Distinguishing Cis and Trans Isomers in Intact Complex lipids using

Subscriber access provided by EAST TENNESSEE STATE UNIV

Article

Distinguishing Cis and Trans Isomers in Intact Complex lipids using Electron Impact Excitation of Ions from Organics (EIEIO) Mass Spectrometry Takashi Baba, J Larry Campbell, J.C. Yves Le Blanc, and Paul R. S. Baker Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b04734 • Publication Date (Web): 14 Jun 2017 Downloaded from http://pubs.acs.org on June 18, 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 31

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

Distinguishing Cis and Trans Isomers in Intact Complex lipids using Electron Impact Excitation of Ions from Organics (EIEIO) Mass Spectrometry Takashi Baba1, J. Larry Campbell1, J. C. Yves Le Blanc1, and Paul R. S. Baker2 1

2

SCIEX, 71 Four Valley Dr., Concord, Ontario L4K 4V8, Canada

SCIEX, 1201 Radio Rd, Redwood Shores, California 64065, USA

Correspondence: Takashi Baba Phone: 289 982 2233

[email protected] Fax: 905 660 2623

Abstract We present a mass spectrometry-based method for the identification of cis and trans double bond isomers within intact complex lipid mixtures using Electron Impact Excitation of Ions from Organics (EIEIO) mass spectrometry. EIEIO involves irradiating singly charged lipid ions with electrons having kinetic energies of 5-16 eV. The resulting EIEIO spectra can be used to discern cis and trans double bond isomers by virtue of the differences in the fragmentation patterns at the carbon-carbon single bonds neighboring the double bonds. For trans double bonds, these characteristic fragments include unique closed-shell and open-shell (radical) products. To explain this fragmentation pattern in trans double bonds, we have proposed a reaction mechanism involving excitation of the double bond’s π electrons followed by hydrogen atom rearrangement. Several lipid standards were analyzed using the EIEIO method, including mixtures of these standards. Prior to EIEIO, some of the lipid species in these mixtures were separated from their isomeric forms by using differential mobility spectrometry (DMS). For example, mixed cis and trans forms of triacylglycerols and phosphatidylcholines were identified

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

by this DMS-EIEIO workflow. With this combined gas-phase separation and subsequent fragmentation, we could eliminate to the need for authentic standards for identification. When DMS could not separate cis and trans isomers completely, as was the case with sphingomyelins, we relied upon the aforementioned diagnostic EIEIO fragment peaks to determine the relative contribution of the trans double bond isomer in the mixed samples. We also applied the DMSEIEIO methodology to natural samples extracted from a ruminant (bovine), which serve as common origins of trans fatty acids in a typical Western diet that includes dairy products.


Page 2 of 31

Page 3 of 31

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


Introduction The identification of cis and trans double bond isomers within complex lipidomics samples remains challenging in spite of the well-known detrimental behaviors of trans fats in biological systems. For example, trans fatty acids (FAs) play an adverse role in the pathophysiology of cardiovascular disease [1] and brain function [2]. When trans FAs are incorporated into a phospholipid bilayer, the transport and fluidity properties can be altered dramatically compared to bilayers normally containing lipids with only cis double bonds or no double bonds at all [3]. The presence of trans FAs in red blood cell membranes (at ~ 0.5-1.5% of total FA content) has also been related to the likelihood of cardiac arrest [4,5], but it has also been linked to greater longevity [6]. The primary dietary source of trans fats is cow milk products;[3,7-9] while the overall relative abundance of trans FAs in milk products is only a few %, very high concentrations (C18:1 = 94.2%, C22:1 = 98.2%, C23:1 = 63.9%, C24:1 = 23.4%) of trans double bonds have been reported in sphingomyelins (SMs) in cow milk [7]. However, trans fats are also widely known from their origins via the catalytic hydrogenation of poly-unsaturated fatty acids [3]. The free radical path to produce trans FAs in membranes has also been reported [3,10]. While many biochemical questions related to cis and trans lipid isomers have garnered attention over the years, advancements in the bioanalytical techniques used to characterize double bond isomerism in lipids has been somewhat lagging [3]. The most common means of trans fat analysis is performed using chromatographic separation (either by gas chromatography (GC) or liquid chromatography (LC)) of free fatty acids (FFAs) or fatty acid esters, which have been hydrolyzed from intact triacylglycerols (TGs) or other lipid classes[11]. The elution times for these samples are compared to authentic standards, thereby ensuring confidence in 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

analysis.[3] The analyses of intact isomeric lipids containing cis or trans double bonds has also been performed using LC [12] and ion mobility spectrometry (IMS) [13,14]. While mass spectrometers alone (without the assistance of chromatography) can provide many details about the molecular structures of lipids, their ability to distinguish and identify isomers (specifically cis and trans double bond isomers) can be quite limited. Often, successful isomer identification by MS comes via the implementation of unique fragmentation or other ion transformation methods. For example, Voinov and coworkers found that subjecting deprotonated FAs to resonant electron capture dissociation provided distinct fragmentation patterns based upon the cis/trans configuration of double bonds in those FAs [15]. Hejazi and coworkers reported distinguishing cis/trans isomers in linolenic FAME using low energy electron ionization [16]. Jensen and coworkers [17], as well as Ji and coworkers, [18] reported that fast atom bombardment also distinguished cis/trans in deprotonated FAs by virtue of the different fragmentation patterns each isomer exhibited upon desorption/ionization via high-energy bombardment. In some cases, chemical derivatization of FAs, followed by conventional collision-induced dissociation, has assisted in the MS-based discrimination of cis and trans FAs [19]. Unfortunately, the bulk of these alternative techniques have been applied only to FAs or their esters, which are the hydrolysis products of larger lipid classes (e.g., glycerophospholipids (GPLs) like phosphatidylcholines (PCs), sphingomyelins (SMs), and triacylglycerols (TGs)) derived from natural complex lipid samples. As a result, structural information of the relative positioning of the constituent fatty acids within these intact lipids (i.e., their speciation and regioisomerism) has been lost. One method, ozone-induced dissociation (OzID) approach, has also shown promise as a tool to distinguish cis/trans isomers in intact phospholipids [20].


Page 4 of 31

Page 5 of 31

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


In a recent study, we demonstrated the "near-complete Structural Characterization of phosphatidylcholines (PC) using Electron Impact Excitation of Ions from Organics (EIEIO)” [21,22]. Therein, diagnostic product ion peaks were found that enabled identification of the lipid head group, its acyl chains, the regioisomer arrangement of the acyl chains, and the position of the fatty acid double bonds – all in a single EIEIO spectrum. However, the identification of cis/trans isomerism at double bonds within the acyl chains of such lipid molecules remained an unsolved challenge for lipid structural identification. After further investigation, we now present in this work, characteristic differences in the EIEIO fragmentation patterns of cis and trans double bond isomers of PCs, SMs, and TGs using synthetic standards. This technique was applied to biological samples extracted from ruminant (bovine) as possible sources of complex lipids with trans double bonds in the chains [7,13].

Methods Materials Synthetic PC standards, PC 16:1(n-7,cis)/16:1(n-7,cis), PC 16:1(n-7,trans)/16:1(n-7,trans), PC 18:1(n-9,cis)/18:1(n-9,cis) (DOPC), PC 18:1(n-9,trans)/18:1(n-9,trans) (DEPC), PC 16:0/18:1(n9,cis) (POPC) and PC 16:0/18:1(n-9,trans) (PEPC), and SM d18:1/24:1(n-9,cis) were purchased from Avanti Polar Lipids, Inc. (Alabaster, Alabama). Standard TG samples, TG 16:1(n7,cis)/16:1(n-7,cis)/16:1(n-7,cis), TG 16:1(n-7,trans)/16:1(n-7,trans)/16:1(n-7,trans), TG 18:1(n9,cis)/18:1(n-9,cis)/18:1(n-9,cis) and TG 18:1(n-9,trans)/18:1(n-9,trans)/18:1(n-9,trans) were purchased from Nu-Chek Prep, Inc. (Elysian, Minnesota). These samples were used for method validation as well as standards to be compared to natural samples. Naturally-sourced lipid extracts, including bovine liver extract, bovine heart extract, bovine milk SM, and porcine brain SM, were purchased from Avanti. An acetone precipitation method was applied to the liver and



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

heart extracts to concentrate polar lipids. Milk fat globular membrane (MFGM) in cow cream was extracted in house (please refer to Supporting Information). All lipid extract stock solutions were dissolved in chloroform. HPLC-grade solvents were purchased from Caledon Laboratory Chemicals (Georgetown, Ontario), while the sodium acetate and ammonium acetate were purchased from Sigma-Aldrich Canada Co. (Oakville, Ontario). HPLC-grade dichloromethane (DCM): methanol (MeOH) (50:50, v/v containing 0.5 mM ammonium acetate) was used to make a working solution for PC and SM analysis. Using this solvent, electrospray ionization (ESI) produced protonated PCs and SMs, [M+H]+. For ionization of TGs as the sodiated form, [M+Na]+, we employed a different ESI solvent - DCM : MeOH (50:50, v/v containing 0.5 mM sodium acetate). Though PCs and SMs were ionized efficiently in the sodiated form using the latter solvent, we did not use sodiated ions because EIEIO on protonated forms produced chain fragments more intense than the sodiated forms. The lipid standards were diluted to a concentration of 1 µg/mL for the experiments. The natural lipid extracts were diluted in the ammonium acetate solvent with a concentration of 100 µg/mL of total lipid content.

Instrumentation Lipid ions were generated by electrospray ionization (ESI) by infusing the working solutions at 0.3 mL/h into a Turbo VTM source (Sciex, Concord, Ontario). The spray voltage was held at +5000V. All experiments were performed using a mass spectrometer with ESI-DMS-Q-ExDTOF configuration, which was reported previously [23]. The DMS (SelexION® technology, SCIEX) was operated at a separation voltage (SV) of 3900 V while the compensation voltage (COV) was ramped, typically from 0 V to +10 V. The DMS cell was heated at 200 ºC, and 2-


Page 6 of 31

Page 7 of 31

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


propanol was mixed into the nitrogen curtain gas (also the DMS transport gas) as a chemical modifier. DMS resolving gas (DR) was also used for resolution enhancement of the DMS measurements.[23] By using DR, the residence time of isomeric ions between the planar DMS electrodes is extended such that resolution of these species is increased although the ion transmission is decreased (Figure SI-1). The EIEIO cell was operated in simultaneous trapping mode (50-ms trap times) [22]. Nitrogen buffer gas was introduced into the EIEIO cell, which is the same pressure as the Q2 collision cell. Electrons were produced by heated yttria-coated iridium disk (ES-525, Kimball Physics, Wilton, New Hampshire).The electrons were accelerated by the electric DC bias between the iridium disk and the ion trapping electrodes. The electrons exhibited a turn-around trajectory or were possibly trapped along the magnetic field lines because the opposite side of the electron source was biased negatively. Vacuum residual gas in the cell was ionized by the electrons in our EIEIO energy range (5-16eV) so that EIEIO spectra were often contaminated by such electron ionization (EI) noise. We always subtracted the EI noise from EIEIO spectra, where the EI noise spectrum was obtained at the same electron energy used in the EIEIO experiment.

Results and discussion Dissociation of chains with a cis or trans double bond by EIEIO To begin our evaluation, we subjected two PC isomers to EIEIO using an electron kinetic energy (eKE) of 8 eV. These two isomers, PC 16:1(n-7, cis)/16:1(n-7,cis) and PC 16:1(n7,trans)/16:1(n-7,trans), provided useful fragmentation patterns (Figure 1 and Figure SI-2) similar to those observed in our earlier studies [22, 24]. We integrated spectra for 10 min each. In this condition, statistical fluctuation of peak heights was 3% in the displayed m/z region in



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

Figure 1. For example, these samples were confirmed as PC 16:1(n-7)/16:1(n-7) by (1) the diagnostic PC peak of m/z 184.1 (head group fragment), (2) a sn-2 diagnostic peak at m/z 465.3, which indicated the sn-2 acyl chain was 16:1. Subtraction of the sn-2 diagnostic m/z from the precursor m/z revealed that the sn-1 chain was also 16:1. EIEIO also identified the double bond position at n-7, which was identified by the 2H shift (Figure 1) and the V-shape intensity profile (Figure SI-2)[22]. This constitutional structure was consistent with the vendor's specification of the samples. Neither contamination of other acyl chain lengths, double bond numbers, nor double bond positons was detected in the samples. The two acquired spectra of the cis and trans PC standards had very similar intensity profiles (Figure SI-2), but characteristic differences were present in the fragmentation patterns generated at the n-6 and n-8 bonds (Figure 1), which were neighboring carbon-carbon single bonds of the double bond at n-7. At the n-6 bond (in the methyl terminal side), the radical fragment (denoted by a black dot) and the hydrogen-loss non-radical fragment had different intensities between cis and trans isomers. At n-8 carbon-carbon bond, intensities of the hydrogen “gain” fragment and the radical fragment were different between cis and trans isomers. This hydrogen gain-type fragment was exclusively observed at the single bond adjacent to a carboncarbon double bond in the head group side. No apparent difference was observed in minor fragments at the double bond position (around m/z 632) and also in other carbon-carbon single bond fragments. For the more detailed survey, electron kinetic energy, eKE, was scanned around 10 eV (Figure 2 (a), (b), and Figure SI-3). As already reported [24], an abundance of the hydrogen lost fragments increased as eKE was increased because higher electron energy induced the secondary hydrogen loss process. Similar patterns on eKE were observed at many positions of carbon-


Page 8 of 31

Page 9 of 31

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


carbon single bonds in the both cis-PC and trans-PC (Figure SI-3), but eKE dependence at n-6 and n-8 bond (Figure SI-3(h) and (n)) was quite different from others as expected from Figure 1. More specifically, eKE dependence of the cis isomer at n-6 position looked similar to other carbon-carbon single bonds, but the trans isomers showed a significant enhancement of the hydrogen lost species. At higher energy, cis and trans isomers were similar in both hydrogen loss and hydrogen gain because extra energy may randomize the isomeric difference. To enhance the differences between the radical and the hydrogen loss peaks, “normalized , is introduced as intensity difference”, ∆I

() = – . ∆

Because the denominator of this formula did not show the difference between cis and trans isomers (Figure 2(c)), the numerator must reflect the difference. This normalization is useful for practical analysis because precursor intensity varies for each analysis. Electron energy at n-6, ∆I dependence of ∆I

(n-6),

is shown in Figure 2(d), where the trans isomer ∆I ratio

appeared in the positive region, but the cis isomer ∆I ratio appeared in negative at lower eKE on the contrary. The cis isomer ∆I ratio crosses zero (negative to positive) at eKE = 13.5 eV when around eKE = 7.0 eV. The eKE was scanned from low to high. The trans sample had zero of ∆I normalized difference of the H gain species was shown in Figure SI-4(b); however, we did not focus on this aspect in this report. Similar to the H lost case, the trans isomer ∆I ratio gave a higher value than the cis isomer. Figure SI-5(b) shows normalized intensity difference of the standard TGs, where ∆I

(n-6)

and ∆I

(n-8)

are displayed for TG 16:1/16:1/16:1 and TG

18:1/18:1/18:1, respectively. The difference of profiles between cis and trans isomers was quite



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 31

similar to the PC cases. The data shows that the same method to distinguish cis and trans isomers in phospholipids is applicable to TGs, as well. Figure 3 shows a comparison between PC 16:0/18:1(n-9, cis) (POPC) and PC (n-8), where a mono-unsaturated acyl is expressed with another 16:0/18:1(n-9, trans) (PEPC) in ∆I saturated acyl in a single molecule. PC 18:1(n-9, trans)/18:1(n-9, trans) (DEPC) and PC 18:1(n-9, cis)/18:1(n-9, cis) (DOPC) are also displayed in Figure 3 for comparison. For the PEPC, eKE dependence was located at the middle between the DEPC and DOPC. For the POPC, in contrast, eKE dependence was close to DOPC. It is easily understood because the saturated acyl at the sn1 site contributes just like a cis-type, and only one 18:1(n-9,trans) acyl group at the sn-2 site contributed as the trans type. Other possible mixtures of cis and trans double bonds both being present in one molecule includes one of the chains containing a trans double bond and the other a cis double bond. Although the double bonds in each chain position the same location, EIEIO can tell that the double bonds are double trans, single trans and cis, or no trans, because only the trans chains show a high intensity at the diagnostic non-radical fragment ion. However, in the case of the single trans and single cis, two regioisomers, PC cis/trans and PC trans/cis are difficult to distinguish because the regioisomer diagnostic peak [22] can distinguish constitutional isomers but not distinguish positional isomers as well as cis/trans isomers.

Proposed reaction mechanism The enhancement of the H gain and the H loss fragment ions forming at the neighboring carboncarbon single bonds at a trans double bond and the similarity of cis double bond and single bond cleavage suggests a reaction mechanism to distinguish cis and trans double bond by EIEIO. At


Page 11 of 31

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


single carbon-carbon bonds, homolytic scission by an energetic electron is the main process for cleavage to produce radical fragments (Scheme 1(f)-(k)) as reported by Yang and coworkers in dissociation of charge-switch derivatized FAs [25]. Hydrogen loss from the radical fragments follows to produce the non-radical fragments. The existence of odd-electron fragments in EIEIO spectra reveal that the radicals are stable, but it is reasonable that this second hydrogen loss process is enhanced by elevating electron energy (Figure SI-3). We also observed fragment intensity enhancement at the next neighbor carbon-carbon single bonds from a double bond [22] (n-5 and n-9 positions in Figure 1). As described by Yang and coworkers [25], these enhancements would be explained by McLafferty rearrangement (Scheme 1(m)) and allylic cleavage (Scheme 1(n)). These components overlap on the previously described hydrogen loss products. We propose that the double bond is also excited by the electron with eKE ~ 10eV to open the π bond of the double bond or the double bond’s π electrons (Scheme 1(a) and (d)). The diradical state can induce hydrogen rearrangement and cleavage. In the case of a trans double bond, hydrogen rearrangement should be possible without rotation of the σ bond (Scheme 1d). The hydrogen rearrangement can be bi-directional because of symmetry, and two types of fragmentation are possible, i.e., H gain process (Scheme 1b) and H gain process (Scheme 1c). The contribution of these processes is superimposed to the regular homolytic bond cleavage process (Scheme 1-i and 1-j). In contrast, the case of the cis double bond inhibits such hydrogen rearrangement process because the rotation of the σ bond needs extra energy and time to overcome the big momentum inertia of the terminal side chain. Such inhibition of cis double bond cleavage explains that the observed eKE dependence of the cis double bond is similar to the carbon-carbon single bonds.



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 31

This cis/trans identification method may not work on poly-unsaturated acyls and conjugated acyls because other processes, including vinyl cleavage, will be overlapped between two double bonds [25]. Additional experimental studies are underway to elucidate EIEIO cleavage between two double bonds.

Taking advantage of differential mobility spectrometry to analyze mixtures of cis/trans isomers Quantitative identification of a mixture of complex lipids that expected to contain both cis/trans isomers is straightforward when EIEIO is combined with a DMS or an LC, and the DMS or LC can separate the cis/trans isomers. Similar to a recent report [14], our DMS also separated cis- and trans-TG isomers (Figure 4(a)) and PC isomers (Figure 5(a)). However, further structural characterization or identification may be required to apply to natural samples. Comparison to authentic standards may be a possible way similar to fatty acid identification using LC-MS, but it may not be promising for complex lipids because DMS may not separate complex cis/trans isomers with various regioisomerism and double bond positions. In addition, full confirmation of these structures may be hindered if authentic standards of all expected isomers are not available. We applied EIEIO to the DMS-separated mixed standard samples, TGs (Figure 4(b) and 4(c)) and POPC/PEPC (Figure 5(b) and 5(c)) at two COV values on the outer shoulders of these convoluted species. In Figure 4(c), EIEIO indicated the double bond was cis because the hydrogen lost species at m/z 737.6 was less intense than the radical peak at m/z 738.6 (i.e., the ∆I ratio was negative). In Figure 4(c), on the other hand, the double bond is trans because the H lost species is more intense than the radical peaks. In Figure 5(b) (COV = 1.75 V), the EIEIO result indicated the double bond was cis because the hydrogen lost species at m/z 646.5 was less


Page 13 of 31

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


ratio was negative). In Figure 5(b) (COV = intense than the radical peak at m/z 646.5 (i.e., the ∆I 2.35 V), on the other hand, the double bond is trans because the H lost species is more intense than the radical peaks. The difference between POPC and PEPC was slight, but it was significant as shown in Figure 5(d), which shows statistical errors of the measurements. Relative constituents of cis and trans isomers can be analyzed by deconvolution of overlapped DMS profile. Figure 5 confirms a possibility of cis/trans isomer identification in intact lipid molecules without standards in the case of PCs with two acyl chains, but we still have difficulty to identify full structures of TGs because various regioisomer possibilities by combinations of three acyl groups with different double bond positions and cis/trans isomerism overlapped on a narrow region of a DMS ionogram.

Analysis of mixtures of cis/trans isomers exclusively by an EIEIO approach for isomeric species that We surveyed another cis/trans identification possibility using ∆I the DMS could not separate. Mixtures of cis/trans PC isomers: PC 16:1(n-7,cis)/16:1(n-7,cis) and PC 16:1(n-7,trans)/16:1(n-7,trans) with various mixing ratio were analyzed by EIEIO. As similar to the mixed acyl PCs in Figure 3, energy dependence of the cis/trans mixtures laid at the n-6 between the pure trans line and the pure cis line. Figure 6 shows the variation of ∆I at two eKE values are bond depending on the mixing ratio of the trans-PC in the cis-PC, ∆I of the pure trans isomer is zero at eKE = 7.0 eV (blue line in Figure 6(d)) so that shown. ∆I at this eKE shows the amount of cis isomer in the mixture. At eKE experimental amount of ∆I of the cis isomer is zero (red line in Figure 6d) so that experimental amount of ∆I at =13.5 eV, ∆I this eKE shows the amount of the trans isomer.



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 31

We obtain the mixing ratio of cis and trans isomers using these calibration curves (Figure 6); however, this method requires authentic cis and trans isomer standards with the same molecular structure because the curves were not universal for any types of lipids. Our test on two types of acyl groups, 16:1 and 18:1 in both PCs and TGs showed difference in energy dependences (Figure SI-5), which gave different zero crossing eKE for cis isomers (12-16eV) value of pure trans isomers at this eKE range. In lower eKE side, ∆I of the pure and various ∆I values at this trans isomers was zero around 7eV, but the pure cis isomers showed various ∆I eKE. Although the eKE dependence is not universal, we can eliminate the requirement of authentic trans isomer standards, which are often more difficult to obtain than cis isomer of trans is ~zero at eKE = 7eV. For more precise standards. We used the evidence that ∆I of 16:1 and 18:1 in PCs and TGs discussion, linear fitting was applied on eKE dependence of ∆I was given by eKE = 6.7±0.5 eV for trans lipids. We approximated (Figure SI-5c), and zero of ∆I ≈ 0 at this electron energy. that any trans double bond in a mono unsaturated acyl group had ∆I value of the same constitutional isomer but the double Next, we experimentally obtained a ∆I bonds are all cis types. In the case of mono acyls in the lipid, such as lyso GPLs with a double = 0 at 7 eV, and ∆I of the mixture of cis/trans isomers lays between bond, trans isomer gives ∆I value. The interpolated ratio from the cis value shows zero and experimentally obtained cis- ∆I the contamination of the trans isomers in the mixture. In the cases of lipid molecules containing two acyl chains, such as most GPLs, SMs, and diacylglycerols, and where each acyl has a double = 0 at 7 eV. The ∆I value of a mixed bond at the same position, the pure trans isomer gives ∆I value of isomers with two isomer sample appears between zero and experimentally obtained ∆I cis acyl chains. The interpolated ratio from the cis value shows the contamination of the trans


Page 15 of 31

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


acyl isomers in the mixture. In another case of lipids with two acyl chains where one chain is of pure trans isomers has completely saturated, such as POPC and PEPC, we can assume that ∆I value of the cis isomer, which can be verified by the half value of the experimentally obtained ∆I of a mixed isomer samples should appear Figure 3 and the proposed reaction mechanism. The ∆I (cis)/2 and ∆I (cis). Triacyls are also analyzed using the same concept. Assumed between ∆I (trans) at eKE=7eV should be ∆I (cis), 2∆I (cis)/3, or ∆I (cis)/3 for three mono-unsaturated acyls, ∆I two mono-unsaturated acyls, and one mono-unsaturated acyl, respectively. However, while this method gives a total constituent of trans acyls in a complex lipids, we do not have a method to assign a trans acyl to a specific site (sn-1, sn-2, or sn-3) when the molecule has multiple monounsaturated acyl groups. Theoretically this method does not need DMS so that the ion transmission for EIEIO analysis will be higher than with the DMS approach. Although the EIEIO efficiency is low at values by extrapolation. For example, we measured eKE=7 eV, we can obtain more confident ∆I values at efficient eKE of 7, 8, 9, 10 and 11 eV for linear fitting, then ∆I at 7 eV was multiple ∆I calculated using linear fitting in the following biological sample analysis.

Analysis of biological samples Ruminant (bovine) samples from different organs or milk were compared to standards or nonruminant organs as the demonstration of the described method. To find biological samples with trans fatty acids, screening by DMS was applied on bovine liver extracts, bovine heart extract and MFGM. We surveyed PC 18:1/18:1, PC 16:0/18:1 in the liver, the heart and MFGM to compare to the standards. The cis and trans isomers were separated by DMS, but the possible trans level was lower than the detection limit (~1%) [10,13] (Figure SI-6 and SI-7). Next, we



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 31

compared SMs in MFGM, which should be rich in trans chains [7], and non-ruminant porcine brain extract, which can contain a minute amount of trans double bonds. The EIEIO spectrum of SMs in the porcine brain (Figure SI-8(a)), where this lipid class was separated from phospholipids by DMS [26], showed that the chains contain an even number of carbons predominantly, which suggested the FAs were of non-bacterial origin. This also implied that the lipids contained trans fatty acids little. Actually EIEIO spectra of a brain SM (m/z 813.7) and the synthetic cis standard, SM d18:1/24:1(n-9,cis), were identical (Figure SI-8(c)). Conversely, MFGM contained FAs with odd numbers of carbons with the same level of FAs with even numbers of carbons (Figure SI-8(b)). This suggests the lipids were of bacterial origin [3]. We tested three precursors with a mono-unsaturated chain, being at m/z 785.7, 799.7, and 813.7 and present as protonated molecules ([M+H]+). The EIEIO analysis of constitutional isomerism was applied, and SMs in the porcine brain were substantially pure as SM d18:1/22:1(n-9), SM d18:1/23:1(n-9), SM d18:1/24:1(n-9) (Figure 7, Figure SI-9). SM in cow milk with m/z of 813.7 was rather pure with SM d18:1/24:1(n-9) (Figure 7). This species should contain 23% of SM d18:1/24:1(n-9,trans), which was confirmed by an earlier reported FA composition in MFGMSMs [7]. An SM of m/z 799.7 was a mixture of SM d18:1/23:1(n-9) and SM d16:1/25:1 (Figure SI-9(b)), and SM with m/z of 785.7 was a crude mixture of many chain lengths but contained SM d18:1/22:1(n-9) little (figure SI-9(a)). We focused on SM d18:1/24:1(n-9) to demonstrate trans SM analysis. Interestingly, DMS on SM d18:1/24:1(n-9) did not show any difference between the porcine brain and cow milk (Figure 7a) although DMS was operated using higher DR flow rates to promote higher resolution DMS separations. Next, we applied EIEIO on both SMs and the cis at eKE = 7 eV standard, SM d18:1/24:1(n-9,cis), at eKE of 7, 8, 9, 10, and 11 eV to obtain ∆I


Page 17 of 31

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


(Figure 7c). DMS was set at low resolution (DR = 0) to obtain higher ion transmission. EIEIO showed a significant difference spectra were accumulated for 24 min at each energy point. ∆I for the porcine brain and between the porcine brain and the cow milk. The linear fitting gives ∆I (trans) can be estimated as the half of porcine’s ∆I value, the the cow milk. Because ∆I constituents of SM d18:1/24:1(n-9,trans) in SM d18:1/24:1(n-9) was calculated as 36% (average of two measurements), which was good agreement to the reported C24:1(n-9,trans) in SMs in cow milk[7]. The same procedure was applied to other two SM species though they were a mixture of multiple constitutional isomers but can be estimated the chains have a double bond at n-9 position (Figure SI-9). The concentration of trans-SM showed ~100% nominally in these cases as expected [7]. We suppose that cis/trans isomers in mixtures of many constitutional isomers should be analyzed by combination with a future DMS that has much higher resolution and transmission.

Sensitivity of cis/trans identification EIEIO spectra display almost full information of structures of complex lipids in a spectrum. However, the yield of cis/trans isomer diagnostic peaks was quite low compared to the initial precursor intensity. The typical intensity of chain fragments is 0.1% to the peak intensity with our present instrument. If we target 1% accuracy of cis/trans difference, at least ~10,000 fragment ions are required by Poisson statistics. This means required initial precursor ions should be 10,000 x 1,000 (1 / 0.1%) = 10 million precursor ions. When we scan eKE at several points, 100 million precursor ions are required. To accumulate this level of signals, we may need several hours because the total ion intensity during the measurement of Figure 7 was 4000 [cps, count



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 31

per second] and this species was not abundant in the MS spectrum. Our EIEIO method is unique that can apply to complex lipids for structural lipid characterization, and we foresee that improvements in sensitivity of the mass analyzer and improvement of EIEIO reaction speed would make this overall workflow a more versatile tool for “complete” structural lipidomics. We have not observed EIEIO fragments of negatively charged lipid ions so far though the system produced negative ion ECD (niECD) fragments [27], which showed that electrons with energy of several eV reached to the trapped negative ions. The surveyed electron energy was 0 - 16eV and maximum reaction duration was a few 100 ms. While negative EIEIO may be observed in future experiments using higher electron energy with longer reaction durations, past examples point to fragmentation patterns that mirror collision-induced dissociation (CID) processes [28], which are not very useful in complete structural lipidomics studies.

Conclusion EIEIO using electron kinetic energy of 5-16 eV distinguished cis and trans isomers of PCs, TGs and SMs, and hydrogen lost non-radical product at a carbon-carbon single bond neighboring to a double bond can be used for diagnostics. Mixed cis/trans isomers were identified by the EIEIO method after isomer separation by DMS. In the cases that DMS does not separate these isomers, the cis/trans isomers intensity of the diagnostic fragments was used for obtaining trans contamination ratio in the mixed samples. These approaches reduce the requirement of authentic standards. In addition, this methodology demonstrated identification of cis/trans isomers in natural samples extracted from ruminants.

Acknowledgement


Page 19 of 31

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


The authors would like to thank Drs. James Hager and Bradley Schneider of SCIEX for helpful discussions and feedback. We also would like to thank one of our manuscript’s peer reviewers for their very productive suggestions. For Research Use Only. Not for use in diagnostic procedures. The trademarks mentioned herein are the property of AB Sciex Pte. Ltd. or their respective owners. AB SCIEXTM is being used under license.

Supporting information available Additional supporting information is available free of charge via the Internet.



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

References

1. Willett, W.C.; Stampfer, M.J.; Manson, J.E.; Colditz, G.A.; Speizer, F.E.; Rosner, B.A.; Hennkeens, C.H.; Sampson, L.A. Lancet 1993, 341, 581-585.

2. Cook, H.W. Biochimica et Biophysica Acta (BBA) - Lipids and Lipid Metabolism 1978, 531, 245–256. 3. Chatgilialoglu,C.; Ferreri, C.; Melchiorre, M.; Torreggiani, A. Chem. Rev. 2014, 114, 255−284. 4. Block, R.C.; Harris, W.S.; Reid, K.J.; Spertus, J.A. Am. Heart J. 2008, 156, 1117-1123. 5. Lemaitre, R. N.; King, I. B.; Raghunathan, T. E.; Pearce, R. M.; Weinmann, S.; Knopp, R. H.; Copass, M. K.; Cobb, L. A.; Siscovick D. S. Circulation. 2002;105, 697-701. 6. Puca, A.A.; Andrew, T.; Novelli, V.; Anselmi, C.V.; Somalvico, F.; Cirillo, N.A.; Chatgilialoglu,C.; Ferreri, C. Rejuvenation research 2008, 11, 63-72. 7. Morrison W.R.; Hay, J.D. Biochim. Biophys. Acta 1970, 202, 460-467. 8. Donato, P.; Cacciola, F.; Cichello, F.; Russo, M.; Dugo, P.; Mondello, L. Journal of Chromatography A, 2011, 1218, 6476– 6482. 9. Shingfield, K. J.; Chilliard, Y.; Toivonen, V.; Kairenius, P.; Givens, D. I. Adv. Exp. Med. Biol. 2008, 606, 3-65. 10. Ferreri, C.; Panagiotaki, M.; Chatgilialoglu, C. Mol Biotechnol 2007, 37, 19–25.


Page 20 of 31

Page 21 of 31

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. Bhandari, K.; Chaurasia, S.P.; Dalai, A.K. International Journal of Applied and Natural Sciences 2013, 2, 1-12. 12. Bird, S. S.; Marur, V. R.; Stavrovskaya, I. G.; Kristal, B. S. Anal. Chem. 2012, 84, 5509−5517. 13. Groessl, M.; Graf, S.; Knochenmuss, R. Analyst, 2015,140, 6904-6911. 14. Wojcik, R.; Webb, I. K.; Deng, L.; Garimella, S. V. B.; Prost, S. A.; Ibrahim Y. M.; Baker, E.S.; Smith, R. D. Int. J. Mol. Sci. 2017, 18, 183-195. 15. Voinov, V.G.; Claeys, M.; Int. J. Mass Spectrom. 2001, 205, 57–64. 16. Hejazi, L.; Ebrahimi, D.; Guilhaus, M.; Hibbert, D. B. J. Am. Soc. Mass Spectrom. 2009, 20, 1272–1280. 17: Ji, H.; Voinov, V. G.; Deinzer, M. L.; Barofsky, D. F. Anal. Chem. 2007, 79, 1519-1522 18. Jensen, N.; Lam. K.; Cody, R. D.; Tamura, J. RCM 1990, 4, 239–241. 19. Huong, T.; Pham, H. T.; Prendergast, M. B.; Dunstan, C. W.; Trevitt, A. J.; Mitchell, T. W.; Julian, R. R.; Blanksby, S. J.; Int. J. Mass Spectrom. 2015, 390, 170–177. 20. 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, 1989-1999.

21. Cody, R. B.; Freiser, B. S. Anal. Chem. 1979, 51, 547-551.

22. Campbell, J.L.; Baba, T. Anal. Chem. 2015, 87, 5837–5845.



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 31

23. Lintonen, T.; Baker, P.R.S.; Suoniemi, M.; Ubhi, B.; Koistinen, K.; Duchoslav, E.; Campbell, J.L.; Ekroos, K. Anal. Chem. 2014, 86, 9662-9669. 24. Baba, T.; Campbell, J.L.; Le Blanc, J.C.Y.; Baker, P.R.S. Journal of Lipid Research 2016, 57, 2015-2027. 25. Yang, K.; Dilthey, B.G.; Gross, R. W. Anal. Chem. 2013, 85, 9742−9750. 26. Baker, P.R.S.; Armando, A.M.; Campbell, J.L.; Quehenberger, O.; Dennis, E.A. J. Lipid Res., 2014, 55, 2432-2442. 27. Yoo, H.J.; Wang, N.; Zhuang, S.; Song, H.; Håkansson, K.; J. Am. Chem. Soc. 2011, 133, 16790–16793. 28. Yoo, H.J; Liu, H.; Håkansson, K.; Anal. Chem. 2007, 79, 7858-7866.


Page 23 of 31

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


Figure and figure captions

Figure 1. Double bond region of EIEIO spectra of protonated PC 18:1(n-9,cis)/18:1(n-9,cis) (DOPC) and PC 18:1(n-9,trans)/18:1(n-9,trans) (DEPC). Cis/trans differences appeared on hydrogen loss species at the n-6 site and hydrogen gain species at the n-8 site.



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 24 of 31

Figure 2. Electron energy dependence of the hydrogen loss product and the radical product at the n-6 site of (a) PC 16:1(n-7,trans)/16:1(n-7,trans) and (b) PC 16:1(n-7,cis)/16:1(n-7,cis). (c) Sum of the radical and hydrogen loss products. (d) Energy dependence of normalized intensity difference of PC 16:1(n-7,trans)/16:1(n-7,trans) and PC 16:1(n-7,cis)/16:1(n-7,cis). The precursors were in protonated form: [M+H]+.


Page 25 of 31

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


Figure 3. Electron kinetic energy dependence of normalized intensity difference of PC 16:0/18:1(n-9,trans) (PEPC), PC 16:0/18:1(n-9,cis) (POPC), PC 18:1(n-9,trans)/18:1(n-9,trans) (PEPC), and PC 18:1(n-9,cis)/18:1(n-9,cis) (DOPC). The precursors were in protonated form: [M+H]+.



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 26 of 31

Figure 4. DMS separation and identification of trans and cis TG 16:1(n-7)/16:1(n-7)/16:1(n-7). Precursor ions were in the sodiated form: [M+Na]+. (a) DMS profile of a mixture of cis/trans-TG in 1: 1 in weight. DR setting was set at 42 (high flow). (b) DMS profile of pure cis-TG (blue line) and pure trans-TG (pink line). (c) EIEIO spectrum of mixed TG at COV=3.80[V]. The lower intensity of the H loss species (m/z 737.6) than the radical species (m/z 738.6) indicated the precursor had cis isomers. (d) EIEIO spectrum of mixed TG at COV=3.80[V]. Higher intensity of the H loss species than radical species indicated the precursor had trans isomers in the acyl chains. Cis/trans identification by EIEIO was consistent with the DMS separation results of the pure standards, Figure 4(b). Accumulation time of spectrum (c) and (d) was 20 min each.


Page 27 of 31

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


Figure 5. POPC/PEPC mixture separation and identification by DMS approach. The precursors were in the protonated form: [M+H]+. (a) DMS ionogram of pure POPC, pure PEPC, and mixed POPC/PEPC. DR setting was 41 (high flow). (b) EIEIO spectra of mixed sample. The spectra were obtained at the two shoulders. (c) Cis and trans isomerism were determined by normalized intensity difference, ∆I, by EIEIO measurement. Accumulation time to obtain (b) and (c) was 20 min.



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 28 of 31

Figure 6. Normalized intensity difference depending on cis/trans solution mixing ratio at two characteristic eKE, i.e., trans contribution is zero at eKE=7.0V and cis contribution is zero at eKE= 13.5 eV (see Figure1(d)). The samples are PC 16:1(n-7,trans)/16:1(n-7,trans) and PC 16:1(n-7:cis)/16:1(n-7:cis) in protonated form: [M+H]+.


Page 29 of 31

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


Figure 7. Cis/trans separation of SM d18:1/24:1(n-9) in cow milk SM using EIEIO approach. (a) DMS ionogram of cow milk SM and porcine brain SM with m/z of 813.68. DR setting was 34(high flow). (b) Comparison of EIEIO spectra of milk SM and brain SM. The inset shows cis/trans diagnostic peaks (H loss). (c) eKE dependence of normalized intensity difference of milk SM, porcine brain SM, and synthetic standard. Trans contamination in this SM species in cow milk was 36% by this analysis.



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

Scheme 1. Proposed mechanism of EIEIO on acyl chains with double bonds.


Page 30 of 31

Page 31 of 31

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


TOC only


Distinguishing Cis and Trans Isomers in Intact Complex lipids using

Recommend Documents