Determination of Double Bond Positions in ... - ACS Publications

Subscriber access provided by University of Florida | Smathers Libraries

Article

Determination of Double Bond Positions in Polyunsaturated Fatty Acids using the Photochemical Paternò-Büchi Reaction with Acetone and Tandem Mass Spectrometry Robert C Murphy, Toshiaki Okuno, Christopher A. Johnson, and Robert M Barkley Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b02375 • Publication Date (Web): 18 Jul 2017 Downloaded from http://pubs.acs.org on July 19, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Determination of Double Bond Positions in Polyunsaturated Fatty Acids using the Photochemical Paternò-Büchi Reaction with Acetone and Tandem Mass Spectrometry

Robert C. Murphy*, Toshiaki Okuno#, Christopher A. Johnson*, and Robert M. Barkley*

*Department of Pharmacology, University of Colorado Denver, Anschutz Medical Campus, 12801 E. 17th Ave, Aurora, CO 80045

#Department of Biochemistry, Juntendo University School of Medicine, 2-1-1 Hongo, Bunkyo-ku, 113-8421, Japan

Corresponding Author: Robert C. Murphy, Department of Pharmacology, University of Colorado Denver, Anschutz Medical Campus, 12801 E. 17th Ave, Aurora, CO 80045, Tel: 303-724-3352, Fax: 303-7243357, Email: [email protected]

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 30

2 Abstract The position of double bonds along the carbon chain of methylene interrupted polyunsaturated fatty acids are unique identifiers of specific fatty acids derived from biochemical reactions that occur in cells. It is possible to obtain direct structural information as to these double bond positions using tandem mass spectrometry after collisional activation of the carboxylate anions of an acetone adduct at each of the double bond positions formed by the photochemical Paternò-Büchi reaction with acetone. This reaction can be carried out by exposing a small portion of an inline fused silica capillary to UV photons from a mercury vapor lamp as the sample is infused into the electrospray ion source of a mass spectrometer. Collisional activation of [M-H]– yields a series of reverse Paternò-Büchi reaction product ions that essentially are derived from cleavage of the original carbon-carbon double bonds that yield an isopropenyl carboxylate anion corresponding to each double bond location. Aldehydic reverse Paternò-Büchi product ions are much less abundant as the carbon chain length and number of double bonds increase. The use of a mixture of D0/D6-acetone facilitates identification of these double bond indicating product ions as shown for arachidonic acid. If oxygen is present in the solvent stream undergoing UV photoactivation, ozone cleavage ions are also observed without prior collisional activation. This reaction was used to determine the double bond positions in a 20:3 fatty acid that accumulated in phospholipids of RAW 264.7 cells cultured for 3 days.


Page 3 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3 Polyunsaturated fatty acids (PUFAs) play a critical role in the mammalian biochemistry. These fatty acids include dienoic, trienoic, tetraenoic, pentaenoic, as well as hexaenoic fatty acids. Not only are these fatty acids essential for proper functioning of cellular membranes, but also other biochemical events within cells. They are substrates for numerous reactions including synthesis of CoA esters as intermediates in the production of triacylglycerols, phospholipids, sphingolipids, and steroid esters. Many of the polyenoic fatty acids are substrates for enzymatic conversion into bioactive lipid mediators such as prostaglandins, leukotrienes, and other bioactive oxylipins which have been the focus of considerable experimental study.1,2 PUFAs are also rather easily oxidized by free radicals and reactive oxygen species such as ozone and nitric oxide to generate additional complex lipid substances, some of which are known to express considerable biological activity. PUFAs are also highly abundant in certain membranes within cells where they serve as antioxidant molecules that facilitate termination of free radical reactions by their close association with small molecular antioxidants such as vitamin E, glutathione, and ascorbic acid. PUFAs are made following the oxidation of saturated fatty acids by the action of oxygen dependant enzymes that catalyze double bond insertion along the fatty acyl chain. However, this is not random, but a highly directed process in mammalian cells.3 For example, the most common oxidase is stearoyl-CoA oxidase that converts stearic acid and palmitic acid into oleic (18:1, n-9) and palmitoleic acid (16:1, n-7).4 The introduction of the single cis double bond nine carbons from the carboxyl moiety is the characteristic activity of this enzyme, but in mammalian systems, this is as distant from the carboxyl group where double bond insertion can take place. The appearance of most PUFAs in cell membranes throughout the entire body is due to dietary intake of linoleic acid (18:2, n-6) and linolenic acid (18:3, n-3). Further biochemical processing



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 30

4 of these essential fatty acids (in terms of dietary intake) takes place by a ∆5 desaturase and ∆6 desaturase, which can lead to the eventual synthesis of arachidonic acid (AA, 20:4, n-6), eicosapentaenoic acid (EPA, 20:5, n-3), and docosahexaenoic acid (DHA, 22:6, n-3).3 In conditions of starvation of dietary essential fatty acids, these two mammalian enzymes can convert oleic acid into an unusual fatty acid called Mead acid which has three double bonds and 20 carbon atoms (20:3, n-9). This fatty acid has been observed to be made in humans5,6 as well as a fatty acid which is synthesized upon linoleate starvation of cells in culture. Thus, it is important to be able to determine double bond positions of PUFAs in order to understand precisely biochemical events taking place in cells. Routine mass spectrometric analysis of fatty acids or fatty acyl containing complex lipids can identify the presence of the total number of fatty acyl carbons and the total number of double bonds, but not the position of the double bonds. There have been several methods developed for gas chromatography/mass spectrometry as well as liquid chromatography/mass spectrometry to carry out this line of lipid analysis. Reactions of double bonds using osmium tetroxide or potassium permanganate has been used as more classic means to determine position of double bonds in fatty acyl chains by GC/MS 7 and more recently, derivatization strategies such as acetonitrile chemical ionization of methyl esters 8 and dimethyl disulfide derivatization and electron ionization 9 have found great utility. Direct mass spectrometric methods suited for LC-MS, have included collision induced dissociation (CID) after high energy collisional activation by the process termed remote site fragmentation,10 later extended to low energy CID of alkylated carboxylate cations,11 the addition of ozone during mass spectrometric analysis or even as a reagent gas for collision induced dissociation (Oz ID),12,13 and the use of ion mobility to isolate unique product ions that further fragment to generate remote-site product ions indicative of double bond positions in


Page 5 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


5 phospholipids.14 A major limitation has been generation of double bond-position indicating product ions in abundance in a field of product ions that do not carry this information. One strategy employed a NaF additive that lead to a unique product ion with a negative charge located within the structural region of the fatty acyl double bonds, a [M-2H+Na]- anion, which yielded MS3 product ions after collisional activation.15 Most recently has been the use of the Paternò-Büchi (P-B) reaction carried out in the electrospray region of ionization, with formation of an oxetane adduct from acetone with each double bond in the fatty acyl chain.16-18 This latter type of double bond location strategy can be readily implemented with very minor changes to instrumentation (see Supplemental Figure S1) and does not require high energy collisions to obtain direct information of the positions of double bonds. However, PUFAs have not been extensively studied and in fact, it has been reported that it is difficult to obtain double bond position information using the P-B reaction for PUFAs such as arachidonic acid as the carboxylate anion with the suggestion that Li+ cations need to be formed prior to CID to limit the number of non-P-B product ions that increase in abundance as the number of double bonds in the fatty acid increase.18 The report presented here addresses the CID of carboxylate anions from the oxetane adducts of acetone with several PUFAs, the formation of specific ions indicating double bond position, and strategies to improve means by which fragment ions can be identified in collision induced decomposition mass spectra, and use of this P-B reaction to identify a PUFA generated in cell culture.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 30

6 Experimental Section Materials. Arachidonic acid, linoleic acid, γ-linolenic acid, α-linolenic acid, dihomolinolenic acid (DHLA), Mead acid, 8,11,14-eicosatrienoic acid, eicosapentaenoic acid (EPA), and docosahexanoic acid (DHA) were purchased from Cayman Chemical (Ann Arbor, MI). Deuterium labeled Mead acid was also obtained from Cayman Chemical. D6-acetone (99.85 atom % excess 2H6) was purchased from Sigma-Aldrich (St. Louis, MO). All of the solvents were HPLC grade and were purchased from Thermo-Fisher Scientific (Waltham, MA). Photochemical activation. Implementation of the Paternò-Büchi reaction was carried out in a fused silica capillary which was constructed from 3 cm fused silica (100 µm ID, Polymicro Technologies, Phoenix, AZ) adapted at either end with a PEEK 1/16 inch capillary fitting and Teflon sleeve to seal the fused silica capillary to a PEEK union (supporting information Figure S-1). Each end of the PEEK union was attached with epoxy cement to an aluminum bracket to prevent any movement of the fused silica. A section of the fused silica (1.5 cm) was then stripped of the polyimide coating by use of a methane-fueled torch until all polyimide was removed. This section of fused silica capillary was then inserted into a line directly interfaced to the electrospray ion source of the mass spectrometer (supporting information Figure S1) rather than modifying the ion source region as previously described.16 Samples were introduced at 10 µL/min using flow injection from a syringe pump. Samples were dissolved in a solvent system consisting of acetone/water (v/v, 75/25). Samples of PUFAs were diluted to 5-10 µM in this solvent system and infused through the constructed photochemical reactor and into the ion source of the mass spectrometer. A low pressure mercury lamp (model 80-1057-01, BHK, Inc, Ontario, CA USA) was placed approximately 5 mm away from the fused silica capillary which had been stripped of the


Page 7 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


7 polyimide. The lamp was allowed to warm for approximately 2-3 min prior to collection of mass spectrometric data. Mass spectrometry. Most experiments were carried out with the 4000 QTRAP (Applied Biosystems, Thornhill, Ontario, Canada) using Q3 as an ion trap in the negative ion mode. Purified lipids were infused into the mass spectrometer electrospray interface using the acetone containing mobile phase with a declustering potential of -90 V, ion spray voltage -4000 V, entrance potential -10 V. Collision induced dissociation of the acetone adducts was carried out between -35 and -40 V for the PB adducts. High resolution analysis of the PB adducts of PUFAs was performed on a Synapt G2-S mass spectrometer (Waters, Manchester, UK) in negative ion mode. Samples were infused directly into the ion source by way of the fused silica capillary photochemical cell at a concentration of 5 µM by a syringe pump (10 µL/min). The settings for the Synapt G2-S mass spectrometer, including operating the instrument in high resolution mode, were ESI voltage of 4000 V, sampling cone of -40 V, source offset potential of -80 V, source temperature 120o C, and desolvation temperature of 150o C. Desolvation gas at 500 L/h and nebulizer gas at 6.5 bar, trap collision energy at -32 V was used for the negative acetone adduct ions. Leucine enkephalin was used as a lock spray compound to calibrate each spectrum for high resolution analysis. Once accurate mass was obtained for each ion in the CID spectrum, tables generated a possible identification of each ion within a +/- 2 ppm with an instrument resolution at 40,000. Eicosatrienoic acid in RAW264.7 cell Phospholipids. Phospholipids were extracted from RAW 264.7 cells (3 x 106 cells) after three days in culture according to the Bligh-Dyer method.19 The organic phase was dried under a stream of nitrogen gas and an aliquot corresponding to 5 x 106 RAW 264.7 cells, was saponified to free fatty acids as previously



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 30

8 described.20 Isooctane was used to extract free fatty acids after neutralization of the saponification mixture. An aliquot of the total extract (20 µL) was taken to dryness then subject to reversed phase LC-MS after dissolving in 20 µL of a mixture of solvent A (8.3 mM ammonium acetate, pH 5.7) and 20 µL solvent B (acetonitrile/methanol, 65/35, v/v). The sample was injected onto an HPLC column (Accucore C18, 50 x 3 mm, 2.6 µm, Thermo Scientific, Waltham, MA) and eluted at a flow rate of 300 µL/min with a linear gradient of HPLC solvent B increased from 50 to 85% in 6.5 min, to 98% for 1 min and held at 98% for a final time of 6.5 min before equilibration to 50% solvent A. The HPLC effluent was directed into an electrospray ion source of a triple quadrupole mass spectrometer (4000 QTRAP) that was continuously scanned from m/z 200 to 600 in 500 ms. After identification of the elution of the 20:3 fatty acids, the remaining portion of saponified extract was dried and resuspended as described above for injection into the reversed phase HPLC column. Fractions of this HPLC effluent were taken at 0.5 min and collected for subsequent analysis. The fraction that was collected at 7.5 min was found upon separate infusion into the mass spectrometer to contain the target 20:3 fatty acid (Supplemental Figure S3). Implementation of the Paternò-Büchi reaction16 was carried out by dissolving an aliquot of the 7.5 min fraction (approximately 20 µM) in acetone/water/ triethylamine (45/45/10, v/v/v) and directly infusing the solution through the photochemical cell attached immediately before the electrospray ion source. The tandem mass spectrometry of the oxetane adducts was carried out on a 4000 QTRAP mass spectrometer using the third quadrupole as a linear ion trap in a negative ion mode. A declustering potential of -90 V, ion spray voltage -4000 V and entrance potential -10 V were used as ion source parameters. Collision induced dissociation was carried out by ramping the


Page 9 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


9 collision energy between -35 and -40 V during elution of the oxetane, using nitrogen as the collision gas.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 30

10 Results and Discussion Octadecatrienoic Acids (18:3). Two 18:3 fatty acids fairly commonly found in nature are 6,9,12-octadecatrienoic acid, called by the trivial name γ-linolenic acid (GLA), and 9,12,15octadecatrienoic acid, called α-linolenic acid (ALA). This latter fatty acid (ALA), is the dietary fatty acid required to synthesize the members of PUFAs of the n-3 family.3 When an ALA solution was exposed to the photochemical P-B reaction conditions containing acetone, a complex series of products was formed (Supplemental Figure S2). Evidence for the acetone adduct of ALA (single P-B product) was observed at m/z 335 (ALA+58) as well as a double P-B adduct at m/z 393 (ALA+116) (Figure 1A), however, abundant ions that corresponded to the carboxylate anions of ALA at m/z 277 and a series of lower mass ions at m/z 171, 211, and 251 were also present. These latter ions exactly corresponded to expected ozone reaction products of ALA (see structure inset Figure 1A) at each of the three double bonds, after double bond cleavage and formation of an aldehyde at the double bond position.13 Evidence for these ions as photochemical reaction products of ALA and ozone was confirmed by their absence when the UV-lamp was turned off or when the solution of ALA was degassed prior to the Paternò-Büchi reaction (data not shown). This suggests that the photochemical reactor employed in this study might be a way to implement the Oz ID approach 13 but this was not further tested.

Additional ions observed were also dependent upon the presence of dissolved oxygen such as m/z 309 and m/z 291 and corresponded to ALA hydroperoxide carboxylate anion (ALA+32) and the expected dehydration product (ALA+14).21 Other UV-derived reaction products were observed ions at m/z 345 (ALA+68) and m/z 359 (ALA+82). These likely were


Page 11 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


11 formed by photochemical reaction with trace solvent (acetone) impurities and were noted to change with new solvents. The isomeric GLA yielded similar products of adducts of acetone and ozone reaction products generated from the dissolved oxygen (Figure 1B). Evidence for the position of the three double bonds in GLA came from the ions at m/z 129, 169, and 209 at the carbon atom position 6, 9, and 12 of GLA (Figure 1B). Since the aldehydic carboxylate anions did not require acetone in the mobile phase, they represent products whose formation was induced by the UV light and revealed double bond positions of this octadecatrienoic acid. Collisional activation of Paternò-Büchi adduct ion of 18:3 at m/z 335 yielded the characteristic products of the reverse Paternò-Büchi reaction and formation of aldehydic carboxylate anions or isopropenyl carboxylate anions (Figure 2) predicted for polyunsaturated fatty acids (supporting information Table S1). All of these double-bond indicating ions were of low abundance as previously noted,16,18 but well above background. The aldehydic carboxylate anions from ALA (supporting information Table S-1) decreased in abundance as the double bond receded from the carboxylate group in ALA (Figure 2A). It should be noted that the ion at m/z 277 could either be the ALA carboxylate anion or the isopropenyl carboxylate product since they are isomeric C18H29O2−. The isopropenyl carboxylate anions from GLA were very obvious at m/z 155, 195, and 235 (Figure 2B). However, the aldehydic carboxylate product ions at m/z 129, 169, and 209 were much less abundant. This trend was observed for other PUFAs studied to an even greater extent in that aldehydic carboxylate anions were not observed when four or more double bonds were present in the PUFA. Other product ions observed when both ALA and GLA acetone adducts



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 30

12 were collisionally activated were the abundant loss of acetone (original carboxylate anion) and loss of H2O (m/z 259), acetone and CO2 (m/z 233). Eicosatrienoic Acid (20:3). Several isomeric 20:3 fatty acids can be found in nature with three isomers commonly found in animal systems. These include 5,8,11-eicosatrienoic acid (n-9) that has a trivial name of Mead acid after it was identified as a chain-elongation and desaturation product of oleic acid.5,6 The isomer 8,11,14-eicosatrienoic acid (n-6) is often referred to as dihomo-γ-linolenic acid or DGLA and is an elongation product of γ-linolenic acid. The third isomer is 11,14,17-eicosatrienoic acid (n-3) and is derived from elongation of α-linolenic acid. These isomeric PUFAs all readily undergo the Paternò-Büchi reaction with formation of m/z 363. Collisional activation of each of these acetone adducts yielded a very unique series of product ions (Figure 3) of low abundance, but clearly recognizable product ions. There were also common ions corresponding to the loss of acetone (m/z 305), loss of acetone and water (m/z 287), and acetone and carbon dioxide (m/z 261) of higher abundance. The most abundant product ion in the CID mass spectrum of Mead acid-acetone adduct (aside from the carboxylate anion) appeared at m/z 233. This was likely the same ion observed after collisional activation of the arachidonate carboxylate anion22 at m/z 231 due to the additional double bond at carbon-14 in arachidonate. The unique abundance of m/z 233 likely arose due to a mechanism where the acetone adduct at carbon 5-6 of Mead acid facilitated loss of ethylene and CO2. However, it might also come directly from the carboxylate anion of Mead acid by the formation of a stabilized anionic site after delocalization over adjacent double bonds (Scheme 1). The other relevant product ions corresponded to the expected isopropenyl carboxylate anions that resulted from the retro Paternò-Büchi with cleavage at carbon atoms 5-6 (m/z 141), position 8-9 (m/z 181), and position 11-12 (m/z 221). The corresponding aldehydic


Page 13 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


13 carboxylate anions were much less abundant (supplementary Table S1). This low abundance of these aldehydic ions was characteristic of PUFA-acetone adducts with more than three double bonds

-

Scheme 1: Proposed mechanism for formation of m/z 233 from the [M-4] of Mead acid.

relatively close to the carboxylate anion.

Insert Scheme 1 here Collisional activation of the Paternò-Büchi reaction product of DGLA had ions common to Mead acid (Figure 3B) with the carboxylate anion from the retro Paternò-Büchi (m/z 305) with the additional loss of water and CO2 as described above. There was an ion observed at m/z 233 just as with Mead acid, but it was somewhat less abundant. Likely this was due to the increase in distance of the charge site of the carboxylate anion from the first double bond of this eicosatrienoic acid photochemical acetone adduct. The double bond indicating isopropenyl carboxylate anion products were observed at m/z 183, 223, and 263. Interestingly, the aldehydic carboxylate anions were now present for this isomer at m/z 157, 197, and 237. The collision induced dissociation of the 11,14,17-eicosatrienoic acid had quite obvious Paternò-Büchi fragment ions as well as the common decomposition ions discussed above (Figure 3C). There was little indication of an ion corresponding to the loss of 72 Da (m/z 233) from the carboxylate anion that was observed with the other two 20:3 isomers. The isopropenyl carboxylate product ions indicating double bond positions were quite abundant at m/z 225, 265,



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 30

14 and 305. This latter ion was isomeric with the carboxylate anion as discussed above for the 18:3 n-3 series. For this eicosatrienoic acid, the aldehydic carboxylate anions were quite noticeable as well as at m/z 199, 239, and 279. Each of the eicosatrienoic acids had a very unique product ion spectrum that specifically identified double bond positions in these polyunsaturated fatty acids. Structural characterization of accumulating 20:3 in cultured RAW 264.7 cells. The characterization of the 20:3 fatty acid that accumulated in phospholipid molecular species after three days in tissue culture of RAW 264.7 cells was carried out using the P-B reaction. The phospholipid extract of RAW 264.7 cells was saponified and free fatty acids extracted using isooctane. Free fatty acids were separated by reversed-phase chromatography and 0.5 min fractions collected. A small aliquot of this extract was subjected to reversed phase LC-MS using the same column and mobile phase conditions (Figure 4A). Analysis of the components eluting from HPLC revealed the fatty acids as expected, including the presence of an abundant 20:3 isomer eluting at approximately 7.6 min, one less abundant isomer eluting at 7.3 min and a minor isomer at 6.7 min (Figure 4A). Infusion of a small portion of the HPLC fractions (fraction called 7.5) confirmed the presence of only the most abundant 20:3 fatty acid isomers eluting around 7.5 min but 18:1 and 16:0 fatty acids were also present in this fraction (Figure S3). An aliquot of the 7.5 min fraction was then mixed with acetone and subjected to the online Paternò-Büchi reaction. Three compounds were present in this fraction in significant amounts of fatty acids, which were 16:0, 18:1, and one 20:3 isomer. The latter two unsaturated fatty acids yielded oxetane adducts with acetone upon UV irradiation, which appeared at m/z 339 and 363, respectively. Collisional activation of the m/z 363 ion yielded a product ion spectrum only consistent with 5,8,11-eicosatrienoic acid, i.e., Mead acid (n-9) (Figure 5B). The isopropenyl carboxylate anions at m/z 141, 181, and 221 were quite abundant and indicative of


Page 15 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


15 the double bond positions 5, 8, and 11, respectively. No other product ions characteristic of the other 20:3 regioisomers were detected. The presence of 18:1 and 16:0 were also observed as product ions at m/z 255 and 281 when the m/z 363 ion was collisionally activated. These were observed likely because of their very high concentration in this HPLC fraction and the formation of adducts that fortuitously yielded m/z 363. For example, the formation of a minor [M+82]adduct ion of PUFAs has been previously observed (Figure 1B) and would be an adduct ion of 18:1 (281+82), that would appear at m/z 363. Other abundant product ions from collisional activation of m/z 363 included m/z 261 and 233, which were observed as product ions of authentic 5,8,11-eicosatrienoic acid, but did not carry any double bond position information. Thus, the Paternò-Büchi photochemical reaction with acetone unambiguously identified the abundant 20:3 present in the RAW 264.7 cells carried in culture for three days as Mead acid (n9). Eicosatetraenoic Acid (20:4). There are only two isomeric eicosatetraenoic acids made in mammalian biochemistry. One is from the n-3 family of PUFAs after elongation and a single desaturation of α-linolenic acid to form 8,11,14,17-eicosatetraenoic acid (n-3). The much more abundant 20:4 is arachidonic acid, 5,8,11,14-eicosatetraenoic acid (n-6) derived from ∆6desaturase and ∆5-desaturase desaturation reaction after elongation of linoleic acid.3 Previous attempts to apply the Paternò-Büchi reaction to determine positions of double bonds and arachidonic acid from the carboxylate anion were considered to be not successful in generating diagnostic ions.18 In part this was due to the large number of product ions normally derived from collisional activation of the [M-H]− from arachidonate22 and the rather low abundance of the Paternò-Büchi product ions. Thus, there was an array of ions of very similar low abundance (Figure 5A). The most abundant product ion following collisional activation of m/z 361 was the



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 30

16 loss of water (m/z 343), loss of acetone in the retro Paternò-Büchi reaction (m/z 303) which was also the carboxylate anion of arachidonic acid, loss of water and acetone (m/z 285), and decarboxylation of the arachidonate anion after loss of acetone (m/z 259). The rather abundant ions at m/z 205 and 231 had been previously described with the mechanism of formation of m/z 205 coming from two separate pathways.22 The ion at m/z 231 was likely due to a loss of 72 Da from the arachidonate carboxylate anion to form an ion stabilized by adjacent double bonds as was observed in the other PUFAs. The formation of this ion might be facilitated by an oxetane adduct at carbon-5,6 with a concerted loss of these small neutral product ions as presented in Scheme 2.

Scheme 2: Proposed mechanism for formation of m/z 231 following collisional activation of the [M-H]- from arachidonic acid oxetane adduct. Other mechanisms are possible that place the anionic site at doubly vinylic positions. Insert Scheme 2 here

In order to facilitate identification of the relevant Paternò-Büchi fragment ions, this reaction was carried out in approximately 1:1 mixture of D0-acetone:D6-acetone. High quality CID spectra could be obtained by alternating collisional activation of m/z 361 and 367 at 400 ms intervals during collection of the product ions in order to measure abundances of each D0 and D6 adduct product ion under conditions identical as possible. The collision spectra of the D6-acetone adduct (Figure 5B) was essentially identical to that of the D0 adduct except for a few minor ions.


Page 17 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


17 These two spectra were sequentially compared, ion by ion, using a computer program that extracted only those ion pairs separated by 6 Da of similar abundance ratios to that of the ratio of m/z 361 (D0) and 367 (D6). The resulting difference plot showing those ion pairs that satisfied these criteria revealed the expected retro Paternò-Büchi fragment ions from arachidonate (Figure 6). The abundant product ions not retaining six deuterium atoms from the D6-acetone reaction were filtered out by this computer routine. Even though the Paternò-Büchi fragment ions were of low abundance, the D0/D6 differential extraction algorithm readily isolated those ions that marked positions of double bonds in arachidonate. Eicosapentaenoic Acid (EPA) and Docosapentaenoic Acid (DHA). Two important and abundant polyunsaturated fatty acids commonly encountered are 5,8,11,14,17-eicosapentaenoic (EPA) acid and 4,7,10,13,16,19-docosahexaenoic acid (DHA). Both of these fatty acids are of the n-3 family and accumulate in various tissue phospholipids. The latter PUFA is particularly abundant in certain nervous tissues such as the retina.23 Collisional activation of the [M-H]- from EPA and DHA led to abundant product ions greater in number and abundance than even PUFAs containing 4 double bonds (e.g. arachidonic acid) which limited the production of highly abundant P-B adduct ions (Supplemental Figure S4). The approach of making the [M-2H+Na]- anions that leads to localization of the negative charge localized within the double bond region, yields recognizable product ions after CID that can be used to localize double bonds even with 22:6 species, 15 making this a preferred method even though not as facile and perhaps as sensitive as the P-B reaction strategy.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 30

18 Conclusions The Paternò-Büchi reaction is quite favorable for polyunsaturated fatty acids reacting with acetone to yield multiple adducts with the most abundant being a single acetone adduct randomly placed at each of the double positions of the PUFA. These adducts can be formed in a flowing solvent system in a fused silica capillary by photoactivation of acetone with each of the double bonds in the PUFA. This photochemical interface is readily transferred from one mass spectrometer to another enabling experiments to be carried out by multiple instruments. Very characteristic product ions can be observed following CID of [M-H]– ions of 18- and 20-carbon PUFAs that reveal positions of double bonds when three or four double bonds are present. The most abundant reverse Paternò-Büchi reaction product anions are the isopropenyl carboxylate anions. Other abundant product ions not related to the Paternò-Büchi reaction increased as the number of double bonds increase in these long chain fatty acids. This reduces the ability to identify unique Paternò-Büchi product ions in fatty acids containing 5 or 6 double bonds. A mixture of D0/D6-acetone enables one to uniquely find isopropenyl carboxylate anions by an appropriate 6 Da mass shift when high quality MS/MS spectra are obtained. This unique photochemical reaction product with acetone as well as the observed isopropenyl carboxylate anions permit an unambiguous assignment of double bond positions for PUFAs with four double bonds or less and fingerprint identification of those PUFAs with more than four double bonds.


Page 19 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


19 Acknowledgements This study was supported in part by a grant from the National Institutes of Health (HL 117798) (R.C.M.) and grants from JSPS KAKENHI Grant Number 16K08596 (T.O.), 15KK0320 (T.O.) and Takeda Science Foundation (T.O.).

Conflict of Interest Disclosure The authors declare no competing financial interest. Supporting Information Available: Supplemental Table S1, Supplemental Figures S1, S2, S3, and S4



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 30

20

References 1.

Samuelsson, B. J Biol. Chem 2012, 287, 10070-10080.

2.

Smith, W. L. Trends Biochem. Sci. 2008, 33, 27-37.

3.

Chilton, F. H.; Murphy, R. C.; Wilson, B. A.; Sergeant, S.; Ainsworth, H.; Seeds, M. C.; Mathias, R. A. Nutrients. 2014, 6, 1993-2022.

4.

Igal, R. A. Biochim. Biophys. Acta 2016, 1861, 1865-1880.

5.

Fulco, A. J.; Mead, J. F. J Biol. Chem 1959, 234, 1411-1416.

6.

Lundberg, W. O. Nutr. Rev. 1980, 38, 233-235.

7.

Murphy, R. C. In Mass Spectrometry of Lipids, The Handbook of Lipid Research; Snyder, F., Ed.; Plenum Press: New York, 1993, pp 88-99.

8.

Michaud, A. L.; Yurawecz, M. P.; Delmonte, P.; Corl, B. A.; Bauman, D. E.; Brenna, J. T. Anal. Chem. 2003, 75, 4925-4930.

9.

Shibamoto, S.; Murata, T; Yamamoto, K. Lipids 2016, 51, 1077-1081.

10.

Jensen, N. J.; Tomer, K. B.; Gross, M. L. Anal. Chem. 1985, 57, 2018-2021.

11.

Hsu, F. F.; Turk, J. J. Am. Soc. Mass Spectrom. 2008, 19, 1673-1680.

12.

Thomas, M. C.; Mitchell, T. W.; Harman, D. G.; Deeley, J. M.; Nealon, J. R.; Blanksby, S. J. Anal. Chem 2008, 80, 303-311.

13.

Brown, S. H. J.; Mitchell, T. W.; Blanksby, S. J. Biochim. Biophys. Acta 2011, 1811, 807817.

14.

Castro-Perez, J.; Roddy, T. P.; Nibbering, N. M. M.; Shah, V.; McLaren, D. G.; Previs, S.; Attygalle, A. B.; Herath, K.; Chen, Z.; Wang S-P.; Mitnaul, L.; Hubbard, B. K.; Vreeken, R. J.; Johns, D.G.; Hankemeier, T. J. Am. Soc. Mass Spectrom. 2011, 22, 1552-1567.


Page 21 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


21 15.

Thomas, M. C.; Altvater, J.; Gallagher, T. J.; Nette, G. W. J. Am. Soc. Mass Spectrom. 2014, 25, 1917-1926.

16.

Ma, X.; Chong, L.; Tian, R.; Shi, R.; Hu, T. Y.; Ouyang, Z.; Xia, Y. Proc. Natl. Acad. Sci. U. S. A 2016, 113, 2573-2578.

17.

Ma, X.; Xia, Y. Angew. Chem Int. Ed Engl. 2014, 53, 2592-2596.

18.

Ma, X.; Zhao, X.; Li, J.' Zhang, W.; Chen, J-X.; Ouyang, Z; Xia, Y. Anal. Chem. 2016, 88, 8931-8935.

19.

Bligh, E. G.; Dyer, W. J. Can. J. Biochem. Physiol. 1959, 37, 911-917.

20.

Quehenberger, O.; Armando, A. M.; Dennis, E. A. Biochim. Biophys. Acta 2011, 1811, 648-656.

21.

MacMillan, D. K.; Murphy, R. C. J. Am. Soc. Mass Spectrom. 1995, 6, 1190-1201.

22.

Murphy, R. C. In Tandem Mass Spectrometry of Lipids: Molecular Analysis of Complex Lipids; Gaskell, S., Ed.; Royal Society of Chemistry: London, 2015; pp 1-39.

23.

Zemski Berry, K. A.; Gordon, W. C.; Murphy, R. C.; Bazan, N. G. J Lipid Res. 2014, 55, 504-515.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 30

22 Figure Legends Figure 1 – Mass spectrum of the infusion of (A) 9,12,15-octadecatrienoic acid (α-linolenic acid, ALA) and (B) 6,9,12-octadecatrienoic acid (γ-linolenic acid, GLA) following the photochemical reaction of acetone with these polyunsaturated fatty acids in the Paternò-Büchi reaction. The locations of double bonds and predicted ozone fragment carboxylate anions and the structures of these ions are indicated. The abundance of the ions aside from the carboxylate anion were increased by a factor of 5 and at where indicated. Figure 2 – Collision induced dissociation of the Paternò-Büchi adduct of acetone as [M-H]– with α-linolenic acid and γ-linolenic acid. (A) Collisional activation of the [M-H]– m/z 335 from αlinolenic acid and the indicated formation of the decomposition fragments of the oxetane at each double bond position as either the aldehydic ion or the isopropenyl ion icon. The fragmentation leading to m/z 237 is indicated in the inset structure. (B) Collisional activation of γ-linolenic acid as the acetone oxetane adduct ion [M-H]– at m/z 335. Formation of each double bond indicating fragment ions are shown. Figure 3 – Collision induced dissociation of the Paternò-Büchi adduct ions of acetone as [M-H]– of three regioisomeric eicosatrienoic acids . (A) Collisional activation of the [M-H]– m/z 363 from Mead acid, 5,8,11-eicosatrienoic acid (n-9) as the acetone oxetane adduct. (B) Collisional activation of the [M-H]– m/z 363 from dihomo-γ-linolenic acid, 8,11,14-eicosatrienoic acid (n-6) as the acetone oxetane adduct. (C) Collisional activation of the [M-H]– m/z 363 from 9,14,17eicosatrienoic acid (n-3) as the acetone oxetane adduct. Formation of the double bond indicating fragment ions are shown as structural icons. Figure 4 (A) Extracted ion chromatograms for m/z 303 (20:4, blue trace) and m/z 305 (20:3, red trace) from reverse phase HPLC separation of saponified fatty acyl groups of RAW264.7 cells in


Page 23 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


23 culture for 3 days. Fraction between 7.5 and 8.0 min was collected for further analysis to determine the double bond positions of the major 20:3 isomer (7.57 retention time). A total of three isomers of 20:3 were separated by this LC system. (B) Tandem mass spectrum of the oxetane adduct of the 20:3 fatty acid (m/z 363.3) in the reverse phase HPLC fraction collected between 7.5 and 8.0 min (figure 4A) obtained by the Paternò-Büchi reaction with acetone. The structure inset is for one of the three different oxetane locations on the 20:3 fatty acid. The ions indicative of the location of the double bonds of the oxetane are noted as product ions retaining the isopropenyl group (

icon). The structures of these product ions are indicated below the

mass spectrum with the theoretical mass. Figure 5 – Collisional activation of the acetone oxetane adduct of arachidonic acid with the isopropenyl ions that indicate double bond positions are annotated. (A) Photochemical reaction using unlabeled acetone with arachidonic acid to form the oxetane adduct [M-H]–. (B) Photochemical reaction using D6-acetone with arachidonic acid leading to formation of the oxetane adduct [M-H]– at m/z 367 for the carboxylate anion. Note the magnification factor of the abundance of ions in the spectra. Figure 6 – Different spectra between D0- and D6-acetone photochemical adduct for those ion pairs separated by 6 Da which have similar abundances to the respective [M-H]– ions. The isopropenyl and the D6-isopropenyl adduct ions are indicated as positive and negative relative abundance respectively.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53

Page 24 of 30

Bare Fused Silica

UV hν


Page 25 of 30

A.

α-18:3 MS PB products

100

18:3 277.2

x 5.0

x 5.0

O

200

B.

18:3 277

233

x 15

309.2

18:3-PB2 359.2 393.4

300 m/z

γ-18:3 MS PB products

100

211 O3

345.2

291.2

127.1 100

18:3-PB1 335.4

251.2

171.1

O

255.2

O

211.2 233.3

Relative Abundance

171 O3

147.0

Relative Abundance

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53


COO-

251 O3

400

500

x 15 169 O3

O3 129

bkg

171 O

169

O

129

209 bkg 212

191

200

209 O3

18:3-PB1 335 345

O

309 255

291 300

COO-

18:3-PB2 393

359 387

417

400

500

m/z

Figure 1



A. α-18:3 CID m/z 335

100

277 [M-H-58]-

Relative Abundance

x 25

237

[M-H-58-44]233

COOO

[M-H-58-18]259 O

171

O

211

157

[M-H]335

237

209

197

200

O

251

[M-H-18]317 300

m/z

x 20

100

[M-H-58-44]233

B. γ-18:3 CID m/z 335

[M-H-58-18]259 277 [M-H-58]-

195 O

Relative Abundance

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53

Page 26 of 30

COO-

205 O

129

[M-H]335

235

195 155

O

O

169

209 200

m/z Environment ACS Paragon Plus

[M-H-18]317 300

Figure 2

Page 27 of 30

x 15

A

[M-H-58-44]

261

Relative Abundance

100 20:3n-9

[M-H-58]

-

287

305

181

O

[M-H]363

259

181

141

221

COO-

-

233

200

345 300

m/z

400 COO-

B

20:3n-6

Relative Abundance

100

223

O

157

263 261

183

139

O

197

[M-H-58-18]-

x 20

305

[M-H-58]O

287

O

233 237

200

[M-H]363

263

345 300

m/z

400

x 20

COO-

[M-H-58-44]-

C

261

20:3n-3

O

225

O

199

200

237 239

100

Relative Abundance

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53


m/z

[M-H-58]-

305

265 287

[M-H]363

O

O

265

279

345 300


400 Figure 3


A

7.57 m/z 305

Extracted ions 20:4 20:3

3-isomers of 20:3

m/z 305 7.33

m/z 303 6.74

m/z 305 6.72

2

4

6

8

10

12

14

16

18

Retention time (min)

B

18:1 281

100

141

O

x 20.0 Relative Abundance

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53

Total ion signal (cps)

5.4e8

Page 28 of 30

[M-H]-

[M-H-(58-44)]261 233

141

255

16:0

159 181

m/z 363.291

[M-H-58]305

221

200 -

COO m/z 141.092

m/z

COO-

300

400 -

COO m/z 181.123

[M-H]363

COO-

m/z 221.155

Figure 4


Page 29 of 30


1 2 3 4 5 6 7 8 9 10 11 12 x 5.0 x 5.0 13 14 303 100 15 16 MS/MS 17 259 m/z 361 18 181 205 [M-H]19 361 231 301 20 221 343 177 21 261 285 191 217 163 141 317 22 261 23 24 120 140 160 180 200 220 240 260 280 300 320 340 360 380 25 x 5.0 x 5.0 26 303 27 100 259 MS/MS 28 D3C CD3 205 305 D3C CD3 m/z 367 29 D3C CD3 30 D3C CD3 231 [M-H]187 31 227 367 267 32 177 217 147 163 33 191 285 349 34 323 35 36120 140 160 180 200 220 240 260 280 300 320 340 360 380 37 m/z 38 39 Figure 5 40 41 42 43 44 45 46 47 48 49 50 ACS Paragon Plus Environment 51 52 53

A

B


AA-PB [M-H]361

2

Percent Relative Abundance

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53

1

Page 30 of 30

[M-H-18]343

141

263 [M-H-44]221 261 317 223

181

0

-1

D3C

147

CD3

229

187

D3C

CD3

227

D3C

-2 100

150

200

323

267 269

CD3

D3C

250

CD3

300

349

350

m/z

Figure 6


D6-AA-PB [M-H]367 400

Determination of Double Bond Positions in ... - ACS Publications

Recommend Documents