1. Introduction
L-lysine is the third most abundantly produced amino acid at the industrial scale. Eighty per cent is produced by fermentation and 20% by chemical synthesis. In the fermentative processes, the culture medium is the major financial input and N-source is usually the most expensive component. Potential N-sources include various inorganic (ammonium or nitrate salt) or organic compounds (yeast extract, corn steep liquor, soybean protein hydrolysate), and various other extracts of vegetal and animal tissues (Bisaria et al., 1997). By-products of seafood have a high potential as cheap N-sources for microbial growth (Vecht-Lifshitz et al., 1990). Silage produced from fish wastes (whole heads, fins, skeletal debris, shrimp by-catch captures, and under-utilised whole species of fish) results in a stable product containing peptides and amino acids with high nutritional value (Raa and Gildberg, 1982). The use of fish silage instead of yeast extract has been reported for L-lysine microbial production (Coello et al., 2000). In the last decades, statistical experimental methods have been applied to media optimisation for industrial purpose. These designs include the blocking and factorial experiments for determining the path of steepest ascent, in order to identify the effect of individual factors and to approach the neighbourhood of the optimum response (Box et al., 1978). Furthermore, response surface methodology (RSM) is suitable for describing a near-optimum region and thus for exactly investigating optimum conditions for a multifactorial system. RSM and central composite designs has been successfully used to produce enzymes, biomass, and metabolites (Udeh and Achremowicz, 1993; Saval et al., 1993; Ergum and Mutlu, 2000). We used two successive fractional factorial designs (FFDs) followed by a RSM to optimise a medium containing fish silage, glucose, and eight inorganic salts for L-lysine production by Corynebacterium glutamicum. A comparison between the original and the optimised media for biomass and L-lysine productions, and glucose consumption, is presented.
