New alternative processes for the continuous production of gluconic acid by Aureobasidium pullulans, using biomass retention by cell immobilization or cross over filtration, are described in the present work. 315 g/l gluconic acid was continuously produced in chemostat cultures at 21 hrs residence time without any biomass retention. 260 g/l gluconic acid was produced in fluidized bed reactor at 21 hrs residence time. 375 g/l gluconic acid was produced under continuous cultivation at 22 hrs of residence time with a formation rate for the generic product of 17 g/(l x h) and a specific gluconic acid productivity of only 0.74 g/(g x h), using biomass retention by cross over filtration. Biomass retention makes it possible to break the existing link between growth and residence time.

As a multifunctional carbonic acid with multiple physiological, sensoric and chemical characteristics, belonging to the bulk chemicals gluconic acid itself, the gluconolatone form and its salts (e.g. alkali metal salts, in especially sodium gluconate) have found extensively versatile uses in the chemical, pharmaceutical (e.g. iron and calcium deficiency), food, beverage, textile, construction (cement additive) and other industries (Singh, 1976; Hustede et al. 1989; Anastassiadis et al. 1999; Fenice et al. 2000; Vassilev et al. 2001; Anastassiadis et al. 2003; Rodriguez et al. 2004; Znad et al. 2004; Anastassiadis et al. 2005; Anastassiadis et al. 2006). Numerous methods have been extensively described for the production of gluconic acid, including chemical and electrochemical catalysis, enzymatic biocatalysis in enzyme bioreactor, microbial production using free growing or immobilized cells of either Gluconobacter oxydans or Aspergillus niger.

The immobilization of whole cells or glucose oxidase enzyme by various techniques has often been reported to be a useful approach to the production of gluconic acid or other microbial metabolites. Fluidized bed reactors have been often used for several fermentation applications, in especially for anaerobic processes. Biomass retention by immobilization on porous sinter glass or by cross over filtration was applied in present work in order to accelerate the continuous production of gluconic acid by Aureobasidium pullulans and to maximize the product concentration.

Microorganism

Aureobasidium pullulans (de Bary) Arnaud isolate Nr. 70 (DSM 7085) used in present work which was isolated from wild flowers (Jülich, Germany) (Anastassiadis et al. 1999; Anastassiadis et al. 2003; Anastassiadis et al. 2005) and hold on yeast malt extract agar plates (YME) at 4ºC. The 10% inoculum was prepared by transferring of cells from agar plates into 500 ml shake flasks with baffles on the cultivation medium with 30 g/l glucose.

Culture conditions

Continuous gluconic acid fermentation with (cross over filtration) or without biomass retention (Figure 1) was carried out in a 5 litre chemostat fermenter (Biostat E, Braun-Diessel) at a working volume of 3 l, 1000 rpm, pH 6.5 (45% NaOH), 120-180% (free growing cells) or 290% air saturation(biomass retention; about 80% filtrate and 20% bleed) and 30ºC or in a fluidized bed reactor (cell immobilization on sinter glass) with 0.9 liter working volume and 350 g of porous sinter glass beads with 1-2 mm diameter and 60% porosity (SIRAN^R) (Schott AG, Mainz, Germany) (Figure 2) on a defined cultivation medium containing (g/l): varying glucose concentration (s. results), NH₄Cl 3 g/l, KH₂PO₄ 1.4 g/l, MgSO₄ x 7 H₂O 0.35 g/l, MnSO₄ x 4 H₂O 5 mM, FeSO₄ x 7 H₂O 1 mM, CuSO₄ x 5 H₂O 4 μM (1 mg/l), ZnSO₄ x 7 H₂O 0.01 g/l, CoSO₄ x 7 H₂O 4 mg/l, H₃BO₃ 0.04 g/l, CaCl₂ 0.1 g/l, NaCl 0.1 g/l, citric acid 2.5 g/l, Na₂MoO₄ x 2H₂O 0.2 mg/l, thiamine-HCl 2 mg/l, biotin 0.25 g/l, pyridoxine-HCl 0.625 mg/l, Ca-D-pantothenate 0.625 mg/l, nicotinic acid 0.5 mg/l [24,25,2]. Vitamins and NH₄Cl were added separately to autoclaved medium (30-60 min at 121ºC) by sterile filtration (Sartorius filter, Göttingen, Germany) (Anastassiadis et al. 2005). Periodically, an antifoaming agent was pumped to the fermenter using a time switch clock (15 sec performance every 1.5 hrs). An automatic dosing system (Sartorius, Germany) was used for the constant feeding rate of fermentation medium (Figure 2). Because of the very high relationship between substrate flow rate and recirculation flow rate, the fluidized bed reactor can be considered as a regular stirrer fermenter.

Analysis

Optical density (OD_{660 nm}), dry biomass (filter method), ammonium nitrogen and the concentration of glucose and gluconic acid were determined as has described in previous works (Anastassiadis, 1993; Anastassiadis et al. 1999; Anastassiadis et al. 2002; Anastassiadis et al. 2003; Anastassiadis et al. 2005).

Superior fermentation processes were developed for the continuous and discontinuous production of gluconic acid by isolated strains of Aureobasidium pullulans during an extensive process development program, using defined fermentation media (Anastassiadis et al. 1999; Anastassiadis et al. 2003; Anastassiadis et al. 2005). Reaction technical investigations were finally accomplished, in order to examine the transferability and feasibility of those findings to fermentation process variants (e.g. batches, fed batches, continuous fermentation without and with biomass retention) using Aureobasidium pullulans isolate 70 (DSM 7085).

Continuous production of gluconic acid by free growing cells in stirrer fermenter

More than 220 g/l gluconic acid were produced continuously (Anastassiadis et al. 2005) using 360 g/l glucose and 315 g/l at 21 h residence time with 6.8 g/l biomass and a formation rate for the generic product of 15 g/(l x h) with 450 g/l glucose in feeding medium and at 155% air saturation. Table 1 illustrates the results that have been achieved at residence times of 21 and 25 hrs. For comparison, a maximum specific productivity of about 9.3 g/(g x h) and formation rate of about 18.6 g/(l x h) were reached in a chemostat under optimized conditions without any biomass retention at very short times of about 11.8 hrs. The fermentation continued to stably run without any technical and microbial stability problems for a more than one year (also using a new strain of A. pullulans, isolated at the Pythia Institute of Biotechnology, Greece) without any microbial stability problems, showing again the stability of Aureobasidium process.

Continuous gluconic acid production by immobilized cells

After the initial inoculation of fluidized bed reactor with a two days culture of Aureobasidium pullulans isolate 70 (DSM 7085) the reactor performed under optimum fermentation condition in batch mode until continuous feeding of optimized medium with 450 g/l glucose started at a biomass concentration of more than 3 g/l, in order to increase biomass loading on the porous glass beads. 260 g/l gluconic acid were produced continuously at 21 hrs RT with a formation rate for the generic product gluconic acid of 12.4 g/(l x h). The overgrowing of glass beads and the adhesive behavior of A. pullulans (biopolymer formation, pullulan) resulted in oxygen diffusion problems and consequently in oxygen limitation of immobilized biomass.

Continuous gluconic acid fermentation with biomass retention by microfiltration

An efficient alternative approach for the acceleration of product concentration maximization is the partial retention of biomass by cross over filtration (microfiltration), which makes feasible process performance resulting in higher biomass specific productivities at very high oxygen concentrations, which would be toxic for cell growth. Future studies should examine the possibility of reaching very high gluconic acid concentrations at RT lower than 10 hrs, applying very high oxygen concentrations. Operating under the same fermentation conditions as without any biomass retention, 375 g/l gluconic acid were continuously produced by biomass retention (cross over filtration with about 80% filtrate and 20% bleed) at 22 hrs RT with 23 g/l of accumulated biomass and 370 g/l at 19 hrs with 25 g/l biomass, reaching complete glucose conversion at 30ºC, pH 6.5, 290% air saturation and 450 g/l medium glucose.

Microbial metabolite production of primary, intermediary and secondary metabolism usually takes place under stress conditions, whereas medium composition and other fermentation parameters (e.g. pH, oxygen, CO₂ concentration and temperature) strongly influence metabolite production. The highest formation rate and biomass specific productivity for the generic product gluconic acid have been achieved at lower residence times at about 12 hrs, decreasing continuously at higher RT and gluconic acid concentrations of more than 230 g/l (Anastassiadis et al. 1999; Anastassiadis et al. 2003; Anastassiadis et al. 2005). In a conventional chemostat culture without any biomass retention where growth and production occur simultaneously, the wash out effect of biomass is the major limiting factor in terms of achieving very high product concentrations at very low RT (high dilution rates).

Biomass retention enables the application of production optima in a continuous fermentation process and maximization of product concentration even at very short residence times that would normally wash out the biomass from bioreactor, uncoupling product formation from growth rate and indicating the feasibility of complete glucose conversion at even very short RT, i.e. less than 7 hrs. Continuous fermentation can thus perform at optimum production parameters, which would not necessarily be optimal for the formation of biomass, thus breaking partially the interaction between growth and production and still keeping very high formation rates and product concentrations. Continuous fermentation performing in a fluidized bed reactor with immobilized biomass on a support carrier may be more convenient from a process point of view, however fermentation with non-supported microorganisms performing without or with biomass retention by cross over filtration has been found to be advantageously.

Early fungal gluconic acid fermentation processes employing species of Penicillium and Aspergillus niger are knowingly difficult to handle and unsuitable for a continuous operation using free growing cells. Improved strains of Aspergillus niger (predominantly) or Gluconobacter suboxidans (keto-acid formation) are used in discontinuous submerged fermentations in gluconic acid industry today (Röhr et al. 1983). The use of genetically engineered microorganisms and the immobilization of isolated glucose oxidizing enzymes or whole cells in specialized reactors appear in literature reports as the promising future developments regarding the advanced continuous production of gluconic acid. However, industrial gluconic acid production by Aureobasidium (former Dematium or Pullularia) pullulans, which is well characterized for the production of pullulan, stayed out of question and investigation. To our knowledge, no comparable results reaching those very high product concentrations of 220-433 g/l can be found in the international bibliography, referring to the continuous gluconic acid production by free growing or immobilized cells of Aspergillus and Penicillium strains or other microorganisms. Even 504 g/l of gluconic acid were achieved in fed-batch fermentations using A. pullulans isolate 70 (Anastassiadis et al. 1999; Anastassiadis et al. 2003; Anastassiadis et al. 2005; Anastassiadis et al. 2006; present work). Continuous operation of A. pullulans shows many advantages over the traditional discontinuous fungi and bacterial processes, proving to be a superior industrial gluconic acid producer, integrating as yeast-like fungus the advantages of fungal and bacterial systems at once, enabled to utilize astonishingly very high glucose concentrations of 600 g/l in continuous single or two stage mode.

The fact that a natural wild strain was used without any classical or genetic engineering mutagenesis improvement clearly emphasizes nature's great latent and still unknown potential for further future biotechnological achievements. The development of present superior fermentation processes has accomplished the highest expectations of an expertise researcher in the field of industrial microbiology and biotechnology, extending the frontiers of known microbial capabilities and emphasizing the importance of deep understanding of classical and industrial microbiology in the field of biotechnology science. The novel processes offer new opportunities and are very promising for industrial future applications, lasting in today's high competitiveness in industry and claiming numerous advantages over the traditional fungi or bacteria processes of the last 100 years. The continuous production of the responsible glucose oxidizing enzyme from A. pullulans appears feasible as well. No strain instabilities have been observed so far for an unproblematic longtime operation with a precondition to keep the system safe from contaminations, because some contaminants can overgrow the production strain.

Such data of reaching continuously very high gluconic acid concentrations at very high product molar and mass selectivity (more than 100%, g/g) with and without any biomass retention have not been published in the international literature before, promising the favorable alternative for the industrial production of gluconic acid compared with the discontinuous fungi processes of the last 100 years. Mathematical models would enable the prediction of experimental results at various stream ratios between filtrate and bleed stream and facilitate future research designs and experimental plans.

We thank Prof. Dr. Christian Wandrey (Institute of Biotechnology 2 of Research Center Jülich, RCJ, Germany; formerly known as Nuclear Research Center Jülich, KFA) for his helpful advices and support.

ANASTASSIADIS, Savas. Determination of organic acids, especially citricacid and isocitric acid, in fermentation solutions and fruit juices. In: HPLCApplications, MACHEREY-NAGEL GmbH & Co. KG (Dόren, Germany), 1993, Application no. 8, p. 4.

ANASTASSIADIS, Savas, AIVASIDIS, Alexander and WANDREY, Christian. Process for the production of gluconic acid with a strain of Aureobasidium pullulans (De Bary) Arnaud. US Patent No. 5,962,286, October 5, 1999.

ANASTASSIADIS, Savas; AIVASIDIS, Alexander and WANDREY, Christian. Citric acid production by Candida strains under intracellular nitrogen limitation. Applied Microbiology Biotechnology, October 2002, vol. 60, no. 1-2, p. 81-87. [CrossRef]

ANASTASSIADIS, Savas; AIVASIDIS, Alexander and WANDREY, Christian. Continuous gluconic acid production by isolated yeast-like mould strains of Aureobasidium pullulans. Applied Microbiology and Biotechnology, April 2003, vol. 61, no. 2, p. 110-117. [CrossRef]

ANASTASSIADIS, Savas; AIVASIDIS, Alexander; WANDREY, Christian and REHM, Hans-Jürgen. Process optimization of continuous gluconic acid fermentation by isolated yeast-like strains of Aureobasidium pullulans. Biotechnology and Bioengineering, August 2005, vol. 91, no. 4, p. 494-501. [CrossRef]

ANASTASSIADIS, Savas and REHM, Hans-Jürgen. Continuous gluconic acid production by the yeast-like Aureobasidium pullulans in a cascading operation of two bioreactors. Applied Microbiology and Biotechnology, 2006, in press

FENICE, Massimiliano; SELBMAN, Laura; FEDERICI, Federico and VASSILEV, Nikolay. Application of encapsulated Penicillium variabile P16 in solubilization of rock phosphate. Bioresource Technology, June 2000, vol. 73, no. 2, p. 157-162. [CrossRef]

HUSTEDE, H., HABERSTROH, H.J., and SCHINZIG, E. Gluconic acid. In: ULLMANN`S Encyclopedia of Industrial Chemistry, Weinheim 1989, vol. A 12, p. 449-456.

RODRIGUEZ, Hilda; GONZALEZ, Tania; GOIRE, Isabel and BASHAN, Yoav.Gluconic acid production and phosphate solubilization by the plant growth-promoting bacterium Azospirillum spp. Naturwissenschaften, November 2004, vol. 91, no. 11, p. 552-555. [CrossRef]

RÖHR, M.; KUBICEK, C.P. and KOMINEK, J. Gluconic acid. In: REHM, H.J. and REED, G. eds. Biotechnology, Verlag Chemie, Weinheim-Deerfield, vol. 3, 1983, p. 456-465.

SINGH, N.B. Effect of gluconates on the hydration of cement. Cement and Concrete Research, July 1976, vol. 6, no. 4, p. 455-460. [CrossRef]

VASSILEV, Nikolay; VASSILEVA, Maria; FENICE, Massimiliano and FEDERICI, Federico. Immobilized cell technology applied in solubilization of insoluble inorganic (rock) phosphates and P plant acquisition. Bioresource Technology, September 2001, vol. 79, no. 3, p. 263-271. [CrossRef]

ZNAD, H.; MARKOS, J. and BALES, V. Production of gluconic acid from glucose by Aspergillus niger: growth and non-growth conditions. Process Biochemistry, July 2004, vol. 39, no. 11, p. 1341-1345. [CrossRef]