Isolation, Identification, and Fermentation Medium Optimization of a Caproic Acid‑Producing Enterococcus casseliflavus Strain from Pit Mud of Chinese Strong Flavor Baijiu Ecosystem

Strong flavor baijiu (SFB), a traditional Chinese distilled liquor, contains more than 1,300 different flavor compounds (Ji et al. 2017; Zou et al. 2018). Among these compounds, ethyl hexanoate is the primary flavoring substance of SFB, and its content can directly affect the quality of SFB (Hong et al. 2020). Caproic acid, also known as hexanoic acid, is a common carboxylic acid that can be used as a precursor for producing liquid fuels and synthetic chemicals such as food and animal feed additives (Van Immerseel et al. 2004; Nabi et al. 2006; Serhan et al. 2016). Caproic acid-producing bacteria (CPB) are generally considered the most critical functional microorganisms in an SFB ecosystem (Liu and Sun 2018). The isolation, screening, identification, and culture of CPB are important to improve the quantity and quality of CPB in pit mud (Wang et al. 2019; Zhao et al. 2019; Tian et al. 2020). In addition to the SFB ecosystem, CPB have been isolated from different ecological niches such as anaerobic digestion sludge (Kim et al. 2015), the rumen of cattle and sheep (Weimer and Stevenson 2012), and silt (Gildemyn et al. 2017). The oxygen demands of different strains of CPB during the production of caproic acid are not consistent. Most of the reported CPBs are anaerobic species, including members of Clostridium (Weimer and Stevenson 2012), Bacillus (Zhao et al. 2012), Ruminococcaceae (Zhu et al. 2015), and Megasphaera (Kim et al. 2018). They can accumulate caproic acid by using ethanol, lactic acid, glucose, and D-galactitol as the main carbon source (Wang et al. 2020).

In the present study, a new strain of Enterococcus casseliflavus (strain BF-1) with high caproic acid yield was isolated from the pit mud of an SFB ecosystem. E. casseliflavus has potential probiotic effects on growth performance, immunity, and disease resistance (Akbari et al. 2021). In this study, the whole genome of E. casseliflavus BF-1 was sequenced and annotated. The safety of the strain at the genetic level and the caproic acid biosynthesis pathway were analyzed. We verified the essential nutrients of E. casseliflavus BF-1 synthesis defect through single-omission growth experiments and added essential nutrients to the fermentation medium to improve the yield of caproic acid. The results showed that this strain could be further used as a reinforced inoculum in the SFB production process.

Screening of caproic acid-producing strains. A total of 41 strains were screened from pit mud, and among these strains, 19 strains showed the ability to produce caproic acid (Fig. 1). Strain BF-1 showed the highest yield of caproic acid (0.88 g/l). Therefore, strain BF-1 was selected for further studies.

Fig. 1

Physiological characteristics and phylogenetic classification of strain BF-1. Strain BF-1 grew well on sodium acetate solid medium, with large, smooth, milky white, protuberant, and round colonies; the cells were non-spore-forming Gram-positive cocci (Fig. 2a and 2b). The results for VP test, gelatin hydrolysis, glycerol test, and nitrate reduction test were negative, while those for MR test; starch hydrolysis; and tests of growth using lactose, sucrose, raffinose, mannitol, fructose, stachyose, cellobiose, xylose, galactose, glycerol, and mannose as the sole carbon source were positive (Table II).

Fig. 2

Based on the physiological, biochemical, and 16S rDNA sequence analyses (Fig. 2c), strain BF‑1 was classified as E. casseliflavus. A physiological tree was then constructed based on the alignment of the 16S rDNA of strain BF-1 with other 16S rDNA sequences in the GenBank database. The results revealed that strain BF-1 was closest to E. casseliflavus FDAARGOS 1120 and similarity is 100 %. Therefore, strain BF-1 was designated as E. casseliflavus BF-1. It is the first study to report that E. casseliflavus is a caproic acid producer (Yuan et al. 2022).

Genome properties of E. casseliflavus BF-1. Genome assembly and annotation of E. casseliflavus BF-1. To understand the physiological and metabolic characteristics of E. casseliflavus BF-1, we sequenced and annotated its genome (Table III). The contig total sequence length was 3,383,020 bp, and the contig N50 was 372,495 bp. The scaffold N50 was 570,209 bp. The E. casseliflavus genome is 2,968,377 bp in length, has a GC content of 43.52%, and contains 3,270 coding sequences, three rRNA genes, and 53 tRNA genes. The nucleotide sequence of E. casseliflavus BF-1 is deposited in the NCBI database under the accession number PRJNA759370.

For the COG annotation (Fig. 3a), a total of 2,819 genes were assigned with COG terms, which account for 85.8% of the total genes in E. casseliflavus BF-1. Function unknown (S) was the largest category (24.76%). Apart from S, carbohydrate transport and metabolism (G) and transcription (K) were the most prominent groups, accounting for 13.4% and 9.5%, respectively. According to the KAAS annotation, a total of 3,155 genes were assigned with KO terms. The largest three metabolic pathway categories in the genome of strain BF-1 were carbohydrate metabolism (428), amino acid metabolism (187), and energy metabolism (124) (Fig. 3b).

Fig. 3

Caproic acid biosynthesis pathway analysis. According to the KAAS annotation, we constructed the metabolic pathway of caproic acid in E. casseliflavus BF-1 according to the KAAS annotation. The process of synthesizing caproic acid from glucose by E. casseliflavus BF-1 is divided into three steps (Fig. 4): (1) E. casseliflavus BF-1 produces acetyl CoA through pyruvic acid by pyruvate ferredoxin oxidoreductase (porA); (2) Acetyl CoA is converted to butyl CoA under the action of NADH, FADH₂, and CoA, and butyl CoA was converted to butyric acid and acetyl CoA under the action of acetic acid and CoA transferase; and (3) Hexanoyl CoA and butyric acid are converted into caproic acid and butyryl CoA through CoA transferase. Because acyl CoA transferase has a wide range of catalytic activities, it has been speculated that hexanoyl CoA may also directly react with acetic acid to synthesize caproic acid. The main enzymes involved in (1) and (3) are completely annotated. In contrast enoyl-CoA hydratase (paaF) and crotonyl-CoA(ccrA) reductase involved in (2) are not annotated, which indicates the inability of this strain to synthesize crotonyl-CoA and butanoyl-CoA. It contradicts the experimental result that caproic acid was produced by strain BF-1. There are two explanations for this contradiction: (i) the existence of other non-annotated functional enzymes and (ii) the substitution of other spectral enzymes. It is also possible that 3-hydroxybutyryl-CoA dehydrogenase may replace enoyl-CoA hydratase and glutaconyl-CoA decarboxylase subunit alpha may replace crotonyl-CoA due to their wide use of substrates.

Fig. 4

Biosynthesis and transport of amino acids and vitamins. According to the KAAS annotation, E. casseliflavus BF-1 cannot synthesize four amino acids (cysteine, methionine, tyrosine, and phenylalanine) and four vitamins (niacin, pantothenate, biotin and folic acid) (Supplementary materials). However, E. casseliflavus BF-1 cells can transport these amino acids and vitamins into their cell membrane through ATP-binding cassette (ABC) transporters (Supplementary materials).

Prediction of virulence factors. The virulence factors in the genome of strain BF-1 were annotated by VFanalyzer in VFDB (Table IV). Among all the potential virulence factors of Enterococcus, experimental results have shown that toxins, exoenzymes, and antiphagocytosis proteins are pathogenic factors (Zischka et al. 2012). The genome of strain BF-1 does not contain the genes virulent toxins and exoenzymes, although it contains the genes (cpsA, cpsB, and cpsJ) antiphagocytic proteins. The latter operon, however, needs a complete gene cluster (cpsA, cpsB, cpsC, cpsD, cpsE, cpsF, cpsG, cpsH, cpsI, cpsJ, and cpsK), or else it cannot form a capsule (Thurlow et al. 2009). Therefore, E. casseliflavus BF-1 is safe at the genetic level and could be further used in regulating baijiu microbial ecosystem.

Verification of E. casseliflavus BF-1 defection. In order to verify the results of KAAS annotation, we carried out the single-omission growth experiments experiment of E. casseliflavus. The results are shown in Fig. 5. L-cysteine, L-methionine, L-tyrosine, L-phenylalanine, niacin, pantothenate, biotin, and folic acid were subtracted from the CDM. The OD_600nm of E. casseliflavus BF-1 decreased to 24%, 18%, 30%, 21%, 20%, 88%, 24%, and 42%, respectively. It is confirmed that E. casseliflavus BF-1 has defects in synthesizing the above eight nutrients. In this case, the synthesis of biofilm, nucleic acid, and protein of BF-1 would be affected, which could further influence the energy metabolism, oxygen stress resistance, and the growth of bacteria. However, E. casseliflavus BF-1 cells can transport these nutrients into the cell through ATP-binding cassette (ABC) transporters (Additional File 1). BF‑1 can assimilate these nutrients for growth by adding the missing nutrients to the medium.

Fig. 5

Optimization of medium for caproic acid production based on KAAS annotation. In different carbon sources conditions, the best carbon source for caproic acid production in strain BF-1 was glucose, followed by sodium acetate and mannose (Fig. 6a). Strain BF-1 did not use starch to produce caproic acid. As shown in Fig. 6c, the optimum concentration of glucose added was 2%. Yeast extract as a nitrogen source was found to significantly promote the production of caproic acid by E. casseliflavus BF-1. Yeast extract was therefore selected as a nitrogen source for the growth of E. casseliflavus BF-1 (Christ and Blank 2019). As shown in Fig. 6d, the optimal concentration of yeast extract added was 2.0%. E. casseliflavus BF-1 showed weak caproic acid production when using an inorganic nitrogen source (Fig. 6b).

Fig. 6

Based on a single-omission growth experiment, we confirmed that E. casseliflavus BF-1 has defects in synthesizing eight nutrients. Because organic nitrogen source contains many complex nutritional components and rich bioactive substances, we added essential nutrients to the fermentation medium to explore its effects on the growth and acid production of E. casseliflavus BF-1.

We added L-cysteine, L-methionine, L-tyrosine, L-phenylalanine, niacin, pantothenate, biotin, and folic acid to the fermentation medium (L-cysteine was removed from the control). In the medium supplemented with L-cysteine and biotin, the OD_600nm and yield of caproic acid of E. casseliflavus BF-1 increased significantly compared with the control. The OD_600nm of E. casseliflavus BF-1 increased by 10.1% and 10.5%, respectively, after 10 days of fermentation (Fig. 7a). The yield of caproic acid increased by 6.1% and 13.1%, respectively (Fig. 7b). The nutrients in yeast extract were complex. For the sake of cost, only L-cysteine and biotin were added in the subsequent optimization experiment. Based on Fig. 8a–8h, we determined the value of each factor when the yield of caproic acid reached the maximum: sodium acetate 6 g/l, MgSO₄ · 7H₂O 0.25 g/l, KH₂PO₄ 6 g/l, K₂HPO₄ 8 g/l, (NH₄)₂SO₄ 1 g/l, L-cysteine 0.6 g/l, biotin 0.002 g/l, and ethanol at a concentration of 2%.

Fig. 7

Fig. 8

Orthogonal experiment. Based on the results of single factor optimization experiment, an L₁₆ (44) orthogonal experiment was designed to study further the interaction between four factors, namely A: glucose addition, B: yeast extract addition, C: sodium acetate addition, and D: ethanol addition, and four factors and four levels. The results of the orthogonal experiment are shown in Table III. According to the experimental results (Table V), the Kn value and R-value were calculated. The best fermentation scheme was as follows: A₂B₂C₄D₁, i.e., glucose 20 g/l, yeast extract 20 g/l, sodium acetate 8 g/l, and ethanol added at a concentration of 1%; A was found to be the most significant factor. The results of the variance analysis of the orthogonal experiment are shown in Table VI. From these results, it can be concluded that glucose, yeast extract, and sodium acetate significantly impacted on the test results, and glucose had the most significant impact on caproic acid production of E. casseliflavus BF-1 strain (p = 0.013).

The following optimum medium conditions for acid production by strain BF-1 were obtained by the orthogonal experiment: glucose 20 g/l, yeast extract 20 g/l, sodium acetate 8 g/l, ethanol 1%, MgSO₄ · 7H₂O 0.25 g/l, KHPO6 g/l, KHPO8 g/l, (NH)SO1 g/l, _{24 24 424} L‑cysteine 0.6 g/l, and biotin 0.002 g/l. The yield of caproic acid was 3.03 g/l after fermentation for 10 days at initial pH 7 and 35°C, which was 3.44-fold of that before optimization. The caproic acid yield of BF-1 was higher than those of most caproic acid bacteria (Zhao et al. 2012; Hu et al. 2015; Dobritsa et al. 2017; Gou et al. 2020). E. casseliflavus BF-1 could be further used as an enhanced inoculum during the SFB production to increase the yield of ethyl hexanoate. In addition, E. casseliflavus BF-1 could be applied in pit mud maintenance, artificial pit mud preparation, and esterification liquid production.

CPB is a crucial functional group in an SFB ecosystem, as they affect the content of caproic acid and ethyl hexanoate in SFB. The present study is the first to report caproic acid production of E. casseliflavus BF-1. The genomic analysis confirmed that this strain is not pathogenic. After adding the predicted essential nutrients based on the KAAS annotation, the optimum medium conditions for the acid production of strain BF-1 were glucose 20 g/l, yeast extract 20 g/l, sodium acetate 8 g/l, ethanol 1%, MgSO₄ · 7H₂O 0.25 g/l, KH₂PO₄ 6 g/l, K₂HPO₄ 8 g/l, (NH₄)₂SO₄ 1 g/l, L-cysteine 0.6 g/l, and biotin 0.002 g/l by orthogonal experiment, and the yield of caproic acid reached 3.03 g/l. Strain BF-1 screened in this study can be used for the daily maintenance of pit mud and the rapid culture of artificial pit mud, which can improve the maintenance effect of pit mud, shorten the aging cycle of artificial pit mud, and reduce production cost. This strain can also be used in future animal feed and other fields.