The characterization of selected drugs with infrared laser desorption/tunable synchrotron vacuum ultraviolet photoionization mass spectrometry

RAPID COMMUNICATIONS IN MASS SPECTROMETRYRapid Commun. Mass Spectrom. 2008; 22: 2515–2520Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/rcm.3639 The characterization of selected drugs with infrared laserdesorption/tunable synchrotron vacuum ultravioletphotoionization mass spectrometry Yang Pan1, Hao Yin2, Taichang Zhang1, Huijun Guo1, Liusi Sheng1 and Fei Qi1*1National Synchrotron Radiation Laboratory, University of Science and Technology of China, Hefei, Anhui 230029, P.R. China2Hefei National Laboratory for Physical Sciences at Microscale, University of Science and Technology of China, Hefei, Anhui 230026, P.R. China Received 17 March 2008; Revised 2 June 2008; Accepted 14 June 2008 Some selected drugs including captopril, fudosteine and racecadotril have been analyzed by infrared(IR) laser desorption/tunable synchrotron vacuum ultraviolet (VUV) photoionization mass spec-trometry (PIMS). The molecular ions of captopril and racecadotril are exclusively observed withoutany fragments at near threshold single-photon ionization (SPI). However, fudosteine easily formsfragments even at a photon energy near the ionization threshold, indicating the instability of itsmolecular ion. For these drugs, a number of fragments are yielded with the increase of photon energy.
The structures of such fragments proposed by IR LD/VUV PIMS are supported by electron ionizationtime-of-flight mass spectrometry (EI-TOFMS) results. Fragmentation pathways are discussed indetail. Copyright # 2008 John Wiley & Sons, Ltd.
Liquid chromatography/mass spectrometry (LC/MS) has They demonstrated that APPI gives higher efficiency in played a key role in the advancement of drug analysis and ionizing non-polar compounds than ESI and APCI while also discovery for the quantitative and qualitative analysis of low producing comparable ion signals for polar ones.10 molecular weight drugs and their metabolites,1–4 particularly In addition to APPI, another photoionization method – since the advent of electrospray ionization (ESI) and two-laser mass spectrometry (L2MS) – has been reported, in atmospheric pressure chemical ionization (APCI). In positive which two laser beams are utilized and the laser-induced ion mode, however, many drugs, especially those that are processes are decoupled, i.e., one laser is used to desorb the non-polar, are not efficiently ionized by ESI and APCI. Thus, analyte on a substrate, while another acts as a ’soft’ ionization such drugs provide only weak positive ion ESI and APCI source.11–16 L2MS has proven to be a powerful analytical signals. In addition, ESI is vulnerable to ionization suppres- technique for characterizing a wide range of molecular sion from biological matrices resulting in inconsistent systems.17,18 We recently constructed an instrument com- analytical results.5,6 APCI is not always suitable for qualita- bining infrared laser desorption with tunable synchrotron tive analysis such as metabolite characterization due to the VUV photoionization mass spectrometry (IR LD/VUV in-source fragmentation of thermally labile molecules.7 PIMS) for organic analysis.19 IR LD/VUV PIMS has been Therefore, more universal ionization techniques are needed successfully used for the analysis of a variety of small compounds with a detection limit of 50 fmol for 9,10- In the early 2000s, a high-sensitivity ionization technique anthraquinone (signal-to-noise (S/N) ¼ 10). Although this named atmospheric pressure photoionization (APPI) was newly developed method has not yet been used for high- introduced by Robb et al.8 APPI uses a vacuum ultraviolet throughput pharmaceutical purposes, it exhibits several (VUV) lamp to ionize compounds, and this overcomes the advantages in providing structural and energetic infor- drawbacks of ESI and APCI. APPI has been widely used to mation, and information on the photo-induced ionization/ study drugs in plasma5 and drug metabolites.9 More dissociation of drugs, especially for those that are thermally recently, Cai et al. compared the performances of APPI, labile and non-volatile. In a study of uridine with IR LD/ APCI and ESI for the ionization of several hundreds of drugs.
VUV PIMS,19 due to the rapid thermal effect of a nanosecondIR laser beam, neutral molecules are desorbed without *Correspondence to: F. Qi, National Synchrotron Radiation decomposition. Subsequently, molecular ions can be obtained Laboratory, University of Science and Technology of China, by near-threshold photoionization (or the molecular ions Hefei, Anhui 230029, P.R. China.
E-mail: [email protected] gradually dissociate to fragment ions with increased photon Contract/grant sponsor: The Chinese Academy of Sciences; energy). In comparison, very strong fragmentation is observed for uridine with electron ionization mass spec- Contract/grant sponsor: The Natural Science Foundation ofChina; contract/grant number: 10705026 and 10675112.
trometry (EI-MS) even at low electron energy, and hardly any Contract/grant sponsor: The Ministry of Science and Technol- molecular ion signal is obtained.20 In addition, with IR LD/ ogy of China; contract/grant number: 2007CB815204.
VUV PIMS the ionization energies (IEs) of molecules and the Contract/grant sponsor: The China Postdoctoral ScienceFoundation; contract/grant number: 20070410793.
appearance energies (AEs) of fragment ions can be accurately Copyright # 2008 John Wiley & Sons, Ltd.
obtained by scanning the photon energy. This provides a high-order harmonic radiation. The average photon flux can method of distinguishing individual components and isomers reach a magnitude of 1013 photons/s. A silicon photodiode in complex mixtures such as gasoline21 or intermediates in (SXUV-100, International Radiation Detectors Inc., Torrance, combustion processes22–24 by measuring photoionization CA, USA) was used to monitor the photon flux for efficiency (PIE) spectra. In the application of IR LD/VUV PIMS to drug analysis, we believe that mixtures of drug candi- To confirm the structural assignments and to propose dates or metabolites can be identified by determining molecular fragmentation pathways, EI (electron energy 70 eV, trap weights (molecular ions and fragments) and corresponding IEs current 10 mA) accurate mass measurements for the com- and AEs without preliminary separation, and that this may be a pounds were obtained using a GCT TOF mass spectrometer complementary strategy in drug discovery.
(Micromass, Manchester, UK). The source temperature was The purpose of this work is to utilize IR LD/VUV PIMS to set at 2208C and the samples were volatilized from a heated investigate the photo-induced fragmentation mechanisms of direct insertion probe in the source. The instrument was three drug compounds: captopril, fudosteine and racecado- calibrated at a mass resolution of 8000 (FWHM) using tril. The drugs mainly exclusively give molecular ions via heptacosafluorotributylamine as internal reference and the near-threshold ’soft’ photoionization, while a variety of single point lock-mass was at m/z 218.9856. Sample analysis, fragment ions can be formed by increasing the photon exact mass measurements and elemental composition energy. The structural assignments of the fragment ions are determination were performed automatically using the supported by results from a commercial EI-TOFMS instru- OpenLynx software within MassLynx (Micromass).
ment. Fragmentation pathways for the three compounds areproposed and discussed in detail.
ChemicalsRacecadotril was obtained from the Jiangsu Yangtze RiverPharmacy Group Co. (Taizhou, China). Fudosteine was supplied by the Yue Kang Co. Ltd. (Beijing, China). Captopril was obtained from the Shangdong Lukang Pharmaceutical Co.
The experiments were carried out at National Synchrotron Ltd. (Jining, China). All the drugs were deposited onto the Radiation Laboratory (NSRL) at Hefei, China. The apparatus stainless steel substrate without any preparation and purifi- has been reported in detail elsewhere.19 In brief, the cation. No organic matrix was used for these experiments.
experiments utilized the 1064 nm output of a pulsed Nd:YAGlaser (Surelite I-20, Continuum Inc., Santa Clara, CA, USA) with a duration of 7 ns for IR laser desorption. The laser pulsewas focused onto a stainless steel substrate with a 40-cm focal Photoionization mass spectra of captopril, fudosteine and length lens, and the central spot was kept at around 1 mm in racecadotril at different photon energies (below 10.5 eV) diameter. The laser power density at the surface of the were measured and are displayed in Figs. 1, 3 and 4. Due to substrate was controlled at 1.82 Â 109 W/cm2 (10 mJ/pulse)to generate intact neutral molecules for near-threshold VUVphotoionization. The VUV light beam is perpendicular to,and overlapping with, the desorption plume in the photo-ionization region. VUV photoionization takes place at adistance of 2–4 mm from the substrate surface, where theplume of molecules formed from the desorption processdisperses and is ionized by VUV light. Ions produced byVUV light were analyzed by a home-made reflectron TOFmass spectrometer with a mass resolution of $1400 (FWHM),which is not high enough to give accurate mass determi-nation. A pulsed voltage of 260 V applied to the repellerplates was used to propel ions into the flight tube. The pulsedvoltage with a frequency of 10 kHz works with a delay of150 ms after the laser fires with a frequency of 10 Hz, with thedelay being controlled by a home-made pulse/delaygenerator. The pressure in the photoionization chamberwas around 1.0 Â 10À4 Pa.
Synchrotron radiation from an undulator beamline of the 800 MeV electron storage ring at the NSRL was mono-chromatized with a 1 m Seya-Namioka monochromatorequipped with a laminar grating (1500 grooves/mm, HoribaJobin Yvon, Longjumeau, France). The grating covers thephoton energy from 7.8 to 24 eV. The monochromator wascalibrated with the known ionization energies of the inertgases. The energy resolving power (E/DE) is about 1000. A Figure 1. Photoionization mass spectra of captopril at gas filter filled with argon or helium was used to eliminate photon energies of (a) 9.0 and (b) 10.5 eV.
Copyright # 2008 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2008; 22: 2515–2520 Table 2. Relative signal intensities of major fragment ions inthe mass spectra of fudosteine Figure 2. Photoionization efficiency spectrum of the capto- EI: electron ionization; PI: photoionization.
For EI, a relative intensity of C2H5S (m/z 61) is assigned to be 100%(see Figure S2 of the Supporting Information).
broad tunability of synchrotron radiation, all the analytes canbe ionized with limited or no fragmentation via near-threshold single-photon ionization (SPI). Fragment ions, after myocardial infarction.25 GC/MS, LC/MS and LC/MS/ which are helpful in the identification of molecular MS methods have been applied for the determination of structures, are selectively yielded by increasing the photon energy. The major fragmentation pathways are described in The photoionization mass spectra of captopril at photon Schemes 1–3. Accurate EI mass measurements were per- energies of 9.0 and 10.5 eV are shown in Fig. 1. The 9.0 eV formed to support the proposed assignments of the photon energy, which is slightly higher than the ionization fragmentation pathways based on IR LD/VUV PIMS. The energy of captopril, yields the molecular ion (Mþ) signal at full EI mass spectra of three drug compounds are provided m/z 217 accompanied by minor fragment ions. The ionization in the Supporting Information. More fragment ions were energy of captopril was determined to be 8.69 Æ 0.05 eV via yielded at an electron energy of 70 eV. The relative signal the measurement of the PIE spectrum, as shown in Fig. 2, intensities of the major fragment ions of the drugs are which is obtained by plotting the integrated peak for m/z 217 summarized in Tables 1–3. Considerable differences for the versus the corresponding photon energy.
relative abundance of the fragment ions are found from Fragment ions can be formed by increasing the photon the SPI and EI spectra at different excitation energies.
energy. As shown in Fig. 1(b), an ion at m/z 173 is produced,which originates from the direct loss of carbon dioxide fromthe molecular ion by C–C bond fission, as described by Salem et al.25 Some relatively high-intensity fragment ions at Captopril, (S)-1-(3-mercapto-2-methyl-1-oxopropyl)-L-proline, m/z 199, 184, 140, 131 are formed at a photon energy of is a specific competitive inhibitor of angiotensin-convertingenzyme, and it is generally used for the treatment ofhypertension, heart failure, and left ventricular dysfunction Table 3. Relative signal intensities of major fragment ions inthe mass spectra of racecadotril Table 1. Relative signal intensities of major fragment ions inthe mass spectra of captopril EI: electron ionization: PI: photoionization.
EI: electron ionization; PI: photoionization.
For EI, a relative intensity of C4H8N (m/z 70) is assigned to be 100% For EI, a relative intensity of C7H7 (m/z 91) is assigned to be 100% (see (see Figure S1 of the Supporting Information).
Figure S3 of the Supporting Information).
Copyright # 2008 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2008; 22: 2515–2520 Scheme 1. Proposed fragmentation pathways for the captopril cation.
10.5 eV. The ion at m/z 199 is assigned to the loss of water after loss of ammonia from Mþ after intramolecular hydrogen an intramolecular hydrogen transfer process, as shown in transfer, which is a similar pathway to the CID behavior of Scheme 1. A relatively low-intensity fragment ion at m/z 171 the fudosteine derivative bis-methylated carbocysteine in is due to the subsequent loss of CO from m/z 199. Loss of the ESI-MS.32 The low-intensity ion at m/z 134 is due to the loss of mercapto radical (SH) from Mþ results in the fragment ion at a hydrocarboxyl radical from Mþ. Cysteine is known to m/z 184. Intramolecular hydrogen transfer and subsequent undergo the loss of a hydrocarboxyl radical to yield its most decarboxylation processes then occur with the loss of carbon abundant fragment ion.31 This discrepancy between the high dioxide from m/z 184 to yield the ion at m/z 140. Another intensity of the fragment ion in cysteine and the low intensity major pathway for formation of the m/z 140 ion occurs with of the m/z 134 ion in fudosteine by both PI and EI may come consecutive loss of carbon dioxide and the mercapto radical from the interference of the –(CH2)3OH terminal group of from the molecular ion. The ion at m/z 131 may be a dissociation product of m/z 173 by cyclic cleavage with Application of VUV photoionization also results in a main concomitant neutral loss of propene. A schematic repres- fragment ion at m/z 105, which is reasonably assigned to the entation of the major fragmentation pathways is given in loss of the CHNH2COOH radical from the Mþ ion. This ion FudosteineFudosteine, (–)-(R)-2-amino-3-(3-hydroxypropylthio)propio-nic acid, is used as a new muco active agent with indicationfor chronic respiratory diseases.28,29 Due to its poor retentionon LC columns, and to its lack of UV absorption andfluorescent functional groups, fudosteine has to be deriva-tized before LC/ESI-MS analysis.29 As shown in Fig. 3(a), fudosteine gives a molecular ion (Mþ) signal at m/z 179 accompanied by an intense m/z 75 ionat a photon energy of 8.5 eV, which is just a little higher thanits IE value. This dominant fragment ion is produced by theloss of a carbene derivative HOðCH Þ cleavage after intramolecular hydrogen transfer from thecarbon atom to the nitrogen atom. The high abundance of them/z 75 ion is due to the high stability of two resonancestructures. The appearance energy of this fragment ion isclose to the IE of fudosteine, indicating that the molecular ion(m/z 179) is not stable. This dominant dissociation pathwaywas also found for cysteine by EI and collision-induceddissociation (CID).30,31 As depicted in Fig. 3(b), in addition to the most abundant fragment ion at m/z 75, other fragment ions are formed atm/z 162, 137, 105, 92, 91, 87, 57, etc., with different yields at aphoton energy of 9.5 eV. Several probable dissociation Figure 3. Photoionization mass spectra of fudosteine at pathways of the fudosteine molecular ion are described in photon energies of (a) 8.5 and (b) 9.5 eV. Part of the spectrum Scheme 2. The m/z 162 ion would be expected to be formed by is amplified by a factor of 3, labeled in the figure.
Copyright # 2008 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2008; 22: 2515–2520 Scheme 2. Proposed fragmentation pathways for the fudosteine cation.
further dissociates with the loss of water to yield the m/z 87 Application of IR LD/VUV PIMS at a photon energy of ion. Simple S–C bond fission leads to the production of the 8.7 eV results in the formation of a molecular ion at m/z 385 fragment ion at m/z 92. The b-alanine cation (m/z 89) can also without any obvious fragment ions. A higher photon be detected from this bond cleavage pathway.33 A second energy (10.0 eV) gives a series of fragment ions, as shown formation pathway of m/z 92 can be initiated by consecutive neutral losses of ammonia and propiolic acid (CHCCOOH) The racecadotril cation (Mþ) eliminates neutral ketene as from Mþ. The ion at m/z 58 is assigned to 2-propenol which one of the initial fragmentation pathways to yield the ion at results from the loss of hydrogen sulfide from the m/z 92 ion.
m/z 343. Subsequently, the m/z 310 ion with a terminal –CH––CH2 double bond is yielded from C–S bond fission of the m/z 343 ion and/or Mþ. At such a photon energy, Mþ readilyeliminates a CH Racecadotril, N-[(R,S)-3-acetylmercapto-2-benzylpropanoyl]- abundant fragment ion at m/z 296. A relative low signal at glycine benzyl ester, is used as an anti-diarrhoea drug. In m/z 294 is reasonably attributed to the loss of a benzyl radical.
peripheral tissues, it is rapidly hydrolyzed to the more potent The fragment ion at m/z 205 may originate from the loss of a C7H5 group from the m/z 294 ion, although this dissociationpathway is unfavorable for most even-electron ions. Thefragment ion at m/z 130 probably originates from the loss ofSHCOCH3 from the m/z 205 ion, as shown in Scheme 3.
Three drugs, captopril, fudosteine and racecadotril, havebeen investigated by using IR laser desorption combinedwith tunable synchrotron VUV photoionization massspectrometry. The mass spectra for these drugs have beenobtained at different photon energies. Captopril andracecadotril are found to form exclusively molecular ions atphoton energies near their respective ionization thresholds.
Fudosteine, however, is prone to dissociation even near itsionization threshold, indicating its unstable nature. Thefragmentation pathways of these drugs in the low photonenergy range have been discussed in detail. The structuralassignments of the fragment ions are supported by accurateEI-TOFMS measurements. The relative intensities of thefragment ions of the drug compounds obtained fromphotoionization have been compared with those from Figure 4. Photoionization mass spectra of racecadotril at 70 eV electron ionization. The results indicate that IR laser photon energies of (a) 8.7 and (b) 10.0 eV. Part of the spec- desorption/tunable VUV PIMS could be a complementary trum is amplified by a factor of 6, labeled in the figure.
Copyright # 2008 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2008; 22: 2515–2520 Scheme 3. Proposed fragmentation pathways for the racecadotril cation.
18. Ferge T, Muhlberger F, Zimmermann R. Anal. Chem. 2005; 77: Additional supporting information may be found in the 19. Pan Y, Zhang TC, Hong X, Zhang YW, Sheng LS, Qi F. Rapid Commun. Mass Spectrom. 2008; 22: 1619.
20. Ptasinska S, Candori P, Denifl S, Yoon S, Grill V, Scheier P, Mark TD. Chem. Phys. Lett. 2005; 409: 270.
21. Wang J, Yang B, Li YY, Tian ZY, Zhang TC, Qi F, Nakajima K.
This work is supported by the Chinese Academy of Sciences Int. J. Mass Spectrom. 2007; 263: 30.
22. Cool TA, Nakajima K, Mostefaoui TA, Qi F, Mcllroy A, (YZ200764), the Natural Science Foundation of China Westmoreland PR, Law ME, Poisson L, Peterka DS, Ahmed (10705026 and 10675112), the Ministry of Science and Tech- nology of China (2007CB815204), and the China Postdoctoral 23. Taatjes CA, Hansen N, Mcllroy A, Miller JA, Senosiain JP, Klippenstein SJ, Qi F, Sheng L, Zhang Y, Cool TA, Wang J, Westmoreland PR, Law ME, Kasper T, Kohse-Hoinghaus K.
Science 2005; 308: 1887.
24. Berkenkamp S, Kirpekar F, Hillenkamp F. Science 1998; 281: 1. Papac DI, Shahrokh Z. Pharm. Res. 2001; 18: 131.
25. Salem II, Saif WA, Jmeian Y, Al Tamimi JI. J. Pharm. Biomed.
2. Cheng XH, Hochlowski J. Anal. Chem. 2002; 74: 2679.
3. Kassel DB. Chem. Rev. 2001; 101: 255.
26. Wong KK, Lan SJ, Migdalof BH. Biochem. Pharm. 1981; 30: 4. Geoghegan KF, Kelly MA. Mass Spectrom. Rev. 2005; 24: 347.
5. Hsieh Y, Merkle K, Wang G, Brisson JM, Korfmacher WA.
27. Franklin ME, Addison RS, Baker PV, Hooper WD.
6. Matuszewski BK, Constanzer ML, Chavez-Eng CM. Anal.
28. Takahashi K, Kai H, Mizuno H, Koda T, Miyata T. J. Pharm.
7. Cox KA, Clarke N, Ridgen D, Korfmacher WA. Am. Pharm.
29. Xu FG, Zhang ZJ, Jiao HY, Tian Y, Zhang BB, Chen Y. J. Mass 8. Robb DB, Covey TR, Bruins AP. Anal. Chem. 2000; 72: 3653.
30. Linstrom PJ, Mallard WG. NIST Chemistry Webbook, Number 9. Yang CM, Henion J. J. Chromatogr. A 2002; 970: 155.
69. National Institute of Standards and Technology: 10. Cai Y, Kingery D, McConnell O, Bach AC II. Rapid Commun.
Gaithersburg, MD, 2005. Available: http://webbook.nist.- 11. Tembreull R, Lubman DM. Anal. Chem. 1986; 58: 1299.
31. Rafii M, Elango R, Courtney-Martin G, House JD, Fisher L, 12. Bucheli TD, Haefliger OP, Dietiker R, Zenobi R. Anal. Chem.
Pencharz PB. Anal. Biochem. 2007; 371: 71.
32. Chen XY, Zhong DF, Han Y, Xie ZY. Rapid Commun. Mass 13. Engelke F, Hahn JH, Henke W, Zare RN. Anal. Chem. 1987; 33. Bi HC, Zhao LZ, Zhong GP, Zhou SF, Li B, Deng Y, Chen X, 14. Fye JL, Nelson HH, Mowery RL, Baronavski AP, Callahan Huang M. Rapid Commun. Mass Spectrom. 2006; 20: 1153.
34. Lecomte JM, Costentin J, Vlaiculescu A, Chaillet P, Marcais- 15. Nir E, Hunziker HE, de Vries MS. Anal. Chem. 1999; 71: 1674.
Collado H, Llorens-Cortes C, Leboyer M, Schwartz JC.
16. Chen YF, Sullards MC, Hoang TT, May SW, Orlando TM.
J. Pharmacol. Exp. Ther. 1986; 237: 937.
35. Xu Y, Huang JC, Liu F, Gao S, Guo QX. J. Chromatogr. B 2007; 17. Haefliger OP, Zenobi R. Anal. Chem. 1998; 70: 2660.
Copyright # 2008 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2008; 22: 2515–2520

Source: http://flame.nsrl.ustc.edu.cn/source/publication/2008/Rapid%20Commun.%20Mass%20Spectrom.%202008;%2022%202515-2520.pdf