Abstract
Abstract
Many flavoring agents on the market are extracted from natural sources or synthesized chemically. Due to the disadvantages of both methods, biotechnology is becoming a promising alternative. In this study, short chain ethyl esters with fruity notes were biosynthesized in UHT whole milk via coupling ethanolic fermentation with lipase (Palatase®) transesterification. Kluyveromyces marxianus, Lactobacillus fermentum and Lb. paracasei were used for fermentation. Milk fat was esterified with in situ produced ethanol by adding lipase at 0, 8 and 24 h of fermentation. Viable cell counts and pH were monitored during 48 h fermentation period. Flavor active ethyl esters, ethanol and free fatty acids were analyzed using headspace SPME-GC. Free fatty acid levels were lower in K. marxianus samples than lactobacilli. K. marxianus produced higher amounts of ethanol and esters than lactic acid bacteria. Viable cell counts decreased after lipase application at 0 and 8 h, possibly due to fatty acid production. Addition of lipase at 24 h resulted in improved cell counts as well as ethanol and ester production in the case of K. marxianus. This study demonstrated that fermenting milk with alcohol producing cultures in conjunction with lipase application can be an alternative to artificial flavorings in fermented milks.
Graphic abstract
Keywords: Ethyl ester, Palatase®, Lactobacillus, Kluyveromyces, Solid phase microextraction (SPME)
Introduction
Flavor is one of the important attributes of foods as it attracts consumers’ attention to choose a product (Khurana and Kanawjia 2007). Although flavor compounds exist naturally in foods, they are also used as food additives. Several categories of chemical compounds including esters contribute towards the flavor of foods. Esters can be extracted from natural sources or produced chemically or biotechnologically. Natural sources of flavor compounds can be expensive due to their low yield, while artificial flavors do not achieve consumers' demand of having healthy lifestyle. Biotechnology, however, provides appropriate methods to produce esters in higher amounts with an advantage of labeling as “natural”. Microbial fermentation and enzymatic synthesis, two common biotechnological routes, are particularly applicable for ester synthesis (Longo and Sanromán 2006).
Lipases (EC 3.1.1.3) are enzymes which have been widely investigated for flavor ester production. Although lipases are hydrolytic enzymes, they can also catalyze reverse reactions to esterify free fatty acid with alcohol or can alcoholyse glycerides to generate esters under certain conditions (Sun and Liu 2015). Palatase® 20,000 L, a purified lipase from Rhizomucor miehei produced by recombinant Aspergillus oryzae, has been developed for food applications (Liu et al. 2009). It is not only a suitable enzyme for food applications (Zhang et al. 2016), but it can also esterify in aqueous environments of foodstuffs, unlike many other lipases (Sun and Liu 2015). Palatase® is selective towards short-chain fatty acids so it is a suitable lipase in flavor ester production (Liu et al. 2013). Short chain fatty acid esters produce pleasant fruity flavor, whereas fatty acids with carbon chain longer than 12 may impart soapy and unpleasant flavors to food products (Liu et al. 2004).
One of the recently developed methods of generating fruity flavors in food is in situ ester production. In most cases, the alcohol applied for esterifying the fatty acids is directly added to the esterification/transesterification medium. Production of fruity flavor in milk and cream as described by Liu et al. (2009) and modification of the flavor of recombined milk as described by Zhang et al. (2016) was previously carried out by addition of exogenous ethanol. Ethanol is an alcohol with frequent utilization in flavor ester synthesis (Sun et al. 2013). Some studies, on the other hand, relied on in situ ethanol production by fermentation. Enzymatic biosynthesis of esters in coconut cream was previously studied by Sun et al. (2013) in such systems. Labeling the food product as “natural” is the advantage of in situ alcohol production coupled with enzymatic biosynthesis (Sun et al. 2013).
Fermented milk is a functional beverage with health benefits. Fermented milks are classified into three categories of fermentation, including lactic fermentation, yeast/lactic fermentation and mold/lactic fermentation. Offering a variety of flavors is one of the ways to improve the marketability of these products. Flavoring agents are one of the additives used in fermented milks (Khurana and Kanawjia 2007). However, the use of biosynthesis to produce flavors would be preferred for their “natural” labeling. Acidophilus milks have unpleasant flavor due to slow growth of Lactobacillus acidophilus (Mital and Garg 1992). Hence, it is plausible that fermented milk can take advantage of coupling fermentation with lipase ester synthesizing activity to naturally provide fruity flavors, since milk fat is comprised of short and medium chain fatty acids (C2–C14) (Zhang et al. 2016).
Therefore, this study was performed to evaluate the biosynthesis of natural fruity flavor esters in fermented whole milk using the lipase, Palatase®, along with in situ ethanol production. A yeast specious, an obligate heterofermentative and a facultative heterofermentative lactic acid bacteria were selected as starter cultures for milk fermentation. Fermentation time was selected based on the viable cell counts and pH values. Since sensory analysis is related to individuals, a more precise method, headspace solid phase microextraction–gas chromatography (SPME–GC), was used for analyzing the esters.
Materials and methods
Materials and microorganisms
Dextrose, chloramphenicol and saturated alkane standards (Sigma-Aldrich, St., MO), MRS (de Man Rogosa Sharpe) broth, potato dextrose agar, yeast extract and bacteriological peptone (Oxoid, Basingstoke, UK), malt extract, MRS agar, HCl and absolute ethanol (Merck, Darmstadt, Germany), Natamax (50 g Natamycin/100 g) (Danisco A/S, Copenhagen, Denmark), and lipase (Palatase® 20,000 L, (CAS number 9001-62-1), a lipase from Rhizomucor miehei produced by recombinant Aspergillus oryzae) (Novozymes, Bagsværd, Denmark) were purchased from their respective manufacturers. UHT whole milk (Devondale, Melbourne, Australia) was purchased from a supermarket in Singapore.
Two lactic acid bacteria, Lactobacillus fermentum PCC (Chr. Hansen A/S, Horsholm, Denmark) and Lactobacillus paracasei L26 (Lallemand, Ontario, Canada) and a yeast culture, Kluyveromyces marxianus NCYC 1425 (National Collection of Yeast Cultures, Norwich, UK) were obtained as pure cultures in the freeze-dried form.
Propagation and enumeration of microorganisms
Freeze-dried Lactobacillus cultures were propagated in MRS broth at 37 °C for 48 h. Freeze dried yeast culture was also inoculated into yeast malt [YM] broth (10 g/L dextrose, 3 g/L yeast extract, 3 g/L malt extract, 5 g/L bacteriological peptone, with pH adjusted to 5.0 with 1 M HCl) and incubated at 30 °C for 48 h. All propagated microbial cultures were stored at − 80 °C as stock cultures.
Starter cultures for milk fermentation were prepared by transferring 5% (v/v) of thawed bacterial and yeast stock cultures into fresh MRS and YM broth, respectively, and incubating the yeasts at 30 °C for 24 h and lactobacilli at 37 °C for 24 h. After two consecutive transfers of the microbial cultures, the cells were then harvested by centrifuging at 8000×g at 4 °C for 10 min, followed by washing the cell pellets twice with physiological saline (8.5 g/L NaCl). The washed cells were then re-suspended in fresh saline with the same volume prior to inoculation into milk.
The plate count method was used for enumeration of viable bacterial and yeast cell counts. MRS agar supplemented with 0.25 g/L natamycin and potato dextrose agar supplemented with 0.1 g/L chloramphenicol were used for plate cultures of bacteria and yeasts, respectively. Incubation of the inoculated agars at 37 °C for bacteria and 30 °C for yeasts was performed until colonies appeared (24–48 h).
Evaluation of fermentation time
Fermentation of UHT whole milk by the microorganisms was performed to evaluate appropriate fermentation time for the subsequent experiments. UHT whole milk was inoculated with the prepared bacterial and yeast inocula at 1% (v/v). The inoculated milk was dispensed in 30-mL aliquots into 50-mL polypropylene tubes and incubated at 30 °C for 72 h. Enumeration of viable microbial cell counts and pH measurements were performed at 0, 24, 48 and 72 h. Fermentation time was selected based on the growth trend.
Whole milk fermentation in the presence of lipase (Palatase®)
To evaluate the suitable enzyme concentration, 0.1%, 0.3% and 0.5% filter sterilized lipase, Palatase® 20,000 L, was added along with 6% (v/v, final) absolute ethanol to UHT whole milk. An overnight incubation at 30 °C determined that 0.1% lipase was enough for subsequent experiments as sample with 0.1% lipase had obvious fruity odor.
Lipase (undiluted) was added at 0.1% (v/v) to inoculated milk tubes at 0, 8 and 24 h from incubation time and sampling from individual tubes was also carried out at 0, 8, 24 and 48 h time points to evaluate the microbial cell count enumerations and pH of the milk samples. Enumeration of cell counts was performed using the above-mentioned methods. The pH measurements were also carried out with a calibrated pH meter (Metrohm, Herisau, Switzerland). Inoculated whole milk without added enzyme was used as a control. Uninoculated whole milk with 0.1% enzyme addition served as another control.
Ester and free fatty acid analysis by headspace solid phase microextraction–gas chromatography–MS/FID
The samples of 48 h fermentation as well as UHT whole milk without any fermentation (with enzyme addition at different times and without enzyme addition) were analyzed for short chain ethyl esters as well as free fatty acids by headspace solid phase microextraction–gas chromatography (SPME–GC).
For this purpose, a gas chromatograph (Agilent, 7890A, CA) equipped with DB-FFAP capillary column (60 m length, 0.25 µm i.d., 0.25 µm film thickness, Agilent, VA) was used. Helium with a flow rate of 1.2 mL/min was used as a carrier gas and an Agilent 5975C inert mass selective detector (MSD) and a flame ionization detector (FID) were used for detection. MS detector was operated in electron ionization mode (70 eV) with ion source temperature of 230 °C. Esters were identified based on the NIST 05 database as well as comparing their linear retention indices (LRI) with the data of NIST WebBook. Relating the retention times of esters to those of C7–C40 saturated alkane standards was performed to derive LRI values of esters. Free fatty acids and ethanol were also identified based on the NIST 05 database. Quantification of esters, free fatty acids and ethanol was also done using the peak areas of their corresponding FID peaks.
Volatile compounds in fermented milks were extracted using the SPME method described by Condurso et al. (2008) with modifications. Two mL of saturated NaCl solution was added to 2.2 g of milk samples in a 20-mL glass headspace vial sealed with a PTFE septum. Vials were incubated for 20 min at 60 °C under 250 rpm/min agitation to accelerate the equilibrium of esters between milk sample and headspace. An 85 µm carboxen/polydimethylsiloxane solid-phase microextraction (SPME) fiber (CAR/PDMS, Supelco) was exposed to the headspace of vials for 30 min to extract the volatile esters, using a Combi Pal autosampler (CTC Analytics, Switzerland). The fiber was then inserted to injection port of GC to be desorbed at 250 °C for 3 min. The initial oven temperature was 50 °C which began to increase after 5 min up to 230 °C at a rate of 5 °C/min and held for 30 min.
Statistical analysis
Data are reported as mean values ± standard deviations of three independent experiments (n = 3). One-way analysis of variance (ANOVA) and Duncan’s Multiple Range test with SPSS 16.0 was used to determine significant differences at P < 0.05.
Results and discussion
Evaluation of fermentation time
Figure 1 shows the changes in the viable cell counts and pH levels of fermented UHT whole milk samples during fermentation. All microbial cultures could grow in the milk samples. K. marxianus reached its highest cell count at 24 h and fermentation after 24 h did not change it. Lb. fermentum and Lb. paracasei continued their growth up to 72 h and 48 h, respectively. The pH level of yeast samples did not show loss after 24 h while two Lactobacillus species decreased pH even up to 72 h. In a previous study conducted by Hamme et al. (2009), K. marxianus decreased pH of crude goat whey up to 24 h and then increased it after 168 h fermentation. Considering the ability of K. marxianus in lactic acid production, they attributed pH increase to the consumption of such acid during the stationary phase by the yeast. The constant pH levels of K. marxianus samples after 24 h in our study can be the result of the same effect. In their study, Lb. rhamnosus decreased pH of crude goat whey during 72 h fermentation and pH remained constant for the following 96 h. This is also in consistence with our results on pH decrease by lactobacilli during 72 h (Hamme et al. 2009). Also, another study demonstrated different behavior of Lactobacillus strains in pH reduction of UHT milk, which can justify prolonged pH decrease by Lactobacillus cultures in our study (Østlie et al. 2003).
Fig. 1.
Changes in a microbial counts (log CFU/mL) and b pH during milk fermentation with yeast and Lactobacillus. Each microorganism and each parameter analyzed separately. Mean values with different lower case letters are significantly different (P < 0.05). Values are expressed as the mean ± SD of three independent experiments (n = 3)
So, with an overall count and pH assessment, 48 h was selected as fermentation time for the next experiments. SPME-GC analysis indicated that all three microorganisms could produce ethanol in the milk samples. So, they were appropriate for using in the enzymatic experiments. Detailed results on the ethanol production are discussed later.
Whole milk fermentation in the presence of lipase (Palatase®)
Table 1 shows the effect of Palatase® and its addition time on the viable cell count of three selected cultures during a 48 h fermentation period in UHT whole milk. Also, Table 2 displays the respective pH values of the corresponding samples. To obtain adequate amounts of ethanol for esterification/transesterification reactions by lipase, microorganisms should reach an adequate count. Addition of lipase to the fermentation medium could affect the growth rate and final viable cell counts. According to Table 1, in the case of K. marxianus, enzyme addition at the earlier stages of fermentation resulted in significant reductions of microbial counts. Decline in count was observed at the sampling point just after enzyme addition. Although enzyme addition at 24 h, significantly affected cell counts compared to control (sample without enzyme), but enzyme addition at 24 h had the lowest effect on the cell growth.
Table 1.
Changes in microbial counts (log CFU/mL) during milk fermentation with and without Palatase®
| Microorganism type | Sampling time (h) | Without enzyme | Enzyme addition time (h) | ||
|---|---|---|---|---|---|
| 0 | 8 | 24 | |||
| Kluyveromyces marxianus | 0 | 5.7 ± 0.0a A | 5.7 ± 0.0a A | 5.7 ± 0.0a A | 5.7 ± 0.0a A |
| 8 | 6.8 ± 0.0e B | 6.2 ± 0.0b B | 6.8 ± 0.0e B | 6.8 ± 0.0e B | |
| 24 | 7.6 ± 0.0g,h C | 6.4 ± 0.0d C | 7.1 ± 0.0f C | 7.5 ± 0.0g C | |
| 48 | 7.7 ± 0.0h C | 6.4 ± 0.0c C | 7.0 ± 0.0f C | 7.5 ± 0.0g C | |
| Lactobacillus fermentum | 0 | 7.2 ± 0.0a A | 7.2 ± 0.0a A | 7.2 ± 0.0a A | 7.2 ± 0.0a A |
| 8 | 7.5 ± 0.1b B | 7.8 ± 0.0b,c,d,e B | 7.6 ± 0.0b,c B | 7.5 ± 0.0b,c A | |
| 24 | 7.6 ± 0.1b,c,d B | 7.9 ± 0.01d,e B,C | 7.8 ± 0.0c,d,e B | 7.6 ± 0.1b,c,d A | |
| 48 | 8.6 ± 0.1h C | 8.1 ± 0.2e,f C | 8.2 ± 0.3f,g C | 8.5 ± 0.4g,h B | |
| Lactobacillus paracasei | 0 | 7.2 ± 0.0a A | 7.2 ± 0.0a A | 7.2 ± 0.0a A | 7.2 ± 0.0a A |
| 8 | 8.1 ± 0.0b B | 8.0 ± 0.0b B | 8.0 ± 0.0b B | 8.0 ± 0.0b B | |
| 24 | 8.6 ± 0.0d C | 8.3 ± 0.0c C | 8.5 ± 0.0d C | 8.6 ± 0.0d C | |
| 48 | 8.8 ± 0.0e D | 8.8 ± 0.0e D | 8.7 ± 0.0e D | 8.8 ± 0.0e D | |
Each microorganism analyzed separately. Upper case letters indicate differences during 48 h fermentation with same enzymatic treatment. Lower case letters show differences during 48 h fermentation and between enzymatic treatments. Mean values with different upper or lower case letters are significantly different (P < 0.05). Values are expressed as the mean ± SD of three independent experiments (n = 3)
Table 2.
Changes in pH during milk fermentation with and without Palatase®
| Microorganism type | Sampling time (h) | Without enzyme | Enzyme addition time (h) | ||
|---|---|---|---|---|---|
| 0 | 8 | 24 | |||
| Kluyveromyces marxianus | 0 | 6.60 ± 0.05e,f D | 6.63 ± 0.05f C | 6.63 ± 0.03f B | 6.63 ± 0.04f D |
| 8 | 6.37 ± 0.03d C | 5.41 ± 0.01b B | 6.43 ± 0.01d,e B | 6.36 ± 0.01d C | |
| 24 | 5.88 ± 0.09c B | 5.19 ± 0.08a A | 5.23 ± 0.07a,b A | 5.83 ± 0.11c B | |
| 48 | 5.76 ± 0.06c A | 5.30 ± 0.2a,b A,B | 5.29 ± 0.2a,b A | 5.33 ± 0.2a,b A | |
| Lactobacillus fermentum | 0 | 6.57 ± 0.01c C | 6.60 ± 0.02c B | 6.59 ± 0.02c B | 6.59 ± 0.03c B |
| 8 | 6.36 ± 0.04c B | 5.29 ± 0.04a A | 6.36 ± 0.02c B | 6.36 ± 0.04c B | |
| 24 | 6.31 ± 0.05c B | 5.09 ± 0.01a A | 5.16 ± 0.03a A | 6.30 ± 0.05c B | |
| 48 | 5.95 ± 0.14b A | 5.17 ± 0.37a A | 5.17 ± 0.35a A | 5.22 ± 0.36a A | |
| Lactobacillus paracasei | 0 | 6.62 ± 0.03c C | 6.60 ± 0.04c B | 6.60 ± 0.04c B | 6.59 ± 0.03c B |
| 8 | 6.43 ± 0.05c B,C | 5.34 ± 0.04a A | 6.45 ± 0.01c B | 6.44 ± 0.03c B | |
| 24 | 6.39 ± 0.02c B | 5.13 ± 0.01a A | 5.18 ± 0.04a A | 6.37 ± 0.01c B | |
| 48 | 6.08 ± 0.19b A | 5.17 ± 0.28a A | 5.14 ± 0.25a A | 5.24 ± 0.33a A | |
Each microorganism analyzed separately. Upper case letters indicate differences during 48 h fermentation with same enzymatic treatment. Lower case letters show differences during 48 h fermentation and between enzymatic treatments. Mean values with different upper or lower case letters are significantly different (P < 0.05). Values are expressed as the mean ± SD of three independent experiments (n = 3)
Palatase® had a less impact on lactobacilli than yeast growth. Addition of enzyme at 24 h did not significantly affect cell counts of Lb. fermentum, whereas, samples with enzyme addition at 0 or 8 h had lower counts compare to control samples. The only exception in cell count decrease was for Lb. paracasei; its growth trend was not affected by lipase under examined conditions during 48 h.
Our results were in accordance with a previous study conducted by Sun et al. (2013) on Saccharomyces cerevisiae fermentation in coconut cream. In their study, the yeast count was decreased upon the addition of Palatase® and the count was even lower when lipase was added in the earlier stage of fermentation. However, the effect of Palatase® on the counts of S. cerevisiae was not significant. They attributed this decrease to the compounds formed from lipase reactions, especially the medium chain fatty acids (Sun et al. 2013). Also in our study, cells underwent such stress caused by fatty acids released from milk fat due to hydrolytic activities of lipase. Other studies have confirmed the antimicrobial effects of fatty acids (Ricke 2003). A study indicated that susceptibility to different antimicrobial agents is species-dependent in Lactobacillus bacteria (Danielsen and Wind 2003). Accordingly, the difference between the resistance of Lb. paracasei and Lb. fermentum to fatty acids is not unusual, although a more detailed study is required to reveal the reasons.
Free fatty acids not only had detrimental effects on cells (Table 1), but also decreased the pH values (Table 2). The reduced pH level can also affect microbial growth rate depending on the type of microorganisms (Rampelotto 2010). However, yeast and lactobacilli have the ability to tolerate low pH (Basso et al. 2011), although less so with lactobacilli. The pH levels in turn are under the influence of microbial growth which produces acids and carbon dioxide (Sun et al. 2013). Reduction of growth following enzyme addition likely led to a slower drop of pH value. However, it is obvious that samples with earlier enzyme addition had lower pH levels despite their lower counts compared to control samples (Table 2). This reduction in pH levels was most likely due to lipolysis which increased free fatty acids in the medium.
Sun et al. (2013) reported pH increase upon addition of Palatase® to coconut cream medium fermented with S. cerevisiae that is in contrast with our results. In their study, after 72 h fermentation the pH value of the control sample was 4.78, and pH values of samples with enzyme addition at 24 and 48 h were 5.01, and 4.92, respectively. They related pH increment to loss of carbon dioxide and acids during fermentation by yeast (Sun et al. 2013). The difference is probably due to microorganism type and fermentation time that affect pH levels. They used the lipase enzyme in its diluted form (15% in phosphate buffer) at 0.1%, so the difference can also be due to enzyme amount. It seems that in our study Palatase® was more effective than fermentation in lowering pH levels so that it encompassed the pH fall caused by fermentation.
The pH value is also an important factor for Palatase® activity. The reported working range of pH for Palatase® is 5.0–9.5 with the optimum pH of 6.5–7.5 (Kurtovic et al. 2011). In our study, however, there were no samples with pH lower than 5.0. Moreover, it has been reported that different pH levels may result in different substrate selectivities of Palatase® due to altered conformation of the enzyme or reduced quantity of ionized free fatty acids which can inhibit enzyme activity (Jensen et al. 1990).
Ester analysis by headspace solid phase microextraction–GC–MS/FID
In this study, ethyl esters of short chain fatty acids (up to 10 carbon chains) were considered. The results reported in Table 3 demonstrate how ethanol and ester levels differed by adding lipase at different times for each microorganism.
Table 3.
Ester and ethanol levels [FID peak area (× 106)] in samples undergone 48 h of fermentation
| Microorganism type | Z (h) | Ethanol | Ethyl acetate | Ethyl butanoate | Ethyl pentanoate | Ethyl hexanoate | Ethyl heptanoate | Ethyl octanoate | Ethyl nonanoate | Ethyl decanoate |
|---|---|---|---|---|---|---|---|---|---|---|
| Kluyveromyces marxianus | 0 | 45.10 ± 7.45a | 3.63 ± 0.82a | 269.66 ± 62.92b | 3.83 ± 1.12a | 347.10 ± 28.49b | 1.57 ± 0.26a | 87.48 ± 14.63b | 1.83 ± 0.42a | 62.47 ± 10.43a |
| 8 | 170.09 ± 12.79b | 7.08 ± 0.07a | 424.37 ± 68.22c | 4.27 ± 1.15a | 393.94 ± 39.34b | 2.49 ± 0.07b | 184.84 ± 7.42c | 4.64 ± 0.37b | 312.17 ± 62.48b | |
| 24 | 405.48 ± 84.69c | 16.15 ± 5.36a | 321.05 ± 82.19b,c | 3.28 ± 0.93a | 312.61 ± 78.09b | 3.90 ± 0.12c | 283.06 ± 9.87d | 6.48 ± 0.74c | 566.81 ± 58.50c | |
| Without Z | 467.31 ± 42.58c | 90.05 ± 15.14b | 10.26 ± 0.56a | nd | 6.52 ± 0.66a | nd | 4.23 ± 0.47a | nd | 2.29 ± 0.65a | |
| Lactobacillus fermentum | 0 | 14.88 ± 4.14b | 1.98 ± 0.51a,b | 46.17 ± 12.14b | 0.80 ± 0.26a,b | 33.27 ± 10.96b | 0.51 ± 0.11b | 18.74 ± 4.57a | nd | 16.61 ± 5.12c |
| 8 | 23.93 ± 2.32c | 2.43 ± 0.77b | 37.89 ± 22.92b | 1.47 ± 0.49b | 28.49 ± 17.31b | 0.44 ± 0.14a,b | 18.67 ± 7.96a | nd | 14.41 ± 4.88b,c | |
| 24 | 6.58 ± 4.08a | 1.65 ± 0.30a,b | 22.35 ± 8.72a,b | 1.42 ± 0.78b | 15.93 ± 8.33a,b | 0.30 ± 0.07a | 9.52 ± 2.64a | nd | 8.06 ± 0.23b | |
| Without Z | 4.21 ± 1.58a | 1.21 ± 0.23a | 3.13 ± 1.36a | nd | Nd | nd | nd | nd | 0.69 ± 0.33a | |
| Lactobacillus paracasei | 0 | 1.15 ± 0.36a | 0.37 ± 0.01b | 3.99 ± 0.23b | nd | 1.40 ± 0.12b | nd | 1.33 ± 0.06a,b | nd | 3.88 ± 0.91b |
| 8 | 0.08 ± 0.01a | 0.17 ± 0.02a | 9.88 ± 0.78d | 0.19 ± 0.17a | 4.68 ± 0.78d | 0.11 ± 0.09a | 2.07 ± 1.51b | nd | 2.90 ± 2.38a,b | |
| 24 | 1.35 ± 0.09a | 0.51 ± 0.01b | 6.84 ± 0.68c | 0.20 ± 0.01a | 2.62 ± 0.21c | 0.16 ± 0.02a | 2.75 ± 0.13b | 0.41 ± 0.03a | 4.04 ± 0.17b | |
| Without Z | 8.15 ± 3.32b | 1.40 ± 0.20c | 1.55 ± 0.07a | nd | 0.45 ± 0.11a | nd | 0.16 ± 0.02a | nd | 0.92 ± 0.01a | |
| Without microorganism | 0 | 1.50 ± 0.22a | 1.45 ± 0.25b | 1.52 ± 0.33a | nd | 2.49 ± 0.36a | nd | 2.63 ± 0.21a | nd | 3.45 ± 0.48a |
| 8 | 1.34 ± 0.34a | 1.45 ± 0.27b | 4.92 ± 1.14b | nd | 3.48 ± 0.79a | 0.17 ± 0.02a | 2.77 ± 0.63a | nd | 3.62 ± 0.40a | |
| 24 | 1.80 ± 0.23a | 1.86 ± 0.28b | 5.92 ± 1.60b | nd | 4.79 ± 0.99b | nd | 3.70 ± 0.37b | nd | 3.54 ± 0.12a | |
| Without Z | nd | 0.60 ± 0.04a | nd | nd | nd | nd | nd | nd | nd | |
| LRI | – | – | 1037.46 | 1129.29 | 1235.18 | 1331.71 | 1439.34 | 1536.48 | 1649.58 |
Each microorganism analyzed separately. Mean values in the same column of each microorganism with different lower case letters are significantly different (P < 0.05). Values are expressed as the mean ± SD of three independent experiments (n = 3), Z enzyme addition time, Without Z without enzyme, nd not detected, LRI linear retention index
Based on Table 3, ethanol as well as ester levels of K. marxianus samples reached their highest levels by lipase addition at 24 h. Previously, researchers reported that ester biosynthesis by Palatase® may require a longer time to achieve an equilibrium and in short reaction times, hydrolysis may dominate (Sun et al. 2012). In our study, the enzyme had less time to catalyze reactions when added at 24 h rather than 8 h since fermentation continued up to 48 h. Nevertheless, higher cell counts of K. marxianus at 24 h (Table 1) caused higher ethanol and ester levels in 24 h samples.
Ethyl acetate levels of the samples without lipase were dramatically higher than the samples with lipase addition at 24 h in K. marxianus (Table 3). Ethyl acetate not only can be produced by lipase (not from fat), but also it is a product of yeast fermentation (Löser et al. 2013). It seems that addition of Palatase® had inhibitory effect on ethyl acetate production by K. marxianus likely due to inhibition of yeast growth.
K. marxianus had the highest ester production in comparison with Lactobacillus species even in the absence of Palatase® (statistical analysis of difference between microorganisms is not shown). This significant difference was likely due to the higher ethanol production by the yeast as compared to the lactobacilli. K. marxianus has previously been applied as an efficient ethanol producer. For instance, in a study 48.98 g/L of ethanol could be obtained in taro waste fermented by a strain of K. marxianus for 22 h (Wu et al. 2016).
Based on Table 3, in the samples without microorganisms and lipase, ethanol was not detected because no fermentation took place. On the other hand, the addition of Palatase® surprisingly produced some ethanol without microorganisms present. This was likely a result of lipase hydrolytic activity that released ethanol from ethyl esters which already existed in the fermentation medium. In the samples without microorganisms, esters were produced following the lipase addition and ethanol generation. In the samples without microorganisms and enzyme (UHT milk without any addition), no esters were present due to the lack of ethanol, except for ethyl acetate. This ester was most probably the metabolite of yeasts, lactobacilli or acetic acid bacteria which had grown in milk before sterilization (Annan et al. 2003; Quigley et al. 2013). It should be noted that all these cases can be caused by carry-overs from the SPME fiber.
Although the ethanol level increased during 8 h fermentation by Lb. fermentum, it decreased at 24 h. Reduction of ethanol level in 24 h samples was not likely due to consumption in ester biosynthesis reactions since ester levels were not significantly different for 8 and 24 h samples. Enzyme addition time had no significant effects on most ester levels of Lb. fermentum samples. However, lipase could have significantly raised ester levels compared to the control samples (without enzyme).
Lb. paracasei is facultatively heterofermentative (Naaber et al. 2004) and some ethanol levels were detected in the samples fermented by this Lactobacillus. Lb. paracasei L26 produced trace amounts of ethanol in a study by Lu et al. (2018) that is in accordance with our results (Lu et al. 2018). Although Palatase® had no significant effect on the growth rate of Lb. paracasei (Table 1), the ethanol amount was statistically different after enzyme addition according to Table 3. The reduction of ethanol level upon enzyme addition was probably due to ethanol consumption in esterification/transesterification reactions to produce esters.
The ethanol level produced by Lb. paracasei and Lb. fermentum in the absence of enzyme was similar. Enzyme addition at 0 or 8 h resulted in higher ester levels of Lb. fermentum samples while there were no significant differences in ester levels in Lb. paracasei and Lb. fermentum samples when enzyme added at 24 h. However, K. marxianus was more effective in ethanol and ester production than both lactobacilli as mentioned before (statistical analysis of difference between microorganisms is not shown).
Detection of esters in the samples containing microorganisms but without enzyme was not unusual since fermentation can produce some levels of these compounds. In the other words, esters are produced by enzymes present in the microbes (Toh and Liu 2017).
In addition to fatty acids of milk fat, lactic acid produced by lactobacilli species was present in the fermentation medium (Naaber et al. 2004) which can be esterified with ethanol in presence of lipase (Sun et al. 2010). Nevertheless, ethyl lactate was not detected in our study. It can be due to lipase type and ability of Palatase® in esterifying ethanol and lactic acid. Palatase® could synthesize ethyl lactate in its immobilized form in a previous study (Ugur Nigiz and Durmaz Hilmioglu 2016) but according to our information, there is no report on ethyl lactate synthesis by Palatase® in aqueous media. The main reason may originate from hydrophilic properties of both ethanol and lactic acid which cause them to stay in the water phase rather than accumulate at the fat–water interface. Therefore, Palatase® could not favorably esterify them. In a related research performed by Liu et al. (2003) ethyl butanoate also could not be synthesized from butanoic acid and ethanol via direct esterification. Yet, tributyrin with higher hydrophobicity could be utilized for ethyl butanoate production by alcoholysis with ethanol (Sun et al. 2013).
Besides, previous studies have demonstrated that low alcohol to acid ratio has an adverse effect on ethyl lactate synthesis (Delgado et al. 2010; Sun et al. 2010). In our study, ethanol was probably not sufficient for esterifying lactic acid since ethanol was gradually used by Palatase® while lactic acid was simultaneously increasing by fermentation.
One important factor that positively affects Palatase® activity is the fat content of milk as it gives hydrophobic nature to the mixture. In a research by Sun and Liu (2015) on the mechanism of ester synthesis, coconut cream was a more appropriate environment than buffer system due to its higher fat content which made the whole system more hydrophobic. Indeed, any substrate in aqueous media can affect lipase activity by influencing the distribution of water between lipase and reaction medium (Sun and Liu 2015).
Previous studies have reported inhibitory and denaturing effects of alcohols on Palatase® (Sun et al. 2012). Subsequent esterification of ethanol with fatty acids may overcome these toxicities. In our study K. marxianus produced higher ethanol levels compared to other microorganisms (statistical analysis is not shown) but even these levels of ethanol could not have adversely affected Palatase® activity since high ester levels were produced following enzyme addition.
In this study, a standard curve was used to evaluate the concentration of ethanol that remained in the product. In the case of K. marxianus that produced the highest ethanol levels, 13.5 mg/g (1.35 g/100 g or 1.35%) ethanol was detected in the absence of lipase. Applying enzyme at 24 h for samples containing K. marxianus led to 11.6 mg/g (1.16%) of ethanol which is almost near the Halal status (below 1%) (Alzeer and Abou Hadeed 2016; Sun et al. 2013).
The produced fermented milk can be consumed directly but since it was highly flavored, two ways are suggested to make its taste milder in order to consume directly: dilution with unflavored milk or applying less culture and enzyme amount in the process. On the other hand, the product can be used as a flavoring agent in the foodstuffs especially in dairy products. For example, fruit flavored yogurts can benefit the naturally flavored milk in their products. This product of our study can also be used to improve the taste of low fat dairies as a flavor enhancer. Weak flavor of acidophilus milks and sweet acidophilus milks that are considered undesirable (Mital and Garg 1992) can be compensated by applying some amounts of fruity flavored milk in their formulations. Fermented milks which have unpleasant flavor due to slow growth of Lb. acidophilus (Mital and Garg 1992) can use alcohol producing cultures along with Palatase® instead of fermentation by Lb. acidophilus as a sole culture. Presence of esters on the other hand can mask off-flavors such as goat-like and mutton-like flavors if present in dairy products (Liu et al. 2004).
Free fatty acid analysis by headspace solid phase microextraction–GC–MS/FID
Table 4 indicates the differences of main fatty acids (up to 10 carbon chains) in samples fermented for 48 h by different microbial cultures with and without enzymatic treatment. Since samples with enzyme addition at 24 h of fermentation showed better ester production results (Table 3), free fatty acid analyses were carried out for these samples. Free fatty acids were detected in samples without enzyme, indicating the ability of all three microbial cultures in fatty acid production. The fatty acid production of lactic acid bacteria and yeasts in dairy products has been reported previously (Leclercq-Perlat et al. 2007). There were no significant differences among microorganisms in terms of free fatty acid production except higher levels of acetic acid for Lb. paracasei. Enzyme addition increased fatty acid levels in all samples due to lipolytic activity of Palatase®. In the presence of enzyme, K. marxianus samples had lower fatty acid levels than lactobacilli which could be the result of higher involvement of fatty acids in ester production due to higher ethanol levels in K. marxianus samples. Free fatty acid accumulation in samples shows higher lipolysis than esterification reactions which is in accordance with the study on coconut cream (Sun et al. 2013). According to the statistical analysis of each microorganism, Lb. paracasei samples had higher levels of butanoic, octanoic, and decanoic acids in comparison with unfermented samples, in presence of enzyme. Although the microorganism produced some levels of fatty acids, but this cannot justify the high differences. Palatase® may affect fatty acid production by Lb. paracasei, but more research is needed. Samples without microbial and enzymatic treatments (only whole milk) had hexanoic and octanoic acids that probably had been produced microbially before milk sterilization.
Table 4.
Free fatty acid levels [FID peak area (× 106)] in samples undergone 48 h of fermentation
| Microorganism | Kluyveromyces marxianus | Lactobacillus fermentum | Lactobacillus paracasei | Without microorganism | ||||
|---|---|---|---|---|---|---|---|---|
| Z (h) | 24 | Without Z | 24 | Without Z | 24 | Without Z | 24 | Without Z |
| Acetic acid | 1.11 ± 0.33a | 1.34 ± 0.30a | 7.88 ± 0.26d | 2.81 ± 1.44b | 9.11 ± 0.41e | 4.26 ± 0.51c | 0.77 ± 0.34a | nd |
| Butanoic acid | 35.75 ± 8.48b | 1.18 ± 0.14a | 108.99 ± 32.02c | 0.76 ± 0.17a | 132.46 ± 11.05c | 2.13 ± 0.86a | 114.24 ± 3.96c | nd |
| Hexanoic acid | 123.63 ± 24.25b | 3.82 ± 0.38a | 478.75 ± 110.25c | 7.69 ± 1.53a | 554.72 ± 17.17d | 7.73 ± 1.99a | 543.54 ± 22.46c,d | 2.05 ± 0.66a |
| Octanoic acid | 74.07 ± 5.89b | 4.01 ± 0.26a | 252.16 ± 28.70d | 8.13 ± 1.34a | 312.74 ± 26.58e | 9.19 ± 2.73a | 221.07 ± 25.83c | 2.20 ± 0.58a |
| Decanoic acid | 84.90 ± 28.98b | 3.11 ± 0.29a | 98.32 ± 19.54b,c | 5.49 ± 1.61a | 125.45 ± 15.90c | 4.72 ± 0.83a | 69.08 ± 22.64b | nd |
Each fatty acid analyzed separately. Mean values in the same row with different lower case letters are significantly different (P < 0.05). Values are expressed as the mean ± SD of three independent experiments (n = 3), Z enzyme addition time, Without Z without enzyme, nd not detected
Conclusion
This study was performed to synthesize short chain flavor esters with fruity notes in fermented UHT whole milk via coupling alcoholic fermentation with lipase (Palatase®) transesterification activity. Kluyveromyces marxianus produced higher amounts of ethanol and esters than the two Lactobacillus species. However, free fatty acid levels were lower in K. marxianus samples than lactobacilli. Addition of Palatase® at earlier stages of fermentation resulted in decreased viable cell counts as well as ethanol and ester production.
This study revealed that in situ flavor ester synthesis can be a suitable alternative for utilization of artificial flavoring agents. In addition to directly use the obtained product as a fruity flavored fermented milk, it can also be utilized as a flavoring agent in other dairy products. The most important aspect of this method is that the product can be labeled as “natural”.
Acknowledgements
It is a great pleasure for us to thank National University of Singapore (NUS) particularly members of the Chemistry Department for providing laboratory and supporting this research. The support from Shiraz University is also sincerely appreciated.
Author contributions
MTG and MHE suggested the subject of study and supervised it; SQL designed and led the study; MSZ performed the experiments, analyzed the data and wrote the manuscript; MT provided the materials and directed the experiments implementation and data analysis; all authors revised and commented on the manuscript.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Maryam Shojaei Zinjanab, Email: mShojaei@shirazu.ac.ir.
Mohammad Taghi Golmakani, Email: golmakani@shirazu.ac.ir.
Mohammad Hadi Eskandari, Email: eskandar@shirazu.ac.ir.
Mingzhan Toh, Email: chmtohm@nus.edu.sg.
Shao Quan Liu, Email: fstLsq@nus.edu.sg.
References
- Alzeer J, Abou Hadeed K. Ethanol and its Halal status in food industries. Trends Food Sci Technol. 2016;58:14–20. doi: 10.1016/j.tifs.2016.10.018. [DOI] [Google Scholar]
- Annan NT, Poll L, Sefa-Dedeh S, Plahar WA, Jakobsen M. Volatile compounds produced by Lactobacillus fermentum, Saccharomyces cerevisiae and Candida krusei in single starter culture fermentations of Ghanaian maize dough. J Appl Microbiol. 2003;94(3):462–474. doi: 10.1046/j.1365-2672.2003.01852.x. [DOI] [PubMed] [Google Scholar]
- Basso LC, Basso TO, Rocha SN. Ethanol production in Brazil: the industrial process and its impact on yeast fermentation. In: Bernardes MAS, editor. Biofuel production-recent developments and prospects. London: IntechOpen; 2011. pp. 85–100. [Google Scholar]
- Condurso C, Verzera A, Romeo V, Ziino M, Conte F. Solid-phase microextraction and gas chromatography mass spectrometry analysis of dairy product volatiles for the determination of shelf-life. Int Dairy J. 2008;18(8):819–825. doi: 10.1016/j.idairyj.2007.12.005. [DOI] [Google Scholar]
- Danielsen M, Wind A. Susceptibility of Lactobacillus spp. to antimicrobial agents. Int J Food Microbiol. 2003;82(1):1–11. doi: 10.1016/S0168-1605(02)00254-4. [DOI] [PubMed] [Google Scholar]
- Delgado P, Sanz MT, Beltrán S, Núñez LA. Ethyl lactate production via esterification of lactic acid with ethanol combined with pervaporation. Chem Eng J. 2010;165(2):693–700. doi: 10.1016/j.cej.2010.10.009. [DOI] [Google Scholar]
- Hamme V, Sannier F, Piot JM, Didelot S, Bordenave-Juchereau S. Crude goat whey fermentation by Kluyveromyces marxianus and Lactobacillus rhamnosus: contribution to proteolysis and ACE inhibitory activity. J Dairy Res. 2009;76(2):152–157. doi: 10.1017/S0022029908003877. [DOI] [PubMed] [Google Scholar]
- Jensen RG, Galluzzo DR, Bush VJ. Selectivity is an important characteristic of lipases (acylglycerol hydrolases) Biocatalysis. 1990;3(4):307–316. doi: 10.3109/10242429008992074. [DOI] [Google Scholar]
- Khurana HK, Kanawjia SK. Recent trends in development of fermented milks. Curr Nutr Food Sci. 2007;3(1):91–108. doi: 10.2174/1573401310703010091. [DOI] [Google Scholar]
- Kurtovic I, Marshall SN, Miller MR, Zhao X. Flavour development in dairy cream using fish digestive lipases from Chinook salmon (Oncorhynchus tshawytscha) and New Zealand hoki (Macruronus novaezealandiae) Food Chem. 2011;127(4):1562–1568. doi: 10.1016/j.foodchem.2011.02.018. [DOI] [Google Scholar]
- Leclercq-Perlat MN, Corrieu G, Spinnler HE. Controlled production of Camembert-type cheeses: part III role of the ripening microflora on free fatty acid concentrations. J Dairy Res. 2007;74(2):218–225. doi: 10.1017/S0022029906002329. [DOI] [PubMed] [Google Scholar]
- Liu SQ, Holland R, Crow VL. Synthesis of ethyl butanoate by a commercial lipase in aqueous media under conditions relevant to cheese ripening. J Dairy Res. 2003;70:359–363. doi: 10.1017/S0022029903006290. [DOI] [PubMed] [Google Scholar]
- Liu SQ, Holland R, Crow VL. Esters and their biosynthesis in fermented dairy products: a review. Int Dairy J. 2004;14(11):923–945. doi: 10.1016/j.idairyj.2004.02.010. [DOI] [Google Scholar]
- Liu SQ, Crow VL, Holland R. Production of natural fruity flavour in dairy foods. Nutr Food Sci. 2009;39:483–489. doi: 10.1108/00346650910992132. [DOI] [Google Scholar]
- Liu SQ, Lee HY, Yu B, Curran P, Sun J. Bioproduction of natural isoamyl esters from coconut cream as catalysed by lipases. J Food Res. 2013;2(2):157–166. doi: 10.5539/jfr.v2n2p157. [DOI] [Google Scholar]
- Longo MA, Sanromán MA. Production of food aroma compounds: microbial and enzymatic methodologies. Food Technol Biotechnol. 2006;44(3):335–353. [Google Scholar]
- Löser C, Urit T, Stukert A, Bley T. Formation of ethyl acetate from whey by Kluyveromyces marxianus on a pilot scale. J Biotechnol. 2013;163(1):17–23. doi: 10.1016/j.jbiotec.2012.10.009. [DOI] [PubMed] [Google Scholar]
- Lu Y, Putra SD, Liu SQ. A novel non-dairy beverage from durian pulp fermented with selected probiotics and yeast. Int J Food Microbiol. 2018;265:1–8. doi: 10.1016/j.ijfoodmicro.2017.10.030. [DOI] [PubMed] [Google Scholar]
- Mital BK, Garg SK. Acidophilus milk products: manufacture and therapeutics. Food Rev Int. 1992;8(3):347–389. doi: 10.1080/87559129209540946. [DOI] [Google Scholar]
- Naaber P, Smidt I, Štšepetova J, Brilene T, Annuk H, Mikelsaar M. Inhibition of Clostridium difficile strains by intestinal Lactobacillus species. J Med Microbiol. 2004;53(6):551–554. doi: 10.1099/jmm.0.45595-0. [DOI] [PubMed] [Google Scholar]
- Østlie HM, Helland MH, Narvhus JA. Growth and metabolism of selected strains of probiotic bacteria in milk. Int J Food Microbiol. 2003;87(1–2):17–27. doi: 10.1016/s0168-1605(03)00044-8. [DOI] [PubMed] [Google Scholar]
- Quigley L, O'Sullivan O, Stanton C, Beresford TP, Ross RP, Fitzgerald GF, Cotter PD. The complex microbiota of raw milk. FEMS Microbiol Rev. 2013;37(5):664–698. doi: 10.1111/1574-6976.12030. [DOI] [PubMed] [Google Scholar]
- Rampelotto PH. Resistance of microorganisms to extreme environmental conditions and its contribution to astrobiology. Sustainability. 2010;2(6):1602–1623. doi: 10.3390/su2061602. [DOI] [Google Scholar]
- Ricke SC. Perspectives on the use of organic acids and short chain fatty acids as antimicrobials. J Poult Sci. 2003;82(4):632–639. doi: 10.1093/ps/82.4.632. [DOI] [PubMed] [Google Scholar]
- Sun J, Liu SQ. Ester synthesis in aqueous media by lipase: alcoholysis, esterification and substrate hydrophobicity. J Food Biochem. 2015;39(1):11–18. doi: 10.1111/jfbc.12104. [DOI] [Google Scholar]
- Sun J, Jiang Y, Zhou L, Gao J. Optimization and kinetic study of immobilized lipase-catalyzed synthesis of ethyl lactate. Biocatal Biotransform. 2010;28(4):279–287. doi: 10.3109/10242422.2010.501893. [DOI] [Google Scholar]
- Sun J, Yu B, Curran P, Liu SQ. Optimisation of flavour ester biosynthesis in an aqueous system of coconut cream and fusel oil catalysed by lipase. Food Chem. 2012;135(4):2714–2720. doi: 10.1016/j.foodchem.2012.06.119. [DOI] [PubMed] [Google Scholar]
- Sun J, Lim Y, Liu SQ. Biosynthesis of flavor esters in coconut cream through coupling fermentation and lipase-catalyzed biocatalysis. Eur J Lipid Sci Technol. 2013;115(10):1107–1114. doi: 10.1002/ejlt.201300144. [DOI] [Google Scholar]
- Toh M, Liu SQ. Impact of coculturing Bifidobacterium animalis subsp. lactis HN 019 with yeasts on microbial viability and metabolite formation. J Appl Microbiol. 2017;123(4):956–968. doi: 10.1111/jam.13571. [DOI] [PubMed] [Google Scholar]
- Ugur Nigiz F, Durmaz Hilmioglu N. Green solvent synthesis from biomass based source by biocatalytic membrane reactor. Int J Energy Res. 2016;40(1):71–80. doi: 10.1002/er.3319. [DOI] [Google Scholar]
- Wu WH, Hung WC, Lo KY, Chen YH, Wan HP, Cheng KC. Bioethanol production from taro waste using thermo-tolerant yeast Kluyveromyces marxianus K21. Bioresour Technol. 2016;201:27–32. doi: 10.1016/j.biortech.2015.11.015. [DOI] [PubMed] [Google Scholar]
- Zhang XM, Ai NS, Wang J, Tong LJ, Zheng FP, Sun BG. Lipase-catalyzed modification of the flavor profiles in recombined skim milk products by enriching the volatile components. J Dairy Sci. 2016;99(11):8665–8679. doi: 10.3168/jds.2015-10773. [DOI] [PubMed] [Google Scholar]