Determination of characteristic compound in Manuka honey. by automatic on-line solid phase extraction-liquid

Transcription

1 October 2017 Chinese Journal of Chromatography Vol. 35 No Research article DOI: /SP.J Determination of characteristic compound in Manuka honey by automatic on-line solid phase extraction-liquid chromatography-high resolution mass spectrometry SHEN Chongyu 1#, GUO Siyan 2#, DING Tao 1*, LIU Yun 1, CHEN Lei 1, FEI Xiaoqing 1, ZHANG Rui 1, WU Bin 1, SHEN Weijian 1, CHEN Lei 1, ZHANG Feng 3, FENG Feng 3, DENG Xiaojun 4, YI Xionghai 4, YANG Gongjun 2, CHEN Guoqiang 5 (1. Animal, Plant and Food Inspection Center, Jiangsu Entry-Exit Inspection and Quarantine Bureau, Nanjing , China; 2. College of Pharmacy, China Pharmaceutical University, Nanjing , China; 3. Chinese Academy of Inspection and Quarantine, Beijing , China; 4. Shanghai Entry-Exit Inspection and Quarantine Bureau, Shanghai , China; 5. Suzhou Entry-Exit Inspection and Quarantine Bureau, Suzhou , China) Abstract: A method for the determination of characteristic compound 3,5-dime thoxybenzoate- 4-diglucoside (leptosperin) in Manuka honey was developed by automatic on-line solid phase extraction-liquid chromatography-high resolution mass spectrometry (SPE-LC-HRMS). The samples were separated on a Dikma Diamonsil Plus C18 column (150 mmx4. 6 mm, 5 xm) using the mobile phases of 0. 1% (v/v) formic acid aqueous solution and acetonitrile with gradient elution. The compound was detected with negative electrospray ionization (ESI-) in Target-MS2 mode. The results showed that the linear range was mg/l, the correlation coefficient was The limit of detection (LOD, S/N 5= 3) and limit of quantification (LOQ, S/N5= 10) of the method was 3 mg/kg and 10 mg/kg, respectively. The recoveries at the spiked levels of 50. 0, , mg/kg (10. 0, 20. 0, mg/kg in black locust samples) were in the range of 82. 0%-95. 2% with the relative standard deviations ranging from 2. 7% to 9. 7% (n = 6). The proposed method was applied to 95 mature honey samples from hives in New Zealand including 12 different kinds and 50 commercial honey

2 samples from four different countries. The method is fast, sensitive and accurate to provide technical support to solve the judgment of the Manuka honey imported from New Zealand. Key words: automatic on-line solid phase extraction-liquid chromatography-high resolution mass spectrometry (SPE-LC-HRMS); characteristic compound; 3,5-dimethoxybenzoate-4- diglucoside (leptosperin); Manuka honey Manuscript received on: 10 July 2017 # Joint first authors *Corresponding author: dingt@jsciq.gov.cn Foundation item: National Key Research and Development Plan (No. 2017YFF ); Shanghai Yangtze River Delta Science and Technology Cooperation Project (No ).

3 Manuka honey is a kind of precious honey peculiar to New Zealand. It originates from the nectar of leptosermum scoparium, a unique shrub in New Zealand. Manuka honey can be distinguished from other kinds of honey in its strong and unique antibacterial activity [1]. Honey features antibacterial activity in general due to its high permeability, strong acidity and contained peroxides. However, the antibacterial activity of Manuka honey does not rely on peroxides, so it is called non-peroxide anti-bacterial activity (NPA) [2]. The excellent antibacterial activity of Manuka honey can be utilised to cure wound infection [3] and to help wound healing [4]. Thus, Manuka honey is quite valuable for medicinal purposes [5]. In the recent 10 years, the export rate of Manuka honey of New Zealand has been growing at 10% annually. In 2013, the total value of honey exported from New Zealand amounted to 128 million New Zealand dollars. According to the statistical data from a New Zealand beekeepers association, about 1,700-2,000 tons of Manuka honey are produced in New Zealand each year. However, 10,000 tons of honey are sold as "Manuka honey" each year around the globe [6]. Using unique Manuka factor (UMF) to grade the activity of Manuka honey is faced with the problems of unstable measurement and great errors [7]. Manuka honey is often illegally adulterated with artificial methylglyoxal (MGO) or dihydroxyacetone (DHA) to increase the activity and make exorbitant profits. Consequently, MGO-based labelling of Manuka honey activity has become less reliable [8, 9]. In 2016, the Unique Manuka Factor Honey Association (UMFHA) introduced a new grading system based on the new characteristic compound of Manuka honey - 3,5-dimethoxybenzoate-4- diglucoside (leptosperin) - found by Kato et. al [10]. UMFHA announced this introduction on its website ( 3,5- dimethoxybenzoate-4-diglucoside is a glycosidic substance transformed from methyl syringate. With good thermal stability, it can be stored for a long time [11]. Previous research involved detecting characteristic compound using the high performance liquid chromatography - UV-detector (HPLC-UV), high performance liquid chromatography - photodiode array detector (HPLC- PDAD), liquid chromatography - tandem mass spectrometry (LC-MS/MS), enzyme immunoassay, immunochromatography, fluorescent spectrometry, and so on [12-16]. Among them, the mass spectrometry features higher sensitivity and selectivity. In this article, a method to rapidly and precisely determine the new characteristic compound in Manuka

4 1 The experiment 1.1 Instruments, reagents and materials SYMBIOSIS SPH1240 liquid chromatography system (AB Sciex Pte Ltd., US), with an online solid-phase extractor, 08xx Oasis HLB solid-phase extraction column (Waters Corporation, US); Q Exactive high-resolution mass spectrometer (Thermo Fisher Scientific, US); ELGA-Q15 high-purity water generator (ELGA LabWater, UK). honey: 3,5-dimethoxybenzoate-4- diglucoside, was developed by automatic on-line solid-phase extraction - liquid chromatography - high resolution mass spectrometry (SPE-LC-HRMS). This article also tested 95 samples of mature raw honey from beehives in New Zealand and 50 commercial honey products from four different countries. Standard 3,5-dimethoxybenzoate-4- diglucoside (purity > 99%) was purchased from Alfa Aesar Fine Chemicals, US; acetonitrile (HPLC grade) was purchased from Merck & Co., Germany; formic acid (HPLC grade) was purchased from Tedia Company, US; and also 0.45µm nylon microporous filtering film (Tianjin Jinteng Experimental Equipment Co., Ltd.). The mature raw honey from beehives were collected from a range areas of New Zealand. The commercial honey products were purchased from major supermarkets. 1.2 Preparing the standard solution We precisely weighed 10mg 3,5- dimethoxybenzoate-4-diglucoside and put it into a 10ml volumetric flask. We added acetonitrile into the flask until it reached the constant volume, so as to prepare the standard reserve solution with the mass concentration of 1.0g/L. We stored the solution in a refrigerator at 4. We diluted the above standard reserve solution progressively with acetonitrile-water (1:9, v/v) to make a series of standard solutions with the mass concentration ranging from 0.5 to 100.0mg/L. 1.3 Pre-treatment of samples We precisely weighed 1.0g honey sample and put it into a 50ml lidded Teflon centrifuge tube. We added water until it reached the volume of 10ml, and homogenised the solution by vortical vibration. We took the top 1ml supernatant liquid to go through the 0.45µm nylon microporous filtering film for instrumental analysis. 1.4 Conditions for analysis Online solid-phase extraction conditions

5 We activated the solid-phase extraction column with 1000µL methanol at the flow rate of 5000µL/min. And then we neutralised it with 1000µL water at the flow rate of 5000µL/min. After washing the sample with 500µL water at the flow rate of 250µL/min, we washed the solidphase extraction column with 500µL water at the flow rate of 1000µL/min. Then we used 300µL methanol to elude with high pressure for dispenser (HPD) under the concentrated elusion mode at the flow rate of 150µL/min. capillary tube : 320 ; flow rate of sheath gas (N 2 ): 40L/min; flow rate of auxiliary gas (N 2 ): 10L/min; flow rate of purging gas (N 2 ): 3L/min; spraying voltage: 3kV; Target MS 2 resolution (R): 17500; max retention time: 100ms; separation window: m/z 2.0; automatic gain control (AGC): 2 x The other mass spectrometric parameters of the characteristic compound are set out in Table Chromatographic conditions Chromatographic column: Dikma Diamonsil Plus C 18 chromatographic column (150 mm x 4.6 mm, 5µm); column temperature: ambient temperature; mobile phase: A is 0.1% (v/v) formic acid solution, and B is acetonitrile. The procedure of gradient elution went as: min, 5% B; min, 5% B-15% B; min, 15% B-70% B; min, 70% B-98% B; min, 98% B; 6.50 to 7.00 min, 98% B-5% B; min, 5% B. Flow rate: 0.8 ml/min; sample injection size: 10µL Mass spectrometric conditions Ion source: electrospray ionization (ESI - ), negative ion mode; testing mode: Target- MS2 mode; heating temperature for the 2 Results and discussion 2.1 Optimisation of automatic online solid-phase extraction conditions Manuka honey contains UMF, methyl glycine and many other active ingredients. Sample injection analysis via direct dilution of Manuka honey found that there were matrix-induced effects and the recoveries were in the range of 120% to 150%. In addition, it was not significantly effective to reduce the matrix-induced effect through changing chromatographic column and adjusting the gradient elution procedure, so we tried solid-phase extraction to purify the samples. The characteristic compound of Manuka honey - 3,5-dimethoxybenzoate-

6 4-diglucoside - is a glucoside which contains polyhydroxy and benzoate, with strong polarity as a whole. Therefore, we selected Oasis HLB solid-phase extraction column with stuffing of lipophilic divinylbenzene and hydrophilic N-vinyl pyrrolidone for online purification of the samples. It was found in the experiment that, due to the high dissolubility of the samples, flow rates had a considerable impact on the retention of the characteristic compound on the solid-phase extraction column during washing. Therefore, we examined the impact of different flow rates ( µL/min) on the recoveries of the characteristic compound in Manuka honey (see Fig. 1). The background level of the compound was 994mg/kg. It can be seen from Figure 1 that, for the washing volume of 500µL, when the flow rate was 1200µL/min, the recovery of the characteristic compound in the samples declined by 10%; when the flow rate was 1500µL/min, the recovery of the characteristic compound was only 70%. It was speculated that there was penetration due to excessively high flow rate. Upon comprehensive consideration on the analysis of efficiency and recovery, the flow rate was finally determined to be 1000µL/min. Under the optimised conditions of online solid-phase extraction, the matrix-induced effect of Manuka honey was effectively controlled, and the recovery rate declined to 90%- 95%. 2.2 Optimisation of chromatographic and mass spectrometric conditions The experimental analysis compared the separation of the target compound when the mobile phase was water-acetonitrile, 5mmol/L ammonium acetate solution - acetonitrile, 0.1% (v/v) formic acid solution-acetonitrile, and 0.1% (v/v) formic acid solution-acetonitrile containing 5mmol/L ammonium acetate. The results show that in the ESI - mode, adding a certain amount of ammonium acetate to the mobile phase could extend the retention time. However, there was tailing and asymmetric phenomena in the chromatographic peak of the target compound. After adding a small amount of formic acid into the water mobile phase, the retention time of the target

7 compound would be earlier, but the peak curve was better. It was finally determined to use 0.1% (v/v) formic acid solution-acetonitrile as the mobile phase. We directly injected the standard solution of the target compound into the mass spectrometer for full scan (m/z ) via flow injection analysis (FIA) method. Then the primary excimer ion peak [M- H] - with m/z at was presented. Through the optimization of spray voltage, collision energy and spray gas flow rate, it was decided to select the secondary daughter ion fragments with the strongest signal strength as the qualitative and quantitative ions, with m/z at and respectively. Figure 2 presents the secondary ion chromatograms of the target compound standard solution (10mg/L). 2.3 Linear range, limit of detection and limit of quantity After preparing the standard solutions with the mass concentration ranging from 0.5 to 100.0mg/L, we injected samples following optimised methods. Linear regression can be done using the peak area of the compound as the vertical ordinate (y) and corresponding content as the horizontal ordinate (x, mg/l). The results show that the linear equation is y = 5.15 x 10 6 x x 10 3, correlation coefficient (R 2 ) is This means that 3,5-dimethoxybenzoate-4- diglucoside has a good linear relationship in the range of mg/L. The limit of detection (LOD, S/N 3) and limit of quantity (LOQ, S/N 10) are 3 mg/kg and 10 mg/kg respectively Recoveries and precisions Low-, medium- and high-level standard 3,5-dimethoxybenzoate-4-diglucoside was added to 3 different kinds of honey samples (Manuka and Kanuka honey samples were added at 50.0, and mg/kg, and acacia honey was added at 10.0, 20.0 and 50.0 mg/ml). Six parallel samples were prepared for each level, using external standard methods for quantification, to carry out a spikeand-recovery experiment (see Table 2). The results show that the spiked recovery

8 of 3,5-dimethoxybenzoate-4-diglucoside was 82.0%-95.2%, and the precision was 2.7%-9.7%. The chromatogram of the spiked acacia honey sample (10.0mg/kg) is presented in Fig Analysis of actual samples It is stipulated by UMFHA that honey with the content of 3,5- dimethoxybenzoate-4-diglucoside 100mg/kg is Manuka honey, or it is not Manuka honey. We tested the following honey (see Table 3): 95 samples of mature raw honey from beehives across New Zealand of 12 different kinds including Manuka, Kanuka, Rewarewa, Pohutukawa, honeydew honey, Tawari, clover honey, Rita, Kamahi, borage honey, Taku and an unknown, as well as 50 commerical honey products from four countries - New Zealand, China, Australia and Malaysia (including 39 samples of various-floral honey, Manuka honey, wild-flower honey, honeysuckle honey, Pohutukawa honey and children's honey from New Zealand; 4 samples of jujube honey, motherwort honey, codonopsis pilosula honey and astragalus honey from China; 6 samples of red eucalyptus honey, Chinese photinia honey, eucalyptus honey and desert gum honey from Australia; 1 sample of acacia honey from Malaysia). The results show that, out of the 39 mature raw honey samples indicated as single-floral Manuka honey, 31 (79.5%) have the characteristic compound level over 100mg/kg, specifically mg/kg, while the other 8 (20.5%) contain the compound at 50-95mg/kg. Out of the 8 mature raw honey samples indicated as multi-floral Manuka honey, 3 (37.5%) have the characteristic compound level above 100mg/kg, while the other 5 (62.5%) contain the compound at 25-96mg/kg. Out of the 48 kinds of hive-

9 sourced non-manuka raw honey, 2 samples (4.2% ) - 1 Pohutukawa honey and another unknown honey - have the characteristic compound level over 100mg/kg, and the level of the other 46 (95.8%) is lower than 100mg/kg. Out of the 20 commercial products of singlefloral Manuka honey, 15 samples (75.0%) have the characteristic compound level ranging from 136 to 311 mg/kg, and the level of other 5 (25.0%) is lower than 100mg/kg, ranging from 12 to 96mg/kg. Out of the 2 commercial products of multi-floral Manuka honey, 1 (50.0%) has the above level lower that 100mg/kg. The level of the compound in the non- Manuka honey is 0-59mg/kg, under the 100mg/kg threshold in all the 28 samples. The results show that, according to the UMFHA criterion, 25% of the commercial pure Manuka honey products will be judged as non-manuka honey. Among the hive-sourced Manuka raw honey indicated as single-flora and multifloral, some samples were found to have the characteristic compound level lower than the 100mg/kg threshold, and some samples in the non-manuka raw honey were found to contain the compound at a level above 100mg/kg. This might be because that there are other honey plants in the nectar-sourced areas, and bees collect nectar from random plants. 3 Conclusion In this article, a method for the determination of the characteristic compound in honey samples: 3,5- dimethoxybenzoate-4-diglucoside, was developed by automatic on-line solidphase extraction - liquid chromatography - high resolution mass spectrometry. This method was fast and simple, providing technical support for the determination and verification of Manuka honey imported from New Zealand. References:

10 [1] Zhang Y Z, Hu F L. Journal of Bee, 2015, 35(7): 1 [2] Su L. Food Industry, 2015(8): 48 [3] Willix D J, Molan P C, Harfoot C G. J Appl Microbio, 1992, 73(5): 388 [4] Visavadia B G, Honeysett J, Danford M H. Brit J Oral Max Surg, 2008, 46(1): 55 [5] Chambers J. Palliat Med, 2006, 20(5); 557 [6] China Chamber of Commerce for Import and Export of Food and Beverage. Journal of Bee, 2015, 35(4): 6 [7] Chen L, Fei X Q, Zhang X Y, et al. Journal of Food Safety and Quality, 2014, 5( 10) : 2967 [8] Chen L, Luan J, Fei X Q, et al. Chinese Journal of Chromatography, 2014, 32(2): 189 [9] Adams C J, Manley-Harris M, Molan P C. Carbohyd Res, 2009, 344(8): 1050 [10] Kato Y, Fiyinaka R, Ishisaka A, et al. J Agric Food Chem, 2014, 62(27): 6400 [11] Bong J, Prijic G, Braggins T J, et al. Food Chem, 2017, 214: 102 [12] Kato Y, Umeda N, Maeda A, et al. J Agric Food Chem, 2012, 60(13) : 3418 [13] Kato Y, Araki Y, Juri M, et al. J Agric Food Chem, 2014, 62(44): [14] Kato Y, Araki Y, Juri M, et al. Food Chem, 2016, 194(3): 362 [15] Stephens J M, Loomes K M, Braggins T J, et al. Fluorescence: A Novel Method for Determining Manuka Honey Floral Purity. ( ). intechopen. com/pdfs-wm/53209.pdf [16] Zhang W W, Wen Y Q, Zhao W, et al. Quality and Safety of Agro-Products, 2017(3): 28 No. 10 SHEN Chongyu et. al Determination of characteristic compound in Manuka honey by automatic on-line solid phase extraction-liquid chromatography-high resolution mass spectrometry Chinse Journal of Chromatography Vol. 35 No. 10 SHEN Chongyu et. al Determination of characteristic compound in Manuka honey by automatic on-line solid phase extraction-liquid chromatography-high resolution mass spectrometry Chinse Journal of Chromatography Vol. 35