1. Introduction
Prunus mume is an important fruit crop owing its health benefits and characteristic sensory attributes [
1]. The common Korean name for the fruit of
P. mume is
maesil, which is also known as Japanese apricot or Chinese plum.
Maesil is used to prepare various value-added products such as pickles, squashes, flavoring agents, alcoholic beverages, and extracts for medicinal use in Korea, China, and Japan [
2,
3]. Fermented
maesil products are gaining popularity and their consumption rate has increased in Korea because of their nutritional and health benefits like antioxidant and anti-osteoporosis activities [
4].
In general,
maesil is not consumed as raw fruit owing to its strong sour taste and the presence of a toxic cyanogenic glucoside D-mandelonitrile-β-D-gentiobioside (amygdalin) [
5]. During physical damage such as chewing and bruising of
maesil, enzymatic degradation of amygdalin occurs, producing hydrogen cyanide (HCN) and a ketone or aldehyde [
6]. Consumption of cyanogenic material from plant-based foods has been reported to cause both acute and sub-acute health issues, such as abdominal cramps, cardiac arrest, failure of the circulatory and respiratory system, coma, dizziness, headaches, mental confusion, nausea, vomiting, weakness, and death in extreme conditions, depending on the dosage [
7].
Currently there is growing public concern about the toxicity of
maesil products. In particular, the toxicity of
maesil sugar syrup is receiving substantial public attention because of the significant amounts being consumed every day [
2].
Maesil sugar syrup is prepared by soaking unripe
maesil in sugar for a long time (more than 3 months) at room temperature for natural fermentation. The seeds of unripe
maesil contain a high amount of amygdalin, which raise public concern to use these fruits for the sugar syrup preparation [
8]. It is believed that using either ripe or destoned fruits, whether ripe or unripe, in the production of
maesil sugar syrup, along with prolonging the fermentation period, will lower the level of amygdalin in the finished product. However, the influence of ripe, unripe, and destoned fruits, and of the fermentation period on the amygdalin content in the
maesil sugar syrup has not been thoroughly studied. Therefore, the present investigation aims to evaluate the effect of
maesil maturity, ripening methods, processing, and fermentation period on the amygdalin content in
maesil sugar syrup.
3. Results and Discussion
Standard amygdalin (50 mg/L) appeared at a characteristic retention time of approximately 10.09 min using 25% methanol mobile phase at 214 nm in the HPLC analysis (
Figure 2). The HPLC instruments have limit of detection (LOD) 0.3516 mg/L and limit of quantification (LOQ) 1.0655 mg/L. Under the analyzed HPLC instrument conditions, amygdalin content was detected in both the sugar syrup and different parts of
maesil. The chromatograms of the samples (
Figures S1–S6) contained a few other unknown peaks, which were attributed to other chemicals in the samples. A wide range of amygdalin content (from 0 to 52.85 g/kg) was detected in different parts of the
maesil with maximum amount in unripe
maesil seeds (52.85 ± 2.00 g/kg), followed by that in ripe
maesil seeds (40.33 ± 3.96 g/kg) (
Table 1). Similarly, unripe fruit pulp contained a significantly higher amount of amygdalin (0.81 ± 0.22 g/kg) than ripe fruit pulp (0.19 ± 0.01 g/kg). The seeds and fruit pulp of unripe fruits contained 98.48% and 1.52% of the amygdalin content, respectively. Similarly, the seeds and fruit pulp of ripe fruits contained 99.52% and 0.48% of the amygdalin content, respectively. Interestingly, no presence of amygdalin was detected in seed shell (endocarp) and seed coats (testa) of both ripen and unripen
maesil (
Table 1). These results indicate that seeds are the major source of amygdalin in both ripe and unripe
maesil. Previous studies have reported higher amounts of amygdalin in seeds (ranging from 100–340 mg/100 g) compared to fruit pulp (ranging from 5–35 mg/100 g) in various cultivars of
P. mume, supporting these findings (Son et al., 2017) [
8]. The amygdalin content in seeds and pulp of ripe fruit was 23.70% and 76.02% lower than that of unripe fruit, consistent with the findings of Son et al. (2017) [
8]. Interestingly, amygdalin was not detected in either ripe or unripe seed coat and seed shell. Similarly, the fruit enlargement stage of
Hikawa hakuho variety peach showed no amygdalin content in the seed shell [
11]. It should be noted that the seed shell is made of fibrous lignin and composes the hard part of the fruit, which may affect the extraction of amygdalin by restricting the transport through the seed shell [
12].
The results of the long-term amygdalin analysis of all
maesil sugar syrups revealed that the amygdalin content decreased during fermentation (
Figure 3). A higher amount of amygdalin (63.31 ± 1.68 to 66.19 ± 3.21 mg/L) was found in
maesil sugar syrups prepared with unripe
maesil (U1–U3), compared with that found in syrup prepared with ripe
maesil (R1–R3) (42.17 ± 0.30 to 50.39 ± 0.84 mg/L), after three months of fermentation. The lowest amount of amygdalin (42.17 ± 0.30 mg/L) was detected in R1 sample. A significantly (
p < 0.05) ~1.5-fold higher prevalence of amygdalin was detected in
maesil sugar syrups prepared with unripe
maesil pulp (U1) compared to those prepared with ripened
maesil pulp (R1) after three months of fermentation (
Figure 3). Consistently, a significant (
p < 0.05) ~1.3-fold higher concentration of amygdalin was detected in
maesil sugar syrups with unripe
maesil (U2 and U3), compared to the similar sugar syrups prepared with ripened
maesil (R2 and R3), respectively (
Figure 3).
Visual observations of the textural integrity of both ripe and unripe processed maesil indicated that osmotic stress induced by excessive sugar had a negative impact. This stress led to irregularities such as texture damage to the fruit skin and partial or complete liquefaction of the fruit pulp. Although neither ripe nor unripe whole fruits underwent liquefaction, they did experience shrinkage due to osmotic stress and water loss. When the seeds were extracted, the fruit was halved, increasing the surface area exposed directly to sugar. The absence of fruit skin on the cut side may have resulted in textural instability, which was visually evident in the deseeded fruit material of the R1 sample.
Due to its unique physical properties, the unripe fruit pulp maintained its texture better than the ripe fruit pulp [
3,
13]. Consequently, no significant changes in structural integrity (except shrinkage) were observed visually in the unripe
maesil, nor in their amygdalin content in the U1–U3 sugar syrups. The observed differences in amygdalin concentration between unripe (U1–U3) and ripe (R1–R3) fruit sugar syrups may be attributed to variations in the original fruit’s amygdalin content, the rate of amygdalin release from the fruit to the
maesil sugar syrups, and the fermentation process.
The findings of the 6-month amygdalin analysis showed a remarkable (47–66%) decrease in the amygdalin content in all tested samples as compared to the amygdalin content observed at 3-month (
Figure 3). The maximum 67% and 66% amygdalin reduction was observed in sample U2 (20.94 ± 3.21 mg/L), and in sample R2 (16.06 ± 0.57 mg/L), which are the sugar syrups from where
maesil was removed after three months of fermentation. Sugar syrup samples U1 and U3 had the highest amounts of amygdalin of 34.32 ± 0.54 mg/L and 34.12 ± 1.85 mg/L, indicating a 47.12% and 48.45% reduction of amygdalin, respectively, compared with the corresponding samples fermented for 3 months. The maximum amount of amygdalin among the ripe
maesil sugar syrups (R1–R3) was noticed in sample R1, with the lowest percentage reduction (36.46%) of amygdalin in the 6-month fermented samples. Sample R3 contained 19.03 ± 0.64 mg/L of amygdalin, corresponding to a 62.23% decline in amygdalin content when compared with the sample fermented for 3 months.
The significant differences between the amygdalin content of the sugar syrup samples are due to different sources of fruit, the rate of amygdalin release, and microbial action during fermentation. The lack of a fruit source in U2 and R2 resulted in no liberation of amygdalin to the
maesil sugar syrup, which is a probable reason for the significant reduction in amygdalin content. The loss of toughness and liquefaction of the fruit tissue and cell structure of ripe
maesil in R1 might be the reason for the high amount of amygdalin release [
6].
Moreover, all tested sugar syrup samples were found to contain 3.8–4.4 log CFU/mL of sugar-tolerant microbes, specifically
Zygosaccharomyces rouxii [
9].
Z. rouxii is the primary microorganism present in all
maesil sugar syrups, leading to alcohol fermentation [
9]. During this fermentation, these microbes utilize amygdalin for their growth, which might explain the reduction of amygdalin in all the
maesil sugar syrups compared to the samples fermented for 3 months [
14].
In general, the enzymatic hydrolysis of amygdalin by glycosidases and hydroxyl nitrile lyases results in the production of glucosides, HCN, and ketone/aldehyde as a defensive mechanism against insect infestation in plants [
15]. The isolated
Z. rouxii in the
maesil sugar syrups possess both β-glycosidases and hydroxynitrile lyase enzymes, which are capable of hydrolyzing amygdalin into benzaldehyde and HCN [
14]. HCN, a toxic hydrolysis product of amygdalin, either evaporates due to its high volatility or is utilized by
Z. rouxii through mineralization as a source of carbon and nitrogen, or is converted into ammonia and carbon dioxide [
14,
16]. Son et al. (2017) reported that the concentration of amygdalin in
maesil sugar syrups declined after fermentation of three months [
8], supporting our findings.
The results of the 9-month analysis revealed a significant reduction in the amygdalin quantity of all tested sugar syrup samples (
Figure 3). The lowest amygdalin content was observed in R1 (3.18 ± 0.13 mg/L), representing a 92.44% and 88.10% reduction compared to the 3- and 6-month samples, respectively. U1 and U2 contained amygdalin (5.61 ± 0.17 mg/L and 5.82 ± 0.13 mg/L) that represented 91.36% and 90.80% and 82.66% and 72.20% decrease in amygdalin content, compared with 3- and 6-months fermented samples, respectively. The whole fruit
maesil sugar syrups (U3 and R3) possessed the highest levels of amygdalin in both unripe and ripe fruit
maesil sugar syrups. Compared to the total amygdalin level of the samples fermented for 3 and 6 months, a reduction by 87.08% and 74.94% was observed in U3, whereas a reduction by 77.92% and 41.53% was observed in R3, respectively. The R2
maesil sugar syrup contained 7.25 ± 0.32 mg/L of amygdalin, which decreased to 84.50% and 54.81% in samples fermented for 3 and 6 months, respectively.
The low level of amygdalin found in the samples fermented for 9 months might be due to the peak release rate of amygdalin occurring at three months. After this period, there was little to no further release, combined with a significant rate of degradation, which lowered the amygdalin content in all
maesil sugar syrups [
8]. Consequently, slightly higher amounts of amygdalin were detected in U3 and R3, where the fruit source remained present throughout the fermentation process.
The present results revealed that the amygdalin content in the fruit pulp (0.4–1.5% of the total; 0.2–0.8 g/kg of dry weight) was much lower than that in the seeds (>99–98% of the total; 42–53 g/kg of dry weight), which is another possible reason for the low amygdalin levels. A decline in the quantity of amygdalin over time indicates that the amygdalin content in
maesil sugar syrups may have been predominantly derived from the fruit pulp and not from the seeds. The tough, fibrous, lignin-based endocarp (seed shell) may prevent the seeds from osmotic damage stress [
17], resulting in no migration of amygdalin from the seeds to the syrup. Thus, the quantity of amygdalin gradually decreases over time in
maesil sugar syrups.
After 12 months of fermentation, all the sugar syrup samples showed more than a 90% decrease in the amygdalin content compared to that in 3-month samples, and at least a 25% reduction compared to the samples fermented for 9 months (
Figure 3). The
maesil sugar syrups prepared from the deseeded fruit in both ripe and unripe
maesil sugar syrups contained a lower amygdalin content than all other
maesil sugar syrups. The lowest amygdalin content (1.45 ± 0.29 mg/L) was found in R1, which also had the greatest decrease in amygdalin content, at 96.54% and 54.26% for the 3- and 9-month samples, respectively, followed by U1 (2.99 ± 0.40 mg/L) with a 95.38% decrease in the amygdalin level compared to the 3-month U1 sample. U2, U3, R2, and R3 contained nominal amounts of amygdalin: 4.37 ± 0.09, 4.82 ± 0.12, 4.45 ± 0.34, and 4.49 ± 0.22 mg/L, respectively, which are not significantly different. Compared with the 3-month fermented samples, amygdalin reduction of 93.09%, 92.72%, 90.49%, and 91.07% was found in U2, U3, R2, and R3, respectively. The amygdalin content decreased steadily in both ripe and unripe
maesil sugar syrups between 3 and 12 months, and the rate was similar for ripe and unripe
maesil sugar syrups. The outcome of other
maesil-based studies revealed that a gradual degradation of amygdalin occurs after 3 months to 1 year of the fermentation process [
18], which supports the present findings.
The decrease in amygdalin concentration from 9 months (3–12 mg/L) to 12 months (1–5 mg/L) was relatively minor compared to the more substantial reduction observed from 6 months (16–35 mg/L) to 9 months (3–12 mg/L). Despite this smaller absolute reduction (in mg/L), the relative decrease in amygdalin content between 9 and 12 months was still significant, ranging from 24% to 60% (
Figure 3). A possible reason for this could be the low concentration of amygdalin in the syrup relative to the number of microbes involved in its degradation. These microbes may require a minimum threshold level of amygdalin to initiate their degradation process. When the amygdalin concentration falls below this threshold, the microbes might switch to utilizing other substrates present in the
maesil sugar syrups. Consequently, microbial growth could continue to increase throughout the study period in all
maesil sugar syrups. This is supported by our previous report, which demonstrated that the growth of
Z. rouxii, the primary microbe involved in
maesil sugar syrup fermentation, increased from 3.6 log CFU/mL at 3 months to 5.6 log CFU/mL at 12 months [
9].
The present results show that the amygdalin content of all
maesil sugar syrups was less than 5 mg/L at the end of 1 year of fermentation, which is at least 27-fold lower than that reported by Go et al. (2018), who revealed 134.98 ± 5.96 mg/L and 166.82 ± 4.16 mg/L amygdalin content in
maesil sugar syrups with and without fruit, respectively [
2]. The difference in amygdalin concentrations of the
maesil sugar syrups might be due to variations in the
P. mume cultivar, fermentation environment, microflora, fruit maturity and processing method and is supported by the findings of Son et al. (2017) [
8]. Although Go et al. (2018) reported a higher quantity of amygdalin (134.98 to 166.82 mg/L) compared to the levels found in the present study (63 to 67 mg/L), they noted that this concentration of amygdalin was not toxic to rats [
2]. According to EFSA (European Food Safety Authority), the complete degradation of 1 g amygdalin can produce 59 mg of HCN [
19]. That mean theoretically, approximately ~17 mg/L amygdalin will produce 1 mg of HCN. In this way, the theoretical yield of HCN from the detected amygdalin quantities (>17 mg/L) in the 9- and 12-month fermented
maesil sugar syrups in our study is below the threshold for acute and chronic toxicity to humans (>1 mg/kg of HCN) as per described by FSANZ (Food Standards Australia New Zealand) [
20]. There is currently no global regulatory entity that has established a maximum allowable limit for amygdalin in food for human consumption. The toxicity of amygdalin, resulting from the release of HCN, can vary depending on individual susceptibility. In general,
maesil sugar syrups are consumed after a 10-fold dilution with water. Thus, it is assumed that our 3- and 6-month fermented
maesil sugar syrups would not cause toxicity after dilution. An adequate reduction (87.35%) in amygdalin content was attained at the 9th month for the overall (both ripe and unripe)
maesil sugar syrups with the least amygdalin content (<13 mg/L). Thus, the present study suggests that a minimum of 9 months of fermentation of
maesil sugar syrups should be conducted, both for ripe and unripe
maesil, to deliver a good traditional taste with low amygdalin content.