Microbial Interactions in a Vitamin C Industrial Fermentation System: Novel Insights and Perspectives

ABSTRACT

In industrial production, the precursor of l-ascorbic acid (L-AA, also referred to as vitamin C), 2-keto-l-gulonic acid (2-KLG), is mainly produced using a classic two-step fermentation process performed by Gluconobacter oxydans, Bacillus megaterium, and Ketogulonicigenium vulgare. In the second step of the two-step fermentation process, the microbial consortium of K. vulgare and B. megaterium is used to achieve 2-KLG production. K. vulgare can transform l-sorbose to 2-KLG, but the yield of 2-KLG is much lower in the monoculture than in the coculture fermentation system. The relationship between the two strains is too diverse to analyze and has been a hot topic in the field of vitamin C fermentation. With the development of omics technology, the relationships between the two strains are well explained; nevertheless, the cell-cell communication is unclear. In this review, based on current omics results, the interactions between the two strains are summarized, and the potential cell-cell communications between the two strains are discussed, which will shed a light on the further understanding of synthetic consortia.

INTRODUCTION

Vitamin C (l-ascorbic acid, L-AA) is an essential water-soluble vitamin in humans, primates, and a few other mammals and is widely used in the pharmaceutical, food, cosmetics, and feed industries for its antioxidant capacity (1). L-AA was first discovered and isolated from plants and the adrenal cortex to prevent scurvy, a common disease among the world’s sailors and navies in the early 19th century, with serious symptoms of bleeding of mucous membranes, anemia, and ultimately, death (2, 3). After the discovery of L-AA, the development of industrial production processes was triggered by the demand of pure L-AA in the early 1930s. Currently, L-AA is produced via a classic two-step fermentation system on an industrial scale and has achieved huge annual economic benefits (4, 5).

During the classic two-step fermentation system, Gluconobacter oxydans converts d-sorbitol to l-sorbose in the first step, and the yield of l-sorbose from d-sorbitol reaches 98%. Then, the consortium consisting of Ketogulonicigenium vulgare and Bacillus megaterium converts l-sorbose to 2-keto-l-gulonic acid (2-KLG, the precursor of L-AA), and the yield of 2-KLG is >97% from l-sorbose (1, 6) (Fig. 1). Although the classic two-step fermentation system has a higher 2-KLG yield and has achieved large-scale industrial production, there are some typical disadvantages, such as higher energy consumption and investment in fermentation equipment due to secondary sterilization, compared to the one-step direct fermentation process (4). Hence, cutting costs with a one-step method rather than the two-step process has been a major target of researchers for more than 30 years, but this has not yet been achieved (1). To break through the bottleneck, it is crucial to understand the interactions between the two strains in the second step of the classic two-step fermentation system and the reasons for the high production of 2-KLG.

FIG 1.

The process of the classic two-step fermentation system of vitamin C. The green dashed line represents substances that are provided to K. vulgare and the surroundings by Bacillus megaterium.

K. vulgare has all the key enzymes for the conversion of l-sorbose to 2-KLG, but with extremely low yields in the monoculture fermentation system, whereas a huge increase in 2-KLG production was achieved in the coculture fermentation system with Bacillus strains (7, 8). Bacillus strains such as Bacillus megaterium, Bacillus endophyticus, Bacillus pumilus, and Bacillus thuringiensis, as helper strains, are Gram-positive, nonmotile, and rod-shaped bacteria that lack all of the key enzymes required to convert l-sorbose to 2-KLG but have been reported to promote the growth and 2-KLG production of K. vulgare (9, 10). The relationship between the two strains has been a hot topic in the field of L-AA microbial fermentation. Currently, with the development of high-throughput sequencing and omics technology, the interactions between the two strains have been well explained. However, the cell-cell communication between the two strains is still unclear. Here, we summarized recent developments in the field of L-AA microbial fermentation systems and discussed the potential cell-cell communications and interactions between the two strains based on quorum sensing.

SUBSTANCE EXCHANGES IN THE COCULTURE FERMENTATION SYSTEM

In natural environments, microorganisms normally live in communities composed of multiple species and are able to perform more diverse and complex behavior than clonal populations since different members of the consortium take on different responsibilities (11). Therefore, it is essential to understand the interactions among the members of the microbial consortium. Generally, microorganisms exchange biomolecules (such as proteins, nucleic acids, and metabolites) and information signals to interact with each other (12, 13). Omics studies (e.g., genomics, transcriptomics, metabolomics, and proteomics) can provide valuable insights into the metabolic cross talk mechanism of the consortium and provide a direction for better understanding the relationship between species (11). Here, we briefly describe the substance exchanges of several biomolecules involved in the L-AA coculture fermentation system based on the available omics results (Fig. 2).

FIG 2.

Schematic representation of the interaction between K. vulgare and Bacillus strains. The red fonts and lines represent the substances that K. vulgare fails to de novo biosynthesize, which are produced and provided by the helper strain and have a vital role in the growth and 2-KLG production of K. vulgare. The blue dashed line represents substances that are provided to K. vulgare by the helper strain.

Amino acids.

Based on the genomic analysis, K. vulgare lacks key enzymes involved in the de novo biosynthesis pathways of l-glycine, l-histidine, l-lysine, l-proline, l-cysteine, l-threonine, l-methionine, l-leucine, and l-isoleucine. l-glycine, l-proline, l-threonine, and l-isoleucine play crucial roles in the growth and 2-KLG production of K. vulgare (3, 9, 14, 15). Among them, l-glycine, l-proline, and l-threonine could be transformed into intermediates of the tricarboxylic acid (TCA) cycle and be used for the generation of energy or other components, and the addition of these three amino acids significantly increased the growth of K. vulgare (16). To further investigate the effect of amino acids on the growth and 2-KLG production of K. vulgare, Pan et al. reconstructed the threonine biosynthesis pathways and increased the 2-KLG productivity of K. vulgare (17). Moreover, Zhou et al. analyzed the interaction mechanism between the two microorganisms in the consortium consisting of K. vulgare and B. megaterium based on metabolomics and showed that these microorganisms interacted with each other by exchanging several metabolites, such as amino acids (8). The contents of amino acids and other metabolites in K. vulgare were rather low compared to those in B. megaterium, but the levels of these compounds were much higher in the medium surrounding K. vulgare than in fresh medium (18). This indicates that the helper strain, such as Bacillus strains, has a complete relative metabolic capacity to participate in the supply of amino acids to K. vulgare, especially l-glycine, l-cysteine, l-methionine, and l-tryptophan, which K. vulgare cannot synthesize on its own, and to promote the growth and increase the 2-KLG production of K. vulgare.

Proteins.

In a previous study, two extracellular proteins of helper strains were found to promote the growth and 2-KLG production of K. vulgare, with molecular weights of 30 to 50 kDa and >100 kDa, respectively (9, 14). To further investigate which proteins of the helper strain promote the growth and 2-KLG production of K. vulgare, the proteins released into the extracellular environment of the helper strain were annotated with the help of metabolomic and proteomic analysis. It was found that in addition to the proteins associated with sporulation and flagella, extracellular esterase, aminopeptidase, and polysaccharide deacetylase capable of digesting large molecular substances in the environment were also detected. Furthermore, the proteins that respond to intracellular oxidative stress, such as superoxide dismutase, catalase, glucose-6-phosphate dehydrogenase, and oxidoreductase, were annotated (3, 8, 9, 18). These discoveries indicated that the helper strain provides key proteins for the growth and 2-KLG production of K. vulgare.

Vitamin B.

B vitamins, including B₁, B₂, B₃, B₅, B₆, B₇, B₉, and B₁₂, are water-soluble vitamins that act as essential cofactors and coenzymes involved in several metabolic pathways of organisms. Among them, vitamins B₁ (thiamine), B₂ (riboflavin), and B₅ (pantothenic acid) are involved in the TCA cycle and fatty acid oxidation. Vitamin B₃ (niacin), a precursor of NAD, is involved in the respiratory chain. Vitamins B₆, B₇ (biotin), and B₉ (folate) are involved in various cellular metabolic processes, including the metabolism of amino acids, lipids, and carbohydrates and the synthesis of DNA and amino acids. Vitamin B₁₂ (cobalamin) catalyzes the synthesis of methionine (19).

Although B vitamins, including B₁, B₂, B₃, B₆, B₇, and B₁₂ have a vital role in promoting the biomass and 2-KLG production of K. vulgare, it was found that the key genes for the de novo biosynthesis of several B vitamins are lacking in K. vulgare based on genomics (9, 14, 15). Helper strains such as B. endophyticus and B. thuringiensis have been reported to have a complete biosynthetic pathway of B vitamins except lipoic acid and B₁₂, which could be released into the environment and absorbed by K. vulgare (16).

Siderophores.

Metal elements have been reported to participate in the growth and 2-KLG production of K. vulgare, especially zinc (Zn), iron (Fe), manganese (Mn), copper (Cu), potassium (K), sodium (Na), magnesium (Mg), calcium (Ca), chromium (Cr), and cadmium (Cd) (15). In addition, the addition of rare earth elements such as lanthanum (La), cerium (Ce), neodymium (Nd), and samarium (Sm) also promotes 2-KLG production (20). Among them, Fe plays a vital role in bacterial metabolism as a cofactor for many enzymes in metabolic pathways (such as iron-sulfur clusters, antioxidant enzymes, and heme proteins) (21). Generally, bacteria secrete siderophores to absorb Fe in the environment. In addition to Fe, siderophores also showed affinity for other metals and can interact with Zn, Cu, Cr, Cd, etc. (22, 23). Based on integrated proteomics and metabolomics, proteins related to inorganic ion transport and metabolism of K. vulgare were extensively detected (18). Furthermore, our lab found that K. vulgare is a nonproducer of siderophores and can use siderophores of the helper strain to promote 2-KLG production (10).

REDUCED OXIDATIVE STRESS IN THE COCULTURE FERMENTATION SYSTEM

In K. vulgare, oxidative stress, which is induced by reactive oxygen species (ROS), inhibits the growth and 2-KLG production (24). ROS such as superoxide radicals (O₂^·–), hydroxyl radicals (OH^·), hydrogen peroxide (H₂O₂), lipid peroxides, and peroxides of proteins and nucleic acids are the consequence of oxygen metabolism in aerobic organisms, which attack DNA, leading to respiration chain breaks and changes in the carbohydrate parts and nitro bases and to point mutations. ROS also cause damage to proteins and lipids (18, 25). Generally, ROS are generated mainly through electron escape from coenzyme Q of the electron-transport chain (ETC). Then, the escaped electrons from the ETC interact with molecular oxygen to produce O₂^·–, which can be further converted to OH^· and H₂O₂ either spontaneously or enzymatically (26). The control of ROS homeostatic levels is achieved not only by their production but also by their elimination. Organisms possess multilevel and sophisticated antioxidant systems to eliminate ROS or minimize negative effects, such as low-molecular-mass antioxidants (<1,000 Da) and high-molecular-mass antioxidants (>10,000 Da). Generally, low-molecular-mass antioxidants include vitamins C and E, polyphenols, anthocyanins, carotenoids, and glutathione, whereas, high-molecular-mass antioxidants include superoxide dismutase (SOD), catalase (CAT), glutathione-dependent peroxidases, glucose-6-phosphate dehydrogenase, and NADP⁺ oxidoreductase (21, 25, 26). Among them, SOD removes O₂^·– by catalyzing its dismutation, and then CAT decomposes H₂O₂. Therefore, the imbalance between ROS generation and elimination in organisms under aerobic conditions is functionally defined as oxidative stress.

In K. vulgare, it was shown that oxidative stress limits cell growth and 2-KLG production (24). Based on metabolomics and proteomics, it is believed that the helper strain releases antioxidants to reduce the oxidative stress of K. vulgare in the coculture fermentation system. Furthermore, the addition of antioxidants, such as glutathione and vitamin C, has been reported to enhance the growth and 2-KLG yield of K. vulgare in the coculture fermentation system (24, 27). What are the reasons for the generation of oxidative stress in K. vulgare? In addition to electron leakage in the ETC of K. vulgare, 2-KLG, an oxidative product of K. vulgare, has been reported to lead to oxidative stress and DNA damage in G. oxydans (4) and to inhibit the growth of the helper strain (28). Wang et al. reported that 2-KLG production is coupled with the respiratory chain via cytochrome c₅₅₁ (29). l-sorbose/l-sorbosone dehydrogenase (SSDH), the main enzyme catalyzing the conversion of l-sorbose to l-sorbosone and ultimately to 2-KLG, as quinoprotein dehydrogenase, contains a cytochrome c domain that is potentially available as a physiological electron acceptor and involved in electron transfer (30, 31). Another quinoprotein dehydrogenase, l-sorbosone dehydrogenase (SNDH), which directly converts l-sorbosone to 2-KLG, as well as L-AA, was expected to transfer electrons from the substrate to the membrane-bound cytochrome c oxidase via pyrroloquinoline quinone (PQQ) and heme c (32). Although SSDHs and SNDHs are quinoproteins and participate in the respiratory chain, the specific electron transport pathways in K. vulgare remain obscure. The electron transport pathway between 2-KLG production and the respiratory chain of K. vulgare was inferred with reference to the electron transfer reactions of acetic acid bacteria (Gluconobacter and Acetobacter) (33, 34) (Fig. 3). These findings suggested that 2-KLG may lead to oxidative stress and inhibit the growth of K. vulgare; conversely, oxidative stress attacks the respiratory chain associated with 2-KLG biosynthesis and limits the production of 2-KLG.

FIG 3.

Schematic representation of the relationship between 2-KLG production and the respiratory chain in K. vulgare.

CELL-CELL COMMUNICATIONS IN THE COCULTURE FERMENTATION SYSTEM

Quorum sensing (QS) is a cell-to-cell communication mechanism that allows bacteria to act as multicellular organisms and control sophisticated behaviors such as competence, virulence factor secretion, biofilm formation, bioluminescence, antibiotic production, motility, and sporulation (35). Generally, bacteria use the production, release, exchange, and detection of extracellular signal molecules (called autoinducers or quormones) to measure their population density and to control their behavior depending on changes in cell numbers (36–38). In the microbial world, all bacteria use QS (Fig. 4), and there are four classes of known bacterial cell-cell signaling molecules that have been detected, N-acylhomoserine lactones (AHLs) (35), autoinducing peptides (AIPs) (39, 40), autoinducer-2 (AI-2) (41–43), and other types (36).

FIG 4.

Generalized model of the quorum sensing receptor mechanism. In Gram-negative bacteria, LuxR-type receptors bind to endogenous AHLs that are produced by LuxI. In Gram-positive bacteria, the QS receptors are mainly cytoplasmic receptors and membrane-bound sensor kinases (such as Rap and ComP), and their ligands are unmodified short and long peptides or modified peptides, respectively. AI-2, the “universal language” in the bacterial world, shares a common precursor, 4,5-dihydroxy-2,3 pentanedione (DPD), which is a product of LuxS. In Gram-negative bacteria, LuxP is the receptor for AI-2.

Although quorum sensing is considered to be widespread in the bacterial world, cell-cell communication in the field of vitamin C microbial fermentation systems remains unclear. Based on genomics, K. vulgare contains LuxR (NCBI SPY12106.1) and lacks LuxI (16) but has an N-acyl-l-homoserine lactone synthetase-like protein (NCBI AOZ55678.1), which indicated that K. vulgare could produce AHLs to participate in AHL-LuxR quorum sensing. Interestingly, most of the helper strains, as Gram-positive bacteria, have only LuxR, which allows them to respond to the AHLs produced by K. vulgare and makes it possible for them to team up or switch to competition against their neighbors (44). Based on LuxI/LuxR quorum sensing, Wang et al. established a programmed cell death module in G. oxydans to reduce the competition between G. oxydans and K. vulgare in a three-species consortium, and promote the growth and 2-KLG production of K. vulgare (45). In previous studies, sporulation and spore ability of the helper strain have also been reported to enhance K. vulgare propagation and 2-KLG production (46). The sporulation of helper strains is controlled by the quorum sensing system, including the two-component pathway and the self-signaling pathway in the RNPP (Rap, NprR, PrgX, and PlcR) family (47). In addition, AI-2, as a “universal language” for cross talk, may even extend to interspecies communication, possibly participating in cell-cell communication and playing a role in motility, virulence, and biofilm formation in the consortium of K. vulgare and the helper strain. Therefore, investigation of the presence or absence of quorum sensing in the vitamin C coculture fermentation system would shed a light on further understanding of the interactions between the consortium members.

CONCLUSIONS AND PERSPECTIVES

Approximately 90 years ago, Reichstein and his colleagues developed the first commercial method for vitamin C production. Subsequently, the classic two-step fermentation process replaced Reichstein’s method for the large-scale industrial production of vitamin C and has been applied for almost 50 years (48). Although the two-step fermentation system has the weakness of two sterilizations and higher costs compared to one-step fermentation, it has greater 2-KLG production. Therefore, the classic two-step fermentation process is still used in industry and achieves huge economic benefits per year. Notwithstanding that, the rise and development of the one-step fermentation process remains an innovative path in the vitamin C industry. To increase the yield of 2-KLG in one-step fermentation systems, in addition to the construction and optimization of engineered strains, a deeper understanding of the cell-cell communication between the two strains in the second step of the two-step fermentation process also plays a crucial role.

In our opinion, to further understand the reasons for the high production of 2-KLG in the coculture fermentation system and the cell-cell communication between the two strains, several subtle and fascinating questions in the field of two-step fermentation research have to be mentioned here. (i) In K. vulgare, the relationship between the respiratory chain and 2-KLG production remains unclear, and the electron transport pathway of the process is still ambiguous. Based on omics technology and combined with bioinformatics analysis, it is possible to explore the electron transfer process of K. vulgare in different fermentation systems. (ii) In the consortium of K. vulgare and the helper strain, there are research gaps concerning the cell-cell communication and quorum sensing system between the two strains, which are essential for understanding the interactions of the synthetic consortia (as described in this paper). (iii) 2-KLG has been reported to cause oxidative stress and osmotic stress, whereas sporulation formation of the helper strain is a response to environmental stress. In addition, the sporulation and spore ability of the helper strain has been reported to promote the growth and 2-KLG production of K. vulgare, since stable spores of the helper strain are favorable for scavenging oxygen radicals to maintain a suitable environment for the growth of K. vulgare in the coculture fermentation system (46). Then, is there a correlation between the production of spores of the helper strain and the high production of 2-KLG of K. vulgare? In other words, does 2-KLG stimulate the formation of spores of the helper strain? The future investigation of these contents would contribute to a better understanding of the interactions between K. vulgare and the helper strain in the consortium and the reasons for the high production of 2-KLG in the coculture fermentation system. And it would further facilitate innovation in the vitamin C industry and provide insights into the study of more sophisticated synthetic consortia.