Uncovering Structural Diversity of Unsaturated Fatty Acyls in Cholesteryl Esters via Photochemical Reaction and Tandem Mass Spectrometry
- 335 Downloads
Mass spectrometry analysis of cholesteryl esters (CEs) faces several challenges, with one of them being the determination of the carbon–carbon double bond (C=C) locations within unsaturated fatty acyl chains. Paternὸ-Büchi (PB) reaction, a photochemical reaction based on the addition of acetone to C=C, is capable of C=C location determination when coupled with tandem mass spectrometry (MS/MS). In this study, the PB reaction conditions were tailored for CEs and subsequent nanoelectrospray ionization (nanoESI). A solvent system containing acetone/methanol/dichloromethane/water (40/30/20/10, volume ratios) and 100 μM LiOH was determined to be optimal, resulting in reasonable PB reaction yield (~30%) and good ionization efficiency (forming lithium adduct of CEs). Collision-induced dissociation (CID) of the PB reaction products produced characteristic fragment ions of CE together with those modified by the PB reactions, such as lithiated fatty acyl ([FA + Li]+) and its PB product ([FA – PB + Li]+). MS3 CID of [FA – PB + Li]+ led to abundant C=C diagnostic ion formation, which was used for C=C location determination and isomer quantitation. A PB-MS3 CID approach was developed and applied for CE analysis from human plasma. A series of unsaturated CEs was identified with specific C=C locations within fatty acyl chains. Absolute quantitation for each CE species was achieved including coexisting C=C location isomers, such as Δ9 and Δ11 isomers of CE 18:1 and ω-6 and ω-3 isomers of CE 18:3. These results show that PB-MS/MS is useful in uncovering structural diversity of CEs due to unsaturation in fatty acyls, which is often undetected from current lipid analysis approach.
KeywordsC=C determination Cholesteryl esters Paternὸ-Büchi reaction Lipidomics Tandem mass spectrometry
Mass spectrometry (MS)-based lipid analysis has been established as a powerful tool for the identification and quantitation of a variety of lipid classes from biological matrixes, offering a combined advantage of high sensitivity, speed, and detailed structural information . Many different MS techniques have been applied to CE analysis with both successes and limitations. Conventionally, the fatty acyl composition of CE is analyzed by gas chromatography (GC)-MS via electron ionization (EI) or chemical ionization (CI). Because of limited volatility, CEs need to be hydrolyzed and the methylated fatty acyls are subjected to GC-MS, the process of which requires relatively long analysis time (in hours) . The development of electrospray ionization (ESI)  has greatly expanded the spectrum of lipids that can be analyzed by MS. Han and Gross were amongst the first to recognize the potential of ESI-MS for global lipid analysis from biological systems . Soon thereafter, two techniques based on ESI-MS were established as the main methods for current lipidomics studies: shotgun lipid analysis of crude lipid extract [19, 20] and ESI-MS coupled with liquid chromatography (LC) separations [21, 22].
Given the nonpolar nature of CE, adduct ion formation is employed to enhance ionization efficiency of CE in ESI, with NH4 +, Li+, and Na+ being frequently used as adduct ions for positive ion mode. Tandem mass spectrometry (MS/MS) via low energy collision-induced dissociation (CID) of CE adduct ions has been developed for the identification and quantitation of CEs from mixtures . For instance, CID of the ammonium adduct of CE generates a dominant fragment ion at m/z 369 (cholestene cation), allowing the use of product ion scan (PIS) for quantitation of CEs from lipid mixtures . Collisional activation of lithiated or sodiated CE ions results in an abundant neutral loss of cholesterene (368 Da) and forms Li+ or Na+ adduct of fatty acyl ions [25, 26]. This fragmentation channel allows fatty acyl composition determination, viz. the number of carbons and degrees of unsaturation, while neutral loss scan (NLS) of 368 Da renders sensitive detection and quantitation of CEs. The above ESI-MS/MS techniques, however, have a major limitation of not being able to provide carbon–carbon double bond (C=C) location information within fatty acyls of CEs. Given that unsaturation in lipids play critical roles in their biochemical and biophysical properties, and many unsaturated fatty acids have multiple C=C location isomers fulfilling distinct biological roles, it is of great importance to characterize unsaturated CEs at the structural level of C=C locations [27, 28, 29].
The need to determine a specific lipid structure from coexisting isomers and isobars has drawn increasing awareness in the field of lipidomics [30, 31]. Some notable MS methods that are suitable for the analysis of different classes of lipids and provide C=C location information include charge-remote fragmentation induced by high energy CID , ozone-induced dissociation (OzID) [33, 34, 35], radical directed dissociation (RDD) [36, 37], helium metastable atom-activated dissociation (He-MAD) [38, 39], and electron impact excitation of ions from organics (EIEIO) [40, 41]. Recently, our group has demonstrated that the Paternὸ-Büchi (PB) reaction , a photochemical reaction via the addition of electronically excited acetone to a C=C, can be applied to C=C location determination and location isomer quantitation for fatty acids and phospholipids when coupled with ESI-MS/MS [43, 44, 45, 46]. In this study, we aim to expand the scope of the PB reaction to neutral lipids, such as CEs, which tend to have high composition of unsaturated fatty acyls. The factors that are key to this development have been identified, including proper solvent system that allows both good PB reaction yield and ionization for CE using ESI, and MS/MS methods that can allow confident structural determination and C=C location isomer quantitation. Commercially available CE standards were used for method development and optimization. The analytical capability of PB-MS/MS was further tested for analyzing CEs in human plasma.
Shorthand notation for CEs is taken from LIPID MAPS . For example, CE 18:1(9Z) detonates a cholesteryl ester containing an 18-carbon fatty acyl chain. The numeral “1” after the carbon number refers to the degree of unsaturation of the fatty acyl. The location(s) of the C=C bond(s) is defined in ∆x nomenclature by counting from the alpha carbon of the fatty acyl and is indicated with the number(s) in parentheses. For polyunsaturated fatty acyl chains, the ω-nomenclature is also used for simplicity in which the location of the first C=C is counted from the methyl end of the fatty acyl chain. Cis- or trans-configuration of C=C bond is denoted with Z and E nomenclature, respectively. For CE analysis from biological extract, the Z/E stereo-configurations are not assigned.
CE standards, including CE 18:1 (9Z), CE 18: 1 (11Z), CE 18:2 (9Z, 12Z), CE 18:3 (9Z, 12Z, 15Z), CE 18:3 (6Z, 9Z, 12Z), CE 20:4 (5Z, 8Z, 11Z, 14Z), and CE 22:5 (7Z, 10Z, 13Z, 16Z, 19Z), and lithium hydroxide (LiOH) were purchased from Sigma-Aldrich (St. Louis, MO USA). Pooled normal human plasma with anticoagulant lithium heparin was obtained from Innovative Research, Inc. (Novi, MI, USA). Organic solvents were all LC grade and ultrapure water was obtained from a purification system at 0.03 μS cm. CEs were extracted from 20 μL of plasma following methyl-tert-butyl ether (MTBE) method with an extraction efficiency of 71.0% ± 0.6% obtained (from three replicates) .
PB-MS/MS for CE Analysis
All samples were analyzed in positive ion mode using a 4000 QTRAP mass spectrometer (Sciex, Toronto, ON, CA) equipped with a home-built nanoESI source and a low pressure mercury lamp (primary emission at 254 nm; BHK Inc., Ontario, CA). CE standards (0.1–10 μM) were dissolved in acetone/methanol/dichloromethane/water = 40/30/20/10 (volume ratios) containing 100 μM LiOH. The PB reaction was facilitated by UV irradiation of a pulled borosilicate glass capillary nanoESI tip, which contained a solution of CE. A scheme of the reaction setup is shown in Supporting Information (SI), Figure S1. Multiple modes of MS/MS, including neutral loss scan (NLS), MS2 via beam-type CID, and MS3 via sequential beam-type CID and on-resonance ion trap CID were employed in CE analysis. For NLS, collision energy (CE) was optimized at 40 eV. In beam-type CID, the precursor ions were isolated by Q1 and accelerated to q2 for collisional activation with collision energy defined by the DC potential difference between Q0 and q2, which was typically within the range 35–40 V. Ion trap CID was carried out in Q3 linear ion trap, where a dipolar excitation was used for on-resonance collisional activation. The activation amplitudes were within the range 40–50 mV for 200 ms. The characteristic parameters of the mass spectrometer during this study were set as follows: spray voltage, 1200–1500 V; curtain gas, 10 psi; and declustering potential, 180 V. Data acquisition, processing, and instrument control were performed using Analyst 1.5 software.
Results and Discussions
Online Coupling of the PB Reaction and NanoESI-MS for CE Analysis
A simple binary mixture of acetone/water solvent system has been used for the analysis of fatty acids and phospholipids with acetone serving as both the PB reaction reagent and co-solvent in subsequent online ESI-MS or nanoESI-MS [43, 45]. This condition works well for both PB reaction and ionization since the above two classes of lipids have reasonable solubility in such polar solvent system. It is worth noting that organic solvent other than acetone is minimized in the solvent system to reduce possible side reactions involving radical intermediates formed from Norrish Type I cleavage of acetone . CEs have very limited solubility in water and they are typically dissolved in methanol/chloroform (4:1 volume ratio) for ESI-MS analysis [23, 25]. Consequently, acetone/water solvent system developed for polar lipids cannot be directly applied for the PB reaction and nanoESI-MS of CEs. Our first task was targeted to developing a solvent system that could provide good PB reaction of CE. Pure acetone was evaluated as the solvent for a model CE compound, CE 18:1(9Z) (5 μM), for various lengths of reaction time (10 s to 3 min) before reconstituting in methanol/chloroform (4:1 volume ratio, 100 μM LiOH) for nanoESI. The PB reaction yield was found low and was accompanied by a large extent of side reactions. It was hypothesized that because of the nonpolar nature of CE molecules, the fatty acyls might prefer strong interactions with their own cholesteryl moieties and therefore have reduced interactions with the more polar acetone solvent molecules. We suspected that the addition of less polar organic solvent into acetone might be necessary to enhance effective interaction between acetone and CE fatty acyls. A series of organic solvents (10%–50% volume ratio relative to acetone) including chloroform (CHCl3), dichloromethane (CH2Cl2), and hexane were examined. Dichloromethane was found as the best solvent to provide reasonable reaction yield and limited degree of side reactions among all organic solvents tested. An example of reaction spectrum involving CHCl3 as a co-solvent is shown in SI, Figure S2, which suffers from a high degree of CE ion signal loss and limited formation of the PB products.
The binary mixture of acetone/dichloromethane, however, was not miscible with even 10% of aqueous solution of salts, i.e., LiOH, the addition of which was necessary to enhance CE adduct ion formation and detection in ESI. Upon various tests, we found that methanol could be added as a co-solvent leading to a homogenous solution and ensuring stable ionization during nanoESI. The commonly used adduct ions for ESI analysis of CE, including lithium (Li+), ammonium (NH4 +), and sodium (Na+) were evaluated for their compatibility with PB-MS/MS. Li+ adduct of CE had combined advantages of relatively high ionization efficiency (relative to Na+ adduct) and forming abundant fatty acyl fragment ion (relative to NH4 + adduct) under CID, and was used for further studies. After optimizations, a solvent system composed of acetone/methanol/dichloromethane/water (40/30/20/10) and 100 μM LiOH was identified as the best solvent condition for conducting online PB-nanoESI-MS for CE analysis.
Figure 2c and d compare MS2 CID of intact CE 18:1(9Z) lithium adduct ([CE + Li]+) versus its PB reaction product ([CE – PB + Li]+). Intact CE ions (m/z 657.6) produce two complementary fragment ions: lithiated fatty acyl ion ([FA + Li]+) at m/z 289.4 and cholestene cation ([Chol]+) at m/z 369.4. These two characteristic fragment ions are also detected, however, with significantly reduced relative intensities in MS2 CID of the PB reaction products (Figure 2d). Instead, ions at m/z 347.4 corresponding to acetone addition to the fatty acyl chain ([FA – PB + Li]+) is the most abundant fragment peak. Interestingly, the fragment containing acetone addition to cholesterol ring (m/z 433.4) is detected at 20% intensity relative to [FA – PB + Li]+ peak. It is worth noting that CE 18:1(9Z) consists of one C=C bond within the cholesteryl ring and one in its fatty acyl chain. The above results suggest that the PB reaction is preferred at a C=C with less steric hindrance. Two pairs of fragments related to the location of C=C in fatty acyl are also detected: ions at m/z 547.5 and 573.5 and ions at m/z 179.3 and 205.3 (structures shown in the inset of Figure 2). The characteristic 26 Da mass difference within each pair of ions clearly suggests that they each result from rupture of the four-membered oxatane rings of the two PB products formed. The former pair consists of intact cholesteryl moiety, whereas the latter pair are likely produced from sequential fragmentation of [FA – PB + Li]+ ions. Indeed, MS3 CID of [FA – PB + Li]+ (m/z 347.4, Figure 2e) leads to abundant formation of ions at m/z 179.3 and 205.3. Although the location of C=C in fatty acyl chain can be readily deduced from MS2 CID as shown in Figure 2d, these C=C diagnostics are much more dominant in MS3 CID, which is later found to be useful in mixture analysis.
Figure 3c shows MS2 beam-type CID of [CE – PB + Li]+ ions at m/z 737.4, the single acetone addition PB product of CE 20:4. [FA + 58 + Li]+ (m/z 369.3) is the most abundant fragment peak, consistent with that observed from PB-MS/MS of CE 18:1. Due to sequential fragmentation of [FA –PB + Li]+, C=C diagnostic ions corresponding to the four C=C bonds in the fatty acyl chain are also present in the MS2 CID spectrum, however, with relatively low ion intensities. These C=C diagnostic ions (m/z 123/149, 163/189, 203/229, 243/269) are more prominent in MS3 ion trap CID (Figure 3d), leading to unambiguous assignment of C=C bond locations at Δ5, 8, 11, and 14, respectively (fragmentation map shown in the inset of Figure 3).
PB-MS3 CID for CE C=C Location Isomer Analysis
Analysis CE Extract from Human Plasma
CEs are relatively abundant species in human plasma and quantitative analyses of CE species in human plasma have been reported using either shotgun or LC-MS methods; the C=C double bond location isomer identification and relative composition have not been determined . We are interested in providing C=C location information for unsaturated CE species, especially detecting coexisting C=C location isomers in human plasma by applying PB-MS3 CID approach (Scheme 1) in the context of shotgun lipid analysis. CEs were extracted from human plasma (20 μL) based on the MTBE method and an extraction efficiency of 71.0% ± 0.6% was achieved, comparable to literature report . The extract was dried and reconstituted into 1 mL acetone/methanol/dichloromethane/water (40/30/20/10) containing 100 μM LiOH for nanoESI-MS analysis.
Quantitative Analysis of Major Unsaturated CE Species in Human Plasma
[M + Li]+
CE 16:1 (9)
98 ± 13
CE 18:1 (total)
414 ± 48
CE 18:1 (9)
382 ± 44
CE 18:1 (11)
32 ± 4
CE 18:2 (9, 12)
(1.30 ± 0.11) × 103
CE 18:3 (total)
24.0 ± 1.6
CE 18:3 (6, 9, 12)/ω-6
9.0 ± 0.6
CE 18:3 (9, 12, 15)/ω-3
15.0 ± 1.0
CE 20:4 (5, 8, 11, 14)
127 ± 4
Free FAs of the same human plasma sample were analyzed by PB-MS/MS and reported previously . Comparing to CE data collected from this study, we notice that the C=C locations within fatty acyls are conserved no matter if they are in free FAs or in more complicated lipid molecules, such as CEs. For instance, CE 18:2 is a pure form with C=C located at Δ9 and Δ12, and the same is detected for FA 18:2. CE 18:1 has Δ9 and Δ11, two location isomers, and their molar ratio (12/1) is very close to that of free FA 18:1 (Δ9/Δ11 = 11/1). Similarly, the same ω-6 and ω-3 isomers of FA 18:3 are detected for CE 18:3. The conservation of fatty acyl identities in different classes of lipids reflects them being common building blocks in lipid biosynthesis. More importantly, this type of information could be used as a guide for the identification of unsaturated lipids of different classes.
In this study, an approach based on PB-MS3 CID has been developed for the analysis of CEs containing unsaturated fatty acyl chains from mixtures. An important step of the development is to determine a proper solvent system that allows for both good PB reaction and ionization via nanoESI. Since CE is nonpolar, adding a nonpolar co-solvent, such as CH2Cl2 (20%), to acetone (40%) is necessary for achieving good yields in the PB reactions, while adding another relatively polar co-solvent, MeOH (30%) helps in forming homogenous solution and maintaining stable nanoESI upon the addition of 10% of LiOH aqueous solution. The PB products of CEs in the form of lithium adduct ions ([CE – PB + Li]+) can be detected abundantly after reaction. Collisional activation of [CE – PB + Li]+ results in prominent formation of PB modified fatty acyl ions ([FA – PB + Li]+), while MS3 CID of [FA – PB + Li]+ leads to C=C diagnostic ion formation, which are employed in the determination of C=C location and of C=C location isomer composition. The analytical utility of PB-MS3 CID was demonstrated with CE analysis from human plasma. A series of CEs containing unsaturated fatty acyl chains was identified and quantified at structural level of the C=C locations, unveiling several coexisting C=C location isomers of CEs. Such detailed information is generally not obtainable using conventional lipid analysis methods. Overall, the PB-MS3 CID approach is simple, fast, and compatible with shotgun lipid analysis workflow. It also offers a distinct advantage of high molecular specificity due to coupling C=C site specific derivatization and MS3. These advantages would be especially attractive for discovering composition changes of CE C=C location isomers related to diseases or physiological changes. Undesirable aspects associated with the current method include low PB reaction yield and unwanted side reactions due to the use of MeOH as a co-solvent, both of which limit sensitivity of the PB-MS3 CID approach. Future studies will focus on developing an improved reaction system for CE and other non-polar lipids.
Financial support from NSF CHE-1308114 and NIH R01GM118484 is greatly appreciated. X.Y. acknowledges ASMS research award for supporting the research on radical ion chemistry.
- 2.Brown, M.S., Ho, Y.K., Goldstein, J.L.: The cholesteryl ester cycle in macrophage foam cells. Continual hydrolysis and re-esterification of cytoplasmic cholesteryl esters. J. Biol. Chem. 255, 9344–9352 (1980)Google Scholar
- 4.Suckling, K.E., Stange, E.F.: Role of Acyl-CoA: cholesterol acyltransferase in cellular cholesterol metabolism. J. Lipid Res. 26, 647–671 (1985)Google Scholar
- 6.Francone, O.L., Gurakar, A., Fielding, C.: Distribution and functions of lecithin: cholesterol acyltransferase and cholesteryl ester transfer protein in plasma lipoproteins. Evidence for a functional unit containing these activities together with apolipoproteins a-i and d that catalyzes the esterification and transfer of cell-derived cholesterol. J. Biol. Chem. 264, 7066–7072 (1989)Google Scholar
- 7.Quehenberger, O., Armando, A.M., Brown, A.H., Milne, S.B., Myers, D.S., Merrill, A.H., Bandyopadhyay, S., Jones, K.N., Kelly, S., Shaner, R.L., Sullards, C.M., Wang, E., Murphy, R.C., Barkley, R.M., Leiker, T.J., Raetz, C.R.H., Guan, Z., Laird, G.M., Six, D.A., Russell, D.W., McDonald, J.G., Subramaniam, S., Fahy, E., Dennis, E.A.: Lipidomics reveals a remarkable diversity of lipids in human plasma. J. Lipid Res. 51, 3299–3305 (2010)CrossRefGoogle Scholar
- 10.Crowe, F.L., Skeaff, C.M., Green, T.J., Gray, A.R.: Serum fatty acids as biomarkers of fat intake predict serum cholesterol concentrations in a population-based survey of New Zealand adolescents and adults. Am. J. Clin. Nutr. 83, 887–894 (2006)Google Scholar
- 12.Yue, S., Li, J., Lee, S.-Y., Lee, H.J., Shao, T., Song, B., Cheng, L., Masterson, T.A., Liu, X., Ratliff, T.L., Cheng, J.-X.: Cholesteryl ester accumulation induced by PTEN loss and pi3k/akt activation underlies human prostate cancer aggressiveness. Cell Metab. 19, 393–406 (2014)CrossRefGoogle Scholar
- 14.German, J.B., Gillies, L.A., Smilowitz, J.T., Zivkovic, A.M., Watkins, S.M.: Lipidomics and lipid profiling in metabolomics. Curr. Opin. Lipid. 18, 66–71 (2007)Google Scholar
- 21.Bird, S.S., Marur, V.R., Sniatynski, M.J., Greenberg, H.K., Kristal, B.S.: Lipidomics profiling by high-resolution LC-MS and high-energy collisional dissociation fragmentation: focus on characterization of mitochondrial cardiolipins and monolysocardiolipins. Anal. Chem. 83, 940–949 (2010)CrossRefGoogle Scholar
- 22.Nygren, H., Seppänen-Laakso, T., Castillo, S., Hyötyläinen, T., Orešič, M.: Liquid chromatography-mass spectrometry (LC-MS)-based lipidomics for studies of body fluids and tissues. Springer (2011)Google Scholar
- 24.Murphy, R.C.: Tandem mass spectrometry of lipids: molecular analysis of complex lipids. Royal Soc. Chem. (2014)Google Scholar
- 29.Rustan, A.C., Drevon, C.A.: Fatty acids: Structures and properties. John Wiley and Sons (2005)Google Scholar
- 35.Pham, H.T., Maccarone, A.T., Thomas, M.C., Campbell, J.L., Mitchell, T.W., Blanksby, S.J.: Structural characterization of glycerophospholipids by combinations of ozone- and collision-induced dissociation mass spectrometry: the next step towards "top-down" lipidomics. Analyst 139, 204–214 (2014)CrossRefGoogle Scholar