Co-cultivation of Lactobacillus zeae and Veillonella criceti for the production of propionic acid
© Dietz et al.; licensee Springer. 2013
Received: 16 March 2013
Accepted: 20 March 2013
Published: 24 May 2013
In this work a defined co-culture of the lactic acid bacterium Lactobacillus zeae and the propionate producer Veillonella criceti has been studied in continuous stirred tank reactor (CSTR) and in a dialysis membrane reactor. It is the first time that this reactor type is used for a defined co-culture fermentation. This reactor allows high mixing rates and working with high cell densities, making it ideal for co-culture investigations. In CSTR experiments the co-culture showed over a broad concentration range an almost linear correlation in consumption and production rates to the supply with complex nutrients. In CSTR and dialysis cultures a strong growth stimulation of L. zeae by V. criceti was shown. In dialysis cultures very high propionate production rates (0.61 g L-1h-1) with final titers up to 28 g L-1 have been realized. This reactor allows an individual, intracellular investigation of the co-culture partners by omic-technologies to provide a better understanding of microbial communities.
Today most industrial processes are mono-culture processes due to a high degree of control, reproducibility and predictability. But mixed cultures arouse more and more interest. The high potential of mixed culture fermentations for industrial applications has been recently reviewed (Bader et al. (2010), Sabra et al. (2010)). The authors stress the advantages of mixed cultures beside others in respect of wide substrate and product spectrum. These advantages may be applied e.g. in the fields of food and bioenergy.
Besides their technological relevance microbial communities have been investigated over the past decades due to their biological relevance in nature. Usually no mono-culture will occur in natural environments. In fact there exist complex microbial communities as biofilms that play an important role e.g. in human health (Bryers (2008)).
One approach in understanding such complex communities is to investigate a defined part of this community, i.e. a defined co-culture of two organisms. In fact there exist a lot of studies with defined co-cultures including two microorganisms (e.g. Gerritse et al. (1990), Mikx and Vanderhoeven (1975), Tatton et al. (1989)). To have reproducible results for metabolic analysis experiments in a continuous stirred tank reactor (CSTR) are preferred. Usually substrate-limited conditions have to be investigated to prevent a wash-out of one organism, beside that a general problem of these studies is that no individualized study of both organisms is possible. To analyse the growth behaviour in defined co-cultures in a single reactor is possible with sophisticated molecular-biology methods (e.g. Schmidt et al. (2007)), but an individualization in different compartments would be preferable to investigate also intracellular processes.
To solve the latter problem, membrane-associated separation of the organisms in one study has been realized (Egland et al. (2004)). In fact it was shown that bacteria do not only communicate by direct cell-cell-contact but also via low molecular metabolites that may pass through dialysis membranes (Egland et al. (2004), Kolenbrander et al. (2010)). One approach to investigate a co-culture with separating the cells by a membrane is involving two reactors connected with each other by a membrane module. Manjarrez et al. (2000) used a hollow fiber module to investigate the amensalistic-type interaction between two Saccharomyces strains. They stress the superiority of this type of system over the EcoloGen system of New Brunswick Scientific (Edison, NJ, USA) described in Tannenbaum and Kornfeld (1974). The advantage is that in their hollow-fiber module system high mixing rates can be established that are important for fast interactions. But a problem for hollow fiber modules comes with their tendency to get blocked at higher cell densities.
The dialysis membrane reactor (Poertner and Maerkl (1998)) solves both problems: High mixing rates and high cell densities can be reached. In fact, up to now it was mostly used for high cell density fermentations (e.g. Markl et al. (1993)). The use and advantages of such a system for the investigation of a defined co-culture have been described (Pestchanker and Ercoli (1997)), but to our best knowledge not realized up to now.
A mixed culture approach with industrial applicability is propionic acid production. Lactobacillus zeae and Veillonella criceti have been described as a defined co-culture for propionate production with a high potential in the industrial environment (Mays and Fornili (1985), Sabra et al. (2012)).
In this work we use the dialysis membrane reactor for a defined co-culture for the first time with the example of the aforementioned co-culture. L. zeae will convert glucose to lactate which is the substrate of V. criceti, producing propionate and acetate. V. criceti is not able to use carbohydrates as a substrate, this leads to a commensalistic co-culture. We show the high potential for metabolic interaction studies with the membrane reactor and discuss about the possibility of the defined co-culture for propionic acid production in a CSTR process. So far these studies have mostly been performed with Propionibacterium species. It was a goal to find out cultivation conditions for this coculture to reach a high propionate concentration and to evaluate its feasibility for an industrial application.
Materials and methods
Strains and media
Veillonella criceti (DSM20734) and Lactobacillus zeae (DSM20178) were obtained from DSMZ (German Collection of Microorganisms and Cell Cultures, Braunschweig). The media for anaerobic precultivation contained KH2PO4 15 g L-1, K2HPO4 5 g L-1, Cystein.HCl × H2O 0,5 g L-1, peptone 2 g L-1, yeast extract 2 g L-1. For V. criceti pottasium lactate 10 g L-1 and resazurin 1 g L-1 were added. Anaerobiosis was achieved by nitrogen sparging and sealing with butyl-rubber stoppers of the serum bottles (50 mL medium in 100 mL bottles). After autoclaving, Na2S × 9 H2O 1 mM was added for V. criceti, for L. zeae glucose 10 g L-1 was added. Precultivation and fermentation were performed at 37°C, pH 6.0.
Fermentation medium was basically of the same composition, but had varying amounts of complex nutrients and carbon source as mentioned in the text. Yeast extract and peptone were added simultaneously in equal amounts to result in a certain concentration (g L-1).
For CSTR fermentations, a 2l foil fermentor from Bioengineering (Wald, Switzerland) was used which is equipped with temperature and pH control. The fermentor was sparged before inoculation with nitrogen for at least 30 min. For titration, a 5 M NaOH solution was used.
Properties of the Cuprophan-membrane (Poertner and Maerkl ( 1998 ))
Type of membrane
Non-porous diffusion membrane
Thickness (μ m)
Area (cm 2 )
Retention capacity (kDa)
(mL water min -1 cm -2 bar -1 )
Permeability coefficient for
glucose a (dm h -1 )
Optical density of the cells was measured at 600 nm. Samples were filtered through 0.2 μ m filters. Glucose, lactate, propionate and acetate were measured by HPLC (Kontron, Germany) with an Aminex HPX-87H column (300 × 7.8 mm) at 60°C and with UV and RI detectors. H2SO4 (5 mM) was used as the mobile phase.
First coculture experiments in a common stirred tank reactor were performed. Experiments started with a low concentration of complex nutrients, carbon source and a low dilution rate. Successively the glucose concentration was increased to reach higher product concentrations. Since nutrient limitations could be expected for these organisms, complex nutrients should be adapted to improve growth of one or both bacteria. Finally, by raising the dilution rate, the productivity of the process should be enhanced.
Cultivated with 0.5 g L -1 yeast extract/peptone, 5 g L -1 glucose in the feed and a dilution rate of D = 0.1 h -1 V. criceti, but not L. zeae, was washed out from the bioreactor. Neither acetic nor propionic acid was produced. Obviously complex nutrients are strongly limited. L. zeae did grow, but the final lactic acid concentration was very low (0.3 g L -1 ), glucose was not consumed completely. This shows the high nutrient demand from both bacteria with an advantage to survive for L. zeae under these cultivation conditions.
After this initial experiment the yeast extract/peptone (YP)-concentration was increased to 3 g L -1 , the lactic acid concentration increased due to production by L. zeae to 3 g 3 g L -1 , but still glucose was not completely consumed. To ensure anaerobic conditions by reducing the redox potential (reactor was not sparged with nitrogen), cystein was added, but this did not influence biomass and lactate production.
This changed clearly after a second inoculation of V. criceti and led to steady state (ST) 1 (only coculture steady states are considered). Now the concentration of complex nutrients was high enough for the cultivation of both organisms simultaneously. Glucose and consequently produced lactic acid were converted to propionic acid. The propionate/acetate-ratio was 1.29 mol mol -1 .
At the half residence time (5 h, ST 3) glucose consumption was incomplete, but the produced lactic acid was completely converted in acetic and propionic acid. The reason for the incomplete glucose consumption was a nutrient limitation, as can be seen after doubling the YP-concentration in the feed (ST 4). The glucose concentration was reduced from 6 to 2 g L -1 by consumption of L. zeae. Lactic acid again was completely consumed.
The glucose was entirely converted to lactic acid at a YP-concentration of 9 g L -1 (ST 5). The propionate concentration rose to 8 g L -1 , the propionate/acetate-ratio was 1.21 mol mol -1 .
Results of CSTR experiments with a defined co-culture of L. zeae und V. criceti under lactate limitation
g L -1
mmol L -1
mmol L -1
mmol L -1
mmol L -1
mmol L -1
mmol L -1
mol mol -1
Experiments in dialysis chamber reactor
L. zeae monoculture
V. criceti monoculture
Results of cultivations of L. zeae and V. criceti in the dialysis reaktor
g L -1
g L -1 h -1 a
mol mol -1
mol mol -1
g h -1
3g L -1 Yeast extract/Peptone
10 g L -1 Yeast extract/Peptone
Experiments with the defined co-culture
P: product; X: biomass. α: parameter for growth-associated production. β: parameter for non-growth associated production.
Improvement of dialysis chamber experiments
In the experiments above neither glucose nor lactic acid have been completely consumed. To improve the co-culture experiments in this respect it was therefore decided to increase the concentration of complex nutrients was increased to 10 g L -1 .
Again lactic acid was not completely consumed. This was surprising since a nutrient limitation was implausible due to the high YP-concentration.
Preeminent result is the very high propionate production rate of 0.61 g L -1 h -1 . Between the 13th and 33rd hour 20 g L -1 propionic acid has been produced in the inner chamber.
In this experiment not only the propionate production rate, but the final biomass and propionate/acetate ratio was higher than in the monoculture experiment.
Since the propionate/acetate-ratio was in no experiment 2 mol mol -1 , the maximal propionate/glucose-yield could not be reached. Interestingly both ratios were not directly correlated. Obviously growth and substrate consumption in L. zeae depends on the complex nutrient availability which is influenced by the co-culture partner.
In this study a defined co-culture of a lactic acid bacterium, L. zeae, and a propionate producer, V. criceti, has been realized and investigated for the first time in a dialysis chamber reactor.
In CSTR experiments an almost linear correlation of production rate and complex nutrient supply was shown (Figure 3). This may be used for further improvement of the process. Perhaps the use of very high concentrations of complex nutrients may lead to a high propionate concentration in CSTR experiments. From the industrial point of view then the economic aspect would become less attractive. For high titers batch or fed-batch processes, perhaps with cell immobilization seem much more favourable to this end (Colomban et al. (1993)).
In the dialysis cultures with high nutrient supply very high propionate production rates (0.61 g L -1 h -1 ) with final titers up to 28 g L -1 have been realized.
In Veillonella parvula this pathway is less effectiv than in Propionibacterium freudenreichii. V. parvula gets 0.33 mol ATP per mol lactate, P. freudenreichii 0.78 mol ATP (Seeliger et al. (2002)).
This less effective energy recovery compensates Veillonella spec. by higher growth and substrate consumption rates (Seeliger et al. (2002)).
The energy yield per mol lactate is then comparable in both organisms. When the nutrients are limited, the hydrogen production rises as can be seen from both the CSTR and the dialysis reactor experiments. Although we did not measure hydrogen directly we could not detect any other metabolite by HPLC. Hydrogen production explains well our results in agreement with similar investigations (Denger and Schink (1992); Seeliger et al. (2002)).
Both organisms have a high nutrient demand. In the experiment with a low nutrient supply, the faster growing L. zeae outcompetes V. criceti for the complex nutrients. V. criceti has a high demand for complex nutrients as vitamins, amino acids and even nucleobases (Durant et al. (1997)). For some of those nutrients V. criceti has obviously a higher demand than L. zeae, since L. zeae can grow well on hydrolyzed wheat straw, but V. criceti cannot (Sabra et al. (2012)).
The metabolism of L. zeae is markedly stimulated by the co-culture partner independent of nutrient limitations and experimental setup. This behaviour could be investigated in more detail in the dialysis chamber reactor due to the local separation of both microorganisms. A reason for the strong improvement of the growth of L. zeae may be, that V. criceti reduces the redox potential of the medium by hydrogen production. The reduced redox potential may have a strong impact on microorganisms (van Hoek and Merks (2012)).
Surprisingly V. criceti stops growth in the co-culture experiment with high nutrient supply when it is cultured in the outer chamber - although carbon and most probably complex nutrients are available. We suggest the accumulation of an inhibiting secondary metabolite, that develops its inhibiting potential not until a critical concentration is reached, probably comparable to quorum sensing-signals. This interesting phenomenon should be investigated in future studies in more details. The dialysis chamber reactor is ideal for the investigation of such phenomena which are based on low molecular weight “communication molecules”. The communication molecule may become only relevant at higher cell densities.
Investigations in the dialysis reactor in CSTR will give more information. Still the experimental requirements, especially for medium (50 L for one steady state assuming 10 medium exchanges as shown in the continuous culture), are very high and we now undergo efforts to realize these experiments.
The overall strength of this system for co-culture investigations is the local separation of microorganisms. This allows not only the investigation of growth behaviour as performed in this study, but also a detailed observation of a singularised mircroorganism by genomic, proteomic and transcriptomic approaches.
With these data, mathematical models can be established to describe the kinetics of cell growth and metabolism of the defined microbial community. For a more fundamental understanding of the microbial community, intracellular metabolic fluxes should be estimated. Metabolic fluxes of microbial communities have been seldom studied, therefore the existing methods for flux estimation need to be adapted and further developed. The results from kinetic and flux analysis should help to identify possible limiting step(s) and key parameters for the development and optimization of the novel bioprocesses for propionic acid production.
- Bader J, Mast-Gerlach E, Popović M, Bajpai R, Stahl U: Relevance of microbial coculture fermentations in biotechnology. J Appl Microbiol 2010,109(2):371–387. 10.1111/j.1365-2672.2009.04659.xPubMedView ArticleGoogle Scholar
- Bryers JD: Medical biofilms. Biotechnol Bioeng 2008,100(1):1–18. 10.1002/bit.21838PubMed CentralPubMedView ArticleGoogle Scholar
- Colomban A, Roger L, Boyaval P: Production of propionic-acid from whey permeate by sequential fermentation, ultrafiltration, and cell recycling. Biotechnol Bioeng 1993,42(9):1091–1098. 10.1002/bit.260420911PubMedView ArticleGoogle Scholar
- Denger K, Schink B: Energy-conservation by succinate decarboxylation in Veillonella parvula. J Gen Microbiol 1992, 138: 967–971. 10.1099/00221287-138-5-967PubMedView ArticleGoogle Scholar
- Durant JA, Nisbet DJ, Ricke SC: Comparison of batch culture growth and fermentation of a poultry Veillonella isolate and selected Veillonella species grown in a defined medium. Anaerobe 1997,3(6):391–397. 10.1006/anae.1997.0129PubMedView ArticleGoogle Scholar
- Egland PG, Palmer RJ, Kolenbrander PE: Interspecies communication in Streptococcus gordonii-Veillonella atypica biofilms Signaling in flow conditions requires juxtaposition. Proc Natl, Acad Sci, U S A 2004,101(48):16,917–16,922.View ArticleGoogle Scholar
- Gerritse J, Schut F, Gottschal JC: Mixed chemostat cultures of obligately aerobic and fermentative or methanogenic bacteria grown under oxygen-limited conditions. FEMS Microbiol Lett 1990,66(1–3):87–89.View ArticleGoogle Scholar
- van Hoek MJ, Merks RM: Redox balance is key to explaining full vs partial switching to low-yield metabolism. BMC Syst Biol 2012., 6: Google Scholar
- Kolenbrander PE, Palmer RJ, Periasamy S, Jakubovics NS: Oral multispecies biofilm development and the key role of cell-cell distance. Nature Rev Microbiol 2010,8(7):471–480. 10.1038/nrmicro2381View ArticleGoogle Scholar
- Luedeking R, Piret EL: A kinetic Study of the Lactic Acid Fermentation - Batch Process at controlled pH. J Biochem Microbiol Technol Eng 1959,1(4):393–412. 10.1002/jbmte.390010406View ArticleGoogle Scholar
- Manjarrez ES, Albasi C, Riba JP: A two-reservoir, hollow-fiber bioreactor for the study of mixed-population dynamics: Design aspects and validation of the approach. Biotechnol Bioeng 2000,69(4):401–408. 10.1002/1097-0290(20000820)69:4<401::AID-BIT6>3.0.CO;2-3PubMedView ArticleGoogle Scholar
- Markl H, Zenneck C, Dubach A, Ogbonna J: Cultivation of Escherichia-coli to high cell densities in a dialysis reactor. Appl Microbiol, Biotechnol 1993,39(1):48–52.View ArticleGoogle Scholar
- Mays TD, Fornili P: Microbial co-culture production of propionic acid. Patent 1985. WO 85/04901Google Scholar
- Mikx FH, Vanderhoeven JS: Symbiosis of Streptococcus mutans and Veillonella alcalescens in mixed chemostat cultures. Arch Oral, Biol 1975,20(7):407–410. 10.1016/0003-9969(75)90224-1View ArticleGoogle Scholar
- Pestchanker LJ, Ercoli EC: A novel membrane reactor design for controlled studies of interacting populations (simulation of the interaction between microorganism and plant suspension cultures). Biotechnol Bioeng 1997,55(4):609–615. 10.1002/(SICI)1097-0290(19970820)55:4<609::AID-BIT3>3.0.CO;2-LPubMedView ArticleGoogle Scholar
- Poertner R, Maerkl H: Dialysis cultures. Appl Microbiol, Biotechnol 1998, 50: 403–414. 10.1007/s002530051312View ArticleGoogle Scholar
- Sabra W, Dietz D, Tjahjasari D, Zeng AP: Biosystems analysis and engineering of microbial consortia for industrial biotechnology. Eng Life, Sci 2010,10(5):407–421. 10.1002/elsc.201000111View ArticleGoogle Scholar
- Sabra W, Dietz D, Zeng AP: Substrate limited co-culture for efficient production of propionic acid from flour hydrolysate. Biotechnol Bioeng Submitted 2012.Google Scholar
- Schmidt JK, Konig B, Reichl U: Characterization of a three bacteria mixed culture in a chemostat: Evaluation and application of a quantitative terminal-restriction fragment length polymorphism (T-RFLP) analysis for absolute and species specific cell enumeration. Biotechnol Bioeng 2007,96(4):738–756. 10.1002/bit.21147PubMedView ArticleGoogle Scholar
- Seeliger S, Janssen PH, Schink B: Energetics and kinetics of lactate fermentation to acetate and propionate via methylmalonyl-CoA or acrylyl-CoA. FEMS Microbiol Lett 2002,211(1):65–70. 10.1111/j.1574-6968.2002.tb11204.xPubMedView ArticleGoogle Scholar
- Tannenbaum M, Kornfeld JM: Multiple Diffusion Chamber. 1974.Google Scholar
- Tatton MJ, Archer DB, Powell GE, Parker ML: Methanogenesis from ethanol by defined mixed continuous cultures. Appl Environ Microbiol 55 1989, 2: 440–445.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License(http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.