Carbon sources affect metabolic capacities of Bacillus species for the production of industrial enzymes: theoretical analyses for serine and neutral proteases and α-amylase

https://doi.org/10.1016/S1369-703X(00)00136-4Get rights and content

Abstract

The metabolic fluxes through the central carbon pathways were calculated for the genus Bacillus separately for the enzymes serine alkaline protease (SAP), neutral protease (NP) and α-amylase (AMY) on five carbon sources that have different reduction degrees (γ), to determine the theoretical ultimate limits of the production capacities of Bacillus species and to predict the selective substrate for the media design. Glucose (γ=4.0), acetate (γ=4.0), and the TCA cycle organic-acids succinate (γ=3.5), malate (γ=3.0), and citrate (γ=3.0) were selected for the theoretical analyses and comparisons. A detailed mass flux balance-based general stoichiometric model based on the proposed metabolic reaction network starting with the alternative five carbon sources for the synthesis of each enzyme in Bacillus licheniformis that simulates the behaviour of the metabolic pathways with 107 metabolites and 150 reaction fluxes is developed. Highest and lowest specific cell growth rates (μ) were calculated as 1.142 and 0.766 h−1, respectively, when glucose that has the highest degree of reduction and citrate that has the lowest degree of reduction were used as the carbon sources. Highest and lowest SAP, NP and AMY synthesis rates were also obtained, respectively, when glucose and citrate were used. Metabolic capacity analyses showed that the maximum SAP, NP, and AMY synthesis rates were, respectively, 0.0483, 0.0215 and 0.0191 mmol g−1 DW h−1 when glucose uptake rate was 10 mmol g−1 DW h−1 and specific growth rate was zero. The amino acid compositions and the molecular weights of the enzyme influence the production yield and selectivity. For SAP and NP oxaloacetate and pyruvate, for AMY oxaloacetate appear to be the critical main branch points. Consequently, for SAP and NP syntheses the fluxes towards the alanine group and aspartate group, and for AMY synthesis the flux towards the aspartate group amino acids need to be high. The results encourage the discussion of the potential strategies for improving productions of SAP, NP and AMY.

Introduction

Serine alkaline protease (SAP, EC 3.4.21.14), neutral protease (NP, EC 3.4.24) and α-amylase (AMY, EC 3.2.1.1) produced by Bacillus sp. are the major industrial enzymes with increasing market demands, whereupon, the genus Bacillus is one of the most important group of industrial microorganisms that are, in fact, the microbioreactors within the bioreactors [1]. In the bacilli, substrate and medium design, and bioreactor operation conditions, e.g. oxygen transfer rate and pH, regulate SAP, NP, AMY formations beginning from t=0 h [2], [3]. Most studies that are aimed at improving the production of proteases [4] and α-amylase [5] by Bacillus species have investigated the effects of various carbon sources, e.g. glucose, acetate and citric acid. Nevertheless, analysis of the existing information on the effects of carbon sources is difficult, since many of the investigations performed on extracellular enzyme production that are subject to carbon repression suffer in-depth systematic comparative study in order to design the bioreactor media. Moreover, in production the ability of the bacilli to secrete other enzymes, amino acids and organic-acids as by-products besides the desired product, e.g. SAP, NP or AMY, needs to be regulated in order to fine tune bioreactor performance in relation with the physiology of the microorganism. In this context, knowing the distribution of the metabolic fluxes and further the capacity of the microorganism is helpful to determine the theoretical ultimate limit of the production capacity of the cell for successful selection and development of the industrial microorganism on a selective substrate, and further for successful application of metabolic engineering for increasing the yield and selectivity by predicting the rate limiting step(s) and the changes in the fluxes in the metabolic pathways leading to the desired enzyme synthesis.

Bioreaction network flux analysis that is the novel application of biochemical reaction engineering principles to metabolic reaction pathways achieved first by Aiba and Matsuoka [6], and understood and applied in the last decade of the 20th century, depends on the stoichiometrically-based analysis of the metabolic reactions for determining the metabolic fluxes in the cells [7], [8]. This analysis can be used to find the critical branch points and bottlenecks in the overall flux distributions, for modifying the medium composition, for improving the bioreactor operation conditions, moreover, for calculating the theoretical metabolic capacities of the microorganism and for selecting the host microorganism. Theoretical metabolic capacity (TC) analysis is used to determine the maximum theoretical yield of a product on a substrate together with the optimum metabolic flux distributions in order to predict the ultimate theoretical yield, selectivity and productivity that can be performed by the microorganism. Nevertheless, maximisation of the carbon yield of a product relative to the carbon consumed by the cell, and consequently minimisation of the by-product formations can be achieved in the first step by influencing the cellular metabolism by designing the bioreactor operation conditions, and further by altering the regulatory controls of the microorganism by using metabolic engineering techniques. The TC analysis has been successfully applied to a number of industrially important microorganisms and products [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24].

As the Bacillus species secrete a large number of extracellular enzymes and produce purine nucleosides, antibiotics, insecticides, vitamins, and are the potential hosts in the fermentation industry for the production of recombinant proteins, attracted the application of the metabolic flux analysis [1], [15], [18], [19], [20], [21], [22], [23], [24]. Çalık and Özdamar [19] and Çalık et al. [1] recently reported the application of metabolic flux analysis for an enzyme for the first time. The authors [19] reported the biochemistry of B. licheniformis and developed a detailed mass flux balance-based stoichiometric model that contains 147 reaction fluxes and 105 metabolites to obtain the intracellular metabolic flux distributions for SAP production; and they used the model for SAP production that minimises the SAP accumulation in the cell using elaborate experimental data in order to explain the metabolic behaviour of the cells throughout the batch bioprocess. The effects of oxygen transfer conditions on metabolic flux distributions and the potential metabolic bottlenecks in SAP synthesis in B. licheniformis were investigated by Çalık et al. [1]; and reported that the flux partitioning in the TCA cycle at α-ketoglutarate (α-KG) towards glutamate group and at oxalacetate (OA) towards aspartate group amino acids were dependent on the oxygen transfer rate. Çalık et al. [18] reported maximum SAP theoretical yields and optimum carbon flux distributions for SAP overproduction in B. licheniformis by using citric acid as the sole carbon source and then applied the model to the SAP fermentation by using experimental data obtained during the batch bioprocess in order to explain the performance of the bacilli. And lastly, Çalık et al. [20] reported the effects of oxygen transfer on the metabolic capacity of B. licheniformis.

The literature related with the metabolic flux analysis shows that there is not any individual published work concerning the effects of carbon sources on any enzyme or protein production with any Bacillus sp. or even with any microorganism. As the genus Bacillus has the potential to produce SAP, NP and AMY in relation to its genetic structure and regulation in response to its designed environment, in the present work, we focused on the effects of carbon sources that have different reduction degrees (γ) on the metabolic capacity of the bacilli for the production of the three enzymes. The carbon sources glucose (γ=4.0), acetate (γ=4.0), and the TCA cycle organic-acids succinate (γ=3.5), malate (γ=3.0), and citrate (γ=3.0) were selected for the theoretical analyses and comparisons, despite the fact that two of the TCA cycle organic-acids, i.e. succinate and malate, are not economic substrates for the bioprocesses. Thus, the theoretical intracellular bioreaction network flux distributions in B. licheniformis in response to each of the carbon sources for the overproduction of each of the three industrial enzymes are calculated to evaluate the effects of carbon sources on the metabolism of the bacilli, and, moreover, to predict the adequate substrate(s) for the media design. Further, depending on the optimised metabolic flux distributions, the potential metabolic bottlenecks in the synthesis of each enzyme are predicted, finally, the general conceptual framework to develop the bioprocesses for SAP, NP and AMY productions is presented.

Section snippets

Mass flux balance based theoretical capacity analysis

The metabolic reaction network of B. licheniformis [19] that contains 105 metabolites and 147 intracellular reactions is used by the addition of the following reactions:

Anaplerotic reactionR148:Pyr+CO2OAcatalysed by pyruvate carboxylase is included to the model for the connection of the glycolysis pathway to the TCA cycle.

Neutral protease synthesis reaction [25]R149:47Ala+14Arg+23Asn+30Asp+24Gln+40Glu+48Gly+18His+32Ile+33Leu+37Lys+8Met+20Phe+17Pro+43Ser+35Thr+4Trp+32Tyr+32Val+5.5ATPNP+5.5ADP

Theoretical metabolic flux distributions for maximum biomass generation: cell growth phase

Metabolic flux distributions of the intracellular reaction network were determined by theoretical metabolic capacity (TC) analysis by using the reaction network model that was set up for B. licheniformis for the cell growth phase by maximising the biomass production (R145) in the objective function Eq. (3) of the generalised model for each of the five substrates, i.e. glucose, acetate, succinate, malate, and citrate. In the calculations, for the direct comparisons, all the biomass flux values

Discussion and conclusions

Since the synthesis of an enzyme and by-product formations in a microorganism in the bioreactor are dependent on the substrate, medium design, and bioreactor operation conditions, an in-depth insight was provided by applying the theoretical metabolic capacity (TC) analysis to determine the theoretical ultimate limits of the production capacities of the bacilli for the enzymes SAP, NP and AMY on the five carbon sources, i.e. glucose, acetate, succinate, malate and citrate, in order to predict

Acknowledgements

This work was supported by the State Planning Organisation (SPO, Turkey) Grant 97K120590. Professor Güzide Çalık (Ankara University) is gratefully acknowledged for her valuable comments and for reading the manuscript.

References (41)

  • P. Çalık et al.

    Metabolic flux analysis for serine alkaline protease fermentation by Bacillus licheniformis in a defined medium: effects of oxygen transfer rate

    Biotechnol. Bioeng.

    (1999)
  • P. Çalık et al.

    Oxygen transfer strategy and its regulation effects in serine alkaline protease production by Bacillus licheniformis

    Biotechnol. Bioeng.

    (1999)
  • P. Çalık, G. Çalık, T.H. Özdamar, Bioprocess development for serine alkaline protease production, Rev. Chem. Eng.,...
  • Y.S. Park et al.

    Enhanced amylase production in recombinant Bacillus brevis by fed-batch culture with amino acid control

    Biotechnol. Bioeng.

    (1996)
  • S. Aiba et al.

    Identification of metabolic model: citrate production from glucose by Candida lipolytica

    Biotechnol. Bioeng.

    (1979)
  • G. Stephanopoulos et al.

    Network rigidity and metabolic engineering in metabolite overproduction

    Science

    (1991)
  • J. Nielsen, J. Villadsen, Bioreaction Engineering Principles, Plenium, New York,...
  • A. Varma et al.

    Biochemical production capabilities of Escherichia coli

    Biotechnol. Bioeng.

    (1993)
  • A. Varma et al.

    Parametric sensitivity of stoichiometric flux balance models applied to wild-type Escherichia coli metabolism

    Biotechnol. Bioeng.

    (1995)
  • W.M. van Gulik et al.

    A metabolic network stoichiometery analysis of microbial growth and product formation

    Biotechnol. Bioeng.

    (1995)
  • Cited by (45)

    • A comparative study on the production of amidase using immobilized and dehydrated immobilized cells of Pseudomonas putida MTCC 6809

      2012, Journal of Genetic Engineering and Biotechnology
      Citation Excerpt :

      with 1% (w/v) alginate at different time intervals. The production capacity of organism depends on successful selection of growth conditions and substrate [4]. The immobilized catalyst may provide good mechanical strength, low protein leachability, and high retention of amidase activity.

    • Extracellular cold-active lipase of Microbacterium luteolum isolated from Gangotri glacier, western Himalaya: Isolation, partial purification and characterization

      2012, Journal of Genetic Engineering and Biotechnology
      Citation Excerpt :

      The aim of the investigation was to isolate potential lipolytic bacteria with novel properties and to improve the production. The production capacity of organism depends on successful selection of growth conditions and substrate [4]. Lipase production by psychrotrophs varies with species, as does the optimum temperature, optimum pH and enzyme specificity [42].

    View all citing articles on Scopus
    View full text