Screening for Organic Phosphorus Compounds in Aquatic Sediments

Anal. Chem. 2008, 80, 6689–6697

Screening for Organic Phosphorus Compounds in Aquatic Sediments by Liquid Chromatography Coupled to ICP-AES and ESI-MS/MS Heidi De Brabandere,† Niklas Forsgard,† Lena Israelsson,† Jean Petterson,† Emil Rydin,‡ Monica Waldeba ¨ ck,† and Per J. R. Sjo¨berg*,† Department of Physical and Analytical Chemistry, Uppsala University, Box 599, 751 24 Uppsala, Sweden, and Department of Ecology and Evolution, Limnology, Uppsala University, Box 573, 751 23 Uppsala, Sweden The structures of organic phosphorous (P) compounds in aquatic sediments are to a large extent unknown although these compounds are considered to play an important role in regulating lake trophic status. To enhance identification of these compounds, a liquid chromatography (LC) method for their separation was developed. The stationary phase was porous graphitic carbon (PGC), and the mobile phases used in the gradient elution were compatible with both inductive coupled plasma atomic emission spectroscopy (ICP-AES) and electrospray ionization tandem mass spectrometry (ESI-MS/MS). With LC-ICP-AES, eight different P containing peaks could be observed in the P chromatogram indicating that at least eight different P compounds were separated. With the setup of an information dependent acquisition (IDA) with ESI-MS/MS, the mass over charge (m/z) of compounds containing a phosphate group (H2PO3-, m/z 97) could be measured and further fragmentation experiments gave additional information on the structure of almost 40 separated P compounds, several were verified to be nucleotides. ICP-AES was very suitable in the development of the LC method and allowed screening and quantification of P compounds. The presented LC-ESIMS/MS technique was able to identify several sediment organic P compounds. Phosphorus (P) is a key element for primary production in aquatic systems and thereby regulates trophic status and subsequently water quality. Although the primary source of P is from the drainage area (external loading of P), P available for primary production also relies on recycling of P, either within the water column or after sedimentation. Even when external P loading is decreased, the recycling of P stored in the sediment into the water column (internal loading) can sustain the P concentration in the water column. Overload of the nutrient P into an aquatic system is therefore a hard to reverse process and can cause long-term eutrophication problems. Sediment P partly consists of organic P * To whom correspondence should be addressed. [email protected]. † Department of Physical and Analytical Chemistry. ‡ Department of Ecology and Evolution, Limnology.

E-mail:

10.1021/ac8006335 CCC: $40.75  2008 American Chemical Society Published on Web 07/30/2008

compounds.1-3 However, the organic P compounds contribute largely to the recycling of P.4 Even so, identification of these organic P compounds has rarely been reported, due to the limitations of analytical techniques for extraction and identification of organic P forms in sediments. Studies focusing on organic P in sediments have in most cases only been able to identify groups of organic P compounds (i.e., phosphate monoesters, phosphate diesters, polyphosphates, pyrophosphate, orthophosphate), using 31 P-nuclear magnetic resonance (31P NMR).5-10 However, a few studies have focused entirely on individual organic P compounds. Suzumura and Kamatani11 extracted inositol phosphates from sediment from Tokyo Bay by hypobromite oxidation followed by isolation with anion exchange chromatography and analysis with 1 H NMR and gas chromatography (GC). The phospholipid phosphatidylcholine, isolated from sediment pore waters and extracts, has been analyzed with chemiluminescence via an enzymatic reaction.12 Additionally, several sediment phospholipids were identified by a flow-blending extraction combined with liquid chromatography electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS).13 To gain a deeper understanding of the sediment P turnover, it would be of interest to characterize and quantify as many as possible of all the individual organic P compounds present in sediment samples at different depth, not only phospholipids or inositole phosphates. Since existing methods focus either on groups of organic P compounds or on one specific P compound, (1) Reitzel, K.; Hansen, J.; Jensen, H. S.; Andersen, F. Ø.; Hansen, K. S Hydrobiologia 2003, 506-509, 781–787. (2) Suzumura, M.; Kamatani, A. Limnol. Oceanogr. 1995, 40, 1254–1261. (3) Organic Phosphorus in the Environment; CABI Publishing: Wallingford, U.K., 2005. (4) McKelvie, I. D. In Organic Phosphorus in the Environment; Turner, B. L., Frossard, E., Baldwin, D. S., Eds.; CABI Publishing: Wallingford, U.K., 2005; pp 1-20. (5) Hupfer, M.; Gachter, R.; Ru ¨ egger, H. Limnol. Oceanogr. 1995, 40, 610– 617. (6) Clark, L. L.; Ingall, E. D.; Benner, R. Nature 1998, 393, 426. (7) Paytan, A.; Cade-Menun, B. J.; McLaughlin, K.; Faul, K. L. Mar. Chem. 2003, 82, 55–70. (8) Hupfer, M.; Rube, B.; Schmieder, P. Limnol. Oceanogr. 2004, 49, 1–10. (9) Ahlgren, J.; Tranvik, L.; Gogoll, A.; Waldeback, M.; Markides, K.; Rydin, E. Environ. Sci. Technol. 2005, 39, 867–872. (10) Reitzel, K.; Ahlgren, J.; De Brabandere, H.; Waldeba¨ck, M.; Gogoll, A.; Tranvik, L. Biogeochem. 2007, 82, 15–28. (11) Suzumura, M.; Kamatani, A. Geochim. Cosmochim. Acta 1993, 57, 2197– 2202. (12) Amini, N.; McKelvie, I. Talanta 2005, 66, 445–452. (13) Zink, K. G.; Mangelsdorf, K. Anal. Bioanal. Chem. 2004, 380, 798–812.

Analytical Chemistry, Vol. 80, No. 17, September 1, 2008

6689

there is a need for analytical methods that determine and characterize several individual P containing compounds simultaneously. Mass spectrometry (MS) is a powerful tool to achieve this goal detecting ions with specific mass over charge ratios with high selectivity and sensitivity. In related studies, electrospray ionization (ESI) has been utilized as the ion source to study dissolved organic matter (DOM),14,15 organic phosphorus in natural waters,16,17 and phospholipids in sediments.13 The popularity of ESI is based on its “soft” ionization, allowing compounds to be observed with minimum fragmentation, which facilitates the characterization of entire compounds. Furthermore, with tandem mass spectrometry (MS/MS) it is possible to increase the detection selectivity, i.e., detect precursor ions of selected fragments after collision induced dissociation (CID). This technique was applied in the determination of phosphopeptides.18,19 A first attempt toward an overall identification with the use of ESI-MS/MS was reported in a previous study.20 A sample preparation method was designed to enable direct infusion of the sediment extract into an ESI-MS/MS instrument. The need for a separation step prior to mass spectrometry detection soon became obvious. A separation step can reduce the need of extensive sample preparation and sample complexity and can improve identification possibilities. Whereas ESI-MS/MS is an excellent technique for identification of unknown compounds, it is difficult to use when it comes to quantification. On the other hand, inductive coupled plasma atomic emission spectroscopy (ICP-AES) has proven to be an excellent technique for quantification of unknowns due to a high element selectivity where even inorganic standards can be used for quantification of organic compounds.21 The high selectivity also facilitates the development of a LC separation method. The combination of LC with ICP-AES could thus give a means of screening and quantifying organic P compounds in sediments from different aquatic environments or sediments of different ages. A comprehensive overview of different separation techniques applied in the natural organic phosphorus area can be found in Turner et al. 2005.4 Size exclusion was one of the first serious attempts for separation of organic phosphorus compounds. It was, e.g., used to proof the presence and importance of inositol phosphates in preconcentrated lake water.22 Ion exchange chromatography (IC) has been used for separation of P compounds in soil solutions23 and for separation of inositol phosphates in sediments by Suzumura and Kamatani 1993.11 Capillary electro(14) Seitzinger, S. P.; Hartnett, H.; Lauck, R.; Mazurek, M.; Minegishi, T.; Spyres, G.; Styles, R. Limnol. Oceanogr. 2005, 50, 1–12. (15) Kujawinski, E. B.; Del Vecchio, R.; Blough, N. V.; Klein, G. C.; Marshall, A. G. Mar. Chem. 2004, 92, 23–37. (16) Llewelyn, J. M.; Landing, W. M.; Marshall, A. G.; Cooper, W. T. Anal. Chem. 2002, 74, 600–606. (17) Cooper, W. T.; Llewelyn, J. M.; Bennett, G. L.; Salters, V. J. M. Talanta 2005, 66, 348–358. (18) Edelson-Averbukh, M.; Pipkorn, R.; Lehmann, W. D. Anal. Chem. 2006, 78, 1249–1256. (19) Huddleston, M. J.; Annan, R. S.; Bean, M. F.; Carr, S. A. J. Am. Soc. Mass Spectrom. 1993, 4, 710–717. (20) De Brabandere, H.; Danielsson, R.; Sjoberg, P. J. R.; Ahlgren, J.; Rydin, E.; Waldeback, M. Talanta 2008, 74, 1175–1183. (21) Svantesson, E.; Pettersson, J.; Markides, K. E. J. Anal. At. Spectrom. 2002, 17, 491–496. (22) Eisenreich, S. J.; Armstrong, D. E. Environ. Sci. Technol. 1977, 11, 497– 501.

6690


phoresis (CE) has also been used for determination of inositol phosphates.24 Gas chromatography is commonly used for determination of organophosphate pesticides.25 Partition chromatography such as normal phase chromatography has mainly been used for separation of phospholipids.13 In this project we opted for a stationary phase withstanding high pH since the sediments were extracted with the highly alkaline sodium hydroxide (NaOH) which is a well established extractant for P compounds from aquatic sediments.9,10,26 However, the fact that NaOH can induce hydrolysis of labile compounds, e.g., RNA (ribonucleic acid), should be considered when interpreting results.27 Porous graphitic carbon (PGC) met the requirement of withstanding high pH and has to our knowledge not been used in the determination of natural organic phosphorus compounds in the aquatic environment. In one report PGC has been used for separation of standard phosphonic acids.28 PGC with plasma spectrometry has also been used in speciation analysis of arsenic,29 selenium,30 boron containing drugs, and metabolites from boron neutron capture therapy patients31 and galliumchelated siderophores.32 The PGC material33 is composed of flat sheets of carbon atoms bound in a hexagonal arrangement and numerous publications describe its properties.34-36 Its special retention behavior for polar analytes has been ascribed to the polar retention effect on graphite35,36 and provides an alternative and attractive complement to IC.37 In this study, a method for separation and quantification of sediment organic P compounds was developed with LC-ICP-AES. The optimized LC method was then run against ESI-MS/MS for identification of the separated compounds. Separating organic P compounds extracted from the sediment prior to detection greatly decreases sample complexity and enables a better characterization of sediment organic P compounds. This is an important step toward the ultimate goal to identify all the organic P compounds in an aquatic sediment extract. EXPERIMENTAL SECTION Sampling. All sediment samples were taken in Lake Erken, a moderately eutrophic lake in Sweden. A detailed description of (23) Espinosa, M.; Turner, B. L.; Haygarth, P. M. J. Environ. Qual. 1999, 28, 1497–1504. (24) Henshall, A.; Harrold, M. P.; Tso, J. M. Y. J. Chromatogr., A 1992, 608, 413–419. (25) Serrano, R.; Lopez, F. J.; Hernandez, F. J. Chromatogr., A 1999, 855, 633– 643. (26) Ingall, E. D.; Schroeder, P. A.; Berner, R. A. Geochim. Cosmochim. Acta 1990, 54, 2617–2620. (27) Nelson, D. L.; Cox, M. M. Lehninger Principles of Biochemistry, 3rd ed.; Worth Publishers: New York, 2000. (28) Mercier, J. P.; Morin, P.; Dreux, M.; Tambute, A. J. Chromatogr., A 1999, 849, 197–207. (29) Mazan, S.; Cretier, G.; Gilon, N.; Mermet, J. M.; Rocca, J. L. Anal. Chem. 2002, 74, 1281–1287. (30) Dauthieu, M.; Bueno, M.; Darrouzes, J.; Gilon, N.; Potin-Gautier, M. J. Chromatogr., A 2006, 1114, 34–39. (31) Svantesson, E.; Capala, J.; Markides, K. E.; Pettersson, J. Anal. Chem. 2002, 74, 5358–5363. (32) Moberg, M.; Nilsson, E. M.; Holmstrom, S. J. M.; Lundstrom, U. S.; Pettersson, J.; Markides, K. E. Anal. Chem. 2004, 76, 2618–2622. (33) Gilbert, M.; Knox, J.; Kaur, B. Chromatographia 1982, 16, 138–146. (34) Ross, P. LCGC North Am. 2000, 18, 14. (35) Forgacs, E. J. Chromatogr., A 2002, 975, 229–243. (36) Hennion, M. C.; Coquart, V. J. Chromatogr., A 1993, 642, 211–224. (37) Hanai, T. J. Chromatogr., A 2003, 989, 183–196.

Lake Erken can be found in Ahlgren et al. 2007.38 The sediment samples were collected with a core sampler (Willner sampler) at a depth of 17 m within the same accumulation bottom area of the lake. In May 2007, two sediment cores were taken from this area. From the two cores, the 0-1 cm upper sediment layer and the 29-30 cm sediment layer were taken and pooled. The samples were kept at 4 °C during transport, and extraction was started 3 h after sampling. Chemicals. The following phosphate standards were purchased from Sigma-Aldrich (St. Louis, MO): adenosine 5′-monophosphate (5′-AMP), adenosine 3′,5′-cyclic monophosphate (cAMP), cytidine 5′-monophosphate (CMP), guanosine 5′-monophosphate (GMP), inosine 5′-monophosphate (IMP), uridine 5′-monophosphate (UMP), and adenosine 3′-monophosphate (3′-AMP). For extraction, ethylenediamine tetraacetic acid (EDTA, disodium salt, Merck, Switzerland) and sodium hydroxide (NaOH, p.a., EKA Bohus-Sweden) were used. LC mobile phases were prepared with acetonitrile (ACN, Merck, Germany), ammonium formate (HCO2NH4, AnalaR, England) and ammonium acetate (CH3CO2NH4, Merck, Germany). Disodiumhydrogen phosphate (Na2HPO4) was purchased from Fluka, Switzerland, and P element standard for atomic spectroscopy from SPECTRASCAN, Teknolab AS, Norway. Water was purified with a Milli-Q purification system (Millipore, Bedford, MA). Sample Preparation. In this study, EDTA, 0.067 M, was used as the pre-extractant and NaOH, 0.1 M, as the main extractant for sediment organic P according to Ahlgren et al. 2007.38 The wet sediment was in both steps mixed with the extractant in a 1:3 w/v ratio. Pre-extraction with EDTA was done for 30 min followed by a 16 h extraction with NaOH, both performed at room temperature. After pre-extraction with EDTA, sediment particles were collected by centrifugation (4302 relative centrifugal force (RCF), Sigma 4-15 Laboratory Centrifuges, Germany). The supernatant was discarded with no substantial P loss.38 This procedure also prevented EDTA from entering the ESI-interface which could cause ion suppression and obscure and complex the phosphate signal. NaOH was then added to the remaining sediment, and the extract was finally collected after centrifugation (4302 RCF). The extract was concentrated 10 times (from a 200 to 20 g sample) with rotary evaporation under reduced pressure and at ambient temperature. The use of rotary evaporation induces a P loss of 10%.39 Particles formed during the concentration step were removed by centrifugation at 10 000 RCF during 10 min (SigmaLabozentrifugen 2K15, Sigma 4-15 Laboratory Centrifuges, Germany). All samples were frozen until analysis since it has been shown that freezing extracts does not alter their P composition.8 Just before analysis, the samples were run through a 17 mm 0.45 µm PVDF (polyvinylidene difluoride, National Scientific Company) syringe filter. Solutions of 1.38 mM 5′-AMP, cAMP, CMP, GMP, IMP, UMP, and 3′-AMP were prepared in mobile phase A (HCO2NH4). A volume of 450 µL of sample was spiked with 4 µL of each of these standard solutions before mass spectral analysis. (38) Ahlgren, J.; De Brabandere, H.; Reitzel, K.; Rydin, E.; Gogoll, A.; Waldeba¨ck, M. J. Environ. Qual. 2007, 36, 892–898. (39) Ahlgren, J.; Reitzel, K.; Tranvik, L.; Gogoll, A.; Rydin, E. Limnol. Oceanogr. 2006, 51, 2341–2348.

Table 1. Gradient Used for Elution and Separation of P Compounds in a EDTA f NaOH Sediment Extract on a PGC Column step

time (min)

mobile phase A (%)

mobile phase B (%)

1 2 3 4 5 6 7

0 3 12.5 60 70 80 95

100 100 95 0 0 100 100

0 0 5 100 100 0 0

Sediment total P content was analyzed in duplicate after acid hydrolysis at 340 °C, according to Murphy and Riley.40 LC. The chromatographic system was an Agilent 1100 (Agilent Technologies, Waldbronn, Germany) consisting of binary pumps and autosampler. The column was a PGC Hypercarb (100 mm × 2.1 mm 5 µm, Hypersil, Cheshire, U.K.). Gradient elution was performed with mobile phases A which consisted of 10 mM HCO2NH4 in water (pH ) 10) with 1% ACN and mobile phase B which consisted of 10 mM HCO2NH4 in MQ (pH ) 10) with 80% ACN. These mobile phases are compatible with both ICP and ESI. The pump was programmed to deliver a gradient (see Table 1) at a flow rate of 200 µL/min. The chosen mobile phases and the gradient were compatible with both ICP and ESI. The injected sample volume was 250 µL when coupled to ICP-AES and 100 µL when coupled to ESI-MS/MS. Two other mobile phases A and B were prepared where 10 mM HCO2NH4 was exchanged to 50 mM CH3CO2NH4. ICP-AES. A Spectroflame P ICP-AES (Spectro Analytical Instruments, Kleve, Germany) equipped with a Meinhard nebulizer (J.E. Meinhard Associates Inc., Santa Ana, CA) followed by a Scott double-pass spray chamber, was used for the ICP-AES measurements. The outlet of the LC column was connected to the nebulizer via 50 cm PEEK tubing with i.d. 0.13 mm. The outlet of the PEEK tubing was encapsulated in 0.5 cm long Teflon tubing for tight connection with the nebulizer inlet. Operating conditions: plasma power 1.25 kW, argon flows 0.7, 1.0, and 15 L/min for nebulizer, auxiliary, and coolant, respectively. The atomic phosphorus line at 178.2 nm was measured, and the transient signal was collected with an integration time of 3 s. Data were collected with the Smart Analyzer 2.1 (Spectro Analytical Instruments, Kleve, Germany), and the obtained chromatographic peaks were integrated using the peak-fitting module Origin 6.0 (Microcal Software Inc., Northampthon, MA). The gas flows were optimized on a daily basis to give optimal signal for P by introducing a 200 ppb P solution in 30% (v/v) acetonitrile. Total P concentration in the sediment extract before separation was measured by direct infusion of 5 mL of the sediment extract into the ICP-AES instrument. A calibration curve was obtained from P standards containing 3, 6, and 10 ppm P. For quantification of the P containing peaks from a 0-1 cm lake sediment extract with a retention time (tR) larger than the dead time (t0), a calibration curve was made with the peak areas of 4, 8, 16, and 32 µM 5′AMP. To quantify the P containing peak eluting at t0, a calibration curve was made with the peak areas of 20, 40, 80, and 160 ppm Na2HPO4 (corresponding to 0.14, 0.28, 0.56, 1.13 mM, respec(40) Murphy, J.; Riley, J. P. Anal. Chim. Acta 1962, 27, 31–36.


6691

Figure 1. Chromatogram of phosphorus compounds extracted from lake sediment (0-1 cm) with EDTA f NaOH detected with LC-ICP-AES.

tively), since most P in this peak is presumed to be free phosphate. For a semiquantitative comparison of the 0-1 cm and the 29-30 cm lake sediment extract, 10 µL of 1.38 mM adenosine 3′,5′-cyclic monophosphate (cAMP, Aldrich) was added as an internal standard to 450 µL of sediment extract. In a separate experiment, the peak eluting at t0 was collected for further analysis with 31P NMR. ESI-MS/MS. The flow from the LC column (200 µL/min) was directed to a grounded connection which led to the pneumatically assisted ESI interface (TurboV ion source) coupled to the 3200 Q-TRAP (LC/MS/MS system, Applied Biosystems MDS SCIEX, Concord, ON, Canada) operating in negative ion mode. The ionspray voltage and declustering potential were typically set to -4.5 kV and -30 V, respectively. The nebulizer and curtain gas (boil off from liquid nitrogen) were set to 60 and 15 psi, respectively. The gas flow through the turbo heaters, heated to 700 °C, was set to 65 psi. An information dependent acquisition (IDA) experiment was performed, which involved during the same chromatographic run: (i) precursor ion scan (Prec) monitoring precursors of either m/z 79 (PO3-) or 97 (H2PO4-), covering a mass range of 90-1200 u in 4 s with a step size of 0.25 u; (ii) enhanced product ion scan (EPI) on the two most intense precursor ions in the mass range above with a collision energy (CE) and collision energy spread (CES) set to -40 and -10 eV, respectively. The collision gas pressure (CAD) was set to high (4 ×10-5 Torr). The linear ion trap fill time was set to 150 ms, and a scan rate of 1000 u/s with a step size of 0.06 u was used. 31 P NMR. 31P nuclear magnetic resonance was used to analyze the nature of the phosphate compounds present in the chromatographic peak eluting at t0. The procedure for sample preparation prior to 31P NMR was according to Ahlgren et al. 2007.38 RESULTS AND DISCUSSION LC-ICP-AES. The developed separation method resulted in eight chromatographic peaks containing phosphorus compounds (Figure 1). Since ICP-AES is element selective and not P compound 6692


Figure 2. A column chart on the areas of P containing peaks 1-8 on the extract of 0-1 cm and 29-30 cm deep lake sediment from LC-ICP-AES analysis. The sediment samples were spiked with standard cAMP to compensate for drifts in the P signal.

selective it is not possible to resolve chromatographically overlapping peaks. Therefore, the chromatographic peaks observed in Figure 1 most likely do not represent individual P compounds but rather several. To preconcentrate the compounds of interest on the beginning of the column, a sample loop with large volume (250 µL) was combined with a linear gradient of the mobile phase containing 1% acetonitrile during 3 min (Table 1) and a flow rate of 200 µL/ min. The on-column concentration step in combination with the concentration step (10 times) prior to injection was advantageous for the P measurements. Although ICP-AES is capable of detecting low P concentrations, the sensitivity for P could vary with the acetonitrile content of the mobile phase due to increased background emission at similar wavelengths as P. The surface (0-1 cm) sediment total P content was 1.7 mg P/g dry weight according to the Murphy and Riley method.40 Peak

Figure 3. LC-ESI-MS/MS data from the separation of an EDTA f NaOH surface (0-1 cm) sediment extract. Panel A shows the total ion chromatogram on the precursors of m/z 97 (H2PO4-). Panel B is a spectrum taken at 26.3 min showing the precursors of m/z 97 detected at that very time. Panel C is an enhanced product ion scan on m/z 346 (panel B).

0 (Figure 1) represents 23% and peaks 1-8 together represent 1.4% of the surface sediment total P content, respectively. Since peak 0 contains compounds that are not retained by the PGC material, these are most likely small compounds and/or highly charged compounds and therefore have little interaction to the hexagonal carbon sheets.34-36 The NaOH extracted 37% of the surface sediment total P content, and according to Ahlgren et al. 38 35% of this is present as orthophosphate. It is thus likely that a large amount of this orthophosphate is present in peak 0. This is confirmed by a 31P NMR experiment on a fraction containing peak 0. The NMR analysis indicates that free phosphate and phosphate monoesters are present (results not shown). It is therefore of great interest to continue the development of a separation step for these P compounds, but this is not addressed in this study. Despite the small phosphorus amount found in peaks 1-8 compared to the sediment total P content, they might still be of great ecological importance. The developed separation method was used for semiquantitative comparison of a 0-1 cm sediment extract and a 29-30 cm sediment extract by adding an internal P standard (cAMP) in both samples before injection on the LC

column. cAMP had a retention time of about 52 min. The internal standard was necessary to compensate for drifts in the ICP-AES instrumentation from run to run. A pronounced decline in the detected P compounds with sediment depth can be seen in Figure 2. The 29-30 cm sediment was in a previous article10 dated to be about 60 years. Only about 15-20% (Figure 2) remains present in the sediment after 60 years. Assuming a constant load of these compounds to the sediment surface, the decline in concentration with sediment depth apparently reflects degradation of these kinds of compounds, potentially recycled back to the water column. LC-ESI-MS/MS. To gain more information regarding the identity of the phosphorus compounds detected by ICP-AES, the same LC separation system was connected to an ESI-MS/MS instrument. The mobile phase used was compatible with ESI. An information dependent acquisition (IDA) experiment was set up with a survey scan consisting of a precursor ion scan monitoring precursors of either m/z 79 (PO3-) or 97 (H2PO4-). In Figure 3A, a representative chromatogram is shown for precursors of m/z 97 (similar data were obtained for precursors of m/z 79). Note the increased number of chromatographic peaks as compared to the LC-ICP-AES chromatogram probably due to both Analytical Chemistry, Vol. 80, No. 17, September 1, 2008

6693

Figure 4. Daughter ion scans on the fragments with m/z 211 (A) and 134 (B) from a LC fraction of an EDTA f NaOH surface (0-1 cm) sediment extract.

higher selectivity and sensitivity. The survey scan triggers the selection of precursor ions and automatically switches to MS/ MS mode during elution of the plausible P compounds (Figure 3B). In Figure 3C, the fragmentation spectra of m/z 346 is shown revealing the characteristic P ion signals at m/z 79 and 97. A number of other signals are also present in the spectrum, i.e.,

m/z 211 and 134. This indicates that the precursor 346 could fragment into two parts; either losing a neutral fragment of 135 u generating the corresponding ion with m/z 211or losing a neutral fragment 212 u generating the corresponding ion with m/z 134. Nucleotides were considered to be present in the sediment extract and for example deprotonated adenosine monophosphate (AMP, see Figure 3C) should give a signal at m/z 346. To gain more knowledge about the identity of m/z 346, fractions in a separate LC run were collected for more extensive analysis. One fraction containing the above 346 ion was reanalyzed by continuous infusion ESI-MS/MS. By increasing the declustering potential, ions generated in the ion source can gain more kinetic energy during the sampling process. As a result of the increased kinetic energy, the internal energy increases when the ions collide with the surrounding gas and thus more fragmentation is obtained. The fragment ions at m/z 211 and 134 were selected as precursor ions, and the results are shown in Figure 4. Fragment ion 211 is the only ion of these two that generates m/z 79 and 97 indicating that this part of the molecule contains the P group. The loss of the neutral group with a mass of 90 u in the mass spectrum of the product ions of m/z 211 could correspond to the pentose group breaking up and resulting in a fragment with m/z 121 as shown in Figure 4A. Even in the spectrum showing the products of m/z 134, the m/z 107 will correspond to a loss of a neutral compound with a mass of 27 u and will likely be HCN (Figure 4B). The loss of the neutral compound with a mass of 42 u corresponds to H2NCN. Upon extraction of the signal corresponding to m/z 346, three chromatographic peaks are apparent in the extracted ion chromatogram (XIC) shown in Figure 5. At least two positional isomers of AMP exist, with the phosphate group attached to the pentose at the 3′ or 5′ position. It was confirmed that the chromatographic peaks at 26.3 min. and 29.9 min. corresponded to 5′-AMP and 3′AMP, respectively, by spiking the sample with 3′-AMP and 5′AMP standards. The compound eluting at 26.3 min shows less fragmentation compared to the compound eluting at 29.9 min, which agrees well with previously reported observations for

Figure 5. XIC (extracted ion chromatogram) on m/z 346 as monitored with LC-ESI-MS/MS on an EDTA f NaOH surface (0-1 cm) sediment extract. 6694



6695

22.0

23.3

24.2

25.1 4.9 × 105 5′-UMP 323 211 121 111 97 79

25.5 2.0 × 105 CMP 322 211 121 110 97 79

26.3

2.8 × 105 5′-GMP 5′-AMP 362 346 319 211 211 151 150 134 133 107 121 97 107 79 97 79

25.5

Verified compound identity in bold, possible compound identity in italic.

2.3 × 104 3.2×104 2.1 × 105 CMP 484 425 322 281 408 211 211 381 121 128 328 110 97 291 97 79 273 79 224 211 159 134 97 79

28.1 6.8 × 105 GMP 362 211 150 133 121 107 97 79

28.7

29.5

2.0 × 104 4.3 × 104 GMP 627 362 609 211 583 150 529 133 436 121 384 97 322 79 304 211 97 79

29.9 4.6 × 105 3′-AMP 346 211 134 107 97 79

31.4 1.7 × 105 AMP 346 211 134 107 97 79

31.9

time (min)

22.2

a

25.5

22.8

23.1

23.6

27.8

34.4

36.1

36.3

37.6

37.8

38.5

39.1

39.5

33.1

39.6

33.7

40.3

1.2 × 104 AMP-CMP 651 633 571 540 436 346 328 322 211 134 97 79

39.8

1.0 × 104 AMP-UMP 652 634 572 554 385 323 305 211 134 97 79

35.1

40.7

9.0 × 103 AMP-GMP 691 673 647 611 593 540 476 458 424 362 328 273 211 150 134 97 79

35.5

41.0

1.9 × 104 AMP-AMP 675 675 595 540 460 442 408 346 328 273 211 134 97 79

527 396 367 196 128 97 79

512 311 181 97 79

510 466 380 135 97 79

482 465 403 281 154 97 79

395 377 351 241 153 139 109 97 393 689 644 530 449 431 393 353 313 287 180 97

361 317 201 182 123 97

465 421 285 259 186 97 79

383 627 513 433 415 384 344 277 271 256 97

353 273 97

397 171 97 79

409 365 329 285 153 113 97

427 409 392 383 365 347 339 259 97

411 368 305 171 153 97 79

441 423 397 379 97

351 207 167 97

397 379 353 329 317 309 250 233 193 164 147 97

411 351 349 271 171 97

2.9 × 104 6.1 × 104 2.6 × 104 3.5 × 104 3.9 × 104 5.2 × 104 4.7 × 104 6.5 × 103 2.4 × 104 2.2 × 104 2.3 × 104 1.0 × 105 7.0 × 104 2.0 × 104 2.8 × 104 1.4 × 104 3.2 × 104 3.6 × 104

m/z of doubly charged compounds in bold.

intensity ([cps]) 2.5 × 104 identity precursor ion, m/z 427 product ions, m/z 409 366 329 227 159 135 134 97 79

32.6

1.5 × 104 3.0 × 104 AMP-CMP 429 651 399 633 385 571 367 553 355 408 228 322 211 304 185 211 97 134 97 79

Table 3. LC-ESI-MS/MS Detected Phosphorus Compounds in an EDTA f NaOH Surface (0-1 cm) Sediment Extracta

a

21.2

7.3 ×104 5′-CMP 322 211 121 110 97 79

time (min)

intensity ([cps]) identity precursor ion, m/z product ions, m/z

Table 2. LC-ESI-MS/MS Detected Nucleotides in an EDTA f NaOH Surface (0-1 cm) Sediment Extracta

Figure 6. XIC (extracted ion chromatogram) from all EPI (enhanced product ion) scans on m/z 211 (main fragment in several nucleotides) as monitored with LC-ESI-MS/MS on an EDTA f NaOH surface (0-1 cm) sediment extract. The verified compound identity is in bold, and the possible compound identity is in italic.

Figure 7. Panels A and B are the total ion chromatograms from the precursors of m/z 97 as monitored with LC-ESI-MS/MS on a EDTA f NaOH sediment extract at depth 0-1 and 29-30 cm, respectively. Panels C and D are the XIC (extracted ion chromatogram) on m/z 346 from panels A and B, respectively. Note the difference in intensity between panels C and D.

deoxynucleotides.41 The exact identity of the compound eluting at 31.4 min remains unknown. A single chromatographic run generates extensive data, and we have explored the possibility to quickly scan through the data for possible nucleotides by selecting a characteristic fragment and extracting that signal from all EPI spectra. As an example XIC of m/z 211, which corresponds to a common fragment for nucleotides, is shown in Figure 6 indicating around 19 nucleotides. For evaluation of the presence of more specific nucleotides other m/z (41) Herna´ndez, H. Rapid Commun. Mass Spectrom. 1996, 10, 1543–1550.

6696


were selected, for example m/z 134 corresponding to the deprotonated adenine fragment. The possible identity of some of those compounds suggests that dinucleotides such as AMP-AMP and AMP-CMP are present. In an effort to increase the sensitivity of nucleotides, the use of 50 mM ammonium acetate instead of 10 mM ammonium formate in the mobile phase was evaluated. In contradiction to previously reported results,42 no increase in sensitivity was (42) Xing, J.; Apedo, A.; Tymiak, A.; Zhao, N. Rapid Commun. Mass Spectrom. 2004, 18, 1599–1606.

obtained. Anyhow a number of nucleotides were detected, and a summary of detected signals and in some cases possible identity is presented in Table 2. As can be seen, some precursor ions, besides m/z 346, with common fragment ions are detected at different retention times indicating the occurrence of different isomers. In total, 19 nucleotide containing compounds were detected, and the identities of 5 were confirmed. A number of other P containing compounds were also detected (see Table 3). Two chromatographic peaks, at 34.4 and 37.6 min, represent doubly charged ions probably due to multiple phosphate groups. Furthermore, some of the signals have m/z 171 as a common fragment, probably phosphoglycerol, possibly originating from hydrolysis of phospholipids. A more detailed study with the LC-ESI-MS/MS method, running sediment extracts from two depths, was performed to gain insight about P compound turnover. The total ion chromatogram on the precursors of 97 (H2PO4-) of a 0-1 cm sediment extract and a 29-30 cm sediment extract, respectively, are shown in Figure 7. From the total ion chromatogram it is not easy to see any clear difference between the samples. However, upon extracting the signal for AMP (m/z 346, Figure 7C,D) a significant difference, about a 10 times decrease, in signal strength was observed for the deep sediment indicating substantial degradation over time in the sediment.

several well separated chromatographic P peaks. Since the separation is compatible with both ICP-AES and ESI-MS/MS, this is a great step forward. ICP-AES showed to be a useful tool for the development of a separation method of organic P compounds extracted from aquatic sediment. It was also shown that the technique in the future could be applied for a first screening and quantification of sediment organic P compounds. With the presented LC-ESI-MS/MS technique, it was possible to identify several sediment organic P compounds which mainly were a different kind of nucleotides. Future aspects include the detailed analysis of several organic P standards possibly found in aquatic sediments with both techniques. It could also include a detailed comparison of organic P compounds of sediments at different depths or surface sediments from different lakes and seas. Milder extraction methods should definitely be considered since NaOH can induce hydrolysis and in this way can produce artifacts. Finally, further development of separation strategies, compatible with both ICP and ESI, that can address the limitation in obtaining retention of the most polar early coeluting compounds seen in the LC-ICP run is desirable.

CONCLUSIONS AND FUTURE ASPECTS In the development of a separation step for P compounds extracted from aquatic sediments, the use of PGC resulted in

Received for review March 28, 2008. Accepted July 8, 2008.

ACKNOWLEDGMENT Thanks to FORMAS and the Malme´ns fund for making this project financially possible.

AC8006335


6697

Screening for Organic Phosphorus Compounds in Aquatic Sediments

Recommend Documents