Vaughan L. Crow

The role of autolysis of lactic acid bacteria in the ripening of cheese

Abstract The importance of autolysis of lactic acid bacteria in cheese ripening is evident from the literature. However, the mechanisms and the consequences still require investigation. The consequences of autolysis of mesophilic starters in Cheddar cheese are discussed and highlights from current physiological and genetic studies on starter autolysis are presented. The relative merits of measuring starter autolysis in cheese by viable starter cell densities, electron microscopic observations and assay of cell-free cytoplasmic enzymes are discussed for cheese studies using different starter strains and added phage to achieve different levels of autolysis. The balance of both the intact and autolysed starter cells in young curd appear to be important in cheese ripening. The intact cells are necessary for physiological reactions such as lactose fermentation and oxygen removal and possibly for a number of flavour reactions. In contrast, the main consequence of autolysed cells in cheese is to accelerate the peptidolytic reactions. The possible influences of autolysis of adventitious lactic acid bacteria during cheese ripening are discussed.

The ecology of non-starter lactic acid bacteria (NSLAB) and their use as adjuncts in New Zealand Cheddar

Abstract This paper reviews the recent New Zealand studies on non-starter lactic acid bacteria (NSLAB) in Cheddar cheese. In New Zealand, cheese adventitious NSLAB were usually strains of homofermentative lactobacilli and occasionally pediococci and heterofermentative lactobacilli. In Cheddar cheese, manufactured in six factories at different times and ripened over 24 months, over 140 NSLAB strains were identified using pulse field gel electrophoresis profiles and fermentation patterns on 22 carbohydrates. The majority of NSLAB isolates identified belong to two species: Lactobacillus paracasei subsp. paracasei and Lactobacillus rhamnosus. The composition of the NSLAB strains in a cheese was variable and dependent on the manufacturing factory, the date of manufacture and the age of the cheese. In most cheeses, no single NSLAB strain predominated during ripening. Up to six strains formed the majority (about 90%) of the NSLAB population, the remainder was represented by up to 15 other strains. In only about 10% of the cheeses investigated was one strain predominant throughout the 24 month ripening period. The acid and salt tolerance, effect of temperature, ability to produce biogenic amines and the lipolytic and proteolytic activities of the NSLAB strains were studied to screen strains for use as adjuncts. Sixty suitable strains were selected for further screening in a cheese model system. The NSLAB were incubated at 30°C for 10 days in the model system and the flavour and biochemical changes were assessed. Twenty-three strains were eliminated as potential adjuncts because they produced defective flavours and a further 13 strains were eliminated on the basis of undesirable biochemical properties (e.g. excess fermentation of glutamate, lactate racemisation). The 24 remaining NSLAB strains that had acceptable microbiological biochemical and flavour properties were used as adjuncts in Cheddar trials. Seven of the 24 adjunct combinations studied (made up from 2–4 strains of NSLAB) provided consistent flavour benefits. Analysis of the cheeses during ripening in these trials provided significant understanding of the ripening mechanisms of the NSLAB adjuncts.

The influence of phage-assisted lysis of Lactococcus lactis subsp. lactis ML8 on cheddar cheese ripening

Cheddar cheese was made with Lactococcus lactis subsp. lactis strain ML8 as starter and two levels of rennet and three levels of homologous phage. The use of the different phage levels in cheese milk resulted in various degrees of starter lysis early in the ripening process. The levels of activity of two starter cytoplasmic enzymes, lysylaminopeptidase and FBP-aldolase, in the cheese matrix were used as a direct measure of lysis and increased with the level of phage added. Elevated starter lysis was associated with an increased rate of formation of amino acids and ammonia but the removal rate of lactose in the cheese was decreased. The relative levels of hydrophobic and hydrophilic peptides in aqueous cheese extracts were also influenced by the extent of starter lysis. Bitter flavour was prominent in cheese with high rennet concentration, but not when there was also high starter lysis. The results suggest that a balance of lysed and intact cells is important for control of cheese ripening; enzymes released on cell lysis accelerate the rate-limiting peptidolytic steps and removal of some bitter peptides, while intact cells are required for lactose removal. The consequences of the relative levels of intact and lysed cells on substrate-enzyme interactions, enzyme stability and flavour profiles are discussed.

The diversity of potential cheese ripening characteristics of lactic acid starter bacteria: 2. The levels and subcellular distributions of peptidase and esterase activities

Abstract The levels and subcellular distributions of various peptidase and esterase activities in a range of lactococcal and Streptococcus thermophilus strains were investigated. There was no correlation between the levels of the enzymes in the different strains and the ability of the strains to produce acid when grown in milk. While considerable differences between individual strains were apparent, average levels of X-prolyldipeptidyl aminopeptidase, dipeptidase and tripeptidase were similar in the Lactococcus lactis subsp. lactis and L. lactis subsp. cremoris strains studied, while that of lysylaminopeptidase (i.e. activity assayed using lysine p-nitroanilide as substrate) in the L. lactis subsp. cremoris strains was approximately double that in the L. lactis subsp. lactis strains. The average levels of lysylaminopeptidase and X-prolyldipeptidyl aminopeptidase in the S. thermophilus strains studied were similar to those in the L. lactis subsp. cremoris strains, while the average levels of dipeptidase and tripeptidase were considerably lower. All peptidases studied were recovered predominantly in the cytoplasmic fraction, although in a few strains there was some evidence to suggest that a part of the tripeptidase activity may be associated with cell structures comprising the particulate fraction. The levels of esterase activity in the strains were considerably different between strains. However, the average level of esterase activity detected in the two lactococcal subspecies was similar, while that in the S. thermophilus strains was more than double the lactococcal average. The subcellular distribution of the esterase in all strains studied showed that a significant proportion of the activity is located on the cell surface.

Purification and properties of pyruvate kinase from Streptococcus lactis.

The pyruvate kinase (ATP: pyruvate 2-O-phosphotransferase, EC 2.7.1.40) of Streptococcus lactis C10 is activated by fructose 1,6-diphosphate (Fru-1,6-P2), activity being a sigmoidal function of activator concentration. The FDP0.5V (Fru-1,6-P2 concentration giving half-maximal velocity) is markedly increased in the presence of low concentrations of inorganic phosphate; 1 mM phosphate increases the FDP0.5V value 6-fold. Although the intracellular level of Fru-1,6-P2 (12-18 mM) in exponentially growing cells on the medium used is much greater than the FDP0.5V for pyruvate kinase (0.2 mM) as determined in triethanolamine-HCl buffer, a much higher Fru-1,6-P2 concentration may be required to activate the enzyme in vivo to overcome phosphate inhibition. Tris and maleate also inhibit the enzyme. At low concentrations of Fru-1,6-P2 (0.1 mM), reaction rate is a sigmoidal function of both phosphoenolpyruvate and adenosine diphosphate (ADP) concentrations; at near saturating concentrations of activator (1 mM) the response to varying ADP is hyperbolic while the response to varying phosphoenolpyruvate becomes much less sigmoidal. The affinity for both substrates (especially phosphoenolpyruvate) is also increased by increasing the concentration of Fru-1,6-P2. The affinity of the enzyme for guanosine disphosphate (GDP) is 12-13 times that for ADP under the assay conditions used. The Streptococcus lactis pyruvate kinase has a molecular weight of 240000 with a subunit molecular weight of 60000.

Cell surface differences of lactococcal strains

Abstract A number of cell surface properties were compared in 15 pairs of lactococcal strains in order to gain an understanding of cell surface diversity and the relationship between the acquisition of the phage-resistance phenotype and alteration of cell surface properties. Each pair comprised a parent strain and a derivative resistant to a phage (O R ) or a number of phages. Three cell surface hydrophobicity patterns were found: (1) three parent strains were more hydrophobic than their O R derivatives; (2) five O R derivatives were more hydrophobic than their parent strains; (3) there were no differences for seven strain pairs. Loosely associated cell surface material was removed without cell lysis, and concentration differences between 28 strains of 40-, 23- and 11-fold were found for the extracted protein, hexose and rhamnose, respectively. These three surface components were extracted in higher concentrations from the O R derivative for seven strain pairs and from the parent strain for three strain pairs, and no differences were observed for four strain pairs. Intracellular and extracellular lipoteichoic acid concentrations varied in four of six strain pairs studied. The extracted protein profiles determined on polyacrylamide gels and by Superose 12 chromatography and the compositions of the extracted polysaccharide were different between most of the strain pairs. In addition, the surface properties, particularly cell hydrophobicity, varied according to growth conditions for some strains. The cell-surface components showed considerable diversity within the 30 lactococcal strains studied, with multiple differences between many of the strain pairs. For example, differences in hydrophobicity, the extracellular lipoteichoic acid concentration, molecular weight profile of proteins and the amount of protein, hexose and rhamnose extracted as loosely associated cell surface material were observed between the strains of pair E8/398. No unifying theme was evident to describe the basis of changes to the cell surface in the phage-resistant derivative strains.

The diversity of potential cheese ripening characteristics of lactic acid starter bacteria: 1. Resistance to cell lysis and levels and cellular distribution of proteinase activities

Abstract By reference to subcellular fraction markers, the resistance to lysis of 23 strains of Lactococcus lactis subsp. cremoris , 30 strains of L. lactis subsp. lactis and five strains of Streptococcus thermophilus and the levels and distribution of proteinase activity in the strains were determined. Strains of L. lactis subsp. cremoris were readily lysed by transfer to hypotonic buffer after treatment with lysozyme alone, whilst strains of L. lactis subsp. lactis and S. thermophilus could be efficiently lysed in this way only after treatment with a combination of lysozyme and mutanolysin. With a few notable exceptions, those strains which gave the fastest rates of acid production also generally presented higher levels of cell surface proteinase, as determined by activity on fluorescein isothiocyanate-labelled β-casein. The highest levels of cell surface proteinase detected were found for strains of L. lactis subsp. cremoris . However, the levels of total proteinase activity in the lactococcal strains did not correlate with the rate of acid production in milk, some slow acid-producers yielding similar or greater total proteinase levels than fast acid-producers. Homology to DNA probes for the lactococcal cell surface proteinase gene and to the conserved region encoding the serine proteinase active site was shown by the fast acid-producing lactococcal strains, but not by most of the slow acid-producing lactococcal strains or by the strains of S. thermophilus . A significant proportion of the total proteinase activity was recovered in the subcellular fractions in which high levels of cytoplasmic marker enzyme activity were found. The total proteinase levels detected in strains of L. lactis subsp. lactis showed a greater range of variation than in the strains of L. lactis subsp. cremoris . High levels of total proteinase activity were found in the slow acid-producers despite the strains having been grown in the presence of yeast extract. For many of the strains, the levels of proteinase released from the cell surface during cell wall degradation with lytic enzyme treatment were higher than those found using whole cells, suggesting that a significant amount of proteolytic activity was either inaccessible to substrate or present in an inactive form.

Comparison of subcellular fractionation methods for lactococcus lactis subsp. lactis and L. lactis subsp. cremoris

Abstract Cells of Lactococcus lactis subsp. lactis proved to be resistant to cell wall digestion by lysozyme using 0·6 M glycylglycine/10 mM MgCl 2 as stabilizing agent, this procedure having been described recently as suitable for subcellular fractionation of cells of L. lactis subsp. cremoris (Coolbear et al. (1992). Int. Dairy J. , 2, 213-32). A procedure has now been developed for L. lactis subsp. lactis , based on the use of a combination of lysozyme and mutanolysin to digest the walls of cells stabilized in 24% sucrose/10 mM MgCl 2 . Although the transfer of cell wall-depleted cells to hypotonic buffer resulted in lysis of most of the cells, a proportion of the cells remaining were permeable to small molecules (but not to proteins). The proportions of permeabilized cells, intact cells and cell wall-membrane complexes in the particulate fraction depended on both the lactococcal strain and the actual protocol used in the fractionation. Further, the concentration of N -acetylglucosamine in the subcellular fractions showed considerable variation between strains and the distribution of N -acetylglucosamine and another cell wall marker, rhamnose, in the fractions did not correlate. Lysylaminopeptidase activity was distributed between subcellular fractions in a similar manner to aldolase, but some evidence was obtained to suggest that a proportion of the activity may be associated with the cell membrane. For the four strains studied in detail (two strains each of L. lactis subsp. lactis and L. lactis subsp. cremoris ), nearly half of the total proteinase and esterase activities were associated with the cell surface. The origin of the remainder of the activities was unclear.

The effect of monovalent and divalent cations on the activity of Streptococcus lactis C10 pyruvate kinase

The pyruvate kinase (ATP: pyruvate 2-O-phosphotransferase, EC 2.7.1.40) from Streptococcus lactis C10 had an obligatory requirement for both a monovalent cation and divalent cation. NH+4 and K+ activated the enzyme in a sigmoidal manner (nH =1.55) at similar concentrations, whereas Na+ and Li+ could only weakly activate the enzyme. Of eight divalent cations studied, only three (Co2+, Mg2+ and Mn2+) activated the enzyme. The remaining five divalent cations (Cu2+, Zn2+, Ca2+, Ni2+ and Ba2+) inhibited the Mg2+ activated enzyme to varying degrees. (Cu2+ completely inhibited activity at 0.1 mM while Ba2+, the least potent inhibitor, caused 50% inhibition at 3.2 mM). In the presence of 1 mM fructose 1,6-diphosphate (Fru-1,6-P2) the enzyme showed a different kinetic response to each of the three activating divalent cations. For Co2+, Mn2+ and Mg2+ the Hill interaction coefficients (nH) were 1.6, 1.7 and 2.3 respectively and the respective divalent cation concentrations required for 50% maximum activity were 0.9, 0.46 and 0.9 mM. Only with Mn2+ as the divalent cation was there significatn activity in the absence of Fru-1,6-P2. When Mn2+ replaced Mg2+, the Fru-1,6-P2 activation changed from sigmoidal (nH = 2.0) to hyperbolic (nH = 1.0) kinetics and the Fru-1,6-P2 concentration required for 50% maximum activity decreased from 0.35 to 0.015 mM. The cooperativity of phosphoenolpyruvate binding increased (nH 1.2 to 1.8) and the value of the phosphoenolpyruvate concentration giving half maximal velocity decreased (0.18 to 0.015 mM phosphoenolyruvate) when Mg2+ was replaced by Mn2+ in the presence of 1 mM Fru-1,6-P2. The kinetic response to ADP was not altered significantly when Mn2+ was substituted for Mg2+. The effects of pH on the binding of phosphoenolpyruvate and Fru-1,6-P2 were different depending on whether Mg2+ or Mn2+ was the divalent cation.