On-Line LC−MS Analysis of Urinary Porphyrins - ACS Publications

Xavier Ausio´ ,† Joan O. Grimalt,*,† Dolors Ozalla,‡ and Carme Herrero‡ ... of Dermatology (IDIBAPS-Hospital Clınic), Villarroel, 170, 08036...
0 downloads 0 Views 55KB Size
Anal. Chem. 2000, 72, 4874-4877

On-Line LC-MS Analysis of Urinary Porphyrins Xavier Ausio´,† Joan O. Grimalt,*,† Dolors Ozalla,‡ and Carme Herrero‡

Department of Enviromental Chemistry (ICER, CSIC), Jordi Girona, 18, 08034-Barcelona, Catalonia, Spain, and Department of Dermatology (IDIBAPS-Hospital Clı´nic), Villarroel, 170, 08036 Barcelona, Catalonia, Spain

A reversed-phase high-performance liquid chromatography-mass spectrometry (LC-MS) method is described for the separation and simultaneous analysis of porphyrins related to disorders of heme biosynthesis (uro-, heptacarboxylic, hexacarboxylic, pentacarboxylic, and coproporphyrins). The method involves initial porphyrin esterification and extraction from urine. Detection and quantification is performed from the extracts by separation with a Hypersil BDS column and on-line detection by MS through coupling with an atmospheric pressure chemical ionization interface. The porphyrin esters are detected as protonated molecules [M + H]+. Their mass spectra also exhibit an [M + Na]+ fragment of lower intensity. The analytical performance of this method is compared with those of LC with UV and fluorescence detection. LC-MS used in selective [M + H]+ ion monitoring provides the lowest detection and quantitation limits. In scan mode, this LC-MS method affords, without further isolation or concentration steps, the measurement of mass spectra of unknown compounds present in the urine of patients with altered porphyrin excretion. Porphyrias are disorders of heme biosynthesis due to the excessive production, accumulation, and excretion of porphyrins and their precursors. Each type of porphyria is caused by specific enzyme deficiencies. Thus, porphyria cutanea tarda (PCT), the most common of the human porphyrias, results from partial deficiency of uroporphyrinogen descarboxylase producing the accumulation of acetyl porphyrins (uroporphyrins and heptacarboxylporphyrins). These liver insufficiencies are usually inherited but they may also be induced by exposure to environmental toxic chemicals.1-4 They exist in clinically manifest and asymptomatic stages which can be recognized on the basis of the amount and characteristic pattern of urinary porphyrin excretion.1,5 Several techniques are available for the analysis of porphyrins in biological samples, e.g., spectrophotometry,6,7 fluorometry,8,9 †

ICER, CSIC. ‡ IDIBAPS-Hospital Clı´nic. (1) Herrero, C.; Ozalla, M. D.; Sala, M.; Otero, R.; Santiago-Silva, M.; Lecha, M.; To-Figueras, J.; Deulofeu, R.; Mascaro, J. M.; Grimalt, J. O.; Sunyer, J. Arch. Dermatol. 1999, 135, 400-404. (2) Elder G. H. In Recent advances in dermatology; Champion, R. H., Pye, R. J., Eds.: Churchill Livingstone: Edinburgh, 1990; pp 55-69. (3) Longnecker, M. P.; Rogan, W. J.; Luier, G. Annu. Rev. Public Health 1997, 18, 211-244. (4) Woods, J. S.; Bowers, M. A.; Holly, A. D. Toxicol. Appl. Pharmacol. 1991, 110, 464-476. (5) Lim, C. K.; Peters, T. J. Clin. Chim. Acta 1984, 139, 55-63. (6) Jones, K. G.; Sweeney, G. D. Clin. Chem. 1979, 25, 71-74.

4874 Analytical Chemistry, Vol. 72, No. 20, October 15, 2000

Table 1. Solvent Elution Program for LC time (min)

% Aa

% Ba

0 2 15 20 27 32

25 25 12.5 15 15 25

75 75 85.5 85 85 75

a (A) acetonitrile/ammonium acetate (1 M pH 5.16; 1:9, v/v); (B) acetonitrile/methanol (1:9, v/v).

thin-layer chromatography,10,11 and high-performance liquid chromatography (LC).5,12-14 Mass spectrometric (MS) techniques have also been reported, namely, fast atom bombardment MS15,16 and liquid secondary ion MS.17 The MS techniques provide structural information on isolated porphyrin compounds. However, combination to LC would potentially afford a robust method for the simultaneous identification and quantification of porphyrin composition in porphyria patients. The present study describes a first combined LC-MS method based on coupling with the atmospheric pressure chemical ionization interface (APCI). Its analytical performance is compared to the more conventional techniques for the analysis of porphyrin composition in urine samples such as LC with UV absorption and UV fluorescence detection, LC-UV and LC-FL, respectively. The method has been applied to the analysis of human urine samples of PCT patients. EXPERIMENTAL SECTION Materials. Type I uroporphyrin, heptaporphyrin, hexaporphyrin, pentaporphyrin, and coproporphyrin methyl ester standards were obtained from chromatographic marker kit vials containing 10 nmol each (Porphyrin Products; Logan, UT). Liquid chromatography grade water, methanol, and acetonitrile were from Merck (7) Buttery, J. E.; Chamberlain, B. R.; Gee, D.; Pannall, P. R. Clin. Chem. 1995, 41, 103-106. (8) Polous, V.; Lockwood, W. Int. J. Biochem. 1980, 12, 1051-1052. (9) Lockwood, W.; Polous, V.; Rossi, E. Clin. Chem. 1985, 31, 1164-1167. (10) Smith, S. G. Br. J. Dermatol. 1975, 93, 291-294. (11) Henderson, M. J. Clin. Chem. 1989, 35, 1043-1044. (12) Gray, C. H.; Lim, C. K.; Nicholson, D. C. Clin. Chim. Acta 1977, 77, 167178. (13) Lim, C. K.; Rideout, J. R.; Wright, D. J. Biochem. J. 1983, 137, 435-438. (14) Lim, C. K.; Famei Li; Peters, T. J. J. Chromatogr. 1988, 429, 123-153. (15) Luo, J.; Lamb, J. H.; Lim, C. K. J. Pharmacol. Biomed. Anal. 1997, 15, 12891294. (16) Guo, R.; Rideout, J. M.; Lim, C. K. Biochem. J. 1989, 264, 293-295. (17) Luo, J.; Lim, C. K. Biomed. Chromatogr. 1995, 9, 113-122. 10.1021/ac0005060 CCC: $19.00

© 2000 American Chemical Society Published on Web 09/21/2000

Figure 1. LC-MS chromatogram of a mixture of type I porphyrin methyl ester standards: A, uroporphyrin octamethyl ester; B, heptaporphyrin heptamethyl ester; C, hexaporphyrin hexamethyl ester; D, pentaporphyrin pentamethyl ester; E, coproporphyrin tetramethyl ester.

(Darmstadt, Germany). Analysis grade concentrated sulfuric acid, glacial acetic acid, and ammonium acetate were also from Merck. Analysis grade chloroform was from Carlo Erba Reagenti (Milan, Italy). Esterification of Porphyrins in Urine. Urine (3 mL) was mixed with 25 mL of methanol/concentrated H2SO4 (90:10 v/v) and left to stand 24 h in the dark at room temperature. The resulting porphyrin methyl esters were extracted with chloroform, which was subsequently washed with saturated sodium hydrogencarbonate and water. The extracts were purified by passage through Pasteur pipets filled with alumina for column chromatography. Porphyrins were eluted with chloroform/methanol (1:1 v/v). The resulting solutions were evaporated to dryness under N2 and redissolved in acetonitrile (100 µL). A 20-µL aliquot of this solution was used for the LC analyses. Instrumental Analysis. LC-UV and LC-MS were carried out with a Hewlett-Packard model 1090A instrument provided with a HP 1040 diode array UV/visible detector (set at 405 nm) serially coupled to a LC/MSD HP 1100. Chromatographic separation was performed with a Hypersil BDS C18 column (250 × 4.6 mm; 5-µm particle diameter) protected with a Hypersil BDS guard column (10 × 4 mm) (Shamdom Scientific Ltd.; Chesire, U.K.). The solvent elution program is reported in Table 1. Flow rate was 1 mL min-1. MS was used in positive ion mode by scanning between m/z 700 and 1000 per second or by selective ion monitoring at m/z 943. APCI vaporizer and drying gas temperatures were 475 and 300 °C, respectively. Drying gas flow was 4 L min-1. Nebulizer pressure was 60 psig. Fragmenter and capillary voltages were 70 and 3000 V, respectively. Corona current intensity was 10 µA. LC-FL was performed with a Perkin-Elmer system equipped with a binary high-pressure pump (model. 250) and a LC 240 FL detector (Buckimghamshire, U.K.). Porphyrins were recorded at 405 and 618 nm for absorption and emission, respectively. The chromatographic column and solvent elution program were the same as described above. RESULTS AND DISCUSSION The LC-MS method described in this study requires the methylation of the free carboxylic groups present in the urine porphyrins. Good recoveries are only obtained when significant excess of methylation agent is used in relation to the amount of urine porphyrins. Treatment with reagent amounts lower than those indicated in the Experimental Section resulted in drastic yield losses.

Figure 2. Mass spectra of type I porphyrin methyl ester standards obtained by LC-MS using an APCI interface. Letters refer to the same compounds as in Figure 1.

This methylation step is needed to generate compounds providing a LC-MS signal of an intensity comparable to those observed by LC-UV and LC-FL. Sensitivity was very poor if the porphyrins were analyzed in their original form, as free carboxylic acid groups, either in positive or negative ion mode. Very poor sensitivity for the free acidic compounds was also observed in LC-MS couplings using other interfaces, e.g., thermospray and electrospray. Separation of all five main porphyrins in methylated form is illustrated in Figure 1. Chromatographic Conditions. Porphyrin analysis as methyl derivatives is performed in conditions very similar to those required for the free carboxylic compounds.1,5 The same column and solvent mixtures are used in both cases (Table 1). The only difference concerns the elution program that in the case of the methyl derivatives ranged between 75 and 85.5% (v/v) of the more Analytical Chemistry, Vol. 72, No. 20, October 15, 2000

4875

Table 2. Repeatibility and Reproducibility (%) of Porphyrin Standards Analyzed with Several LC Configurations

uroporphyrin I octamethyl ester heptaporphyrin I heptamethyl ester hexaporphyrin I hexamethyl ester pentaporphyrin I pentamethyl ester coproporphyrin I tetramethyl ester a

MS SIM

MS scan

MS SIM (spiked samples)

UV

FL

5.8a

4.7 5.8 5.4 5.3 6.7 6.9 6.2 6.6 7.2 8.7

6.2 6.8 5.5 6.7 4.0 5.1 3.8 4.0 2.7 3.2

1.2 1.3 0.56 1.4 1.0 1.4 1.4 2.8 2.1 2.8

3.6 3.8 3.3 5.2 3.6 5.5 4.0 5.0 4.6 5.9

6.8b 5.6 6.1 4.8 5.7 4.2 5.5 2.4 6.8

Repeatibility. b Reproducibility.

apolar mixture (acetonitrile/methanol, 1:9, v/v) (Table 1) whereas the range of this solvent is 0-65% in the program of the free carboxylic acid porphyrins.1 Both solvent elution programs afford the elution of all porphyrins of interest, from uroporphyrin to coproporphyrin, in 30 min. However, uroporphyrin octamethyl esters I and III and heptacarboxylporphyrin heptamethyl esters I and III cannot be separated in the more apolar program. These compounds are also coeluting in the form of free carboxylic acids upon analysis with the more polar solvent program.1,5 The solvent program developed in the present study is used for comparison of all above-mentioned LC setups. No significant differences in terms of detection properties, e.g., sensitivity or linear range, are observed between LC-UV and LC-FL upon comparison of the methylated and nonmethylated methods. Optimization of the APCI-MS conditions. Sensitivity optimization was performed by flow injection of a uroporphyrin I octamethyl ester standard (10 nmol mL-1) in the positive ion mode. Thus, the two eluents, (A) 1 M ammonium acetate (pH 5.16)/acetonitrile (9:1 v/v) and (B) methanol/acetonitrile (9:1 v/v), were combined in a proportion of 30:70 (v/v). Peaks were detected in scan mode (m/z 700-1000 every second). The nebulizer pressure depends on column flow. Flow rates equal or higher than 800 µL min-1 require nebulizer pressures varying from 45 to 60 psi. The optimum value was found at 60 psi. The drying gas flow was evaluated in the range of 3-5 L min-1, and maximum response was found at 4 L min-1. Optimal drying gas temperature was 300 °C (tested range 250-350 °C). The maximum standard signal was found at a vaporizer temperature of 475 °C. The response of the analyte signal was also significantly dependent on corona current. The highest values were found at 10 µA for a test range between 4 and 12 µA. Capillary voltage was monitored from the m/z 943 ion, [M + H]+, in a range of 2500-5000 V; maximum response was found at 3000 V. In these conditions, the sodium adduct [M + Na]+, m/z 965, was also observed. These two ions are those dominating the mass spectra of all isomers, from uroporphyrin to coproporphyrin (Figure 2). Analytical Properties. LC-MS in scan (m/z 700-1000 every second) and selected ion monitoring modes were examined for linearity using mixtures of type I porphyrin ester standards in the 0.1-5 nmol mL-1 range. SIM mode involved the use of the [M + H]+ ions of the main porphyrins, m/z 711, 769, 827, 885, and 943. The calibration curve was linear in this range, with correlation coefficients of 0.9984-0.9994 and 0.9894-0.9951 in SIM and scan 4876 Analytical Chemistry, Vol. 72, No. 20, October 15, 2000

Figure 3. Comparison of the LC chromatograms of the methylated urine extract from an individual affected by porphyria cutanea tarda using different detectors (UV, FL, and MS in SIM and scan modes): 1, uroporphyrin I + III octamethyl ester; 2, unknown (see Figure 5); 3, heptaporphyrin III heptamethyl ester; 4, hexaporphyrin I hexamethyl ester; 5, hexaporphyrin III hexamethyl ester; 6, pentaporphyrin I pentamethyl ester; 7, pentaporphyrin III pentamethyl ester; 8, coproporphyrin I tetramethyl ester; 9, coproporphyrin III tetramethyl ester.

modes, respectively. These correlation coefficients were similar to those observed for LC-UV and LC-FL when calibrated with the same standards for a concentration range between 0.33 and 10 nmol mL-1. The coefficients of LC-UV, 0.9990-0.9998, were slightly higher than those obtained in SIM mode and those of LC-FL, 0.9977-0.9987, were intermediate between those in scan and SIM modes.

Limits of detection and quantification were calculated by analysis of the type I porphyrin standards in the concentration range of 1-330 pmol mL-1. These limits were determined from signal/noise ratios of 3 and 5, respectively, and calculated by reference to the analysis of 3 mL of urine, the sample volume analyzed currently. The lowest detection and quantification limits were those of LC-MS in the SIM mode, 0.07 pmol mL-1 for the detection limit of all compounds and 0.2 pmol mL-1 for the quantification limit of uro- to pentaporphyrins (0.4 pmol mL-1 for coproporphyrin). These limits are even lower than those observed for LC-FL, 0.2 and 0.4 pmol mL-1 for detection and quantification, respectively. As expected, the detection and quantification limits of LC-UV, 0.4 and 0.7 pmol mL-1, respectively, are higher than those of LC-FL but still lower than those of LC-MS in scan mode, 3 and 7 pmol mL-1, respectively. Repeatibility and reproducibility were calculated from replicate analysis of the mixture of type I porphyrin standards at a concentration of 3.3 nmol mL-1 (Table 2). In the case of LC-MS in SIM mode, one standard spiked sample was also used for comparison. The repeatability and reproducibility values for LCMS are in the order of 2-6 and 5-8% for LC-MS in SIM and scan modes, respectively. No major differences are observed when using standard solutions or spiked samples for the calculations. The LC-MS SIM values are similar to those of LC-FL, 3-6% and higher than those of LC-UV, 0.6-3%. Analysis of Urine Samples. The chromatograms corresponding to one urine sample from a PCT patient analysed by all the methods considered in this study are shown in Figure 3. LCMS is the technique providing better defined peaks for the minor compounds, which is consistent with its low limits of detection and quantification. The dominance of uroporphyrins and heptacarboxylporphyrin of this sample is in agreement with those previously reported for individuals with PCT patterns.1 This composition strongly contrasts with porphyrin distributions in the regular population, which are dominated by coproporphyrins I and III (Figure 4). LC-MS in the scan mode allows the mass spectral identification of the porphyrins found in these chromatograms. The mass spectra obtained from the samples included in these examples are identical to those reported in Figure 2. This technique also allows one to obtain mass spectra of unknown compounds that are significant, for instance, in the PCT profiles. One of these compounds is shown in Figure 5 and corresponds to peak 2 in the chromatograms of Figure 2. Its base peak, m/z 769, is equivalent to [M + H]+ of a pentacarboxylporphyrin pentamethyl ester but elutes at an earlier retention time. In addition, the mass spectrum exhibits a m/z 726 fragment corresponding to a loss of 43 mass units. Both relative retention and secondary fragment agree with a structure of hexacarboxylporphyrin hexamethyl ester in which one propionate group is substituted by a formaldehyde. This aldehyde function could be stabilized by conjugation of the carbonyl double bond with the unsaturated system of the porphyrin ring. However, the absolute identification of this compound requires further work, which is beyond the scope of the present study. CONCLUSIONS LC-MS interfaced with APCI is a viable method for the direct analysis of porphyrins related to disorders of heme biosynthesis

Figure 4. Examples of LC-MS (scan mode) chromatograms of the methylated urine extracts from healthy (top) and ill (bottom) individuals. Peak identification like in Figure 3.

Figure 5. Mass spectra of compound 2 in the chromatograms shown in Figure 3.

in human urine mixtures upon derivatization as methyl esters. This technique, in the SIM mode, provide lower detection and quantitation limits than LC-UV and LC-FL. In terms of repeatability and reproducibility, it provides values similar to those observed for LC-FL. The MS of the porphyrins obtained from the urine mixtures are characterized by an [M + H]+ fragment as base peak and an additional [M + Na]+ ion of 5% relative intensity. The technique is also useful for providing mass spectral data on unknown porphyrin compounds present in the urine mixtures of porphyria patients. No additional isolation or concentration is needed for suitable measurement of these spectra by LC-MS in the scan mode. Received for review May 2, 2000. Accepted August 3, 2000. AC0005060 Analytical Chemistry, Vol. 72, No. 20, October 15, 2000

4877