Study On Gujingtribute Microbial Community Structure And Its Association With Liquor Flavor Molecules

Gujing Tribute liquor,one of eight famous liquor brands in China,owns strongflavor features that derive from its specific fermentation microbial community.Analysis of the specific Gujing Tribute liquor fermentation microbial community structure can help to know its microbial composition;the comparison between the good-quality fermentation microbial community and the ordinary-quality fermentation microbial community can be used to pinpoint the key different species or genus;both the flavor compositional difference between good-quality liquor and ordinary-quality liquor,and the metabolic compositional difference between goodquality pitmud and ordinary-quality pitmud help to explain the difference between good-quality microbial community and ordinary-quality microbial community from the point of view of fermentation process.This study is of great practical significance to provide theoretical basis for the optimization of the oridinary-quality fermentation microbial community and the improving of liquor quality.The microbial composition of Gujing Starters,Pitmud,and Zaopei(also known as fermented grains)were illuminated by near full-length 16 S rDNA/ITS TA clone library sequencing,Illumina Miseq 16 S rDNA V3-V4 region sequecing and metagenomics de nove sequencing:(1)There are five relatively specific species in Daqu: Pantoea agglomerans,Pantoea vagans,Talaromyces thermophiles,Thermomyces lanuginosus,and Thermoascus aurantiacus;(2)The microbial community of Gujing Tribute pitmud is very complex and contains a large number of un-cultured/unknown bacteria;(3)In the fermented grains,Lactobacillus was the monopoly genus,and the most abundant bacteria was Lactobacillus acetotolerans.The microbial compositional differences between good-quality and ordinaryquality pits were analyzed through principal component analysis(PCA),difference significant analysis and correlation analysis,and the following conclusions were obtained:(1)There were big differences between good-quality and ordinary-quality pitmud microbial communities at both the phylum and the genus levels,and the most significant difference was the content of Lactobacillus,which was about 34% higher in ordinary-quallity pitmud;(2)The content difference of six genera between goodquality and ordinary-quality pitmud was also obvious: the content of Klebsiella was relatively higher in ordinary-quality pitmud;the contents of Clostridium,Syntrophomonas,Methanoculleus,Aminobacterium and Sedimentibacter were relatively higher in good-quality pitmud;(3)There was almost no microbial compositional difference between good-quality and ordinary-quality afterfermentation-Zaopei samples,but their internal relationships between the OTUs were different.The species-specific primers of seven species(Lactobacillus plantarum,Bacillus subtilis,Lactobacillus fermentum,Bacillus licheniformis,Staphylococcus saprophyticus,Enterobacter asburiae,and Pseudomonas putida)were successfully designed according to their genome sequence,and the specificity of the 7 primers was validated,then the content of seven species in several bacterial communites were determined by absolute quantitative qPCR method.In the future,with more and more species' whole genome sequence published,it is expected to design more speciesspecific primers based on whole genome sequence information and to achieve rapid monitoring of the microbial community fluctuation using absolute qPCR method,early warning of liquid quality.The composition of 56 typical flavor substances in 127 liquor samples of two fermentation batches from 3 plants was analyzed by gas chromatography.The most taste-relevant flavors and most pit-age-relevant flavors were obtained through correlation analysis.The relationship between flavors and corresponding microbial communities was established by CCA(Canomical Correlation Analysis)method.The quality of liquor samples was evaluated by qualified panelists according to the national standard scoring,and the Least Absolute Shrinkage and Selection Operator(Lasso)linear regression model was used to establish the relationship between the flavors of liquor samples and their taste scores.The two Lasso regression models were established based on two batches of liquor samples separately,and the two groups of flavor substances kept in the two Lasso regression models were surprisingly different.The liquid chromatography-mass spectrometry(LC-MS)was used to detect the detailed metabolic components in pitmud samples with different age /quality from three plants.A total of 3193 metabolic components were analyzed,the information of 30% metabolic components was determined,of which the acids were the most abundant.It was found that the metabolic components of the high-quality pitmud samples of A and B plants were similar,while the metabolic components of the ordinary-quality pitmud samples of C plant were obviously different from those of the A and B plant,and the content of many metabolites in pitmud samples of C plant were more than that of the A/B plant.The differences in the content of the metabolic components and flavor compositons between good-quality and ordinary-quality samples indicated that their metabolic transformation processes were different.The metabolic function difference between high-quality and ordinary-quality pits remains to be further studied.