Metabolite Cross-Feeding between Rhodococcus ruber YYL and Bacillus cereus MLY1 in the Biodegradation of Tetrahydrofuran under pH Stress

Rhodococcus species have been discovered in diverse environmental niches and can degrade numerous recalcitrant toxic pollutants. However, the pollutant degradation efficiency of these strains is severely reduced due to the complexity of environmental conditions and limitations in the growth of the pollutant-degrading microorganism. In our study, Bacillus cereus strain MLY1 exhibited strong stress resistance to adapt to various environments and improved the THF degradation efficiency of Rhodococcus ruber YYL by a metabolic cross-feeding interaction style to relieve the pH stress. These findings suggest that metabolite cross-feeding occurred in a complementary manner, allowing a pollutant-degrading strain to collaborate with a nondegrading strain in the biodegradation of various recalcitrant compounds. The study of metabolic exchanges is crucial to elucidate mechanisms by which degrading and symbiotic bacteria interact to survive environmental stress.

ABSTRACT

Bacterial consortia are among the most basic units in the biodegradation of environmental pollutants. Pollutant-degrading strains frequently encounter different types of environmental stresses and must be able to survive with other bacteria present in the polluted environments. In this study, we proposed a noncontact interaction mode between a tetrahydrofuran (THF)-degrading strain, Rhodococcus ruber YYL, and a non-THF-degrading strain, Bacillus cereus MLY1. The metabolic interaction mechanism between strains YYL and MLY1 was explored through physiological and molecular studies and was further supported by the metabolic response profile of strain YYL, both monocultured and cocultured with strain MLY1 at the optimal pH (pH 8.3) and under pH stress (pH 7.0), through a liquid chromatography-mass spectrometry-based metabolomic analysis. The results suggested that the coculture system resists pH stress in three ways: (i) strain MLY1 utilized acid metabolites and impacted the proportion of glutamine, resulting in an elevated intracellular pH of the system; (ii) strain MLY1 had the ability to degrade intermediates, thus alleviating the product inhibition of strain YYL; and (iii) strain MLY1 produced some essential micronutrients for strain YYL to aid the growth of this strain under pH stress, while strain YYL produced THF degradation intermediates for strain MLY1 as major nutrients. In addition, a metabolite cross-feeding interaction with respect to pollutant biodegradation is discussed.

IMPORTANCE Rhodococcus species have been discovered in diverse environmental niches and can degrade numerous recalcitrant toxic pollutants. However, the pollutant degradation efficiency of these strains is severely reduced due to the complexity of environmental conditions and limitations in the growth of the pollutant-degrading microorganism. In our study, Bacillus cereus strain MLY1 exhibited strong stress resistance to adapt to various environments and improved the THF degradation efficiency of Rhodococcus ruber YYL by a metabolic cross-feeding interaction style to relieve the pH stress. These findings suggest that metabolite cross-feeding occurred in a complementary manner, allowing a pollutant-degrading strain to collaborate with a nondegrading strain in the biodegradation of various recalcitrant compounds. The study of metabolic exchanges is crucial to elucidate mechanisms by which degrading and symbiotic bacteria interact to survive environmental stress.

INTRODUCTION

Microbial degradation is one of the most important methods used for the mineralization of hazardous substances present in contaminated environments. A large number of bacteria have been isolated from polluted environments and evaluated for their potential to degrade pollutants and investigated for the associated degradation mechanisms (1). Nevertheless, the bacterial consortium is the primary unit that performs biological functions, and most bacteria that survive in bacterial consortia depend on communication or substance exchange with other bacteria. Therefore, many degradation-associated microorganisms are not always found in monoculture (2–4) but rather occur in close association with other bacteria, forming microbial consortia that exhibit better biodegradation performance than single species (5, 6). As reported previously, a microbial consortium composed of Gordonia sp. strain JDC-2 and Arthrobacter sp. strain JDC-32 exhibited the ability to completely degrade di-n-octyl phthalate, where Gordonia sp. strain JDC-2 rapidly degrades di-n-octyl phthalate to phthalic acid followed by complete mineralization of phthalic acid by Arthrobacter sp. strain JDC-32 (7). Although some bacteria in microbial consortia cannot degrade target compounds, these bacteria exhibit the ability to change the structures of the contaminants or protect the degrading strains in the presence of toxic compounds or under environmental stress (8–10). Therefore, cooperation between microbes plays a significant role in the degradation of hazardous substances in situ.

Tetrahydrofuran (THF), one of the most polar ethers and a versatile solvent that is widely used in the synthesis of polymers, has been detected in groundwater, causing contamination and health problems. Moreover, THF has been classified as a refractory compound over time (11). To date, only a few strains of specific genera, including Rhodococcus (12–14), Pseudonocardia (15, 16), and Pseudomonas (17), have been reported to utilize THF as a sole carbon source. In our previous studies, one of the most efficient THF-degrading strains, named Rhodococcus ruber YYL, was isolated from activated sludge from a pharmaceutical factory. Strain YYL can tolerate THF at a concentration of 200 mM, which is nearly five times the concentration tolerated by other described THF-degrading strains. In addition, the THF degradation rate of strain YYL reached 137.60 mg THF h⁻¹ g⁻¹ dry weight, indicating that strain YYL has the highest THF degradation ability of the strains reported to utilize THF as a carbon source (13). Nevertheless, the optimum degradation performance is typically determined under optimized conditions, and the environment is not sufficiently stable to attain ideal conditions. Thus, degradation efficiency may be reduced due to changes in culture conditions, such as the pH value, heavy metal concentration, and temperature. Two bacilli, designated Bacillus cereus MLY1 and Bacillus aquimaris MLY2, were isolated simultaneously with strain YYL and supported it for successful bioaugmentation in an activated sludge reactor (18). Further studies revealed that strain MLY1 is conditionally helpful to strain YYL under harsh environmental conditions. The THF degradation efficiency of strain YYL was significantly reduced in the hostile environment, but the efficiency clearly recovered when strain YYL was cocultured with MLY1 (19). However, the interaction mechanism between the THF-degrading strain YYL and the helper strain MLY1 under stressful pH conditions remains unknown. Generally, metabolic exchanges are the primary ways by which microbes in a consortium interact with each other (20). Therefore, to elucidate the interaction mechanism between strains YYL and MLY1, it is necessary to elucidate the metabolic communication between them.

Based on our previous study (19), metabolic interactions were hypothesized to exist in a coculture system of strains YYL and MLY1. To identify the metabolic interactions between strains YYL and MLY1, it is necessary to determine the metabolic exchange and response profile between these strains under pH stress. In this study, strain YYL was mono- or cocultured with strain MLY1 at the optimum pH and at a lower pH. The strains were subjected to a physiological study combined with an extracellular metabolomics analysis to identify the metabolic exchange and response profile between the strains. Our results indicated that cross-feeding cooperation between strains YYL and MLY1 buffers the environmental pH and improves THF degradation by strain YYL.

RESULTS

THF degradation and bacterial growth in incubation systems.

In a previous study, the pH value was observed to decrease significantly with THF degradation by strain YYL, but the THF degradation rate was increased by coculture with strain MLY1 (13). In this study, physiological analyses were repeated to demonstrate the previously observed physiological effect. To this end, strains YYL and MLY1 were co- and monocultured at pH 7.0 and 8.3 (see Fig. S2 in the supplemental material). The results demonstrated that all of the coculture systems had higher THF degradation efficiencies than the monoculture systems. In addition, this difference was greater at pH 7.0 than at pH 8.3. The relative abundance of thm was much higher in the coculture system than in the monoculture systems (see Fig. S3A). In addition, the relative optical density at 600 nm (OD₆₀₀) value of strain MLY1 increased with the growth of strain YYL in the coculture at initial pH values of 7.0 and 8.3, but the relative proportion of strain YYL was higher at an initial pH of 7.0 than at pH 8.3 (Fig. S3B). Notably, the pH value was higher in the coculture system than in the monoculture systems when the OD₆₀₀ or THF concentration was the same. This finding strongly suggests that strain MLY1 can increase the pH value of the coculture system.

pH value test.

To test whether adjusting the pH value in the culture systems was the only way by which strain MLY1 assists strain YYL, the pH value was adjusted to 7.0 using a 1 M NaOH solution when the pH of the culture system decreased to 5.4 after 72 h of cultivation. Subsequently, the cooperator strain MLY1 was added to the culture systems. As predicted, compared to the group of monocultures, the THF degradation rate notably increased in the coculture system with or without adjusting the pH value, while the THF degradation rate of the strain YYL monoculture system decreased significantly when the other conditions were the same (Fig. 1A and B). Surprisingly, the coculture system exhibited the ability to completely and easily degrade 20 mM THF at pH 8.3, whether the pH was adjusted or not (Fig. 1A). Meanwhile, in pH 7.0, strain YYL performed much better when cocultured with MLY1 than in the monoculture, whether the pH was adjusted or not (Fig. 1B). These results showed that strain MLY1 may assist strain YYL not merely via adjusting pH in the coculture system.

FIG 1.

THF degradation curves of strain YYL mono- and cocultured with strain MLY1 after adjusting the pH value at 72 h with initial pH conditions of 8.3 (A) and 7.0 (B); triangle symbols represent mono- and cocultures without pH adjustment starting from 72 h of cultivation, and circle symbols represent mono- and cocultures with pH adjustment starting from 72 h of cultivation. (C) THF degradation curves of strain YYL mono- and cocultured with MLY1 in modified glass vessels. (D) The OD₆₀₀ (left y axis) and pH (right y axis) change curves of strain MLY1 cultured in the supernatant of strain YYL cultured after 72 h, with the bacteria removed using a 0.22-μm vacuum bottle filter. Significance was analyzed by Student’s t test (n = 4): *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Interaction mode between strains YYL and MLY1.

To determine the mode of interaction between strains YYL and MLY1, the two strains were separately cultured, allowing only extracellular metabolite exchange. The results showed that the separately cultured strain MLY1 still grew and had the ability to help strain YYL degrade THF under pH stress (initial pH 7.0) (Fig. 1C), indicating that strain YYL could interact with MLY1 without physical contact. In other words, extracellular metabolic interactions occurred during the cooperation of strain YYL with MLY1. Based on this finding, we proposed a noncontact interaction mode between strains YYL and MLY1; thus, an extracellular metabolomics analysis to assess the metabolic exchange and response profile between strains YYL and MLY1 was performed to elucidate the mechanism of interaction between the two strains.

Intermediate metabolites of THF used to culture strain MLY1.

Strain MLY1 was incubated and apparently grew in the supernatant of strain YYL, with the OD₆₀₀ value reaching 0.14 (Fig. 1D). Furthermore, the pH value of the system increased significantly from 5.2 to 6.4 in the presence of strain MLY1. Both strains YYL and MLY1 utilized 2-hydroxytetrahydrofuran and γ-butyrolactone as sole carbon sources to survive. Strain YYL utilized 5 mM γ-butyrolactone to reach an OD₆₀₀ of 1.102, but the highest OD₆₀₀ value reached was only 0.214 when 5 mM 2-hydroxytetrahydrofuran was used as the sole carbon source. In contrast to strain YYL, strain MLY1 exhibited OD₆₀₀ values of 0.201 and 0.753 using γ-butyrolactone and 2-hydroxytetrahydrofuran as sole carbon sources, respectively (see Fig. S4).

Extracellular metabolites in mono- and coculture systems.

Liquid chromatography-mass spectrometry (LC-MS) was performed in the positive electrospray ionization (ESI+) and negative electrospray ionization (ESI−) modes to identify the metabolites in the samples. In this study, 1,187 and 2,085 metabolites were detected in the ESI+ and ESI− modes, respectively. Subsequently, 84 and 102 metabolites from the ESI+ and ESI− modes, respectively, were identified and annotated after removing the redundant metabolites using Metlin (see Tables S3 and S4).

Metabolites that exhibited significantly different levels between the strain YYL monoculture and the coculture with strain MLY1 in the ESI+ and ESI− modes were identified by orthogonal partial least-squares discriminant analysis (OPLS-DA), and the selected metabolites exhibited variable importance in projection (VIP) of >1 between the two groups (see Tables S5 and S6). In addition, to visualize metabolic differences between the sample groups, multivariate statistical analysis with PLS-DA (partial least-squares discriminant analysis) was performed on all the extracellular metabolites identified by the ESI+ and ESI− modes at three time points (60 h, 96 h, and 120 h) except for the outliers (see Fig. S5). The results showed that the groups of the monoculture systems clearly separated at the three time points, especially at 120 h at different initial pH values. However, none of the plots of the groups in the coculture system separated completely, even parts of samples mixed together. This finding demonstrated that the pH value significantly influenced the metabolism of strain YYL in monoculture from the early exponential phase to the stationary phase. Moreover, the coculture with strain MLY1 clearly reduced the effect of pH stress.

DISCUSSION

The environmental pH plays a vital role in bacterial growth and metabolism (21). However, the knowledge of how bacteria resist low-pH stress with the help of other strains is poorly understood. The THF-degrading strain YYL exhibited weak resistance to low-pH stress and showed a low growth rate and THF degradation efficiency under low-pH stress. Considering that the coculture of strain YYL with the non-THF-degrading strain MLY1 notably improved the THF degradation efficiency under stressful pH conditions, it is important to elucidate the interaction mechanism between these strains to promote THF degradation. In this study, a noncontact interaction mode between strains YYL and MLY1 was ensured by separately culturing the two strains while maintaining extracellular metabolite exchange. It is noteworthy that metabolite communication has been widely shown to mediate contact-independent interactions between bacteria (22–24). Based on these findings, we focused on elucidating the metabolic interaction mechanism between strains YYL and MLY1 by examining the changing profiles of metabolites in the mono- and coculture systems at the optimal pH of 8.3 and a stressful pH of 7.0. Their metabolites exhibited significantly different patterns in the mono- and coculture systems, demonstrating the metabolic interactions between strains YYL and MLY1 or the different responses to environmental stress during THF degradation. These results indicated that strain MLY1 obtained nutrients from THF degradation by strain YYL and helped strain YYL to resist low-pH stress by a series of metabolic exchanges. Thus, YYL exhibited better THF-degrading activity in the coculture system than in the monoculture system under stressful pH conditions.

Strain MLY1 utilizes acidic metabolites to alleviate pH stress for strain YYL.

The pH value was determined to be a key factor that influences THF degradation by strain YYL, and the alkaline environment was observed to be more suitable for THF degradation by strain YYL (see Fig. S2 in the supplemental material). The production of acidic metabolites has been consistently shown to reduce the pH value in the culture system along with THF degradation (19). In this study, multiple acidic metabolites, including pyruvic acid, l-ascorbic acid, citramalic acid, and acetoacetic acid, were detected in the ESI− mode that resulted in a decrease in the pH value. Alternatively, strain MLY1 may obtain energy from the degradation products generated by strain YYL, as it is unable to utilize THF as a carbon source. Several acid metabolites generated from THF degradation by strain YYL were utilized by strain MLY1, supporting the growth of this strain in the supernatant of strain YYL along with the increase in the pH value (Fig. 1C). This result explained why the coculture system exhibited a higher pH value than the monoculture system under the same THF concentration or at the same OD₆₀₀ value. Moreover, these four metabolites had significantly higher relative abundances in the monoculture system than in the coculture systems (Fig. 2A to C). Indeed, strain YYL was unable to degrade THF at pH values of less than 4.0; thus, pH 4.0 was the limit for THF degradation by strain YYL (Fig. S2). Furthermore, 20 mM THF was not completely degraded by strain YYL, because the pH value decreased rapidly with THF degradation and the optimized medium described in a previous study is a type of oligotrophic medium that does not have sufficient buffering capacity. The coculture system had the ability to completely degrade the 20 mM THF when the pH value was adjusted to 7.0 after cultivation for 72 h (Fig. 1A and B), indicating that the cooperator strain MLY1 could consume the acidic metabolites in the coculture to alleviate the pH stress for strain YYL. Unexpectedly, adjusting the pH value using an NaOH solution did not completely improve the THF degradation rate in the YYL monoculture system compared to that of the coculture system. Therefore, the consumption of acidic metabolites by strain MLY1 is not the only benefit that strain YYL obtains from strain MLY1.

FIG 2.

The relative abundances of acid metabolites when strain YYL was mono- and cocultured with strain MLY1 at initial pH values of 7.0 and 8.3 at 60 h (A), 96 h (B), and 120 h (C). Significance was analyzed by Student’s t test (n = 4): *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Strain MLY1 consumes intermediate metabolites of THF to promote THF degradation.

THF is degraded to 2-hydroxytetrafuran under oxidation by THF monooxygenase and then transformed to γ-butyrolactone (25). Strain MLY1 degraded 2-hydroxytetrafuran better than γ-butyrolactone as a sole carbon source (Fig. S3), whereas strain YYL utilized γ-butyrolactone well but not 2-hydroxytetrafuran. The accumulation of intermediate metabolites such as 2-hydroxytetrafuran may inhibit THF degradation or be toxic to the THF-degrading strain YYL, but the consumption of intermediate metabolites by strain MLY1 may reduce the inhibitory effect and allow strain YYL to degrade THF. This finding demonstrates the complementarity in the metabolite utilization by the two strains, promoting a desirable distribution of energy sources and increasing the THF degradation efficiency.

Nutrient metabolite exchange between strain YYL and strain MLY1.

Strain MLY1 does not possess the ability to utilize THF as a carbon source but has the ability to consume the energy source generated by strain YYL to survive. Succinic aldehyde is the metabolite that links THF metabolites and the tricarboxylic acid (TCA) cycle, because succinic acid is generated from succinic aldehyde by succinic aldehyde dehydrogenase (26). The energy from the metabolites involved in the TCA cycle is more easily used than other energy sources. Succinic aldehyde levels exhibited no significant difference among the samples. Furthermore, strain MLY1 was unable to utilize succinic aldehyde as an energy source, because there is no succinic aldehyde dehydrogenase gene present in the genome of this organism (genome accession number NC_004722). Remarkably, metabolites involved in the TCA cycle, including succinic acid, fumaric acid, pyruvate, and oxoglutaric acid, were detected in this study. Among these metabolites, only pyruvate exhibited a significantly higher abundance in the monoculture systems than in the coculture system, with succinic and fumaric acids having slightly higher abundances in the coculture system and oxoglutaric acid levels exhibiting almost no difference among the systems (see Fig. S6). Consequently, the major nutrient for strain MLY1 in the TCA cycle may be pyruvate. The metabolic data also showed several other types of nutrient metabolites with higher abundances in the monoculture systems than in the coculture system, including citramalic acid, d-glucose-6-phosphate, amino l-valine, 3-hydroxy-l-proline, l-leucine, N-acetyl-l-phenylalanine, pyroglutamic acid, β-alanine, 2-phenylglycine, and l-ascorbic acid (Fig. S6). To survive in the coculture system, strain MLY1 may use pyruvate, a highly abundant acid metabolite from the TCA cycle, and the other types of nutrient metabolites mentioned above as energy sources.

In contrast, some nutrients that are not required but are important to bacteria exhibited higher abundances in the coculture system than in the monoculture systems, including 2-phenyl butyramide, farnesylcysteine, N-acetylleucine, sedoheptulose, stearamide, 5-methylcytidine, and (R)-5-phosphomevalonate (Fig. 3). Among these nutrients, 2-phenyl butyramide acts as an inhibitor of cell receptors, farnesylcysteine, 5-methylcytidine, and (R)-5-phosphomevalonate are required for cell survival, sedoheptulose is associated with the cell structure, farnesylcysteine and its analogs have high affinity for methyltransferase and function as specific inhibitor of the COOH-terminal S-farnesyl-cysteine methyltransferase, stearamide and cresol are antibacterial compounds, and (R)-5-phosphomevalonate is the synthetic precursor of some biomolecules in the phosphomevalonate pathway (27, 28). All of these metabolites play important roles in the basal metabolism of microorganisms.

FIG 3.

The abundances of seven metabolites, namely, 2-phenyl butyramide (A), farnesylcysteine (B), N-acetylleucine (C), sedoheptulose (D), stearamide (E), 5-methylcytidine (F), and (R)-5-phosphomevalonate (G), in mono- and coculture systems under initial pH values of 7.0 and 8.3 at 60 h, 96 h, and 120 h. Significance was analyzed by Student’s t test (n = 4): *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Based on the pattern of nutrient changes observed in different culture systems, the nutrient exchange between strains YYL and MLY1 was elucidated. Strain MLY1 utilized large amounts of basal nutrients, such as various amino acids and other intermediates, to survive in coculture systems, whereas strain MLY1 produced micronutrients to help strain YYL grow and resist environmental stress. Remarkably, 2-phenyl butyramide, farnesylcysteine, sedoheptulose, stearamide, and (R)-5-phosphomevalonate had higher abundances in the pH 7.0 coculture system than in the pH 8.3 coculture system, which may explain why strain MLY1 was able to help strain YYL at pH 7.0 but not at pH 8.3.

High proportion of glutamine suggests a better acid resistance capacity.

Amino acids protect cells in several ways, including by acting as osmolytes, which affects cellular water retention and increases the stability of proteins (29, 30). Glutamate and glutamine are compatible solutes and play important roles in acid resistance (31–34). This study revealed that both glutamate and glutamine were present at high abundances in all of the samples (Fig. S6A and B). The glutamate abundance decreased over time in all of the culture systems, which positively correlated with the pH variation, and this change is conducive for the resistance of acid damage in cellular metabolism. Furthermore, the glutamate/glutamine abundance ratio was lower in the coculture system than in the monoculture systems, and the coculture system at pH 7.0 had the lowest glutamate/glutamine abundance ratio, especially at 120 h (Fig. 4). These results illustrated that the consortium formed by strains YYL and MLY1 has a better acid resistance capacity in response to changes in environmental pH than the monocultured strain YYL; thus, the bacterial consortium is much stronger than the individual strains under conditions of acid stress.

FIG 4.

The proportions of l-glutamate and l-glutamine under different conditions. Strain YYL monoculture system at initial pH values of 7.0 (indicated by uppercase A) and pH 8.3 (indicated by C); strains YYL and MLY1 in a coculture system at initial pH values of 7.0 (indicated by B) and pH 8.3 (indicated by D). 1, 2, and 3 indicate the three time points of 60, 96, and 120 h, respectively.

In summary, the THF-degrading strain YYL and the cooperator strain MLY1 had several types of extracellular metabolite communication depending on the contact-independent interaction mode under different initial pH values. Based on the results above, we propose that the strains MLY1 and YYL exhibit a cross-feeding type of metabolic cooperation (Fig. 5). In the coculture system with THF as the sole carbon source, strain YYL supplied major nutrients to strain MLY1, and in return, strain MLY1 produced some important micronutrients that helped strain YYL grow under pH stress. Furthermore, strain MLY1 utilized the intermediate metabolites of THF to eliminate the product limitation of strain YYL and decrease the abundance of acidic metabolites in the coculture system. Additionally, the presence of strain MLY1 influenced the proportion of glutamine, contributing to the elevated intracellular pH under acidic environmental conditions, helping strain YYL resist acid stress. Based on the observed interactions between strains YYL and MLY1, the symbiotic system based on metabolic cross-feeding could be stabilized to completely degrade the THF.

FIG 5.

The proposed schematic depiction of the cross-feeding mechanism of interspecific metabolism between strains YYL and MLY1 during THF degradation. Strain YYL degrades THF to intermediates (purple diamonds) and produces acidic metabolites (red diamonds) that are used as carbon sources by strain MLY1. Strain MLY1 influences the glutamate/glutamine ratio to resist acid stress (blue circles and yellow diamonds). Furthermore, strain MLY1 provides important micronutrients (green circles) to strain YYL to help the growth of this strain under pH stress.

MATERIALS AND METHODS

Strains, culture conditions, and coculture experiment.

The THF-degrading strain YYL was cultured in 100 ml of base mineral medium (BMM) supplemented with 20 mM THF (13). Strain MLY1 was cultured in 100 ml of BMM with 1.0 g/liter yeast extract. The cells of strains YYL and MLY1 from the late logarithmic growth stage were centrifuged (5 min at 8,000 × g), the growth medium was decanted, and the cells were washed three times with BMM. The washed cells of strains YYL and MLY1 were resuspended with BMM at pH values of 8.3 and 7.0, respectively. Then, 100 ml of BMM (pH 8.3 or 7.0) with 20 mM THF was inoculated with 2 ml of strain YYL (OD₆₀₀ of 1.5) and 1 ml of MLY1 (OD₆₀₀ of 1.5) to attain an original cell ratio of 2:1. To ensure that the volume of the monoculture system was the same as the coculture system, an additional 1 ml of BMM was added to the monoculture systems instead of 1 ml of strain MLY1. One Erlenmeyer flask containing culture was used for each sample, and ten replicates were prepared for each sampling time point in the four treatment groups. The interaction mode of the coculture was tested using modified glass vessels, each holding 500 ml, with 100 ml of BMM on the right and left sides. A 0.22-μm nylon membrane filter (Membrane Solutions, USA) was used to separate the two culture chambers, allowing for communication to occur only via extracellular metabolites (see Fig. S1 in the supplemental material). Modified glass vessels were used only to verify noncontact interactions, while other experiments were performed using contacting cocultures.

Sample collection.

Two-milliliter samples were collected from each Erlenmeyer flask every 12 h in the exponential growth phase and every 24 h in the lag and stationary phases. Then, 1 ml of each sample was used to measure the optical density at 600 nm and the pH value. The remaining 1 ml was centrifuged at 4°C at 5,000 × g for 10 min, and the supernatant was subjected to THF concentration determination using gas chromatography (GC-2014C; Shimadzu, Japan). Two milliliters of each sample was collected to extract DNA and RNA. Three sampling time points, 60 h (early stage of the exponential phase), 96 h (middle stage of the exponential phase), and 120 h (early stage of the stationary phase), were used for each treatment. Five-milliliter samples were collected to measure the metabolite levels in each culture system. The samples were chilled on ice to quench the metabolic activity and then centrifuged at 4°C at 5,000 × g for 10 min. Subsequently, the supernatant and bacteria were separately collected, flash frozen in liquid nitrogen, and stored at −80°C.

To detect the growth of strain MLY1 in mono- and coculture with strain YYL, the ratio of the abundances of the genes thm and gerM at the DNA level was quantified by quantitative PCR (qPCR) (35, 36). The gene thm encodes the THF monooxygenase of YYL, which is responsible for the first step in THF degradation, as described in a previous study (37). The gerM gene encodes a lipoprotein that stabilizes the GerA-GerQ complex via an interaction with the remodeled cell wall in spore formation by bacilli (38). The primers to assay thm and gerM by qPCR are shown in Table S1. The samples used to test for the interaction mode were collected on the right and left sides of each glass vessel every 24 h to measure the THF concentration, OD₆₀₀, and pH values.

pH calibration assays.

To determine whether strain MLY1 only assists strain YYL by improving the pH condition, strain YYL was cultured in 100 ml of BMM with 20 mM THF at an initial pH of 8.3 or 7.0. When the pH value of the culture systems decreased below 5.5, NaOH (1 M) was used to adjust the pH to 7.0. Then, strain MLY1 was added to the culture systems with or without adjusting the pH value. The samples were collected every 12 h to measure the THF degradation rate.

Intermediate metabolites of THF used to culture strain MLY1.

Strain YYL was cultured in 100 ml of BMM supplemented with 20 mM THF for 72 h at an initial pH of 8.3. Strain MLY1 was added to the supernatant after removing the bacteria using a 0.22-μm vacuum bottle filter (BIOFIL, China). Samples were collected every 24 h to measure the OD₆₀₀ and pH values. Furthermore, 2-hydroxytetrahydrofuran and γ-butyrolactone, two intermediate metabolites of THF, were chosen to test the ability of strain MLY1 to use these compounds. Strain YYL or MLY1 was cultured in 100 ml of BMM with 5 mM 2-hydroxytetrahydrofuran or γ-butyrolactone at an initial pH of 8.3, with samples collected every 6 or 12 h, respectively, to measure the OD₆₀₀.

Metabolite extraction.

The samples were thawed at room temperature, and 1.5 ml of each sample was transferred to centrifuge tubes (1.5 ml) and extracted with 4.5 ml of methanol. All of the samples were vortexed for 30 s and centrifuged at 12,000 × g at 4°C for 15 min. The supernatant was transferred to centrifuge tubes (1.5 ml) and centrifuged at 3,000 × g at room temperature. Then, 300 μl of acetonitrile (10 μg/ml of internal standard, dl-o-chlorophenylalanine) was dissolved in the supernatant, which was followed by centrifugation at 12,000 × g at 4°C for 15 min. Then, 200 μl of the supernatant was transferred to a vial for LC-MS analysis.

LC-MS measurements and data processing.

All LC-MS spectra of the extracellular metabolites were recorded using an Ultimate 3000 LC Orbitrap Elite (Thermo, USA) and a Hypergod C₁₈ column (100 mm by 4.6 mm, 3 μm). The chromatographic separation conditions were as follows: column temperature, 40°C; flow rate, 0.3 ml/min; mobile phase A, water with 0.1% formic acid; mobile phase B, acetonitrile with formic acid; injection volume, 4 μl; automatic injector temperature, 4°C. The gradient of the mobile phase is summarized in Table S2. Two modes were used for measurements. For the positive ionization mode (ESI+), the following conditions were used: heater temperature, 300°C; sheath gas flow rate, 45 arbitrary units; aux gas flow rate, 15 arbitrary units; sweep gas flow rate, 1 arbitrary unit; spray voltage, 3.0 kV; capillary temperature, 350°C; S-lens RF level, 30%. For the negative ionization mode (ESI−), the following conditions were used: heater temperature, 300°C; sheath gas flow rate, 45 arbitrary units; auxiliary gas flow rate, 15 arbitrary units; sweep gas flow rate, 1 arbitrary unit; spray voltage, 3.2 kV; capillary temperature, 350°C; and S-lens RF level, 60%.

The data were subjected to feature extraction and preprocessing using SIEVE software (Thermo) and then normalized and edited into a two-dimensional data matrix using Excel 2010, including retention time (RT), compound molecular weight (compMW), observations (samples), and peak intensity. The features in the ESI+ and ESI− modes in this study were subjected to multivariate analysis (MVA) using SIMCA-P (Umetrics AB, Umea, Sweden) after editing.

RNA and DNA isolation, cDNA generation, qPCR, and RT-qPCR.

Total RNA of all the samples was extracted separately using a Bacterial RNA kit (Omega, USA), with contaminating DNA in the RNA samples removed using DNase I (TaKaRa, Dalian, China). The DNA of all the samples was isolated using a Bacterial DNA kit (Omega, USA). The quality and yield of the DNA and RNA samples were measured via 1% agarose gel electrophoresis and a NanoDrop 2000 spectrophotometer (Thermo Scientific, Wilmington, DE). cDNA synthesis was performed using a Prime Script reverse transcriptase reagent kit (TaKaRa).

The qPCR and reverse transcriptase quantitative PCR (RT-qPCR) reactions were performed in 20-μl volumes containing 10 μl SYBR Premis Ex Taq (TaKaRa), 0.4 μl of each forward and reverse gene-specific primers (10 μM), and 2.0 μl of DNA or cDNA (1:10 dilution). The specific primers for the genes were designed using Beacon Designer (v7) (see Table S1). Two housekeeping genes were chosen as internal standards: the RNA polymerase beta subunit (rpoB) (39) and gyrase A (gyrA) (40) genes. All RT-qPCR experiments were performed on a Rotor-Gene Q instrument (Qiagen, Hilden, Germany) according to the user’s guide. Three independent DNA or cDNA samples were assayed, and four repetition operations were performed for each sample. The 2^−ΔΔCT method (where C_T is the threshold cycle) was used to calculate the relative gene expression levels for strain YYL in the mono- and coculture systems. The genes' expression at 60 h was normalized to 1.