Transcriptome and Metabolome Analysis Revealing the Improved ε-Poly-l-Lysine Production Induced by a Microbial Call from Botrytis cinerea

ABSTRACT

ε-Poly-l-lysine (ε-PL) is a wide-spectrum antimicrobial agent, while its biosynthesis-inducing signals are rarely reported. This study found that Botrytis cinerea extracts could act as a microbial call to induce a physiological modification of Streptomyces albulus for ε-PL efficient biosynthesis and thereby resulted in ε-PL production (34.2 g/liter) 1.34-fold higher than control. The elicitors could be primary isolated by ethanol and butanol extraction, which resulted in more vibrant, aggregate and stronger mycelia. The elicitor-derived physiological changes focused on three aspects: ε-PL synthase, energy metabolism, and lysine biosynthesis. After elicitor addition, upregulated sigma factor hrdD and improved transcription and expression of pls directly contributed to the high ε-PL productivity; upregulated genes in tricarboxylic acid (TCA) cycle and energy metabolism promoted activities of citrate synthase and the electron transport system; in addition, pool enlargements of ATP, ADP, and NADH guaranteed the ATP provision for ε-PL assembly. Lysine biosynthesis was also increased based on enhancements of gene transcription, key enzyme activities, and intracellular metabolite pools related to carbon source utilization, the Embden-Meyerhof pathway (EMP), the diaminopimelic acid pathway (DAP), and the replenishment pathway. Interestingly, the elicitors stimulated the gene transcription for the quorum-sensing system and resulted in upregulation of genes for other antibiotic production. These results indicated that the Botrytis cinerea could produce inducing signals to change the Streptomyces mycelial physiology and accelerate the ε-PL biosynthesis.

IMPORTANCE This work identified the role of microbial elicitors on ε-PL production and disclosed the underlying mechanism through analysis of gene transcription, key enzyme activities, and intracellular metabolite pools, including transcriptome and metabolome analysis. It was the first report for the inducing effects of the “microbial call” to Streptomyces albulus and ε-PL biosynthesis, and these elicitors could be potentially obtained from decayed fruits infected by Botrytis cinerea; hence, this may be a way of turning a biohazard into bioproduct wealth. This study provided a reference for application of microbial signals in secondary metabolite production, which is of theoretical and practical significance in industrial antibiotic production.

INTRODUCTION

The biopolymer ε-poly-l-lysine (ε-PL) is a natural cation consisting of 25 to 35 l-lysine residues (1). It is mainly produced through submerged fermentation by Actinomyces species, which is not only harmless to humans but can also inhibit the growth of G⁺/G⁻ bacteria and fungi, as well as virus replication. It is of good solubility, thermostability, biodegradability, and edibility and has been widely used as biological preservative in Japan, the United States, Europe, and South Korea for several decades (2). In addition, it serves as drug carrier in disease therapy, as a lipase inhibitor in food therapy, and as a special material in biosensor construction (3). Therefore, it is of great significance to promote the commercial production of ε-PL.

To meet the great demand for this compound, many strategies have been carried out in terms of optimal media developments (4–6), precise process regulation (7, 8), producing strain immobilization (9), and in situ product removal (10). All of these strategies were performed in acidic environments (pH < 4.5). Since a short term of acidic pH shock (pH 3.0 to 3.5) significantly promoted the ε-PL productivity (11), a low-pH environment has received continued attention as a key factor for ε-PL biosynthesis in recent years (12–15). A possible explanation is that the acidic signal stimulates the producing strain to secrete more ε-PL so as to relieve the acidic stress through ε-PL-H⁺ binding (15). However, except for the acidic signal, research concerning other inducing factors for ε-PL biosynthesis is still absent.

The complex habitats of ε-PL-producing strains in nature are filled with microbial interactions and competitions, which are good sources for the discovery of new inducing factors (16). The main ε-PL-producing strains, such as Streptomyces albulus (17), Streptomyces noursei (5), Streptomyces griseofuscus (18), and Kitasatospora sp. (9), show slower growth than bacteria and weaker ability of spore diffusion than fungi. The significance of ε-PL secretion may have been highly competitive during the evolution period of a billion years. For economic considerations, ε-PL will not be overproduced until the producing strain receives threatening signals from other microorganisms. Therefore, it is an interesting topic to study whether other microorganisms can produce signals to induce ε-PL overproduction by the producing strains.

In recent years, studies have reported that biological elicitor could enhance the production of bioactive secondary metabolites in fungi (19–21) and in Actinomyces (22–26). For plant elicitors, salicylic acid results in higher fungal taxol production by Pestalotiopsis microspore and Paraconiothyrium variabile, and so does serine addition in Epicoccum nigrum cultivation (19–21). Gynostemma pentaphyllum herb elicitors, such as ginseng saponin and puerarin, increased antibacterial polysaccharide production by herb endophytic fungus Chaetomium globosum CGMCC 6882, and a coculture of host herb and endophytic fungus has been designed to achieve a further production promotion (27). Some small plant chemicals can also promote the production of antiplasmodial substances (20). Similar inducing effects are observed in microbial elicitors. Luti and Mavituna found that the addition of dead cells of Bacillus subtilis and Staphylococcus aureus can stimulate the biosynthesis of undecylprodigiosin by Streptomyces coelicolor (22). In addition, fungal elicitors also show positive influences on natamycin production by Streptomyces natalensis HW-2 (23, 24, 26) and rimocidin production (25) by Streptomyces rimosus M527. In plants and fungal cells, the underlying mechanisms are more reported (28–31), while relatively little research is carried out in Streptomyces species, and systematic studies are urgently needed.

In the present study, the influences of microbial elicitors on ε-PL production were evaluated, and this “microbial call” was employed to establish an efficient approach to ε-PL production. The underlying mechanisms were finally systematically disclosed by analysis of transcriptome, metabolome, and key physiological parameters.

RESULTS

Effects of microbial elicitors on ε-PL production.

As shown in Table 1, ε-PL strongly inhibited the growth of bacteria (B. subtilis and Escherichia coli), and less inhibited the fungi (Saccharomyces cerevisiae and Botrytis cinerea). However, whether microorganisms could influence ε-PL production is not clear. Fig. 1A and C shows that the broth filtrate of S. cerevisiae and B. cinerea could slightly promote ε-PL production and dried cell weight (DCW) (21% addition, vol/vol), while filtrate from B. subtilis and E. coli decreased the ε-PL accumulation (above 15% addition, vol/vol). By biomass extract addition (Fig. 1B and D), there is a greater tendency for E. coli extract (above 18 g wet cells/liter) inhibition of the ε-PL biosynthesis and cell growth, while extracts of S. cerevisiae and B. cinerea (above 18 g wet cells/liter) resulted in extremely higher ε-PL production. The extracts of the optimal strain (B. cinerea) led to 0.99 g/liter ε-PL production, which was 1.74-fold higher than that of the control (0.57 g/liter).

TABLE 1.

ε-PL susceptibility of four microorganisms in this study^a

	ε-PL concn (μg/mL)a
Test microorganisms	0.1	0.5	1	10	100	200	300	500	1,000
B. subtilis	+	−	−	−	−	−	−	−	−
E. coli	+	−	−	−	−	−	−	−	−
S. cerevisiae	+++	+++	+++	+++	+++	++	−	−	−
B. cinerea	+++	+++	+++	+++	+++	+++	+++	+++	++

^aThe data indicate the growth of microorganisms in a petri plate with ε-PL addition of different concentrations. ε-PL, ε-poly-l-lysine.

FIG 1.

Influences of elicitors from different microorganisms on ε-poly-l-lysine (ε-PL) production. (A) Effects of culture broth filtrate on ε-PL production. (B) Effects of biomass extracts on ε-PL production. (C) Effects of culture broth filtrate on cell growth. (D) Effects of biomass extracts on cell growth. DCW, dried cell weight.

Primary isolation of B. cinerea elicitors.

Primary elicitor isolation of B. cinerea, the optimal elicitor producer, was carried out in Fig. 2A, and the effects of each extract fraction on ε-PL biosynthesis was studied in Fig. 2B. After ethanol extraction (75% ethanol), the addition of the S_n fraction (supernatant) increased ε-PL production, while no significant difference was observed in the culture with the addition of the S_d fraction (sediment). The S_n fraction was extracted with ethyl acetate, chloroform, butyl alcohol, and petroleum ether, individually. Fractions E and B promoted ε-PL production, in which fraction B showed higher ε-PL titer, and fractions C and P resulted in no significant differences. Therefore, the positive elicitor for ε-PL biosynthesis could be dissolved in water, 75% ethanol, and butanol; hence, it is a small polar molecule rather than a macromolecule.

FIG 2.

(A) Primary isolation of elicitors in B. cinerea. (B to E) Influences of extraction fragments on ε-PL production (B), petri plate colony morphology (C), mycelial scanning electron microscopy (SEM) morphology (D), and CTC staining images (E). CK, without elicitor addition; EG, with addition of B. cinerea elicitors (36 g wet cells/liter).

Cell morphology changes after B. cinerea elicitor addition.

The macroscopic morphology of petri plate colonies and the microcosmic morphology of submerged cultured mycelia were compared between culture groups of elicitor group (EG) and control check (CK). In petri plate incubation, the S. albulus in EG exhibited bigger colonies and aggregation state at 2 days of incubation, and this trend of aggregation continued, which resulted in a distinct morphology at 5 days of incubation (Fig. 2C). In submerged cultivation, the cells in EG showed thicker and stronger mycelia (Fig. 2D). The image of laser scanning confocal microscope after 5-cyano-2, 3-ditolyl tetrazolium chloride (CTC) staining revealed that the S. albulus tend to form aggregate and bigger mycelia pellets in submerged culture with elicitor addition, and the respiration activity was also significantly improved. These differences in macroscopic morphology and microcosmic morphology implied that the B. cinerea elicitors resulted in significant physiological changes in S. albulus cells, which might be self-defensive responses to external biological signals.

Optimization of key parameters in ε-PL production with elicitor addition.

Flask’s culture was employed to optimize key parameters in ε-PL production by adding B. cinerea elicitors. Fig. 3A shows that a time range of 0 to 36 h was available for elicitor addition, and the addition at 24 h resulted in the maximum ε-PL titer of 1.21 g/liter. The ε-PL titer was improved with the increasing culture age of B. cinerea during 0 to 4 days and reached the highest ε-PL titer at the cultivation time of 4 days (Fig. 3B). The appropriate range of B. cinerea elicitors concentration was 18 to 48 g wet cells/liter, and the optimal concentration was observed at 36 g wet cells/liter (Fig. 3C).

FIG 3.

(A, B) Optimization of B. cinerea elicitors adding time (A) and B. cinerea culture age (B). (C) Elicitor concentration for ε-PL production. (D to G) Comparison of culture profiles (D, E, G) and kinetics parameters (F) of batch/fed-batch cultures between EG (B. cinerea elicitor addition, 36 g wet cells/liter) and CK (without elicitor addition) for ε-PL production. Arrows indicate the elicitor addition time or equal volume of distilled water.

Batch and fed-batch ε-PL fermentation with elicitor addition in a 5-liter fermenter.

Fig. 3D and E shows the difference of parameters profile between cultures with (EG) and without (CK) elicitor addition. The elicitor addition at 24 h led to rapid cell growth within a short time at the first stage (at 24 to 48 h) and subsequently accelerated the ε-PL biosynthesis at the second stage (at 36 to 54.6 h) (Fig. 3F). Significantly high specific cell growth rates and ε-PL formation rates at these two stages implied that the elicitors resulted in great metabolic change, which was an interesting phenomenon that was worthy to be physiologically disclosed. In Fig. 3G, a fed-batch culture was carried out to confirm the superior performance of elicitor addition in ε-PL production. The ε-PL titer reached a high level of 34.2 g/liter at 168 h, which is 1.34-fold higher than that of the control (25.5 g/liter). Therefore, the strategy of B. cinerea elicitor addition was an effective approach to improving ε-PL production.

Transcriptome analysis of the differences between cultures of EG and CK.

To understand the physiological changes after elicitor addition, transcriptome analysis and comparison were carried out between cultures of EG and CK. As shown in Fig. 4A, the elicitor addition resulted in 1,330 upregulated genes and 232 downregulated genes. In Gene Ontology (GO) clustering analysis (Fig. 4D), the most 20 significant differentially expressed genes concentrated on the aspects of stimulus response, signal transduction, transmembrane transport, ATPase activity, process of carbohydrate metabolic and cellular lipids, etc. In KEGG clustering analysis (Fig. 4C), the most 20 significant differentially expressed genes concentrated on the aspects of signal transduction, secondary metabolite biosynthesis, global metabolism in carbon, fatty acids, amino acids, and energy cofactors, etc. The distribution of upregulated genes focused on the aspects of quorum sensing, two-component system, microbial metabolism in diverse environment, biosynthesis of secondary metabolites, ABC transports, and metabolisms of amino/nucleotide sugar, pyruvate, fatty acid, propanoate, and cofactors (Fig. 4B).

FIG 4.

Transcriptome analysis of samples withdrawn from cultures of EG and CK. (A) Volcano diagrams showing the number of differentially expressed genes in EG versus CK comparison. (B) Significantly enriched KEGG pathways (padj < 0.05) in EG versus CK comparison. (C) Top 20 significantly enriched (padj < 0.05) KEGG pathways in EG versus CK comparison. (D) Top 20 significantly enriched Gene Ontology terms (padj < 0.05) in EG versus CK comparison. CK, without elicitor addition; EG, with addition of B. cinerea elicitors (36 g wet cells/liter).

These different expressed genes highlighted five categories of physiological functions, including cell propagation, lipid catabolism, carbon source utilization, secondary metabolite biosynthesis, and signal transduction. We further summarized the relevant genes, which are listed in Tables 2 to 5.

TABLE 2.

Comparation of gene transcription related to cell propagation between cultures of EG (elicitor group) and CK (control check)

Category	Gene IDa	Log2(EG/CK)	P value	Description
De novo synthesis of pyrimidine nucleotides	GM001714 (N1H47_08480)	4.02	5.07E−4	Aspartate carbamoyltransferase
	GM006389 (N1H47_32130)	3.65	1.04E−7	Carbamoyltransferase
	GM004844 (N1H47_24430)	4.90	1.16E−6	Uridine kinase
De novo synthesis of purine nucleotides	GM006445 (N1H47_32395)	3.00	3.65E−5	Phosphoribosylglycinamide synthetase
DNA replication	GM002040 (N1H47_10160)	2.02	2.89E−3	scpA segregation and condensation protein A
	GM002044 (N1H47_10180)	3.23	1.12E−5	DNA polymerase III, delta prime subunit L
	GM004255 (N1H47_21430)	5.52	3.95E−9	Helicase YprA
Cell division	GM006155 (N1H47_30945)	2.84	6.96E−2	Septum formation initiator
	GM004312 (ftsH)	3.17	1.02E−4	Cell division protease FtsH

^aThe former is the Gene ID in our laboratory, and the latter in bracket is the ID in NCBI database.

TABLE 3.

Comparation of gene transcription related to lipid catabolism between cultures of EG (elicitor group) and CK (control check)

Category	Gene IDa	Log2(EG/CK)	P value	Description
Phospholipid hydrolyzation	GM007399 (N1H47_37145)	7.24	4.78E−5	Lysophospholipase
	GM005353 (N1H47_26930)	4.70	5.52E−6	Phospholipase C
	GM001022 (N1H47_05045)	3.63	5.25E−3	Acetyl esterase
β-Oxidation	GM007289 (N1H47_36595)	3.47	7.86E−3	Long-chain fatty-acid CoA ligase
	GM006210 (N1H47_31220)	4.42	4.06E−5	3-Hydroxybutyryl-CoA dehydrogenase
	GM000786 (N1H47_03880)	3.90	1.08E−3	Coenzyme A transferase
	GM000355 (N1H47_01760)	4.99	6.40E−7	Acetyltransferase
	GM005511 (N1H47_27740)	4.77	4.01E−6	Acyl-CoA dehydrogenase
	GM000871 (N1H47_04290)	3.34	3.21E−5	Enoyl-CoA hydratase

^aThe former is the Gene ID in our laboratory, and the latter in bracket is the ID in NCBI database.

TABLE 4.

Comparation of gene transcription related to carbon source utilization between cultures of EG (elicitor group) and CK (control check)

Category	Gene IDa	Log2(EG/CK)	P value	Description
Carbon source utilization	GM002383 (N1H47_11895)	8.48	3.96E−8	Glucokinase
	GM007497 (xylB)	3.90	1.08E−3	Xylulose kinase

^aThe former is the Gene ID in our laboratory, and the latter in bracket is the ID in NCBI database.

TABLE 5.

Comparation of gene transcription related to secondary metabolite biosynthesis between cultures of EG (elicitor group) and CK (control check)

Category	Gene IDa	Log2(EG/CK)	P value	Description
Other secondary metabolite biosynthesis	GM007534 (N1H47_37810)	5.48	6.45E−9	B. thuringiensis toxin
	GM006361 (N1H47_31990)	3.78	2.14E−7	Polyketide synthase
	GM005225 (N1H47_26295)	4.96	8.61E−7	Phthiocerol/phenolphthiocerol synthesis polyketide synthase
Quorum sensing	GM005453 (lepB)	2.25	2.39E−4	lepB signal peptidase I
	GM007791 (N1H47_39060)	2.91	4.85E−3	Toxoflavin biosynthesis protein ToxC
	GM007792 (N1H47_39065)	2.84	8.72E−4	Toxoflavin biosynthesis protein ToxD
	GM005326 (N1H47_26795)	3.35	4.19E−4	Peptide/nickel transport system permease protein
Signal transduction for biosynthesis of ε-poly-l-lysine	GM003371 (N1H47_16885)	3.28	5.70E−5	Sigma factor HrdD
	GM001033 (N1H47_05100)	4.03	3.89E−6	LysR family transcriptional regulator
	GM007888 (N1H47_39545)	7.54	8.28E−6	LysR family transcriptional regulator
	GM006259 (N1H47_31460)	7.48	8.28E−6	LysR family transcriptional regulator

^aThe former is the Gene ID in our laboratory, and the latter in bracket is the ID in NCBI database.

Table 2 shows that the de novo synthesis of pyrimidine and purine nucleotides resulted in significantly higher gene transcription. Meanwhile, the transcription of genes related to DNA replication was upregulated. Higher gene transcriptions were also observed in cell division (N1H47_30945, septum formation initiator; ftsH, cell division potease FtsH). The above upregulated genes in cell propagation indicated that the B. cinerea elicitors could induce a rapid cell growth, which was in accordance with the results in Fig. 3F.

The elicitors also activated lipid catabolism. As shown in Table 3, upregulated genes were observed in terms of phospholipid hydrolyzation (N1H47_37145, N1H47_26930, and N1H47_05045) and β-oxidation (N1H47_36595, N1H47_31220, N1H47_03880, N1H47_01760, N1H47_27740, and N1H47_04290). It is worth noting that significant transcription improvement of genes for carbon source utilization was induced by the B. cinerea elicitors (Table 4). In EG, the gene coding for glucokinase (N1H47_11895) reached a higher level of transcription [Log₂^(EG/CK) = 8.48] than that in CK. The xylulose kinase gene (xylB) was also upregulated after elicitor addition. These improvements were beneficial to increasing carbon skeleton uptake and energy production.

The quorum-sensing system was closely involved with survival competition among different microbial cells/species, high biomass concentration, and antibiotic biosynthesis (32, 33). Interestingly, the elicitors activated the quorum-sensing system in S. albulus, resulting in significant transcriptional improvement of genes coding for lepB signal peptidase I (lepB), toxoflavin biosynthesis proteins ToxC and ToxD (N1H47_39060 and N1H47_39065), and peptide/nickel transport system permease protein (N1H47_26795). The elicitors also stimulated biosynthesis of several antibiotics, such as Bacillus thuringiensis toxin (N1H47_37810), putative polyketide (N1H47_31990), and phthiocerol/phenolphthiocerol (N1H47_26295), another metabolite produced through polyketide synthase (Table 5). In ε-PL production, HrdD, a key sigma factor for ε-PL biosynthesis (34), was also upregulated in EG (Table 5). This factor proved to be a positive regulator for ε-PL synthase, the direct enzyme for l-lysine assembly to produce ε-PL with ATP pyrophosphorolysis. LysR family transcriptional regulators (N1H47_05100, N1H47_39545, and N1H47_31460), the transcriptional regulators for l-lysine biosynthesis, were also improved, which implied that precursor l-lysine biosynthesis was activated in EG.

Transcriptional comparison of key genes for ε-PL biosynthesis in EG.

As shown in Fig. 5, quantitative real-time PCR (qRT-PCR) assay was employed to understand the transcriptional influences of elicitors on ε-PL biosynthesis metabolism. The elicitors significantly increased the transcriptional levels of genes coding for glucose-6-phosphate dehydrogenase (zwf), phosphoenolpyruvate carboxylase (N1H47_01765), aspartate kinase (N1H47_21215), and citrate synthase (N1H47_14640) at 30 h. An extraordinarily higher transcription level of N1H47_01765 (phosphoenolpyruvate carboxylase) was observed after elicitor addition, which was 204.6- and 92.8-fold higher than the levels in CK (without elicitor) at 30 and 42 h, respectively. In the EG, N1H47_16885 (sigma factor HrdD) showed transcription 7.5-fold higher than that in the CK, which led to 3.4-fold improvement of N1H47_34205 (ε-PL synthase) transcription at 30 h. Similar gene transcriptional improvements were also observed in N1H47_16135 (diaminopimelate decarboxylase) and its transcriptional regulator, N1H47_05100 (LysR family transcriptional regulator [LTTR]) (35), which had 7.8- and 7.2-fold higher levels than in the CK, individually. However, all transcriptional improvements of these genes exhibited a decreasing trend 12 h later (42 h), implying that the inducing effects were time efficient in ε-PL production.

FIG 5.

Effects of B. cinerea elicitor addition (24 h) to the transcription of key genes in ε-PL biosynthesis metabolism and relevant signal genes by quantitative real-time PCR (qRT-PCR) assays, where the gene transcription in CK was set as 1 (dotted lines), and those genes relative transcription in EG was calculated based on the levels of CK. CK, without elicitor addition; EG, with addition of B. cinerea elicitors (36 g wet cells/liter). LTTR, LysR family transcriptional regulator; NS, not significant; TCA, tricarboxylic acid; ASK, aspartate kinase; DAPD, diaminopimelate decarboxylase.

Profiles of intracellular key enzymes, amino acids, and energy cofactors for ε-PL biosynthesis in EG.

As shown in Fig. 6, even though key enzyme activities decreased with the time extension (in accordance with the transcriptional performances) (Fig. 5), the elicitors significantly increased the activities of phosphoenolpyruvate carboxylase, aspartate kinase, citrate synthase, and ε-PL synthase after 36 h. ε-PL synthase, the direct enzyme for ε-PL biosynthesis, showed significantly higher activity after elicitor addition, which reached high levels at 36 and 42 h. In the EG, the aspartate kinase activity maintained higher levels at 36 to 56 h than in CK, implying that the elicitors could activate the l-lysine biosynthesis for a long time. The activities of phosphoenolpyruvate carboxylase in the EG reached extremely higher levels at 36 and 48 h (Fig. 6), which were 2.4- and 4.6-fold higher than those in CK. Similar improvement of phosphoenolpyruvate carboxylase was also reported in natamycin production after adding Penicillium chrysogenum elicitors (24). The intracellular pools of l-aspartic acid and l-glutamic acid were correspondingly enlarged in EG, which supplied sufficient precursors and a source of NH₂ for l-lysine synthesis and finally led to a bigger intracellular pool of l-lysine. In addition, the activity of citrate synthase, the key enzyme for the tricarboxylic acid (TCA) cycle, increased significantly during 36 to 48 h in the EG, which accelerated the NADH formation for subsequent ATP biosynthesis and ε-PL production. Higher levels of NADH, ATP, and electron transport system (ETS) activity were observed in EG at 36 to 54 h, which was in accordance with the increased citrate synthase activity. These improved key enzyme activities, precursor amino acids, and energy cofactors confirmed the superior inducing effects of the B. cinerea elicitors on ε-PL biosynthesis.

FIG 6.

Effects of B. cinerea elicitor addition (24 h) to the intracellular activities of key enzymes, energy cofactors, and key amino acids in ε-PL biosynthesis metabolism.

Metabolic profile of intermediates in pathways for ε-PL biosynthesis.

Nontargeted metabolomics analysis of intracellular metabolites after elicitor addition was carried out by an ultraperformance liquid chromatography-electrospray ionization-mass spectrometry (UPLC-ESI-MS) system. A total metabolite of 8 in the Embden-Meyerhof pathway (EMP), 9 in TCA cycle, 3 in the pentose phosphate pathway (PPP), 5 in the diaminopimelic acid pathway (DAP), 7 in cofactors, and 19 in amino acids was identified in S. albulus. The metabolome was analyzed using principal-component analysis (PCA) and orthogonal projections to latent structures discriminant analysis (OPLS-DA). The score plot (Fig. 7A) shows that the EG was distinctly separated from CK. The two principal components extracted by the model explained 83.2% of the total variance. The OPLS-DA score scatterplots indicated that EG and CK were clearly separated (Fig. 7B). As shown in Table 6 and Fig. 7C, the metabolite pools in central carbon metabolism were significantly enlarged after elicitor addition. The relative content of metabolites in the EMP, TCA cycle, and DAP were significantly higher than that in CK. The levels of 6 of 8 EMP metabolites, 6 of 9 TCA cycle compounds, 4 of 5 DAP intermediates, and 5 of 7 cofactors in EG were higher than those in CK. In contrast, the levels of 14 of 19 amino acids were lower than those in CK. Notably, compared to those in CK, the EG group showed significantly higher relative contents of aspartic acid and lysine but lower contents of threonine, isoleucine, and methionine. These results suggested that the elicitors resulted in global enhancements of central carbon metabolism, energy biosynthesis metabolism, and metabolic changes of amino acids biosynthesis, which was beneficial to ε-PL production.

FIG 7.

Metabolome analysis of samples withdrawn from cultures of EG and CK. (A) Principal-component analysis (PCA) score plot of CK and EG. (B) Orthogonal projections to latent structures discriminant analysis (OPLS-DA) score plot of EG versus CK comparison group. (C) Heat map for the Z score of important intermediate metabolites in ε-PL biosynthesis metabolism. The relative content of significantly difference metabolites was calculated according to the OPLS-DA model (P < 0.05).

TABLE 6.

Relative fold change of compounds in S. albulus after elicitor addition according to UPLC-ESI-MS analysis^a

ID	tR (min)	Measured mass	MS/MS irons	Formula	Name	Fold change
						CK/CK	EG/CK
					Embden-Meyerhof pathway
1	0.573	169.9976	55, 73, 100, 153	C3H7O6P	Glyceraldehyde 3-phosphate	1.00	1.30**
2	1.105	338.9886	79, 97, 159	C6H14O12P2	Fructose 1,6-bisphosphate	1.00	1.38**
3	1.153	88.0107	43, 87	C3H4O3	Pyruvic acid	1.00	1.32**
4	1.245	259.0223	59, 97, 119	C6H13O9P	Fructose 6-phosphate	1.00	0.99ns
5	1.428	261.0369	117, 127, 128, 163	C6H13O9P	Glucose 6-phosphate	1.00	1.53***
6	1.505	186.0448	79, 97	C3H7O7P	2-Phosphoglyceric acid	1.00	2.19***
7	5.482	171.1492	73, 153, 171	C3H7O6P	Dihydroxyacetone phosphate	1.00	0.62***
8	10.482	181.0089	145, 163, 181	C6H12O6	Glucose	1.00	1.34**
					Tricarboxylic acid cycle
9	0.565	116.1073	45, 53, 61, 99	C4H4O4	Fumaric acid	1.00	3.01***
10	1.117	146.0298	55, 101, 111, 129	C5H6O5	Oxoglutaric acid	1.00	1.04ns
11	1.153	191.0894	85, 87, 111	C6H8O7	Citric acid	1.00	1.41***
12	1.245	133.0129	71, 115	C4H6O5	Malic acid	1.00	1.20*
13	1.268	102.0341	73, 101	C4H6O4	Succinic acid	1.00	1.46***
14	2.905	175.0231	115, 129, 157, 175	C6H6O6	cis-Aconitic acid	1.00	21.93***
15	5.190	809.1254	79, 134	C23H38N7O17P3S	Acetyl-CoA	1.00	6.09***
16	8.755	112.9846	44, 87	C4H4O5	Oxalacetic acid	1.00	0.30***
17	10.353	191.0197	111, 173	C6H8O7	Isocitric acid	1.00	0.46***
					Pentose phosphate pathway
18	1.165	275.018	78, 96	C6H13O10P	6-Phosphogluconic acid	1.00	0.80*
19	1.707	177.0397	57, 59, 71	C6H10O6	Gluconolactone	1.00	1.94***
20	3.228	231.0281	125, 143, 185	C5H11O8P	Ribulose 5-phosphate	1.00	0.67***
					Diaminopimelic acid pathway
21	0.580	117.0826	55, 72	C4H7NO3	Aspartate-semialdehyde	1.00	2.73***
22	1.117	170.0412		C7H9NO5	(2S,4S)-4-Hydroxy-2,3,4,5-tetrahydrodipicolinate	1.00	0.96ns
23	1.305	191.1027	82, 128	C7H14N2O4	Diaminopimelic acid	1.00	1.31**
24	1.365	134.0447	88, 116	C4H7NO4	Aspartic acid	1.00	1.74***
25	2.812	172.0605	126, 152, 170	C7H9NO4	Tetrahydrodipicolinate	1.00	1.79***
					Cofactors	1.00
26	1.543	427.0318	136	C10H15N5O10P2	ADP	1.00	4.29***
27	1.873	744.0901	604, 622	C21H29N7O17P3	NADP	1.00	0.80***
28	4.713	664.1266	134, 448	C21H29N7O14P2	NADH	1.00	1.80***
29	4.730	348.0694	136	C10H14N5O7P	AMP	1.00	0.60***
30	5.615	664.1132	524, 542	C21H28N7O14P2	NAD	1.00	1.49***
31	8.730	507.2744		C10H16N5O13P3	ATP	1.00	2.03***
32	13.938	832.6128	744	C21H30N7O17P3	NADPH	1.00	7.10***
					Amino acids
33	1.195	147.1122	129, 147	C6H14N2O2	Lysine	1.00	1.57***
34	1.293	156.076	60, 116, 156	C6H9N3O2	Histidine	1.00	1.11*
35	1.305	175.1188	70, 116, 175	C6H14N4O2	Arginine	1.00	0.46***
36	1.362	147.0749	130, 147	C5H10N2O3	Glutamine	1.00	1.32**
37	1.385	120.0657	56, 84	C4H9NO3	Threonine	1.00	0.73**
38	1.402	238.9315	74, 120	C6H12N2O4S2	Cystine	1.00	0.46***
39	1.428	119.0495	86, 87	C3H7NO3	Serine	1.00	0.62***
40	1.520	116.0711	70, 116	C5H9NO2	Proline	1.00	0.49***
41	1.882	147.1132	102, 130	C5H9NO4	Glutamic acid	1.00	0.53***
42	2.218	150.0585	104, 133	C5H11NO2S	Methionine	1.00	0.37***
43	3.362	130.0869	130	C6H13NO2	Isoleucine	1.00	0.39***
44	3.420	132.1013	86, 115	C6H13NO2	Leucine	1.00	0.47***
45	5.357	118.0867	55, 72, 83	C5H11NO2	Valine	1.00	0.42***
46	5.550	166.086	73, 120	C9H11NO2	Phenylalanine	1.00	0.41***
47	5.835	182.0814	119, 163, 180	C9H11NO3	Tyrosine	1.00	0.60***
48	6.588	205.0971	188	C11H12N2O2	Tryptophan	1.00	0.59***
49	11.792	90.0553	44, 90	C3H7NO2	Alanine	1.00	7.82***
50	12.363	133.1013	70, 114	C4H8N2O3	Asparagine	1.00	0.31***

^aUPLC-ESI-MS, ultraperformance liquid chromatography-electrospray ionization-mass spectrometry. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant.

DISCUSSION

Streptomyces species usually produce multiple antifungal metabolites, which have been widely applied to plant fungicidal disease control (36–40). However, whether the biosynthesis of antifungal metabolites could be adversely influenced by fungus or bacteria is not clear. In this study, an interesting phenomenon was observed that ε-PL overproduction could be induced by elicitors from microorganism (B. cinerea) in its inhibition spectrum. Elicitors from fungus showed positive influences on ε-PL production, while those from bacteria performed negatively. It is known that ε-PL performed strong inhibition of bacteria and weak inhibition of molds (Table 1), but speculation was raised that the ε-PL-producing strain tend to produce more antimicrobial agent (ε-PL) to fight against molds, which might happen in complex natural habitats. Interestingly, this “elicitors’ call” not only resulted in ε-PL overproduction but also promoted the transcription of genes for biosynthesis of other antibiotics (Table 5), such as B. thuringiensis toxin (N1H47_37810), and metabolites produced through polyketide synthase (N1H47_31990 and N1H47_26295); hence, it is a signal for global antibiotics overproduction after receiving “threatening message” from molds. The same activated quorum-sensing system (Table 5) was speculated as one of the reasons for global antibiotics gene upregulation as a crucial role in stimulating antibiotic synthesis (32, 33).

Primary isolation indicated that the threatening message from B. cinerea biomass was more likely to be a small polar molecule rather than a macromolecule or nonpolar structure. The elicitors resulted in aggregate colonies in petri plate culture, thicker and stronger mycelia, and aggregate and bigger mycelia pellets in submerged cultivation (Fig. 2C and D), which might be more like stress-defensive responses. Similar mycelial morphology was reported in S. rimosus M527 after the addition of living cells of Fusarium oxysporum f. sp. cucumerinum, which showed mature, plump, smoother surfaces mycelia in scanning electron microscopy (SEM) observation (25). This morphology was also observed after P. chrysogenum elicitor addition, which resulted in improved natamycin production (24). The CTC-staining image also revealed a respiration activity improvement in EG. These differences in macroscopic morphology and microcosmic morphology implied that the B. cinerea elicitors resulted in significant physiological changes in S. albulus cells, which might be self-defensive responses to external biological signals.

Therefore, this raised an interesting question regarding how the S. albulus responded to the elicitors’ call from B. cinerea. To understand the underlying physiological basis, transcriptome analysis, qRT-PCR, determination of key enzyme activities, intracellular pools of energy and amino acids, and metabolome analysis were performed. All these results after elicitor addition were of mutual consistency and pointed to these facts as follows: (i) the elicitors could result in higher levels of carbon source utilization, including glucose transmembrane transport, glucose phosphorylation, and xylose phosphorylation, which guaranteed sufficient provision of carbon skeleton and energy; (ii) the biomass generation was temporally activated in terms of 5-phosphate ribose synthesis, nucleotides synthesis, DNA replication, and cell division, which was beneficial to cell growth and biomass accumulation; (iii) the energy metabolism was accelerated at aspects of gene’s transcription and activity enhancements of citrate synthase, enlargements of NADH and ATP pools, as well as ETS activation and lipid β-oxidation, which produced sufficient ATP for ε-PL biosynthesis through l-lysine assembly; (iv) metabolic changes after elicitor addition led to an enhanced biosynthesis of l-lysine through inhibition of branched-chain amino acid metabolism, pool expansion of l-aspartate and l-glutamate, improvements of gene transcription, and activity of phosphoenolpyruvate carboxylase and aspartate kinase, as well as transcriptional increase of N1H47_16135 (diaminopimelate decarboxylase) and its regulator, N1H47_05100 (LTTR); and (v) direct ε-PL biosynthesis was improved based on higher gene transcriptional levels of HrdD and ε-PL synthase, higher ε-PL synthase activity, and greater pools of l-lysine and ATP.

Part of the above results was inconsistent with the physiological changes in S. rimosus after F. oxysporum f. sp. cucumerinum elicitor addition and S. natalensis after P. chrysogenum elicitor addition, including higher activity of phosphoenolpyruvate carboxylase (24), mycelia of plump and smoother surfaces (24, 25), improvement of gene transcription in secondary metabolite biosynthesis, and branch-chained amino acid degradation (26). However, there were two different points: (i) the growth of S. natalensis was found to be inhibited after P. chrysogenum elicitor addition (24), while we found that the B. cinerea elicitors first increased and then inhibited the S. albulus growth; and (ii) the citrate synthase in S. natalensis was relatively inhibited after elicitor addition, while both enzyme activity and gene transcription were improved in this study. It was implied that microbial elicitors performed similar mechanisms to influence Streptomyces species, and adding a dose might be a crucial element for physiological regulation.

Based on this new discovery, a high-efficiency strategy for ε-PL production was proposed by exogenous addition of B. cinerea elicitors at 24 h, which resulted in 34.2 g/liter ε-PL production after 168-h cultivation (Fig. 3G), a 33.6% improvement of ε-PL titer over that without elicitors. The elicitor addition led to rapid cell growth at the first stage (24 to 36 h) (Fig. 3D), while gradual decline of cell growth rate sharply promoted the specific ε-PL formation rate (Fig. 3E), which might be due to target transition of ATP and l-lysine from biomass building to ε-PL synthesis. The successful application of B. cinerea elicitors in ε-PL production could also be regarded as a potentially harmless disposal for decayed fruits infected by B. cinerea in the agricultural industry, indicating that these fruits could be recycled as a source for microbial elicitors to enhance ε-PL production.

In fact, interactions among microorganisms and the applications on secondary metabolite production have been reported. Serrano-Carreón et al. found that the productivity of 6-pentyl-α-pyrone can be significantly improved in Trichoderma harzianum by exogenous addition of both viable and nonviable mycelium of Rhizoctonia solani (41). Likewise, the biosynthesis of ganoderic acid and polysaccharides by Ganoderma lucidum can be significantly improved by protein elicitors isolated from Tuber melanosporum but inhibited by its polysaccharide’s elicitors (42, 43). Similar improvement is also observed in rimocidin production by adding broth and living cells of S. cerevisiae and F. oxysporum f. sp. cucumerinum, which is due to transcriptions increase of 10 genes involved in rimocidin biosynthesis (25). In addition, some fungi (Aspergillus niger and P. chrysogenum) can generate elicitors to promote the biosynthesis of natamycin by S. natalensis HW-2 (23, 24, 26), which might be attributed to an increase of precursor supply through the regulation of branch-chain amino acid metabolism by elicitors.

In this study, after systematic physiological analysis, we found that the S. albulus tended to globally activate its central carbon metabolism to rapidly produce antimicrobial metabolites and made its mycelia stronger to compete with another microorganism, even though there was only a “call” from B. cinerea (a widespread plant-pathogenic mold) rather than actual threats brought from real cells. Except the acidic signal, the microbial elicitors were reckoned as another signal for ε-PL overproduction. The results depicted a physiological map of ε-PL overproduction after elicitor addition, and an efficient strategy of exogenous addition of the elicitors was successfully applied in industrial ε-PL production, which was of theoretical and practical significance. Further research is underway for signal molecules isolation, structural identification, and signal-cell physiology connection, so as to answer a new question: what does the elicitors’ call look like (molecular structure) and how does it function? Relevant results are expected to provide guidance on metabolic engineering (44) in future revolving around the key signal genes.

MATERIALS AND METHODS

Microorganism and culture media.

The ε-PL-producing strain of S. albulus IFO 14147 (CICC 11022) and elicitors producing strains of B. subtilis (CICC 10002), E. coli (CICC 10389), and S. cerevisiae (CICC 1302) were derived from the China Center of Industrial Culture Collection (CICC). B. cinerea CGMCC 3.3790 was purchased from China General Microbiological Culture Collection Center (CGMCC).

An agar slant medium was prepared for the activation of B. subtilis, E. coli, and S. cerevisiae, as well as spore generation of B. cinerea and S. albulus IFO 14147. It was also used for the susceptibility evaluation of above four microorganisms to ε-PL. It contained 10 g/liter glucose, 5 g/liter yeast extract, 5 g/liter peptone, and 20 g/liter agar with an initial pH of 7.5 by adding 2 M NaOH.

Medium 3G (M3G) with an initial pH of 6.8 was used for both the seed preculture and ε-PL production in this study. It was comprised of 60 g/liter glucose, 5 g/liter yeast extract, 10 g/liter (NH₄)₂SO₄, 1.36 g/liter KH₂PO₄, 0.8 g/liter K₂HPO₄, 0.5 g/liter MgSO₄·7H₂O, 0.04 g/liter ZnSO₄·7H₂O, and 0.03 g/liter FeSO₄·7H₂O.

LB medium was used for the cultivations of B. subtilis and E. coli, which consist of 10 g/liter tryptone, 5 g/liter yeast extract, and 10 g/liter NaCl with an initial pH of 7.4 by adding 2 M NaOH. YPD medium was applied to the cultures of S. cerevisiae and B. cinerea. It contained 10 g/liter yeast extract, 20 g/liter peptone, and 20 g/liter glucose with an initial pH of 6.5 by adding 2 M NaOH. The above media were autoclaved at 121°C for 20 min, and the glucose was separately sterilized to prevent Maillard reaction.

Susceptibility evaluation of four microorganisms to ε-PL.

The suspension (4 × 10⁵) of B. subtilis, E. coli, S. cerevisiae, and spores of B. cinerea were prepared by washing, collecting, and proper dilution of precultured cells on petri plate. The suspension of above the microorganisms was inoculated on agar slant medium with ε-PL addition in a concentration range from 0.1 to 300 μg/mL. The number of colonies on agar slant medium was counted after 4 days of incubation at 30°C.

Influences of elicitors from different microorganisms on ε-PL production.

B. subtilis, E. coli, and S. cerevisiae were inoculated in the agar medium and incubated at 30°C for 2 days for strain activation. For spore preparation, S. albulus IFO 14147 was inoculated on agar slant medium and incubated at 30°C for 9 days until the spores emerged. The spores of B. cinerea were obtained after incubation at 23°C for 5 days in agar medium.

In 250-mL flasks, the cultures of B. subtilis and E. coli were carried out in a rotary shaker at 200 rpm and 30°C for 15 h, while the S. cerevisiae was cultivated at 200 rpm and 30°C for 24 h. For the culture of B. cinerea, four loops of spores (6 × 10³ spores) were inoculated in 60 mL YPD medium in 500-mL flasks and then placed into in a rotary shaker at 160 rpm and 23°C for 4 days. After these cultivations, the broths were centrifuged at 5,000 × g for 20 min. The supernatant and sediment were collected individually. The supernatant was further centrifuged at 10,000 × g for 20 min and sterilized at 115°C for 15 min. The cell pellets were washed twice by deionized water and treated by an ultrasonic crusher for biomass disruption with ice bath. Then, the extract was centrifuged at 10,000 × g for 20 min and sterilized at 115°C for 15 min to obtain the biomass extract.

For the seed preculture for ε-PL production, one loop of spores (5 × 10⁵ spores) was inoculated in 60 mL M3G in 500-mL flasks and incubated in a rotary shaker at 200 rpm and 30°C for 24 h. The formal ε-PL production was initiated by 8% seed broth and performed in 250-mL flasks with 30 mL M3G for 2 days at 200 rpm and 30°C. The elicitor extracts from different microorganisms (B. subtilis, E. coli, S. cerevisiae, and B. cinerea) were added in the medium with broth filtrate ratios of 3, 6, 9, 15, 21, and 27% and final biomass extract concentrations of 6, 12, 18, 24, 30, and 36 g wet cell/liter at 24 h, when the pH had naturally dropped to 4.0. Extra sterilized water was added into the medium to make the adding volume consistent. The final ε-PL titer and dried cell weight (DCW) in broth were measured after fermentation.

Primary isolation of the elicitors from B. cinerea biomass.

The elicitors were isolated from the biomass of B. cinerea after 4 days of cultivation. The biomass was obtained by filtration with four layers of gauze and washed four times by deionized water. The biomass was subsequently suspended in water, grinded with quartz sand, and centrifuged at 5,000 × g for 20 min to obtain the supernatant of biomass extract. The extract was subjected to ethanol precipitation. Briefly, the extract solution was mixed with cool ethanol with a final concentration of 75% (vol/vol). After standing for 12 h, the mixture was centrifuged at 5,000 × g for 20 min to obtain the supernatant (fraction S_n) and sediment (fraction S_d). The supernatant (fraction S_n) was evaporated to remove the ethanol. The remaining solution was divided into four parts with equal volumes and extracted by 3-fold volume of ethyl acetate, chloroform, butyl alcohol, and petroleum ether, individually. The organic phase was collected with a separating funnel and treated with a rotary evaporator to remove the organic reagent. The extracts after ethyl acetate, chloroform, butyl alcohol, and petroleum ether treatment were named fractions E, C, B, and P, respectively. The inducing effects of these fractions were evaluated in ε-PL production by exogenous addition at 24 h with pH control by adding sodium citrate buffer (pH 4.0, 10 g/liter final concentration) in 250-mL flasks.

Optimization of key parameters for ε-PL production with elicitor addition.

The biomass extract of B. cinerea was prepared by ethanol-butyl alcohol extraction (mentioned above) and sterilized by an autoclave at 121°C for 20 min. The adding time of elicitors in ε-PL production was studied, including 0, 6, 12, 18, 24, 30, 36, and 42 h, while the culture age of elicitors’ producing strain was investigated, including 1, 2, 3, 4, and 5 days. The optimal concentration of biomass extract was varied from 0 to 84 g wet cells/liter. When one of the parameters was evaluated, the other factors were maintained as 24 h of elicitors adding time, 4 days of B. cinerea culture age and 30 g wet cells/liter of biomass extract concentration. To maintain a constant broth pH, sterilized buffer of sodium citrate (pH 4.0) was added to the medium at 24 h with a final concentration of 10 g/liter. The ε-PL titer was measured after batch culture of 48 h.

ε-PL production with exogenous addition of B. cinerea elicitors in 5 liters of fermenter.

The performances of B. cinerea elicitors on ε-PL production were evaluated through batch and fed-batch cultures in a 5-liter jar fermenter (Baoxing Corp., Shanghai, China) with a 3-liter working volume. Profiles of batch/fed-batch cultures without elicitor addition (control check [CK]) were compared with those with elicitor addition (experimental group [EG]) at 24 h. The fermenter was equipped with two turbine agitators, as well as auto-controlling systems of temperature, dissolved oxygen (DO), and broth pH value. The cultures were initialized after inoculation of seed broth with a ratio of 8% and then performed at 30°C. Agitation of 200 to 800 rpm and aeration of 1.0 vvm were employed to maintain the DO at 30% during the whole culture period. When the broth pH naturally dropped to 4.0, 12.5% (wt/vol) NH₃·H₂O was automatically added into the broth to retain a constant pH of 4.0. The batch cultures were terminated when the glucose was exhausted. For the fed-batch cultures, when the residual glucose decreased below 10 g/liter, 70% of sterilized glucose was automatically added in to control a concentration of 5 to 15 g/liter. The ammonia nitrogen (NH₄⁺-N) concentration was also controlled at about 0.5 g/liter by automatic addition of sterilized (NH₄)₂SO₄ solution (40%, wt/vol). Broth samples were withdrawn from the bioreactor at specific times for assays of offline parameters, including ε-PL titer, DCW, and residual glucose and NH₄⁺-N.

Effects of elicitors on the morphology of S. albulus IFO 14147.

The influence of elicitors on S. albulus morphology was evaluated through observation and comparison of the petri plates colony (macroscopical) morphology and mycelial image in submerged culture by SEM (microscopic). For the observation of colony morphology, petri plates with agar slant medium were prepared by adding elicitors of fraction B with a final concentration of 36 g wet cells/liter. The spore suspension of S. albulus IFO 14147 was diluted and added onto these petri plates. The colony morphology with elicitor addition (EG) was observed in a benchtop at 2 days and 5 days and compared with that without elicitors (CK). For the observation of mycelial morphology, the samples were withdrawn at 45 h from the submerged culture with elicitor addition at 24 h. The samples were treated with glutaraldehyde, phosphate buffer, and ethanol and observed using scanning electron microscopy (Sigma 300, ZEISS, Germany). A cell viability assay was performed by using a CTC rapid staining kit (Dojindo Laboratories, Japan) as described in our previous study (45), and the image of mycelial morphology after CTC staining was captured by a laser scanning confocal microscope (LEXT OLS4500, Olympus, Japan).

RNA extraction and transcriptome analysis.

In Fig. 3, the main differences of cultures CK and EG mainly appeared between 36 and 54 h. Therefore, the mycelia pellets at 36 h were more representative to be used as the samples for RNA extraction and the transcriptome sequencing (RNA-seq) procedure. The cell samples were taken from three independent cultures at 36 h, and they were merged. The broth was immediately centrifuged at 7,000 × g for 1 min, and the precipitates were washed once before frozen by liquid nitrogen and stored at −80°C overnight for RNA isolation (preparation of total RNA). The procedures of RNA isolation and RNA-seq were carried out as described previously (17), and relevant genes were mapped to the genome of S. albulus IFO14147 (CP104098.1). Only more than 2-fold differential transcription [Log₂^(EG/CK) > 1; P value < 0.05] was considered to be significant, and these genes were selected. The transcriptions of genes in cell propagation, lipid catabolism, carbon source utilization, and secondary metabolite biosynthesis were studied to obtain a deeper insight into the fermentation enhancements with elicitor addition (EG).

Transcriptional levels of genes related to ε-PL biosynthesis.

The qRT-PCR was employed to evaluate the key genes related to the ε-PL biosynthesis. The samples were taken from the fermenter at 30 and 42 h in batch cultures of CK and EG. The extraction of total RNA from S. albulus IFO 14147 was carried out as described previously (17). The transcriptional levels of genes (zwf, N1H47_01765, N1H47_21215, N1H47_14640, N1H47_34205, N1H47_16885, N1H47_05100, and N1H47_16135) were measured with a real-time fluorescent quantitative PCR (ABI StepOnePlus, Applied Biosystems, USA) with SG Fast qPCR Master Mix (High Rox) (Bio Basic Inc., Toronto, Canada). The primer pairs sequences are shown in Table 7.

TABLE 7.

Primer pairs sequences for quantitative real-time PCR (qRT-PCR) assay

Target genea	Primer name	Primer sequence (5′ to 3′)
16S rRNA	16S rRNA forward	GCACAAGCAGCGGAGCAT
	16S rRNA reverse	CCCAACATCTCACGACACGA
GM002188 (zwf)	GM002188 forward	CGTCCACGAGGTCTTCCC
	GM002188 reverse	GGAGGAGGTGGTTCTGGATG
GM000356 (N1H47_01765)	GM000356 forward	CTAACTCAGCAGGCACTGTGTC
	GM000356 reverse	GACGTGGTAGGCGTTCTCC
GM004214 (N1H47_21215)	GM004214 forward	GTCCGCTCCTCCTTCTCG
	GM004214 reverse	CCGACGACCGTGACCTTC
GM002933 (N1H47_14640)	GM002933 forward	GCCTTCAAGCAGGACATCAC
	GM002933 reverse	GGTTGTGGCTGTCCTGGTAG
GM003371 (N1H47_16885)	GM003371 forward	GAGCTGTCCCAGACCATCG
	GM003371 reverse	GTTGGACCGGATGAAGACG
GM006807 (N1H47_34205)	GM006807 forward	CCCTGTGGTCGTCGTTCG
	GM006807 reverse	GAAGAGGTGGGTCTGCAGGA
GM001033 (N1H47_05100)	GM001033 forward	GACCTGCTGCTGGTGGGA
	GM001033 reverse	GTTCGACGTCCACCCAGC
GM003225 (N1H47_16135)	GM003225 forward	GGTGCCGCAGGACATCG
	GM003225 reverse	CAGCGAGTTGTAGGCGAAGTG

^aThe former is the Gene ID in our laboratory, and the latter in bracket is the ID in NCBI database.

Activity assays of key enzymes and electron transport system in ε-PL biosynthesis.

For the measurement of key enzymes and ETS, the broth samples were withdrawn from the fermenter at 36, 42, 48, and 54 h. The cell extracts preparation and activity assays of phosphoenolpyruvate carboxylase, citrate synthase, aspartate kinase, and ε-PL synthase were measured by following the procedures described previously (46). The activity of ETS was measured as described by Herrera et al. (47). The method was usually used to measure the ETS activity in hydroplankton (48, 49), while it was proved to be applicable to microorganisms. For ETS activity determination, a biochemical reaction mixture was prepared that contained 50 μL biomass extracts, 60 μL 100 mM Tris-HCl buffer (pH 7.5), 15 μL 0.835 mM NADH, 15 μL 0.24 mM NADPH, 30 μL 0.133 M sodium succinate, 30 μL 1% Triton, and 30 μL 4 mM iodonitrotetrazolium. The absorbance of 490 nm was recorded for 3 min, and the ETS activity (φ, μL·O₂/h/g protein) was calculated as follows:

$$\phi = \frac{60\times \Delta A_{490}\times V_{\mathrm{Total}}}{1.42\times \Delta t\times V_{\mathrm{Cell}\mathrm{extract}}\times H}$$ (1)

where ΔA₄₉₀ is the 490-nm absorbance, V_Total is the value of the total reaction volume (μL), Δt is the reaction time (min), V_{Cell extract} is the cell extract volume (μL), and H is the reaction mixture height (mm). All activity assays of the key enzymes and ETS were carried out at 30°C.

Assay of intracellular energy cofactors and amino acids.

The broth samples were withdrawn from the fermenter at 36, 42, 48, and 54 h and centrifuged at 7,000 × g for 1 min, and the cell precipitates were washed twice with 0.9% NaCl. For the determination of intracellular energy cofactors (ATP, NADH), the cell precipitates were treated with HClO₄ extraction, ultrasonic disruption, and centrifugation, and the supernatant was used for determination by high-pressure liquid chromatography (HPLC) as described previously (46). For the measurement of intracellular amino acids, the precipitates were resuspended in 10% trichloroacetic acid at 37°C for 10 min. These samples were boiled for 30 min and centrifuged at 12,000 × g for 20 min to remove cell debris. l-Asparaginic acid, l-glutamic acid, and l-lysine in the supernatant were finally analyzed by HPLC as depicted by Fountoulakis et al. (50).

Metabolome analysis and data processing.

The UPLC-ESI-MS analysis was carried out as follows: The cell samples were taken from three independent cultures at 42 h in CK and EG groups, and they were merged independently. The mycelia were washed twice with distilled water at 4°C and frozen by liquid nitrogen. The cell samples were thawed at room temperature, resuspended with solution of acetonitrile (ACN):methanol:H₂O mixed solution (2:2:1, vol/vol/vol) (stored at 4°C), immersed in liquid nitrogen for rapid freezing for 5 min, thawed at room temperature, and installed into the tissue grinder for 2 min of grinding at 55 Hz. These procedures were repeated three times before centrifugation for 10 min at 12,000 rpm and 4°C. The supernatant was concentrated and dried. The samples were redissolved by solution of acetonitrile:2-amino-3-(2-chloro-phenyl)-propionic acid (4 ppm) with 0.1% formic acid (1:9, vol/vol) (stored at 4°C), filtered by 0.22-μm membrane, and transferred into the detection bottle for LC-MS detection.

The LC analysis was performed on a Vanquish UHPLC System (Thermo Fisher Scientific, USA). Chromatography was carried out with an ACQUITY UPLC HSS T3 (150 × 2.1 mm, 1.8 μm) (Waters, Milford, MA, USA). The column was maintained at 40°C. The flow rate and injection volume were set at 0.25 mL/min and 2 μL, respectively. For LC-ESI (+)-MS analysis, the mobile phases consisted of 0.1% formic acid in acetonitrile (vol/vol) and 0.1% formic acid in water (vol/vol). Separation was conducted under the following gradient: 0 to 1 min, 2% C; 1 to 9 min, 2 to 50% C; 9 to 12 min, 50 to 98% C; 12 to 13.5 min, 98% C; 13.5 to 14 min, 98 to 2% C; and 14 to 20 min, 2% C. For LC-ESI(−)-MS analysis, the analytes were carried out with acetonitrile and ammonium formate (5 mM). Separation was conducted under the following gradient: 0 to 1 min, 2% A; 1 to 9 min, 2 to 50% A; 9 to 12 min, 50 to 98% A; 12 to 13.5 min, 98% A; 13.5 to 14 min, 98 to 2% A; and 14 to 17 min, 2% A (51). Mass spectrometric detection of metabolites was performed on Q Exactive (Thermo Fisher Scientific, USA) with ESI ion source. Simultaneous MS1 and MS/MS (full MS-ddMS2 mode, data-dependent MS/MS) acquisition was used. The parameters were as follows: sheath gas pressure, 30 arb; auxilliary gas flow, 10 arb; spray voltage, 3.50 kV and −2.50 kV for ESI(+) and ESI(−), respectively; capillary temperature, 325°C; MS1 range, m/z 81–1,000; MS1 resolving power, 70,000 full width at half maxima (FWHM); number of data-dependent scans per cycle, 10; MS/MS resolving power, 17,500 FWHM; normalized collision energy, 30%; and dynamic exclusion time, automatic (52). The metabolites profiles were analyzed by Mass Profiler Professional 13.0 software (Agilent, US). The metabolic compounds were identified according to the retention time and mass spectrum by the self-built database of our laboratory. The PCA and OPLS-DA models were established to analyze the differences of metabolites between samples (SIMCA 14.1, Umetrics AB, Umea, Sweden). The Z score was calculated to visualize the metabolic profiling.

Analytical methods.

The broth samples were treated by centrifugation at 5,000 × g for 10 min. The supernatant was used for the concentration measurements of ε-PL titer, residual glucose, and NH₄⁺-N, while the sediment was used for DCW determination. The measurements of ε-PL titer, residual glucose, dried cell weight, and NH₄⁺-N were carried out as described in our previous study (32).

Calculations.

The average values of specific cell growth rate, specific ε-PL formation rate, and specific glucose consumption rate were calculated as follows:

$$\mathrm{Average}\mathrm{specific}\mathrm{cell}\mathrm{growth}\mathrm{rate}=\hspace{0.17em}\frac{1}{\frac{1}{2}[c{\left(X\right)}_{t_2}+c{\left(X\right)}_{t_1}]}\times \frac{c{(X)}_{t_2}-c{(X)}_{t_1}}{t_2-t_1}$$ (2)

$$\mathrm{Average}\mathrm{specific}\epsilon \mathrm{-PL}\mathrm{formation}\mathrm{rate}=\hspace{0.17em}\frac{1}{\frac{1}{2}[c{\left(X\right)}_{t_2}+c{\left(X\right)}_{t_1}]}\times \frac{c{(P)}_{t_2}-c{(P)}_{t_1}}{t_2-t_1}$$ (3)

$$\mathrm{Average}\mathrm{specific}\mathrm{glucose}\mathrm{consumption}\mathrm{rate}=\frac{1}{\frac{1}{2}[c{\left(X\right)}_{t_2}+c{\left(X\right)}_{t_1}]}\times \frac{c{(S)}_{t_2}-c{(S)}_{t_1}}{t_2-t_1}$$ (4)

The above parameters are mean values in a time range of 12 h, where t, c(X), c(P), and c(S) represent the culture time (h), DCW (g/liter), ε-PL titer (g/liter), and residual glucose concentration (g/liter) of the t (h), individually.

Statistical analysis.

All assays in this study were carried out at least in triplicate. The figures were produced using GraphPad prism 9.0 (San Diego, CA, USA), and the statistical analyses were performed using SPSS version 23.0 (SPSS, Inc., Chicago, IL, USA). The results are presented as means ± SD (n ≥ 3), and those at P < 0.05 are considered statistically significant.