Objective: To study the growth and product formation kinetics of Acetobacter suboxydans in fed-batch cultivation using different nutrient feeding strategies (by model simulation and experiment).
Theory: Basic concept and need of fed-batch cultivation
Fed-batch cultivation –
The most simple category of fermentation is batch cultivation where in the substrate is taken in the beginning of cultivation and nothing is added or withdrawn during the fermentation.
However the yield and productivity is lower in these cultivations mainly because, either the substrate &/or product inhibition occurs and the product accumulation in never optimal. Fed-batch cultivation can provide the solution to substrate inhibition problem by slow feeding of nutrients to the bioreactor; however it can still not address the severe inhibition problem due to accumulating high product concentrations. The optimal design of fed-batch cultivation has to take in to account several factors in to consideration for example time to start the fresh nutrient feed (in the end or when the culture is exponentially growing) what should be the substrate concentration in the feed and its rate of addition and when to finish the nutrient feeding so that the highest concentration of product is produced and no unconverted substrate when the reactor is full. It is rather impossible to do trial and error experimentation with so many “open ended” variables (as described above) which may play key role in the overall performance of the fed-batch cultivation.
Modelling of Fed-batch cultivation –
A batch mathematical model is therefore developed using the kinetics of the fermentation which is extrapolated to fed-batch cultivation by incorporating the mass balance terms in the model equations. A large number of fed-batch simulations are then done off line by the developed model particularly to ensure that there is significant improvement in product formation and productivity by selection of high concentration of nutrients in the feed, during selected time periods, so that unnecessary dilution of the broth can be avoided and fresh feeding is done for maximum working volume of the reactor and also towards the end of fermentation the reactor should be having highest product concentration with no unconverted substrate.
Different types of Fed-batch cultivation –
Following scenario of nutrient feeding can contribute in the elimination of substrate inhibition to yield high productivity of the product.
Add substrate when low –
This is the simplest type of fed-batch cultivation where in the fresh feeding of the nutrient is done when substrate has become limiting towards the end of the batch cultivation. At this point of time if no feeding of fresh nutrient is done for some time, the culture dies out. A step input of substrate (predesigned concentration and its rate) is identified by the mathematical model and is administered to the dying culture in the bioreactor which instantaneously raises the concentration of substrate and thereby gives an “installment” of life to the starving culture for few hours which results in product formation also for some more time than the batch cultivation. This cycle can be repeated number of times till the reactor is full of medium. In fact different combinations of substrate concentration, its rate and time of feeding can be chosen and used in the mathematical model to arrive at the best possible cultivation protocol for highly productive fermentation. The best offline simulated protocol can then be taken in to the lab and implemented to optimize the production.
Constant feeding of substrate –
Significant Improvement in product concentration and improvement in yield /productivity is possible if the nutrient feeding is done during the exponential phase of the cell growth when the maximum cell population is young and growing. This may be suitably selected by the study of batch kinetics. The mathematical model can then be used to simulate number of possibilities of start /stop time of nutrient feed, substrate concentration, its rate and so on. The simplest feeding profile could be constant feeding of suitably selected nutrient concentration and its pre identified rate such that it does not yield increased concentration of substrate than the initial substrate concentration at any time during the feeding in the bioreactor. Model can very easily facilitate the identification of above specific feeding strategy. The advantage of above strategy is that it is very simple it does not required computer coupled peristaltic pump to implement the feeding strategy. But the disadvantage is that it leads to build up of substrate which need to fermented by yet another batch cultivation so that in the end no unconverted substrate is left in the bioreactor.
Linearly or exponentially increasing nutrient feeding strategy –
In this fed-batch cultivation the nutrient feed is linearly or exponentially increased at predetermined time of cultivation. If the feeding rate coincides with the growth rate in the bioreactor by model simulations it may be possible to achieve non limiting non inhibitory concentration of substrate during the feeding period. However after the termination of feeding it may be necessary to do a secondary batch cultivation to consume the residual substrate in the bioreactor.
Decreasing rate nutrient feeding strategy –
The disadvantage of above feeding strategy can be overcoming by suitable selection of nutrient feeding strategy where in the feeding starts in exponential phase of fermentation and nutrient feeding rate is so designed that feeding is high when the culture is young and there after the rate is gradually decreased as the culture gets older & is diluted by incoming feed nutrients. It is possible to arrive at a simulation where the feeding rate of nutrient gradually decreases & stops when the reactor is full. In this there will not be any requirement of secondary batch fermentation.
Pseudo steady state of substrate or biomass –
It is possible to design the nutrient feeding at a particular substrate concentration so that its concentration is neither high enough to inhibit the fermentation nor it is too low to limit the growth in the bioreactor. To arrive at the feeding profile, which might result the constant substrate concentration in the broth, the first derivative (say ds/dt) is made zero and the model equations (because for S to be constant its first derivative has to be made zero) are rearranged to calculate the corresponding feeding rate of the substrate The feeding of the nutrient thus calculated can then be implemented to achieve pseudo steady state of any variable e.g., S, X etc.
Generalized feeding profile of nutrient feeding in fed-batch cultivation –
In this type of fed-batch cultivation design the start/stop time of feeding is first identified and the nutrient feeding concentration rate / profile is so designed that it gives highest product concentration when the reactor is full at the end of the feeding. Non linear regression is used to find the coefficients of polynomial feeding profile, and the identified profile is then used with the model equations to simulate the maximum final product concentration. This may then be implemented experimentally to optimize the fermentation.