Citric acid is a biotechnological commodity. It is required mainly for the food industry where is appreciated as natural acidulant, taste enhancer and chelating agent. Its annual production is around of one million tons which are mostly obtained by fermentation with the filamentous fungus Aspergillus niger. In part, the history of citric acid production is parallel to the history of the biotechnology development in the last century, either regarding to the technological aspects (submerged fermentation) or the biological aspects (biochemistry, physiology, genetics of microorganism).

As in other commodities, the process improvements are oriented to obtain higher rates of production rather than best yields. Technologically this goal is restricted mainly to the process operation optimization and any potential gain will be economically significative even when the relative improvement were low. However, best results could be obtained if optimization were focussed on metabolic microorganism capacities. Since the performance of microorganisms are near their theoretical maximum, a redesign of their metabolism would be required. This is feasible but not an easy task. In the last 50 years the strain improvement was based in the assay and error approach through random mutagenesis and screening. The penicillin production can be taken as paradigm were many years were required to achieve the today's productivities. Moreover, the reason of the observed improvement were usually not well understood, the process becoming unstable and poorly controlled.

Nevertheless, this first "ancient" procedure of classical genetics has paved the way to the new methods of strain improvement, based in DNA recombinant technology, such as the directed mutagenesis and cloning. Today, our capacity to introduce changes in DNA information is quite ample and it is not the limiting factor in our capacity to redirect cellular activities. A verification of this assertion would be the entire determination of the genome of several species. Thus the question can should be posed in different terms: what it should be genetically modified to obtain the overproducing strains?

The answer to these questions can be found in a new approach, the cell metabolic engineering. This discipline, is not limited to an specific microorganism or product, but its of general application either for vegetables as for animal cells, bacteria, yeast or fungus. its applications ranges from the overproduction of interferon or ethanol. Which is the paradigm of this approach?

XX Century life science was characterized but a lack of the holistic, integrative vision. Most of the research has been done with the aim of characterize and solve a closely defined problem but isolated of their natural context. Although effective in many instances as a first approximation to tackle complex problems, it fails to capture the system's behavior when dealing with living beings: it happens that the whole is much more than the addition of its isolated parts. This philosophical and methodological approach is illustrated in Figure 1.

It is clear that besides of the classical cycle of the scientific activity of inference-measurements it is possible today to extend it by including computer simulations and mathematical modeling. This allows us to organize the available information; to test its inner consistency and interactions; to explore new relations and made inferences to be experimentally verified. Modeling is a iterative task. The new conclusions feed new hypothesis, that lead to new experiments, that again feed the model and so on.

In the case of citric acid a lot of basic information is available (Kubicek et al. 1994; Kristiansen et al.1998; Röhr, 1998). Efforts have been done in order to integrate the core of this knowledge in highly structured dynamical models (Torres et al. 1995; Torres et al. 1998; Alvarez-Vazquez et al. 2000). However, no recent attempts are reported about to gain insight trough macroscopic and energetic modeling strategies. Based in metabolic flux analysis -a formalism derived from macroscopic approach- we have developed a mathematical model of the process (Figure 1 and Figure 2). This model, was aimed not only to the better understanding and description of the process but also to help in the design of the best genetic strategies leading to the optimization of citric acid production rate. The macroscopic approach is not based in detailed biochemistry knowledge, but based in basic physico-chemical laws, as the energy (entalphy, Gibss energy), electric charge (reductance grade), and material (mass, chemical identity) conservation principles. An important hypothesis developed in the present model is the existence of a close energetic coupling between the citric acid production and the intracellular pH regulation. This is due to the evidence of the strong acidic conditions that are required for A. niger to produce citric acid (extracelullar pH=2). Accordingly, we have included the quantitative description of the proton motive force operating among the cellular compartments as well as their interplay with the H⁺-antiport and H⁺-symport transport mechanisms. Other metabolic processes such as polyols and amino acids excretion are also analyzed in this physiological context.

From the metabolic analysis of A. niger metabolism at 120 hours idiophase stage, when in citric acid accumulating conditions the following conclusion were reached:

• A. niger idiophase, contrary to the current understanding, do not correspond to a unique physiological state, and therefore to a unique metabolic steady state. Therefore, previous analysis unaware of this fact could be affected by inaccuracies.

• All the metabolic process analyzed, even those not previously measured, could be quantified and validated against macroscopic data and bio-energetic available constraints without assuming any detailed biochemical mechanism. The Hexose Monophosphate Pathway (HMP) accounts for 16% of the glucose input (carbon basis), the Krebs cycle for 13% and the citric acid synthesis for the remaining 71%. The GABA shunt and the NH₄⁺/NH₃ cycle showed no significance at all.

• The established flux distribution (Figure 2) implies the existence of an operative glycerol-P shuttle that recycles 93% of the cytosolic glycerol-P to cytosolic dihidroxyacetone phosphate (DHAP), thus coupling the transformation of cytosolic NADH to mitochondrial FADH₂. This shuttle would be critical to keep a close cellular balance among the reductive and the bio-energetic (ATP) cellular status. According with our model, the citric acid synthesis reaction alone produces up to 43% of the total NADH but consuming only 12% of the total ATP. This explain why a 70% of respiration in A. niger occurs by the alternative respiratory system (not associated to ATP generation) and only the remain 30% is used for respiration through the normal system. This also explain why A. niger must excrete polyols. Consequently, a cellular maintenance energy of 3.7 mmol ATP/g_·h was determined. It would be spent in fueling cytoplasmatic (1.4 mmol H⁺/g·h) and mitochondrial (1.8 mmol H⁺/g·h) H⁺-ATPase pumps with efficiencies of 0.65 and 1.2 mmol H⁺/mmol ATP respectively. These value are supported by the pH and potential membranes gradients observed among cellular compartments (Figure 3).

• Practically the most important conclusions refer to the identification of the best target for genetic manipulation in order to achieve higher citric acid productivity. The potential targets were ranked according with its potential for citric acid rate production improvement. The most promising was the glucose carrier. According to our estimations, a gain ratio of 47% in citric acid productivity could be obtained by doubling the glucose influx, which is a feasible manipulation. Also the model explains why some cloning experiences executed were not successful to obtain overproducing strains. Some controversy remains about if a glucose carrier actually exist, since recently it was speculated that glucose influx occurs by diffusion.

Bailey, 1998. Mathematical modeling and analysis in Biochemical engineering: past accomplishments and future opportunities. Biotechnology Progress 14:8-20.

Schilling, Ch. H., Edwards, J.E. and Palsson, B.O. 1999. Toward metabolic phenomics: Analysis of genomic data using flux balances. Biotechnology Progress 15:288-295

Alvarez-Vazquez, F., González-Alcón, C. and Torres, N.V. (2000). Metabolism of citric acid production by Aspergillus niger: Model definition, steady state analysis and constrained optimization of the citric acid production rate. Biotechnology and Bioengineering 70:82-108.

Kristiansen, B., Mattey, M. and Linden, J. (1998). Citric acid biotechnology. Kristiansen, B., Mattey, M. and Linden, J. eds. Taylor and Francis, London, UK. 250 p.

Kubicek, C.P, Witteveen, C.F.B. and Visser, J. (1994). Regulation of organic acid production by Aspergilli. In: Powell, K.A. ed. The Genus Aspergillus. Plenum Press. New York. pp. 35-145.

Röhr, M. (1998). A century of citric acid fermentation and research. Food Technology and Biotechnology 36:163-171.

Torres, N.V., Voit, E.O., González-Alcón, C. and Rodríguez, F. (1998). A novel approach for design an overexpression strategy for metabolic engineering. Application to the carbohydrate metabolism in the citric acid producing mould Aspergillus niger. Food Technology and Biotechnology 36:177-1184.

Torres, N.V., Voit, E.O. and González-Alcón, C. (1995). Optimization of nonlinear biotechnological process with linear programming: application to citric acid production by Aspergillus niger. Biotechnology and Bioengineering 49:247-258.