The analysis of 2′-fucosyllactose concentration in Korean maternal milk using LC–MS/MS

Abstract

Despite health benefits reported recently, 2′-fucosyllactose (2′-FL) concentration in maternal milk was not conclusively reported because it varies between countries and mothers. Particularly, its distribution among Korean mothers was not obtained from a reliable sample group yet. Thus, a dynamic range for 2′-FL concentration in Korean mothers’ milk was investigated from 102 samples. A quantitative method using multiple reaction monitoring (MRM) by triple-quadrupole-mass spectrometry has been evaluated by a standard procedure of method validation. The 2′-FL concentration was in the range of 0.4 to 2.6 g/L overall. While the samples from secretor mothers (n = 80) contained 1.0 to 2.8 g/L of 2′-FL, the maternal milk from non-secretor mothers (n = 22) had 0.01 to 0.06 g/L of 2′-FL only. In addition to the genetic variation of mothers, the lactation period impacted the 2′-FL concentration. The average 2′-FL concentration of the late-stage group (> 60 days) was 78% of that obtained from the first month of postpartum mothers.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10068-022-01154-4.

Introduction

Human milk is an ideal food for infants due to its unique composition. Aside from energy nutrients, it contains numerous functional metabolites, including human milk oligosaccharides (HMOs) (Ballard & Morrow, 2013). HMOs comprise a group of unconjugated glycans which are highly abundant and unique to human milk. Compared to human milk, animal milks from goats, cows, and sheep contain free oligosaccharides at much lower levels and with less diversity (van Leeuwen et al., 2020).

2′-Fucosyllactose (2′-FL) has been demonstrated to have multiple health benefits for infants. It is a natural prebiotic that may selectively promote Bifidobacterium growth (Hirvonen et al., 2019; Lewis et al., 2015). Furthermore, studies have reported that 2′-FL is effective against bacterial (Good et al., 2016; Weichert et al., 2013) and viral infections (Moore et al., 2021), inhibits pathogenic microbial adhesion, and reduces the risk of necrotizing enterocolitis in premature infants (Good et al., 2016). In addition, an important property of 2′-FL is immunomodulation. It may help reduce food allergies (He et al., 2016) and inhibit inflammatory pathways (He et al., 2016).

Among HMOs, 2′-FL occurs at a high concentration in secretor’s mother milk. However, the concentration of 2′-FL and HMO concentrations depend on several factors. Approximately 20% of women are non-secretors who produce milk with negligible amounts of 2′-FL due lack of fucosyltransferase 2 (FUT2) gene (Kelly et al., 1995). In addition, the concentration of 2′-FL depends on lactation stages (Thum et al., 2021). Furthermore, 2′-FL concentrations in human milk appear to vary between different countries, such as Japan, Malaysia, China, Germany, USA, and some European countries, and have been reported in the range of 1.7–3.7 g/L (Thum et al., 2021).

Uptaking 2′-FL in sufficient amounts could provide many health benefits for infants. Moreover, 2′-FL concentration in human milk is a gold standard for setting recommended nutrient intake (RNI) values for babies. Therefore, in this study, we sought to provide a dynamic range for 2′-FL levels in Korean mother’s milk and a dataset of 2′-FL concentrations to support childcare nutritional programs.

Materials and methods

Standards and materials

2′-FL standard was purchased from Carbosynth Ltd. (Berkshire, UK). Acetonitrile (HPLC grade) was obtained from Merck (Darmstadt, Germany). Formic acid and trifluoroacetic acid were purchased from Thermo Fisher Scientific Korea Co. Ltd. (Seoul, Korea). Porous graphitic carbon cartridges (Bond Elute Carbon, 250 mg) were purchased from Agilent (Santa Clara, CA, USA).

Study design and participants

This study was approved by the Institutional Review Board of Chungnam National University (Korea) and Maeil Dairies Co. Ltd. (Korea). Human milk collection was performed using relevant guidelines and regulations. Written informed consent was obtained from all participants before sample collection.

Human milk was collected from 102 healthy Korean mothers, regardless of diet, supplementation, or nutritional status. Human milk was directly collected into sterilized 50 mL conical tubes. The sample was then packaged in a gel ice pack in well-insulated containers to maintain low temperatures, and delivered to the laboratory. Samples were stored at − 80 °C before analysis.

Sample preparation for liquid chromatography- tandem mass spectrometry (LC–MS/MS) analysis

2′-FL extraction

2′-FL was extracted and purified using a previous method (Xu et al., 2017), with minor modifications. Human milk was thawed overnight at 4 °C before analysis. A 20 µL sample was diluted 20–50 fold in distilled water. Then, a 25 µL diluted aliquot was mixed with 25 µL distilled water. The lipid was removed by the Folch method (Folch et al., 1957). A 200 µL Folch solution (chloroform/methanol in a 2:1 ratio) was added to the mixture, vortexed for 10 s. Then, the samples were centrifuged at 14,000 rpm at 4 °C for 10 min to separate the chloroform layer. The top layers were collected and protein precipitation was performed by adding 700 µL chilled ethanol. The sample was centrifuged at 14,000 rpm at 4 °C for 20 min. Finally, the supernatant was collected and completely evaporated under vacuum.

2′-FL reduction and purification

2′-FL was reduced to an alditol form using 1 M NaBH₄ (Sigma-Aldrich, Merck) at 65 °C for 1 h. Then, 2′-FL was purified by solid-phase extraction (SPE) using porous graphitic carbon cartridges (Bond Elut Carbon, 250 mg, 6 mL, Agilent). The cartridge was activated by adding 6 mL 80% acetonitrile (containing 0.1% trifluoroacetic acid) and conditioned by 6 mL of distilled water. After sample loading, cartridges were washed by adding 6 mL distilled water. 2′-FL was eluted by 40% acetonitrile (containing 0.05% trifluoroacetic acid). The eluted fraction was evaporated and resuspended in distilled water for LC–MS/MS analysis.

2′-FL standard preparation

A stock solution was prepared at 100 mg/L in distilled water. After reduction and purification by NaHB₄ and SPE as above method, respectively. The stock solution was later diluted to obtain 0.05–10 mg/L working standards.

Quantitative analysis of 2′-FL by LC–MS/MS

For quantitation, 2′-FL concentrations were analyzed by LC–MS/MS (6470 Triple Quadrupole System, Agilent). Separation was performed on an LC analytical column packed with porous graphitized carbon (Hypercarb™ Porous Graphitic Carbon-PGC, 3 μm, 2.1 × 100 mm, Thermo Scientific, USA). The column temperature was 40 °C. A binary solvent system consisted of distilled water (solvent A) and acetonitrile (solvent B), containing 0.1% formic acid. The flow rate was 0.3 mL/min. The gradient was: 0–1 min, 3% B; 1–15 min, 4%–15% B; 15–20 min, 15%–40% B; 20–22 min, 40%–90% B; 22–27 min, 90% B; and 27.5–40 min 3% B.

The mass spectrometry was operated in positive mode. Drying gas and sheath gas temperatures were 250 °C and 350 °C, respectively. Drying and sheath gas flow rates were 6 L/min and 8 L/min, respectively. Nebulizer pressure was 40 psi. Capillary voltage was 4000 V. Multiple reaction monitoring mode (MRM) transition for 2′-FL were 491.2 → 183.1 m/z (for quantification) and 491.2 → 329.0 m/z (for qualification).

Results and discussion

Method validation

Standard method validation has been performed in precision, accuracy, linearity, selectivity, matrix effect, and 2′-FL recovery from milk.

Intra- and inter-day precision and accuracy

The validation of precision and accuracy were performed by analyzing six replicates of the low, medium, and high concentrations of 2′-FL. The inter-day validation was obtained by repeating the same set of experiments on 3 different days. Samples at all concentrations did not deviate (%CV) by more than ± 15% from their nominal concentration. The relative error, which indicated the accuracy of the analysis, was less than ± 15% of the nominal concentration (Supplementary Table S1).

Linearity of standard curve and back-calculation

The standard curve was obtained from seven different concentrations in the range of 50–10,000 mg/L. The correlation coefficient of the calibration curve was 0.9998 exhibiting strong linearity (Supplementary Figure S1). Back-calculated concentrations of calibrators were within ± 15% of nominal concentrations, except at the point below the limit of quantitation (LOQ) (± 20%). We observed that 75% of calibrator levels met these criteria in validation runs (Supplementary Table S2). The limit of detection (LOD) and quantitation (LOQ) were 10 and 50 ppb, respectively.

Selectivity

Since no blank matrix was available, a 2′-FL standard was added to maternal milk samples (after sample purification) to validate the selectivity. The MRM mode of mass spectrometric analysis demonstrated outstanding specificity. No interference peaks were observed in the chromatogram except for the selected mass peak.

Among HMOs, 3-FL has the same mass (m/z value) as 2′-FL, however, 2′-FL and 3-FL were completely separated by the PGC column showing different retention times. 2′-FL had a retention time of 9 min followed by the 3-FL which was eluted in 2.8 min. Additionally, the standard addition method to the real sample confirmed the peak separation and retention times (Supplementary Figure S1).

Recovery and reproducibility

To ensure the efficiency and reproducibility of sample preparation, recovery experiments were performed as follows. The 2′-FL was extracted by liquid–liquid extraction (LLE) from human milk and then cleaned up by SPE. A blank matrix for the recovery test was prepared using the same human milk sample but the reduction of oligosaccharides was not performed. The 2’-FL standard, which was reduced by NaHB₄, was spiked into the blank matrix at the final concentration of 1 and 5 mg/L levels. For the calculation of recovery, the spiked 2’-FL concentration (reduced form) of the instrumental measurement was compared with the added concentration. Our method demonstrated a high recovery between 92 and 104% (Supplementary Table S3) which was suitable for the quantitative analysis of 2’-FL in human milk. The average CV (%) was less than ± 15% indicating the high reproducibility of the assay.

Matrix effect

The matrix effect was evaluated using five replicates of samples which were spiked with a low and high concentration of 2′-FL. Unspiked and standard spiked samples were analyzed under the same conditions. Differences between the values measured by the instrument and the initial added values were <  ± 5% (Supplementary Table S4) which suggests no or, at least, a minimum matrix effect.

Sample stability

The stability of 2′-FL during the instrumental analysis has been determined. Two conditions were evaluated: short term-storage at the condition in the auto-sampler (4 °C) for 24 h and long-term storage which is at − 60 °C for 4 weeks. Purified samples were stable with a variation of less than ± 10% at both short-term and long-term storage (Supplementary Table S5).

Biological information of mothers and milk macronutrients

Biological information of mother-infant and macronutrients in their human milk is summarized (Table 1). On average, the age of mothers was 32.2 ± 3.3 years old and the body mass index (BMI) was 22.2 ± 2.8 kg/m². No significant differences were observed between lactation periods.

Table 1.

Mothers’ characteristics and macronutrients present in milk

Lactation stage (day)	Count	Mother characteristics		Milk macronutrients (g/100 mL)a
		Mother BMI	Mother age (year)	Fat	Protein	Lactose	Total solid
Total
0–30	4	22.2 ± 3.9	32.5 ± 5.4	3.9 ± 0.6	1.4 ± 0.1	6.6 ± 0.2	12.8 ± 0.6
30–60	22	22.3 ± 2.3	32.1 ± 2.7	3.3 ± 0.9	1.3 ± 0.1	6.7 ± 0.2	12.3 ± 1.0
> 60	76	22.2 ± 3.0	32.3 ± 3.3	3.4 ± 1.2	1.2 ± 0.1	6.6 ± 0.2	12.1 ± 1.2
Total	102	22.2 ± 2.8	32.2 ± 3.3	3.4 ± 1.1	1.2 ± 0.2	6.6 ± 0.2	12.2 ± 1.1
Secretor
0–30	4	22.2 ± 3.9	32.5 ± 5.4	3.9 ± 0.6	1.4 ± 0.1	6.6 ± 0.2	12.8 ± 0.6
30–60	17	22.1 ± 2.5	32.2 ± 2.0	3.3 ± 1.0	1.3 ± 0.1	6.7 ± 0.2	12.2 ± 1.0
> 60	59	22.3 ± 3.1	32.6 ± 3.4	3.4 ± 1.2	1.1 ± 0.2	6.6 ± 0.2	12.0 ± 1.2
Total	80	22.3 ± 3.0	32.5 ± 3.2	3.4 ± 1.1	1.2 ± 0.2	6.6 ± 0.2	12.1 ± 1.1
Non-secretor
0–30	0
30–60	5	22.9 ± 2.0	31.4 ± 4.8	3.5 ± 0.9	1.4 ± 0.2	6.7 ± 0.2	12.5 ± 0.9
> 60	17	21.8 ± 2.5	31.2 ± 3.0	3.7 ± 1.1	1.2 ± 0.1	6.5 ± 0.3	12.4 ± 1.0
Total	22	22.1 ± 2.4	31.2 ± 3.4	3.7 ± 1.1	1.2 ± 0.2	6.6 ± 0.3	12.4 ± 1.0

^aMilk macronutrients were analyzed by infrared spectrometry using MilkoScan FT-2

Milk macronutrients slightly decreased during lactation stages. Fat concentrations slightly decreased from 3.9 ± 0.6 to 3.4 ± 1.2 g/100 mL, corresponding to 0–30 days to > 60 days of postpartum. Protein concentration was also slightly decreased from 1.4 to 1.2 g/100 mL. As consequence, a decrease in the total solid content from 12.8 ± 0.6 to 12.1 ± 1.2 g/100 mL was observed with the increase in the postpartum periods.

2′-FL concentration in Korean maternal milk

On average, the 2′-FL concentration in Korean mothers’ milk was 0.4 to 2.6 g/L (n = 102). The median value of 2′-FL concentration (mg/L) was 1.6 g/L. Secretors' milk from the secretor mother contained high 2′-FL concentrations in the range of 1.0 to 2.8 g/L (n = 80). Theoretically, non-secretor mothers, whose FUT2 gene was inactivated, cannot synthesize α-1,2- fucosylated HMOs. However, 2′-FL concentrations in maternal milk from non-secretor mothers were not zero. It was distributed between 0.01 and 0.11 g/L with an average of 35.3 ± 24.9 mg/L, which was 1%–10% of the levels observed in secretors’ milk. The median, maximum, minimum, 25th, and 75th percentile values for 2′-FL in secretor and non-secretor groups are shown (Table 2).

Table 2.

2′-FL concentrations (mg/L) in Korean mothers’ milk

Lactation stage (day)	Count	Average ± STD	Median	Minimum	Maximum	25th percentile	75th percentile
Total
0–30	4	2300.6 ± 279.1	2282.0	1932.2	2706.0	2139.2	2443.4
30–60	22	1589.2 ± 1054.6	1752.8	8.0	4079.2	1100.8	1994.8
> 60	76	1398.3 ± 1114.1	1420.7	12.7	5919.1	481.8	1971.6
Total	102	1474.8 ± 1096.2	1581.8	8.0	5919.1	552.4	1989.9
Secretor
0–30	4	2300.6 ± 279.1	2282.0	1932.2	2706.0	2139.2	2443.4
30–60	17	2046.6 ± 720.1	1851.4	1069.3	4079.2	1603.0	2429.4
> 60	59	1790.9 ± 953.7	1776.2	361.9	5919.1	1143.8	2049.5
Total	80	1870.7 ± 897.4	1836.3	361.9	5919.1	1351.1	2230.5
Non-secretor
0–30	0
30–60	5	33.9 ± 20.5	30.5	8.0	65.4	18.1	47.4
> 60	17	35.7 ± 26.0	22.1	12.7	112.0	18.4	53.3
Total	22	35.3 ± 24.9	24.0	8.0	112.0	18.2	51.9

Among the groups with a different lactation period, 2′-FL concentrations showed a trait of decrease with an increase in the postpartum time. Compared to the first-month lactation group 2′-FLconcentration decreased from 2.3 g/L to 1.4 g/L after two months of breastfeeding. From the first month (0–30 days) to the second (30–60 days), 2′-FL concentrations in secretor mothers’ milk decreased to 13% and further dropped to 15% in the next stage (> 60 days).

Due to the small number of lactation samples, non-secretor samples in the early stage were not observed. However, 2′-FL concentrations in non-secretor mothers’ milk were stable from 30 to over 60 days of breastfeeding showing constant at 35.3 ± 24.9 mg/L.

2′-FL levels in human milk vary between different countries and/or geographical regions (McGuire et al., 2017). A study on human milk from seven European countries showed that the 2′-FL concentration in Europe is relatively higher than in Korea. Meanwhile, 2′-FL concentration dropped from 3.7 to 1.6 g/L over the first four months of postpartum which is similar to Korean mothers (Samuel et al., 2019). 2′-FL level of Singaporean maternal milk has been reported the decrease with the lactation period as well showing 2.2 to 1.4 g/L and 26 to 11 mg/L at 1 and 4 months of postpartum, respectively (Sprenger et al., 2017). Meanwhile, China exhibited a wide range of 2′-FL concentrations from 0.9 to 2.9 g/L during the colostrum to over 120 days of lactation (Zhou et al., 2021).

In this study, the absolute quantitation of 2′-FL was analyzed to provide the reference data of 2′-FL distribution in Korean maternal milk. 2′-FL concentrations from secretor and non-secretor mothers were investigated and exhibited that the non-secretor mother was able to synthesize 2′-fucosylated glycan, though the quantity was extremely low compared to secretor mothers. In addition, changes in 2′-FL levels at different stages of lactation were also demonstrated. As such, our data could facilitate the establishment of RNI (Recommended Nutrient Intake) of 2′-FL values for Korean babies.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

This research was supported by the National Research Foundation of Korea (NRF) grant, funded by the Korean government (MSIT) (2020R1A2C1011944) and the Maeil Dairies Co., Ltd.

Declarations

Conflict of interest

The authors declare no conflict of interest.