Expanding the Versatility of a Quantitative Determination

in-source CID would provide a versatile system for bioanalysis if it exhibits similar effects under other *Corresponding author: Nariyasu MANO Receive...

0 downloads 12 Views 929KB Size
Chromatography 2017, 38, 59-63

Original Paper

Expanding the Versatility of a Quantitative Determination Range Adjustment Technique Using In-Source CID in LC/MS/MS Hideaki ISHII1,2, Hiroaki YAMAGUCHI1,3, Nariyasu MANO*1,3 1

Graduate School of Pharmaceutical Sciences, Tohoku University, 1-1 Seiryo-machi, Aoba-ku, Sendai 980-8574, Japan Laboratory of Racing Chemistry, 1731-2 Tsuruta-machi, Utsunomiya 320-0851, Japan

2 3

Department of Pharmaceutical Sciences, Tohoku University Hospital, 1-1 Seiryo-machi, Aoba-ku, Sendai 980-8574, Japan

Abstract The aim of this study was to investigate the applicability of a new technique using in-source collision-induced dissociation (CID) for improving quantitative linear range of various compounds in liquid chromatography-tandem mass spectrometry (LC/MS/MS). To determine whether the linear range shift due to in-source CID occurs under various MS conditions, we investigated the quantitative linear ranges of reserpine, indomethacin, and furosemide in both electrospray ionization (ESI) and atmospheric pressure chemical ionization (APCI) in positive- and negative-ion detection modes. We observed 3–30-fold linear range shifts upon changing the declustering potential to adjust in-source CID in positive- and negative-ion detection modes in both ESI and APCI. These results indicated that this new technique could be applied for arbitrary adjustment of linear range regardless of ionization process and polarities in LC/MS/MS. Therefore, this technique could be applied to simultaneous quantification, in biological fluids, of compounds with quite different sensitivities under various MS conditions. Keywords: In-source CID; Linear range shift; LC/MS/MS; ESI; APCI 1. Introduction Liquid chromatography-tandem mass spectrometry (LC/MS/MS) is now used for analysis of drugs, endogenous substances, and environmental chemicals because of its selectivity and therefore high sensitivity. The quantification ranges in LC/MS/MS are often limited due to saturation of ionization [1] and/or detection by an electron multiplier [2]. When non-linearity appears in the quantification range to measure due to such saturation, the operators must reduce the injection volume or dilute the sample [3]. Consequently, multiple analyses are required when analyzing more than two analytes, such as a drug and its metabolites, with extremely different responses and/or at different targeted concentrations. Recently, we reported a new technique for shifting the linear range using in-source collision-induced dissociation

(CID) in negative-ion detection mode in electrospray ionization (ESI) for LC/MS/MS [4]. In addition, we succeeded in applying the technique to the quantification of a drug and its metabolites, which had different sensitivities on ESI-MS/MS, in a complex biological fluid [5]. Although two useful techniques, monitoring isotopologue transitions in selected reaction monitoring mode (i-SRM) [6,7] and monitoring secondary product ions in SRM mode (s-SRM) [8,9] have already been developed for the same purpose by limiting ion amounts into electron multiplier, our technique using in-source CID was more effective for improvement of linearity because of controlling ion amounts at the ion inlet of orifice to widen the linear quantification range [4]. Therefore, the linear range shifting technique using in-source CID would provide a versatile system for bioanalysis if it exhibits similar effects under other

* Corresponding author: Nariyasu MANO Tel: +81-22-717-7525; Fax: +81-22-717-7545 E-mail: [email protected]

Received: 3 March 2017 Accepted: 19 April 2017 J-STAGE Advance Published: 30 April 2017 DOI: 10.15583/jpchrom.2017.004 - 59 -

Chromatography 2017, 38, 59-63

Table 1. SRM transitions and MS parameters. Compound Polarity Precursor ion Product ion (m/z) (m/z) Reserpine Positive 609.0 397.0 Reserpine (s-SRM) 609.0 577.2 (6.6%)* Indomethacin 357.9 138.9 Indomethacin (s-SRM) 357.9 173.8 (3.3%)* Diclofenac (IS) 295.7 213.9 Indomethacin Negative 355.9 312.0 Indomethacin (s-SRM) 355.9 297.1 (1.7%)* Furosemide 328.8 204.8 Furosemide (s-SRM) 328.8 125.9 (3.0%)* Diclofenac (IS) 293.8 249.9 *: Intensity relative to regular transition. DP: declustering potential; CE: collision energy; CXP: collision cell exit potential.

conditions. In this study, we investigated whether the linear range shifts by in-source CID could be observed under various conditions, including ESI/atmospheric pressure chemical ionization (APCI) and positive-/negative-ion detection modes, using reserpine, indomethacin, and furosemide as model compounds. 2. Experimental procedures 2.1. Materials Diclofenac (internal standard [IS]), furosemide, and indomethacin were purchased from Sigma-Aldrich Chemical (St. Louis, MO, USA). Reserpine was obtained from Fluka (Buchs, Switzerland). Formic acid and ammonium formate were acquired from Wako Pure Chemical Industries (Osaka, Japan). LC/MS-grade acetonitrile was purchased from Kanto Chemical (Tokyo, Japan). High-purity water was obtained using a Milli-Q purification system (Millipore, Bedford, MA, USA). 2.2. Instruments and conditions A Shimadzu Nexera series liquid chromatograph consisting of binary pumps, an online degasser, an autosampler, and a column oven (Shimadzu, Kyoto, Japan) was linked to a linear ion trap quadrupole mass spectrometer equipped with an APCI or turbo ionspray interface (QTRAP 4500, AB SCIEX, Framingham, MA, USA). Analytes were separated using a Kinetex Biphenyl (2.1 mm i.d. × 100 mm, 2.6 μm, Phenomenex, Torrance, CA, USA) whose temperature was maintained at 40°C. A 5-minute isocratic analysis was performed with a mobile phase consisting of 50% solvent A (0.1% formic acid and 10 mmol/L ammonium formate in water) and 50% solvent B (acetonitrile) at a flow rate of 0.3 mL/min. All analytes were monitored under the SRM mode described in Table 1. Instrument settings for ESI analysis were as follows: heater temperature, 700°C; curtain gas, 20 psi; nebulizer gas, 40 psi; heater gas, 80 psi; collisionally activated dissociation gas, 10 arbitrary units; ionspray voltage, −4500 V or 5500 V; dwell time, 40 ms. Instrument settings

DP (V) 135 135 80 80 66 −40 −40 −60 −60 −30

CE (V) 39 39 29 29 45 −12 −12 −30 −30 −18

CXP (V) 6 6 4 4 10 −9 −9 −15 −15 −15

for APCI analysis were as follows: ceramic tube temperature, 500°C; curtain gas, 20 psi; nebulizer gas, 20 psi; collisionally activated dissociation gas, 10 arbitrary units; nebulizer current, −2 μA or 1 μA; dwell time, 40 ms. The other common parameters among ESI and APCI are shown in Table 1. All peak integrations were performed using the Analyst 1.6.2 software. The ratios of the peak area of each analyte to nominal analyte concentrations were fitted by least-squares linear regression using 1/y2 weighting factors. 2.3. Sample preparation Stock solutions (1.0 mg/mL) were prepared independently in methanol. Standard working solutions (0.1, 0.3, 1, 3, 10, 30, 100, 300, 1000, 3000, 10000, 30000, 100000 ng/mL) were prepared by diluting the stock solutions in water/acetonitrile (1:1, v/v). IS working solution (1 μg/mL) was prepared by diluting stock solution with water/acetonitrile (1:1, v/v). To investigate linearity in the concentration range 0.1–100000 ng/mL, 500 μL aliquots of standard working solution and 100 μL IS working solution were transferred to a vial and mixed. Two-microliter aliquots of this solution were injected in triplicate into the LC/MS/MS system. 2.4. Investigation of linearity Linearities of reserpine, indomethacin, and furosemide were investigated under various MS conditions. To investigate whether saturation occurred at the electron multiplier or under ion source (ESI or APCI), product ions with different relative intensities observed in product ion spectra of those compounds (Figs. 1A, 1B, 1D, and 1E) were also monitored in SRM mode (s-SRM). Next, 10 different declustering potentials (DP) were used, with 15-V increments from 135 V for reserpine; 10-V increments from 80 V or 10-V decrements from −40 V for indomethacin; and 10-V decrements from −60 V for furosemide. Diclofenac, used as an IS, was monitored using the transition from m/z 295.7 to 213.9 under positive-ion detection mode and from m/z 293.8 to 249.9 under negative-ion detection mode to

- 60 -

Chromatography 2017, 38, 59-63

Relative intensity (%)

(A) 100

609

195 50

174

0

236

200

365

448

400

m/z 174

Relative intensity (%)

m/z 448 N

N H H

O

577

O

600

139

O

m/z 236

O

m/z 174 111

O

Cl

3. Results and discussion During full scan analysis using both ESI and APCI, protonated molecules ([M+H]+) for reserpine, indomethacin, and diclofenac, and deprotonated molecules ([M−H]−) for indomethacin, furosemide, and diclofenac, were detected as base peaks. Product ion spectra derived from those protonated and/or deprotonated molecules as precursor ions were obtained by infusion analysis using each standard solution at a concentration of 1 µg/mL, as shown in Fig. 1. Diclofenac yielded simple product ion spectra under positive- and negative-ion detection modes, whereas reserpine, indomethacin, and furosemide yielded multiple product ions with different relative intensities. Although the most abundant peaks in the product ion spectra were used for SRM transitions, the relatively low-abundance peaks with intensities of a few percent were used for s-SRM transitions, as shown in Table 1. Typical SRM chromatograms obtained with ESI were shown in Fig. 2. Furosemide, reserpine, diclofenac, and indomethacin were detected at retention times of 1.10, 1.72, 2.71, and 2.87 min, respectively. To check the fragmentation of the target molecule by in-source CID, we investigated its effect using reserpine; the results are shown in Fig. 3. The higher DP promoted the fragmentation of reserpine, and other compounds exhibited almost the same tendencies (data not shown).

N O

174

HO

0

O

m/z 111

m/z 139

50

O

O

O

m/z 577

m/z 195

H

H O

(B) 100

normalize responses of analytes. Criteria for linearity were set as follows: (1) Regression lines should be composed of at least four successive points in triplicate. (2) Accuracy (mean back-calculated concentration, expressed as a percentage of nominal concentration) must be within ±15% (or within ±20% at the lower limit of quantification [LLOQ]) [10]. (3) Precision, expressed as the percentage of the coefficient of variance (CV), cannot exceed 15% (or 20% at the LLOQ) [10].

m/z 397 m/z 365

397

200

400

O

600

Relative intensity (%)

(C) 214

100

m/z 214 Cl

50 Cl

NH

OH O

0

200

400

600

Relative intensity (%)

(D) 312

100

50 270 0

297

m/z 270

356

200

400

m/z 297 m/z 312

600

Relative intensity (%)

(E) 205

100

50

0

m/z 78 m/z 285

78 126

m/z 126

285

200

400

600

m/z 205

250

100

m/z 250

50

0

2.71

8×104

200

400

Intensity (cps)

Relative intensity (%)

(F)

600

m/z

Fig. 1. Product ion mass spectra of protonated molecules of (A) reserpine, (B) indomethacin, and (C) diclofenac (IS), and deprotonated molecules of (D) indomethacin, (E) furosemide, and (F) diclofenac (IS). Conditions: sample concentrations, 1 μg/mL in water-acetonitrile (1:1, v/v); infusion flow rate, 5 μL/min; instrument, QTRAP 4500; ionization, ESI; spray voltage, (A-C) 5500 V, (D-F) −4500 V; precursor ions: (A) m/z 609.0, (B) m/z 357.9, (C) m/z 295.7, (D) m/z 355.9, (E) m/z 328.8, (F) m/z 293.8; DP: (A) 135 V, (B) 80 V, (C) 66 V, (D) −40 V, (E) −60 V, (F) −30 V; CE: (A) 39 V, (B) 29 V, (C) 45 V, (D) −12 V, (E) −30 V, (F) −18 V; CXP: (A) 6 V, (B) 4 V, (C) 10 V, (D) −9 V, (E) −15 V, (F) −15 V.

1.10

1.72

2.71 2.87 2.87

0

0.0

(C) (A) (B)

Time (min)

(D) (E)

(F)

4.0

Fig. 2. SRM chromatograms of (A) reserpine, (B) diclofenac, (C) indomethacin, (D) furosemide, (E) diclofenac and (F) indomethacin. Conditions: injected mass, 50 pg each; instrument, QTRAP 4500; ionization, ESI; analytical column, Kinetex Biphenyl (2.1 mm i.d. × 100 mm, 2.6 µm); mobile phase, 0.1 v/v% formic acid and 10 mmol/L ammonium formate in water/acetonitrile (1:1, v/v); isocratic flow, 0.3 mL/min; SRM: (A, positive) m/z 609.0→397.0, (B, positive) m/z 295.7→213.9, (C, positive) m/z 357.9→138.9, (D, negative) m/z 328.8→204.8, (E, negative) m/z 293.8→249.9, (F, negative) m/z 355.9→312.0.

- 61 -

Chromatography 2017, 38, 59-63

ESI

300

m/z

609 500

700

Fig. 3. Promotion of fragmentation of protonated reserpine by in-source CID. Conditions: injected mass, 2 ng each; instrument, QTRAP 4500; scan mode, full scan; ionization, ESI in positive-ion mode; DP, (A) 135 V, (B) 150 V, (C) 175 V; analytical column, Kinetex Biphenyl (2.1 mm i.d. × 100 mm, 2.6 µm); mobile phase, 0.1 v/v% formic acid and 10 mmol/L ammonium formate in water/acetonitrile (1:1, v/v); isocratic flow, 0.3 mL/min.

(D)

1000

1E+077 1×10

100

100

10

10

1

1

0.1

0.1

100000

100000

10000

10000

1000

1000

100

100

10

10

1

1

0.1

0.1

1E+055 1×10

1E+033 1×10

s-SRM

1E+044 1×10

8 1E+08 1×10 7 1E+07 1×10 6 1E+06 1×10 5 1E+05 1×10 4 1E+04 1×10 3 1E+03 1×10

100000

100000

1E+088 1×10

10000

10000

1000

1000

1E+077 1×10

100

100

10

10

1

1

0.1

0.1

DP (V)

s-SRM

1E+066 1×10

80 90 100 110 120 130 140 150 160 170

Intensity (cps)

1000

135 150 165 180 195 210 225 240 255 270

Intensity (cps) s-SRM

135 150 165 180 195 210 225 240 255 270

s-SRM

1E+088 1×10

10000

DP (V)

s-SRM

0 100

100000

10000

s-SRM

397

174

1E+044 1×10

-40 -50 -60 -70 -80 -90 -100 -110 -120 -130

(C)

195

1E+055 1×10

100000

+ H+

ESI

1E+066 1×10

1E+033 1×10

s-SRM

4×106

Concentration (ng/mL)

Intensity (cps)

(C)

0.1

Intensity (cps)

0

0.1

s-SRM

397

1

Intensity (cps)

O

10

1

s-SRM

(B)

609

10

80 90 100 110 120 130 140 150 160 170

+ H+

100

-40 -50 -60 -70 -80 -90 -100 -110 -120 -130

O

195

H

100

-60 -70 -80 -90 -100 -110 -120 -130 -140 -150

O

H

1E+077 1×10

s-SRM

N

N H H

1000

s-SRM

O

4×106

10000

1000

s-SRM

0

10000

135 150 165 180 195 210 225 240 255 270

397

1E+088 1×10

80 90 100 110 120 130 140 150 160 170

O

100000

-40 -50 -60 -70 -80 -90 -100 -110 -120 -130

195

609

O

O O

(B) Intensity (cps)

O

APCI

No linearity

100000

-60 -70 -80 -90 -100 -110 -120 -130 -140 -150

O

(A)

Concentration (ng/mL)

O

O

Concentration (ng/mL)

4×106

H

H

Good linearity

+ H+

Concentration (ng/mL)

Intensity (cps)

(A)

N

N H H

-60 -70 -80 -90 -100 -110 -120 -130 -140 -150

APCI O

1E+06 1×106 1E+055 1×10 1E+044 1×10 1E+033 1×10

DP (V)

Fig. 4. Effect of ion sources, polarities, and in-source CID on linear ranges using APCI (left) and ESI (middle). (A) reserpine (positive), (B) indomethacin (positive), (C) indomethacin (negative), and (D) furosemide (negative). The graphs on the right represent intensity at ULOQ using APCI (○) and ESI (♦).

We analyzed standard solutions for all tested compounds in the concentration range of 0.1–100000 ng/mL using LC/MS/MS, and investigated the effect of DP values on the linear range (Fig. 4). In APCI, the use of the initial value of DP (135 V), which was most appropriate for production of the intense precursor ion, provided wider and higher linear ranges for reserpine and indomethacin in comparison with ESI (Figs. 4A, 4B, and 4C). On the other hand, in APCI, furosemide yielded a relatively narrow linear range (Fig. 4D). Because other study of quantification of furosemide using APCI reported similar characteristics [11], this behavior might be a consequence of the physicochemical characteristics of furosemide. Comparison between APCI and ESI of the intensities obtained from the CAL-1000 sample, which contained each analyte at the concentration of 1000 ng/mL with the fixed concentration of IS, revealed that the ESI yielded an equal or more intense response for all compounds (Table 2). Reserpine exhibited good linearity in the range of 3–30000 ng/mL under positive APCI with DP of 135 V, whereas at DP 180 V the linear range was 3–100000 ng/mL. Those intensities were 8–9 × 106 counts per second (cps)

for the 30000 ng/mL sample with DP of 135 V and the 100000 ng/mL sample with DP of 180 V, respectively, suggesting that the upper detection limit by the electron multiplier seemed to be approximately 1 × 107 cps. Although s-SRM analysis of reserpine using the transition from m/z 609.0 to 577.2 (relative intensity, 6.6%) also exhibited good linearity in the range of 10–100000 ng/mL, the intensity of the 100000 ng/mL sample at the upper limit of quantification (ULOQ) was 2 × 106 cps. Because ion counts were one-fifth lower than the detection limit of the electron multiplier (1 × 107 cps), analysis of a higher concentration (300000 ng/mL) could extend the linear range. On the other hand, the linear range of reserpine at DP 135 V in ESI was 1–3000 ng/mL, narrower than in the case of APCI, albeit with higher sensitivity. At a DP of 165 V, the ULOQ shifted to 10000 ng/mL, and remained constant at higher voltage. Because increasing DP was associated with decreasing intensity of reserpine, the lack of an increase in sensitivity might have been due to ionization saturation during the ESI process. For analysis of indomethacin in positive APCI, the linear range was 3–10000 ng/mL at DP 80 V, and shifted to

- 62 -

Chromatography 2017, 38, 59-63

Table 2. Intensity obtained from CAL-1000 using APCI and ESI. Compound Polarity Precursor ion Product ion (m/z)

(m/z)

DP

Intensity (cps)

(V)

APCI

ESI

Reserpine Positive 609.0 397.0 135 3.7 × 105 1.3 × 106 5 Indomethacin Positive 357.9 138.9 80 9.2 × 10 9.2 × 105 4 Indomethacin Negative 355.9 312.0 −40 3.5 × 10 4.3 × 105 Furosemide Negative 328.8 204.8 −60 4.1 × 105 1.4 × 106 Conditions: injected mass, each 2 ng; instrument, QTRAP 4500; analytical column, Kinetex Biphenyl (2.1 mm i.d. × 100 mm, 2.6 µm); mobile phase, 0.1 v/v% formic acid and 10 mmol/L ammonium formate in water/ acetonitrile (1:1, v/v); isocratic flow, 0.3 mL/min.

30–100000 ng/mL at DP of 110 V (Fig. 4B). ESI analysis at DP 80 V exhibited linearity in the range 1–10000 ng/mL, and the linear range was shifted by in-source CID in a manner similar to APCI analysis. Because both ESI and APCI at ULOQ gave several million intensities at DP between 80–110 V, the main factor defining the linear range would be the limit of detection of the electron multiplier. It should be noted that a wider quantitative range of 1–100000 ng/mL was achieved by single ESI analysis with two DPs (80 V and 110 V) without dilution of samples whereas maximum linear ranges in i-SRM and s-SRM were from 4 ×102 to 2 ×104 [6,7] and from 7.5 ×103 to five orders of magnitude [8,9], respectively. For indomethacin analysis under negative-ion detection mode, the linear ranges were 300–100000 ng/mL on APCI and 3–3000 ng/mL on ESI, respectively (Fig. 4C), and the LLOQs in both APCI and ESI were quite different. ESI exhibited 100-fold higher sensitivity than APCI, although in both, the intensities of ULOQ samples were approximately 1 × 107 cps. These data supported the idea that the incorporation limit of ions into the electron multiplier dramatically improved linearity without saturation of the ionization process. APCI was not suitable for ionization of furosemide under negative-ion detection mode. The linear range was constant (30–3000 ng/mL) at DP from −60 V to −110 V, with very low intensities of ULOQ, approximately 2 × 105 cps. In ESI analysis, the linear range of furosemide was 1–10000 ng/mL at DP −60 V, and the intensity at ULOQ was 1 × 107 cps. The decrease in intensity with increasing DP indicates in-source CID-related decrease in the amount of ions introduced into mass spectrometer without saturation on the electron multiplier. 4. Conclusion In this study, we investigated the applicability of the linear range shifting technique using in-source CID under various conditions in ESI and APCI under positive- and negative-ion detection modes. Linear range shifts were observed under various conditions in a predictable manner, although the phenomena seemed to depend on the physicochemical properties of individual compounds. The

linear range shifting technique using in-source CID is applicable to simultaneous quantitation, in biological fluids, of various compounds with quite different sensitivities using LC/MS/MS. Moreover, by using multiple DPs, this methodology could achieve a wider range, beyond the general limitations of the linear range of LC/MS/MS. This approach would be especially useful for the determination of samples with a broad range of unknown concentrations. References [1] Tang, K.; Page, J. S.; Smith, R. D. J. Am. Soc. Mass Spectrom. 2004, 15, 1416-1423. [2] Collings, B. A.; Dima, M. D.; Ivosev, G.; Zhong, F. Rapid Commun. Mass Spectrom. 2014, 28, 209-216. [3] Remaud, G.; Boisdron-Celle, M.; Morel, A.; Gamelin, A. J. Chromatogr. B 2005, 824, 153-160. [4] Ishii, H.; Yamaguchi, H.; Mano, N. Chem. Pharm. Bull. 2016, 64, 356-359. [5] Ishii, H.; Shimada, M.; Yamaguchi, H.; Mano, N. Biomed. Chromatogr. 2016, 30, 1882-1886. [6] Liu, H.; Lam, L.; Dasgupta, P. K. Talanta 2011, 87, 307-310. [7] Tsuji, M.; Matsunaga, H.; Jinno, D.; Tsukamoto, H.; Suzuki, N.; Tomioka, Y. J. Chromatogr. B 2014, 953-954, 38-47. [8] Yuan, L.; Zhang, D.; Jemal, M.; Aubry, A. F. Rapid Commun. Mass Spectrom. 2012, 26, 1465-1474. [9] Curtis, M. A.; Matassa, L. C.; Demers, R.; Fegan, K. Rapid Commun. Mass Spectrom. 2001, 15, 963-968. [10] US FDA. Guidance for Industry, Bioanalytical Method Validation, U.S. Department of Health and Human Services, FDA, Center for Drug Evaluation and Research, Center for Veterinary Medicine 2001. [11] Abdel-Hamid, M. E. Farmaco 2000, 55, 448-454.

- 63 -