Solid-state fermentation : modelling fungal growth and activity

<p>In solid-state fermentation (SSF) research, it is not possible to separate biomass quantitatively from the substrate. The evolution of biomass dry weight in time can therefore not be measured. Of the aiternatives to dry weight available, glucosamine content is most promising.<br/>Glucosamine is the monomer of the cell-wall component chitin. Glucosamine content of a fermented substrate is therefore related to the biomass present. The concentration of glucosamine in biomass, however, might vary in time end with culture conditions.<br/>Instead of using the glucosamine content to calculate how much biomass dry weight is present as was done previously, in this research biomass growth and activity are directly related to glucosamine. With these descriptions a mathematical model is constructed which allows prediction of biomass glucosamine and temperature patterns in an SSF bed.<p>The research is done with Trichoderma reesei QM9414 growing on wheat bran as a model SSF. The fermentations are carried out in Petri dishes containing 5 g moistened inoculated and sterilized wheat bran which are placed in an incubator with constant temperature and ambient relative humidity. Samples for analysis ware drawn by taking Petri dishes from the incubator. In this way, accurate measurement of dry-matter weight loss, respiration activfty and glucosamine was possible with a standard deviation of less than 7%. Measurement of ATP and cellulase activity proved not to be as accurate. This was attributed to handling of the fermented substrate during necessary pretreatment procedures for ATP measurement and interactions between enzyme and substrate, respectively.<br/>Respiration activities, i.e. oxygen consumption rate and carbon-dioxide production rate, were measured simuitaneously. During a 125 h fermentation ca. 9 mmol CO2, and 02 per gram initial dry matter were produced and consumed, respectively. The decrease in dry matter in this period amounted to ca. 0.20 g per gram initial dry matter. The increase in glucosamine could be described with a logistic equation, with initial and final level of 0.02 and 8.1 mg glucosamine per gram initial dry matter, respectively. The maximum specific growth rate amounted to 0.123 per hour.<br/>The specific respiration activities were calculated per quantity of glucosamine. The correlation with maximum specific growth rate deviated from Pirts linear-growth model. These deviations ware attributed to the different forms in which fungal biomass can be present (active growing and active non-growing). These deviations are, however, of minor importance in modelling fungal activity since they appear only at the initial stage of fermentation where the amount of fungal biomass is small.<br/>There was a pronounced decline in respiration activity after growth has stopped. This decline, called inactivation, was ascribed to a decrease in amount of active non-growing biomass. The rate of decline seemed constant in time under isothermal conditions, but increased exponentially with increasing temperature above the maximum temperature for growth.<p>The influence of temperature on specific growth rate, maximum attainabie biomass glucosamine level end yield of glucosamine on oxygen consumption or carbon-dioxide production could be described with a (modified) Ratkowsky equation or a Gaussian curve. The influence of temperature on the maintenance coefficient was negligible.<p>These mathematical equations were combined with conservation laws describing mess and heat transfer in a simulation model. lnactivation, as described under isothermal conditions, was implemented in three different ways, which resuited in the models M <sub>part</sub> , M <sub>temp</sub> , and M <sub>cont</sub> . In the M <sub>part</sub> , model, inactivation starts when the specific growth rate becomes equal to or less than 1 % of its maximum value. The inactivation continues until no activity is left. In the M <sub>temp</sub> model, inactivation starts for the same reason, but stops when the specific growth rate is more than 1 % of its maximum value again. In the M <sub>cont</sub> model, inactivation is continuously effective, immediately from the start of fermentation.<br/>The temperature end biomass evolution in a tray SSF predicted by these three models are compared with a model in which inactivation is omitted. The results show the importance of describing the inactivation in modelling SSF. Without this, irrealistic temperature and biomass glucosamine patterns are predicted.<br/>The M <sub>part</sub> , model, representing inactivation due to substrate limitation or product toxicity, predicts temperature and blomass patterns described in the literature: restricted biomass growth end decrease in temperature at the end of fermentation. The mathematical models reported thus far in the literature predict restricted growth or a decrease in temperature.<br/>Although the combination of the logistic equation, for describing fungal biomass in time, and the linear-growth model, for correlating the respiration activity to biomass, is most often used in scientific literature on modelling SSF, it is insufficiently detailed. The logistic equation is shown to be applicabie to the glucosamine measurement results, but the glucosamine determination cannot distinguish between active growing, active non-growing end inactive biomass. The total<br/>amount of these three types of biomass is described by the logistic equation, while the linear-growth model only describes the first two types. It is therefore suggested to focus research efforts on the description of the evolution of these three different forms of fungal biomass in time. Sensitive image analysis techniques, combined with computer technology, applied to fungal biomass growing under SSF conditions have the potential of being powerful tools to accomplish this target