Identification of the prepared foods promising for dietary folate intake in Beijing, China

Abstract The aim of the present study was to analyze folate content and composition in foods consumed daily by Chinese people. The concentration of seven folate derivatives in sixty‐four selected foods was determined by a liquid‐chromatography tandem‐mass spectrometry method. The total folate levels ranged from 0.28 to 129 µg/100 g fresh weight, with an average of 21.18 μg/100 g. The highest folate content was found in boiled egg yolk and waxy corn (>120 µg/100 g), abundant folate levels in cooked vegetables such as hot pepper, spinach, soybean sprout, stem lettuce, coriander, and broccoli (44–72 µg/100 g), and lowest in Coca Cola (0.28 µg/100 g). 5‐Methyl‐tetrahydrofolate was the major folate derivative in various foods, accounting for 72% of the total folates on average, with the highest being 90% in egg yolk. These data will enable estimation of the daily folate intake and allow dietary recommendations to improve folate status in humans.


| INTRODUC TI ON
Folates (vitamin B9), comprising tetrahydrofolate (THF) and its derivatives, are vital cofactors for all organisms. By acting as donors and acceptors in one-carbon metabolism, folates provide methyl groups for the biosynthesis of nucleotides, amino acids, formyl-methionyl tRNA, and pantothenate (Blancquaert et al., 2010). Folates are important for maintaining health and preventing disease.
Folate deficiency is linked to several human disorders including neural tube defects (NTDs) in infants and megaloblastic anemia, cardiovascular disease, and cancers in adults (Li et al., 2016;Smith & Refsum, 2016;Stover, 2004). Mammals cannot synthesize folates de novo and therefore dependent on dietary intake, mainly in the form of plant-based foods (Hanson & Gregory, 2011) . Thus, assessing the natural folates in cooked/prepared foods becomes necessary for estimation of the daily folate intake. Folate deficiency is a worldwide problem. China has a high prevalence of folate deficiencies, with 20% of the population being considered deficient and 18,000 pregnancies affected by NTDs annually (De Steur et al., 2010;Zhao et al., 2009). The recommended daily allowance (RDA) of folates is 400 μg/day for adults and 600 μg/day for pregnant women (Strobbe & Van Der Straeten, 2017). The RDA of folates can be met either by folic acid supplementation or by adequate intake of food-sourced folates. Folic acid supplementation has been reported to be effective in preventing folate malnutrition in humans (Czeizel et al., 2013;Wang et al., 2016); however, an excess of folic acid intake may result in adverse effects, such as increased cancer risk and progression within certain patient groups, insulin resistance in children, interaction with epilepsy medications, and masking vitamin B12 deficiency and hepatotoxicity (Patel & Sobczynska-Malefora, 2017). In light of food-sourced folate intake, the extent to which folates are absorbed by general population remains questionable due to the lack of values of folate content in processed foods and lack of information on the folate forms in the diets. Folate levels vary significantly among plant-based foods, and raw fresh vegetables and legumes are considered as good sources of folates (Blancquaert et al., 2014;Delchier et al., 2016). However, as much as 10%-64% of folate loss was observed during storage, industrial processing, and cooking due to the sensitivity of folates to heat, light, or oxidation (Dong et al., 2011;Hefni et al., 2015;Maharaj et al., 2015). Therefore, investigation of the folate retention in processed foods will enable an accurate estimation of dietary folate intake.
Microbiological assay and the high-performance liquid-chromatography coupled with tandem-mass spectrometry (HPLC-MS/ MS) are widely adopted for folate determination (Arcot & Shrestha, 2005;. The microbiological assay could not distinguish various folate derivatives, revealing only the total amount of folates. Moreover, because of different responses and stabilities of the folate vitamers, the microbiological assay may give inaccurate results (Ringling & Rychlik, 2017).
In the present study, we have investigated the folate content and composition in 64 foods commonly consumed in China, including cereals, vegetables, fruits, eggs, meat, fish, and beverages.
Seven folate derivatives in the foods were determined by LC-MS/ MS.

| Selection of foods
A total of 64 frequently consumed ready-for-consumption foods were selected based on the China Food Composition 2018 (main food category and composition of Chinese foods included) and were categorized into five groups: cereal/carbohydrate-based foods (11), vegetables (22), fruits (13), eggs/meat/fish (9), and beverages (9). The foods were purchased from three canteens in Beijing, China. Fruits and beverages were purchased from a local supermarket in Beijing. Sampling and analysis of the foods and beverages were performed in August 2019, and there were three replicates per sample. All of the foods were ground and homogenized before analysis.

| Beverages
Soya milk (15 g of soya powder in 100 ml of hot water), whole milk, skimmed milk, yoghurt, orange juice, Nescafe (15 g of Nescafe dissolved in 250 ml of hot water), Coca Cola, Pepsi, and Sprite.

| Sample preparation and extraction
Cooked rice, noodles, egg yolk, and egg albumen were ground into a fine paste using a mortar and pestle. The other foods were homogenized individually in a domestic kitchen blender. Half of the composite samples (~6 g) were used for determination of moisture content and the rest for folate analysis. Folate extraction was performed as described previously with slight modification (Riaz et al., 2019).
MTX at a final concentration of 20 ng/ml was used as internal standard and added to extraction buffer at the beginning of the extraction procedure.
For vegetables and fruits, 1 ml of extraction buffer was added to 50 mg of fine paste/powder and mixed; for beverages, 500 µl of 2 × extraction buffer was mixed with 500 µl of sample. The mixture was immediately boiled for 10 min in a water bath, cooled on ice, 30 µl of rat serum was added, and incubated at 37°C for 4 hr to convert polyglutamated into monoglutamated folates. The samples were boiled for 10 min to inactivate rat conjugase, cooled on ice for 10 min, centrifuged at 15,000 g at 4°C for 10 min, and the supernatants were transferred to 3 kDa ultra-filtration tubes (Millipore) for clean-up and centrifuged at 15,000 g at 4°C for 20 min. The resulting solution was used directly for folate analysis.
For cereal/carbohydrate-based foods, additional α-amylase treatment was used. 1 ml of extraction buffer was added to 50 mg of fine powder, mixed, boiled for 10 min, and cooled on ice to room temperature. Next, 20 µl of α-amylase (40 mg/ml) was added, mixed, and incubated at 37°C for 30 min. Subsequently, the samples were boiled for 5 min to deactivate α-amylase, cooled on ice, 30 µl of rat serum was added, and incubated at 37°C for 4 hr. The samples were boiled for 5 min, cooled on ice for 10 min, and centrifuged at 15,000 g at 4°C for 10 min. The supernatants were transferred to 3 kDa ultra-filtration tubes (Millipore) for clean-up and centrifuged at 15,000 g at 4°C for 20 min. The filtrate was used for folate analysis.
The mobile phases were 0.1% (v/v) formic acid in water (phase A) and 0.1% (v/v) formic acid in acetonitrile (phase B) with a gradient program of 20 min. The initial mobile phase B was set at 5% at a flow rate of 0.3 ml/min. The proportion of mobile phase B increased linearly from 5% to 9% over 2 min. In the following 7 min, phase B increased to 9.6%, then sharply increased to 20% over 0.2 min. After holding at 20% for 4 min, the proportion of phase B decreased to 5% in 0.8 min followed by an equilibration time of 6 min. Mass analysis and folate quantification were operated in an Aglient 6420 triplequadruple MS coupled to an electron spray ionization interface system. The mass spectrometer was operated in a positive ion mode.
The multiple reaction monitoring (MRM) parameters of each folate derivative including the precursor ion, the product ion, and collision energy (eV) were optimized with a gas temperature of 320°C, drying gas flow at 11 L/min, nebuliser pressure at 35 psi, and capillary voltage at 3,500 V. One major product ion for each folate was

| Determination of moisture content
Moisture content was determined in triplicate using a vacuum oven at 70°C overnight (AOAC, 2007). The foods (2 g) or beverage (2 ml) were initially weighed, incubated overnight in a vacuum oven at 70°C, and weighed. The moisture content was calculated as the difference between the two weights.

| Statistical analysis
All data were presented as the mean and standard deviation (SD) of three biological replicates. The folate data were subjected to analysis of principal component analysis (PCA) and Pearson's correlation coefficients (r) to find differences in folate content of different foodstuffs, identify folate vitamers with high discriminative properties, and the association between the various folate derivatives.
Data analysis and visualization were performed by using R 3.6.2 (R Foundation for Statistical Computing).

| Total folates in selected foods
A total of 64 foods/preparations commonly consumed by Chinese population were selected for folate analysis. Among them, 11 were cereal/carbohydrate-based preparations, 22 vegetables, 13 fresh fruits, 9 egg/meat/fish items, and 9 beverages. All of the foods were in ready-for-consumption forms.
The folate levels of these foods ranged from 0.28 ± 0.08 µg/100 g to 129 ± 10.4 µg/100 g fresh weight (FW), with an average of 21.1 μg/100 g (Table 1). The mean folate contents of cereal/carbohydrate-based foods, vegetables, fruits, and egg/meat/fish were 27.2, 28.8, 15.5, and 19.1 µg/100 g, respectively, and that of beverages was 5.4 µg/100 ml. Therefore, the folate contents varied considerably among different foods/preparations, with vegetables and cereal/carbohydrate-based foods averaging higher in folates than the other foods.
A principal component analysis (PCA) was conducted to examine the variation in folate concentration among all the selected foods. Foodstuffs were scattered in biplots by origin, with folate derivatives indicated as vectors, and five overlapped clusters were illustrated ( Figure 1). The first two principal components together explained 59.5% (PC1, 39.3% and PC2, 20.2%) of the total observed variation. 5-M-THF and THF were closely associated, and they were the major contributors to the variation in PC1; 5-F-THF has a close correlation with FA, and they contributed to the variation in PC2.

| Cereal/carbohydrate-based foods
Cereal/carbohydrate-based foods are usually consumed as staple foods. It was observed that the total folates ranged from 126 ± 18.5 µg/100 g to 3.29 ± 0.26 µg/100 g among the 11 selected cereal/carbohydrate foods (Table 2). Boiled waxy corn contained the highest level of folates (126 ± 18.5 µg/100 g), followed by sweet corn (47.6 ± 7.69 µg/100 g). The folate contents detected in corns in this study were comparable to that indicated in the USDA database (United States Department of Agriculture, 2020), where waxy corn and sweet corn were shown to contain 100 µg/100 g and 42 µg/100 g of folates, respectively.
Wheat is widely planted and commonly consumed as a staple food in North China (Riaz et al., 2019); unfortunately, wheat flour contains folates as low as 4-20 µg/100 g (Patring et al., 2009). In the present study, several foods are made of wheat flour, including bread, steamed bun, fermented pancake, plain pancake, noodles, and gluten. Out of these, bread had a highest level of folates (31.2 ± 4.2 µg/100 g), followed by steamed bun (19.9 ± 4.3 µg/100 g) and pancakes (12.5 ± 1.78 µg/100 g and 15.4 ± 1.92 µg/100 g), with noodles and gluten being the least (4.02 ± 1.51 µg/100 g and 3.29 ± 0.26 µg/100 g). The folate level detected in bread was in agreement with the previous study, which reported a folate level of 29.8 µg/100 g (Pfeiffer et al., 1997). Compared to noodles and gluten, the higher level of folates in bread and steamed bun may attribute to the yeast used for fermentation during food preparation, a good tool to increase folates in foods (Saubade et al., 2017).

| Vegetable-based preparations
The folate levels of 26 fresh vegetables commonly consumed in China were determined previously, ranging from 14.8 µg/100 g to 146 µg/100 g and averaging 62 µg/100 g, with the highest in leafy vegetables such as pak choi and spinach (Shohag et al., 2012).
In this study, folates in 22 cooked/prepared vegetables were examined. It was observed that the folate levels varied from 72 ± 6.4 µg/100 g for hot pepper to 2.65 ± 0.5 µg/100 g for potato, with an average of 28.82 µg/100 g (Table 3). Cooked hot pepper, spinach, soybean sprout, lettuce, and broccoli were found containing a considerably high level of folates (>44 µg/100 g), ensuring 11%-25% of the recommended daily folate intake upon 100-g serving.
Previously, it was reported that the folate contents of raw spinach and broccoli were 143.99 and 110.67 µg/100 g, respectively (Shohag et al., 2012). In the present study, the folate levels were 69.6 µg/100 g for spinach and 44.4 µg/100 g for broccoli, indicating a 52%-60% folate loss upon cooking. In fact, a similar folate loss upon boiling was also observed in spinach (51%) and broccoli (66%) in a previous study (McKillop et al., 2002). The cooked stem lettuce was also found as a good source of folates in this study (57.7 ± 6.2 µg/100 g).
In this study, it was found that the boiled soybean sprout contained folates at a level of 64.05 ± 9.73 µg/100 g and the boiled tofu 14.31 ± 0.59 µg/100 g fresh weight. Obviously, the big difference of folate levels in tofu was due to the calculations based on dry weight or on fresh weight.
The folate content of tomatoes served in green salad was 11.4 ± 0.5 µg/100 g, whereas 8.27 ± 1.1 µg/100 g in cooked form, both lower than that in red ripe tomatoes (18.1 µg/100 g) (Tyagi et al., 2015). Potato has been grown as a staple crop in certain region, but is a poor source of folates (19 µg/100 g in raw samples) (Blancquaert et al., 2014). In this study, the cooked potato chips had the lowest level of folates (2.65 ± 0.51 µg/100 g), thus confirming the conclusion mentioned above.

| Fresh fruits
The folate levels in 13 fresh fruits showed a broad range of variations, with the highest being 44.58 ± 0.62 µg/100 g in lemon and lowest being 3.78 ± 0.91 µg/100 g in mandarin orange, averaging 15.50 µg/100 g (Table 4). Recently, a range of folate levels of TA B L E 2 Folate content and composition of cereal/carbohydrate-based preparations (µg/100 g)   (Striegel et al., 2019). In this study, most of the fruits were temperate or subtropical species, indicating a significant difference in folate levels between nontropical and tropical fruits.
The folate level in the apple was 21.08 ± 2.7 µg/100 g, in good agreement with that stated in the USDA database 2019 (19 µg/100 g). Melons and grapes are the fruits in market supply in August in the North of China. The folate contents of watermelon and kyoho grape were 6.48 ± 0.97 and 6.77 ± 0.37 µg/100 g, respectively, and those of the other varieties of seedless grape and green melon were 29.54 ± 0.92 and 23.53 ± 5.02 µg/100 g, respectively. Thus, the seedless grape and green melon were good sources for folate nutrition improvement compared to watermelon and kyoho grapes. Navel oranges and mandarin oranges were also commonly consumed by general population, the folate levels were 15.1 ± 3.4 µg/100 g and 3.78 ± 0.91 µg/100 g, respectively (Table 4).

| Meat/egg/fish preparations
Usually, plant protein and animal foods, such as meat, egg, and fish, contain high level of good quality proteins. In this study, the cooked/ prepared chicken, beef, mutton, pork, fish, shrimp, egg yolk, tofu, soya milk, and egg albumen were analyzed for folates and a big variation was observed among these preparations (Table 5). The boiled egg yolk contained the highest level of folates (129 ± 10.4 µg/100 g), consistent with the data of the USDA database (116 µg/100 g).
Thus, egg yolk is a good source of folates, with 5-M-THF accounting for up to 90% of the total; similar 5-M-THF abundance was also reported in a recent investigation on folate-enriched egg yolks (Sun et al., 2020). In contrast, egg albumen contained a low level 13.4 ± 1.8 µg/100 g. An extremely low level of folates was observed in fish and the selected meat (1.3-5.34 µg/100 g), more or less the same as that in the USDA database (2-8 µg/100 g). Our results were in agreement with the recent literature, which reported that the folate levels in chicken breast and pork tenderloin were 5 ± 2 µg/100 g and 1 ± 1 µg/100 g, respectively, and therefore, meat is not a good source of folates (Loznjak et al., 2020).
Soya milk made of soybean contained the highest level and Coca cola contained the lowest level. 5-F-THF and 5-M-THF were the two dominant folate derivatives in soya milk, contributing equally to the total folates (11.5 ± 1.74 µg/100 ml vs. 12.7 ± 1.88 µg/100 ml).
Whole milk and yoghurt contained folates at levels of 2.1 ± 0.3 µg/100 ml and 6.85 ± 0.3 µg/100 g, respectively. A previous study also examined folate contents in milk and yoghurt, with a range of 4-10 µg/100 g due to seasonal variations and different starter cultures (Forssen et al., 2000); in comparison, a higher level of folates, 20 ± 7 µg/100 g, was determined by a microbiological assay in the home-made yoghurt (Vishnumohan et al., 2009). As expected, a low level of folates was observed in Pepsi, Sprite, Nescafe, and Coca Cola (<1 µg/100 ml) with 5-M-THF only being detected, and in agreement with the data indicated in the USDA database.

| CON CLUS ION
Dietary folate intake by humans is dependent on food consumption.
This is the first report on natural folates presented in cooked foods/ preparations in China. The data of this study demonstrate that the level of folates vary greatly among foods, with boiled egg yolk, waxy corn, cooked hot pepper, spinach, soybean sprout, stem lettuce, broccoli, coriander, lemon, and soya milk being optimal sources of folates. Boiled waxy corn has been found most abundant in folates among the foods examined in this study. Given a high popularity with favorable flavor and nutritious ingredients in Chinese population, waxy corn, including sweet corn, is strongly recommended as daily food for dietary folate intake. To conclude, the data generated in the study will be of help for an accurate estimation of daily folate intake of Chinese people and for establishment of an optimal dietary pattern to meet the daily folate requirement as well.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

E TH I C A L A PPROVA L
This study does not involve any human or animal testing.