Rapid Identification of Long-Chain Polyunsaturated Fatty Acids in a

Jun 11, 2012 - File failed to load: https://cdn.mathjax.org/mathjax/contrib/a11y/accessibility-menu.js ..... Fatty acid standards (Cayman Chemical, An...
1 downloads 9 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Rapid Identification of Long-Chain Polyunsaturated Fatty Acids in a Marine Extract by HPLC-MS Using Data-Dependent Acquisition Michael C. Thomas,†,‡ Simon R. Dunn,‡ Jens Altvater,† Sophie G. Dove,‡ and Geoffrey W. Nette*,† †

Independent Marine Biochemical Research, Moreton Bay Research Station, Dunwich Qld 4183, Australia ARC Centre for Excellence in Coral Reef Studies, School of Biological Sciences, University of Queensland, St. Lucia Qld 4072, Australia

‡

S Supporting Information *

ABSTRACT: The collision-induced dissociation (CID) of a range of deprotonated fatty acid standards was studied using linear ion trap mass spectrometry. Neutral losses of 78, 98, and 136 Da were consistently observed for fatty acids with five or more double bonds. Comparison of the MS/MS spectra of docosahexaenoic acid (DHA) and universally 13C-labeled DHA allowed the molecular formulas for these neutral losses to be determined as C6H6, C5H6O2, and C8H8O2. Knowledge of fatty acid fragmentation processes was then applied to identify fatty acids from a sea anemone, Aiptasia pulchella, and dinoflagellate symbiont, Symbiodinium sp. extract. Using HPLC-MS, fatty acids were separated and analyzed by tandem mass spectrometry in data-dependent acquisition mode. Neutral loss chromatograms for 78, 98, and 136 Da allowed the identification of long-chain fatty acids with five or more double bonds. On the basis of precursor ion m/z ratios, chain length and degree of unsaturation for these fatty acids were determined. The application of this technique to an Aiptasia sp.−Symbiodinium sp. lipid extract enabled the identification of the unusual, long-chain fatty acids 24:6, 26:6, 26:7, 28:7, and 28:8 during a single 40 min HPLC-MS analysis.

F

High-performance liquid chromatography (HPLC) is easily coupled to ESI-MS and, as such, HPLC-MS is considered to be an extremely powerful tool for the analysis of a range of compounds including fatty acids.20−30 Derivatization is commonly employed to enhance the ionization of fatty acids.20,22,23,27 Alternatively, fatty acids may be ionized as barium adduct ions by the post-column addition of barium acetate.25 The ESI of fatty acids in the negative ion mode results in the formation of abundant [M−H]− ions. Therefore, fatty acids may be analyzed directly by HPLC-ESI-MS without chemical derivatization or cationization with metal ions.21,26,28−30 This is advantageous because additional sample preparation is not required. Multiple reaction monitoring (MRM) is commonly employed to increase the specificity and sensitivity of fatty acid identification and quantification.21,26,30 The long-chain polyunsaturated fatty acids arachidonic acid (20:4 n-6), DHA (22:6 n-3), and EPA (20:5 n-3) have been analyzed using the MRM transitions resulting from the neutral loss of carbon dioxide.21,26 These MRM transitions are attractive because the loss of carbon dioxide is known to be a major fragmentation process for highly unsaturated fatty acids.31 The use of MRM and the identification of unknown fatty acids requires

atty acids are arguably the most important lipid class and are the building blocks for the majority of lipid classes.1 Polyunsaturated fatty acids (PUFAs), including docosahexaenoic acid (DHA, 22:6 n-3) and eicosapentaenoic acid (EPA, 20:5 n-3), are of particular interest to human health due to their important roles in brain function2 and preventing cardiovascular disease.3 Very-long-chain (>22 C) PUFAs are also known to exist and have been identified in the retina4 and in the brains of patients with Zellweger’s disease.5 Marine algae are rich in the highly unsaturated fatty acids, DHA and EPA. In addition, fatty acids with both greater degrees of unsaturation and longer chain lengths have also been identified in marine algae.6−10 These very-long-chain PUFAs from marine algae can be traced through the food chain and have been identified in marine animals such as fish,6,11−13 invertebrates,14−16 and seals.17 The analysis of fatty acids is routinely performed by gas chromatography−mass spectrometry (GC-MS) after derivatization to form fatty acid methyl esters (FAMEs).18 By comparison of retention times (tR) to that of authentic standards it is possible to identify most common fatty acids. In the case of unusual fatty acids, GC-MS of picolinyl ester and 4,4-dimethyloxazoline derivatives has proven very effective in structural characterization.19 While GC-MS is still an excellent method for the analysis of fatty acids, mass spectrometers utilizing electrospray ionization (ESI) are now more common in laboratories than GC-MS instruments. Consequently, there is a need for methods for the analysis of fatty acids by ESI-MS. © 2012 American Chemical Society

Received: March 6, 2012 Accepted: June 11, 2012 Published: June 11, 2012 5976

dx.doi.org/10.1021/ac3006523 | Anal. Chem. 2012, 84, 5976−5983

Analytical Chemistry

Article

following components: micro degasser (G1379B), binary pump SL (G1312B), high-performance autosampler SL (G1367C), thermostatted column compartment SL (G1316B), and a diode array detector SL (G1315C). Chromatography was performed using a Luna 3 μm C18(2) 100 Å 75 × 4.60 mm column (Phenomenex, Torrance, CA). Water and acetonitrile were used as the A1 and B1 solvents, respectively. For sample analysis, 10 μL of A. pulchella sp.−Symbiodinium sp. lipid extract was injected using the autosampler. The components of A. pulchella sp.−Symbiodinium sp. lipid extracts were then analyzed in a 40 min run with a HPLC flow rate of 0.8 mL/min. A gradient elution was preformed as shown in Table S-2 of Supporting Information. The mass spectrometer was run in negative ion mode with the following instrument settings: spray voltage −3.2 kV, capillary voltage −28 V, tube voltage −75 V, and capillary temperature 320 °C. The nitrogen flow rates at the ESI source were set as follows: sheath gas 40 (arbitrary units) and auxiliary gas 20 (arbitrary units). No sweep gas was used. Nitrogen was sourced using a Domnick Hunter LCMS30-1-E nitrogen generator (Parker Hannifin Ltd., Industrial Division, England). Universally 13C-labeled [U-13C] DHA was produced by culturing A. pulchella sp.−Symbiodinium sp. within artificial seawater containing 13C-labeled sodium bicarbonate (Supporting Information). The MS/MS spectrum for [U-13C] DHA was acquired using a normalized collision energy of 22% with an activation time of 30 ms following chromatographic separation using the gradient elution provided in Table S-2 of Supporting Information. The retention time of [U-13C] DHA was found to be identical to unlabeled DHA. The data-dependent acquisition (DDA) experiment was constructed so that an MS scan was performed from m/z 200− 550 followed by three MS/MS scans. MS/MS spectra were obtained for the three most intense ions observed in the preceding MS scan. MS/MS spectra were acquired using a normalized collision energy of 25% and an isolation width of 1 Th. Multiply charged ions and contaminant ions of m/z 204, 218, and 240 were excluded from MS/MS analysis. Dynamic exclusion was not enabled to allow multiple MS/MS spectra to be acquired for the same fatty acid and possible isomers.

knowledge of the fragmentation processes of fatty acid [M− H]− ions. The low-energy collision-induced dissociation (CID) of fatty acid anions reported by Kerwin and co-workers using triple quadrupole tandem mass spectrometry showed positional isomers of monounsaturated fatty acids (MUFAs), and some PUFAs could be differentiated based on their MS/MS spectra.31 However, the fragmentation of unsaturated fatty acids was unpredictable, and the localization of doubles bonds was difficult. For these reasons, structural characterization by low-energy CID of fatty acid anions has not been widely utilized. Recently, however, Yang and co-workers reported that isomeric fatty acids could be differentiated via tandem mass spectrometry of deprotonated fatty acid anions.32 This suggests that low-energy CID of fatty acid [M−H]− ions is capable of providing higher levels of structural information than previously acknowledged. This study is part of ongoing research into the regulation of the symbiotic relationship between the cnidarian host, Aiptasia pulchella, and its dinoflagellate symbiont, Symbiodinium sp. The LC-MS/MS protocol described here utilizes data-dependent acquisition (DDA) on a linear ion trap mass spectrometer. A series of very long-chain, highly unsaturated fatty acids were identified within an A. pulchella sp.−Symbiodinium sp. lipid extract on the basis of neutral losses characteristic for this fatty acid class.

■

MATERIALS AND METHODS Materials. Fatty acid standards (Cayman Chemical, Ann Arbor, MI, via the Australian distributor Sapphire Bioscience, Waterloo, NSW, Australia) were used without further purification and are listed in Supporting Information (Table S-1). NaH13CO3 (isotopic purity 99%) was purchased from Sigma Aldrich (Sydney, NSW, Australia). Experimental details regarding the isotopic incubation and preparation of A. pulchella sp.−Symbiodinium sp. lipid extracts are provided in Supporting Information. Deionized water was obtained from a Milli-Q Plus filtration system (Millipore, Billerica, MA), and HPLC grade acetonitrile was obtained from Lomb Scientific (Taren Point, NSW, Australia). Mass Spectrometry of Fatty Acid Standards. Fatty acid standards were made to a concentration of 1 ng/μL in acetonitrile/water (60:40, v/v). An LTQ XL linear ion trap mass spectrometer (Thermo Fisher Scientific, Waltham, MA) fitted with an IonMax electrospray ionization source was used. Fatty acid standard solutions were directly infused at a flow rate of 10 μL/min. The mass spectrometer was operated in the negative ion mode with the following settings: source voltage −3.2 kV, tube lens −75 V, capillary voltage −28 V, capillary temperature 320 °C, sheath gas 2 (arbitrary units) and auxiliary gas 8 (arbitrary units). For tandem mass spectrometry, an isolation width of 1 Th was used for ion selection, and the Q value was kept at the default value of 0.25. Collision energy was adjusted to promote extensive fragmentation of the selected precursor ion (precursor ion relative intensities were typically between 2% and 40%). For isomeric fatty acids, the same collision energies were used so that the observed fragment ion relative intensities were directly comparable. High-Performance Liquid Chromatography−Mass Spectrometry (HPLC-MS). A. pulchella sp.−Symbiodinium sp. lipid extracts were analyzed using an Agilent 1200 Series HPLC instrument (Agilent Technologies, Santa Clara, CA) coupled to an LTQ XL linear ion trap mass spectrometer (Thermo Fisher Scientific). The 1200 Series HPLC had the

■

RESULTS AND DISCUSSION MS/MS Spectrum of Docosahexaenoic acid (DHA, 22:6 n-3). The MS/MS spectrum of the [M−H]− anion of 22:6 n-3 acquired on an ion trap mass spectrometer is shown in Figure 1a. In this spectrum, an abundant fragment ion is observed at m/z 283 with a normalized abundance of 85% representing a neutral loss of 44 Da (Table 1). The neutral loss of 44 Da corresponds to the loss of carbon dioxide and has previously been observed for polyunsaturated fatty acid anions.31 An ion of m/z 309 is also observed corresponding to the [M−H−H2O]− fragment ion with a normalized abundance of 2% (Table 1). Lower mass fragment ions are also observed at m/z 191, 229, and 249 representing the neutral losses of 136, 98, and 78 Da, respectively. In the ion trap MS/ MS spectrum of 22:6 n-3, the [M−H−136]−, [M−H−98]−, and [M−H−78]− fragment ions are the only fragment ions with a normalized abundance greater than 1% resulting from fragmentation of the acyl chain. In this respect, the ion trap MS/MS spectrum is significantly different from the MS/MS spectrum presented by Kerwin and co-workers acquired using a triple quadrupole mass spectrometer where extensive fragmentation of the acyl chain was observed.31 5977

dx.doi.org/10.1021/ac3006523 | Anal. Chem. 2012, 84, 5976−5983

Analytical Chemistry

Article

dominant fragmentation pathway for the saturated fatty acids 14:0, 16:0, and 18:0 (normalized abundance of 97−99%). The presence of the [M−H−H2O]− and [M−H−CO2]− fragment ions in a wide range of the fatty acid standards analyzed suggests that these fragment ions are not specific enough for identification of long-chain highly unsaturated fatty acids. Curiously, the neutral loss of 2 Da corresponding to molecular hydrogen is also observed for fatty acids with 3−5 double bonds (Table 1). The molecular hydrogen neutral loss is particularly prominent from triunsaturated fatty acids. The normalized abundance of the [M−H−H2]− fragment ion is as high as 49% for the 22:3 n-3 fatty acid. Interestingly, the normalized abundance of the [M−H−H2]− fragment ion is greater with increasing distance between the three methyleneinterrupted double bonds and the carboxylate group. There is no observed trend in the normalized abundance of the [M−H−98]− fragment ion with varying degrees of unsaturation. The [M−H−98]− fragment ion is observed in the MS/MS spectra of four of the five monounsaturated fatty acids analyzed in addition to the fatty acid standards with 3−6 double bonds (Table 1). Conversely, the [M−H−78]− and [M−H−136]− fragment ions are only observed for the highly unsaturated fatty acids standards (Table 1). The [M−H−78]− fragment ion is observed at very low abundance for the tetraunsaturated fatty acids 18:4 n-3, 20:4 n-3, 20:4 n-6, and 22:4 n-6 (normalized abundances of 0.02−0.1%). For the 20:5 n-3, 22:5 n-3, and 22:6 n-3 fatty acid standards, however, the normalized abundance of the [M−H−78]− fragment ion is significantly higher (normalized abundances of 1−2%). Similarly, the [M−H−136]− fragment ion is only observed for the 20:5 n-3, 22:5 n-3, and 22:6 n-3 fatty acid standards. The normalized abundance of the [M−H−136]− fragment ion is greater for 22:6 n-3 (normalized abundance of 8%) than for 20:5 n-3 and 22:5 n-3 (normalized abundances of 0.4%). The fatty acids 20:5 n-3, 22:5 n-3 and 22:6 n-3 are the only standards studied which formed all three of the [M−H−78]−, [M−H−98]−, and [M−H−136]− fragment ions and therefore supports the initial hypothesis that these ions are characteristic of long-chain, highly unsaturated fatty acids. Characterization of the [M−H−78]−, [M−H−98]−, and [M−H−136]− Fragment Ions. Experiments were conducted to gain further insight into the distinct [M−H−78]−, [M−H− 98]−, and [M−H−136]− fragment ions. The ion trap mass spectrometer used in these experiments does not have the high resolution and mass accuracy to determine the acurate mass of the m/z 191, 229, and 249 fragment ions observed in the MS/ MS spectrum of 22:6 n-3 (Figure 1a). Using an alternative approach, the molecular formulas of the neutral fragments were determined by comparison of the MS/MS spectrum of 22:6 n-3 (Figure 1a) to the MS/MS spectrum of the universally 13Clabeled [U-13C] 22:6 n-3 isotopologue (Figure 1b). In the MS/ MS spectrum of [U-13C] 22:6 n-3, the precursor [M−H]− ion is observed at m/z 349. This is 22 Da greater than the [M−H]− precursor ion of non-13C labeled, monoisotopic 22:6 n-3 indicative of the 22 13C atoms. High-mass fragment ions are observed at m/z 331 and 304 corresponding to the neutral loss of 18 and 45 Da. The m/z 331 is 22 Da greater than the analogous m/z 309 ion observed for unlabeled 22:6 n-3 (Figure 1a) and indicates that no 13C atoms are lost in the 18 Da neutral loss. This is expected because the neutral loss of 18 Da is attributed to H2O. Conversely, the base peak at m/z 304 is only 21 Da greater than the m/z 283 ion observed in the MS/ MS spectrum of unlabeled 22:6 n-3, indicating that a single 13C

Figure 1. The MS/MS spectrum of the [M−H]− ions of (a) docosahexaenoic acid (22:6 n-3) and (b) universally 13C-labeled docosahexaenoic acid ([U-13C] 22:6 n-3).

Comparison of the MS/MS Spectrum of Docosahexaenoic Acid (DHA, 22:6 n-3) to a Range of Fatty Acid Standards. It was hypothesized that the [M−H−136]−, [M− H−98]−, and [M−H−78]− fragment ions may be characteristic of long-chain, highly unsaturated fatty acids. To test this hypothesis, MS/MS spectra were acquired for a range of fatty acid standards with varying degrees of unsaturation and double bond positions. Representative MS/MS spectra for saturated, monounsaturated, and polyunsaturated standards using ion mass spectrometry are given in Figure S-1 in Supporting Information. The normalized abundances for the [M−H−H2]−, [M−H−H2O]−, [M−H−CO2]−, [M−H−78]−, [M−H−98]−, and [M−H−136]− fragment ions for the fatty acid standards studied are shown in Table 1. The intensity of the precursor ion was excluded from the calculation of normalized abundances to reduce the impact of precursor ion intensity variation on the calculated fragment ion normalized abundance. The abundant [M−H−CO2]− fragment ion observed in the MS/MS spectrum of 22:6 n-3 is also observed for the triunsaturated, tetraunsaturated, and pentaunsaturated fatty acid standards in high abundance (Table 1). For fatty acids with 0−2 double bonds, however, the fragment ion resulting from the neutral loss of 44 Da is either not observed or observed at very low abundance (less than 0.5% normalized abundance). Conversely, the [M−H−H2O]− fragment ion is observed for all fatty acid standards regardless of the number of double bonds (Table 1). The normalized abundance, however, increases with decreasing degrees of unsaturation and is the 5978

dx.doi.org/10.1021/ac3006523 | Anal. Chem. 2012, 84, 5976−5983

Analytical Chemistry

Article

Table 1. Normalized Abundances of Selected Fragment Ions from the Negative Ion MS/MS Spectra of a Series of Fatty Acid Standards fragment ion neutral fragment 22:6 n-3 22:5 n-3 20:5 n-3 22:4 n-6 20:4 n-3 20:4 n-6 18:4 n-3 22:3 n-3 20:3 n-3 20:3 n-6 20:3 n-9 18:3 n-3 18:3 n-6 20:2 n-6 18:2 n-6 9Z,11E-18:2 9E,11E-18:2 10E,12Z-18:2 22:1 n-9 20:1 n-6 18:1 n-9 cis 18:1 n-9 trans 16:1 n-7 22:0 20:0 18:0 16:0 14:0 a

a

[M−H−2]−

[M−H−18]−

[M−H−44]−

[M−H−78]−

[M−H−98]−

[M−H−136]−

H2 0.0 0.1 0.0 1.9 1.3 0.5 0.2 48.5 31.7 18.3 10.3 17.5 10.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

H2O 1.5 3.4 3.0 7.8 6.5 6.3 4.6 17.9 19.3 16.9 11.1 17.3 12.7 88.0 88.0 88.6 90.0 87.3 92.3 92.1 90.0 96.6 91.6 83.1 96.6 97.2 98.2 98.6

CO2 85.4 84.3 88.2 74.8 76.6 80.0 81.8 23.6 37.3 54.6 69.0 55.5 67.7 0.0 0.0 0.0 0.1 0.1 0.3 0.3 0.0 0.3 0.1 0.1 0.2 0.2 0.6 0.2

C6H6 1.3 2.3 0.9 0.11 0.11 0.03 0.02 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

C5H6O2 2.7 6.9 5.2 11.9 12.5 11.6 11.3 0.8 1.9 3.2 4.9 3.2 4.1 0.0 0.0 0.0 0.0 0.0 0.9 0.3 0.4 0.8 0.0 0.0 0.0 0.0 0.0 0.0

C8H8O2 7.5 0.4 0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Molecular formulas of the neutral fragments, with the exception of H2, were determined for DHA.

involves the loss of 6 C atoms. The mass of 6 C atoms is 72 Da which implies that 6 H atoms are lost in the 78 Da neutral loss. Consequently, the 78 Da neutral loss is determined to have a molecular formula of C6H6. The neutral loss of C6H6 may suggest that the neutral fragment is benzene. The aromatic stability of benzene may mean that this gas-phase rearrangement is thermodynamically favorable. The MS3 spectrum of the [M−H−78]− fragment ion at m/z 249 is shown in Figure 2a. Within this spectrum, abundant fragment ions are observed at m/z 247, 231, 205, and 151, corresponding to the neutral losses of 2, 18, 44, and 98 Da, repectively. These neutral losses are also observed in the MS/ MS spectra for all of the triunsaturated acids studied (Table 1). In fact, the MS3 spectrum of the [M−H−78]− fragment ion is remarkably similar to the MS/MS spectra of 18:3 n-6 (Figure 2b) and 20:3 n-9 (which are Δ6 and Δ5 fatty acids, repectively). This suggests that the m/z 249 ion is a 16:3 carboxylate anion. Data-Dependent HPLC-MS/MS of a Marine Extract. It was proposed that the 78, 98, and 136 Da neutral losses could possibly be used to identify unknown, highly unsaturated fatty acids from complex biological extracts. The [M−H−78]−, [M− H−98]−, and [M−H−136]− fragment ions are of low abundance; however, the signal-to-noise ratio achieved by using ion trap mass spectrometry is excellent. In addition, the abundant [M−H−44]− fragment ion is also indicative of fatty acids with three or more double bonds (Table 1) and was

atom is lost in the 45 Da neutral loss. This is consistent with the assignment of the 44 Da neutral loss from the [M−H]− ion of non-13C labeled 22:6 n-3 as CO2. The same approach was applied to the [M−H−136]−, [M− H−98]−, and [M−H−78]− fragment ions. In the MS/MS spectrum of [U-13C] 22:6 n-3, the analogous ions are observed at m/z 205, 246, and 265, respectively. The m/z 205 ion is 14 Da greater than the m/z 191 fragment ion observed in the MS/ MS spectrum of unlabeled 22:6 n-3. Therefore, the [M−H− 136]− fragment ion has 14 C atoms, and the neutral loss consists of 8 C atoms. C8 gives a mass of 96 Da; therefore, the remaining 40 Da is the result of a combination of hydrogen and oxygen atoms. The 22:6 n-3 carboxylate anion has a molecular formula of C22H31O2. This gives rise to two possible combinations of O and H atoms to give 40 Da: 1 O atom and 24 H atoms, or 2 O atoms and 8 H atoms. The first of these combinations is not chemically possible because even an 8 C fully saturated alcohol or ether has only 18 H atoms. Therefore, the neutral loss of 136 Da is assigned the molecular formula of C8H8O2. When the same methodology to the [M− H−98]− fragment ion is applied, the 98 Da neutral loss is determined to have the molecular formula of C5H6O2. Both of these neutral losses involve the loss of the carboxylate group, resulting in the formation of a carbanion. The m/z 265 fragment ion in the MS/MS spectrum of [U-13C] 22:6 n-3 is 16 Da greater than the equivalent m/z 249 [M−H−78]− ion observed in the MS/MS spectrum of unlabeled 22:6 n-3. Therefore, the neutral loss of 78 Da 5979

dx.doi.org/10.1021/ac3006523 | Anal. Chem. 2012, 84, 5976−5983

Analytical Chemistry

Article

136 Da neutral loss chromatograms because under low-energy CID these fatty acids fragment via the neutral loss of water only. This reveals numerous chromatographic peaks that are obscured in the TIC (Figure 3a and Figure S-2a). From 0 to 5 min, several minor compounds are observed in the 44, 78, 98, and 136 Da neutral loss chromatograms (Figure S-3b−e). The structures for many of these compounds are currently unknown. With the short retention times on the reverse phase HPLC column, however, there is the possibility that these compounds are oxygenated fatty acid derivatives. At 5.3 min, a chromatographic peak is observed in the 44, 78, 98, and 136 Da neutral loss chromatograms. This corresponded to the elution of a compound with an m/z of 273 and an MS/MS spectrum consistent with that of a highly unsaturated fatty acid (Figure S-3, Supporting Information). The compound eluting at tR of 5.3 min is therefore assigned as the fatty acid 18:5. In the 44 Da neutral loss chromatogram, the abundant polyunsaturated fatty acids 18:4, 20:5, 18:3, 22:6, 20:4, and 22:5 are observed (in order of elution, see Figure S-2b). The increase in specificity observed with the 44 Da neutral loss chromatogram also reveals the presence of the 20:3 (tR 18.3 min) and 22:4 (tR 20.1 min) fatty acids within the A. pulchella sp.−Symbiodinium sp. lipid extract (Figure 3b). These fatty acids could be identified by comparison of the retention times and MS/MS spectra to the available fatty acids standards. Interestingly, five unidentified ions are observed in the 18− 24 min range in the 44, 78, 98, and 136 Da neutral loss chromatograms (Figure 3b−e). This would suggest that these compounds are highly unsaturated fatty acids. The longer retention times would also suggest that the chain-length for these fatty acids is greater than 22 carbon atoms. The MS/MS spectra from these putative fatty acids are displayed in Figure 4a−e. These MS/MS spectra of A. pulchella sp.−Symbiodinium sp. fatty acids are strikingly similar to the MS/MS spectrum of the 22:6 n-3 fatty acid standard shown in Figure 1a. From the mass-to-charge ratio of the precursor ions, these fatty acids are assigned as follows: 24:6 (tR 18.9 min), 26:7 (tR 20.4 min), 28:8 (tR 21.3 min), 26:6 (tR 22.1 min), and 28:7 (tR 23.1 min). Assignment of double bond position within these fatty acids was not possible from the observed fragmentation of the deprotonated fatty acid anions. However, the results presented by Yang and co-workers32 together with the results presented in Table 1 for the isomeric polyunsaturated fatty acid standards, suggest that discrimination of fatty acid isomers may be possible if authentic standards were available. The fatty acid 28:8 is presumably a methylene-interrupted n3 fatty acid as previously characterized in marine dinoflagellates.6,10 Van Pelt and co-workers proposed that 28:8 n3 is synthesized by chain elongation and desaturation of 24:6 n3.6 The biosynthetic pathway proposed by Van Pelt et al. also includes the intermediate fatty acids 26:6, 26:7, and 28:7.6 Curiously, within the data-dependent acquisition experiment of the A. pulchella sp.−Symbiodinium sp. lipid extract, all four of these very-long-chain, highly unsaturated fatty acids were identified. This may suggest that 24:6, 26:6, 26:7, and 28:7 identified within the A. pulchella sp.−Symbiodinium sp. lipid extract are n-3 fatty acids derived from the abundant 22:6 n-3. The n-6 fatty acid, 28:7 n-6, has, however, been identified in marine dinoflagellates.33 Therefore, it is possible that both n-3 and n-6 biosynthetic pathways may lead to the formation of very-long-chain PUFAs within the A. pulchella sp.−Symbiodinium sp. symbiosis.

Figure 2. (a) The MS3 spectrum of the [M−H−C6H6]− fragment ion from the collision-induced dissociation of docosahexaenoic acid (22:6 n-3). (b) The negative ion MS/MS spectrum of the triunsaturated fatty acid γ-linolenic acid (18:3 n-6).

therefore also investigated as a potentially more sensitive identifier of highly unsaturated fatty acids. Data-dependent acquisition (DDA) was applied to an A. pulchella sp.−Symbiodinium sp. lipid extract in conjunction with HPLC for the chromatographic separation of fatty acids. The total ion chromatogram (TIC) for the HPLC-DDA-MS experiment is shown in Figure 3a (see Figure S-2a of Supporting Information for the full TIC from 0 to 40 min). A series of chromatographic peaks are observed from 5 to 25 min corresponding to the abundant fatty acids present within the A. pulchella sp.−Symbiodinium sp. lipid extract. From the observed masses and comparison of the retention times to that of authentic fatty acid standards, these chromatographic peaks were identified as follows: 18:4 (tR 9.0 min), 20:5 (tR 11.5 min), 14:0 (tR 12.7 min), 18:3 (tR 12.9 min), 22:6 (tR 14.5 min) 20:4 (tR 15.6 min), 22:5 (tR 16.7 min), 16:0 (tR 19.7 min), and 18:0 (tR 23.4 min). An elevated baseline is observed in the TIC from 0 to 29 min (Figure S-2a) resulting from a high level of background chemical noise. This high background obscures minor fatty acids present within the sample. The ion chromatograms for the 44, 78, 98, and 136 Da neutral losses are shown in Figure 3b−e for the 18−24 min time range (see Figure S-2b−e of Supporting Information for the chromatograms from 0 to 40 min). For these neutral loss chromatograms, the elevated baseline observed in the TIC is not present. Moreover, the saturated fatty acids 14:0, 16:0, and 18:0 present in the TIC are not observed in the 44, 78, 98, and 5980

dx.doi.org/10.1021/ac3006523 | Anal. Chem. 2012, 84, 5976−5983

Analytical Chemistry

Article

Figure 3. (a) The total ion chromatogram from the data-dependent acquisition experiment on an A. pulchella sp.−Symbiodinium sp. lipid extract shown from 0 to 25 min and 18−24 min. From the same experiment, chromatograms for the (b) 44 Da, (c) 78 Da, (d) 98 Da, and (e) 136 Da neutral losses were extracted. The neutral loss chromatograms are shown only from 18 to 24 min.

■

SUMMARY

136 Da) may be used to identify unknown, very-long-chain, highly unsaturated fatty acids present within the sample. Within the A. pulchella sp.−Symbiodinium sp. lipid extract analyzed during this present study, five very-long-chain, highly unsaturated fatty acids were identified. This was achieved without prior chromatographic procedures to enrich this particular class of fatty acid. While not used in the present

The fragmentation of fatty acids with five or more double bonds and chain lengths of greater than 20 carbon atoms proceeds in a similar manner with respect to the observed neutral losses. When data-dependent acquisition is applied to a fatty acid mixture, characteristic neutral losses (44, 78, 98, and 5981

dx.doi.org/10.1021/ac3006523 | Anal. Chem. 2012, 84, 5976−5983

Analytical Chemistry

Article

Figure 4. The MS/MS spectra of (a) 24:6 (tR 18.9 min), (b) 26:7 (tR 20.4 min), (c) 28:8 (tR 21.3 min), (d) 26:6 (tR 22.1 min), and (e) 28:7 (tR 23.1 min) from an A. pulchella sp.−Symbiodinium sp. lipid extract. The sample was analyzed by HPLC-MS using data-dependent acquisition.

etry of methyl esters,6,35 or ozone-induced dissociation of fatty acid sodium adducts36 could be implemented.

study, it is possible that software could be developed to automate the identification and characterization of very-longchain, highly unsaturated fatty acids from data-dependent HPLC-MS/MS experiments. This would require MS/MS spectra matching based on the characteristic neutral losses and the determination of chain length and number of double bonds based on the mass-to-charge ratio of the precursor ion. We suggest that the HPLC data-dependent acquisition method presented here is suitable to survey lipid samples for very-long-chain, highly unsaturated fatty acids. If information regarding the locations of double bonds is required, more targeted analyses may be pursued. This could involve the use of traditional techniques such as the GC-MS of picolinyl ester and 4,4-dimethyloxazoline derivatives.19 Alternatively, more contemporary methods including CID of dilithiated cations,34 acetonitrile adduct chemical ionization tandem mass spectrom-

■

ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in the text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail [email protected]. Notes

The authors declare no competing financial interest. 5982

dx.doi.org/10.1021/ac3006523 | Anal. Chem. 2012, 84, 5976−5983

Analytical Chemistry

■

Article

(32) Yang, K.; Zhao, Z. D.; Gross, R. W.; Han, X. L. Anal. Chem. 2011, 83, 4243. (33) Mansour, M. P.; Volkman, J. K.; Jackson, A. E.; Blackburn, S. I. J. Phycol. 1999, 35, 710. (34) Hsu, F. F.; Turk, J. J. Am. Soc. Mass Spectrom. 2008, 19, 1673. (35) Van Pelt, C. K.; Brenna, J. T. Anal. Chem. 1999, 71, 1981. (36) Poad, B. L. J.; Pham, H. T.; Thomas, M. C.; Nealon, J. R.; Campbell, J. L.; Mitchell, T. W.; Blanksby, S. J. J. Am. Soc. Mass Spectrom. 2010, 21, 1989.

ACKNOWLEDGMENTS This study was supported by the Australian Research Council through CE0561435 and linkage LP110100566 grant funding. The authors acknowledge Thomas Gallagher (Independent Marine Biochemical Research) for laboratory assistance, Assoc. Prof. Stephen Blanksby (University of Wollongong) for helpful discussions when preparing this manuscript, and Dr. Mei Bai (CSIRO Livestock Industries) for careful proofreading prior to resubmission.

■

REFERENCES

(1) Fahy, E.; Subramaniam, S.; Brown, H. A.; Glass Christopher, K.; Merrill Alfred, H., Jr.; Murphy Robert, C.; Raetz Christian, R. H.; Russell David, W.; Seyama, Y.; Shaw, W.; Shimizu, T.; Spener, F.; van Meer, G.; Vannieuwenhze Michael, S.; White Stephen, H.; Witztum Joseph, L.; Dennis Edward, A. J. Lipid Res. 2005, 46, 839. (2) Horrocks, L. A.; Yeo, Y. K. Pharmacol. Res. 1999, 40, 211. (3) Lavie, C. J.; Milani, R. V.; Mehra, M. R.; Ventura, H. O. J. Am. Coll. Cardiol. 2009, 54, 585. (4) Berdeaux, O.; Juaneda, P.; Martine, L.; Cabaret, S.; Bretillon, L.; Acar, N. J. Chromatogr., A 2010, 1217, 7738. (5) Poulos, A.; Sharp, P.; Singh, H.; Johnson, D.; Fellenberg, A.; Pollard, A. Biochem. J. 1986, 235, 607. (6) Van Pelt, C. K.; Huang, M. C.; Tschanz, C. L.; Brenna, J. T. J. Lipid Res. 1999, 40, 1501. (7) Rezanka, T.; Nedbalova, L.; Sigler, K. Phytochemistry 2008, 69, 2391. (8) Mansour, M. P.; Frampton, D. M. F.; Nichols, P. D.; Volkman, J. K.; Blackburn, S. I. J. Appl. Phycol. 2005, 17, 287. (9) Rezanka, T.; Nedbalova, L.; Sigler, K. Phytochemistry 2008, 69, 2849. (10) Mansour, M. P.; Volkman, J. K.; Holdsworth, D. G.; Jackson, A. E.; Blackburn, S. I. Phytochemistry 1999, 50, 541. (11) Linko, R. R.; Karinkan., H. J. Am. Oil Chem. Soc. 1970, 47, 42. (12) Fukuda, Y.; Ando, Y. Fish. Sci. (Tokyo, Jpn.) 2011, 77, 875. (13) Ota, T.; Chihara, Y.; Itabashi, Y.; Takagi, T. Fish. Sci. (Tokyo, Jpn.) 1994, 60, 171. (14) Takagi, T.; Kaneniwa, M.; Itabashi, Y. Lipids 1986, 21, 430. (15) Mansour, M. P.; Holdsworth, D. G.; Forbes, S. E.; Macleod, C. K.; Volkman, J. K. Biochem. Syst. Ecol. 2005, 33, 659. (16) Nichols, P. D.; Danaher, K. T.; Koslow, J. A. Lipids 2003, 38, 1207. (17) Kakela, R.; Ackman, R. G.; Hyvarinen, H. Lipids 1995, 30, 725. (18) Eder, K. J. Chromatogr., Biomed. Appl. 1995, 671, 113. (19) Christie, W. W. Lipids 1998, 33, 343. (20) Johnson, D. W. Rapid Commun. Mass Spectrom. 2000, 14, 2019. (21) Bowen, C. L.; Kehler, J.; Evans, C. A. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2010, 878, 3125. (22) Yang, W.-C.; Adamec, J.; Regnier, F. E. Anal. Chem. 2007, 79, 5150. (23) Li, X.; Franke, A. A. Anal. Chem. 2011, 83, 3192. (24) Salm, P.; Taylor, P.; Kostner, K. Ther. Drug Monit. 2007, 29, 849. (25) Zehethofer, N.; Pinto, D. M.; Volmer, D. A. Rapid Commun. Mass Spectrom. 2008, 22, 2125. (26) Schiesel, S.; Lammerhofer, M.; Lindner, W. Anal. Bioanal. Chem. 2010, 397, 147. (27) Pettinella, C.; Lee, S. H.; Cipollone, F.; Blair, I. A. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2007, 850, 168. (28) Spitsmeister, M.; Adamberg, K.; Vilu, R. J. Microbiol. Methods 2010, 82, 288. (29) Lacaze, J.; Stobo, L. A.; Turrell, E. A.; Quilliam, M. A. J. Chromatogr., A 2007, 1145, 51. (30) Perret, D.; Gentili, A.; Marchese, S.; Sergi, M.; Caporossi, L. Rapid Commun. Mass Spectrom. 2004, 18, 1989. (31) Kerwin, J. L.; Wiens, A. M.; Ericsson, L. H. J. Mass Spectrom. 1996, 31, 184. 5983

dx.doi.org/10.1021/ac3006523 | Anal. Chem. 2012, 84, 5976−5983