pg each of [15N5]dG-gx-dC and [15N5]dG-gx-dA was added nuclease P1 (0.5 unit)

122

and calf spleen phosphodiesterase II (0.0025 unit) in 30 mM sodium acetate (pH 5.0) and

123

incubated at 37 °C for 6 h. The solution was added snake venom phosphodiesterase I (0.025

124

unit) and alkaline phosphatase (10 units) in 0.15 M Tris-HCl (pH 8.9) and incubated at 37 °C

125

for 4 h, followed by addition and incubation with adenosine deaminase (0.5 unit) at 37 °C for

126

an additional 0.5 h.24 The amount of dG contained in the hydrolysate was analyzed by a

127

portion (1/20) of the DNA hydrolysate by an HPLC-UV system consisting a Hitachi L-7000

128

pump system with a D-7000 interface (Hitachi, Tokyo, Japan), a Rheodyne injector, and a

129

reversed phase C18 column [5 TC-C18(2), 4.6 mm × 250 mm, 5 µm, Agilent Technologies,

130

Palo Alto, CA]. The column was eluted at a flow rate of 1.0 mL/min with water (mobile

131

phase A) and methanol (mobile phase B) with a linear gradient from 0% to 100% methanol in

132

30 min. dG was eluted at 11.6 min and quantified using a calibration curve constructed from

133

0.1 to 1.0 µg dG. The amount of dG released from digestion was nearly quantitative.

134

Removal of Hydrolytic Enzymes from the DNA Hydrolysate. Three different

135

methods were used to remove hydrolytic enzymes in the DNA hydrolysate (131 µL, final

136

volume). The DNA hydrolysate was filtered by (A) 0.22 µm Nylon syringe filter or (B) 0.22

137

µm PTFE spin column and centrifuged at 2,400 ×g for 5 min. In method (C), DNA

138

hydrolysate was added cold acetone (1180 µL), allowed to stand at −20 °C for 20 min, and

139

removed the precipitate by centrifugation at 21,600 ×g for 20 min. 9



140

SPE Enrichment. The cross-linked adducts dG-gx-dC and dG-gx-dA in the DNA

141

hydrolysate was enriched by a reversed phase C18 SPE column [Bond Elut Bonded phase

142

C18, 100 mg, 1 mL, Varian (Harbor City, CA)]. The SPE column was conditioned with 5 mL

143

of methanol, followed by 5 mL of water. The sample was loaded and the SPE column was

144

washed with 3 mL of water and 1 mL of 5% aqueous MeOH. Then, the cross-linked adducts

145

were eluted with 1 mL of 30% aqueous MeOH. The eluent was evaporated to dryness and

146

reconstituted in 10 µL of 0.1% aqueous acetic acid (v/v) before analysis by

147

nanoLC−NSI/MS/MS.

148

nanoLC System. The LC system consists of a 2 µL injection loop connected to a

149

six-port switching valve injector into an UltiMate 3000 Nano Nano and Capillary LC system

150

(Dionex, Amsterdam, The Netherlands) and a reversed phase tip-column packed in-house

151

(Magic C18AQ, 5 µm, 200 Å, Michrom BioResource, Album, CA). System A employed a 75

152

µm × 11 cm column with mobile phase A of 0.1% acetic acid (v/v, pH 3.2) and mobile phase

153

B of 0.1% acetic acid (v/v) in acetonitrile eluting at the flow rate of 300 nL/min. System B

154

used a 180 µm × 11 cm column with mobile phase A being0.1% acetic acid (v/v, pH 3.2) and

155

mobile phase B being0.1% acetic acid (v/v) in methanol. The elution started with a linear

156

gradient of 1% mobile phase B (v/v/) to 100% mobile phase B from 0 to 30 min and it was

157

maintained at 100% B in the next 20 min at the flow rate of 500 nL/min.

158

NSI/MS/MS Analysis. The column effluent was subjected to analysis by a triple 10


Page 10 of 32

Page 11 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


159

quadrupole mass spectrometer, TSQ Quantum Ultra EMR mass spectrometer (Thermo

160

Electron Corp., San Jose, CA), equipped with a nanospray ionization (NSI) interface and

161

performed under the positive-ion mode. The source temperature was at 220 °C, and the spray

162

voltage was 1.5 kV. For the MS/MS experiments, argon was the collision gas with the

163

collision energy of 10 V, and the pressure of the collision cell was 1.5 mTorr. The

164

adduct-enriched sample was analyzed by either nanoLC-NSI/MS/MS with the transition from

165

the parent ion [M + H]+ focused in quadrupole 1 (Q1) and dissociated in a collision cell (Q2),

166

yielding the product ion analyzed in quadrupole 3 (Q3) under the highly selective reaction

167

monitoring (H-SRM) mode. The mass width at Q1 was 0.2 m/z and that at Q3 was 0.7 m/z,

168

and the dwell time was 0.1 s.

169

For dG-gx-dC and [15N5]dG-gx-dC, the SRM 1monitoring Q1 and Q3 were m/z 535.2 →

170

m/z 419.0 and m/z 540.2 → m/z 424.0, respectively, with the collision energy of 10 V. For

171

dG-gx-dA and [15N5]dG-gx-dA, the SRM 1 monitored Q1 and Q3 were m/z 559.2 → m/z

172

443.0 and m/z 564.2 → m/z 448.0, respectively, with the collision energy of 15 V. The SRM 2

173

monitored Q1 and Q3 for dG-gx-dC and [15N5]dG-gx-dC at m/z 535.2 → m/z 303.0 and m/z

174

540.2 → m/z 308.0, respectively, with the collision energy of 15 V. For dG-gx-dA and

175

[15N5]dG-gx-dA, the SRM 2 monitored Q1 and Q3 were m/z 559.2 → m/z 327.0 and m/z

176

564.2 → m/z 332.0, respectively, with the collision energy of 15 V (Table 1).

177

Calibration curve. The solutions containing 100 pg each of [15N5]dG-gx-dC and 11



178

[15N5]dG-gx-dA with various amounts (0, 0.05, 0.1, 0.2, 0.5, 1.0, 10, and 50 pg) of the

179

dG-gx-dC and dG-gx-dA were prepared in triplicates. Each sample went through a C18 SPE

180

column as described above. The fraction containing these cross-linked adducts was

181

evaporated to dryness and reconstituted in 10 μL of 0.1% acetic acid (pH 3.2), and an

182

aliquot (2 µL) was analyzed by the nanoLC−NSI/MS/MS system described above.

183

Study-Subjects. Collection of blood from type 2 diabetic mellitus patients was

184

approved by Institutional Review Board of Buddhist Dalin Tzu Chi General Hospital (IRB

185

No. B10203014). The patients were 14 males and 24 females. The mean (±SD) age was 58.4

186

± 12.1 (ranging from 23 to 77) years and the mean body mass index (BMI) was 27.6 ± 4.0.

187

The mean (±SD) HbA1C level was 9.6 ± 1.8%.

188

The non-diabetic subjects included 27 males and 12 females, recruited from students and

189

employees of the National Chung Cheng University under the approval of the Institutional

190

Review Board of the National Chung Cheng University (IRB No. 100112902). The mean (±

191

SD) age was 34.6 ± 15.3 (ranging from 21 to 60) years and the average BMI was 24.2 ± 6.2.

192

The subjects were given a written warranty stating that the information they provided was for

193

research purposes only and that their personal information would be kept confidential.

194

Statistical Analysis. Statistical analyses were performed by GraphPad InStat version

195

3.00 for Windows 95, GraphPad Software (San Diego, CA, www.graphpad.com). The

196

nonparametric Mann−Whitney test was used to analyze the difference in the adduct levels 12


Page 12 of 32

Page 13 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


197

between diabetic patients and non-diabetic controls. The multiple regression analysis was

198

used to correlate the level of each cross-linked adduct with HbA1c, age, number of cigarettes

199

smoked per day, and BMI.

13



200

Results and Discussion

201

Synthesis and Characterization of Stable Isotope-Labeled Standards. Standard

202

[15N5]dG-gx-dC and [15N5]dG-gx-dA were individually synthesized from [15N5]gx-dG with

203

dCyd and dAdo, respectively, using the reported procedures for dG-gx-dC and dG-gx-dA.24

204

To avoid cross-contamination, the isotope standards were purified through a disposable SPE

205

column. [15N5]dG-gx-dC and [15N5]dG-gx-dA were characterized by their distinctive UV

206

spectra and retention time on HPLC and by comparing their collision-induced dissociation

207

mass spectra

208

Information).24 The purity of the two isotopomers [15N5]dG-gx-dC and [15N5]dG-gx-dA was

209

confirmed by nanoLC−NSI/MS/MS analysis (described below), showing no signals for

210

dG-gx-dC and dG-gx-dA. The quantities of [15N5]dG-gx-dC and [15N5]dG-gx-dA were

211

estimated by the respective UV molar responses of standard dG-gx-dC and dG-gx-dA and

212

calibrated by their peak areas in the nanoLC−NSI/MS/MS.24 [15N5]dG-gx-dC and

213

[15N5]dG-gx-dA were chemically pure with an isotopic purity of 100% and 99.8%,

214

respectively (Figures S2, Supporting Information).

with those of standard dG-gx-dC and dG-gx-dA (Figure S1, Supporting

215

Method Development for Detecting DNA Cross-Linked Adducts by Stable Isotope

216

Dilution Approach. The procedures for analyzing dG-gx-dC and dG-gx-dA include the (1)

217

isolation and quantification of DNA, (2) addition of isotope-labeled [15N5]dG-gx-dC and

218

[15N5]dG-gx-dA to DNA as internal standards, (3) enzyme hydrolysis of DNA to its 14


Page 14 of 32

Page 15 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


219

nucleosides, (4) enrichment of the cross-linked adducts through a disposable reversed phase

220

SPE column, and (5) nanoLC−NSI/MS/MS analysis under the highly selected reaction

221

monitoring (H-SRM) mode (Scheme 1).

222

To achieve accurate quantification of trace analyte in an assay, an appropriate internal

223

standard is needed to compensate for the analyte loss in the multi-step assay procedures as

224

well as for the matrix effect. The ideal internal standard in MS-based analysis is the stable

225

isotope analog of the analyte, since it has identical physical and chemical properties as the

226

analyte except for its mass.25-27 The isotope-labeled internal standard functions to identify the

227

analyte peak in the complex chromatogram. It also serves as a carrier for the small amount of

228

analyte through the multi-step procedures. As the amount of analyte in samples can vary

229

widely, and a trace amount of analyte would suffer from poor recovery without the presence

230

of the isotope-labeled internal standard.28

231

A fixed amount of [15N5]dG-gx-dC and [15N5]dG-gx-dA (100 pg each) was added to a

232

solution containing DNA, which was then enzymatically hydrolyzed to its nucleosides using

233

the appropriate enzymes under the optimized conditions.24 The hydrolytic enzymes were

234

removed by three different methods: (A) 0.22 µm nylon syringe filter, (B) 0.22 µm PTFE

235

spin column, or (C) precipitation with cold acetone. Method C was chosen for the assay

236

because the recovery of [15N5]dG-gx-dC and [15N5]dG-gx-dA (65% and 68%, respectively)

237

was the highest and the chromatograms showed the least interference. 15



238

Two DNA extraction kits from different venders (Qiagen and Viogen) were used in this

239

study. The efficacies of these two kits on the same blood sample were evaluated, and similar

240

adduct levels were obtained (data not shown). The average amount of DNA isolated from

241

these kits was 23.1 ± 7.9 (± SD) µg/mL in 30 blood samples. The isolated leukocyte DNA

242

was enzymatically hydrolyzed to its nucleosides. Because of structural alteration of the DNA

243

by the cross-linking, hydrolysis of the cross-linked adducts by the hydrolytic enzymes can be

244

much less efficient than that of unmodified normal nucleosides. The hydrolysis efficiency

245

varies greatly according to our previous finding.24 The optimized conditions employed in this

246

study were modified from those used for other DNA cross-links.29,30

247

The isotope-labeled standards can adjust for loss of the target analytes during the

248

hydrolysis and enrichment steps by a SPE column. Two SPE columns were examined for

249

their performance, namely, a reversed phase C18 column and a polymer-based Strata-X SPE

250

column. The former gave cleaner chromatograms for dG-gx-dC and dG-gx-dA in the DNA

251

hydrolysate and was therefore used for enrichment of these cross-linked adducts. The

252

advantage of using SPE enrichment before LC-MS/MS analysis is that the sample-to-sample

253

carryover interference is avoided.

254

Method Performance and Validation. To achieve high sensitivity and specificity of the

255

assay, a nanoflow LC system was connected to a triple quadrupole mass spectrometer

256

monitored under the H-SRM mode, in which a narrow window (0.2 Da) was used for the 16


Page 16 of 32

Page 17 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


257

precursor ion at the first quadrupole. This setting allows fewer ions to pass through the first

258

quadrupole and thus lowers the background signals, which is particularly helpful in analyzing

259

complex mixture. The adducts were analyzed by nanoLC−NSI/MS/MS under the H-SRM

260

mode using the transition from the parent ion [M + H]+ focused in the first quadrupole (Q1)

261

and dissociated in a collision cell (Q2) to yield the product ion [M + H − deoxyribose]+ (SRM

262

1) or [M + H − 2 deoxyribose]+ (SRM 2), which was then analyzed in the third quadrupole

263

(Q3). The SRM transitions with loss of dual or single deoxyribose for dG-gx-dC, dG-gx-dA,

264

[15N5]dG-gx-dC, and [15N5]dG-gx-dA are listed in Table 1. The sensitivity of dG-gx-dC and

265

dG-gx-dA using SRM 1 was approximately twice of that using SRM 2.24 The transitions with

266

less interference were chosen as quantifier and the other transitions were used for qualitative

267

assurance. As a result, dG-gx-dC was quantified by monitoring the dual loss of deoxyribose

268

and dG-gx-dA was quantified by monitoring the loss of single deoxyribose.

269

Assay Sensitivity. Using

nanoLC−NSI/MS/MS system A with a 75 µm-id column, the

270

limit of detection (defined as a signal-to-noise ratio ≥3) of the standard dG-gx-dC and

271

dG-gx-dA in aqueous solution was 0.1 fg (0.19 amol) and 0.5 fg (0.89 amol), respectively.

272

which was 400 and 80 times lower, respectively, than that using capillary

273

LC−NSI/MS/MS.24 The on-column detection limit of dG-gx-dC and dG-gx-dA was 1.0 fg

274

(1.9 amol) and 3.0 fg (5.4 amol), respectively, using a 180 µm-id column (nanoLC system

275

B). The calibration curves were constructed by adding a fixed amount (100 pg each) of the 17



276

internal standards [15N5]dG-gx-dC and [15N5]dG-gx-dA and various amounts of dG-gx-dC

277

and dG-gx-dA (0−50 pg) to calf thymus DNA (50 µg), in which both adducts are not

278

detectable. The mixture was subjected to enzyme hydrolysis, SPE enrichment, and

279

nanoLC−NSI/MS/MS analysis. Only an equivalent of 10 µg of DNA hydrolysate was

280

subjected to analysis by nanoLC-NSI/MS/MS. The lower limit of quantification was 50 fg

281

for both adducts in either nanoLC system, corresponding to 0.056 dG-gx-dC and 0.065

282

dG-gx-dA in 108 normal nucleotides in 50 µg DNA (Figure S3, Supporting Information),

283

which was 26 and 8 times, respectively, more sensitive than the published report.24 The limit

284

of quantification was not improved by using nanoLC system A, mainly due to the ratio of

285

analyte vs. isotope standard in the positive control experiments.

286

Assay Validation. The precision of the assay was evaluated by analyzing dG-gx-dC and

287

dG-gx-dA in human leukocyte DNA (50 µg) in triplicates per day for experiment performed

288

on three different days. The relative standard deviations of intraday analyses for both adducts

289

were within 10% and those for interday analyses were within 5%, indicating the high

290

reproducibility of the assay (Table 2). The accuracy of the assay was validated by adding

291

known amounts of dG-gx-dC (2.0, 4.0, and 6.0 pg) and dG-gx-dA (6.0, 12, and 18 pg) to

292

human placental DNA (50 µg) and analyzing for their levels in triplicates as described above.

293

Linear regression gave a y-intercept of 0.46 pg and 2.35 pg for dG-gx-dC and dG-gx-dA,

294

respectively, with high correlation coefficients (Figure S4, Supporting Information). These 18


Page 18 of 32

Page 19 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


295

amounts are very close to 0.44 pg and 2.33 pg of dG-gx-dC and dG-gx-dA, respectively, in

296

50 µg of human placental DNA without the addition of standards, which correspond to 0.52

297

and 2.65 adducts in 108 total normal nucleotides.

298

Quantification of dG-gx-dC and dG-gx-dA in Human Leukocyte DNA. Measuring

299

DNA adducts from circulating blood leukocyte DNA can offer a minimally invasive

300

biomarker to reflect the extent of DNA damage of the whole body. The use of leukocyte DNA

301

for biomonitoring requires a highly specific, sensitive, and accurate assay because of the low

302

adduct levels. Only 50 µg leukocyte DNA was used in this stable isotope dilution

303

nanoLC−NSI/MS/MS system.

304

As shown in Figure 1A, the stereoisomers of dG-gx-dC separated into two peaks eluting

305

at 20.85 and 21.15 min, whereas those of dG-gx-dA eluted as one peak at 24.26 min. A

306

selected reaction monitoring transitions from m/z 535.2 303.0 ( [MH]+ [MH − 2

307

deoxyribose]+) was chosen for quantification of dG-gx-dC, whereas m/z 559.2 443.0

308

( [MH]+ [MH − deoxyribose]+) was used for dG-gx-dA because they showed the lowest

309

amount of interference in the chromatograms of human leukocyte samples. The

310

corresponding transitions were used for their isotopomers. Thus, dG-gx-dC and

311

[15N5]dG-gx-dC were quantified by the loss of two deoxyribose moieties, whereas dG-gx-dA

312

and [15N5]dG-gx-dA were quantified by loss of a single deoxyribose unit. The adduct levels

313

of this sample were 4.05 dG-gx-dC and 1.68 dG-gx-dA in 108 normal nucleotides. Another 19



314

set of SRM transitions was used for qualitative assurance (qualifier) for the presence of these

315

adducts (Scheme 1 and Figure 1B).

316

As we have reported previously, the types and amounts of hydrolytic enzymes used can

317

affect the amount of DNA adducts released.31-33 Thus, fixed amounts of DNA and hydrolytic

318

enzymes were used for each sample throughout this study. Because of the low levels of the

319

cross-linked adducts, 50 µg of leukocyte DNA was used, which can be obtained from 2−3 mL

320

of blood, making the assay clinically feasible. Leukocyte DNA was isolated from venous

321

blood of 38 T2DM patients and 39 nondiabetic control subjects, using a blood DNA

322

extraction kit, and quantified. A positive control experiment was performed, which contained

323

the buffer solutions and the enzymes, but not DNA, and was incubated under the same

324

conditions as the samples. After the DNA had been analyzed using the assay procedures

325

described above, the adduct levels were quantified by subtracting the ratio of the analyte

326

versus the internal standard of the sample from that of the positive control experiment, and

327

interpolated into the calibration curve as listed in Table S1, Supporting Information. The

328

presence of a small but detectable amount of adducts in the positive control experiment

329

indicates that the adducts may have been formed artificially during sample work-up. Samples

330

with a ratio of analyte versus the internal standard equal to or less than that of the positive

331

control experiment were marked as being not detectable (ND) and calculated as zero for the

332

mean and for statistical analysis. All samples with a quantifiable adduct level also showed 20


Page 20 of 32

Page 21 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

333


detectable adducts in the SRM transitions for the qualifier.

334

Statistical analysis showed that the levels of dG-gx-dC and dG-gx-dA are significantly

335

higher in the T2DM patients than in the healthy subjects (p

A Stable Isotope Dilution Nanoflow Liquid Chromatography Tandem

Recommend Documents