Bacteria Community Structure Analysis of Daqu in Different Storage Periods of SheDian Liquor

yanbo liu Henan University of Animal Husbandry and Economy https://orcid.org/0000-0002-9017-5861 Shangping Jin Henan University of Animal Husbandry and Economy Zhijun Zhao Henan University of Animal Husbandry and Economy Junying Fu Henan University of Animal Husbandry and Economy Xian Wang Henan University of Animal Husbandry and Economy Xiyu Sun Henan University of Animal Husbandry and Economy Suna Han Henan University of Animal Husbandry and Economy Chunmei Pan (  chunmeipan66@hotmail.com ) Henan University of Animal Husbandry and Economy

and large biomass, which easily lead to rapid temperature rise and excessive total acids in fermented grains (Song et al. 2017). Moreover, too long storage period will result in severe loss of starches from Daqu masses (Ståhl et al. 2012). So far, the maturity of Daqu during production is mainly decided according to operating experience, which is limited by certain risks and is unfavorable for the prediction of subsequent production.
The Daqu at different storage periods has been mainly studied by traditional isolated culture or polymerase chain reaction-denaturing gradient gel electrophoresis (PCR-DGGE). However, the microbial strain isolation is feasible only for research on separable species in Daqu (Pace et al. 1997). The PCR cloning library, PCR-DGGE, and single-strand conformation polymorphism ampli cation are limited by huge workloads and low sensitivity and thus cannot correctly re ect the varying trends of microbial communities(Dalmasso et al. 2016). In comparison, high-throughput sequencing can rapidly analyze the diversity of complex microbial communities and detect the unculturable microorganisms with low existing rates (Tang et al. 2017). This technique has been extensively applied to research on biological In recent years, high-throughput sequencing has been gradually used into research on liquor production, but relevant research on bacterial community variation during storage period of Daqu is concentrated in Sichuan, but not in Henan. In fact, the Daqu from different regions is largely different in microorganism species and composition owing to the discrepancy in raw material composition, technology and climate.
Hence, in this study, Daqu in Shedian liquor at different storage periods was studied. Total DNA was extracted from Daqu microorganisms and after PCR ampli cation, the bacterial sequences of Daqu were analyzed via high-throughput sequencing. The differences in bacterial community compositions at different storage periods were investigated. From the perspectives of basic composition and abundance variation of bacteria, the ndings will theoretically underlie the reasonable judgement of optimal Daqu storage time and Luzhou-avor liquor starter propagation process.
From Daqu of the same quality and different batches, a ve-point sampling method was adopted, and then the samples were crushed, mixed, sealed and stored at -20°C.

DNA extraction
Total DNA was extracted by using a FastDNA SPIN Kit for Soil according to the manual. DNA concentration and purity were detected on a NanoDrop2000 instrument, and DNA quality was analyzed via 1% agarose gel electrophoretic analysis.
Quantitative analysis was conducted via QuantiFluor™-ST. The library establishment and high-throughput sequencing were nished by Sangon Biotech (Shanghai) Co., Ltd. (China).

Data processing
The original sequences were sent to quality control on Trimmomatic and spliced on Flash. The operational taxonomic units (OTUs) were clustered on Uparse at the similarity of 97%, and chimaeras were deleted on Uchime. The species of each sequence were classi ed and annotated with an RDP classi er and compared with the Silva database (SSU123) at the threshold of 70%.

Sequence number and diversity indices
The PE reads obtained from Miseq sequencing were rst spliced as per the overlap relationship, and the sequences were quality-controlled and ltrated. After that, the sequences were sent to OTU clustering and species taxonomic analysis. Based on OTUs, diversity indices and sequencing depth were analyzed. As the sequence number increased, the Shannon index dilution curves of all samples became attened (Fig. 2). The Coverage rates of all samples exceeded 99% (Table 1), indicating the sequencing depth of this study was enough for analysis of bacterial community composition and diversity. After quality control, 75402, 47143 and 49840 sequences were identi ed from the 3-, 6-and 9-month-old Daqu respectively. OTU clustering analysis at the similarity above 97% returned 458, 469 and 237 OTUs respectively. The Alpha diversity indices usually consist of Shannon index and Chao1 index. A larger Shannon index indicates a higher diversity of the tested community, and the Chao1 index often estimates total species number, and a larger Chao1 means higher species abundance. Results showed the Shannon index and Chao1 both maximized in the 6-month-old Daqu (3.843066 and 414.1 respectively), indicating the diversity of bacterial communities was the highest in the 6-month-old Daqu compared with the other two types of Daqu.

OTU Venn diagram
Venn diagrams can statistically re ect the numbers of shared OTUs and exclusive OTUs in a certain sample and visually display the similarity and overlap of OTUs.

Phylogenetic tree analysis
The sequences corresponding to the top 50 species in terms of abundance were selected and thereby a phylogenetic tree as per the maximum likelihood method was built on FastTree. The phylogenetic tree was plotted with R-language (Fig. 6).
Based on the phylogenetic tree, the biological evolution can be deduced so as to uncover the evolutionary history and mechanism.

Discussion
The bacterial communities of Daqu at different storage periods were analyzed with high-throughput sequencing, which uncovered totally 20 phyla, 38 classes, 84 orders, 138 families, 268 genera and 396 species. With the threshold at relative abundance > 0.5% at the phylum level, 6 predominant bacterial The exclusive genera found in the 6-month-old Daqu were Terrisporobacter, Clostridium sensu stricto 1, and Ruminococcaceae UCG-005, which were caproic acid bacteria, metabolites of which are the major fragrant substances of Luzhou-avor liquors. The Lactobacillus, Lactococcus and Weissella identi ed from the 3-month-old Daqu were all lactic acid bacteria (Horvath et al. 2009). During the production of white liquors, lactic acid bacteria can ferment carbohydrates, produding abundant lactic acids, but excessive lactic acids will lead to a bitter taste of liquors and decrease the liquor quality. Moreover, the acid environment will produce bioamine (Linares et al. 2011), but the intake of excessive bioamine into the body will cause adverse reactions (Mah et al. 2002). The Clostridium sensu stricto 1 identi ed from the 6-month-old Daqu belongs to Clostridia, which is a dominant class of caproic acid bacteria during the production of Luzhou-avor liquors (Hu et al. 2015). Terrisporobacter and Ruminococcaceae are also caproic acid bacteria (Zheng et al. 2013). The Enterobacter found in the 9-month-old Daqu can ferment glucoses to form lactic acid (Li et al. 2015) and is also a bioamine-producing bacterium in Daqu ( Totally 396 bacterial species were identi ed through high-throughput sequencing, and a resource library of Daqu bacteria from Luzhou-avor liquors after different storage periods was preliminarily established. The information of bacteria was richer compared with traditional isolating culture. This study enriches the knowledge on the diversity of bacterial communities in the Daqu of Luzhou-avor liquors and will theoretically underlie the rational judgement of Daqu storage time during practical productions and promote the standardized and scienti c Daqu production.

Declarations
Ethics approval and consent to participate: This article does not contain any studies with human participants or animals performed by any of the authors.
Consent for publication: Not applicable.
Availability of data and materials: The datasets used or analysed during the current study are available from the corresponding author on reasonable request.    Phylogenetic tree