Journal of the American Oil Chemists' Society

, Volume 90, Issue 12, pp 1819–1829

Identification of TAG and DAG and their FA Constituents in Lesquerella (Physaria fendleri) Oil by HPLC and MS

Authors

    • Western Regional Research Center, Agricultural Research Service, U.S. Department of Agriculture
  • Grace Q. Chen
    • Western Regional Research Center, Agricultural Research Service, U.S. Department of Agriculture
Original Paper

DOI: 10.1007/s11746-013-2313-1

Cite this article as:
Lin, J. & Chen, G.Q. J Am Oil Chem Soc (2013) 90: 1819. doi:10.1007/s11746-013-2313-1

Abstract

Castor oil has many industrial uses because of its high content (90 %) of the hydroxy fatty acid, ricinoleic acid (OH1218:19). Lesquerella oil containing lesquerolic acid (Ls, OH1420:111) is potentially useful in industry. Ten molecular species of diacylglycerols and 74 molecular species of triacylglycerols in lesquerella (Physaria fendleri) oil were identified by electrospray ionization mass spectrometry as lithium adducts of acylglycerols in the HPLC fractions of lesquerella oil. Among them were: LsLsO, LsLsLn, LsLsL, LsLn–OH20:2, LsO–OH20:2 and LsL–OH20:2. The structures of the four new hydroxy fatty acid constituents of acylglycerols were proposed by the MS of the lithium adducts of fatty acids as (comparing to those in castor oil): OH1218:29,14 (OH1218:29,13 in castor oil), OH1218:39,14,16 (OH18:3 not detected in castor oil), diOH12,1318:29,14 (diOH11,1218:29,13 in castor oil) and diOH13,1420:111 (diOH20:1 not detected in castor oil, diOH11,1218:19 in castor oil). Trihydroxy fatty acids were not detected in lesquerella oil. The differences in the structures of these C18 hydroxy fatty acids between lesquerella and castor oils indicated that the polyhydroxy fatty acids were biosynthesized and were not the result of autoxidation products.

Keywords

Hydroxy fatty acidsDi- and tri-acylglycerolsLesquerella oilMass spectrometryLesquerella fendleriPhysaria fendleri

Introduction

Ricinoleate (R, OH1218:19) has many industrial uses such as the manufacture of biodegradable plastics, nylon, plasticizers, lubricants, cosmetics, paints and surfactants, because of its hydroxyl group on the fatty acid (FA) chain. Castor oil is the only commercial source of ricinoleate. However because the castor bean contains the toxin, ricin, as well as potent allergens, it is hazardous to grow, harvest and process castor. It is desirable to produce ricinoleate in an oilseed of a transgenic plant lacking these toxic components.

Lesquerella [Physaria fendleri (A. Gray) O’Kane & Al-Shehbaz, formerly Lesquerella fendleri (A. Gray) S. Wats., Brassicaceae] [1] contains no such toxic components. Its oil contains ricinoleate and thus the plant can be developed to produce high level of ricinoleate for industrial uses. Physaria fendleri is a new industrial oilseed crop in the southwestern region of the U.S., valued for its lesquerolate (Ls, OH1420:111). Lesquerolate, a C20 homolog of ricinoleate, is the most abundant FA in lesquerella oil and can also be used for industry similar to those of ricinoleate. Physaria fendleri is amenable to Agrobacterium-mediated transformation [24]. To develop a transgenic Physaria fendleri plant capable of producing high levels of ricinoleate, lesquerolate and other useful FA, one must understand the components and the biosynthesis of lesquerella oil. We have recently reported the hydroxy FA biosynthesis and lipid gene expression during seed development in Physaria fendleri [5]. We have also recently proposed the biosynthetic pathway of castor oil [6].

Twenty five species of lesquerella were studied with respect to their FA and triacylglycerols (TAG) in seed oils including Physaria fendleri [79]. TAG and the contents (%) in Physaria fendleri oil were reported to have the highest content of TAG as OH20–18–OH20 (73.3 %, double bond not specified) by supercritical fluid chromatography [8]. The FA and their contents in Physaria fendleri oil reported [79] were similar. Atmospheric pressure chemical ionization mass spectrometry (APCI-MS) coupled to reversed-phase high-performance liquid chromatography (HPLC) was used for direct analysis of intact acylglycerols (AG) from the seed oils of Physaria fendleri, Physaria gordonii and castor [10]. We would like to report here the identification of TAG and diacylglycerols (DAG) in the seed oil of Physaria fendleri using electrospray ionization mass spectrometry (ESI–MS, multiple stages).

Experimental Procedure

Physariafendleri Oil

The Physaria fendleri seeds, WCL-LY2 [11], were kindly provided by Dr. David Dierig (USDA-ARS, National Center for Genetic Resources Preservation, Fort Collins, CO 80521, USA). Plants were germinated and grown in a greenhouse at temperatures between 28 °C (day) and 18 °C (night), with supplemental metal halide lighting to provide 15-h-day length (1,000–1,250 μmol m−2 s−1). Mature female flowers were hand-pollinated and the seeds were harvested at about 49 days after pollination. The seeds were vacuum dried overnight at 50 °C and ground. The ground seeds (0.03 g) in 10-mL glass tube were extracted with 2 mL hexane/2-propanol (8:2) containing 50 μg/mL butylated hydroxytoluene (BHT) to prevent oxidation during extraction. The extraction took place at 55 °C for 30 min with shaking every 10 min. The extract was filtered and dried over sodium sulfate, and the solvent was evaporated under a stream of nitrogen.

Materials

Lithium acetate was obtained from Sigma (St. Louis, MO, USA). High purity methanol and 2-propanol (Honeywell Burdick & Jackson) for LC-MS were purchased from VWR International (West Chester, PA, USA). High purity nitrogen (Praxair, Oakland, CA, USA) was used as the sheath gas. Research grade helium (Praxair) was used as the collision gas.

HPLC of the Molecular Species of Acylglycerols in Lesquerella Oil

HPLC was carried out on a liquid chromatograph (Waters Associates, Milford, MA, USA), using an absorbance detector (Waters, Model 2487) at 205 nm. Molecular species of AG were separated using a C18 analytical column (Gemini, 250 × 4.6 mm, 5 μm, C18, Phenomenex, Torrance, CA, USA) with a linear gradient from 100 % methanol to 100 % 2-propanol in 40 min, at 1 mL/min flow rate. For fractionation, 1 mg of lesquerella (Physaria fendleri) oil in ethanol (50 μl) was chromatographed at 22 °C (room temperature) (Fig. 1). Fractions were collected every 30 s and corresponding fractions were pooled from 8 HPLC runs. HPLC fractions were used for MS studies. The final methanol solutions of samples were prepared for direct infusion into the mass spectrometer by combining approximately one fourth or half of each HPLC fraction with 50 μL of methanol solution of 100 mM lithium acetate and diluting to a total volume of 250 μL.
https://static-content.springer.com/image/art%3A10.1007%2Fs11746-013-2313-1/MediaObjects/11746_2013_2313_Fig1_HTML.gif
Fig. 1

HPLC chromatogram of lesquerella oil (1 mg) for fractionation (0.5 min/fraction). HPLC conditions: Detection at 205 nm, C18 analytical column (Gemini, 250 × 4.6 mm, 5μ), linear gradient from 100 % methanol to 100 % 2-propanol in 40 min, flow rate at 1 mL/min. The variation of the retention times of the eight HPLC fractionations (in the same day) pooled together was less than 0.1 min. For retention times of minor HPLC peaks, refer to AG and HPLC fraction numbers in Table 1

Electrospray Ionization Mass Spectrometry (ESI-MS)

An LCQ Advantage ion-trap mass spectrometer (MS 2.0) with Xcalibur 2.0 SR2 software (ThermoFisher Scientific, San Jose, CA, USA) was utilized for MS analysis of the various molecular species of AG. The infusion at a 2.5 μL/min flow rate from a syringe (250 μL) pump produced stable singly-charged lithiated parent ions which were subsequently fragmented for MS2, MS3 and MS4 analysis. ESI source conditions were as follows: sheath gas (nitrogen) flow rate, 10 arbitrary units (au); aux/sweep gas flow rate, 0 au; spray voltage, 4 kV; capillary temperature, 200 °C; capillary voltage, 5 V; tube lens offset, 15 V. Scan conditions were as follows: isolation width, 1.5 m/z; normalized collision energy, 27–42 %; scan ranges, 100–1,500 m/z. Acquire time was usually 3 min.

Results and Discussion

Lesquerella (Physaria fendleri) oil was fractionated (0.5 min/fraction) by HPLC as shown in Fig. 1. The methanol solution of individual HPLC fractions together with lithium acetate was infused into the mass spectrometer. MS2 fragmentation data were used for the identification of AG. Table 1 shows the mass-to-charge ratios (m/z) of the molecular ions of the lithium adducts of AG and their fragment ions (MS2) as well as the relative abundances of the fragment ions. We have identified many molecular species of DAG and TAG in lesquerella oil by MS2 (Table 1). The identification of TAG depended on the fragment ions produced from the neutral losses of three constituent FA, [M + Li − FA]+. The identification of DAG depended on both [M + Li − FA]+ and [FA + Li]+ because some of the DAG in Table 1 shows only one (not two) of the [M + Li − FA]+. We have recently reported the fragmentation of the lithium adducts of DAG [12]. The more hydroxylated FA in a DAG could not be detected as [M + Li − FA]+ (or trace only) and the less hydroxylated FA in a DAG could not be detected as [FA + Li]+ (or trace only) (Table 1). The identification of tetraacylglycerols (acylglycerol estolides) was reported separately [13]. The names of AG used in this report show the FA constituents only, without the indication of FA locations on the glycerol backbone.
Table 1

MS2 characteristic ion (m/z) and their relative abundance (% in parenthesis) of diacylglycerols and triacylglycerols in the HPLC fractions of the seed oil of Physaria fendleri

Acylglycerols (HPLC fraction #)

[M + Li]+

[M + Li − FA]+

Other ions

Ls–OH18:2 (11)

685.6

359.2(2)a, 389.2(3)b

333.2(4)d, 303.2(3)e, 571.4(5)g, 667.6(100)i

LsR (12)

687.6

361.2(13)a, 389.2(22)b

333.2(18)d, 305.2(16)e, 573.4(52)g, 686.6(100)

Ls–OH20:2 (12)

713.5

387.2(82)a, 389.2(67)b

333.2(61)d, 331.2(72)e, 599.4(100)g, 601.4(94)h

LsLs (13, 14)

715.6

389.2(72)a

333.2(65)d, 601.4(100)g

Ln–OH20:2(16)

665.4

387.1(79)a

331.2(86)e, 553.3(100)h

LsLn (17, 18)

667.5

389.2(71)b

333.2(80)d, 553.4(100)g, 649.4(15)i

L–OH20:2 (18)

667.5

387.2(17)a

331.2(20)e, 553.4(100)g, 649.4(15)i

LsLs–diOH18:1 (18)

1011.7

685.5(100)a, 697.5(17)c

897.6(1)g, 667.5(10)

LsLs–diOH18:2 (19)

1009.7

683.5(100)a, 697.5(28)c

319.2(2)f, 895.5(1)g

LsL (19, 20)

669.5

343.1(1)a, 389.2(66)b

333.2(79)d, 555.4(100)g,651.4(10)i

O–OH20:2 (21)

669.5

387.1(74)a, 345.1(1)b

331.2(86)e, 557.4(100)h, 555.4(15)g, 651.4(14)i

LsLs–OH18:3 (21, 22)

991.6

665.4(100)a, 697.5(22)c

333.2(1)d, 389.2(4), 877.5(1)g

Ls–OH20:2–OH18:2 (21, 22)

991.6

665.4(100)a, 667.5(26)b, 695.5 (13)c

333.2(1)d, 389.2(4), 877.5(1)g

LsO (22,23)

671.5

389.2(73)b

333.2(81)d, 557.4(100)g,653.4(10)i

LsR–OH20:2 (23)

993.8

667.4(100)a, 695.5(11)b, 669.5(22)c

333.1(1)d, 389.1(4), 879.7(2)g

LsLs–OH18:2 (23, 24)

993.8

667.4(100)a, 697.5(28)c

333.1(2)d, 389.1(6), 879.7(1)g

LsLsR (25, 26)

995.8

669.5(100)a, 697.5(32)c

333.2(2)d, 389.2(4), 881.6(2)g

LsLs–OH20:2 (26)

1021.8

695.5(100)a, 697.5(40)c

389.2(2), 907.6(1)g

LsLn–diOH20:0 (28)

993.8

667.5(100)a, 715.5(29)b, 649.5(28)c

351.3(7)f, 879.5(1)g

LsLsLs (28, 29)

1023.8

697.5(100)a

389.2(2), 909.6(1)g

LsLn–OH18:2 (29)

945.7

619.4(100)a, 667.5(51)b, 649.4(80)c

333.2(1)d, 303.1(1)f, 695.5(10)

RLn–OH20:2 (29)

945.7

647.4(9)a, 667.5(51)b, 621.4(30)c

831.5(2)g, 833.5(1)h

Ln–OH20:2–OH20:2 (31)

971.7

693.5(22)a, 647.4(100)b

331.2(1)e, 859.5(1)h

LsRLn (31)

947.6

621.4(100)a, 649.4(37)b, 669.5(25)c

333.2(1)d, 833.5(2)g

LsL–diOH20:0 (31)

995.7

669.5(100)a, 715.5(20)b, 651.5(20)c

351.3(4)f, 881.5(1)g

LsO–diOH20:1 (31)

995.7

669.5(100)a, 713.5(13)b, 653.5(15)c

349.2(2)f, 881.5(1)g

LsO–diOH18:0 (31)

969.7

643.5(100)a, 687.5(34)b, 653.5(46)c

323.2(5)f, 855.6(1)g

LsLn–OH20:2 (32, 33)

973.6

647.3(100)a, 695.5(32)b, 649.4(84)c

331.2(1)f, 859.2(1)g

LsLsLn (34–36)

975.7

649.4(100)a, 697.5(20)c

333.1(1)d, 861.5(1)g

LsLsPo (36)

951.6

625.4(100)a, 697.5(18)c

333.2(1)d, 837.5(1)g

LsL–OH20:2 (36)

975.6

649.4 (100)a, 695.4(24)b, 651.4(62)c

333.1(2)d, 331.1(1)f, 861.5(3)g

LsLsL (37, 38)

977.6

651.4(100)a, 697.5(18)c

333.2(1)d, 863.4(1)g

LsO–OH20:2 (38, 39)

977.6

651.4(100)a, 695.4(40)b, 653.4(88)c

331.2(2)f, 863.4(1)g, 865.4(1)h

LsLsP (40, 41)

953.8

627.3(100)a, 697.4(25)c

333.1(2)d, 839.5(2)g

LsLsO (40, 41)

979.8

653.4(100)a, 697.5(19)c

333.1(1)d, 865.5(1)g

LsLnLn (43, 44)

927.7

601.3(4)a, 649.4(100)b

333.2(7)d, 813.4(1)g, 595.2(4)j

LsO–OH20:0 (44)

981.6

655.4(100)a, 699.5(24)b, 653.4(64)c

333.1(1)d, 867.5(2)g, 627.4(5)

LsO-23:0 (44, 45)

1007.6

681.4(100)a, 725.5(40)b, 653.4(85)c

333.1(1)d, 361.1(1)f, 893.5(3)g

LsLsS (44, 45)

981.6

655.4(100)a, 697.4(29)c

333.1(1)d, 867.5(2)g

LsS–OH18:3 (45)

949.6

623.4(77)a, 665.4(100)b, 655.5(62)c

301.1(1)f, 835.5(2)g, 891.5(3)

LsLLn (46-47)

929.5

603.3(6)a, 649.4(63)b, 651.4(100)c

333.2(13)d, 815.4(2)g, 597.2(7)j

OLn–OH20:2 (47)

929.6

647.4(89)a, 651.4(100)b, 605.3(3)c

331.1(13)f, 817.4(2)h, 599.2(7)l

LsLPo (48)

905.6

579.4(4)a, 625.4(58)b, 651.4(81)c

333.2(30)d, 791.4(4)g, 573.3(17)j

LsO-16:2 (48)

905.6

579.4(4)a, 623.4(56)b, 653.4(43)c

333.2(30)d, 791.4(4)g, 573.3(17)j

LsLnP (48, 49, 50)

905.6

579.4(4)a, 627.4(43)b, 649.4(100)c

333.2(30)d, 791.4(4)g, 573.3(17)j

LsLL (48, 49, 50)

931.6

605.3(3)a, 651.4(100)b

333.2(8)d, 817.4(1)g, 599.3(5)j

LsOLn (49, 50)

931.6

605.3(6)a, 649.4(100)b, 653.4(72)c

333.2(16)d, 817.4(3)g, 599.3(8)j

LsOPo (51)

907.6

581.4(4)a, 625.4(45)b, 653.4(36)c

333.2(21)d, 793.4(3)g, 575.3(10)j

SLn–OH20:2 (51)

931.6

647.4(100)a, 653.4(46)b, 607.4(2)c

331(13)f, 819.5(2)h, 601.3(6)l

OL–OH20:2 (51)

931.6

649.4(15)a, 651.4(4)b, 607.4(2)c

331(13)f, 819.5(2)h, 601.3(6)l

LsLP (51, 52)

907.6

581.4(4)a, 627.4(39)b, 651.4(100)c

333.2(21)d, 793.4(3)g, 575.3(10)j

LsOL (52, 53)

933.6

607.4(5)a, 651.4(100)b, 653.4(50)c

333.1(13)d, 819.5(5)g, 601.3(6)j

LsSLn (53, 54)

933.7

607.3(5)a, 649.4(100)b, 655.5(39)c

333.1(15)d, 819.5(3)g, 601.3(9)j

LsLn-20:1 (53, 54)

959.6

633.4(5)a, 681.5(39)b, 649.4 (100)c

333.2(12)d, 845.5(3)g, 627.3(7)j

LsOP (55)

909.6

583.3(2)a, 627.4(40)b, 653.4(100)c

333.2(14)d, 795.5(2)g, 577.3(2)j

LsOO (55, 56)

935.6

609.3(2)a, 653.4(100)b

333.1(8)d, 821.3(2)g, 603.2(5)j

LsLS (56, 57)

935.6

609.3(5)a, 655.4(37)b, 651.4(100)c

333.1(21)d, 821.4(3)g, 603.2(14)j

LsL-20:1 (56, 57)

961.6

635.5(5)a, 681.5(38)b, 651.4(100)c

333.2(11)d, 847.4(2)g, 629.3(7)j

LsLn-20:0 (58)

961.7

635.4(4)a, 683.5(40)b, 649.4(100)c

333.2(12)d, 847.5(3)g, 629.3(8)j

LsOS (59, 60)

937.6

611.3(2)a, 655.4(38)b, 653.4(100)c

333.2(13)d, 823.5(2)g, 605.3(7)j

LsO-20:1 (59, 60)

963.8

637.4(3)a, 681.4(45)b, 653.4(100)c

333.2(13)d, 849.6(2)g, 631.4(11)j, 907.7(4)

LsL-20:0 (60, 61)

963.7

637.4(4)a, 683.5(37)b, 651.4(100)c

333.2(12)d, 849.5(3)g, 631.3(8)j

OLnPo (61)

859.5

577.3(100)a, 581.3(85)b, 605.3(60)c

571.3(20)j, 575.3(22)j, 599.3(13)j

LnPPo (61)

833.5

555.3(79)a, 577.3(100)b, 579.3(55)c

549.2(22)j, 571.3(20)j, 573.3(10)j

OLLn (62)

885.7

603.3(82)a, 605.3(100)b, 607.3(67)c

597.2(14)j, 599.2(19)j, 601.2(15)j

LsLn-24:1(62)

1015.6

689.4(6)a, 737.5(39)b, 649.4(100)c

333.2(9)d, 901.5(2)g, 683.3(7)

LLnP (62)

859.5

579.4(58)a, 581.3(79)b, 603.3(100)c

773.4(10)

OPoPo (63)

835.6

553.3(73)a, 581.3(100)b

547.2(13)j, 575.3(18)k

LsO-20:0 (63, 64)

965.7

639.4(2)a, 683.5(40)b, 653.4(100)c

333.2(11)d, 851.5(3)g, 633.3(8)j

OPLn (65)

861.5

579.4(60)a, 605.3(100)b, 583.4(48)c

573.4(16)j, 599.2(25)k, 577.3(23)l

LLP (65)

861.5

581.4(40)a, 605.3(100)c

575.3(12)j, 599.2(25)k

OOLn (65, 66)

887.6

605.3(100)a, 609.3(49)c

599.2(22)j, 603.3(30)l

OLL (65, 66)

887.6

605.3(100)a, 607.3(16)b

599(22)j, 601.2(3)k

LSLn (65, 66)

887.6

607.3(16)a, 603.3(30)b, 609.3(49)c

601.2(3)j, 597.2(3)k, 603.3(30)l

OPPo (66, 67)

837.6

555.3(82)a, 581.3(100)b, 583.4(43)c

549.2(27)j, 575.3(26)k, 577.3(17)l

OOPo (67)

863.6

581.3(100)a, 609.4(27)c

575.2(28)j, 603.4(8)l

OOL (68)

889.6

607.3(100)1, 609.4(36)c

601.3(25)j, 603.3(11)l

OLP (68)

863.7

581.4(100)a, 583.4(59)b, 607.4(93)c

575.2(27)j, 577.2(16)k, 601.3(23)l

OSLn (70)

889.6

607.4(75)a, 605.4(100)b, 611.4(53)c

601.3(16)j, 559.2(17)k, 605.412

OOP (70, 71)

865.7

583.3(100)a, 609.4(51)c

577.3(34)j, 603.3(16)l

OSPo (70, 71)

865.7

583.3(100)a, 581.3(37)b, 611.3(20)c

577.3(34)j, 575.3(13)k, 605.3(4)l

OOO (71, 72)

891.7

609.4(100)a

603.3(11)j

OLS (72)

891.7

609.4(100)a, 611.4(28)b, 607.4(47)c

603.3(19)j, 605.3(5)k, 601.3(8)l

OOS (75)

893.7

611.4(100)a, 609.4(72)b

605.4(24)j, 603.4(15)l

a[M + Li − FA]+ for the first FA shown on the names of di- and triacylglycerols on first column on the Table

b[M + Li − FA]+ for the second FA shown on the names of di- and triacylglycerols

c[M + Li − FA]+ for the third FA shown on the names of triacylglycerols

d[FA + Li]+ for the first FA shown on the names of di- and triacylglycerols

e[FA + Li]+ for the second FA shown on the names of di- and triacylglycerols

f[FA + Li]+ for the third FA shown on the names of triacylglycerols

g[M + Li − C6H13CHO (114)]+

h[M + Li − C6H11CHO (112)]+

i[M + Li − H2O]+

j[M + Li − FALi]+, for the first FA shown on the names of triacylglycerols

k[M + Li − FALi]+, for the second FA shown on the names of triacylglycerols

l[M + Li − FALi]+, for the third FA shown on the names of triacylglycerols

The HPLC elusion order of LsLs–diOH18:1 (HPLC fraction #18) and LsLs–diOH18:2 (HPLC fraction #19) in Table 1 seemed to contradict our earlier report [14] that the addition of a double bond on TAG made the TAG eluted earlier on a reversed-phase column. The extra double bond on the TAG made the elusion order reversed. A hydroxyl group is much more polar than a double bond. The addition of a double bond made the orientation of the hydroxyl groups different and the polarity of the TAG different.

An example of the identification of TAG using MS2 is given here. Figure 2 shows the MS2 fragmentation of the lithium adduct of TAG, LsO–diOH18:0, in the HPLC fraction #31 of lesquerella oil (Table 1). The MS spectrum shows three major fragment ions from the neutral losses of the constituent FA, [M + Li − Ls]+ at 643.5, [M + Li − diOH18:0]+ at 653.4, and [M + Li − O at 687.5. Table 1 lists the fragment ions, [M + Li − FA]+, as a column with the superscripts a, b and c representing the losses of FA in the order of FA on the names of AG. A minor fragment ion, [diOH18:0 + Li]+ at 323.2, was detected in Fig. 2 while [Ls + Li]+ at 333.2 and [O + Li]+ at 289.2 were not detected. The more hydroxylated FA in the TAG molecule tended to be shown as [FA + Li]+ more abundantly than those of the less hydroxylated FA (Table 1). Table 1 lists the fragment ions, [FA + Li]+, with the superscripts d, e and f representing the FA lithium adducts in the order of FA on the names of AG. A very minor fragment ion is also shown in Fig. 2 as m/z at 855.5, [M + Li − C6H13CHO (114)]+ (superscript g in Table 1). This was from lesquerolic acid constituent (Fig. 5) similar to that of ricinoleate reported earlier [15, 16].
https://static-content.springer.com/image/art%3A10.1007%2Fs11746-013-2313-1/MediaObjects/11746_2013_2313_Fig2_HTML.gif
Fig. 2

The MS2 spectrum of [M + Li]+ at m/z 969.7 from LsO–diOH18:0 in the HPLC fraction #31

Figure 2 is simple with very few fragment ions and does not show the fragment ions from the neutral losses of FA lithium salts [M + Li − FALi]+. The presence of [M + Li − FALi]+ might interfere with other important fragment ions and made the MS spectrum more complicated to interpret. Twenty TAG in Table 1 contain three normal FA (without hydroxy FA). The fragment ions [M + Li − FALi]+, are indicated in Table 1 by the superscripts j, k and l. The MS2 spectra of all of these twenty TAG showed the three fragment ions from the three constituent FA [M + Li − FALi]+. The abundances of these fragment ions were lower than those of [M + Li − FA]+. The MS2 spectra of 28 TAG containing two or three hydroxy FA in Table 1 showed no fragment ions of [M + Li − FALi]+. The MS2 spectra of the 26 TAG containing one hydroxy FA in Table 1 showed one fragment ion from the hydroxy FA, [M + Li − FALi]+. The fragment ions from the other two normal FA of the 26 TAG were only a trace or not detectable, and are not listed in Table 1. Among the ten of the DAG in Table 1 containing one or two hydroxy FA, the fragment ions [M + Li − FALi]+ were not detected. This was similar to those of DAG in castor oil recently reported [12]. We also observed that DAG containing two normal FA produced abundant fragment ions [M + Li − FALi]+ [12]. There were differences of fragmentations between AG containing hydroxy FA and AG containing no hydroxy FA. It is interesting to know why.

Hsu and Turk outlined the advantages of using Li+ (rather than Na+ or NH4+) for TAG analysis, in particular for the location of double bonds (based on dilithiated FA ions) and for the regioisomeric position on the glycerol [17, 18]. However, we did not detect the dilithiated FA ions [FA − H + 2Li]+ in Fig. 2 and Table 1 as well as many MS2 and MS3 spectra of TAG and DAG we obtained. We used lithium acetate in the sample solution for infusion and Hsu and Turk used lithium hydroxide [17]. Lithium hydroxide formed the ion of FA lithium salt as [FALi + Li]+ instead of [FA + Li]+.

We have identified ten molecular species of DAG and 74 molecular species of TAG in Physaria fendleri oil (Table 1). Using the same method, we previously identified 13 DAG and 66 TAG in castor oil [6, 12, 16, 19, 20]. Byrdwell and Neff [10] identified 11 DAG and 44 TAG in Physaria fendleri oil on the peaks of the reconstructed ion chromatogram. Atmospheric pressure chemical ionization mass spectrometry (APCI-MS) coupled to HPLC was used for the identification of TAG from Physaria fendleri oil [10]. Single quadrupole APCI-MS spectra were used for the identification and the spectra were complicated [10]. MS2 spectra of AG of lithium adducts of ESI-MS were simple as shown in Fig. 2.

Figure 3 is the MS4 spectrum of [OH18:2 + Li]+ from LsLs–OH18:2 in the HPLC fraction #24 of lesquerella oil (Table 1). Ls-OH20:2–OH18:2 in the HPLC fractions #21, 22 and LsLn–OH18:2 in the HPLC fraction #29 (Table 1) also contain OH18:2 and their MS4 spectra of [OH18:2 + Li]+ were similar to Fig. 3. We have recently reported the mass spectrum of [OH18:2 + Li]+ from castor oil (Fig. 4 of Ref. [6]) as well as the proposed fragmentation pathway of its fragment ion at m/z 219.0. The fragment ion was used to propose the location of the second double bond on OH18:2 in castor oil at C-13 as OH1218:29,13. We depended on the proposed fragmentation pathways to propose the structures of new FA. The fragment ion at m/z 219.0 was not detected in Fig. 3 of [OH18:2 + Li]+ from lesquerella oil. Therefore, two OH18:2 from castor oil and lesquerella oil were different.
https://static-content.springer.com/image/art%3A10.1007%2Fs11746-013-2313-1/MediaObjects/11746_2013_2313_Fig3_HTML.gif
Fig. 3

The MS4 spectrum of the fragment ion [OH18:2 + Li]+ at m/z 303.1 from LsLs–OH18:2 in the HPLC fraction #24 of lesquerella oil. The MS2 was from the molecular ion [M + Li]+ at m/z 993.8. The MS3 was from the fragment ion [M + Li − Ls]+ at m/z 667.4. The proposed fragmentation pathways of fragment ions at m/z 231.0 and m/z 191.0 are also shown here

The proposed fragmentation pathways of the fragment ions [OH18:2 + Li − C6H11CHO (112)]+ at m/z 191.0 and [OH18:2 + Li − C5H12]+ at m/z 231.0 are also given in Fig. 3. The fragment ion at m/z 231 was a ketene. The pathways involved with ketene have been reported [16, 19, 21]. The proposed fragmentation pathways suggested that the second double bond of OH18:2 was at the C-14 position as OH1218:29,14. The fragment ions [M + Li − C6H11CHO (112)]+ shown in Table 1 designated with superscript h were associated with OH20:2 (Fig. 6). Densipolic acid, OH1218:29,15, identified earlier in Physaria densipila oil [22] was not detected in Physaria fendleri oil ([9] and the current study).

We have proposed a fragmentation pathway of m/z 219 from OH1218:29,13 in Fig. 4 of our recent report [6]. However there was another possible pathway of m/z 219 from OH1218:29,14 similar to m/z 191 of the Fig. 3 produced by fragmentation on the other side of the hydroxyl group. We assumed that the location of the second double bond (newly biosynthesized) was next to the hydroxyl group as diOH11,1218:29,13, and triOH11,12,1318:29,14 in castor oil [16, 20]. The structure of OH1218:29,13 in Fig. 4 of our recent report [6] was the likely structure, not OH1218:29,14.

Figure 4 is the MS4 spectrum of [OH18:3 + Li]+ at m/z 301.1 from LsLs–OH18:3 in the HPLC fraction #22 of lesquerella oil (Table 1). Similar to Fig. 3 of [OH18:2 + Li]+, Fig. 4 of [OH18:3 + Li]+ also shows the presence of a fragment ion at m/z 231.0. The structure of OH18:3 was similar to that of OH18:2 in lesquerella oil except the third double bond. The proposed fragmentation pathway of the fragment ion at m/z 243.0 is also given in Fig. 4. The third double bond was proposed at the C-16 position. The structure of OH18:3 in lesquerella oil was proposed as OH1218:39,14,16. We did not detect OH18:3 in castor oil [6].
https://static-content.springer.com/image/art%3A10.1007%2Fs11746-013-2313-1/MediaObjects/11746_2013_2313_Fig4_HTML.gif
Fig. 4

The MS4 spectrum of [OH18:3 + Li]+ at m/z 301.1 from LsLs–OH18:3 in the HPLC fraction #22 of lesquerella oil. The MS2 was from the molecular ion [M + Li]+ at m/z 991.8. The MS3 was from the fragment ion [M + Li − Ls]+ at m/z 665.4. The proposed fragmentation pathway of fragment ion at m/z 243.0 is also shown here

Figure 5 is the MS3 spectrum of lesquerolic acid [OH20:1 + Li]+ from LsLn in the HPLC fraction #17. The spectrum was similar to that of ricinoleic acid (OH18:1) (Fig. 7 of Ref. [16]). The major fragment ions in Fig. 5, at m/z 219.0, 236.3 and 315.1 were the same as those of ricinoleic acid plus C2H4 (28 amu). There were many continuous minor fragment ions between m/z 219 and m/z 236.3 (Fig. 5) which were similar to those of ricinoleic acid [16], and we do not know what these fragment ions were. Lesquerolic acid (OH1420:111) is the most abundant FA in lesquerella oil. The proposed fragmentation pathway of the fragment ion at m/z 219.0 is also shown in Fig. 5.
https://static-content.springer.com/image/art%3A10.1007%2Fs11746-013-2313-1/MediaObjects/11746_2013_2313_Fig5_HTML.gif
Fig. 5

The MS3 spectrum of lesquerolic acid [Ls + Li]+ at m/z 333.1 from LsLn in the HPLC fraction #17 of lesquerella oil. The MS2 was from the molecular ion [M + Li]+ at m/z 667.5. The proposed fragmentation pathway of fragment ion at m/z 219.0 is also shown here

Figure 6 is the MS4 spectrum of [OH20:2 + Li]+ at m/z 331.1 from LsLs–OH20:2 in the HPLC fraction #26 of lesquerella oil (Table 1). There are fifteen molecular species of AG containing OH20:2 (auricolic acid) in Table 1. Presumably this OH20:2 is auricolic acid (OH1420:211,17) identified recently in Physaria fendleri oil with a content of 3.3 % [9]. Auricolic acid was not identified in castor oil [6].
https://static-content.springer.com/image/art%3A10.1007%2Fs11746-013-2313-1/MediaObjects/11746_2013_2313_Fig6_HTML.gif
Fig. 6

The MS4 spectrum of [OH20:2 + Li]+ at m/z 331.1 from LsLs–OH20:2 in the HPLC fraction #26 of lesquerella oil. The MS2 was from the molecular ion [M + Li]+ at m/z 1,021.8. The MS3 was from the fragment ion [M + Li − Ls]+ at m/z 695.5. The structure of the FA was auricolic acid (OH1420:211,17)

Figure 7 is the MS4 spectrum of [diOH18:1 + Li]+ at m/z 321.2 from LsLs–diOH18:1 in the HPLC fraction #18 of lesquerella oil, the only AG containing diOH18:1 in Table 1. We have reported the mass spectrum of [diOH11,1218:19 + Li]+ at m/z 321.2 from castor oil (Fig. 5; Ref. [16]). The major fragment ions were [diOH18:1 + Li − C5H11CH=C=O (112)]+ at m/z 209.2 (relative abundance 100 %) and [diOH18:1 + Li − C6H13CHO (114)]+ at m/z 207.1 (relative abundance 36 %) from castor oil [16]. However Fig. 7 from lesquerella oil shows the major fragment ion [diOH18:1 + Li − 112]+ at m/z 209.1, and the fragment ion [diOH18:1 + Li − 114]+ at m/z 207.1 was only detected slightly. In our recent studies of minor acylglycerols in castor oil [6], the fragment ion of [diOH18:1 + Li − 114]+ at m/z 207 in the MS4 spectra of [diOH18:1 + Li]+ at m/z 321 might be detected as a minor fragment ion (unpublished). We proposed the structure of diOH18:1 in lesquerella oil as diOH11,1218:19 which was the same as that in castor oil [16]. The proposed fragmentation pathway of the fragment ion m/z at 209.1 was given as Fig. 6B of Ref. [16].
https://static-content.springer.com/image/art%3A10.1007%2Fs11746-013-2313-1/MediaObjects/11746_2013_2313_Fig7_HTML.gif
Fig. 7

The MS4 spectrum of [diOH18:1 + Li]+ at m/z 321.2 from LsLs–diOH18:1 in the HPLC fraction #18 of lesquerella oil. The MS2 was from the molecular ion [M + Li]+ at m/z 1011.8. The MS3 was from the fragment ion [M + Li − Ls]+ at m/z 685.5. The structure was proposed as diOH11,1218:19

Figure 8 is the MS4 spectrum of [diOH18:2 + Li]+ at m/s 319.2 from LsLs–diOH18:2 in the HPLC fraction #19 in lesquerella oil, the only AG containing diOH18:2 in Table 1. We have reported the mass spectrum of [diOH18:2 + Li]+ at m/z 319.2 from castor oil (Fig. 8; Ref. [16]). The fragment ions at m/z 205 and 209 were used to proposed the structure of diOH11,1218:29,13 in castor oil [16]. The fragment ion at m/z 205 is not detected in Fig. 8 and fragment ion at m/z 209 is only slightly detected in Fig. 8. The fragment ions at m/z 231 and 249 in Fig. 8 were not detected in Fig. 8 of Ref. [16]. So these two diOH18:2 were different. The proposed fragmentation pathways of fragment ions at m/z 231 and 249 are given in Fig. 8. We propose the structure of diOH18:2 in lesquerella oil as diOH12,1318:29,14.
https://static-content.springer.com/image/art%3A10.1007%2Fs11746-013-2313-1/MediaObjects/11746_2013_2313_Fig8_HTML.gif
Fig. 8

The MS4 spectrum of [diOH18:2 + Li]+ at m/s 319.2 from LsLs–diOH18:2 in the HPLC fraction #19 in lesquerella oil. The MS2 was from the molecular ion [M + Li]+ at m/z 1009.8. The MS3 was from the fragment ion [M + Li − Ls]+ at m/z 683.4. The proposed fragmentation pathways of fragment ions at m/z 231.1 and m/z 249.1 are also shown here

Figure 9 is the MS4 spectrum of [diOH20:1 + Li]+ at m/z 349.2 from LsO–diOH20:1 in the HPLC fraction #31 of lesquerella oil, the only AG containing diOH20:1 in Table 1. Figure 9 shows the fragment ion [M + Li − 114]+ at m/z 235.0 without the detection of the fragment ion [M + Li − 112]+ at m/z 237.0. This is different from Fig. 7 showing the major fragment ion [M + Li − 112]+ at m/z 209.0. However, fragment ion [M + Li − 114]+ at m/z 235.0 in Fig. 9 was shown in the MS spectrum of diOH11,1218:19 in the Fig. 5 of [16]. Unidentified larger peaks at m/z 136.9 and 154.3 in Fig. 9 are also shown in Figs. 7 and 5 of [16]. We proposed the structure of diOH20:1 in lesquerella oil as diOH13,1420:111. This structure agreed with our proposed diOH11,1218:19 in castor oil [16] and current study of lesquerella oil (Fig. 7). The proposed fragmentation pathway of fragment ion at m/z 235.0 was similar to that of Fig. 6A of [16] and Fig. 5.
https://static-content.springer.com/image/art%3A10.1007%2Fs11746-013-2313-1/MediaObjects/11746_2013_2313_Fig9_HTML.gif
Fig. 9

The MS4 spectrum of [diOH20:1 + Li]+ at m/z 349.2 from LsO–diOH20:1 in the HPLC fraction #31 of lesquerella oil. The MS2 was from the molecular ion [M + Li]+ at m/z 995.7. The MS3 was from the fragment ion [M + Li − Ls]+ at m/z 713.6. The proposed fragmentation pathway of fragment ion at m/z 235.0 is also shown here

We have proposed the structures of hydroxy FA using the MS4 spectra of FA from TAG in lesquerella oil as our earlier reports of castor oil [6, 16, 19, 20]. We would like to be sure that these MS4 spectra of FA from TAG could be compared with the MS2 spectra of FA standards at different concentrations and isolation widths. We have obtained MS2 spectra of ricinoleate standard (Sigma) at various sample concentrations (from 50 μg to 4 mg in 250 μl sample solution) and two different isolation widths (1.0 and 1.5). The MS4 spectrum of ricinoleate was from RR-OH18:2 in the HPLC fraction #17 of castor oil of our recent report [6]. All of these spectra appeared to be identical. In our experimental conditions, the MS4 spectra of FA from TAG and the MS2 spectra of FA standards can be compared.

We have proposed the structures of hydroxy FA in the seed oil of Physaria fendleri (comparing them to those of castor oil) as: OH1218:29,14 (new FA, OH1218:29,13 in castor oil), OH1218:39,14,16 (new FA, OH18:3 not detected in castor oil), auricolic acid OH1420:211,17 (OH20:2 not detected in castor oil, OH1218:29,13 in castor oil), diOH11,1218:19 (same as in castor), diOH12,1318:29,14 (new FA, diOH11,1218:29,13 in castor oil) and diOH13,1420:111 (new FA, diOH20:1 not detected in castor oil, diOH11,1218:19 in castor oil). Trihydroxy FA were identified in castor oil [19, 20] but not in the current study of lesquerella oil. There were differences in the locations of double bonds and hydroxyl groups on the C18 FA in lesquerella and castor oils. The newly discovered FA with 9,14-dienes in lesquerella oil was not detected in castor oil. The FA with 9,13-dienes in castor oil was not detected in lesquerella oil. If the di- and trihydroxy FA were the autoxidation products, the structures of the hydroxy FA in both oils must be the same. The differences indicated that the polyhydroxy FA were biosynthesized in both lesquerella and castor, and were not the results of autoxidation products.

The present proposed structures of three different 9,14 diene FA can be tentative structures. Some future work to support position 14 of the 2nd double bond would be welcome. It would be interesting to see if (silylated) FA methyl esters of the OH-9,14-dienes could be separated from their isomeric 9,13-dienes, with capillary gas chromatography, to prove that they are indeed coexisting separate molecules, i.e., coexisting prior to MS.

Many lesquerella FA appear to be chain elongation products of castor FA. However, Ricinus and Lesquerella belong to two totally different plant families with totally different evolutionary history. It may therefore not be surprising, that the structures of FA and the biosynthetic pathways may also be different. Five minor epoxy- and hydroxy-FA reported earlier [23] were not detected in this study as well as other studies [710]. Perhaps because the sample in this earlier study [23] was exposed to higher temperature than in other studies [710].

Copyright information

© AOCS (Outside the USA) 2013