
==== Front
Heliyon
Heliyon
Heliyon
2405-8440
Elsevier

S2405-8440(24)11864-6
10.1016/j.heliyon.2024.e35833
e35833
Research Article
Quality control of pollen products in the market by quantitative analysis of total amino acids with liquid chromatography
Li Linqiu a1
Zhang Xintong a1
Wang Xin a
Qiu Yuan a
Li Weiqiao a
Guo Lianxian a
Shen Qing sq@zjgsu.edu.cn
b⁎⁎
Dai Juanxiu daijuanxiu@gdmu.edu.cn
a⁎
a Dongguan Key Laboratory of Public Health Laboratory Science, School of Public Health, Guangdong Medical University, Dongguan, China
b Zhejiang Province Joint Key Laboratory of Aquatic Products Processing, Institute of Seafood, Zhejiang Gongshang University, Hangzhou, China
⁎ Corresponding author. daijuanxiu@gdmu.edu.cn
⁎⁎ Corresponding author. sq@zjgsu.edu.cn
1 These authors contributed equally to this work.

06 8 2024
30 8 2024
06 8 2024
10 16 e3583324 12 2023
4 8 2024
5 8 2024
© 2024 The Authors. Published by Elsevier Ltd.
2024

https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
A rapid high-performance liquid chromatography (HPLC) protocol for the determination of amino acids with 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC) derivatization was successfully developed for assessing amino acid levels in six species of representative commercial bee pollen. Based on a poroshell column, a favorable chromatographic separation of seventeen amino acids was achieved in approximately 10 min with satisfactory resolution. The LOD and LOQ of this method were less than 0.034 μg/mL and 0.232 μg/mL, and the intra- and inter-day RSDs ranged between 0.86–5.28 % and 3.21–6.50 %, respectively. The matrix effect (ME) ranged from −8 to 3, implying that the matrix effect was not significant. Under the optimum conditions, the established method was adopted to determine amino acids in six types of bee pollens. The results showed that the total amino acid content ranged from 151.94 mg/g (Rosa rugosa) to 214.52 mg/g (Leonurus artemisia) in the six bee pollen species. Notably, proline (Pro), valine (Val), leucine (Leu), and phenylalanine (Phe) were abundant in the majority of samples. To identify the suspicious samples, principal component analysis (PCA) was performed, and each type of bee pollen was differentiated. Results showed that, in the market, the qualification rate of RR was 100 %, but that of NN was merely 62.5 %, revealing that a few of them were counterfeit. This method offers advantages such as high speed, low cost, and outstanding performance.

Graphical abstract

An efficient HPLC method has been developed for the detection of amino acids in bee pollen. Using the poroshell column and the analysis time was reduced to 10 min. Six types of bee pollen [Leonurus artemisia (LA), Rosa rugosa (RR), Schisandra chinensis (SC), Nelumbo nucifera (NN), Camellia japonica (CJ), Brassica campestris (BC)] was analyzed. RR had a qualification rate of 100 % and NN had a rate of 62.5 %.Image 1

Highlights

• 17 amino acids were separated within 10 min with good resolution.

• Six bee pollens were differentiated, and some were found to be fake.

• This method are high speed, low cost, and excellent performance.

Keywords

Bee pollen
Amino acids
Poroshell column
Liquid chromatography
6-Aminoquinolyl-N-Hydroxysuccinimidyl-carbamate
==== Body
pmc1 Introduction

Bee pollens, comprising pollen from various plants, refer to the male reproductive cells of flower stamens gathered by honeybees [1]. Pollen preparations are distributed worldwide for dietary purposes and as a diet supplement. Commercial bee pollen in the Chinese market is popularly consumed directly as fresh pollen granules. As a healthy apicultural product, bee pollen is considered the “only perfectly complete food”, given that it contains a significant amount of essential amino acids pivotal for humans [2]. Of note, the amino acids in whole dry pollen fluctuate between 10–13 % (26.88 % protein or albuminous substances), which is equivalent to 5–7 times that of beef, milk, eggs, or cheese. Indeed, the nutritional and medicinal value of bee pollen has been established for centuries [3]. Recently, information on the chemical composition and physiological effects of bee pollens has garnered extensive interest, attracting worldwide attention [4,5]. As for the amino acid profile in bee pollen, current research mainly focuses on the free amino acids [6,7]. To better elucidate the nutritional value of this functional food, it is necessary to conduct an analysis of the total amino acids present.

Ascribed to fluctuations in pollen availability and price, adulteration is employed to increase the economic benefits by incorporating low-priced pollen during production and processing. Hence, there is an urgent need to pioneer a robust method to efficiently control the quality of bee pollens. Analyzing the types and contents of amino acids in pollen can offer key nutritional information to the public and manufacturers. However, the absence of specific chromophores, high polarity, and structure diversity of amino acids poses challenges in their identification and characterization [8]. Currently, there exist standardized methodologies for identifying total amino acids in foodstuffs, including the official national standard in China. Nevertheless, this conventional approach tends to be rather time-consuming and involves the utilization of ninhydrin as a derivatization reagent, which is toxic [9]. At present, techniques designed for qualitative and quantitative analysis of amino acids include ion-exchange chromatography [10], gas chromatography [11], capillary electrophoresis [12], and LC coupled with tandem mass spectrometry (MS/MS) [13,14], among which the most popular one involves a LC system coupled to a fluorescence or an ultraviolet detector [[15], [16], [17]]. In this procedure, the hydrophobicity of amino acids is enhanced using chromophores or fluorophores prior to separation on a reversed-phase column. Recently, several attempts have been made to optimize the effect of derivation with regents like 2,4-dinitrofluorobenzene, 9-fluorenylmethyloxycarbonyl chloride, diethylethoxymethylenemalonate, o-phthaldialdehyde, and so on [8,[18], [19], [20]]. However, these derivative methods are either toxic, slow, or produce a large number of by-products. In 1993, Cohen and Michaud first introduced 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC), which can react with both primary and secondary amines in seconds with strong fluorescence intensities with minimal matrix interference [21]. Afterward, efforts have been made by several research groups to enhance and adapt the AQC method to different sample matrices, such as beers, cheeses, fish, juices, and wines [[22], [23], [24], [25], [26]]. However, the method for the determination of amino acids in bee pollen remains considerably limited.

Traditional chromatographic procedures developed for a comprehensive measurement of amino acids require long running times (roughly 30 min) [27]. A sub-2 μm particle-packed column, accompanied by an ultra-high pressure LC (UPLC) system, can offer high peak efficiency, but the UPLC system and the column are not cost-effective for most research labs globally [[28], [29], [30]]. Recently, the use of superficially porous particles (sub-3 μm particle size) has received considerable attention. The distinctive feature of this column is that the particles consist of solid cores surrounded by a porous silica shell measuring about 0.5 μm, allowing for high peak efficiency (even at higher flow rates), low solvent consumption, and short handling time, thereby rendering this method economical and less environmentally hazardous [31,32].

In the present study, a methodology to simultaneously separate and quantify the total amino acids in several types of bee pollens, using an AQC derivation protocol followed by poroshell column-based HPLC analysis with DAD detection, was developed to address this gap. In this work, two types of column were compared, and the elution conditions were optimized to achieve a rapid separation of amino acids of interest. The optimized method was validated with matrix effects (ME), precision, linearity, sensitivity, and accuracy. Finally, the established method was applied to the quantitation of total amino acids in different bee pollens, and with the help of principal component analysis (PCA), the fake products were identified.

2 Materials and methods

2.1 Chemicals and reagents

Internal standard norleucine (Nor) and seventeen amino acid standards including phenylalanine (Phe), leucine (Leu), isoleucine (Ile), methionine (Met), valine (Val), proline (Pro), tyrosine (Tyr), cystine (Cys2), alanine (Ala), threonine (Thr), glycine (Gly), glutamic acid (Glu), serine (Ser), aspartic acid (Asp), arginine (Arg), histidine (His) and lysine (Lys) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Waters AccQ-Tag Chemistry Package consisted of AccQ-Fluor reagent kits containing borate buffer (0.4 mol/L), derivatizing reagent powder 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC), reagent diluents, and AccQ-Tag Ultra Eluent A Concentrate (containing 140 mmol/L sodium acetate, 17 mmol/L triethylamine, 1 mmol/L calcium disodium EDTA, pH 5.05) was supplied by Waters Co. Ltd. (Milford, MA, USA). The acetonitrile was of HPLC grade and procured from Merck (Darmstadt, Germany). Milli-Q water made by a purification system (Millipore, Bedford, MA, USA) was used throughout the experiment.

2.2 Sample preparation

Since the outer package of pollen only indicates the production date without batch number, it is believed that the same production date means the same batch. Six batches of bee pollens (24 samples in each batch), including Leonurus artemisia (LA), Rosa rugosa (RR), Schisandra chinensis (SC), Nelumbo nucifera (NN), Camellia japonica (CJ), and Brassica campestris (BC) were acquired from Sinohive Co. Ltd., Hubei, China, authenticated by Prof. Zhifeng Zhang (School of Pharmacy, Macau University of Science and Technology) (Table 1). An additional twelve suspicious samples, whose outer packagings indicated the sample names as NN or RR, were bought from a market in mainland China and were authenticated using the method developed in this study. The preparation steps involved grinding the bee pollen samples into a fine powder and subsequent screening through a sieve with a particle size of <250 μm prior to extraction.Table 1 Information of pollen samples (n = 24).

Table 1Sample name	Date of manufacture	Specification	Origin	
Leonurus artemisia (LA)	2022/10/9, 2022/10/14, 2022/10/21, 2022/11/1, 2022/11/9, 2022/11/18	230g per package	Shiyan, Hubei, China	
Rosa rugosa (RR)	2022/7/24, 2022/7/28, 2022/8/1, 2022/8/6, 2022/8/25, 2022/9/15	230g per package	Shiyan, Hubei, China	
Schisandra chinensis (SC)	2022/11/1, 2022/11/3, 2022/11/28, 2022/12/3, 2022/12/9, 2022/12/17	230g per package	Shiyan, Hubei, China	
Nelumbo nucifera (NN)	2023/3/25, 2023/4/3, 2023/4/8, 2023/4/13, 2023/4/28, 2023/5/5	230g per package	Shiyan, Hubei, China	
Camellia japonica (CJ)	2022/7/1, 2022/7/21, 2022/7/25, 2022/8/1, 2022/8/5, 2022/8/15	230g per package	Shiyan, Hubei, China	
Brassica campestris (BC)	2022/11/2, 2022/11/19, 2022/12/28, 2023/1/7, 2023/1/16, 2023/1/26	230g per package	Shiyan, Hubei, China	

To obtain total amino acids, the powdered bee pollen samples (1.0000 g) were accurately weighed and mixed with 10 mL of 6 mol/L HCl containing 0.5 % phenol and hydrolysed for 24 h at 110 ± 2 °C under constant nitrogen in flame-sealed pyrex tubes. Afterward, the hydrolysate was successively filtered, and the filtrate was lyophilized and re-dissolved in 20 mL water.

2.3 Amino acid derivatization

Pre-column AQC derivatization was performed as follows: 10 μL of the filtered sample or standard solution was initially transferred to a 1.5 mL amber glass vial using a micropipette (Eppendorf AG, Hamburg, Germany). To generate a proper pH environment (8.2–10.0), 70 μL of borate buffer was introduced, and the solution was briefly vortexed. Then, 20 μL of 10 mmol/L AQC in acetonitrile was added, and the mixture was revortexed before being heated at 55 °C for 10 min in an oven. Prior to injection, samples were filtered through a 0.2 μm regenerated cellulosic membrane (Sartorius, Goettingen, Germany).

2.4 HPLC analysis

The chromatographic setup consisted of an Agilent 1260 Series HPLC system (Agilent Technologies, Palo Alto, CA, USA) equipped with a vacuum degasser unit, autosampler, quaternary pump, column heating assembly, and diode-array detector. Separation of amino acid derivatives was performed on an Agilent poroshell 120 EC-C18 column (50 × 2.1 mm, i. d., 2.7 μm) with a pre-connected 0.2 μm column inline filter. The mobile phases were composed of solvents A and B (60 % acetonitrile). Solvent A was diluted 10 times from AccQ-Tag Eluent A Concentrate. Derivatized amino acid standards or bee pollen samples were injected (2 μL injection volume) onto the column according to the following gradient: initial 8 % B; 0–4 min/8 %–10 % B; 4–8 min/10 %–32 % B; 8–12 min/32 %–45 % B; 12–15 min/45 %–100 % B. The detection wavelength was set at 254 nm, the flow rate at 0.4 mL/min, and the column temperature at 55 °C. Data acquisition was achieved using the Agilent Chemstation software package.

2.5 Statistical analysis

Statistical analysis and calculation of the mean, standard deviation, and level of significance were performed using Microsoft Excel 2007. Principal component analysis (PCA) was employed to determine the primary sources of variability within the datasets and to establish the relationship between samples (objects) and amino acids (variables). Chromatographic data were imported into the statistical software package MarkerView, version 1.2.1 (AB Sciex), to identify differences among the species of amino acids, retention time, and abundance.

3 Results and discussion

3.1 HPLC method optimization

An ideal fast liquid chromatography method aims to minimize separation time while maintaining high column efficiency. In order to obtain a desirable chromatographic profile with satisfactory retention and peak shape, mobile phase, gradient elution conditions, flow rate, and column temperature were investigated.

For mobile phase, solvent A was prepared according to the suggestions of the user manual of Waters AccQ-Tag Chemistry Package, and therefore, no optimization was needed. For solvent B, acetonitrile was selected due to its better elution effect than methanol [30], and its concentration was optimized from 30 %–100 % with 10 % increment. Comprehensive considering the separation efficiency, peak shape and RT, 60 % acetonitrile was selected. With the same criterions (resolution, peak shape and RT), gradient composition and run time were studied, and the ultimate gradient program was applied as follows: initial 8 % B; 0–4 min/8 %–10 % B; 4–8 min/10 %–32 % B; 8–12 min/32 %–45 % B; 12–15 min/45 %–100 % B. Column temperature is a crucial factor as well. When the temperature was changed from 25 °C to 40 °C (5 °C increment), no obvious difference on separation efficiency and RT was observed. However, when the temperature was higher than 40 °C, both separation efficiency and RT were remarkably improved, and a best separation was achieved at 55 °C. After that, further increase in temperature resulted in an unstable baseline. Therefore, 55 °C was applied. Flow rate was investigated from 0.2 mL/min – 0.5 mL/min. Too low flow rate led to a bad separation and long RT, while too high rate caused high pressure in the column as well as a deterioration in separation. Hence, 0.4 mL/min flow rate was preferred.

Under the optimized conditions, a satisfactory separation was achieved in 10 min (Fig. 1), which was apparently more efficient than the reported methods [28–30].Fig. 1 HPLC chromatograms of seventeen amino acids, including 1. Asp, 2. Glu, 3. Ser, 4. His, 5. Gly, 6. Arg, 7. Thr, 8. Ala, 9. Pro, 10. Cys2, 11. Tyr, 12. Val, 13. Met, 14. Lys, 15. Ile, 16. Leu and 17. Phe.

Fig. 1

3.2 Method validation

The result indicated that AQC reacts with both primary and secondary amino acids, and no significant interference was noted in the only major fluorescent by-product, 6-aminoquinoline (AMQ). The proposed method has already been validated in terms of matrix effect, precision (intra- and inter-day repeatability), linearity, sensitivity (limit of detection [LOD] and limit of quantification [LOQ]), and accuracy (recovery) in a previous study [33].

Method linearity was evaluated based on the analysis of the data with five levels of concentration by linear regression with intercepts and a 1/x weighting factor. As summarized in Table 2, the correlation coefficients (R2) of the calibration curves were in the range of 0.9956–0.9999, which reflected acceptable linearity between the concentration (y) and the peak ratio of analytes to the internal standard (x).Table 2 Parameters of calibration curves, linearity, LOD and LOQ.

Table 2Analyte	Calibration curve	R2	Linearity range μg/mL	LOD μg/mL	LOQ μg/mL	
Asp	y = 0.0504x + 0.015	0.9950	0.5–100	0.049	0.466	
Glu	y = 0.0395x + 0.0149	0.9956	2–100	0.113	1.017	
Ser	y = 0.0467x - 0.0029	0.9993	0.5–100	0.038	0.342	
His	y = 0.0901x - 0.006	0.9987	2.5–100	0.189	1.739	
Gly	y = 0.0529x + 0.0005	0.9995	1.5–100	0.077	0.689	
Arg	y = 0.1945x + 0.162	0.9999	1–150	0.054	0.459	
Thr	y = 0.0581x + 0.0094	0.9962	1–100	0.035	0.301	
Ala	y = 0.0747x + 0.0011	0.9993	2–100	0.163	1.483	
Pro	y = 0.0973x - 0.0028	0.9999	2–150	0.078	0.749	
Cys2	y = 0.0801x - 0.0022	0.9999	5–100	0.233	2.144	
Tyr	y = 0.0264x + 0.009	0.9977	2–200	0.134	1.430	
Val	y = 0.0509x + 0.0063	0.9982	1.5–100	0.071	0.568	
Met	y = 0.0271x + 0.026	0.9981	1.5–100	0.072	0.678	
Lys	y = 0.0203x + 0.0195	0.9981	1–100	0.05	0.455	
Ile	y = 0.0603x + 0.014	0.9976	1–200	0.034	0.232	
Leu	y = 0.0598x - 0.0026	0.9990	1.5–200	0.066	0.514	
Phe	y = 0.0802x - 0.0025	0.9993	1–200	0.083	0.705	

Sensitivity was investigated using LOD and LOQ values, which were calculated on the basis of signal-to-noise ratio (s/n = 3 for LOD and s/n = 10 for LOQ) on the chromatography of the standard mixture. The LOD and LOQ were less than 0.034 μg/mL and 0.232 μg/mL (Table 2), signifying adequate sensitivity for conducting the determination process.

The precision of the developed method was determined in terms of intra- and inter-day variations (Table 3). Six replicates of the standard mixtures were analyzed within a day to evaluate intraday variability and were examined in triplicate for three consecutive days to assess inter-day variability. To confirm repeatability, six sample solutions were independently prepared using the previously described method. The overall intra- and inter-day variations (RSDs) of the target analytes ranged between 0.86 % ∼ 5.28 % and 3.21 % ∼ 6.50 %, respectively, showcasing the precision of this method.Table 3 The method precision, recovery and matrix effect for poroshell based HPLC analysis of amino acids from bee pollen samples.

Table 3Analyte	Precision (RSD, %, n = 6)	Recovery (%)	Matrix effect (%)	
Intra-day	Inter-day	Mean	RSD	
Asp	3.48	3.75	94.5	2.12	−3	
Glu	2.89	4.26	91.8	2.31	−5	
Ser	3.82	4.36	96.3	2.08	−2	
His	2.46	6.50	92.5	3.65	0	
Gly	3.25	3.42	101.5	4.87	2	
Arg	2.92	3.21	98.3	1.92	−6	
Thr	5.28	5.32	95.6	2.88	−2	
Ala	2.01	3.44	92.2	4.49	0	
Pro	1.17	4.25	104.3	2.04	−3	
Cys2	3.10	6.30	96.4	2.48	1	
Tyr	1.84	3.22	99.2	2.22	−6	
Val	0.98	4.53	100.3	1.84	−5	
Met	0.86	3.87	94.8	2.43	3	
Lys	2.34	5.27	93.1	1.02	−2	
Ile	1.48	5.51	98.7	2.78	−4	
Leu	1.93	4.38	97.4	4.42	−8	
Phe	1.54	5.89	102.4	3.98	−1	

The accuracy of the method was examined via recovery experiments. Specifically, standards corresponding to high (150 %), middle (100 %), and low (50 %) levels were spiked into 1 g of previously analyzed samples (Table S1), in which the specific addition amount of “high”, “middle” and “low” was determined based on the content of each amino acid in the sample. The spiked samples were then extracted, processed, and quantified in accordance with the aforementioned methods, with the experiments performed in triplicates at each level. The average recoveries were estimated using the formula recovery (%) = [(amount found − original amount)/amount added] × 100. As depicted in Table 3, the relative recoveries of amino acids ranged from 91.8 % to 104.3 % at the three concentration levels, with RSD values lower than 4.87 %, indicating that all the analytes were well recovered. The results were consistent with the accepted recovery values at certain analyte concentration levels recommended by the FDA Guidelines for the validation of Chemical Methods in the Foods Program [34].

While the separation of standard materials is relatively achievable, daily analytical tasks generally involve complex samples. The bee pollen is complicated due to its composition. Specifically, it contains not only proteins and amino acids but also vitamins, trace metals, and carbohydrates. Interferences were eliminated following sample preparation, and the established method was successfully applied to the final extracts acquired from six bee pollen samples. In this work, matrix effect (ME) was adopted to evaluate the background interference during the analysis procedure [35]. ME was defined as the suppression or enhancement of the intensity of analytes and was calculated as follows: ME (%) = (1 – Areaspiked standards in sample/Areastandards in pure solvent) × 100 %, wherein Areaspiked standards in sample = Areatotal amount – Areaoriginal amount. ME = 0 indicated the absence of matrix effect, whereas ME > 0 reflected intensity suppression, and ME < 0 represented enhancement. The higher the absolute ME value, the stronger the matrix interference. The matrix effect was observed in this study. As listed in Table 3, the signals of Gly, Cys2, and Met were enhanced, while the remaining signals were suppressed. The ME ranged from −8 to 3, implying that the matrix effect was not significant. As delineated in the chromatograms in Fig. 2, under the optimum conditions, the seventeen amino acids were well separated. Although the amino acid profiles substantially varied across bee pollen samples, the obtained chromatograms were of high quality, enabling amino acid profiling and quantification throughout samples, irrespective of the bee pollen species.Fig. 2 HPLC chromatograms of amino acids, including 1. Asp, 2. Glu, 3. Ser, 4. His, 5. Gly, 6. Arg, 7. Thr, 8. Ala, 9. Pro, 10. Cys2, 11. Tyr, 12. Val, 13. Met, 14. Lys, 15. Ile, 16. Leu and 17. Phe, in bee pollens.

Fig. 2

3.3 Pollen samples analysis

The content of amino acids in each bee pollen sample was calculated and summarized in Table 4. The values reported in this table were collected using the internal standardization method, wherein amino acid concentrations were correlated to the peak areas of amino acids and internal standards. The results demonstrated that the total amino acid contents of the six bee pollens ranged from 151.94 to 214.52 mg/g, with the highest LA content, followed by SC, CJ, BC, NN, and RR. The primary amino acids were Pro, Leu, and Phe for the LA, SC, and CJ, and Ser and Phe for RR and BC. In addition, NN was significantly different from the other five bee pollen species. For example, the content of Asp was 19.58 mg/g in NN and accounted for approximately 12.07 % of total amino acids, whereas it accounted for approximately 6.10–8.83 % in the other five bee pollen species. Bee pollens are typically rich in amino acids and thus can be used as nutritional substances that can supplement essential life activities and potentially prevent or cure several diseases. Therefore, further studies are warranted to investigate the relationship between amino acids and the pharmacological effects of bee pollens.Table 4 Content of each amino acid in the six bee pollen species (mg/g).

Table 4Analyte	Leonurus artemisia	Rosa rugosa	Schisandra chinensis	Nelumbo nucifera	Camellia japonica	Brassica campestris	
Mean value	SD	Mean value	SD	Mean value	SD	Mean value	SD	Mean value	SD	Mean value	SD	
Asp	15.61	2.00	11.97	1.38	12.93	1.40	19.58	2.33	12.06	1.45	17.09	2.04	
Glu	12.07	1.43	9.03	0.98	10.34	1.29	14.63	1.68	10.59	1.24	13.90	1.58	
Ser	16.60	1.88	14.25	1.57	14.51	1.90	12.08	1.58	14.50	1.62	14.32	1.61	
His	9.08	1.12	8.37	0.87	8.96	1.12	7.44	0.82	9.94	1.14	7.43	0.90	
Gly	8.00	0.95	5.60	0.73	7.06	0.78	6.28	0.70	7.09	0.74	7.44	0.82	
Arg	14.45	1.56	10.78	1.27	14.37	1.59	9.80	1.12	14.29	1.46	11.97	1.40	
Thr	15.48	1.69	11.55	1.39	14.33	1.73	10.58	1.22	14.02	1.55	13.83	1.37	
Ala	9.73	1.07	5.86	0.69	7.75	0.95	7.63	0.90	7.84	0.92	8.85	0.98	
Pro	21.61	2.66	11.76	1.31	27.01	3.01	7.53	0.95	17.24	2.06	13.18	1.59	
Cys2	3.69	0.40	1.99	0.26	3.79	0.49	1.81	0.22	2.52	0.32	3.23	0.38	
Tyr	9.91	1.11	7.25	0.90	10.80	1.14	4.95	0.57	8.83	1.04	7.20	0.85	
Val	16.69	2.19	12.49	1.43	15.05	1.72	14.62	1.66	17.62	1.90	16.59	1.86	
Met	2.89	0.33	1.50	0.18	3.90	0.37	0.50	0.06	3.21	0.38	3.25	0.40	
Lys	7.47	0.82	3.48	0.43	7.51	1.05	8.18	0.90	6.29	0.76	8.57	0.99	
Ile	12.66	1.54	7.82	0.95	12.44	1.39	10.24	1.21	12.51	1.34	12.36	1.57	
Leu	17.81	2.09	12.38	1.32	15.97	1.89	14.86	1.84	17.85	2.04	17.94	2.05	
Phe	20.77	2.40	15.86	1.70	22.54	2.68	11.54	1.35	21.28	2.47	16.48	1.92	
Total	214.52	6.73	151.94	4.41	209.26	6.56	162.25	5.29	197.68	6.18	193.63	6.09	

3.4 Statistical analysis

Five amino acids (Gly, Ala, Pro, Val, Leu) with better chromatographic separation and higher content were selected for analysis of variance (ANOVA), and the results are shown in Fig. 3. Among the five pollens, LA had the highest Gly content, whereas RR had the lowest. Overall, there were minimal differences in Gly content among the various pollens, as shown in Fig. 3A. Likewise, LA exhibited the highest Ala content, which was 66 % greater than the lowest in RR, as depicted in Fig. 3B. As shown in Fig. 3C, there were significant variations in Pro content among the five pollens. Notably, SC had the highest content, at 27.01 mg/g, which was 259 % higher than the lowest content in NN (7.53 mg/g). Except for NN, which had a Val content of 4.95 mg/g, there were minimal differences in Val content among the remaining four pollens (Fig. 3D). The Leu content was similar in LA, CJ, and BC, nearing 18 mg/g (Fig. 3E).Fig. 3 Difference analysis of contents of amino acids in different pollens, including A: Gly, B: Ala, C: Pro, D: Val, and E: Leu; a ∼ f: different alphabet meant significance difference (p < 0.05).

Fig. 3

Principal component analysis (PCA) is a commonly used method to explore relationships among samples, aiming to extract a small number of latent components that summarize the measured data with minimal information loss in a reduced-dimension plot. To evaluate variations among bee pollen samples, PCA was performed using the contents of amino acids from HPLC profiles. The first two principal components (PC1 and PC2), which accounted for a cumulative percent (cum %) of 66.3 of the total variance in the data, were extracted for analysis. Among them, PC1 and PC2 accounted for 44.9 % and 21.4 % of the total variance, respectively. The remaining principal components were neglected, which had a marginal impact on the model. Fig. 4A illustrates the sample scatter plot, wherein each marker represents a bee pollen sample. As can be deduced, the samples were evidently clustered into several domains. NN, RR, and BC were significantly different from the other bee pollen species. LA, CJ, and SC could be partially differentiated from each other, given that some samples overlapped in the PCA plot.Fig. 4 (A). Principal component analysis of the seventeen amino acids in authenticated bee pollen samples. (B). Principal component analysis for fake bee pollen products detection.

Fig. 4

In the market, NN and RR are the two most expensive bee pollens. Additionally, they are usually adulterated with some low-priced bee pollens, such as BC. Rece ntly, adulterated and counterfeit bee pollen has raised concerns, especially for the higher-priced NN, prompting an urgent need for identifying adulteration in bee pollen in the food industry. Therefore, data from the twelve suspicious samples (8 NN and 4 RR) were imported into the PCA plot in order to authenticate the bee pollens sold in the market. As shown in Fig. 4B, all four suspicious RR samples (blue-green points) fell within the cluster of RR, validating their authenticity. On the other hand, only five samples (green points) among the eight NN samples fell within the NN cluster, yielding a qualification rate of 62.5 %. The other three samples (blue points) were identified as counterfeit products, two of which were BC, whilst the remaining one was suspected to be CJ. The results indicated that this method is efficient and assists in the authentication of bee pollens in the market. Meanwhile, strict regulations concerning bee pollen should be developed. The bee pollen industry and government regulatory agencies should take proactive measures to safeguard the safety and quality of bee pollen products.

4 Conclusions

In summary, a fast and efficient HPLC method was developed for the determination of total amino acids in bee pollens. The application of the poroshell column lowered the analysis time to approximately 10 min on traditional HPLC equipment. The method has been fully validated and found to be precise, sensitive, and accurate. Moreover, the contents of total amino acids in six species of bee pollens, including LA, RR, SC, NN, CJ, and BC, were analyzed. After statistical analysis, the two most expensive bee pollen species, NN and RR, were clustered. The sampling survey determined that, in the market, the qualification rate of RR was 100 %, but that of NN was merely 62.5 %. This method holds considerable promise for the quality control and authentication of bee pollens and their related products.

Ethical approval

This article does not contain studies with human participants. All applicable international, national, and/or institutional guidelines for the care and use of animals were followed.

CRediT authorship contribution statement

Linqiu Li: Writing – review & editing, Visualization, Validation, Supervision, Funding acquisition. Xintong Zhang: Writing – original draft, Data curation. Xin Wang: Formal analysis, Data curation. Yuan Qiu: Formal analysis, Data curation. Weiqiao Li: Funding acquisition, Formal analysis. Lianxian Guo: Funding acquisition. Qing Shen: Formal analysis, Project administration. Juanxiu Dai: Conceptualization, Investigation, Project administration, Supervision, Validation, Visualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Abbreviation Used

AQC 6-Aminoquinolyl-N-Hydroxysuccinimidyl-Carbamate

AMQ 6-aminoquinoline;

HPLC High-performance liquid chromatography

LOD Limit of detection

LOQ Limit of quantity

PCA Principal component analysis

LA Leonurus artemisia

RR Rosa rugosa

SC Schisandra chinensis

NN Nelumbo nucifera

CJ Camellia japonica

BC Brassica campestris

Appendix A Supplementary data

The following is the Supplementary data to this article:Multimedia component 1

Multimedia component 1

Acknowledgment

This work was supported by the 10.13039/501100003453 Medical Scientific Research Foundation of Guangdong Province (A2021036 ), 10.13039/501100016077 Dongguan Sci-tech Commissioner Program (20221800500752 , 20221800500762 ), 10.13039/501100003785 Guangdong Medical University Research Startup Foundation (B2017022 ), 10.13039/501100016077 Dongguan Science and Technology of Social Development Program (20231800939972 ).

Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.heliyon.2024.e35833.
==== Refs
References

1 Li D.-F. Han W.-L. Renner S.S. Huang S.-Q. Touch-sensitive stamens enhance pollen dispersal by scaring away visitors Elife 11 2022 e81449 10.7554/eLife.81449
2 Khalifa S.A.M. Elashal M.H. Yosri N. Du M. Musharraf S.G. Nahar L. Sarker S.D. Guo Z. Cao W. Zou X. Abd El-Wahed A.A. Xiao J. Omar H.A. Hegazy M.-E.F. El-Seedi H.R. Bee pollen: current status and therapeutic potential Nutrients 13 2021 1876 10.3390/nu13061876 34072636
3 Giacomini J.J. Leslie J. Tarpy D.R. Palmer-Young E.C. Irwin R.E. Adler L.S. Medicinal value of sunflower pollen against bee pathogens Sci. Rep. 8 2018 14394 10.1038/s41598-018-32681-y
4 Xi X. Li J. Guo S. Li Y. Xu F. Zheng M. Cao H. Cui X. Guo H. Han C. The potential of using bee pollen in cosmetics: a review J. Oleo Sci. 67 2018 1071 1082 10.5650/jos.ess18048 30111679
5 Inacio L.D.J. Merlanti R. Lucatello L. Bisutti V. Contiero B. Serva L. Segato S. Capolongo F. Pyrrolizidine alkaloids in bee pollen identified by LC-MS/MS analysis and colour parameters using multivariate class modeling Heliyon 6 3 2020 e03593 10.1016/j.heliyon.2020.e03593
6 Ares A.M. Martín M.T. Toribio L. Bernal J. Determination of free amino acids in bee pollen by liquid chromatography with fluorescence detection Food Anal. Methods 15 2022 2172 2180 10.1007/s12161-022-02281-8
7 Martín-Gómez B. Salahange L. Tapia J.A. Martín M.T. Ares A.M. Bernal J. Fast chromatographic determination of free amino acids in bee pollen Foods 11 2022 4013 10.3390/foods11244013 36553756
8 Chen L. Chen Q. Zhang Z. Wan X. A novel colorimetric determination of free amino acids content in tea infusions with 2,4-dinitrofluorobenzene J. Food Compos. Anal. 22 2 2009 137 141 10.1016/j.jfca.2008.08.007
9 National Health and Family Planning Commission of the People's Republic of China, China Food and Drug Administration National Standard of the People's Republic of China GB/T 5009.124-2016 National Food Safety Standard - Determination of Amino Acids in Foods 2016 China Standard Press Beijing
10 Scott A.I. Mendelsohn B.A. Le A. Cowan T.M. The antibiotic cefepime interferes with amino acid analysis by ion-exchange chromatography Clin. Chim. Acta 456 2016 149 150 10.1016/j.cca.2016.03.003 26947969
11 Santos I.C. Smuts J. Crawford M.L. Grant R.P. Schug K.A. Large-volume injection gas chromatography-vacuum ultraviolet spectroscopy for the qualitative and quantitative analysis of fatty acids in blood plasma Anal. Chim. Acta 1053 2019 169 177 10.1016/j.aca.2018.12.007 30712563
12 Hartmann A. Murauer A. Ganzera M. Quantitative analysis of mycosporine-like amino acids in marine algae by capillary electrophoresis with diode-array detection J. Pharm. Biomed. Anal. 138 2017 153 157 10.1016/j.jpba.2017.01.053 28213175
13 Zielińska S. Dryś A. Piątczak E. Kolniak-Ostek J. Podgórska M. Oszmiański J. Matkowski A. Effect of LED illumination and amino acid supplementation on phenolic compounds profile in Agastache rugosa in vitro cultures Phytochem. Lett. 31 2019 12 19 10.1016/j.phytol.2019.02.029
14 Tang L. Swezey R.R. Green C.E. Lee M.S. Bunin D.I. Parman T. A tandem liquid chromatography and tandem mass spectrometry (LC/LC–MS/MS) technique to separate and quantify steroid isomers 11β-methyl-19-nortestosterone and testosterone J. Chromatogr. B Biomed. Sci. Appl. 1193 2022 123165 10.1016/j.jchromb.2022.123165
15 Ebada S.S. Ebrahim W. Quinoisobutyride A, an acyclic antibacterial tetrapeptide incorporating an unprecedented heterocyclic amino acid residue from the hypersaline lake-derived fungus Penicillium simplicissimum strain WSH17 Phytochem. Lett. 36 2020 95 98 10.1016/j.phytol.2020.01.022
16 Peng L.-Q. Cao J. Du L.-J. Zhang Q.-D. Shi Y.-T. Xu J.-J. Analysis of phenolic acids by ionic liquid-in-water microemulsion liquid chromatography coupled with ultraviolet and electrochemical detector J. Chromatogr. A 1499 2017 132 139 10.1016/j.chroma.2017.03.086 28390669
17 Duca Z.A. Craft K.L. Cable M.L. Stockton A.M. Quantitative and compositional analysis of trace amino acids in icy moon analogues using a microcapillary electrophoresis laser-induced fluorescence detection system ACS Earth Space Chem. 6 2022 333 345 10.1021/acsearthspacechem.1c00297
18 Li L.Q. Baibado J.T. Shen Q. Cheung H.Y. Determination of the authenticity of plastron-derived functional foods based on amino acid profiles analysed by MEKC J. Chromatogr. B 1070 2017 23 30 10.1016/j.jchromb.2017.10.036
19 Pereira V. Pontes M. Câmara J.S. Marques J.C. Simultaneous analysis of free amino acids and biogenic amines in honey and wine samples using in loop orthophthalaldeyde derivatization procedure J. Chromatogr., A 1189 2008 435 443 10.1016/j.chroma.2007.12.014 18164022
20 Gómez-Alonso S. Hermosín-Gutiérrez I. García-Romero E. Simultaneous HPLC analysis of biogenic amines, amino acids, and ammonium ion as aminoenone derivatives in wine and beer samples J. Agric. Food Chem. 55 2007 608 613 10.1021/jf062820m 17263449
21 Cohen S.A. Michaud D.P. Synthesis of a fluorescent derivatizing reagent, 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate, and its application for the analysis of hydrolysate amino acids via high-performance liquid chromatography Anal. Biochem. 211 1993 279 287 10.1006/abio.1993.1270 8317704
22 Kabelová I. Dvořáková M. Čížková H. Dostálek P. Melzoch K. Determination of free amino acids in beers: a comparison of Czech and foreign brands J. Food Compos. Anal. 21 2008 736 741 10.1016/j.jfca.2008.06.007
23 Kabelová I. Dvořáková M. Čížková H. Dostálek P. Melzoch K. Determination of free amino acids in cheeses from the Czech market Czech J. Food Sci. 27 2009 143 150 10.17221/1415-CJFS
24 Hou S. He H. Zhang W. Xie H. Zhang X. Determination of soil amino acids by high performance liquid chromatography-electro spray ionization-mass spectrometry derivatized with 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate Talanta 80 2009 440 447 10.1016/j.talanta.2009.07.013 19836501
25 Hernández-Orte P. Ibarz M.J. Cacho J. Ferreira V. Amino acid determination in grape juices and wines by HPLC using a modification of the 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC) method Chromatographia 58 2003 29 35 10.1365/s10337-003-0002-1
26 Shen Q. Guo R. Dai Z. Zhang Y. Investigation of enzymatic hydrolysis conditions on the properties of protein hydrolysate from fish muscle (Collichthys niveatus) and evaluation of its functional properties J. Agric. Food Chem. 60 2012 5192 5198 10.1021/jf205258f 22530872
27 Qasrawi D.O. Petrotchenko E.V. Borchers C.H. Amino acid analysis for peptide quantitation using reversed-phase liquid chromatography combined with multiple reaction monitoring mass spectrometry Anal. Bioanal. Chem. 415 2023 5261 5267 10.1007/s00216-023-04840-2 37468754
28 Gritti F. Wahab F. Extraction of intrinsic column peak profiles of narrow-bore and microbore columns by peak deconvolution methods Anal. Chim. Acta 1180 2021 338851 10.1016/j.aca.2021.338851
29 Berg H.S. Seterdal K.E. Smetop T. Rozenvalds R. Brandtzaeg O.K. Vehus T. Lundanes E. Wilson S.R. Self-packed core shell nano liquid chromatography columns and silica-based monolithic trap columns for targeted proteomics J. Chromatogr., A 1498 2017 111 119 10.1016/j.chroma.2017.03.043 28385266
30 Liang X. Wu H. Su W. A rapid UHPLC-PAD fingerprint analysis of Chrysanthemum morifolium ramat combined with chemometrics methods Food Anal. Methods 7 2014 197 204 10.1007/s12161-013-9618-4
31 Bui T.T. Nguyen D.C. Hua S.H. Chun H. Kim Y.S. Sonochemical preparation of a magnet-responsive Fe3O4@zif-8 adsorbent for efficient Cu2+ removal Nanomaterials 12 2022 753 10.3390/nano12050753 35269241
32 Qi Y. Wan M. Abd El-Aty A.M. Li H. Cao L. She Y. Shao Y. Jin F. Wang S. Wang J. A “half” core-shell magnetic nanohybrid composed of zeolitic imidazolate framework and graphitic carbon nitride for magnetic solid-phase extraction of sulfonylurea herbicides from water samples followed by LC-MS/MS detection Microchim. Acta 187 2020 279 10.1007/s00604-020-04243-5
33 Shen Q. Gong L. Baibado J.T. Dong W. Wang Y. Dai Z. Cheung H.-Y. Graphene based pipette tip solid phase extraction of marine toxins in shellfish muscle followed by UPLC–MS/MS analysis Talanta 116 2013 770 775 10.1016/j.talanta.2013.07.042 24148472
34 Liu Y. Lu F. Adulterated pharmaceutical chemicals in botanical dietary supplements: novel screening approaches Rev. Anal. Chem. 36 2017 10.1016/j.talanta.2013.07.042 10.1515/revac-2016-0032
35 Li L.L. Cai Y. Yu K.M. Yang M. Shen Q. Cheung H.Y. Improvement of reproducibility and sensitivity by reducing matrix effect in micellar electrokinetic chromatography for determination of amino acids in turtle jelly Electrophoresis 36 2015 1186 1195 10.1002/elps.201400533 25781419
