Simultaneous Quantification of Mixed-Acid Triacylglycerol Positional Isomers and Enantiomers in Palm Oil and Lard by Chiral High-Performance Liquid Chromatography Coupled with Mass Spectrometry

Palm oil and lard are edible fats which are rich in palmitic (P) and oleic acids (O). In this study, triacylglycerol (TAG) positional isomers (symmetric and asymmetric isomers) and enantiomers (asymmetric isomers) in palm oil and lard were quantified simultaneously by using liquid chromatography/mass spectrometry. The CHIRALPAK IF-3 column used in our previous study recognized the difference of TAG isomers consisting of P and O in palm oil and lard, separated sn-OPP/sn-PPO/sn-POP and sn-OPO/sn-OOP/sn-POO into each isomer peak, and enabled the quantification of these TAG isomers with good recovery (95–120%). Although sn-POP and sn-OPO were the major TAGs in palm oil and lard, a comparison of the abundance ratios of TAG enantiomers such as sn-PPO/sn-OPP and sn-OOP/sn-POO revealed that there were slightly more TAG enantiomers with O at the sn-1 position and P at the sn-3 position in palm oil and P at the sn-1 position and O at the sn-3 position in lard. These results were consistent with previous reports for the positional distribution of fatty acids of palm oil and lard. This is the first study that has enabled all TAG isomers consisting of P and O in natural oils and fats to be individually quantified by mass spectrometry.


Introduction
Oils and fats originated from plants and animals are widely used in a variety of industries. Triacylglycerol (TAG) is a main component of edible oils and fats, and their physical and nutritional properties depend on TAG compositions. TAG is composed of one glycerol and three fatty acids [1] and not only their composition, but also binding positions on the glycerol backbone, affect their properties. TAG compositions are specific to each oil and fat species. The binding positions of fatty acids on the glycerol backbone of TAG are distinguished as sn-1, 2 and 3, as shown in Figure 1 [2]. The possible permutations of three fatty acids with repetition on the glycerol backbone of TAG composed of two kinds of fatty acids, A and B, are AAA, sn-AAB, sn-ABA, sn-ABB, sn-BAA, sn-BAB, sn-BBA, and BBB. 'sn-' is a prefix meaning stereospecific numbering. When 'sn-' is prefixed to abbreviated TAG (ex. AAB), sn-AAB means the TAG molecule binding two As and B at the sn-1, 2 and 3 positions in this order. Therefore, sn-AAB and sn-BAA are in an enantiomeric relationship. sn-ABA is a positional isomer for sn-AAB (or sn-BAA). AAB without 'sn-' means TAG including two As and one B without considering the binding positions.
Symmetry 2020, 12, x FOR PEER REVIEW 2 of 7 The possible permutations of three fatty acids with repetition on the glycerol backbone of TAG composed of two kinds of fatty acids, A and B, are AAA, sn-AAB, sn-ABA, sn-ABB, sn-BAA, sn-BAB, sn-BBA, and BBB. 'sn-' is a prefix meaning stereospecific numbering. When 'sn-' is prefixed to abbreviated TAG (ex. AAB), sn-AAB means the TAG molecule binding two As and B at the sn-1, 2 and 3 positions in this order. Therefore, sn-AAB and sn-BAA are in an enantiomeric relationship. sn-ABA is a positional isomer for sn-AAB (or sn-BAA). AAB without 'sn-' means TAG including two As and one B without considering the binding positions. Palm oil is the most heavily produced edible plant oil in the world [3], and is used in many applications, including food in the world. The major fatty acids consisting of palm oil is palmitic acid (P) and oleic acid (O), which are representative saturated and unsaturated fatty acids that are widespread in nature. Lard is also composed of P and O and used as edible fat. However, palm oil and lard have been reported to have different binding positions on the glycerol backbone for P and O. The positional distributions of fatty acids on the glycerol backbone of fats and oils analyzed so far indicate that palm oil has more P in the sn-1,3 position and more O in the sn-2 position, and lard has more O in the sn-1,3 positions and more P in the sn-2 position [4,5]. In fact, palm oil is known to contain high levels of sn-POP [6], and lard is known to contain high levels of sn-OPO [7]. Therefore, these two fats and oils were sometimes compared in terms of nutrition [5,8]. In addition, adulteration of palm oil with lard can be a problem [9][10][11] because both palm oil and lard are used as semi-solid fat in food use. For example, differential scanning calorimetry (DSC) and near-infrared (NIR) spectroscopy were used to detect the adulteration of oils and fats with different physical properties such as lard and palm oil. This difference is due to the unique TAGs of palm oil and lard, which can be measured to distinguish between different types of fats and oils. Silver-ion high performance liquid chromatography (HPLC), which can separate TAG positional isomers, enables us to determine symmetric TAGs such as sn-POP and sn-OPO in distinction from asymmetric TAGs.
In 2011, we developed a reversed phase (RP) HPLC method using Sunrise C28 column (ChromaNik Technologies Inc., Osaka, Japan) [12] and a chiral HPLC method using CHIRALCEL OD-RH column (Daicel Corporation, Osaka, Japan) [13], which enable to separate TAG positional isomers and enantiomers, respectively. Sunrise C28 column separated TAG positional isomers containing two saturated and one unsaturated fatty acids such as sn-POP/sn-PPO (or sn-OPP). On the other hand, CHIRALCEL OD-RH column recognized the difference between saturated and unsaturated fatty acids at the sn-1 and 3 positions of the pair of TAG enantiomers; sn-PPO/sn-OPP, sn-OOP/sn-POO, and sn-PPL/sn-LPP and separated these pairs into each enantiomer peak for the first time. Moreover, the combination of the RP HPLC and chiral HPLC enabled the quantification of naturally occurring TAG positional isomers and enantiomers [14][15][16]. However, the use of two kinds of separation modes made the analysis procedure complicated and the analysis time very long. For this reason, we have developed a new chiral HPLC method using CHIRALPAK IF-3 column (Daicel Corporation) which can simultaneously separate TAG positional isomers and enantiomers such as sn-OPP/sn-PPO/sn-POP and sn-OPO/sn-OOP/sn-POO [17]. In this study, we applied this chiral HPLC with mass spectrometry (MS) to the simultaneous quantification of naturally occurring TAG isomers, sn-OPP/sn-PPO/sn-POP and sn-OPO/sn-OOP/sn-POO in palm oil and lard. No study has ever separated all these TAG positional and enantiomers in natural oils and fats by chiral HPLC and quantified them individually by MS. Palm oil is the most heavily produced edible plant oil in the world [3], and is used in many applications, including food in the world. The major fatty acids consisting of palm oil is palmitic acid (P) and oleic acid (O), which are representative saturated and unsaturated fatty acids that are widespread in nature. Lard is also composed of P and O and used as edible fat. However, palm oil and lard have been reported to have different binding positions on the glycerol backbone for P and O. The positional distributions of fatty acids on the glycerol backbone of fats and oils analyzed so far indicate that palm oil has more P in the sn-1,3 position and more O in the sn-2 position, and lard has more O in the sn-1,3 positions and more P in the sn-2 position [4,5]. In fact, palm oil is known to contain high levels of sn-POP [6], and lard is known to contain high levels of sn-OPO [7]. Therefore, these two fats and oils were sometimes compared in terms of nutrition [5,8]. In addition, adulteration of palm oil with lard can be a problem [9][10][11] because both palm oil and lard are used as semi-solid fat in food use. For example, differential scanning calorimetry (DSC) and near-infrared (NIR) spectroscopy were used to detect the adulteration of oils and fats with different physical properties such as lard and palm oil. This difference is due to the unique TAGs of palm oil and lard, which can be measured to distinguish between different types of fats and oils. Silver-ion high performance liquid chromatography (HPLC), which can separate TAG positional isomers, enables us to determine symmetric TAGs such as sn-POP and sn-OPO in distinction from asymmetric TAGs.
In 2011, we developed a reversed phase (RP) HPLC method using Sunrise C28 column (ChromaNik Technologies Inc., Osaka, Japan) [12] and a chiral HPLC method using CHIRALCEL OD-RH column (Daicel Corporation, Osaka, Japan) [13], which enable to separate TAG positional isomers and enantiomers, respectively. Sunrise C28 column separated TAG positional isomers containing two saturated and one unsaturated fatty acids such as sn-POP/sn-PPO (or sn-OPP). On the other hand, CHIRALCEL OD-RH column recognized the difference between saturated and unsaturated fatty acids at the sn-1 and 3 positions of the pair of TAG enantiomers; sn-PPO/sn-OPP, sn-OOP/sn-POO, and sn-PPL/sn-LPP and separated these pairs into each enantiomer peak for the first time. Moreover, the combination of the RP HPLC and chiral HPLC enabled the quantification of naturally occurring TAG positional isomers and enantiomers [14][15][16]. However, the use of two kinds of separation modes made the analysis procedure complicated and the analysis time very long. For this reason, we have developed a new chiral HPLC method using CHIRALPAK IF-3 column (Daicel Corporation) which can simultaneously separate TAG positional isomers and enantiomers such as sn-OPP/sn-PPO/sn-POP and sn-OPO/sn-OOP/sn-POO [17]. In this study, we applied this chiral HPLC with mass spectrometry (MS) to the simultaneous quantification of naturally occurring TAG isomers, sn-OPP/sn-PPO/sn-POP and sn-OPO/sn-OOP/sn-POO in palm oil and lard. No study has ever separated all these TAG positional and enantiomers in natural oils and fats by chiral HPLC and quantified them individually by MS.

Preparation of Standard Solutions
Twenty-five milligrams of sn-POP and 50 mg of rac-PPO ('rac-' means racemic mixture, regarded as 25 mg of sn-PPO and 25 mg of sn-OPP) were dissolved in acetone and their concentrations were adjusted to 1000 µg/mL using a 25 mL-measuring flask. Then, 2.5 mL of 1000 µg/mL standard solution was diluted with acetonitrile and its concentration was adjusted to 50 µg/mL using a 50 mL-measuring flask. Next, 0.1, 0.2, 1, 2, 3, 4, or 5 mL of 50 µg/mL standard solution and 2 mL of 50 µg/mL internal standard solution (C11C11C11) were diluted with acetonitrile and each of standard's concentration was adjusted to 0.5, 1, 5, 10, 15, 20, or 25 µg/mL with 10 µg/mL of C11C11C11 using a 10 mL-measuring flask. Standard solutions of sn-OPO and rac-OOP were prepared in the same way.

Preparation of Sample Solution
Approximately 40 mg of palm oil (or lard) was weighed and dissolved in acetone and filled up using 25 mL-measuring flask (1600 µg/mL). One milliliter of 1600 µg/mL sample solution was diluted with acetonitrile and filled up using 25 mL measuring flask with 5 mL of 50 µg/mL internal standard (C11C11C11) solution to be adjusted the concentration of the sample solution to 64 µg/mL with 10 µg/mL internal standard. In the case of the recovery test, 5 mL of 50 µg/mL standard solution was spiked to the sample solution.

LC/MS Analysis Using MRM (Multiple Reaction Monitoring) Mode
A setup of LC/MS system used for the quantification is as follows: an HPLC system, Alliance e2695 (Waters Corporation, Milford, MA, USA); a MS with electrospray ionization (ESI) probe, Quattro micro API tandem quadrupole system (Waters Corporation); and a post column pump to promote ionization, LC-10AD (Shimadzu Corporation, Kyoto, Japan). LC/MS conditions were as follows: column, CHIRALPAK IF-3 (2.1 mm i.d. × 250 mm, 3 µm, Daicel Corporation); column temperature, 25 • C; mobile phase, acetonitrile; flow rate, 0.2 mL/min; injection volume, 5 µL; ionization mode, ESI-positive; capillary voltage, 3 kV; source block temperature, 120 • C; desolvation temperature, 450 • C; cone gas flow rate, 50 L/h; desolvation gas flow rate, 800 L/h; cone voltage, 35 V; collision energy, 25 V; data acquisition mode, MRM mode (m/z 850.8 > 577.5 for ammonium adduct of PPO, m/z 876.8 > 577.5 for ammonium adduct of OOP, and m/z 614.7 > 411.5 for ammonium adduct of C11C11C11); and dwell time, 0.1 s. Eluent from HPLC was mixed with 0.l mL/min of 0.1 M ammonium formate methanol solution using a tee-connector in front of ESI probe. The measurements were repeated three times for each standard and sample solution.

Calculation of the Concentration and Recovery of Each TAG Isomer in Palm Oil and Lard
The values of slope (a), intercept (b), and R 2 of calibration curves were calculated from the concentrations of standard solutions and the peak areas of samples divided by internal standard, C11C11C11 (A/A IS ) shown below. The concentrations of each TAG isomer in sample solutions were calculated using Equation (1). Recoveries were calculated by the Equation (2). The concentration of the standard solution spiked into each sample solution was 10 µg/mL for the recovery test. The contents of each TAG in palm oil or lard were obtained by Equation (3).

Results and Discussion
In general, natural fats and oils contain various types of TAGs, and each TAG should exhibit different physical properties. Although fats and oils are aggregates of many TAGs, it is expected that the physical properties of semi-solid fats such as palm oil and lard can be evaluated by examining the isomers of the major TAGs that affect the overall physical properties. In this study, TAG isomers composed of P and O were selected as the analytes for quantitative analysis, because P and O are two fatty acids that are most widely distributed in organisms and are particularly prevalent in palm oil and lard. The TAG isomers of interest here consist of both saturated and unsaturated fatty acids and exhibit a semi-solid state at room temperatures.
For the quantification of these TAG isomers, we selected the MRM mode because of its high selectivity and sensitivity. Chiral HPLC using CHIRALPAK IF-3 column (2.1 mm i.d.) which has the same chiral stationary phase as that we used in our previous study (4.6 mm i.d.) was able to separate sn-OPP/sn-PPO/sn-POP (Figure 2a) and sn-OPO/sn-OOP/sn-POO (Figure 2b) into individual isomers, although the peaks were slightly overlapping. The chromatograms of palm oil and lard spiked with 10 µg/mL of each standard TAG isomer were shown in supplementary data ( Figures S2 and S3). The elution orders of the TAG isomers were confirmed from our previous study, but the retention times were slower overall than our previous ones. This would be due to the difference of the condition of the chiral stationary phase.

Results and Discussion
In general, natural fats and oils contain various types of TAGs, and each TAG should exhibit different physical properties. Although fats and oils are aggregates of many TAGs, it is expected that the physical properties of semi-solid fats such as palm oil and lard can be evaluated by examining the isomers of the major TAGs that affect the overall physical properties. In this study, TAG isomers composed of P and O were selected as the analytes for quantitative analysis, because P and O are two fatty acids that are most widely distributed in organisms and are particularly prevalent in palm oil and lard. The TAG isomers of interest here consist of both saturated and unsaturated fatty acids and exhibit a semi-solid state at room temperatures.
For the quantification of these TAG isomers, we selected the MRM mode because of its high selectivity and sensitivity. Chiral HPLC using CHIRALPAK IF-3 column (2.1 mm i.d.) which has the same chiral stationary phase as that we used in our previous study (4.6 mm i.d.) was able to separate sn-OPP/sn-PPO/sn-POP (Figure 2a) and sn-OPO/sn-OOP/sn-POO (Figure 2b) into individual isomers, although the peaks were slightly overlapping. The chromatograms of palm oil and lard spiked with 10 μg/mL of each standard TAG isomer were shown in supplementary data ( Figures S2 and S3). The elution orders of the TAG isomers were confirmed from our previous study, but the retention times were slower overall than our previous ones. This would be due to the difference of the condition of the chiral stationary phase. Calibration curves for each TAG in Table 1 showed good linearity (R 2 > 0.995) in the range of 0.1-25 μg/mL. The limit of quantification (signal-to-noise ratio >10) of each TAG was approximately 0.05 μg/mL (data not shown). The recovery rates were in the range of 95 to 120% for these TAGs (Tables 2 and 3). The present analytical conditions were considered to be applicable for the quantitative analysis of TAG isomers consisting of P and O of palm oil and lard. However, the relative standard deviations (RSD) of the recovery rates of sn-OPO and sn-OOP in palm oil and sn-OPO in lard were greater than those of other TAG isomers. It was presumably due to insufficient separation of sn-OPO and sn-OOP in the MRM chromatograms of palm oil and lard. In this case, the peak areas Calibration curves for each TAG in Table 1 showed good linearity (R 2 > 0.995) in the range of 0.1-25 µg/mL. The limit of quantification (signal-to-noise ratio >10) of each TAG was approximately 0.05 µg/mL (data not shown). The recovery rates were in the range of 95 to 120% for these TAGs (Tables 2 and 3). The present analytical conditions were considered to be applicable for the quantitative analysis of TAG isomers consisting of P and O of palm oil and lard. However, the relative standard deviations (RSD) of the recovery rates of sn-OPO and sn-OOP in palm oil and sn-OPO in lard were greater than those of other TAG isomers. It was presumably due to insufficient separation of sn-OPO and sn-OOP in the MRM chromatograms of palm oil and lard. In this case, the peak areas were determined by vertically dividing partially overlapping peaks, which is not inherently a correct method. Therefore, even if the peaks on the same MRM chromatogram, the slope of the calibration curve for TAG positional isomers is different. If possible, the column should be extended to improve the separation, but in this study, the data were obtained with this resolution.  The values of TAG contents (wt%) in lard are described as MEAN ± SE (n = 3).
Among these TAGs, sn-POP was the most common in palm oil at 19.0% ( Table 2). The total content of 22.9% for sn-POP, sn-OPP, and sn-PPO was close to the previously reported 24.5% as TAG molecular species, which means the sum of all the isomers [6]. The sn-PPO/sn-OPP and sn-OOP/sn-POO ratios in palm oil were about 1 and 3/2, respectively, which was very close to the sn-OOP/sn-POO ratio measured in our previous study [13]. In the case of lard, 12.8% of OPO was detected in the present analysis ( Table 3). The total content of sn-OPO, sn-OOP, and sn-POO was 19.6%, which was in good agreement with the previously reported value 20.8% [7]. A comparison of the abundance ratios of TAG enantiomers (sn-PPO/sn-OPP and sn-OOP/sn-POO) between palm oil and lard suggested that P tends to be present in the sn-3 position in palm oil, while it is present in the sn-1 position in lard, which is in good agreement with previous reports for the positional distribution of fatty acids of palm oil and lard [4]. These differences of the binding positions of P and O on the glycerol backbone between palm oil and lard would be caused by the substrate selectivity of each acyltransferase in glycerol-3-phosphate pathway [18], which would result in the synthesis of nutritionally and physically suitable TAGs for the energy storage tissues of each organism. In future study, the evaluation of various tissues of organisms by this chiral HPLC method will reveal the characteristic features of such TAG synthesis in detail.

Conclusions
Chiral HPLC on a CHIRALPAK IF-3 column enabled all TAG isomers consisting of P and O in palm oil and lard to be separated and individually quantified by MS. Although the resolution of sn-OPO and sn-OOP needed to be improved, it was confirmed that this analytical method has a high potential for application to other natural lipid analyses.