Dynamic changes of the content of biogenic amines in Chinese rice wine during the brewing process

Authors


Abstract

This study examined the biogenic amine content of the brewing raw material (wheat Qu, seed starter, mash or liquid) during fermentation and during the maturation and storage process. Pre-column derivatization and high-performance liquid chromatography were employed to analyse the dynamic changes in biogenic amines during the Chinese rice wine brewing process. Some aliphatic amines (putrescine, cadaverine, spermine) were already found to be present in the glutinous rice and wheat Qu, but the content was below 2.88 mg/kg. None of the amines were detected in the brewing water. The biogenic amine content in the seed starter exhibited a large variation range (16.43–87.72 mg/L), which could influence the total content of biogenic amines prior to fermentation. A much more obvious increase in biogenic amines stemmed from the fermentation process, with the presence of a biogenic amine-producing bacteria and precursor amino amines. In the first 2 days, the biogenic amine content increased sharply, followed by a slow increase over the next 4 days, and the content reached a maximum value on day 6; subsequently there was a decrease or fluctuation in the concentration. After clarification and cooking, the biogenic amine content increased significantly; however, a slight decrease was observed during storage, especially for putrescine and histamine. Putrescine and tyramine were the most prevalent amines in all of the samples. Overall, the content of biogenic amines was found to be within the safe level for human health, ranging from 54.52 to 146. 35 mg/L. Copyright © 2013 The Institute of Brewing & Distilling

Abbreviations used
A

glutinous rice

B

wheat Qu 1

C

wheat Qu 2

D

seed starter

E

water

F

canning

G

main fermentation

H

post fermentation

I

clarification

J

cooking

K

pottery jar

BAs

biogenic amines

MF

main fermentation

PF

post fermentation

Introduction

Chinese rice wine is a low-alcohol drink, of ancient Chinese origin, and is one of the most popular alcoholic beverages in China, with an annual consumption of more than 2 million kL. Chinese rice wine is typically fermented with a variety of microorganisms (mycetes, yeast, bacteria), using glutinous rice, which contains abundant nutrients and bioactive compounds, such as amino acids, active polypeptides, vitamins, polyphenols and minerals. Hence, its nutritional value is high. Consumers choose Chinese rice wine for its health properties, but they must also pay attention to its safety. As the fermentation process is traditional and spontaneous, involving fermentation by multiple fungi over a long time period, the quality depends mainly on the raw materials, yeast strains, brewing conditions and maturation conditions. The biogenic amines (BAs) found in Chinese rice wine are a fermentation by product and have been detected in several studies [1-3].

Biogenic amines are low-molecular-weight compounds, generally found in fermented foods and fishery products, such as dairy products [4], meat products [5], aquatic products [6], beer [7], wine [8-10], Sake rice wine [11], Korean turbid rice wine [12] and Chinese rice wine. The BAs of most concern in beverages are histamine, tyramine, cadaverine, 2-phenylethylamine, tryptamine, putrescine, spermine and spermidine. The presence of BAs is becoming increasingly important to consumers, producers and researchers alike, for two reasons: first, the levels of BAs can be used as a useful indicator of poor hygienic conditions during the brewing procedure [13-15]; second, BAs are one of the causes of headaches and vertigo, indicating intoxication, and in certain instances poisoning may occur when ingested foods contain relatively high levels of BAs, especially as the combined action of alcohol, acetaldehyde and BAs can result in some sensitive consumers requiring monoamine oxidase-inhibiting drugs. While the most extreme reaction to tyramine is at concentrations of 100–800 mg/kg, an intake of 8–40 or 40–100 mg may cause slight or intermediate poisoning, respectively [16-18]. Currently in the European Union, discussions are ongoing regarding the regulation of BAs in imported wine [19].

Studies have confirmed that the content of BAs in Chinese rice wine is much higher than what is found in some other wines. The BAs in wine can originate from the grapes and are mostly formed by the action of active decarboxylases in yeast and lactic acid bacteria during wine-making, aging and storage [20-24]. However, there are few similar studies investigating BAs in Chinese rice wine, and the small number of studies currently ongoing have focused on the qualitative and quantitative analysis of BAs in commercial Chinese rice wines. Cadaverine, histamine, tyramine, spermidine and spermine were first detected in different rice wines by Lu et al. [1], with the average total of BAs found to be 114.45 µg/mL. Zhong et al. [3] investigated the level of nine amines in two types of Chinese rice wine (semi-dry and semi-sweet), with the most prominent amines being serotonin, putrescine and tyramine; the total content was variable, ranging from 29.3 to 260 mg/L, with 115 mg/L on average. Since both indigenous bacteria and abundant amino acids are widely present in the brewing process [25-28], these serve as a basis for BA biosynthesis.

The aim of this study was to investigate, using high-performance liquid chromatography (HPLC), how the BA content changes during the brewing process of a Chinese rice wine, and to present a theory detailing how to decrease the BA content.

Materials and methods

Samples and vinification

Shaoxing rice wine, manufactured in the Shaoxing region of the Zhejiang province, is a well-known Chinese rice wine. It is fermented from steamed glutinous rice and water from Jianhu Lake, with a starter culture of ‘wheat Qu’ as the saccharification agent and fermenting yeasts. Glutinous rice (labelled A) is the best raw material because of its amylopectin structure, as the arrangement of molecules is relatively loose, and so it is easy to paste. ‘Wheat Qu’, similar to the koji of sake, is made from wheat. According to the different inoculation strains, two types of wheat Qu, natural inoculation (wheat Qu 1; labelled B) and single strain inoculation (wheat Qu 2; labelled C) are used to make Shaoxing rice wines.

Wheat Qu also functions as a flavouring and toner agent, and contributes to aroma as well as flavour formation. The seed starter (labelled D) is unique to alcohol beverage fermentation. Its preparation involves mixing steamed sticky rice, 10% (w/w, weight ratio of wheat Qu and sticky rice) wheat Qu and 200% Jianhu Lake water (labelled E) at 25°C, then 3% of a selected pure yeast is added to the mixture after adjusting the pH of the fermenting mash to pH 4 with lactic acid. Subsequently, the mixture (labelled F) is cultured at about 25°C for 48 h, where saccharification and fermentation occur simultaneously [29-31], a characteristic typical of Chinese rice wine fermentation.

The main fermentation (labelled MF, in the 35 m3 tank) is typically carried out at 28–30°C for 4 days with intermittent oxygen filling. On the fourth day the MF is complete with the product temperature dropping to 15°C (G1–G4), and the post-fermentation (labelled PF, in the 125 m3 tank, merging four 35 m3 main fermenta) is carried out at room temperature for 20 days or so (H1–H3).

After PF, the rice wine mash is first filtered by pressing, and then the slightly turbid rice wine is pumped to the fining tank for clarification (labelled I). Caramel, which determines the particular colour of Chinese rice wine, is added to the clarified rice wine, and then fresh rice wine is cooked using steam at 88–90°C for 3 min (labelled J).

Finally, the cooked rice wine is aged in sealed pottery jars to develop a balanced aroma. Most Chinese rice wines are aged for 13 years, but some are aged for 3–5 years or a little more (labelled K1, K2). The brewing process for Chinese rice wine is shown in Fig. 1.

Figure 1.

The brewing process for Chinese rice wine.

Samples were taken during fermentation, maturation and the storage process (the above, marked with capital letters, are summarized in Table 1). Samples were taken in triplicate, with solid samples stored at 4°C and liquid samples stored, after filtration, at −20°C. The fermentation mash was regularly sampled from the top, middle and base of the mixed fermentation. The main fermented samples were taken respectively from four tanks merging into one post fermenta.

Table 1. Samples taken during the production of Chinese rice wine
NumberSampleNumberSample
AGlutinous riceG3Day 3 in main fermentation
BWheat Qu 1G4Day 4 in main fermentation
CWheat Qu 2H1Day 6 in post fermentation
DSeed starterH2Day 13 in post fermentation
EJianhu Lake waterH3Day 21 in post fermentation
FThe mash without fermentationIFresh rice wine after clarification
G1Day 1 in main fermentationJFresh rice wine after cooking
G2Day 2 in main fermentationK1Rice wine in pottery jar – 3 years old
  K2Rice wine in pottery jar – 6 years old

Chemicals and standards

Tryptamine, 2-phenylethylamine, cadaverine dihydrochloride, putrescine dihydrochloride, histamine, spermidine, spermine 1,7-diaminoheptane (internal standard, IS) and dimethylamino naphthalene-5-sulfonyl chloride (derivating agent) were obtained from Sigma (Switzerland, USA). Methanol, acetone, n-butyl alcohol, chloroform and trichloroacetic (all HPLC grade) were supplied by J. T. Baker (USA). The other reagents (sodium hydroxide, sodium chloride, sodium bicarbonate, sodium glutamate, hydrochloric acid and diethyl ether) were supplied from Merck (Darmstadt, Germany).

Sample preparation

A 10 g aliquot of a powdered solid sample was accurately transferred into a 100 mL conical flask, with a cover; 20 mL of trichloroacetic (5%) and 2.0 mL of 1,7-diaminoheptane (100 mg/L) were added and mixed with vibration for 60 min. The samples were transferred into a 50 mL centrifuge tube, centrifuged at 3600 rpm/min, and then the supernatant transferred into a 50 mL volumetric flask. The samples were extracted again in a constant volume with trichloroacetic (5%). Finally, in order to filter the extracted solution, the method of Li et al. [9] was used for the preparation of liquid samples and derivatization of all the extracted solutions, except that 1 mL of the derivatizing reagent (5 mg/mL) was added to 0.5 mL of sample.

HPLC determination of biogenic amines

The content of the BAs for the 51 samples was determined according to the method of Li et al. [9]. The chromatographic conditions employed were as follows: an Agilent 1100 HPLC system with diode array detector, a C18 column (150 × 4.6 mm i.d., particle size 5 µm) was obtained from Agilent. The column temperature was 30°C with a 1.5 mL/min flow velocity. A 20.0 μL aliquot of sample was used and the measurement wavelength was 254 nm. The elution system is shown in Table 2.

Table 2. Gradient elution programme for biogenic amines analysis
  Time (min)
  0714202730353645
AMethanol (%)55657070901001005555
BWater (%)4535303010004545

Determination of fermentation parameters

Total sugar, acidity and the alcohol content of the fermentation mash were measured according to the National Standard of the People's Republic of China for Chinese rice wine [32].

The bacterial counting method

The colony count was determined after the gradient dilution of 1 mL fermentation broth to 10−1, 10−2, 10−3, 10−4, 10−5 and 10−6, drawing a 100 μL sample of each dilution onto improved MRS medium [33] (honey 10 g/L, beef extract powder 7 g/L, peptone 10 g/L, yeast extract powder 5 g/L, arginine 1 g/L, folic acid 2.0 g/L, sodium acetate 1.5 g/L, glucose 20 g/L, malt root extract 0.6 g/L, cycloheximide 0.04 g/L, MgSO4 0.10 g/L, MnSO4 0.04 g/L, agar 12 g/L) and incubatiing at 30°C for 2 days under anaerobic conditions.

Statistical analysis

All data were expressed as means ± SD. Statistical analyses were performed with the SPSS 10.0 and Origin B8.0 statistics package programs.

Results and discussion

Method verification

The samples were identified by comparison with the retention times of BAs and standard solutions, that is, the retention times were determined by injecting individual and mixed reference standards. Fig. 2(A and B) demonstrates the good peak resolution. Quantification was performed using the internal standard method, based on linear regression analysis of the area ratio (BA and the inner standard) vs the concentration of the standard solutions. The regression equation, linear range and linearity are shown in Table 3. This method was linear at concentrations ranging from 1.25 to 80 mg/L, except for tyramine (0.05 to 15 mg/L) and histamine (0.50–80 mg/L), and all linearities were >0.99.

Figure 2.

Typical chromatograms of (top) biogenic amines (BAs) in standard solution and (bottom) in one sample.

Table 3. The performance characteristics of the method for the determination of biogenic amines (BAs)
BAsRange (mg/L)Linear equationLinearity (R2)Limit of detection (signal-to-noise ratio = 3) (mg/L)Repeatability (RSD%)Recovery (%)
  1. His, Histamine; Tyr, tyramine; Put, putrescine; Cad, cadaverine; Try, tryptamine; Phe, 2-phenylethylamine; Spd, spermidine; Spm, spermine.

Try0.50–20y = 0.76x + 0.00080.99990.191.84101.7
Phe1.25–80y = 2.08x − 0.02400.99900.171.4795.3
Put1.25–80y = 4.26x − 0.00820.99940.102.5597.5
Cad1.25–80y = 4.46x − 0.03130.99800.112.6494.9
His0.50–80y = 3.64x + 0.05990.99970.103.11105.9
Tyr1.25–80y = 3.22x + 0.14660.99930.161.94102.1
Spd1.25–80y = 4.74x + 0.06530.99840.122.8192.4
Spm1.25–80y = 3.71x − 0.03720.99960.131.9696.8

The limit of detection (LOD, defined as three times the signal-to-noise ratio) of eight amines ranged from 0.10 to 0.19 mg/L. The precision of this method was confirmed by the significant figures for repeatability (1.47–3.11%, n = 4) and for recovery (92.4–105.9%). The repeatability was estimated by relative standard deviation (RSD) for five times pre-treatments of the same sample. The recovery was examined in a rice wine sample, with an added 10 mg/L of mixed reference standards. Overall, this method is well defined for the determination of these eight amines in Chinese rice wine [1, 2], although serotonin (lacking any definitive evidence of causing migraine headaches) could not be detected [34].

Technological parameters

The technological parameters are shown in Table 4. The level of acidities met the requirements of brewing (seed starter ≤ 4.6, mash after MF ≤ 6.1, mash after PF ≤ 6.4). Except for the alcohol of the mash after MF, the requirements concerning alcohol content were met (seed starter ≥ 9.0, mash after MF ≥ 14.0, mash after PF ≥ 16.0). A previous study demonstrated that significant correlations were detected between the BA concentrations and acidities [35].

Table 4. The fermentation parameters
StepPractical data
Total sugar (g/L)Alcohol content (v/v%)Total acidity (g/L)
Seed starter 10.63.4
Mash after Main fermentation12.513.64.4
Mash after post-fermentation4.417.83.8

Biogenic amine in raw material, wheat Qu and the seed starter

The results are shown in Fig. 3. No amines were detected in the water of Jianhu Lake, while putrescine and spermine were detected in wheat Qu 2. Only putrescine, cadaverine and spermine were found in glutinous rice and wheat Qu 1; other amines (histamine, tyramine, tryptamine, 2-phenylethylamine and spermidine) were not found. Eight types of amines were found in all seed starters.

Figure 3.

(A) Putrescine distribution map in samples A–D; (B) spermine distribution map in samples A–D; (C) cadaverine distribution map in samples A, B and D; (D) total BA distribution map in samples A–D; (E) total BA distribution map in sample D.

The content of the BAs did not exceed 2.88 mg/kg in glutinous rice and wheat Qu. However, the percentage composition of putrescine among the eight amines was highest, ranging from 69.46 to 90.65%. The boxplot shows the graphical dispersion of the dataset and it can be seen that the level of dispersion within the glutinous rice dataset was higher than that in the two wheat Qus; hence, the difference in the BA content depended on the different glutinous rice samples.

Some aliphatic amines (putrescine, cadaverine, spermine) were already present in the glutinous rice and wheat Qu, but production of aliphatic amines is closely related to hygiene conditions [36]. Similarly, these amines were detected in soybean, where the levels were affected by cultivars in different ways over consecutive years [37], and the barley variety had a detectable effect on the amine content [38]; putrescine and other polyamines may already also be present in grapes [39, 40].

Fig. 3(E) shows the boxplot graph of eight amines and the total content dataset for seed starters, illustrating the mean, the interquartile range and the dispersion. Putrescine and tryptamine were the most prevalent amines, followed by tyramine; their average contents were 21.89, 11.06 and 8.08 mg/L, respectively. The high level of dispersion led to a high level of BA content in the different seed starters; the total content range was 16.43–87.72 mg/L.

To calculate the total BA content in the 35 m3 main fermenta, the amounts of added raw materials and the BA content are listed in Table 5. The added content of BAs from the seed starter was 79.24 g; hence, this is a major factor influencing the content of BAs before MF, followed by the glutinous rice, and the effect of the wheat Qu is the smallest.

Table 5. Content of BAs in the raw materials going into the 35 m3 main fermenter
 ABCDE
  1. A, Glutinous rice; B, wheat Qu 1; C, wheat Qu 2; D, seed starter; E, water.

Concentration of BAs (mg/kg)2.880.790.7666.030
Additive amount (kg)105001620168120013130
Content of BAs (g)30.241.280.1379.240

Changes in BAs in Chinese rice wine during fermentation

Table 6 shows the profile of BAs during the fermentation of Shaoxing rice wine. Eight amines were all present in 12 samples with 100% detection. Putrescine and tyramine were the most prevalent amines, followed by histamine, cadaverine and tryptamine. The percentages of putrescine and tyramine were 32–54 and 15–41%, respectively; the percentages of other amines were <10%.

Table 6. Change in BAs during Shaoxing rice wine fermentation (mg/L ± standard deviation)
SampleTryPhePutCadHisTyrSpdSpmTotal
  1. Different letters in columns denote significant differences (p < 0.05).

  2. F, Canning; G, main fermentation; H, post fermentation.

F2.55 ± 0.24a2.33 ± 0.49a19.49 ± 2.12a3.15 ± 0.94a4.35 ± 0.36a17.55 ± 3.16a2.37 ± 0.65a2.74 ± 0.46b54.51a
G16.22 ± 0.27b4.20 ± 0.90b41.31 ± 3.90bc4.40 ± 0.57b6.83 ± 0.57b34.05 ± 2.99b4.24 ± 0.37b3.69 ± 0.27c103.94b
G27.74 ± 0.83bc4.98 ± 0.48c52.54 ± 4.33cd5.81 ± 0.90c7.11 ± 1.04c36.81 ± 2.70bc5.60 ± 0.65d4.28 ± 0.29d125.87c
G38.95 ± 0.84cde5.63 ± 0.42d52.73 ± 5.95cd5.85 ± 0.95c8.01 ± 0.47c37.93 ± 3.13bc5.89 ± 0.12d4.53 ± 0.56d130.52cd
G410.40 ± 1.39de5.67 ± 0.02d55.67 ± 3.04cd6.00 ± 0.77c9.20 ± 0.06d44.65 ± 5.32c5.88 ± 0.23d4.65 ± 0.49d142.13d
H110.81 ± 0.13 e5.88 ± 0.02d55.80 ± 3.M4d5.61 ± 0.03c8.30 ± 0.25c50.70 ± 0.33c4.93 ± 0.27c4.32 ± 0.28d146.35d
H29.24 ± 0.21cde5.33 ± 0.53cd37.51 ± 2.74b4.45 ± 0.02b6.83 ± 0.67b47.16 ± 1.98c2.43 ± 0.07a2.53 ± 0.06b115.48bc
H38.81 ± 0.42cd4.10 ± 0.53b56.70 ± 3.18d5.64 ± 0.05c7.08 ± 0.53b18.48 ± 0.98a2.04 ± 0.04a1.31 ± 0.04a104.18b

Fig. 4 displays the changing curve of the eight amines and the total BA content. In general, the concentration of the eight amines and total BA content sharply increased at the early stage, followed by a slower increase, but a reduction or fluctuation occurred as the fermentation proceeded. At 0–2 days, the eight amines all underwent an increase in concentration, accompanied by a significant difference (p < 0.05), in the next 3–6 days, the rate of increase becoming smaller and smaller. Only spermidine increased to a maximum content on the third day (G3). The content of cadaverine, histamine and spermine increased to a peak on the fourth day and the others reached a maximum content on day 6, except for putrescine. Diverse changes occurred depending on the amines, especially after reaching the maximum content, for example, the putrescine content presented a significant difference (p < 0.05) between day 13 (H13) and day 21 (H21).

Figure 4.

Changes in BAs during fermentation for Chinese rice wine.

The mean total BA content was 54.51 mg/L before MF (F) and 104.18 mg/L after PF (H3), giving a statistically significant increase in the region of 91%. Hence, BA biosynthesis during fermentation was the primary cause of the high content of BAs in the Chinese rice wine; this conclusion could be confirmed by the above analysis of seed starter, as its preparation was a process of saccharification and fermentation. Similarly, the formation of amines in wine was observed during alcoholic fermentation (by the action of yeast) [41] and malolactic fermentation (by the action of lactic bacteria) [21]; the production of BAs depended on the presence of both a biogenic amine-producing microorganism and the amino acid precursors [24, 42].

Biogenic amine production by lactic acid bacteria was reported in alcoholic beverages, and lactic acid bacteria are one of the main bacteria in a rice wine fermentation mash [20, 21, 23, 43-46]. The varying bacterial count during the fermentation is shown in Fig. 5. It can be seen that the bacterial count rapidly increased in the early days; a substantial reduction followed, and it kept to a low level after day 6. Fig. 5 also highlights the varying tendencies of the BA content and the bacterial count, which were similar at the beginning, and that the continual increase in the BAs was accompanied by a higher bacterial count, indicating that the abundant bacteria could be one of the main reasons for the high content of BAs. These bacteria probably mainly originate from the seed starter, wheat Qu, the surroundings and the surfaces of the equipment. Elucidating which strains produce BAs and the metabolic mechanism of the BA-producing strains could be key to controlling amine formation [23, 47]. Likewise, changes in the amino acid content during fermentation are dynamic; on the one hand, the content increased with continuous proteolysis and yeast autolysis, while on the other hand, the amino acids were consumed to meet the needs of the growing yeast and bacteria. Previous research has indicated that abundant amino acids did not limit BA production [28, 29].

Figure 5.

Changes in bacterial count in fermentation for Chinese rice wine.

Biogenic amines can be oxidized to aldehydes by monoamine and diamine oxidases, which are widely present in these microorganisms [48]. The degradation may bring about irregular decreases. Histamine oxidase and tyramine oxidase have been indicated in preventing the accumulation of histamine and tyramine in fermented foods, and the action of these enzymes would be hindered by an acid pH and oxygen [49-51]. Hence, intermittent filling with oxygen during MF can inhibit the activity of oxidases, while the inhibition can be removed during PF by refraining from filling with oxygen. Ohya [52] found Shaoxing rice wine (Chinese rice wine) to contain 4-methylspinaceamine, a Pictet − Spengler condensation reaction product of histamine with acetaldehyde. From the genetic angle, Atkinson et al. [53] studied the histidine utilization operon in Bacillus subtilis, and Pessione et al. [54] proposed the existence of an operon (homologous to the histidine utilization operon) in Lactobacillus sp. (30a and w53).

Changes in BAs in Chinese rice wine during the maturation and storage process

Table 7 summarizes the content of the BAs during maturation and storage. The putrescine concentration ranged from 33.99 to 62.60 mg/L, that of tyramine from 15.70 to 36.37 mg/L and that of histamine from 2.68 to 9.39 mg/L. Putrescine was still the most abundant amine, just as in the glutinous rice, seed starter and fermentation mash, whereas tyramine predominated in the six-year-old rice wine (K2). This result was not surprising as it is an intermediate in the biosynthesis of polyamines [55].

Table 7. Change of BAs in Chinese rice wine during maturation and storage (mg/L ± standard deviation)
SampleTryPhePutCadHisTyrSpdSpmTotal
  1. Different letters in columns denote significant differences (p < 0.05).

  2. H, Post fermentation; I, clarification; J, cooking; K, pottery jar.

H218.81 ± 0.42c4.10 ± 0.53a56.70 ± 3.18c5.64 ± 0.05a7.08 ± 0.53b18.48 ± 0.98b2.04 ± 0.04c1.31 ± 0.04a104.18b
I9.52 ± 0.34cd4.25 ± 0.06a58.62 ± 2.96cd6.89 ± 0.37bc7.90 ± 0.74c33.20 ± 1.23c0.52 ± 0.02a1.34 ± 0.03a122.26c
J10.12 ± 0.26d5.28 ± 0.13c62.60 ± 3.01d6.65 ± 0.42c8.83 ± 0.31c43.00 ± 0.65e0.96 ± 0.02b1.56 ± 0.36ab139.01d
K10.85 ± 0.15a8.00 ± 0.43d50.45 ± 1.19b15.79 ± 0.32d9.39 ± 0.41c15.70 ± 0.91a1.12 ± 0.34b1.51 ± 0.09b102.80b
K24.37 ± 1.08b5.41 ± 0.68c33.99 ± 2.80a7.11 ± 0.28b2.68 ± 0.99a36.37 ± 1.40d3.07 ± 0.37d2.48 ± 0.88b95.46a

After clarification and cooking, the total BA content increased significantly (p < 0.05), largely in the form of tyramine, followed by putrescine; all the other amines, except spermidine, increased slightly. Microbial amino acid metabolism may occur when using an environmentally friendly transformation for the production of BAs, [22]; in addition, there is the possibility that the concentration gradually increased owing to evaporation or filtration. However, it was observed that, as the storage time increased, the total BA content decreased, along with a dramatic reduction in putrescine and histamine (p < 0.05), probably owing to their degradation by the action of oxidases. The same pattern of histamine stability in Rihaakuru was observed by Naila et al. [56]. In the case of spermidine and spermine, a continuous increase in the content was observed with aging. For tyramine, tryptamine and 2-phenylethylamine, there was no apparent correlation between aging and concentration. From these results it can be seen that the process of storage is advantageous in controlling the content of BAs in Chinese rice wines.

To date, there have been no cases of BA poisoning from Chinese rice wine and all of the results from this research were below the harmful level of 1000 mg/kg; however, owing to the potential harm to the consumer, close attention must be paid to this limit. Therefore, reliable methods are needed to support the efforts of manufacturers in limiting amine formation, including adequate sanitary conditions. More importantly, it is necessary to select autochthonous strains, which are incapable of producing BAs. This would enable the creation of multi-strain starters for use in industrial production and allow successful fermentation, while ensuring that the customer is not exposed to any toxic effects. Commercial Oenococcus oeni starter cultures, which were unable to produce BAs [21], have been inoculated into malolactic fermentation for the control of BAs in wine, and the combined use of adequate preservatives and starter cultures (Lactobacillus plantarum together with Kocuria varians) allowed the production of safer salami products [57]. Further studies are currently underway to address this issue.

Acknowledgements

This work was supported by the Ministry of Science and Technology, People's Republic of China under grant no. 2012BAK17B11, and the Beijing Natural Science Foundation (5132029).

Ancillary