Structure and Function of Bovine and Camel Chymosin: A Protein Chemical Study of the Differences in Milk-Clotting Abilities

Research output: Book/ReportPh.D. thesis

  • Jesper Langholm Jensen
The central step in cheese making is the separation of milk into curd and whey. This can be done enzymatically by hydrolysis of the Phe105-Met106 bond or nearby bonds in bovine κ-casein, which releases its hydrophilic C-terminal leading to coagulation of the milk. The preferred enzyme for this action is bovine chymosin from the calf's stomach, as it has a high activity towards the Phe105-Met106 bond, and a low activity towards other bonds in the milk proteins, as the latter can lead to loss of protein in the whey and release of peptides with a bitter taste. Chymosin was isolated from camel and
characterised, and turned out to have an even higher activity and specificity towards the Phe105-Met106 bond than bovine chymosin. The sequences of bovine and camel chymosin are 85% identical, and yet they have significantly different cheese making properties. The aim of the project was to explain this difference through the study of the
structures of bovine and camel chymosin, and preparation of catalytically inactive enzymes in complex with substrate. Their milk-clotting activities was determined using the traditional assay on skimmed milk, and a fluorescence resonance energy transfer (FRET) assay was developed and used to measure Michaelis-Menten kinetics towards a κ-casein derived peptide and to determine the inhibition constants of pepstatin towards the enzymes. In addition to this, the commercial products made by recombinant
expression in Aspergillus niger (A. niger) were subjected to a detailed characterisation.
The characterisation of the commercial products lead to the isolation of several glycosylation variants.
The structure of native camel chymosin was determined to 1.6 A resolution, the structure of native bovine chymosin was re-determined to 1.8 A resolution, higher than the previously known structures, and the structure of bovine chymosin in complex with the inhibitor pepstatin was solved to 1.6 A resolution.
The fold of camel chymosin differs with respect to the conformation of the N-terminal residues 1-16. The β-strand of the central sheet from the bovine enzyme is disrupted in camel chymosin, and its Nterminus is oriented towards the active site. The consequences of this are an increased flexibility of the β-barrel domains leading to a more open binding cleft, and possibly regulation of the flap or partial inhibition of the enzyme. The analysis of the domain movements upon inhibitor binding matches previous studies suggesting that the domains move by a sliding mechanism inside their hydrophobic cores.
A comparison of the surface charges show that camel chymosin contains additional positive charges, which may improve interaction with the negatively charged C-terminal of κ-casein. The complexes of bovine chymosin, human pepsin, and endothiapepsin with the inhibitor pepstatin were examined to explain pepstatin’s lower affinity towards bovine chymosin. Human pepsin has a loop insert and a more hydrophobic binding cleft leading to a larger interaction surface with the hydrophobic pepstatin. Endothiapepsin forms additional hydrogen bonds with pepstatin’s backbone and also has a more hydrophobic binding cleft compared to bovine chymosin. The inhibition assay showed that pepstatin has a lower affinity towards camel chymosin probably due to its more open binding cleft and the disordered N-terminus. The affinity increases as pH is lowered.
The FRET assay was based on a labelled peptide derived from the 98-112 sequence of bovine κ-casein. This was used to determine the Michaelis-Menten kinetics of bovine and camel chymosin under milkclotting like conditions. The results showed that camel chymosin has a higher substrate affinity and a higher turnover number than bovine chymosin. This can be explained by improved substrate binding in the more open binding cleft of camel chymosin, the increased flexibility of the domains, and the replacement of Lys221 and Val223 in the binding cleft of bovine chymosin with Val221 and Phe223 in
camel chymosin. The latter would avoid electrostatic repulsion of the substrate's His residues and improve interaction with the hydrophobic substrate residues adjacent to the scissile bond. In summary, this work shows that the improved milk-clotting properties of camel chymosin are due to several factors. First, additional positively charged patches on the surface of camel chymosin improve its association with the negatively charged C-terminal part of κ-casein. Second, the disordered N-terminus of camel chymosin leads to a greater flexibility of the domains opening the binding cleft for improved binding of the substrate. The disordered N-terminus may also play a role in regulation of the flap and
interact with the substrate. Third, the identified differences in the binding cleft of camel chymosin leads to an improved binding of the stretch of κ-casein that is cleaved during milk-clotting.
Original languageEnglish
PublisherDepartment of Chemistry, Faculty of Science, University of Copenhagen
Number of pages135
Publication statusPublished - 2013

ID: 96653045