High Resolution NMR Spectroscopy as a Structural and Analytical Tool for Unsaturated Lipids in Solution

Abstract

Mono- and polyunsaturated lipids are widely distributed in Nature, and are structurally and functionally a diverse class of molecules with a variety of physicochemical, biological, medicinal and nutritional properties. High resolution NMR spectroscopic techniques including 1H-, 13C- and 31P-NMR have been successfully employed as a structural and analytical tool for unsaturated lipids. The objective of this review article is to provide: (i) an overview of the critical 1H-, 13C- and 31P-NMR parameters for structural and analytical investigations; (ii) an overview of various 1D and 2D NMR techniques that have been used for resonance assignments; (iii) selected analytical and structural studies with emphasis in the identification of major and minor unsaturated fatty acids in complex lipid extracts without the need for the isolation of the individual components; (iv) selected investigations of oxidation products of lipids; (v) applications in the emerging field of lipidomics; (vi) studies of protein-lipid interactions at a molecular level; (vii) practical considerations and (viii) an overview of future developments in the field.

Keywords: NMR; chemical shifts; coupling constants; polyunsaturated fatty acids (PUFAs); unsaturated lipids.

Figure 1

Representative structures of various classes of lipids. Adopted, with permission, from [14]. Copyright 2012, Wiley VCH.

Figure 2

Stereospecific numbering of sn of triacylglycerols. Adopted, with permission, from [13]. Copyright 2002, Marcel Dekker, Inc.

Figure 3

Schematic overview of ω-3 and ω-6 PUFA-derived lipid mediators. Adopted, with permission, from [35]. Copyright 2014, Springer International Publishing, AG.

Figure 4

Metabolic pathways of endogenous conjugated linoleic acid synthesis in ruminants. Adopted, with permission, from [45]. Copyright 2014, Royal Society of Chemistry.

Figure 5

500 MHz ¹H-NMR spectrum of the lipid extract of a milk sample in CDCl₃ (298 K, acquisition time (AQ) = 4.0 s, relaxation delay (RD) = 3.6 s, total experimental = 1 h 5 min). The inset shows 2048× magnification of the spectrum in order to display some resonances of the minor species 9-cis, 11-trans, 18:2 CLA and caproleic acid.

Figure 6

Computer simulated ¹H-NMR signals for alkenyl and allylic protons in isolated cis and trans double bonds. Adopted, with permission, from [54]. Copyright 1975, North-Holland Publishing Company.

Figure 7

Separation of ω-3 from ω-6, ω-9 and saturated fatty acids on the basis of ¹H-NMR chemical shifts of –CH₃ groups. Free download from: http://www.ilps.org/index.php/Standardized_Methods.html.

Figure 8

¹³C-NMR simultaneous detection of free fatty acids (left side) and their alkyl esters (right side). Free download from: http://www.ilps.org/index.php/Standardized_Methods.html.

Figure 9

¹³C-NMR of the carbonyl region of docosahexaenoic acid-enriched egg yolk phosphatidylcholine. Adopted, with permission, from [21]. Copyright 2001, Wiley-VCH.

Figure 10

¹³C-NMR chemical shifts of the terminal methyl groups allow the identification and quantification of ω-3, ω-6, ω-7, ω-9 and saturated fatty acids of a salmon oil. Free download from: http://www.ilps.org/index.php/Standardized_Methods.html.

Figure 11

¹³C-NMR chemical shifts of the methylene C-2 carbons allow the identification of the EPA and DHA ω-3 fatty acids. Free download from: http://www.ilps.org/index.php/Standardized_Methods.html.

Scheme 1

Reaction of hydroxyl and carboxyl groups with the phosphorus reagent 2-chloro-4,4,5,5-tetramethyldioxaphospholane (I) and the phosphitylated product (II).

Figure 12

Signal-to-noise (SNR) increase by a factor of ~10 was obtained after suppression of multiple dominating lipid signals due to the increase of the receiver gain. Number of scans and total experimental time were kept constant, i.e., NS = 32 and 4 min, respectively. Adopted, with permission, from [73]. Copyright 2011, Elsevier Ltd.

Figure 13

(a) 500 MHz 1D ¹H-NMR spectrum of 20 mM solution of the (9-cis,11-trans) 18:2 CLA in CDCl₃ (T, 298 K; acquisition time, 4.3 s; relaxation delay, 5 s; number of scans, 256; and experimental time, ~25 min). (b–e) Selective 1D TOCSY spectra of the above solution using a mixing time of τ_m = 33 ms (b), 70 ms (c), 200 ms (d), and 400 ms (e). The asterisk denotes the selected H11 resonance that was excited. For panels (b–e), the magnetization transfer network is illustrated. Adopted, with permission, from [78]. Copyright 2015, American Chemical Society.

Figure 14

500 MHz 1D ¹H-NMR spectrum of caproleic acid in CDCl₃ (concentration: 5 mM, T: 298 K, acquisition time: 2.7 s, relaxation delay: 8 s, number of scans: 128, experimental time: ~23 min).

Figure 15

Selected region of 500 MHz 2D ¹H-¹H TOCSY spectrum of caproleic acid in CDCl₃ (solution conditions the same as in Figure 14). Mixing time τ_m = 100 ms, number of scans (ns) = 32, number of increments = 256, total experimental time = 5 h 19 min.

Figure 16

500 MHz 1D ¹H-NMR spectrum of (9-cis, 11-trans) 18:2 CLA in CDCl₃ (concentration: 10 mM, T: 298 K, acquisition time: 2.7 s, relaxation delay: 8 s, number of scans: 128, experimental time: ~23 min).

Figure 17

Selected region of 500 MHz 2D ¹H-¹H TOCSY NMR spectrum of (9-cis, 11-trans) 18:2 CLA (solution conditions the same as in Figure 16). Mixing time τ_m = 100 ms, number of scans (ns) = 16, number of increments = 256, total experimental time = 2 h 39 min.

Figure 18

850 MHz band selective ¹H-¹³C gHSQC spectrum of the aliphatic region of a fish oil supplement. Adopted, with permission, from [60]. Copyright 2015, Royal Society of Chemistry.

Figure 19

600 MHz HSQC-DEPT NMR spectrum of DAG olive oil in CDCl₃ solution, showing one bond correlations between the glycerol backbone protons and carbons; negative (blue color) signals for the CH₂ carbons and positive (red color) signals for the CH carbons. Adopted, with permission, from [59]. Copyright 2011, Springer AOCS.

Figure 20

600 MHz gHSQC-TOCSY spectrum of DAG oil in CDCl₃ solution, showing consecutive connectivities between carbons and protons along a common coupling pathway (for the notation system for ¹H resonances see the original article). Adopted, with permission, from [59]. Copyright 2011, Springer AOCS.

Figure 21

Selected region of 500 MHz 2D ¹H-¹³C HMBC spectrum of the (9-cis, 11-trans) 18:2 CLA isomer in CDCl₃ (solution conditions the same as in Figure 16). Total experimental time = 7 h 41 min.

Figure 22

Comparison of the 850 MHz ¹H-¹³C gHMBC spectrum of a fish oil supplement acquired over a 200 ppm spectral width (A) and the band selective constant time HMBC spectrum of the same sample acquired with a 2 ppm spectral width (B), in CDCl₃ solution. Adopted, with permission, from [60]. Copyright 2015, Royal Society of Chemistry.

Figure 22

Comparison of the 850 MHz ¹H-¹³C gHMBC spectrum of a fish oil supplement acquired over a 200 ppm spectral width (A) and the band selective constant time HMBC spectrum of the same sample acquired with a 2 ppm spectral width (B), in CDCl₃ solution. Adopted, with permission, from [60]. Copyright 2015, Royal Society of Chemistry.

Figure 23

¹H-NMR spectrum of a salmon oil which allows the simultaneous detection and quantification of ω-3, EPA and DHA. Free download from: http://www.ilps.org/index.php/Standardized_Methods.html.

Figure 24

¹H-NMR spectra of TAG and phospholipids from krill oil after preparative separation. The neutral lipids fraction contains the branched phytanic acid which can also be quantified. Free download from: http://www.ilps.org/index.php/Standardized_Methods.html.

Figure 25

¹H- and ¹³C-NMR chemical shifts, in ppm, of C9 to C12 carbons and their attached protons for the four geometric isomers of 18:2 CLA. Adopted, with permission, from [56]. Copyright 2014, Elsevier B.V.

Figure 26

500 MHz ¹H-¹H TOCSY NMR spectrum of the lipid fraction of a lyophilized milk sample in CDCl₃ with suppression of two major lipid proton resonances at 1.27 and 5.35 ppm. Experimental conditions: 298 K, 32 repetitions of 256 increments, total experimental time 5 h 5 min. Adopted, with permission, from [56]. Copyright 2014, Elsevier B.V.

Figure 27

Selected regions of 500 MHz ¹H-¹³C HSQC spectra of the lipid fraction of a lyophilized milk sample of Figure 26. (A) With ¹³C shielding spectral range from 0 to 160 ppm and (B) with reduced ¹³C spectral range from 112 to 142 ppm. Experimental conditions: 298 K, 40 repetitions of 256 increments, total experimental time 4 h 55 min and 5 h 49 min for (A,B), respectively. Adopted, with permission, from [56]. Copyright 2014, Elsevier B.V.

Figure 28

Spin chromatogram of the lipid fraction of a lyophilized cheese sample in CDCl₃. (a) 500 MHz ¹H-NMR spectrum of the lipid fraction of a lyophilized cheese sample in CDCl₃ (T, 298 K; number of scans, 256; acquisition time, 4.3 s; relaxation delay, 5 s; and total experiment time, ~25 min). The major lipid resonances are denoted (sn1, sn2, and sn3 indicate the stereospecific numbering of esterified glycerol). The inset shows 512× magnification of the spectrum to display resonances from the 18:2 CLA and other minor species. (b–d) 1D TOCSY spectra of panel with τ_m = 400 ms (number of scans, 256; and total experiment time, ~25 min). The asterisks denote the resonances that were excited by the use of a selective pulse. Adopted, with permission, from [78]. Copyright 2015, American Chemical Society.

Figure 29

Selected regions of (a) Figure 28a, (b) 1D TOCSY spectrum that demonstrates the spin system of the glycerol moiety in 2-MAG (in panels b1 and b2, the selective excitation pulse was set on the 1′,3′-CH₂OH (δ = 3.82 ppm) and 2′-CHOCOR (δ = 4.92 ppm) peaks in 2-MAG, respectively), and (c) 1D TOCSY spectrum of the spin system of the glycerol moiety in 1-MAG (in panels c1 and c2, the selective excitation pulse was set on the 3′-CH₂OH (δ = 3.67 ppm) and 2′-CH₂OH (δ = 3.92 ppm) peaks in 1-MAG, respectively). Adopted, with permission, from [78]. Copyright 2015, American Chemical Society.

Figure 30

(a) Selected region of the 1D ¹H-NMR spectrum of the solution of Figure 28a. The asterisks denote the position of the selected target resonances which were excited. (b,c) selective 1D TOCSY spectra of the above solution. In (b) τ_m = 400 ms was used for magnetization transfer from H11 to H9 of the (9-cis, 11-trans) 18:2 CLA. In (c) τ_m = 70 ms was used in order to optimize magnetization transfer from H10, 11 to H9, 12 of the (9-trans, 11-trans) 18:2 CLA. The asterisks denote the resonances that were excited with the use of a selective pulse.

Figure 31

Oxidized products of methyl linoleate. Adopted, with permission, from [143]. Copyright 1983, Oil Chemists’ Society (AOCS).

Figure 32

Some of the primary oxidation compounds derived from (a) oleic and (b) linolenic acyl groups, that can be detected by ¹H-NMR. Adopted, with permission, from [26]. Copyright 2014, Institute of Food Technologists.

Figure 33

Some of the secondary or further oxidation compounds derived from linoleic acyl groups that can be detected by ¹H-NMR. Adopted, with permission, from [26]. Copyright 2014, Institute of Food Technologists.

Figure 34

Total aldehydes in 72 oils and 38 cosmetics. The boxplot shows a statistical distribution whiskers: minimum and minimum (max 1.5 times the length of the inner quartiles; data points outside are outliers). Adopted, with permission, from [168]. Copyright 2010, John Wiley & Sons Ltd.

Figure 35

NMR spectrum of an authentic rancid lipstick sample (shown in insert) compared to a non-oxidized sample. Adopted, with permission, from [168]. Copyright 2010, John Wiley & Sons Ltd.

Figure 36

600 MHz ¹H-NMR spectra of: (A) an organic lipid fraction of an Asiago cheese from an alpine farm with the assignment of the major components (A, olefinic protons of all unsaturated chains; B, bis-allylic protons; C, methylenic protons bonded to C2 of all fatty acid chains; D, allylic protons; E, methylenic protons bonded to C3; F, methylenic protons; G, methyl protons of linolenic acid; H, methyl protons of butyric acid; I, all other methyl protons. The insets illustrate a comparison of the lipid fraction of an alpine farm and an industrial factory. (B) Expansion of a region in which some important olefinic resonances from CLA are illustrated. Adopted, with permission, from [189]. Copyright 2008, American Chemical Society.

Figure 37

PCA of the ¹H-NMR resonances of lipid fractions: (a) score plot of PC1 vs. PC2 [training set (◆) alpine farms; (•) lowland industrialized factories, (▲) mountain industrialized factories; test set (◊) alpine farms; (○) lowland industrialized factories, (△) mountain industrialized factories]; (b) loadings profile (letters correspond to Figure 36). Adopted, with permission, from [189]. Copyright 2008, American Chemical Society.

Figure 38

¹³C-NMR aliphatic region (32.0–31.3 ppm) of lipids extracted from salmon of four different origins: (from top) wild salmon from Norway (NW), Scotland (SW), and Ireland (IW) and farmed salmon from Norway (NF). Adopted, with permission, from [193]. Copyright 2009, American Chemical Society.

Figure 39

¹³C-NMR spectrum of krill oil in the region of glycerol carbon atoms (A). Selected ¹³C regions of triacylglycerol (TAG) (B) and phosphatidylchorine (PC) (C). Adopted, with permission, from [197]. Copyright 2016, AOCS.

Figure 40

Semi-selective 800 MHz HSQC-TOCSY spectrum of the double bond region of blood plasma lipids. (a) Correlations to the allylic Δ–CH₂–Δ protons (ca. 2.8 ppm) (b); Δ–¹CH protons (ca. 2.05 ppm). In the Δ–CH₂–Δ region the 18:2 signals are well separated from the higher polyunsaturated fatty acids, whereas in the Δ–1 region the 18:1 signals are separated from all other signals. The 20:4 fatty acid shows additional well separated signals at 2.13 ppm. Adopted, with permission, from [89]. Copyright 1998, John Wiley & Sons, Ltd.

Figure 41

Representative ¹H-NMR spectrum (500 MHz) of human blood. Inset shows a schematic view of the micellar chylomicron structure. Adopted, with permission, from [202]. Copyright 2010, BioMed Central Ltd.

Figure 42

Binding of fatty acid to HSA induces significant conformational changes in the protein. (a) Crystal structure of unliganded HSA [223] (PDB ID, 1ao6). The subdomains are color-coded as follows: IA, red; IB, light-red; IIA, green; IIB, light-green; IIIA, blue; IIIB, light-blue. (b) Crystal structure of HSA-myristate [224] (PDB ID, 1bj5). Both crystal structures have been fully refined to 2.5 Å resolution. The six myristate molecules are numbered 1–6 and shown in a space-filling representation. Adopted, with permission, from [225]. Copyright 1999, Elsevier Science B.V.

Figure 42

Binding of fatty acid to HSA induces significant conformational changes in the protein. (a) Crystal structure of unliganded HSA [223] (PDB ID, 1ao6). The subdomains are color-coded as follows: IA, red; IB, light-red; IIA, green; IIB, light-green; IIIA, blue; IIIB, light-blue. (b) Crystal structure of HSA-myristate [224] (PDB ID, 1bj5). Both crystal structures have been fully refined to 2.5 Å resolution. The six myristate molecules are numbered 1–6 and shown in a space-filling representation. Adopted, with permission, from [225]. Copyright 1999, Elsevier Science B.V.

Figure 43

Correlation of the ¹³C-NMR and crystal structure data for HSA. (a) Crystal structure of HSA complexed with palmitate. Labels indicate the FA binding sites and their assigned chemical shifts. The effects of various drugs on these chemical shifts are noted. (b) Spectrum of [¹³C] palmitate bound to HSA at a FA/HSA mole ratio of 4:1. Although a broad shoulder signal may be present at ∼182.6 ppm, the peaks for FA sites 1 and 6 are typically not observed in spectra of wild-type HSA at the FA/HSA ratios investigated. The positions of these two peaks (derived from multiple competition experiments) are indicated in the spectrum for [¹³C]-palmitate bound to wild-type HSA at a ratio of 4:1. (c) Spectrum of [¹³C] palmitate bound to PFL-HSA at a FA/HSA mole ratio of 4:1. Adopted, with permission, from [236]. Copyright 2013, Elsevier B.V.

Figure 43

Correlation of the ¹³C-NMR and crystal structure data for HSA. (a) Crystal structure of HSA complexed with palmitate. Labels indicate the FA binding sites and their assigned chemical shifts. The effects of various drugs on these chemical shifts are noted. (b) Spectrum of [¹³C] palmitate bound to HSA at a FA/HSA mole ratio of 4:1. Although a broad shoulder signal may be present at ∼182.6 ppm, the peaks for FA sites 1 and 6 are typically not observed in spectra of wild-type HSA at the FA/HSA ratios investigated. The positions of these two peaks (derived from multiple competition experiments) are indicated in the spectrum for [¹³C]-palmitate bound to wild-type HSA at a ratio of 4:1. (c) Spectrum of [¹³C] palmitate bound to PFL-HSA at a FA/HSA mole ratio of 4:1. Adopted, with permission, from [236]. Copyright 2013, Elsevier B.V.

Figure 44

Summary of oleic acid (OA)-HSA competition experiments. (A) Crystal of HSA bound to seven labeled OA molecules. Drugs and endogenous compound primary and secondary binding sites are listed (PDB entry 1E7G). (B) ¹H-¹³C HSQC spectrum of the 4:1 OA–HSA complex. FA-2, -4, and -5 (blue) are primary sites for FA and are generally not impacted by competition binding. The low-affinity sites, FA-3, -6, and -7 (red), were identified through drug competition experiments. The locations of the competition of the drugs and endogenous compounds for their primary and secondary binding are listed. The spectrum was recorded at 500 MHz, 25 °C, and pH 7.4 in a 50 mM phosphate, 50 mM NaCl buffer. Adopted, with permission, from [235]. Copyright 2013, by the American Chemical Society.

Figure 45

(A) Comparison of various extraction methods in respect to the total (9-cis, 11-trans), (9-trans, 11-cis) and (9-cis, 11-cis) 18:2 CLA isomers and caproleic acid integrals. (B) Efficiency of the extraction of the milk lipid fraction using the Bligh and Dyer 1:2 method in respect to mg of lyophilized milk sample; (■), the relative integrals of the composite (9-cis, 11-trans), (9-trans, 11-cis) and (9-cis, 11-cis) 18:2 CLA isomers in respect to the standard reference compound as above. (○), the amount of the extracted composite (9-cis, 11-trans), (9-trans, 11-cis) and (9-cis, 11-cis) 18:2 CLA isomers expressed in mg g⁻¹ of lyophilized milk sample. Adopted, with permission, from [56]. Copyright 2014, by the Elsevier B.V.

Figure 45

(A) Comparison of various extraction methods in respect to the total (9-cis, 11-trans), (9-trans, 11-cis) and (9-cis, 11-cis) 18:2 CLA isomers and caproleic acid integrals. (B) Efficiency of the extraction of the milk lipid fraction using the Bligh and Dyer 1:2 method in respect to mg of lyophilized milk sample; (■), the relative integrals of the composite (9-cis, 11-trans), (9-trans, 11-cis) and (9-cis, 11-cis) 18:2 CLA isomers in respect to the standard reference compound as above. (○), the amount of the extracted composite (9-cis, 11-trans), (9-trans, 11-cis) and (9-cis, 11-cis) 18:2 CLA isomers expressed in mg g⁻¹ of lyophilized milk sample. Adopted, with permission, from [56]. Copyright 2014, by the Elsevier B.V.