Changes in Odor Compounds of a Lactobacillus-fermented Dairy Beverage during Two Weeks of Refrigerated Storage

Abstract

Lactobacillus paracasei MCC1849 (LP) has the potential to modulate immune function. To develop a functional LP-fermented dairy beverage, changes in intensities of odorants during two weeks of refrigerated storage were investigated by gas chromatography-olfactometry (GC-O, CharmAnalysis™). A fermented dairy beverage containing 1 × 10⁸ viable LP cells/mL, 0.8% non-fat milk solids, and 0.1% milk fat was prepared, and then stored at 10°C for 0, 1, or 2 weeks before solvent extraction. GC-O of the volatiles in the resulting extracts detected 34 odorants, of which 29 were identified. Among the identified compounds, 4-(4′-hydroxyphenyl)-2-butanone (raspberry ketone) was also detected by GC-mass spectrometry, representing the first identification of this compound as an odorant in fermented dairy products to our knowledge. Principal component analysis of the GC-O data permitted discrimination among the 3 stored samples according to duration of storage, and clarified the major odors contributing to the change in odor characteristics during storage.

Introduction

Lactic acid bacteria (LAB) have long been used around the world for the fermentation of foods, as exemplified by yogurt and cheese. LAB in yogurt has been shown to enhance intestinal function by improving the intestinal microflora. More recently, several food products have been launched based on claims of additional biomodulation functions associated with specific bacterial strains.

Based on its immunostimulatory effects, Lactobacillus paracasei MCC1849 (LP) was selected by Morinaga Milk Industry Co., Ltd. (2014; 2015) from among many species of LAB; LP material has been proposed for use in food and beverage products, and as a dietary supplement. For instance, Maruyama et al. (2016) described the effects of non-viable LP on immune function in the elderly. In other work, non-viable LP was reported to enhance the function of follicular helper T cells (The University of Tokyo and Morinaga Milk Industry Co., Ltd., 2016).

Currently, LP is widely used in a range of comestibles, ranging from dairy products such as yogurt and ice cream to non-dairy products such as supplements, miso soup, chocolate, beverages, and bread in the Japanese market. These uses are based on the observation that LP has an immunoactive function even in a heat-killed bactericidal (non-viable) form, while also permitting easy handling in manufacturing and storage of the used products.

However, LP-based yogurt and a corresponding fermented dairy beverage (LAB beverage) are generated using fermentation, therefore requiring the use of viable LAB. While a defined bacterial density is needed to ensure functionality, the resulting fermented product may exhibit problematic changes in the taste and aroma (flavor) profile during subsequent storage.

We continue to pursue the development of functional LP-fermented dairy beverages. Therefore, this study was conducted to reveal changes in odor compounds of the beverage, given that such changes may affect the flavor during refrigerated storage.

This study consisted of 3 parts. First, odor compounds of samples collected following refrigerated storage for up to two weeks were analyzed using GC-O (CharmAnalysis™). Secondly, the changes in the odor compounds during storage were investigated. Finally, changes in the odor characteristics during storage and the odors contributing to the change were assessed by principal component analysis (PCA).

Materials and Methods

Specification design of an LP-fermented dairy beverage    Our program targets the development of 100-mL LAB beverages (non-fat milk solids [NFMS] < 3% w/w). Under the Food Sanitation Act of Japan, such a beverage (NFMS < 3% w/w) needs to contain viable LAB at a density exceeding 1 × 10⁶ cells/mL. In addition, LP of ≥ 1 × 10¹⁰ cells per 100-mL product (≥ 1 × 10⁸ cells/mL) is required for immunostimulatory activity.

The pH of the beverage product was set at 3.8, which is the optimal pH of the stabilizer used for the product. When the pH was less than 3.8, the survival of LP during storage was poor, acidity increased, and the resulting beverage became less palatable. Based on this preliminary observation, a pH of 3.8 or higher is needed in terms of product design. In addition, the acidity of the product was set after considering the acid taste of the beverage.

Sample preparation    A fermented dairy beverage containing 1 × 10⁸ cells/mL viable LP (Morinaga Milk Industry Co., Ltd.) with 0.8% NFMS and 0.1% milk fat was prepared. The number of LP per a 100-mL product was set at 1 × 10¹⁰ cells, which is the number required to exhibit the potential for immune function modulation. High-fructose starch syrup containing sugar NF55S (12.70% w/w; Showa Sangyo Co., Ltd., Tokyo, Japan), skim milk (0.89% w/w; Morinaga Milk Industry Co., Ltd.), unsalted butter (0.17% w/w; Morinaga Milk Industry Co., Ltd.), maltodextrin MAX1000 (0.09% w/w; Matsutani Chemical Industry Co., Ltd., Hyogo, Japan), citric acid (0.16% w/w; San-Ei Gen F.F.I., Inc., Osaka, Japan), monosodium fumarate (0.02% w/w; San-Ei Gen F.F.I., Inc.), soybean polysaccharides SM-1200 (0.15% w/w; San-Ei Gen F.F.I., Inc.), and high-intensity sweeteners SANSWEET^® SA-5050 (0.002% w/w; San-Ei Gen F.F.I., Inc.) were each dissolved in water, and then blended together. The blended mixture was warmed to 60°C, homogenized at 14 MPa using a homogenizer (type H3-1D INV; 200 L/H, Sanmaru Machinery Co., Ltd., Shizuoka, Japan), and then batch-pasteurized at 80°C for 10 min. Following pasteurization, the mixture was cooled to 10°C, and frozen viable LP was added. The resulting suspension was transferred aseptically into 2-L polyethylene terephthalate (PET) bottles (Toyo Seikan Co., Ltd., Tokyo, Japan), which have lower gas permeability among plastic containers. The containers were stored at 10°C for 0, 1, or 2 weeks.

General analysis

LAB count:    LAB counts were determined according to the methods for measuring LAB in cultured milk and LAB beverages as defined by the “Ministerial Ordinance on Milk and Milk Products Concerning Compositional Standards, etc.” of the Food Sanitation Act of Japan. Briefly, samples of LAB dairy beverages stored at 10°C for 0, 1, or 2 weeks were subjected to 10-fold serial dilution, and each diluted sample was mixed with bromocresol purple (BCP) plate count agar medium (BCP-PCA; Eiken Chemical Co., Ltd., Tokyo, Japan). After incubation of plates at 35–37°C for 72 ± 3 h, the number of yellow-colored colonies among the colonies that grew was counted as LAB. The measurements were performed in triplicate. Means of the measured counts are shown in Table 1.

Table 1. Lactic acid bacterial count, pH, and acidity of LP-fermented dairy beverages stored for 2 weeks at 10°C

	0 week	1 week	2 weeks
Lactic acid bacterial count (×107/mL)	9.8 ± 1.0	4.6 ± 2.9	4.4 ± 0.6
pH	3.78 ± 0.02	3.78 ± 0.00	3.73 ± 0.01
Acidity (lactic acid%)	0.24 ± 0.00	0.25 ± 0.00	0.26 ± 0.00

The values are means ± S.D. (n = 3).

pH measurement:    Measurements of pH and acidity were performed as described by Murakami et al. (2010). pH values were measured using a pH meter (F72; Horiba, Ltd., Kyoto, Japan) at a sample temperature of 20°C.

Acidity measurement:    Each sample (10 g) was weighed into a 100-mL beaker. The sample was titrated, with stirring, to an end-point of pH 8.3 using 0.1 mol/L NaOH. Lactic acid acidity (% w/w) was calculated based on the titer of NaOH, using Eq. 1:

  

Solvent extraction of volatile compounds    A sample (200 g) of the 0, 1, or 2-week stored beverage was transferred to a 500-mL conical flask. Volumes of 4-octanol (2 µg/mL in methanol; 1 mL) and 3-heptanol (10 µg/mL in methanol; 2 mL) were added as internal standards, and dichloromethane (DCM, 200 mL) was then added slowly. The mixture was stirred gently with a magnetic stirrer for 1 h at room temperature, then separated in a separating funnel and dried over anhydrous sodium sulfate. The resulting extract was concentrated in a rotary evaporator to a final volume of approximately 1 mL.

GC-O analysis    The concentrated extract was subjected to 3-fold serial dilution (1:3 to 1:81) with DCM. GC-O analysis of the extract and of the DCM dilutions thereof was performed in triplicate using the CharmAnalysis™ with an Agilent 6890GC (modified by DATU, Inc., Geneva, NY, USA) (Acree et al., 1984). A DB-WAX fused-silica capillary column (15 m × 0.32 mm, 0.25 µm film thickness; Agilent Technologies, Santa Clara, CA, USA) was employed with a helium carrier gas flow rate of 3.2 mL/min. The oven temperature was initially programmed to 40°C, increased at 6°C/min to 230°C, and then held at 230°C for 30 min. The injection port and detector were maintained at 225°C, and a 1-µL sample was injected in splitless mode. The GC effluent gas of each sample was sniffed in humidified air by a trained panelist with 13 years of experience in GC-O sniffing. The odor activities of volatile components obtained by GC-O dilution analyses were represented as a charm value (CMV) as well as by aroma descriptors (Acree et al., 1984). The CMV is an index of odor intensity that is calculated by integrating the time length and the dilution level, using Eq. 2:   

where F is the dilution factor, n is the number of dilutions, and di is the length of time.

Identification of volatile compounds and standard compounds    Volatile compounds were identified by comparing their mass spectra and linear retention indices (using C₆-C₂₈ n-alkanes) with those of standard compounds. The standard compounds for identification of volatile compounds were purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan), Sigma-Aldrich Japan Co., LLC (Tokyo, Japan), Frutarom (UK), Ltd. (Hartlepool, United Kingdom), or Soda Aromatic Co., Ltd. (Tokyo, Japan).

GC-mass spectrometry (MS) analysis    GC-MS analysis of the concentrated extract was performed in triplicate on a Model 5973 mass selective detector (Agilent Technologies) with a fused silica capillary column DB-WAX (60 m × 0.25 mm, 0.25 µm film thickness; Agilent Technologies). The flow of the helium carrier gas was 1.6 mL/min. The oven temperature was initially programmed to 50°C for 2 min, then increased at 3°C/ min to 220°C, and held at 230°C for 75 min. The injection port was maintained at 250°C, and a 5-µL sample was injected. The inlet was operated in split mode (split ratio 20:1).

GC-MS analysis of 4-(4′-hydroxyphenyl)-2-butanone (raspberry ketone, RK)    An aliquot (2000 g) of the 0-week sample was transferred to a 5000-mL conical flask. Volumes of 3-heptanol (10 µg/mL in methanol; 2 mL) and 4-octanol (1 µg/mL in methanol; 2 mL) were added as the internal standards, and 2000 mL DCM was added slowly. The mixture was stirred gently with a magnetic stirrer for 1 h at room temperature, then separated in a separating funnel and dried over anhydrous sodium sulfate. The resulting extract was carefully concentrated in a rotary evaporator to a final volume of approximately 0.5 mL. GC-MS analysis was performed on a Model 5973 mass selective detector (Agilent Technologies) with a fused silica capillary column DB-WAX (60 m × 0.25 mm, 0.25 µm film thickness; Agilent Technologies). The flow of the helium carrier gas was 1.6 mL/min. The oven temperature was initially programmed to 50°C for 2 min, then increased at 3°C/min to 220°C, and held at 220°C for 75 min. The injection port was maintained at 250°C. The inlet was operated in splitless mode. Ions were monitored in SIM (selected ion monitoring) mode with m/z values as follows: 107, 121, 164 for RK; 59, 69, 87 for 3-heptanol; and 69, 73, 87 for 4-octanol.

The RK odorant could not be monitored in SIM mode. Instead, for positive identification, this compound was analyzed by heart-cutting two-dimensional GC-MS using ¹D/²D GC-MS (Ochiai et al., 2012, 2014). The ¹D/²D GC-MS was configured with a Model 5975C mass-selective detector (Agilent Technologies) and Model 7890A gas chromatograph (hostGC; Agilent Technologies) equipped with a CTS2 cryo-trap system (GERSTEL, Mülheim an der Ruhr, Germany), dual low thermal mass (LTM) column modules (5 inch; Agilent Technologies), splitter, pressure control module (PCM), and Deans switch.

Separation of RK was performed using a DB-WAX column (30 m × 0.25 mm, 0.25 µm film thickness; Agilent Technologies) in the first dimension (¹D) and a DB-5 column (10 m × 0.18 mm, 0.18 µm film thickness; Agilent Technologies) in the second dimension (²D). The column temperature for ¹D DB-WAX was initially programmed to 50°C for 2 min, then increased at 5°C/min to 220°C, and held at 220°C for 30 min. After a 60-min retention time, the sample matrix was back-flushed. The column temperature for ²D DB-5 was held at 50°C during ¹D GC analysis, and then programmed to 50°C for 0.8 min, increased at 20°C/min to 230°C, and held at 230°C for 15 min for ²D GC analysis. The hostGC oven was held at a constant temperature of 250°C. The inlet pressure was 394 kPa and the pressure of PCM auxiliary was 19 kPa. The Deans switch pressure (for the ²D column) of the PCM was set at 304 kPa. The inlet was operated in splitless mode. The MS was operated in SIM mode.

Statistical analysis    PCA was conducted for 0, 1, and 2-week samples (n = 3 for each) and for the mean sample for each storage duration using the JMP8 software package (SAS Institute, Inc., Cary, NC, USA). Twelve data of total mean CMVs in each odor description were used as variables.

Results and Discussion

General characteristics of an LP-fermented dairy beverage during refrigerated storage for up to 2 weeks    LAB viable counts, pH, and lactic acid acidity of an LP-fermented dairy beverage were investigated in an effort to assess the effect of fermentation on the quality characteristics of the beverage during refrigerated storage for up to 2 weeks. The results are shown in Table 1.

The LAB viable count in the 0-week sample was 9.8 × 10⁷ cells/mL, which was lower than the intended level of 1 × 10⁸ cells/mL. After 1- and 2-week storage at 10°C, LAB viable counts decreased to 4.6 × 10⁷ cells/mL and 4.4 × 10⁷ cells/mL, respectively. Thus, although the density of viable LP decreased by 55% at 2-week storage (compared with that at 0 week), the remaining density at 2 weeks still exceeded the minimum value (1 × 10⁶ cells/mL) required by the product specifications of the LAB beverage (NFMS < 3%). Furthermore, despite the observed drop in viable cell count during storage, the functional effect of LP is retained whether the bacteria are killed or viable (Taverniti and Guglielmetti, 2011).

At 0, 1, and 2 weeks, pH values were observed to be 3.78, 3.78, and 3.73, respectively. In addition, lactic acidity exhibited a gradual but nominal increase from 0.24% at 0 week to 0.26% at 2 weeks. These results suggested that the continued LP fermentation had little influence on pH and acidity during 2 weeks of refrigerated storage. On the other hand, the sensory scores of some odor attributes such as “fermented odor”, “impure milk odor”, and “aftertaste of fermented odor” were significantly changed by LP fermentation during the 2-week storage (data not shown).

Thus, the viable cell count showed a tendency to decrease over the 2-week refrigeration period, while pH and acidity were apparently unchanged during this period. However, changes in sensory characteristics were observed. This implied that the beverage continued to ferment during 2 weeks of refrigerated storage, although this could not be estimated from pH and acidity. The results of these general analyses indicated that changes during 2 weeks of storage did not represent problems in terms of product development in the context of compliance with the regulations regarding the number of viable LAB. This interval also did not alter product stability in a pH-dependent manner, nor did the acidity affect the flavor palatability. Therefore, the properties (NFMS, pH, and acidity) of these experimental samples continued to meet the product specifications, permitting further development and analysis of changes in the odor profile of the fermented beverage.

GC-O analysis of potent odorants    To detect potent fermentation odorants, GC-O analysis using CharmAnalysis™ was conducted on LP-fermented dairy beverages stored at 10°C for 0, 1, or 2 weeks. Table 2 lists 34 potent odor compounds detected in the solvent extract obtained from the beverage samples. These 34 compounds were classified into 12 groups according to odor descriptions such as buttery-oily, coconut, potato-like, sulfur, vinegar, rancid, fatty-metallic, sweet caramel, phenolic, floral-fruity, sweet-fruity, and woody (Avsar et al., 2004; Akiyama et al., 2007, 2009; Carunchia Whetstine and Drake, 2007; Preininger et al., 2009; Cheng, 2010; Michishita et al., 2010; Antón et al., 2014; Morita et al., 2015). Twenty-nine of the 34 odorants were identified by comparing retention index and odor characteristics with standard compounds; the other 5 odorants that were not identified (indicated as “unknown” in Table 2) included one coconut odorant, one sweet-caramel odorant, and 3 phenolic odorants.

Table 2. Potent odorants in the solvent extract of the LP-fermented milk drink during 2 weeks of refrigerated storage.

No.		Description	Ref.a	Component	Retention index	Charm value
						0 week		1 week		2 weeks
						Meanb	RSDc (%)	Mean	RSD (%)	Mean	RSD (%)	Origin	Ref.d
1	-1	Buttery-oily	A	2,3-Butanedione	1030	0	-	7	48	140	15	Yogurt	A
2	-1	Coconut	B	5-Octanolide	1930	30	77	5	33	17	58	Cheese	E
	-2			Unknown	2150	0	-	15	27	27	44
	-3		B	5-Decanolide	2150	750	21	700	12	1200	19	Yogurt	B
	-4		C	4-Dodecanolide	2340	0	-	0	-	45	41	Yogurt	A
	-5		C	(Z)-6-Dodecen-4-olide	2360	250	26	210	10	810	21	Yogurt	B
	-6		C	5-Dodecanolide	2390	10	30	0	-	30	63	Yogurt	A
				Total		1040		930		2129
3	-1	Potato-like	D	2-Methyl-3-furanthiol	1300	22	32	0	-	0	-	Cheese	E
4	-1	Sulfur	D	3-(Methylthio)propanal	1430	44	27	0	-	0	-	Yogurt	A
5	-1	Vinegar	E	Acetic acid	1430	0	-	46	25	110	20	Yogurt	A
6	-1	Rancid	E	Hexanoic acid	1830	140	25	140	29	160	8	Yogurt	A
	-2		E	Octanoic acid	2030	260	10	250	8	230	8	Yogurt	A
	-3		E	Butyric acid	1610	1400	25	1600	13	1600	12	Yogurt	A
	-4		E	2-/3-Methylbutyric acid	1650	120	10	92	11	120	5	Yogurt	A
				Total		1920		2082		2110
7	-1	Fatty-metallic	D	(E)2-Nonenal	1490	150	26	120	28	270	29	Yogurt	A
	-2		D	(E,E)2,4-Decadienal	1790	61	30	140	26	180	49	Cheese	E
	-3		D	trans-4,5-Epoxy-(E)-2-decenal	1970	38	37	290	41	320	23	Kefir	D
				Total		249		550		770
8	-1	Sweet-caramel	F	3-Hydroxy-2-methyl-4H-pyran-4-one	1930	42	50	17	50	46	18	Kefir	D
	-2			Unknown	1980	130	41	54	46	69	25
	-3		A	4-Hydroxy-2,5-dimethyl-3(2H)-furanone	2010	970	23	1300	23	1000	21	Yogurt	C
	-4		A	3-Hydroxy-4,5-dimethyl-2(5H)-furanone	2160	740	17	500	18	550	14	Cheese	E
				Total		1882		1871		1665
9	-1	Phenolic	B	p-Cresol	2060	0	-	17	19	28	22	Kefir	D
	-2			Unknown	2120	51	25	0	-	24	17
	-3			Unknown	2140	73	12	14	48	95	16
	-4		A	4-Ethenyl-2-methoxyphenol	2180	0	-	13	27	38	50	(Goat milk) Cheese	F
	-5		G	2-Methoxy-4- [(Z)-1-propenyl]phenol	2230	0	-	6	37	28	36	(Goat milk) Cheese	F
	-6			Unknown	2260	0	-	54	69	150	27
	-7		G	2-Methoxy-4- [(E)-1-propenyl]phenol	2320	32	27	180	28	320	32	(Goat milk) Cheese	F
	-8		H	3-Methyl-1H-indole	2440	310	19	10	55	9	57	Cheese	E
				Total		466		294		692
10	-1	Floral-fruity	D	1-(2-Aminophenyl)ethanone	2180	70	29	0	-	0	-	Cheese	E
	-2		D	Phenylacetic acid	2520	1600	12	1500	26	1900	24	Cheese	E
				Total		1670		1500		1900
11	-1	Sweet-fruity	I	Vanillin	2520	320	6	270	11	280	28	Cheese	E
	-2		I	4-(4′-Hydroxyphenyl)-2-butanone	2950	180	16	180	38	170	32
				Total		500		450		450
12	-1	Woody	D	3-Phenylpropionic acid	2590	360	10	980	16	1400	29	Cheese	G
				Total		8153		8710		11366

^a  References: A. Michishita et al., 2010; B. Carunchia Whetstine and Drake, 2007; C. Avsar et al., 2004; D. Morita et al., 2015; E. Cheng, 2010; F. Preininger et al., 2009; G. Antón et al., 2014; H. Akiyama et al., 2009; I. Akiyama et al., 2007

^b  Mean (n=3).

^c  RSD: Relative standard deviation.

^d  References: A. Cheng, 2010; B. Zellner et al., 2008; C. Friedrich and Acree, 1998; D. Preininger et al., 2009; E. Carunchia Whetstine et al., 2005; F. Guillén et al., 2004; G. Lamparsky et al. 1981

Eight of the 34 compounds [consisting of 2,3-butanedione (no. 1-1), unknown compound (no. 2-2), 4-dodecanolide (no. 2-4), acetic acid (no. 5-1), p-cresol (no. 9-1), 4-ethenyl-2-methoxyphenol (no. 9-4), 2-methoxy-4-[(Z)-1-propenyl]phenol (no. 9-5), and unknown (no. 9-6)] were not detected in the 0-week sample, but were detected in the 1- and/or 2-week samples. On the other hand, five compounds detected in the 0-week sample [5-dodecanolide (no. 2-6), 2-methyl-3- furanthiol (no. 3-1), 3-(methylthio)propanal (no. 4-1), unknown (no. 9-2), and 1-(2-aminophenyl)ethanone (no. 10-1)] were not detected in the 1-week sample or in the 1- and 2-week samples.

In terms of the odor intensity (total mean CMVs) of each odor description, duration of 2-week storage was associated with increased intensities of buttery-oily, coconut, vinegar, rancid, fatty-metallic, phenolic, floral-fruity, and woody odors, and with decreased intensities of potato-like, sulfur, sweet-caramel, and sweet-fruity odors. Among the odors that increased in intensity from 0 week to 2 weeks, coconut, phenolic, and floral-fruity odors tended to decrease from 0 week to 1 week, and then to increase from 1 week to 2 weeks. Total intensity of all odor descriptions also increased with storage time, showing a large increase from 1 week to 2 weeks. These changes from 1 week to 2 weeks were presumably attributed to LP adaptation to the beverage storage conditions (pH, acidity, temperature, etc.) (Kaneko, 1994). Among those odors, rancid, sweet-caramel, and floral-fruity odors had much larger intensities than the other odors. Coconut and woody odors showed much larger increases in odor intensities with storage duration than did the other increased odors.

Buttery-oily, potato-like, sulfur, vinegar, and woody odors each corresponded to a single odorant [2,3-butanedione (no. 1-1), 2-methyl-3-furanthiol (no. 3-1), 3-(methylthio)propanal (no. 4-1), acetic acid (no. 5-1), and 3-phenylpropionic acid (no. 12-1), respectively]. Coconut odor originated from a combination of six odorants, with the intensity of this odor depending primarily on 5-decanolide (no. 2-3) and (Z)-6-dodecen-4-olide (no. 2-5). The intensity of rancid odor originated from four compounds and could be attributed primarily to butyric acid (no. 6-3). All three compounds that generated fatty-metallic odor increased during storage. Four of eight phenolic odorants were not detected in the 0-week sample; however, the corresponding odor was detected at 1 week and became stronger by 2 weeks. Although the odorants were not detected at 0 week, they showed an increasing tendency as observed for 2-methoxy-4-[(E)-1-propenyl]phenol (no. 9-7). All the compounds, including 2-methoxy-4-[(E)-1-propenyl]phenol, have a phenolic skeleton. From the above, this increasing tendency appeared to correlate with the structure. Among the floral fruity odorants, 1-(2-aminophenyl) ethanone (no. 10-1) was no longer detected by smell as the storage interval increased. Therefore, phenylacetic acid (no. 10-2) became the primary floral fruity odorant. The sweet-fruity odor consisted of two odorants, one of which (RK) was further characterized as described below. The intensities of these 2 sweet-fruity odorants showed virtually no changes during storage.

GC-MS analysis of RK    Many of the odorants detected by GC-O analysis in the present work have been reported previously in relation to fermented dairy products, as shown in Table 2 (Lamparsky and Klimes, 1981; Friedrich and Acree, 1998; Guillén et al., 2004; Carunchia Whetstine et al., 2005; Zellner et al., 2008; Preininger et al., 2009; Cheng, 2010). For example, 13 of these odorants were detected previously in yogurt (Cheng, 2010; Friedrich and Acree. 1998; Zellner et al., 2008), and 3 were detected previously in kefir (Preininger et al., 2009). In addition, 12 of these odorants were identified from either of two other fermented dairy products: cheddar cheese or goat milk cheese (Lamparsky and Klimes, 1981; Guillén et al., 2004; Carunchia Whetstine et al., 2005). Odors in fermented dairy products corresponded to odorants common to fermented dairy products or to odorants that were characteristic to each product (Friedrich and Acree, 1998).

Among the 34 potent odorants detected by GC-O in the present work, RK, which, to our knowledge, had not previously been identified in fermented dairy products, was further characterized by GC-MS analysis. Specifically, RK was detected by heart-cutting two-dimensional GC-MS, revealing RK to be a novel odorant component of fermented dairy products.

RK has a characteristic aroma of raspberry, and the compound was previously shown to be generated from L-phenylalanine via p-coumaric acid in raspberry fruits (Borejsza-Wyscoski and Hrazdina, 1994). Notably, L-phenylalanine is present in dairy products such as milk and cheese (Belitz et al., 2009), as is p-coumaric acid (Waterstraat et al., 2016). Based on these observations, we infer that RK is produced in dairy products through bacterial biosynthesis via p-coumaric acid.

PCA of GC-O data    PCA was performed using the 12 data of total mean CMVs for each odor description as variants in order to clarify: (1) the change and difference in overall odor characteristics during 2 weeks of refrigerated storage, and (2) the odors and odorants contributing to the change and difference in the odor profile. Four PCs that had eigenvalues exceeding 1 were obtained from the PCA. Two-dimensional (2D) scatter plots of the PC1 and PC2 scores are shown in Figure 1. The first PC (PC1) and the second PC (PC2) accounted for 53.3% and 19.4% (contribution ratios; respectively) of the total GC-MS information. The cumulative contribution ratio of both PCs was 72.7%. The samples plotted on the 2D map were markedly distinguished by duration of storage. The PC1 scores of the samples increased with storage. The PC2 scores of the 0- and 2-week samples were very similar, but the PC2 score of the 1-week sample was smaller. Each of the three samples with a given storage duration was plotted within a shared region of the graph; these samples were readily discriminated from the samples with the other storage durations.

Fig. 1.

Two-dimensional scatter plot of principal component (PC) scores by PC analysis of GC-O data obtained from a LP-fermented dairy beverage.

0 week: 0-week stored samples, 1 week: 1-week stored samples, 2 weeks: 2-week stored samples

Black symbols and open symbols correspond to actual measurement data and mean data, respectively.

The PC1 and PC2 loadings are shown in Table 3. Odors such as vinegar (no. 3, PC1 loading 0.95), woody (no. 12, 0.90), fatty-metallic (no. 5, 0.88), buttery-oily (no. 1, 0.86), and coconut (no. 7, 0.84) had large positive PC1 loadings (more than 0.8). Conversely, the odors having large negative PC1 loadings included potato-like (no. 2, -0.82) and sulfur (no. 4, -0.82). Given that 7 out of the 12 examined odor descriptions contributed to PC1, this component appeared to be the main evaluation axis of overall odor characteristics. Changes in the odor characteristics during two weeks of storage were affected by the increases in intensities of the odors with large positive PC1 loadings, and by the decreases in intensities of the odors with large negative PC1 loadings. In contrast, only the phenolic odor (no. 9, 0.81) exhibited a high positive PC2 loading. This result indicated that the decrease in the intensity of phenolic odor affected the odor characteristics of the 1-week sample only, with the decrease caused by unknown compounds (no. 9-2 and no. 9-3) and 3-methyl-1H-indole (no. 9-8).

Table 3. Principal component (PC) loadings obtained from PC analysis of GC-O data

		PC loading
No.	Odor description	PC1	PC2
1	Buttery-oily	0.86	0.44
2	Potato-like	−0.82	0.43
3	Vinegar	0.95	−0.01
4	Sulfur	−0.82	0.45
5	Fatty-metallic	0.88	−0.28
6	Rancid	0.53	−0.31
7	Coconut	0.84	0.52
8	Sweet-caramel	−0.53	0.18
9	Phenolic	0.43	0.81
10	Floral-fruity	0.49	0.54
11	Sweet-fruity	−0.38	0.57
12	Woody	0.90	0.01
	Eigenvalue	6.40	2.33
	Cumulative contribution ratio (%)	53.32	72.73

Conclusions

Lactobacillus paracasei    MCC1849 (LP), which has the potential to modulate immune function, has been employed as a supplement in a variety of food products. To facilitate the development of a functional LP-fermented dairy beverage, the present study investigated, using GC-O (CharmAnalysis™), changes in the intensities of odor compounds during two weeks of refrigerated storage of the beverage. Thirty-four odorants were detected by GC-O analysis of the stored samples. Twenty-nine of the detected odorants were identified. Among the detected odorants, 4-(4′-hydroxyphenyl)-2-butanone (raspberry ketone) was identified by GC-MS (SIM); this detection represents the first identification of this compound as an odorant in fermented dairy products. The 34 detected odorants were classified into 12 odor descriptions. Changes in the intensities of those odors during storage were evaluated. Notably, the total intensity of all odor descriptions increased with storage.

Odor characteristics of the 3 stored samples (0, 1, and 2 weeks) could be discriminated according to the duration of storage, and odor properties contributing largely to the changed odor characteristics were investigated. The changes during two weeks of storage were affected by increases in the intensities of vinegar, woody, fatty-metallic, buttery-oily, and coconut odors and decreases in the intensities of potato-like and sulfur odors. The one-week sample was distinguished by a decrease in the intensity of phenolic odor.

The information regarding the odorants and odor intensities obtained in this study is expected to be useful for improvement of the odor profile of LAB beverage products during storage. Further work will be needed to investigate changes in the sensory characteristics of the product during up to 2 weeks of refrigerated storage, and to further clarify the flavor compounds responsible for these changes.

References

•  Acree,  T.E.,  Barnard,  J., and  Cunningham,  D.G. (1984). A procedure for the sensory analysis of gas chromatographic effluents. Food Chem., 14, 273-286.
•  Akiyama,  M.,  Murakami,  K.,  Ikeda,  M.,  Iwatsuki,  K.,  Wada,  A.,  Tokuno,  K.,  Onishi,  M., and  Iwabuchi,  H. (2007). Analysis of the headspace volatiles of freshly brewed arabica coffee using solid-phase microextraction. J. Food Sci., 72, C388-C396.
•  Akiyama,  M.,  Murakami,  K.,  Ikeda,  M.,  Iwatsuki,  K.,  Wada,  A.,  Tokuno,  K.,  Onishi,  M.,  Iwabuchi,  H., and  Sagara,  Y. (2009). Analysis of freshly brewed espresso using a retronasal aroma simulator and influence of milk addition. Food Sci. Technol. Res., 15, 233-244.
•  Antón,  M.J.,  Valles,  B.S.,  Hevia,  A.G., and  Lobo,  A.P. (2014). Aromatic profile of ciders by chemical quantitative, gas chromatography-olfactometry, and sensory analysis. J. Food Sci., 79, S92-S99.
•  Avsar,  Y.K.,  Karagul-Yuceer,  Y.,  Drake,  M.A.,  Singh,  T.K.,  Yoon,  Y., and  Cadwallader,  K.R. (2004). Characterization of nutty flavor in cheddar cheese. J. Dairy Sci., 87, 1999-2010.
•  Belitz,  H.-D.,  Grosch,  W., and  Schieberle,  P. (2009). 10 Milk and Dairy Products. In “Food Chemistry 4^th revised and extended Edition,” ed. by  H.-D.  Belitz,  W.  Grosch, and  P.  Schieberle. Springer, Berlin Heidelberg, pp. 498-521.
•  Borejsza-Wyscoski,  W. and  Hrazdina,  G. (1994). Biosynthesis of p-hydroxyphenylbutan-2-one in raspberry fruits and tissue cultures. Phytochemistry, 35, 623-628.
•  Carunchia Whetstine  M.E.,  Cadwallader,  K.R., and  Drake,  M. (2005). Characterization of aroma compounds responsible for the rosy/floral flavor in cheddar cheese J. Agric. Food Chem., 53, 3126-3132.
•  Carunchia Whetstine,  M.E., and  Drake,  M. (2007). The flavor and flavor stability of skim and whole milk powders. In “Flavor of Dairy Products,” ed. by  K.R.  Cadwallader,  M.  Drake, and  R.J.  McGorrin, ACS symposium series 971, American Chemical Society, Washington, DC, pp. 217-252.
•  Cheng,  H. (2010). Volatile flavor compounds in Yogurt: A Review. Crit. Rev. Food Sci. Nutr., 50, 938-950.
•  Friedrich,  J.E. and  Acree,  T.E. (1998). Gas chromatography olfactometry (GC/O) of dairy products. Int. Dairy J., 8, 235-241.
•  Guillén,  M.D.,  Ibargoitia,  M.L.,  Sopelana,  P.,  Palencia,  G., and  Fresno,  M. (2004). Components detected by means of solid-phase microextraction and gas chromatography/mass spectrometry in the headspace of artisan fresh goat cheese smoked by traditional methods. J. Dairy Sci., 87, 284-299.
•  Kaneko,  T. (1994). Adaptability of lactic acid bacteria to unusual environment. Japanese Journal of Dairy and Food Science, 43, A-25-34.
•  Lamparsky,  D. and  Klimes,  I. (1981). Cheddar cheese flavour-its formation in the light of new analytical results. In: “Flavour 81,” ed. by  P.P.  Schreier, Third Weurman Symp., Proc. Int. Conf., Munich, April 28-30. Walter de Gruyter & Co, Berlin, Germany, p. 557.
•  Maruyama,  M.,  Abe,  R.,  Shimono,  T.,  Iwabuchi,  N.,  Abe,  F., and  Jin-Zhong  Xiao. (2016). The effects of non-viable Lactobacillus on immune function in the elderly: a randomized, double-blind, placebo-controlled study. Int. J. Food Sci. Nutr., 67, 67-73.
•  Michishita,  T.,  Akiyama,  M.,  Hirano,  Y.,  Ikeda,  M.,  Sagara,  Y., and  Araki,  T. (2010). Gas chromatography/olfactometry and electronic nose analyses of retronasal aroma of espresso and correlation with sensory evaluation by an artificial neural network. J. Food Sci. 75, S477-S489.
• Morinaga Milk Industry, Co., Ltd., inventor. (2014). Novel lactic acid bacteria, and medicines, foods and beverages, and feed containing novel lactic acid bacteria, feeds. Japanese patent. JP 5646124, Nov. 14.
• Morinaga Milk Industry, Co., Ltd., inventor. (2015). Novel lactic acid bacteria, and medicines, foods and beverages, and feed containing novel lactic acid bacteria, feeds. Published unexamined patent application. JP WO2015/004949, Jan. 15.
•  Morita,  A.,  Araki,  T.,  Ikegami,  S.,  Okaue,  M.,  Sumi,  M.,  Ueda,  R., and  Sagara,  Y. (2015). Coupled stepwise PLS-VIP and ANN modeling for identifying and ranking aroma components contributing to the palatability of cheddar cheese. Food Sci. Technol. Res. 21, 175-186.
•  Murakami  K.,  Akiyama  M.,  Sumi  M.,  Ikeda  M.,  Iwatsuki  K.,  Nishimra  O., and  Kumazawa  K. (2010). Differences in flavor characteristics of coffee drinks originating from thermal sterilization process. Food Sci. Technol. Res. 16, 99-110.
•  Ochiai,  N.,  Sasamoto,  K.,  Hoffmann,  A., and  Okanoya,  K. (2012). Full evaporation dynamic headspace and gas chromatography-mass spectrometry for uniform enrichment of odor compounds in aqueous samples. J. Chromatogr. A. 1240, 59-68.
•  Ochiai,  N.,  Mitsui,  K.,  Sasamoto,  K.,  Yoshimura,  Y.,  David,  F., and  Sandra,  P. (2014). Multidimensional gas chromatography in combination with accurate mass, tandem mass spectrometry, and element-specific detection for identification of sulfur compounds in tobacco smoke J. Chromatogr. A. 1358, 240-251.
•  Preininger  M.,  Gimelfarb,  L.,  LI,  H.-C.,  Dias,  B.E.,  Fahmy,  F., and  White,  J. (2009). Identification of dihydromaltol (2,3-dihydro-5- hydroxy-6-methyl-4H-pyran-4-one) in Ryazhenka kefir and comparative sensory impact assessment of related cycloenolones. J. Agric. Food Chem. 57, 9902-9908.
•  Taverniti,  V. and  Guglielmetti,  S. (2011). The immunomodulatory properties of probiotic microorganisms beyond their viability (ghost probiotics: proposal of paraprobiotic concept). Genes Nutr., 6, 261-274.
• The University of Tokyo and Morinaga Milk Industry Co., Ltd. inventors. (2016-6-23). An agent for increasing follicular helper T cell. Published unexamined patent application. JP 2016-113378 A.
•  Waterstraat,  M.,  Hildebrand,  A.,  Rosler,  M., and  Bunzel,  M. (2016). Development of a QuEChERS-based stable-isotope dilution LC-MS/ MS method to quantitate ferulic acid and its main microbial and hepatic metabolites in milk. J. Agric. Food Chem. 64, 8667-8677.
•  Zellner,  B.d'A.,  Dugo,  P.,  Dugo,  G., and  Mondello,  L. (2008). Gas chromatography-olfactometry in food flavour analysis. J. Chromatography A. 1186, 123-143.