Professional Documents
Culture Documents
https://doi.org/10.1007/s00253-017-8733-3
Abstract
Lactic acid bacteria (LAB) are essential microbiota for the fermentation and flavor formation of Shanxi aged vinegar, a famous
Chinese traditional cereal vinegar that is manufactured using open solid-state fermentation (SSF) technology. However, the dy-
namics of LAB in this SSF process and the underlying mechanism remain poorly understood. Here, the diversity of LAB and the
potential driving factors of the entire process were analyzed by combining culture-independent and culture-dependent methods.
Canonical correlation analysis indicated that ethanol, acetic acid, and temperature that result from the metabolism of microorgan-
isms serve as potential driving factors for LAB succession. LAB strains were periodically isolated, and the characteristics of 57
isolates on environmental factor tolerance and substrate utilization were analyzed to understand the succession sequence. The
environmental tolerance of LAB from different stages was in accordance with their fermentation conditions. Remarkable correla-
tions were identified between LAB growth and environmental factors with 0.866 of ethanol (70 g/L), 0.756 of acetic acid (10 g/L),
and 0.803 of temperature (47 °C). More gentle or harsh environments (less or more than 60 or 80 g/L of ethanol, 5 or 20 g/L of acetic
acid, and 30 or 55 °C temperature) did not affect the LAB succession. The utilization capability evaluation of the 57 isolates for 95
compounds proved that strains from different fermentation stages exhibited different predilections on substrates to contribute to the
fermentation at different stages. Results demonstrated that LAB succession in the SSF process was driven by the capabilities of
environmental tolerance and substrate utilization.
Keywords Lactic acid bacteria . Succession sequence . Shanxi aged vinegar . Environmental tolerance . Substrate utilization
Introduction fermented foods, such as vinegars (Li et al. 2016b) and cheese
(Wolfe et al. 2014), are driven by multispecies microbial com-
Solid-state fermentation (SSF) is one of the oldest and most munities. In an open work environment, succession of micro-
economical ways of producing and preserving foods; more- bial communities is always consecutive and can affect the char-
over, SSF may improve the nutritional values, sensory proper- acteristics and quality of the final fermented food (Li et al.
ties, and functional qualities of raw materials (Lu et al. 2016). 2016b). Many microbiological studies have been conducted
The SSF processes of most known natural and industrial to reveal the diversity and succession of microbial communities
during these open-SSF processes (Nie et al. 2015; Wu et al.
Electronic supplementary material The online version of this article
2012); however, gaps remain between the community forma-
(https://doi.org/10.1007/s00253-017-8733-3) contains supplementary tion and function of these complex microbial communities dur-
material, which is available to authorized users. ing the open-SSF processes.
Vinegars have been used worldwide as acerbic table condi-
* Min Wang ments, preservatives, health products, and medicine for thou-
minw@tust.edu.cn sands of years. In China, vinegars are traditionally made from
1 cereals, such as sorghum, sticky rice, and wheat bran (Wu et al.
State Key Laboratory of Food Nutrition and Safety. Key Laboratory
of Industrial Fermentation Microbiology, Ministry of Education. 2010; Li et al. 2016b). Shanxi aged vinegar (SAV) is one of the
College of Biotechnology, Tianjin University of Science & famous Chinese vinegars, which is produced with a typical SSF
Technology, Tianjin 300457, People’s Republic of China technology involving Daqu, alcohol fermentation (AF), acetic
Appl Microbiol Biotechnol
acid fermentation (AAF), smoking pei, and aging stages (Liu analysis (CCA) methods. LAB strains were periodically iso-
et al. 2012; Wu et al. 2012). Various microorganisms, including lated according to the brewing process to understand the suc-
molds, yeasts, and bacteria, are spontaneously enriched and cession sequence of LAB. Characteristics of LAB strains on
disappeared in the SSF process (Nie et al. 2015; Solieri and environmental adaption and substrate utilization were ex-
Giudici 2009) (Fig. 1). Such microorganisms are significant plored to reveal the succession sequence of LAB throughout
in the production and characteristics of SAV. Smoking the fermentation process. To our knowledge, this is the first
pei, in which microorganisms scarcely remain due to the report to perform comparative studies to investigate the suc-
high-temperature environment (80 to 90 °C), is a special tech- cession sequence of the LAB community by combining
nique for SAV that is performed after AAF stage. Daqu, a type culture-independent and culture-dependent methods. The re-
of mixed-culture starter, is used as inoculum for the SSF, and sults will greatly increase the scientific understanding of the
the dominant beneficial microorganisms in Daqu are molds, of brewing mechanism in the open-SSF process and may render
the genera Aspergillus, Rhizopus, and Mucor, and yeasts of the rational strategies for regulating the SSF process.
genus Saccharomyces (Nie et al. 2013). Yeasts of the genera
Saccharomyces and Saccharomycopsis typically dominate the
AF stage because they can rapidly consume low-molecular- Materials and methods
weight sugar and thus increase ethanol concentration (Li et al.
2015b). Acetic acid bacteria (AAB) and lactic acid bacteria Samples
(LAB) are crucial to the success of SAV. AAB, especially
Acetobacter, dominate the AAF stage because of their excel- Samples of Daqu, Jiulao (samples from AF stage with mi-
lent capability of rapid conversion of ethanol into acetic acid, crobes and substrates), Cupei (samples from AAF stage with
which is the major source of total acid and dominant flavor microbes and substrates), and aging stages were collected
components of vinegars (Nie et al. 2017; Solieri and Giudici from a SAV factory (Shanxi Zilin Vinegar Industry Co.,
2009). LAB, particularly Lactobacillus, are the most relevant Ltd., Qingxu, China) according to the production process.
bacteria throughout the SSF process of SAV and greatly con- Daqu was collected from a mature Daqu storage room. Four
tribute to the production of organic acids, which are the main Jiulao samples were collected at different times (1st, 4th, 7th,
flavor substances of SAV and contribute a soft taste (Li et al. and 10th day), along with the AF stage. Five Cupei samples
2016b; Nie et al. 2015, 2017), as well as other flavor sub- were collected at the 0th, 1st, 3rd, 5th, and 7th day, along with
stances. The composition of LAB constantly changes during the AAF stage. Aging samples were collected at the aging
the fermentation process, from Daqu to AF, AAF, and aging stage of SAV. Parts of these samples were used to determine
stages due to the open-SSF technology and varied conditions physicochemical indictors and were dealt with as follows:
(Nie et al. 2015; Wu et al. 2012). Thus, studies on how highly Samples of Jiulao were centrifuged at 4629×g for
variable environmental conditions affect LAB succession in 10 min after homogenization, and the supernatant was
the SSF process of SAV are becoming increasingly important. used for detection. For Cupei pretreatment, 5 g of samples was
Recent studies have suggested that bacterial communi- diluted into 45 mL of distilled water, shaken for 30 min
ties are often formed under controlled conditions in dis- completely at room temperature, and centrifuged. The super-
crete units, which allow the measurement and manipula- natant was collected for analysis.
tion of migration into the community, environmental con-
ditions, and growth substrates (Li et al. 2015a, 2016a; Analyses of the diversity and succession of LAB
Wang and Xu 2015; Wolfe et al. 2014; Zheng et al. community
2014). Others have argued that bacterial dynamics is large-
ly driven by individual traits (Nemergut et al. 2016). These The succession of LAB along with the entire fermentation
studies generally have implied a degree of mathematical process was analyzed using high-throughput sequencing. For
analysis in the driving factors of bacterial community suc- DNA extraction, each sample was homogenized using liquid
cession. However, knowledge on how environmental con- nitrogen (flash-frozen in liquid nitrogen and then rapidly
ditions, substrates, and individual traits affect the microbial thawed in a water bath at 65 °C for 2 min, these procedures
community structure has not been identified yet. The suc- were repeated three times), and approximately 500 mg of each
cessional trajectory of LAB communities and the mecha- sample was used for genomic DNA extraction which was
nism controlling the dynamics at a scale that is relevant for performed according to a previous method (Nie et al. 2015).
these organisms are significant to understand considering Bacterial 16S rDNA genes (V3–V5 variable regions) were
the importance of LAB for the SAV SSF process. amplified from the total genomic DNA by using primers 341
In this work, the succession of LAB and its potential driv- F and 926 R, which incorporated 454 FLX titanium adapters
ing factors in the open-SSF process of SAV were revealed by and a sample barcode sequence (Nie et al. 2015). Sequencing
using high-throughput sequencing and canonical correlation was performed using a 454 GS FLX titanium system (Roche,
Appl Microbiol Biotechnol
America) at The Beijing Genomics Institute (BGI, Shenzhen, metadata obtained by high-throughput sequencing (relative
China). 16S rDNA gene sequences were processed and mod- abundance), main fermentation parameters (pH, total acid,
ified following previously described methods (Jeong et al. reducing sugar, and temperature), and metabolites (ethanol,
2013; Kostic et al. 2012). Pyrosequencing data were proc- lactic acid, and acetic acid) were taken. The analysis
essed using a data curation pipeline implemented in Mothur was conducted using Canoco for Windows v4.51 and
(Schloss et al. 2009). Sequencing reads were sorted to specific Canodraw (Wageningen UR, Netherlands) (Nie et al. 2017).
samples on the basis of unique barcodes, and the barcodes Data standardization was performed using IBM SPSS 19.0
were subsequently trimmed. Sequencing reads were removed (http://www.spss.com.hk/).
if the reads were < 200 or > 1000 bp, had a non-exact barcode
match, exceeded two ambiguous bases (N), exceeded eight Isolation of LAB
consecutive identical bases, or had average quality scores <
25. A nearest-alignment space termination-based sequence A total of 5 g of each sample was added to 45 mL of physio-
aligner was used to align sequences to a custom reference on logic saline solution and homogenized in an incubator with
the basis of SILVA alignment (Kostic et al. 2012; Schloss et al. shaking at 200 rpm for 5 min. The blending suspensions were
2009). Chimeric sequences were identified and filtered by serially diluted with a sterilized physiologic saline solution,
quality. Chimera-free sequences were clustered in operational and 150 μL of appropriate diluted solution was spread onto
taxonomic units (OTUs) defined by 97% similarity using a Man–Rogosa–Sharpe (MRS) plates (Beijing land and
complete-linkage clustering tool. Representative sequences bridge, China) with 0.1% (w/v) nystatin dihydrate for
per OTU were classified according to previously described inhibiting fungus. Plates were incubated at 37 °C for 2
methods (Kostic et al. 2012). Ribosomal Database Project to 3 days, and colonies with different phenotypes were select-
naïve Bayesian rDNA classifier was used to hierarchically ed and purified. All isolates were submitted to Gram staining,
classify high-quality sequencing reads at the genus level with spore staining, catalase activity testing, and acid production
a confidence threshold of 80% (Wang et al. 2007). The OTUs testing according to LAB definition (Waters et al. 2015) and
that belonged to LAB were selected for the biodiversity anal- were stored at 4 °C. Sequence accession numbers for the iso-
ysis of LAB in the SAV-brewing process. LAB biodiversity lates used in the following experiments are provided in
indices, including Berger–Parker (d), Margalef (d Ma ), Supplemental Table S1.
Simpson index (D), Shannon index (H′), and Pielou evenness
index (Je), which represent species dominance, richness, di- Genotype comparison
versity probability, diversity level, and uniformity, respective-
ly, were calculated from data of relative quantities of OTUs The LAB isolates were cultured in MRS media for 12 to 18 h.
that belong to LAB (Nie et al. 2015). Genome was extracted with MiniBEST Bacterial Genomic
CCA was performed to determine the relationship between DNA Extraction Kit (Takara, Dalian, China) according to
LAB community and fermentation parameters. Nest to the instruction manual.
All LAB isolates in this study were subjected to enterobac- growth in different stresses of acetic acid, ethanol, and tem-
terial repetitive intergenic consensus (ERIC)–polymerase perature. Correlation analysis between cell growth of LAB
chain reaction (PCR) to dislodge the potentially same strain from each stage under environmental stress and the environ-
by comparing the fingerprints. Amplifications were conducted mental parameters that varied along with fermentation was
in 25 μL of reaction mixture containing 0.6 μL of Premix Ex performed using SPSS.
Taq (Takara, Dalian, China) and 1 μM of each primer of
ERIC1 (5′-ATGTAAGCTCCTGGGGATTCAC-3′) and Characteristics of LAB regarding substrate
ERIC2 (5′-AAGTAAGTGACTGGGGTGAGCG-3′) (Wu
et al. 2011). PCR was performed as follows: an initial dena- The capability of LAB to utilize substrates was assessed with a
turation at 94 °C for 5 min; 35 cycles consisting of denatur- Gram-positive (GP) plate (GEN III Microstation, Biology,
ation at 94 °C for 30 s; annealing at 52 °C for 90 s, with Hayward CA) that contained 95 carbon sources, among which
extension at 72 °C for 3 min; and a final extension at 72 °C nearly 60 were detected in SAV fermentation in a pre-
for 10 min. PCR products were examined by 1.5% agarose gel vious study (Li et al. 2016b). The capability of LAB to
electrophoresis, and the amplicon fingerprint was analyzed decompose a particular substrate was evaluated accord-
using Quantity One 4.6.2 (Bio-Rad, Hercules, CA). ing to the color of the GP plate. Data standardization
Cluster analysis was performed with Minitab 16 software was performed using SPSS. PCA was used to evaluate
(http://www.minitab.com/en-us/) (Minitab, America) using the metabolic characteristics of 57 isolates. The hierarchical
the unweighted pair–group method with arithmetic means, cluster analysis (HCA) managed by software HCE3 (http://
and a dendrogram was produced on the basis of each simple www.cs.umd.edu/hcil/hce/) was performed to explore
matching matrix. various LAB catabolism capabilities. A heat map was drawn
according to HCA data.
Identification of isolates
Composition and succession of LAB communities higher than those of the Daqu and AAF stages. OTU 6, OTU
9, OTU 11, and OTU 12, which belonged to Lactobacillus,
High-throughput sequencing was used to investigate the com- existed in the AF and AAF stages. Relative abundances of
position and succession of bacterial communities during SAV OTU 9 and OTU 12 constantly increased from the AF to the
fermentation. At least 12 OTUs in 4 genera, including AAF stages. OTUs of LAB decreased from 11 to 5 with fer-
Lactobacillus, Weissella, Pediococcus, and others, were iden- mentation from the AF stage to the AAF stage, which indicat-
tified as LAB (shown in Fig. 2b). The relative abundance of ed the vanishing of LAB species during SAV fermentation.
Lactobacillus reached 98.7%, in addition to small amounts of However, the succession sequence of LAB remained unclear.
Weissella (0.8%), Pediococcus (0.4%), and others (0.1%) of
LAB. OTUs of LAB in the AF stage (11 OTUs) were richer Relationships between environmental factors
than those in Daqu (3 OTUs) and AAF (5 OTUs) stages, and LAB community succession
which conformed to the results of LAB diversity indices
(shown in Table 1). Species dominance (d), richness (dMa), LAB existing in the brewing process of SAV spontaneously
Simpson (D), and Shannon (H′) indices of the AF stage were propagated to form a particular LAB community under the
Fig. 2 Dynamics of physicochemical indicator and LAB in SAV community structure with environmental factors. Arrows indicate the
fermentation and their correlation analysis. a Time curves of SAV direction and magnitude of environmental factors associated with LAB
fermentation. b Dynamics of relative abundance of LAB analyzed by community structure
high-throughput sequencing. c CCA analysis of the correlation of LAB
Appl Microbiol Biotechnol
effects of nutrition and environmental factor changes, espe- (Nie et al. 2015, 2017). In the present study, LAB strains were
cially temperature, pH, total acid, and reducing sugar (Li et al. isolated according to the brewing process to determine the real
2017). Effects of these environmental factors and dominant characteristics of LAB from different stages. LAB are a group
metabolites (ethanol, lactic acid, and acetic acid) on the LAB of GP, catalase-negative, non-motile, non-respiring, and non-
community succession were evaluated using the CCA meth- spore bacteria that can make bromocresol turn green yellow
od, as shown in Fig. 2c. The two axes explained 86.4% of the (Wu et al. 2011; Waters et al. 2015). A total of 85 strains of
variation in LAB community differentiation, which suggested LAB were isolated from different brewing stages of SAV ac-
a remarkable correlation between LAB structure and these cording to their morphology on plates and under an optical
factors. Reducing sugar, ethanol, pH, and lactic acid showed microscope. All isolates were screened using the ERIC-PCR
strongly positive correlation with LAB distribution in the AF method, which has been proven to be an effective method for
stage but showed negative correlation in the AAF stage. differentiating LAB at the species, subspecies, and strain
Ethanol showed a stronger correlation than those of reducing levels because of its high repeatability and accuracy to char-
sugar, pH, and lactic acid. The remarkable correlation of eth- acterize and differentiate each strain (Gevers et al. 2001;
anol with LAB community indicated that ethanol, as the main Masco et al. 2003; Wu et al. 2012). The ERIC-PCR molecular
product in the AF stage, potentially played a major role in the fingerprint clustering of all strains to reflect microbial diver-
diversity and succession of LAB and corresponded to a pre- sity is shown in Fig. 3 (Yuan et al. 2010). Strains isolated from
vious report (Li et al. 2015b). Total acid and temperature, as the same sample with an identical genotype were considered
the most important physiochemical indicators for AAF, the same strain, and only one was selected for further study.
showed positive correlation with LAB composition in the For instance, among the AAF3-2, AAF3-4, AAF3-6, AAF3-
AAF stage (Li et al. 2016b). By contrast, they were negatively 7, and AAF3-8 that were isolated from sample AAF3d, only
correlated with LAB in the AF stage. The same result was AAF3-2 was selected. However, AAF7-4, AAF3-2, AAF1-4,
observed for the correlation of acetic acid with LAB compo- and AAF0-4 that were identified as Lactobacillus casei in
sition. Acetic acid and temperature showed an almost equal different samples were selected. A total of 57 isolates (listed
correlation with LAB distribution in the AAF stage, which in Supplemental Table S1) were accordingly selected and
indicated that acetic acid and temperature might play impor- identified by 16S rDNA combined with morphological phys-
tant roles in LAB structure in the AAF stage. Thus, exploring iological analysis. A phylogenetic tree was constructed on the
the environmental factor tolerance of LAB, including ethanol, basis of 16S rDNA sequences (Supplemental Fig. S1).
temperature, and acetic acid, would be a rational direction to ERIC-PCR fingerprints demonstrated the diversity of LAB
reveal the succession sequence of LAB throughout the strains throughout the fermentation process. The LAB com-
brewing process of SAV. position changed along with the fermentation process due to
distinct predilections to circumstance. Ethanol concentration
Isolation and identification of LAB increased in the AF stage, and ethanol concentration de-
along with fermentation process creased and acidity and temperature became the main envi-
ronmental stressors in the AAF stage according to the CCA
In previous reports, the succession of microbial community result (Fig. 2c). Thus, some LAB could exist continuously
was mainly analyzed using a culture-independent method throughout the fermentation process; for example,
Lactobacillus helveticus appeared almost in the entire fermen-
tation process. By contrast, Pediococcus acidilactici was iso-
Table 1 Microbial diversity indices of LAB communities. d: Berger-
Parker index, dMa: Margalef index, D: Simpson index, H′: Shannon lated only from samples of Daqu and AAF stages, and
index, Je: Pielou index Lactobacillus fermentum was isolated from the AF stage.
This result might be due to their different adaptability to en-
Sample number d dMa D H′ Je
vironmental stresses that were detected. Furthermore,
Daqu 2.45 0.43 0.66 1.55 0.98 Lactobacillus plantarum and Lactobacillus coryniformis were
AF1d 3.91 1.95 0.85 2.93 0.88 only detected only in AF1d, and Lactobacillus rhamnosus
AF4d 4.34 2.17 0.86 2.97 0.86 was detected only in the aging sample due to its extremely
AF7d 4.45 2.17 0.85 2.89 0.84 acidic and irradiant environment.
AF10d 4.10 1.52 0.82 2.59 0.86
AAF0d 2.87 0.87 0.74 2.06 0.89 Characteristics of LAB regarding environment
AAF1d 2.02 0.87 0.65 1.75 0.75
AAF3d 1.84 0.65 0.61 1.54 0.77 Ethanol and acetic acid, as the most significant products of AF
AAF5d 1.62 0.65 0.52 1.27 0.63
and AAF stages, respectively, were mainly converted by
AAF7d 1.54 0.43 0.47 1.00 0.63
yeasts from sugars and by AAB from ethanol (Nie et al.
2015). In addition to ethanol and acidity, high temperature
Appl Microbiol Biotechnol
also affected the LAB distribution, which was mainly due to correlations were determined between the LAB growth at
microorganism metabolism. The most adaptive temperatures each stage and environmental factors that varied with fermen-
of LAB were 30 to 40 °C (Waters et al. 2015), and the peak tation. These findings agreed with CCA results. The correla-
temperature of AAF reached 47 °C (Fig. 2a). Therefore, the tion coefficients between cell growth and ethanol, acetic acid,
characters of LAB on environmental tolerance were studied and temperature were 0.866, 0.756, and 0.803, respectively.
with acetic acid, ethanol, and temperature as indicators to Results demonstrated that ethanol, acetic acid, and tempera-
reveal the succession sequence of LAB throughout the ture are important environmental factors that drive the succes-
brewing process of SAV. sion of the LAB community and other microorganisms. The
As shown in Fig. 4a–c, cell growth under environmental growth of some LAB strains with a low tolerance against
stress was distinguished along with the fermentation process. environmental factors was inhibited, whereas LAB strains
The growth capacity of LAB from different stages against with a high environmental tolerance grew properly and be-
environmental stress was in accordance with the trend of en- came the predominant LAB community for respective stages.
vironmental factors throughout the fermentation. Remarkable LAB from AAF grew better than those from AF stages under
10 g/L of acetic acid or 47 °C stress, whereas LAB from AF influenced by acetic acid and temperature. LAB in AAF0d
grew better than those from AAF stages under 70 g/L of eth- were influenced by ethanol, acetic acid, and temperature.
anol stress. Moreover, we compared the growths of LAB un- Acetic acid showed almost the same inhibition effect com-
der gentle and harsh environments, including acidity stress of pared with that of temperature, which agreed with the CCA
5 or 20 g/L, ethanol stress of 60 or 80 g/L, and thermal stress results.
of 30 or 55 °C. However, no distinct differences in growths
were observed among the LAB strains from different stages Characteristics of LAB regarding substrate
due to low inhibition conditions (5 g/L of acetic acid, 60 g/L of
ethanol, and 30 °C) or excessive inhibition conditions (20 g/L Nutrition was another factor that affected the succession of
of acetic acid, 80 g/L of ethanol, and 55 °C) (data not shown). LAB community in addition to other environmental factors.
Tolerance variations of LAB against environmental factors For SAV fermentation, Jiulao and Cupei samples contained
were roughly matched with the variations in physiochemical abundant sugars, alcohols, amino acids, organic acids, and
indicators (Fig. 4a–c). other materials (Li et al. 2016b). Their changes in composition
PCA was used to evaluate LAB tolerance against environ- were closely related to microbial metabolism. A GP plate was
mental stresses. In Fig. 4d, the two axes explain 88.5% of the used to analyze the diversity of LAB in substrate metabolism,
overall environmental effect on LAB growth. LAB in AF1d, which could indicate the substrate utilization capabilities of
AF4d, AF7d, and AF10d were greatly influenced by ethanol. LAB strains. Figure 5 describes the heat map of the
LAB in AAF1d, AAF3d, AAF5d, and AAF7d were greatly substrate-decomposing capability of LAB strains. More
Fig. 4 Characteristics of LAB tolerance against environmental factors. a fermentation process; solid curves indicate the average growth of LAB
Environmental tolerance of LAB against acetic acid (10 g/L) throughout under different environmental stresses; dashed curves indicate the
the fermentation process. b Environmental tolerance of LAB against parameter change throughout the fermentation process. d PCA of LAB
ethanol (70 g/L) throughout the fermentation process. c Environmental tolerance on environmental factors; arrows indicate the direction and
tolerance of LAB against temperature (47 °C) throughout the magnitude of environmental factors associated with the LAB tolerance
Appl Microbiol Biotechnol
substrates could be catalyzed by the strains from the AAF substrate catalytic activity. Substrates in group A showed
stage than those from AF, which indicated higher substrate high-catalysis activities by LAB from AF4d and AAF7d.
utilization by LAB in AAF than that in the AF stage. Moreover, lactic acid could be utilized by most LAB. This
Substrates were clustered into four groups on the basis of result was consistent with the concentration change of lactic
acid in the fermentation process, which first increased and A scatter plot was created with the PCA results (Fig. 6a),
then decreased (as shown in Fig. 2a). The concentration of stating that 72.0% (18 over 25) of LAB in the first group were
acetic acid continuously increased; therefore, the total acid mainly isolated from AF, and 59.3% (16 over 27) of LAB in
consistently increased (Li et al. 2016b). Sugars, lactulose, ino- the second group were mainly isolated from AAF stages. The
sine, uridine, and D-galactose exhibited strong positive reac- plot suggested similar metabolic characteristics of LAB from
tions in AF4d. In the AF stage, sugars were metabolized into each fermentation stage. Particularly, group 3 was isolated
ethanol by yeasts, but they could also be metabolized by LAB from AAF7d, the later period of AAF with high acidity.
into multiple organic acids and other metabolites (Nie et al. Calculation of the average overall metabolism of LAB for
2015). Substrates in group B were mainly catalyzed by LAB each day indicted that PCA could explain 73.9% of the infor-
from AAF7d and the aging stage, including lactulose that was mation, as shown in Fig. 6b. LAB from AF and AAF stages
utilized by many LAB (Devos and Vaughan 1994). For group were divided into two groups, except LAB from AAF7d that
C, LAB from AAF5d could metabolize most substrates. L- were far from others due to the high acidity and low sugar
Arabinose, glycerin, pyruvic acid, and α-methyl-D-galactose concentration at the seventh day of AAF (as shown in Fig. 2a).
glucoside showed strongly positive reactions from LAB in
AF10d. D-Cellobiose, 5′-uridine phosphate, and D-sorbitol
showed strongly positive reactions from LAB in AAF7d. Discussion
Substrates in group D exhibited strongly positive reac-
tions from LAB in Daqu, AAF7d, and the aging stages. The importance of microbial communities for the fermenta-
Similar reaction profiles between LAB in AAF7d and tion and flavor formation of SAV SSF process is becoming
the aging stage were mainly because AAF7d was affected increasingly clear (Li et al. 2016b); however, dissecting the
by the aging process. formation mechanisms of these communities remains poorly
Principal component comprehensive score was used to identified due to their high species diversity, low cultivability,
evaluate the substrate utilization capability of LAB according and incapability to easily simulate their natural environment
to Biology system data (GEN III Microstatian, Biolog, (Wolfe et al. 2014). Recent studies have applied culture-
Hayward CA). Table 2 shows that the scores of the first 18 independent methods, including denaturing gradient gel elec-
compounds are mainly influenced by sugars, in addition to D- trophoresis and high-throughput sequencing, to evaluate the
lactic acid methyl ester, lactic acid, and pyruvic acid, which microbial ecology of traditional fermented food, including
indicated that those compounds could be utilized by most vinegars, with advantages of high speed, accuracy, and ability
isolates. Sugars, including glucose, mannose, melibiose, su- to determine uncultured microorganisms (Li et al. 2015a, b).
crose, and maltose, are considered important factors that affect However, results of such culture-independent methods are
LAB succession and are the predominant in carbon nutrition sometimes influenced by some strains that are inactivated or
in SAV fermentation for cell growth and metabolism (Vos even dead. Therefore, in addition to advances in the direct
et al. 1998). Dextrin is a polycarbohydrate, is detected in study of complex microbial communities, culture-dependent
SAV fermentation, and is the product of starch hydrolyses methods can help in the identification and characterization of
from raw materials, sorghum, and bran (Bernal et al. 2002). functional strains and further facilitate works toward under-
L-Lactic acid is a common metabolic product of LAB, but it standing the succession mechanism (Wolfe et al. 2014; Wu
could also be utilized by LAB under the catalysis of lactate et al. 2011). This research revealed the diversity and succes-
dehydrogenase (Koo et al. 2005). D-Lactic acid methyl ester sion sequence by combining culture-dependent and culture-
could be hydrolyzed by methyl lactate hydrolase to produce independent methods.
lactate and methyl groups (Murphy et al. 2016) and then be High-throughput sequencing described the succession of
utilized by LAB. Pyruvic acid is an important intermediate LAB community. CCA indicated that ethanol, acidity, and
metabolite, which has been proven important for LAB acidity temperature, which are natural products from microorganism
resistance (Viana et al. 2005). 2,3-Butanediol could be cata- metabolism accumulated from the fermentation process, are
lyzed by some LAB strains, especially from AAF7d, which potential important factors that influence the succession
could be catalyzed into acetoin, a flavor compound in SAV of LAB community, as shown in Fig. 2c. Our findings
fermentation, by dehydrogenation of LAB (Moens et al. are consistent with those of recent studies that indicated
2014). Acetoin could then react with amino compounds to that microbial communities are largely affected by envi-
form tetramethylpyrazine (Lu et al. 2016), a physiological ronmental variables and substrates (i.e., moisture is the best
activity ingredient. predictor of cheese rind community composition) (Li et al.
PCA was employed to assess the metabolic characteristics 2016a; Wolfe et al. 2014).
of 57 isolates (Fig. 6a) and LAB from different fermentation LAB strains were isolated periodically according to the
stages (Fig. 6b). Figure 6a shows that the PCA for the propor- fermentation process to reveal the real diversity of LAB in
tion of total variance could explain 63.7% of the information. different fermentation stages. After repeated strains were
Appl Microbiol Biotechnol
Pediccoccus pentosaceus, from aging SAV samples with ap- Compliance with ethical standards
proximately 8% (w/v) acidity (Wu et al. 2011). Therefore, the
Conflict of interest The authors declare that they have no conflict of
action of microbial metabolism on SAV aging should be fur-
interest.
ther studied.
Moreover, we focused on the LAB community succession Ethical statement This article does not contain any studies with human
sequence driven by environmental variables. The environ- participants or animals performed by any of the authors.
mental tolerance experiment proved that strains from the AF
stage showed better tolerance against ethanol, which was the
main product in AF (Solieri and Giudici 2009), and LAB from References
the AAF stage showed better tolerance against acetic acid and
temperature, which are the main products in AAF. These re- Altschul SF, Madden TL, Schäffer AA, Zhang JH, Zhang Z, Miller W,
sults indicated that environmental disturbances leave strains Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new gener-
with excellent environmental tolerance and became dominant ation of protein database search programs. Nucleic Acids Res
25(17):3389–3402. https://doi.org/10.1093/nar/25.17.3389
in each stage, to result in the succession of the LAB commu-
Bernal MJ, Periago MJ, Ros G (2002) Effects of processing on dextrin,
nity in SAV fermentation. Only appropriate environmental total starch, dietary fiber and starch digestibility in infant cereals. J
stresses could drive the succession of the LAB community. Food Sci 67(3):1249–1254. https://doi.org/10.1111/j.1365-2621.
More gentle or harsh environments (less or more than 60 or 2002.tb09485.x
80 g/L of ethanol, 5 or 20 g/L of acetic acid, and 30 °C or Cocchi M, Durante C, Grandi M, Lambertini P, Manzini D, Marchetti A
(2006) Simultaneous determination of sugars and organic acids in
55 °C temperature) would not affect the LAB succession. aged vinegars and chemometric data analysis. Talanta 69(5):1166–
Although individual traits were previously determined to scale 1175. https://doi.org/10.1016/j.talanta.2005.12.032
up the bacterial community succession (Salles et al. 2009), our Devos WM, Vaughan EE (1994) Genetics of lactose utilization in lactic
results provided insights into the environmental- and acid bacteria. FEMS Microbiol Rev 15(2–3):217–237. https://doi.
org/10.1016/0168-6445(94)00092-d
metabolic-driven succession sequence of the LAB com-
Gevers D, Huys G, Swing J (2001) Applicability of rep-PCR fingerprint-
munity assemblies during the open-SSF process. Such ing for identification of Lactobacillus species. FEMS Microbial Lett
environmental- and metabolic-driven succession changes 205(1):31–36. https://doi.org/10.1111/j.1574-6968.2001.tb10921.x
might also apply to other microorganisms in SAV fer- Jeong SH, Jung JY, Lee SH, Jin HM, Jeon CO (2013) Microbial succes-
mentation. As reported previously, LAB are confronted sion and metabolite changes during fermentation of dongchimi, tra-
ditional Korean watery kimchi. Int J Food Microbiol 164(1):46–53.
with abiotic and biotic stresses during food fermentation https://doi.org/10.1016/j.ijfoodmicro.2013.03.016
and in complex ecosystems (Papadimitriou et al. 2016). In the Koo OK, Jeong DW, Lee JM, Kim MJ, Lee JH, Chang HC, Kim JH, Lee
majority of studies on LAB stress physiology, strains could HJ (2005) Cloning and characterization of the bifunctional alcohol/
adapt to stresses by appropriate molecular responses, includ- acetaldehyde dehydrogenase gene (adhE) in Leuconostoc
ing those of genes and proteins, to ameliorate negative effects mesenteroides isolated from kimchi. Biotechnol Lett 27(7):505–
510. https://doi.org/10.1007/s10529-005-2541-z
and restore growth or survival potential (Papadimitriou et al. Kostic AD, Gevers D, Pedamallu CS, Michaud M, Duke F, Earl AM,
2016). Thus, unraveling gene expression, which is related to Ojesina AI, Jung J, Bass AJ, Tabernero J, Baselga J, Liu C,
the succession of LAB community, via metatranscriptomic Shivdasani RA, Ogino S, Birren BW, Huttenhower C, Garrett WS,
approaches is ongoing in further studies to understand the Meyerson M (2012) Genomic analysis identifies association of
Fusobacterium with colorectal carcinoma. Genome Res 22(2):
mechanism. 292–298. https://doi.org/10.1101/gr.126573.111
Theoretically, the thermo-, acid-, or temperature-tolerant Li P, Liang H, Lin WT, Feng F, Luo L (2015a) Microbiota dynamics
LAB obtained in this research might improve SAV fermenta- associated with environmental conditions and potential roles of cel-
tion due to their increased metabolism activity under environ- lulolytic communities in traditional Chinese cereal starter solid-state
fermentation. Appl Environ Microbiol 81(15):5144–5156. https://
mental stresses (Li et al. 2017). Therefore, a bioaugmenting
doi.org/10.1128/AEM.01325-15
fermentation technology using environment-tolerant LAB as Li P, Lin W, Liu X, Wang X, Gan X, Luo L, Lin WT (2017) Effect of
inoculum is being performed in our laboratory to improve the bioaugmented inoculation on microbiota dynamics during solid-
substrate utilization and production of SAV. state fermentation of Daqu starter using autochthonous of Bacillus,
Pediococcus, Wickerhamomyces and Saccharomycopsis. Food
Microbiol 61:83–92. https://doi.org/10.1016/j.fm.2016.09.004
Acknowledgments We appreciate the Shanxi Zilin Vinegar Industry Co. Li P, Lin W, Liu X, Wang X, Luo L (2016a) Environmental factors
for providing samples. affecting microbiota dynamics during traditional solid-state fermen-
tation of Chinese Daqu starter. Front Microbiol 7(253):1237. https://
Funding information This work was supported by the National Key doi.org/10.3389/fmicb.2016.01237
R&D Program of China (2016YFD0400505), the National Natural Li S, Li P, Feng F, Luo LX (2015b) Microbial diversity and their roles in
Science Foundation of China (31671722), the Tianjin Municipal the vinegar fermentation process. Appl Microbiol Biotechnol
Science and Technology Commission (16YFZCNC00650), and the 99(12):4997–5024. https://doi.org/10.1007/s00253-015-6659-1
Program for Changjiang Scholars and Innovative Research Team in Li S, Li P, Liu X, Luo L, Lin W (2016b) Bacterial dynamics and metab-
University (IRT15R49). olite changes in solid-state acetic acid fermentation of Shanxi aged
Appl Microbiol Biotechnol
vinegar. Appl Microbiol Biotechnol 100(10):4395–4411. https://doi. Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB,
org/10.1007/s00253-016-7284-3 Lesniewski RA, Oakley BB, Parks DH, Robinson CJ, Sahl JW, Stres
Liang J, Xie J, Hou L, Zhao M, Zhao J, Cheng J, Wang S, Sun BG (2016) B, Thallinger GG, Van Horn DJ, Weber CF (2009) Introducing
Aroma constituents in Shanxi aged vinegar before and after aging. J mothur: open-source, platform-independent, community-supported
Agr Food Chem 64(40):7597–7605. https://doi.org/10.1021/acs. software for describing and comparing microbial communities.
jafc.6b03019 Appl Environ Microbiol 75(23):7537–7541. https://doi.org/10.
Liu DR, Zhu Y, Beeftink R, Ooijkaas L, Rinzema A, Chen J, Tramper J 1128/AEM.01541-09
(2012) Chinese vinegar and its solid-state fermentation. Food Rev Solieri L, Giudici P (2009) Vinegars of the world. Springer, Milan. https://
Int 20(4):407–424. https://doi.org/10.1081/FRI-200033460 doi.org/10.1007/978-88-470-0866-3
Lu ZM, Liu N, Wang LJ, LH W, Gong JS, Yu YJ, Li GQ, Shi JS, Xu ZH Viana R, Perez-Martinez G, Deutscher J, Monedero V (2005) The glyco-
(2016) Elucidating and regulating the acetoin production role of lytic genes pfk and pyk from Lactobacillus casei are induced by
microbial functional groups in multispecies acetic acid fermentation. sugars transported by the phosphoenolpyruvate:sugar phosphotrans-
Appl Environ Microbiol 82(19):5860–5868. https://doi.org/10. ferase system and repressed by CcpA. Arch Microbiol 183(6):385–
1128/AEM.01331-16 393. https://doi.org/10.1007/s00203-005-0003-6
Masco L, Huys G, Gevers D, Verbrugghen L, Swings J (2003) Vos WM, Hols P, Kranenburg RV, Luesink E, Kuipers O, Oost JVD,
Identification of Bifidobacterium species using rep-PCR fingerprint- Kleerebezem M, Hugenholtz J (1998) Making more of milk sugar
ing. Syst Appl Microbiol 26(4):557–563. https://doi.org/10.1078/ by engineering lactic acid bacteria. Int Dairy J 8(3):227–233. https://
072320203770865864 doi.org/10.1016/s0958-6946(98)00049-1
Moens F, Lefeber T, De Vuyst L (2014) Oxidation of metabolites highlights Wang HY, Xu Y (2015) Effect of temperature on microbial composition of
the microbial interactions and role of Acetobacter pasteurianus during starter culture for Chinese light aroma style liquor fermentation. Lett
cocoa bean fermentation. Appl Environ Microbiol 80(6):1848–1857. Appl Microbiol 60(1):85–91. https://doi.org/10.1111/lam.12344
https://doi.org/10.1128/AEM.03344-13 Wang Q, Garrity GM, Tiedje JM, Cole JR (2007) Naive Bayesian classi-
Murphy BM, Letterio MP, Xu B (2016) Catalytic dehydration of methyl fier for rapid assignment of rRNA sequences into the new bacterial
lactate: reaction mechanism and selectivity control. J Catal 339:21– taxonomy. Appl Environ Microbiol 73(16):5261–5267. https://doi.
30. https://doi.org/10.1016/j.jcat.2016.03.026 org/10.1128/AEM.00062-07
Nemergut DR, Knelman JE, Ferrenberg S, Bilinski T, Melbourne B, Jiang Waters DM, Mauch A, Coffey A, Arendt EK, Zannini E (2015) Lactic
L, Violle C, Darcy JL, Prest T, Schmidt SK, Townsend AR (2016) acid bacteria as a cell factory for the delivery of functional biomol-
Decreases in average bacterial community rRNA operon copy num- ecules and ingredients in cereal-based beverages: a review. Crit Rev
ber during succession. ISME J 10(5):1147–1156. https://doi.org/10. Food Sci 55(4):503–520. https://doi.org/10.1080/10408398.2012.
1038/ismej.2015.191 660251
Nie ZQ, Zheng Y, Wang M, Han Y, Wang YN, Luo JM, Niu DD (2013) Wolfe BE, Button JE, Santarelli M, Dutton RJ (2014) Cheese rind com-
Exploring microbial succession and diversity during solid-state fer- munities provide tractable systems for in situ and in vitro studies of
mentation of Tianjin duliu mature vinegar. Bioresour Technol microbial diversity. Cell 158(2):422–433. https://doi.org/10.1016/j.
148(8):325–333. https://doi.org/10.1016/j.biortech.2013.08.152 cell.2014.05.041
Nie ZQ, Zheng Y, Du HF, Xie SK, Wang M (2015) Dynamics and diver- Wu JJ, Gullo M, Chen FS, Giudici P (2010) Diversity of Acetobacter
sity of microbial community succession in traditional fermentation pasteurianus strains isolated from solid-state fermentation of cereal
of Shanxi aged vinegar. Food Microbiol 47:62–68. https://doi.org/ vinegars. Curr Microbiol 60(4):280–286. https://doi.org/10.1007/
10.1016/j.fm.2014.11.006 s00284-009-9538-0
Nie ZQ, Zheng Y, Xie SK, Zhang XL, Song J, Xia ML, Wang M (2017) Wu JJ, Ma YK, Zhang FF, Chen FS (2011) Culture-dependent and
Unraveling the correlation between microbiota succession and me- culture-independent analysis of lactic acid bacteria from Shanxi
tabolite changes in traditional Shanxi aged vinegar. Sci Rep-UK aged vinegar. Ann Microbiol 62(4):1825–1830. https://doi.org/10.
7(1):9240. https://doi.org/10.1038/s41598-017-09850-6 1007/s13213-011-0396-1
Oguntoyinbo FA, Narbad A (2012) Molecular characterization of lactic Wu JJ, Ma YK, Zhang FF, Chen FS (2012) Biodiversity of yeasts, lactic
acid bacteria and in situ amylase expression during traditional fer- acid bacteria and acetic acid bacteria in the fermentation of BShanxi
mentation of cereal foods. Food Microbiol 31(2):254–262. https:// aged vinegar^, a traditional Chinese vinegar. Food Microbiol 30(1):
doi.org/10.1016/j.fm.2012.03.004 289–297. https://doi.org/10.1016/j.fm.2011.08.010
Papadimitriou K, Alegria A, Bron PA, de Angelis M, Gobbetti M, Yuan W, Chai TJ, Miao ZM (2010) ERIC-PCR identification of the
Kleerebezem M, Lemos JA, Linares DM, Ross P, Stanton C, spread of airborne Escherichia coli in pig houses. Sci Total
Turroni F, van Sinderen D, Varmanen P, Ventura M, Zuniga M, Environ 408(6):1446–1450. https://doi.org/10.1016/j.scitotenv.
Tsakalidou E, Kok J (2016) Stress physiology of lactic acid bacteria. 2009.12.019
Microbiol Mol Biol R 80(3):837–890. https://doi.org/10.1128/ Zheng XW, Yan Z, Nout MJ, Smid EJ, Zwietering MH, Boekhout T, Han
MMBR.00076-15 JS, Han BZ (2014) Microbiota dynamics related to environmental
Salles JF, Poly F, Schmid B, Roux XL (2009) Community niche predicts conditions during the fermentative production of Fen-Daqu, a
the functioning of denitrifying bacterial assemblages. Ecology Chinese industrial fermentation starter. Int J Food Microbiol 182-
90(12):3324–3332. https://doi.org/10.1890/09-0188.1 183:57–62. https://doi.org/10.1016/j.ijfoodmicro.2014.05.008