Gerda Urbach

Contribution of lactic acid bacteria to flavour compound formation in dairy products

Abstract Cheese flavour cannot be produced without starter bacteria. Lactic acid bacteria convert lactose to lactic acid and this together with their production of diacetyl and acetaldehyde are their main contributions to the flavour of cultured milks and fresh cheeses. In matured cheeses, the starter bacteria die out quickly and the rate at which they lyse and release their enzymes into the system has an influence on the rate at which free amino acids are formed. Rennet alone is mainly responsible for the formation of large, medium and small peptides but, without interaction with other enzymes, is capable of producing only methionine, histidine, glycine, serine and glutamic acid at quantifiable levels. Free amino acids in Cheddar cheese are mainly the result of microbial peptidase activity. These amino acids, together with the products of glycolysis, form substrates for secondary flora, the nature of which, in many cases, determines the cheese variety. They also form substrates for enzymes from the milk, e.g. the production of H 2 S appears to be dependent on milk enzymes. Methionine, which is released by rennet, is further metabolized by starter enzymes with the production of methanethiol which plays a major role in cheese flavour possibly as a potentiator for other flavours. α—Dicarbonyls, particularly methylglyoxal and diacetyl, and bacteria which can produce them, appear to play a crucial role in the formation of cheese flavour, both the desirable flavour of full-fat cheese and the meaty-brothy off-flavour of low-fat cheese. Although, theoretically, there are many compounds in cheese which could react purely chemically to form flavour compounds, these reactions are also mediated by enzymes in the cheese system and it seems unlikely that straight out chemical reactions play a major role in the production of cheese flavour. The role of the secondary flora is likely to be much more important than that of chemical reactions. Particularly in Cheddar and Emmental it has been shown that good quality cheeses have a low oxidation-reduction potential. This is more likely to be an indicator for the establishment of the anaerobic conditions required for the flavour forming reactions to proceed than an active causal agent of flavour formation. The function of glutathione is more likely to be as some sort of facilitator in enzyme reactions than as an agent for the reduction of oxidation-reduction potential. The ability of bacteria to accumulate glutathione from their media is likely to be one of the indicators of flavour generating capacity. Suitable selected strains of adjunct bacteria increase the rate and intensity of formation of Cheddar cheese flavour but unsuitable adjuncts can also cause off-flavours.

Relations between cheese flavour and chemical composition

Abstract Five Australian groups have attempted to correlate various compounds with the development of Cheddar flavour. Perret (1978) used vacuum distillation of grated mature Cheddar cheese followed by gas chromatography and mass spectrometry to reveal more than 150 components, including acids, alcohols, esters, aldehydes, ketones, sulphur compounds and lactones. He determined H2S, thiols and disulphides chemically to obtain the actual concentrations of sulphur compounds in the cheese rather than in the headspace. Commercial cheeses with flavour defects described as soapy, eggy, oniony and fruity were found to contain ‘out of balance’ concentrations of fatty acids, hydrogen sulphide, methane thiol and ethyl esters, respectively. Aston (1981, M. Appl. Sc. thesis, Queensland Institute of Technology, Brisbane, Queensland) compared normal Cheddar cheeses with cheeses manufactured using (a) mutant starters, (b) higher initial ripening temperature and (c) a combination of (a) and (b) to accelerate ripening. The cheeses were analysed at 1, 3, 6 and 9 months for (1) flavour, (2) volatile sulphur compounds (by GC/FPD of the headspace), (3) amino acids (phosphotungstic acid-soluble nitrogen) and (4) free amino acids and trichloracetic acid-soluble peptides (TCA-soluble tyrosine). Only (3) and (4) correlated well enough with flavour to be considered useful as repening indicators. Barlow et al. (1989a, b, Aust. J. Dairy Technol., 44, 7–18) examined 54 batches of Cheddar cheese over the period of maturation. At each sampling time the cheeses were submitted to a taste panel and analysed for H2S, methane thiol, dimethyl sulphide, free fatty acids from acetic to hexanoic acids, d - and l -lactic acid, water soluble amino-nitrogen (WSN), and volatiles in the headspace over grated Cheddar cheese. The best Cheddar flavour was associated with 45–50 mg/kg butyric and 20–25 mg/kg hexanoic acids. WSN, lactic acid and H2S correlated highly with Cheddar flavour. A reasonable prediction of flavour at 12 months could be made from flavour at 3 months plus H2S plus WSN. Wood (1989, M. Appl. Sc. thesis, Queensland Institute of Technology, Brisbane, Queensland) developed a method for analysis of the volatile headspace components from cheese using cryogenic sample collection. In a 20-month old cheese, methanol, ethanol, acetone, propyl formate, carbon disulphide, propan-1-ol, butanone, butan-2-ol, 3-methyl-butanal, pentan-2-one, pentanal, dimethyl disulphide, toluene, diethyl carbonate, hexanal, ethyl butyrate, ethyl benzene, o-, m- and p-xylene, α-pinene and an α-pinene isomer, dimethyl trisulphide, β-pinene were identified. The effect of, (1) addition of neutrase, (2) use of mutant starter, (3) increased storage temperature, and various combinations of these on accelerating maturation and on the profile of headspace volatiles during maturation were examined. Dimos (1992, M.Sc. thesis, Latrobe University, Bundoora, Victoria) compared the volatiles from full-fat and low-fat (7%) Cheddar cheeses, manufactured on 4 separate occasions at the CSIRO pilot plant, over a 26-week maturation period and found a significant straight-line relationship between actual flavour and fitted flavour calculated from: y = b0 + b1x1 + b2x2 + b3x3 + b4x4 + b5x5 where y is the fitted flavour, b0–b5 are constants and x15 are the logs of the concentrations of H2S, heptan-2-one, butanone, δ-decalactone and propan-2-ol. This relationship held for both the full-fat and the low-fat cheese. Increases in butanone and butan-2-ol with maturation are widely reported, although high quality, mature Cheddar does not necessarily contain these compounds. Sulphur compounds, particularly CH3SH, have been shown to be essential to Cheddar flavour but are likely to be breakdown products of a highly unstable Cheddar flavour compound or compounds.

THIN-LAYER CHROMATOGRAPHY OF ALIPHATIC 2,4-DINITROPHENYLHYDRAZONES.

Abstract Thin-layer chromatographic procedures are described for the separation of 2,4-dinitrophenylhydrazones into: 1. Individual members of homologous series by a partition system between 2-phenoxyethanol supported on kieselguhr G and light petroleum (b.p. 100–120°), 2. n-Aldehyde, n-alkan-2-one and n-alk- i -en-3-one derivatives on aluminium oxide G with 4% diethyl ether in light petroleum (b.p. 30–40°) as the solvent, 3. n-Alkanal, n-alk-2-enal, n-alka-2,4-dienal, n-nona-trans-2,trans-6-dienal and n-nona-trans-2,cis-6-dienal derivatives on plates of aluminium oxide G containing 25% AgNO3 (w/w) with 16% diethyl ether in light petroleum (b.p. 30–40°) as the developing solvent. These procedures have been combined in two-dimensional techniques to give a separation of mixtures of the 2,4-dinitrophenylhydrazones of the normal homologous series of alkan-2-ones, alk- i -en-3-ones, alkanals, alk-2-enals, alka-2,4-dienals, and alka-2,6-dienals.

Volatile compounds in butter oil: II. Flavour and flavour thresholds of lactones, fatty acids, phenols, indole and skatole in deodorized synthetic butter

The flavours of the following compounds were determined in synthetic butter prepared from 84% steam-deodorized butter oil and 16% water: C 6, 8, 10, 12, 14 δ-lactones, C 12 γ-lactone, C 2, 4, 6, 8, 10, 12, 14 n -alkanoic acids, phenol, m - and p -cresols, o -methoxyphenol, indole and skatole. C 6, 8, 10 δ-lactones, C 2, 4, 6, 8, 10, 12, 14 n -alkanoic acids, phenol, m - and p -cresols, o-methoxyphenol and volatile fractions from butter oil were also evaluated in bland butter. None of the compounds listed constitute basic butter flavour on their own, but C 8, 10 δ-lactones, C 10 acid, phenol, p -cresol, indole and skatole contribute to butter flavour.

Dynamic headspace gas chromatography of volatile compounds in milk

A method is described for investigating volatile compounds in milk. The volatiles are removed from milk by a stream of helium swept at 100 ml/min over the surface of the milk at 70 degrees C. They are trapped on 40 mg of NIOSH charcoal and then desorbed by heat and re-trapped on the front of a chromatographic column of Tenax-GC coated with 1% OV-275, the column being maintained at room temperature during trapping. An amount of 40 mg NIOSH charcoal under these conditions traps over 90% of the total quantity of the lowest boiling compounds swept from the milk, such as acetaldehyde and ethanol, and retains 100% of the total quantity of acetone, propanol and higher boiling compounds from the gas stream. The volume of milk and its temperature affect the ratios of volatiles collected and these factors are useful in increasing the proportion of a volatile of particular interest. The addition of potassium carbonate increases the yield of volatiles from 100 ml aqueous phase but not from 10 ml.

Volatile compounds in butter oil: IV. Quantitative estimation of free fatty acids and free δ-lactones in butter oil by cold-finger molecular distillation

Free fatty acids (FFA) and free δ-lactones in butter oil were quantitatively estimated by cold-finger molecular distillation at 35 °C followed by silicic-acid column chromatography of the cold-finger distillate and gas chromatography of the fatty acid and lactone fraction. Typical values for FFA and free δ-lactones in butter oil prepared from export standard sweet-cream salted Australian butter were in the range 0·07–0·24, 0·75–2·4, 4·5–8·6, 34–40 and 44–89 ppm for the C 4,6,8,10,12 fatty acids and 0·09–0·18, 1·2–4·9 and 3·0–5·8 ppm for the C 8,10,12 δ-lactones. These values are in the range of organoleptically desirable levels.

Volatile compounds in butter oil: III. Recovery of added fatty acids and δ-lactones From volatile-free butter oil by cold-finger molecular distillation

More than 90% of the C 2,4,6,8,10 fatty acids and C 8,10 , δ-lactones added at quantities between 0·5 and 40 ppm were recovered from volatile-free butter oil by cold-finger molecular distillation at 35°C, while approximately 65% of the C 12 fatty acid and of the C 12 δ-lactone was recovered. Cold-finger molecular distillation at 50°C increased the recovery of the C 12 compounds to more than 90%. The acids and lactones were estimated quantitatively as the free compounds by gas chromatography on a mixture of Carbowax 20 M and isophthalic acid as the stationary phase.

The effect of diet on the γ- and δ-lactone and methyl ketone potentials of bovine butterfat

When caprine butterfat was heated in the presence of water vapour and the absence of air, γ- and δ-lactones and methyl ketones were produced which were qualitatively the same as those from bovine butterfat. The saturated δ-lactone and methyl ketone potentials of the caprine butterfat were of the order of one third those of bovine butterfat when the goats and cows received similar rations. γ-Dodecanolactone, γ-dodec- cis -6-enolactone, δ-tetradec- cis -8-enolactone and δ-tetradec- trans -8-enolactone potentials of the caprine butterfat were in the same range as for bovine butterfat when the animals were on similar diets. The γ- and δ-lactone and methyl ketone potentials and the fatty acid composition of caprine butterfat varied with the diet of the goat. Supplementing the diet with polyunsaturated fat, whether or not protected against biohydrogenation in the rumen, lowered the saturated δ-lactone and methyl ketone potentials as well as the proportion of C 10,12,14,16 fatty acids in the caprine butterfat. Inclusion of crushed oats in the diet increased the γ-dodecanolactone potential. The γ-dodecanolactone and they γ-dodec- cis -6-enolactone potentials were also enhanced by feeding seed oil supplements which were not protected against ruminal hydrogenation. To prevent the development of the sweet off-flavour associated with γ-dodecalactones in heated, ruminant-derived meat and butterfat with elevated levels of linoleic acid, it appears that oats should be excluded from the basal ration of the ruminant and the polyunsaturated oil supplement should have the highest possible degree of protection against biohydrogenation in the rumen.