Modelling of enzymatic protein hydrolysis

Seleme, Raphahle Nthabiseng (2019-04)

Thesis (MEng)--Stellenbosch University, 2019.

Thesis

ENGLISH ABSTRACT: Enzymatic hydrolysis is a method which is used to produce nutritional supplements, aquaculture feed, plant fertilisers and food ingredients. Mathematical modelling of enzymatic protein hydrolysis is a valuable tool for predicting complex protein hydrolysis reactions, but is highly complex. Previous models are not adequate due to their simplified description of the reaction. Protein hydrolysis involves different substrates and reactions. The products can also act as new substrates resulting in continuous hydrolysis. This study was aimed at developing and validating a theoretical model for protein hydrolysis. To develop a mathematical model to describe enzymatic protein hydrolysis, the current study proposes a population balance approach along with a modified Michaelis-Menten equation. Population balance modelling was used to predict the population of polymer chains while the modified Michaelis-Menten equation describes the hydrolysis of proteins to form shorter chains. Three kinds of catalytic actions were considered: Endopeptidase, exopeptidase, and endopeptidase-exopeptidase. In order to validate the proposed models for endopeptidase and endopeptidase-exopeptidase, hydrolysis of whey protein using commercial enzymes, papain and bromelain, were studied at different operating conditions. The optimum conditions of the enzymes were determined experimentally using a central composite experimental design. The effect of enzyme concentration and buffer type on protein hydrolysis were studied using a multilevel factorial design. The degree of hydrolysis and total heat flow during hydrolysis was measured using o-phthaldialdehyde and isothermal microcalorimetry methods. The heat flow and degree of hydrolysis data were used to calculate the model parameters. The optimum conditions under the studied experimental conditions: Temperature, pH, substrate concentration, enzyme-to-substrate ratio, were 65 ºC, 5.15, 6 % (w/v), 3% (w/w) respectively for papain, and 50 ºC, 6, 10 % (w/v), 3 % (w/w) for bromelain. The heat flow results showed that the heat measured using isothermal microcalorimetry is the apparent heat rather than a heat of the reaction. The comparison between the two techniques for measuring hydrolysis indicated that isothermal microcalorimetry is more accurate and easy to use than spectroscopy method. The parameters of the models were estimated using nonlinear regression analysis to obtain the model predictions. The model predictions from the proposed models and a model by Marquez-Moreno and Fernandez-Cuadrado (1993) were compared to the experimental data. The proposed model for endopeptidase was on average 60.5 % and 37.4 % better than the model found in the literature for heat flow and degree of hydrolysis experimental data. The model for endopeptidase-exopeptidase was on average 55.4 % and 46.5 % better than the model found in literature. This implied that the proposed models presented a promising approach for modelling a protein-peptidase system and proved to predict the experimental data better than the empirical model in literature. Sensitivity analysis was performed to determine the parameters that have maximum impact on protein hydrolysis. The results showed that two parameters had a great influence on hydrolysis. In conclusion, the proposed models can be used to predict complex enzymatic protein hydrolysis reactions with reasonable certainty. Knowledge of the mathematical model of protein hydrolysis is important for process optimization and process control.

AFRIKAANSE OPSOMMING: Ensimatiese hidroliese is ʼn metode wat gebruik word om voedingsaanvullings, awkakultuurvoer, plant bemesting en voedselbestanddele te produseer. Wiskundige modellering van ensimatiese proteïen hidrolise is ʼn waardevolle instrument om komplekse proteïen hidrolise reaksies te voorspel, maar is hoogs kompleks. Vorige modelle is nie voldoende nie as gevolg van hul vereenvoudigde beskrywings van die reaksie. Proteïen hidrolise behels verskillende substrate en reaksies. Die produkte kan ook optree as nuwe substrate wat kontinue hidrolise tot gevolg het. Hierdie studie is gerig op die ontwikkeling en validering van ʼn teoretiese model vir proteïen hidrolise. Om ʼn wiskundige model te onwikkel wat ensimatiese proteïen hidrolise beskryf, stel die huidige studie ʼn populasie balans benadering saam met ʼn gewysigde Michaelis-Menten vergelyking voor. Populasie balans modellering is gebruik om die populasie polimeerkettings te voorspel terwyl die gewysigde Michaelis-Menten vergelyking die hidrolise van proteïne wat korter kettings vorm, beskryf. Drie soorte katalitiese aksies is oorweeg: endopeptidase, eksopeptidase, en endopeptidase-eksopeptidase. Om die voorgestelde modelle vir endopeptidase en endopeptidase-eksopeptidase te valideer, is hidrolise van weiproteïen deur kommersiële ensieme, papaïen en bromelien, bestudeer by verskillende bedryfstoestande. Die optimale toestande van die ensieme is eksperimenteel vasgestel deur ʼn sentrale samestelling eksperimentele ontwerp te gebruik. Die effek van ensiemkonsentrasie en buffer tipe op proteïen hidrolise is bestudeer deur ʼn multivlak faktoriaalontwerp te gebruik. Die grade van hidrolise en totale hittevloei is gemeet deur o-tartaardialdehied en isotermiese mikrokalorimetriemetodes te gebruik. Die hittevloei en grade van hidrolise data is gebruik om die model parameters te bereken. Die optimale toestande in die bestudeerde eksperimentele toestande was temperatuur, pH, substraatkonsentrasie, ensiem-tot-substraatverhouding van 65 °C, 5.15, 6% (w/v), 3% (w/w) onderskeidelik vir papaïen, en 50 °C, 6, 10% (w/v), 3% (w/w) vir bromelien. Die hittevloei resultate het gewys dat die hitte wat deur isotermiese mikrokalorimetrie gemeet is, is die oënskynlike hitte eerder as die hitte van die reaksie. Die vergelyking tussen die twee tegnieke om hidrolise te meet het aangedui dat isotermiese mikrokalorimetrie die metode is wat meer akkuraat en makliker om te gebruik, is. Die parameters van die modelle is beraam deur nie-liniêre regressie analise te gebruik om die model voorspellings te verkry. Die model voorspellings van die voorgestelde modelle en ʼn model deur Marquez-Moreno en Fernandez-Cuadrado (1993) is vergelyk met die eksperimentele data. Die voorgestelde model vir endopeptidase het die kleinste fout van 18.04% gehad en 14.67% minder as dié van die model uit literatuur vir hittevloei en grade van hidrolise korrelasie met eksperimentele data. Die model vir endopeptidase-eksopeptidase het die kleinste fout van 16.16% gehad en 23.11% minder as dié van die literatuurmodel. Dit het geïmpliseer dat die voorgestelde modelle ʼn belowende benadering lewer vir modellering van ’n proteïen-peptidase-stelsel en het bewys dat dit die eksperimentele data beter voorspel as die empiriese model in literatuur. Sensitiwiteitsanalise is uitgevoer om die parameters te bepaal wat maksimum impak op proteïen hidrolise het. Die resultate het gewys dat twee parameters ʼn beduidende invloed op hidrolise gehad het. Ten slotte, die voorgestelde modelle kan gebruik word om komplekse ensimatiese proteïen hidrolise reaksies te voorspel. Kennis van die wiskundige model van proteïen hidrolise is belangrik vir proses optimalisering en prosesbeheer.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/106159
This item appears in the following collections: