Towards sustainable Cleaning-in-Place (CIP) in dairy processing: Exploring enzyme-based approaches to cleaning in the Cheese industry.

Cleaning-in-place (CIP) is the most commonly used cleaning and sanitation system for processing lines, equipment, and storage facilities such as milk silos in the global dairy processing industry. CIP employs thermal treatments and nonbiodegradable chemicals (acids and alkalis), requiring appropriate neutralization before disposal, resulting in sustainability challenges. In addition, biofilms are a major source of contamination and spoilage in dairy industries, and it is believed that current chemical CIP protocols do not entirely destroy biofilms. Use of enzymes as effective agents for CIP and as a more sustainable alternative to chemicals and thermal treatments is gaining interest. Enzymes offer several advantages when used for CIP, such as reduced water usage (less rinsing), lower operating temperatures resulting in energy savings, shorter cleaning times, and lower costs for wastewater treatment. Additionally, they are typically derived from natural sources, are easy to neutralize, and do not produce hazardous waste products. However, even with such advantages, enzymes for CIP within the dairy processing industry remain focused mainly on membrane cleaning. Greater adoption of enzyme-based CIP for cheese industries is projected pending a greater knowledge relating to cost, control of the process (inactivation kinetics), reusability of enzyme solutions, and the potential for residual activity, including possible effects on the subsequent product batches. Such studies are essential for the cheese industry to move toward more energy-efficient and sustainable cleaning solutions.

Invited review: Starter lactic acid bacteria survival in cheese: New perspectives on cheese microbiology.

The importance of starter cultures to cheese manufacture and ripening is well known. Starters are inoculated into cheese milk at a level of ∼106 cfu/mL either from a bulk culture or using commercial direct-to-vat cultures. Before ripening, starters grow in the milk to reach populations of 107 to 109 cfu/g of curd depending on processing variables such as cook temperature, inclusion of washing steps, degree of partitioning with curds and whey, and importantly salt addition rate. Inherent strain-related properties also determine final populations in the curd following manufacture and include temperature sensitivity, salt sensitivity, presence of prophage, autolytic and permeabilization properties (which are influenced by processing steps), presence and type of cell envelope proteinase, and metabolic activity. Ripening of important industrial cheese varieties such as Cheddar, Dutch, Swiss, and Italian-type cheese varieties is characterized by extended storage under temperature-controlled conditions enabling characteristic flavor and texture development to occur. Over ripening, microbiological, biochemical and enzymatic changes occur with a decline in starter viability, release of intracellular enzymes, hydrolysis of proteins, carbohydrates and lipids, and formation of a range of volatile and nonvolatile flavor components. Recent reports suggest that starter strains may be present during the later stages of ripening and therefore their potential role needs to be reconsidered. This review will focus on our current understanding of starter viability and vitality during cheese ripening and will also review the area of starter permeabilization, autolysis, and enzyme release.

Gaps in the assortment of rapid assays for microorganisms of interest to the dairy industry.

This review presents the results of a study into the offering of rapid microbial detection assays to the Irish dairy industry. At the outset, a consultation process was undertaken whereby key stakeholders were asked to compile a list of the key microorganisms of interest to the sector. The resultant list comprises 19 organisms/groups of organisms divided into five categories: single pathogenic species (Cronobacter sakazakii, Escherichia coli and Listeria monocytogenes); genera containing pathogenic species (Bacillus, Clostridium, Listeria, Salmonella; Staphylococcus); broad taxonomic groupings (Coliforms, Enterobacteriaceae, fecal Streptococci, sulfite reducing bacteria/sulfite reducing Clostridia [SRBs/SRCs], yeasts and molds); organisms displaying certain growth preferences or resistance as regards temperature (endospores, psychrotrophs, thermodurics, thermophiles); indicators of quality (total plate count, Pseudomonas spp.). A survey of the rapid assays commercially available for the 19 organisms/groups of organisms was conducted. A wide disparity between the number of rapid tests available was found. Four categories were used to summarize the availability of rapid assays per organism/group of organisms: high coverage (>15 assays available); medium coverage (5-15 assays available); low coverage (<5 assays available); no coverage (0 assays available). Generally, species or genera containing pathogens, whose presence is regulated-for, tend to have a good selection of commercially available rapid assays for their detection, whereas groups composed of heterogenous or even undefined genera of mainly spoilage organisms tend to be "low coverage" or "no coverage." Organisms/groups of organisms with "low coverage" by rapid assays include: Clostridium spp.; fecal Streptococci; and Pseudomonas spp. Those with "no coverage" by rapid assays include: endospores; psychrotrophs; SRB/SRCs; thermodurics; and thermophiles. An important question is: why have manufacturers of rapid microbiological assays failed to respond to the necessity for rapid methods for these organisms/groups of organisms? The review offers explanations, ranging from the technical difficulty involved in detecting as broad a group as the thermodurics, which covers the spores of multiple sporeforming genera as well at least six genera of mesophilic nonsporeformers, to the taxonomically controversial issue as to what constitutes a fecal Streptococcus or SRBs/SRCs. We review two problematic areas for assay developers: validation/certification and the nature of dairy food matrices. Development and implementation of rapid alternative test methods for the dairy industry is influenced by regulations relating to both the microbiological quality standards and the criteria alternative methods must meet to qualify as acceptable test methods. However, the gap between the certification of developer's test systems as valid alternative methods in only a handful of representative matrices, and the requirement of dairy industries to verify the performance of alternative test systems in an extensive and diverse range of dairy matrices needs to be bridged before alternative methods can be widely accepted and adopted in the dairy industry. This study concludes that many important dairy matrices have effectively been ignored by assay developers.

Protein A-Mediated Binding of Staphylococcus spp. to Antibodies in Flow Cytometric Assays and Reduction of This Binding by Using Fc Receptor Blocking Reagent.

Staphylococcus aureus and other coagulase-positive Staphylococcus spp. bind the Fc region of IgG antibodies through expression of protein A (SpA). These species have consequently been a source of false-positive signals in antibody-based assays designed to detect other target bacteria. Here, flow cytometry was used to study the influence of a number of factors on the SpA-mediated binding of single cells to an anti-human IgG antibody, including strain, heat killing, overnight storage, growth phase, cell physiology, surface adhesion, and growth in model food systems. Through the costaining of antibody-stained cells with the permeability dye propidium iodide and calcein violet AM, the cell physiological status was related to SpA-mediated antibody binding. Generally, permeabilized cells lacking esterase activity did not strongly bind antibody. The binding of a number of commercially available polyclonal IgG antibodies to non-Staphylococcus spp. was also characterized. Not all SpA-expressing species showed strong binding of mouse IgG, and one species not known to express SpA showed strong binding. Most SpA-expressing strains bound rabbit IgG antibodies to some extent, whereas only one strain bound goat IgG. To reduce or eliminate SpA-mediated IgG binding, the following products were evaluated as blocking reagents and applied prior to staining with primary or secondary antibody: normal rabbit serum, mouse IgG isotype control, goat IgG, and a commercial FcR blocking reagent. Only the FcR blocking reagent consistently reduced SpA-mediated binding of Staphylococcus spp. to antibodies against other species and could be recommended as a blocking reagent in immunoassays designed to detect non-Staphylococcus species.IMPORTANCE This study characterizes a widespread but little-studied problem associated with the antibody-based detection of microbes-the Staphylococcus protein A (SpA)-mediated binding of IgG antibodies-and offers a solution: the use of commercial FcR blocking reagent. A common source of false-positive signals in the detection of microbes in clinical, food, or environmental samples can be eliminated by applying this study's findings. Using flow cytometry, the authors demonstrate the extent of heterogeneity in a culture's SpA-mediated binding of antibodies and that the degree of SpA-mediated antibody binding is strain, growth phase, and food matrix dependent and influenced by simulated food processing treatments and cell adherence. In addition, our studies of SpA-mediated binding of Staphylococcus spp. to antibodies against other bacterial species produced a very nuanced picture, leading us to recommend testing against multiple strains of S. aureus and S. hyicus of all antibodies to be incorporated into any immunoassay designed to detect a non-Staphylococcus spp.

Heat and Chemical Treatments Affect the Viability, Morphology, and Physiology of Staphylococcus aureus and Its Subsequent Antibody Labeling for Flow Cytometric Analysis.

The effects of heat and chemical treatments on Staphylococcus aureus viability and physiology and their subsequent effects on antibody binding ability and cell morphology were measured. Treatments included lethal and sublethal heat; exposure to organic acids, salt, and sodium hydroxide; and freeze-thawing. Strain-related differences in viability were noted depending on treatment and were reflected in changes in physiology as monitored by flow cytometry (FCM) using three different staining protocols: SYTO 9/propidium iodide (PI), DiOC2(3), or calcein acetoxymethyl ester (calcein-AM)/PI. Treatments that resulted in significant losses in viability as measured by plate counting were reflected better by the first two staining combinations, as intracellular calcein-AM uptake may have been impaired by certain treatments. FCM analysis using labeling by commercial anti-S. aureus antibodies indicated that differences in cell physiology as a result of treatments influenced immunofluorescence detection. The ratio of the mean fluorescence intensities of stained cells to those of unstained cells [MFI/MFI(us)] varied with treatment, five of these treatments, including freeze-thaw, citric acid, oxalic acid, NaCl, and NaOH treatments, resulted in significantly lower fluorescence values compared to controls.IMPORTANCE FCM data indicated that cells conventionally considered to be dead and which would not give rise to CFU in a plate count assay, e.g., cells heated to 80°C, were labeled by antibody staining. This finding suggests that without the inclusion of a live/dead discriminating dye, these cells would be erroneously detected as viable within an FCM assay. Reductions in antibody staining due to physicochemical treatment were strain related, reflecting the complexity of the phenomenon under study and illustrating that substantial validation of any new antibody detection-based method, including physiological staining and cell sorting, should be undertaken. Researchers should be aware of physicochemical treatments causing false-negative results: in this study, freeze-thawing severely reduced antibody binding without affecting the viability of a substantial percentage of cells. Scanning electron microscopy carried out on treated cells revealed a range of morphological changes resulting from physicochemical treatments which may have hindered antibody binding.

Application of Flow Cytometry to the Detection of Pathogenic Bacteria.

Outbreaks of infections have emphasized the necessity for rapid and economic detection methods for pathogens in samples ranging from those of clinical origin to food products during production and retail storage, and increasingly, in environmental samples. Flow cytometry (FCM) allows the rapid acquisition of multi-parametric data regarding cell populations within fluidised samples. However, the application of FCM to pathogen detection depends on the availability of specific fluorescent probes such as antibodies and RNA probes capable of detecting and isolating pathogens from these diverse samples. A particular issue for FCM methodology is the ability to recover and discriminate bacteria from the sample matrix which may pose a major technical hurdle towards accurate and sensitive analysis. This review article focuses on detection of pathogens using FCM in samples originating from food, water, environmental and clinical sources and outlines the current state of the art and potential future applications.

Control of oxidation-reduction potential during Cheddar cheese ripening and its effect on the production of volatile flavour compounds.

In cheese, a negative oxidation-reduction (redox) potential is required for the stability of aroma, especially that associated with volatile sulphur compounds. To control the redox potential during ripening, redox agents were added to the salted curd of Cheddar cheese before pressing. The control cheese contained only salt, while different oxidising or reducing agents were added with the NaCl to the experimental cheeses. KIO3 (at 0·05, 0·1 and 1%, w/w) was used as the oxidising agent while cysteine (at 2%, w/w) and Na2S2O4 (at 0·05 and 0·1%, w/w) were used as reducing agents. During ripening the redox potential of the cheeses made with the reducing agents did not differ significantly from the control cheese (E h ≈ -120 mV) while the cheeses made with 0·1 and 0·05% KIO3 had a significantly higher and positive redox potential in the first month of ripening. Cheese made with 1% KIO3 had positive values of redox potential throughout ripening but no starter lactic acid bacteria survived in this cheese; however, numbers of starter organisms in all other cheeses were similar. Principal component analysis (PCA) of the volatile compounds clearly separated the cheeses made with the reducing agents from cheeses made with the oxidising agents at 2 month of ripening. Cheeses with reducing agents were characterized by the presence of sulphur compounds whereas cheeses made with KIO3 were characterized mainly by aldehydes. At 6 month of ripening, separation by PCA was less evident. These findings support the hypothesis that redox potential could be controlled during ripening and that this parameter has an influence on the development of cheese flavour.

Measurement of pH micro-heterogeneity in natural cheese matrices by fluorescence lifetime imaging.

Cheese, a product of microbial fermentation may be defined as a protein matrix entrapping fat, moisture, minerals and solutes as well as dispersed bacterial colonies. The growth and physiology of bacterial cells in these colonies may be influenced by the microenvironment around the colony, or alternatively the cells within the colony may modify the microenvironment (e.g., pH, redox potential) due to their metabolic activity. While cheese pH may be measured at macro level there remains a significant knowledge gap relating to the degree of micro-heterogeneity of pH within the cheese matrix and its relationship with microbial, enzymatic and physiochemical parameters and ultimately with cheese quality, consistency and ripening patterns. The pH of cheese samples was monitored both at macroscopic scale and at microscopic scale, using a non-destructive microscopic technique employing C-SNARF-4 and Oregon Green 488 fluorescent probes. The objectives of this work were to evaluate the suitability of these dyes for microscale pH measurements in natural cheese matrices and to enhance the sensitivity and extend the useful pH range of these probes using fluorescence lifetime imaging (FLIM). In particular, fluorescence lifetime of Oregon Green 488 proved to be sensitive probe to map pH micro heterogeneity within cheese matrices. Good agreement was observed between macroscopic scale pH measurement by FLIM and by traditional pH methods, but in addition considerable localized microheterogeneity in pH was evident within the curd matrix with pH range between 4.0 and 5.5. This technique provides significant potential to further investigate the relationship between cheese matrix physico-chemistry and bacterial metabolism during cheese manufacture and ripening.

Growth and location of bacterial colonies within dairy foods using microscopy techniques: a review.

The growth, location, and distribution of bacterial colonies in dairy products are important factors for the ripening and flavor development of cheeses, yogurts, and soured creams. Starter, non-starter, spoilage, and pathogenic bacteria all become entrapped in the developing casein matrix of dairy foods. In order to visualize these bacterial colonies and the environments surrounding them, microscopy techniques are used. The use of various microscopy methods allow for the rapid detection, enumeration, and distribution of starter, non-starter and pathogenic bacteria in dairy foods. Confocal laser scanning microscopy is extensively utilized to identify bacteria location via the use of fluorescent dyes. Further study is needed in relation to the development of micro- gradients and localized ripening parameters in dairy products due to the location of bacteria at the protein-fat interface. Development in the area of bacterial discrimination using microscopy techniques and fluorescent dyes/tags is needed as the benefits of rapidly identifying spoilage/pathogenic bacteria early in product manufacture would be of huge benefit in relation to both safety and financial concerns.

Development of defined microbial population standards using fluorescence activated cell sorting for the absolute quantification of S. aureus using real-time PCR.

In this article, four types of standards were assessed in a SYBR Green-based real-time PCR procedure for the quantification of Staphylococcus aureus (S. aureus) in DNA samples. The standards were purified S. aureus genomic DNA (type A), circular plasmid DNA containing a thermonuclease (nuc) gene fragment (type B), DNA extracted from defined populations of S. aureus cells generated by Fluorescence Activated Cell Sorting (FACS) technology with (type C) or without purification of DNA by boiling (type D). The optimal efficiency of 2.016 was obtained on Roche LightCycler(®) 4.1. software for type C standards, whereas the lowest efficiency (1.682) corresponded to type D standards. Type C standards appeared to be more suitable for quantitative real-time PCR because of the use of defined populations for construction of standard curves. Overall, Fieller Confidence Interval algorithm may be improved for replicates having a low standard deviation in Cycle Threshold values such as found for type B and C standards. Stabilities of diluted PCR standards stored at -20°C were compared after 0, 7, 14 and 30 days and were lower for type A or C standards compared with type B standards. However, FACS generated standards may be useful for bacterial quantification in real-time PCR assays once optimal storage and temperature conditions are defined.

Responses of Escherichia coli, Listeria monocytogenes, and Staphylococcus aureus to simulated food processing treatments, determined using fluorescence-activated cell sorting and plate counting.

Three common food pathogenic microorganisms were exposed to treatments simulating those used in food processing. Treated cell suspensions were then analyzed for reduction in growth by plate counting. Flow cytometry (FCM) and fluorescence-activated cell sorting (FACS) were carried out on treated cells stained for membrane integrity (Syto 9/propidium iodide) or the presence of membrane potential [DiOC2(3)]. For each microbial species, representative cells from various subpopulations detected by FCM were sorted onto selective and nonselective agar and evaluated for growth and recovery rates. In general, treatments giving rise to the highest reductions in counts also had the greatest effects on cell membrane integrity and membrane potential. Overall, treatments that impacted cell membrane permeability did not necessarily have a comparable effect on membrane potential. In addition, some bacterial species with extensively damaged membranes, as detected by FCM, appeared to be able to replicate and grow after sorting. Growth of sorted cells from various subpopulations was not always reflected in plate counts, and in some cases the staining protocol may have rendered cells unculturable. Optimized FCM protocols generated a greater insight into the extent of the heterogeneous bacterial population responses to food control measures than did plate counts. This study underlined the requirement to use FACS to relate various cytometric profiles generated by various staining protocols with the ability of cells to grow on microbial agar plates. Such information is a prerequisite for more-widespread adoption of FCM as a routine microbiological analytical technique.

Physiological response of Bacillus cereus vegetative cells to simulated food processing treatments.

Vegetative cells of the spore-former Bacillus cereus were exposed to a number of treatments commonly used in commercial food preparation or during equipment cleaning and decontamination. Treated suspensions were then analyzed for reductions (CFU per milliliter) by plate counting and changes in levels of ATP and ADP released from cells with a bioluminescence-based assay. With the use of flow cytometry (FCM), the physiological status of individual cells before and after exposure to treatments was determined by staining of control and treated cells with three pairs of physiological dyes (SYTO 9/propidium iodide, carboxyfluorescein diacetate/Hoechst 33342, and C12-resazurin/SYTOX Green). Good agreement was found between plate counting and FCM. In general, treatments giving rise to the highest count reductions also had the greatest effects on cell membrane permeability (measured with the use of propidium iodide or SYTOX Green), esterase activity (measured with carboxyfluorescein diacetate), or redox activity (C12-resazurin). FCM data demonstrated the extent of heterogeneity of vegetative cell responses to treatments in, for example, the treatment with 5% H2O2, which caused a 6-log reduction in which approximately 95% of the population was composed of membrane-damaged cells (as reflected by their permeability to SYTOX Green), whereas in treatment with 0.09% (wt/vol) potassium sorbate, which caused only a 1-log reduction, not more than 40% of cells were membrane damaged. The approaches described in this work can be applied to gain a greater understanding of bacterial responses to food control measures, generate more accurate inactivation models, or screen novel prospective food control measures.

Monitoring growth phase-related changes in phosphatidylcholine-specific phospholipase C production, adhesion properties and physiology of Bacillus cereus vegetative cells.

The physiological status and metabolic heterogeneity of Bacillus cereus cells within a culture during an 8-h batch fermentation process was measured using flow cytometry (FCM). Concurrently, production of the toxin, PC-PLC, and the extent of cell adhesion of live and dead cells were monitored using novel fluorescent assays. Flow cytometry analysis detected growth phase-related changes in the physiological profiles of cells over the course of the fermentation, with variation in the percentage of cells displaying membrane damage and intracellular esterase and redox activities. As the exponential phase proceeded, populations became more uniform in terms of protein content as measured using FCM in tandem with a cell tracking dye, with the majority of cells becoming membrane intact, esterase positive and redox active. PC-PLC activity appeared strongly related to cell density. Permeabilisation of cells was accompanied by a loss in adherent properties, while 25-100% of cells with intracellular esterase activity possessed adhesion properties. Cells in late exponential phase appeared to have reduced adherence properties compared to cells in early exponential or lag phase. As well as demonstrating the utility of FCM for measuring heterogeneity in terms of cell physiological status throughout the course of batch cultures, the methods utilised in this study could be used to relate processes such as toxin production or cell adhesion to cell physiological state.

Bacillus cereus endospores exhibit a heterogeneous response to heat treatment and low-temperature storage.

Bacillus cereus endospores were challenged by heat treatments simulating typical domestic/industrial cooking regimes and the resulting effects on germination, viability and sub-lethal heat damage determined using differential plate counting on a rich versus selective medium, flow cytometry (FCM), beta-D-glucuronidase (GUD) activity and OD(600) measurement. Additionally, these techniques were used to investigate the effect on endospores of storage in a non-nutrient medium at 4 degrees C for 1 month. Plate counting revealed that heating generated sub-populations of sub-lethally damaged endospores, with the more severe heat treatments generating larger proportions of sub-lethally damaged endospores. These findings were also reflected in FCM analyses, which detected large amounts of heterogeneity among the populations of heat-treated endospores and uncovered differences in the proportions of membrane-damaged endospores and those displaying esterase activity pre- and post-treatment. Plate count data suggested that both the control and heat-treated endospores lost viability during storage, with FCM data indicating that the proportion of membrane-damaged endospores increased and those displaying the esterase activity decreased. The FCM, GUD and OD(600) data suggested that germination rates decreased with the increasing severity of heat treatment. This study demonstrates that a combination of plate counting and FCM can be used to detect heterogeneity in the response of endospores to insults.

The use of flow cytometry to study the germination of Bacillus cereus endospores.

BACKGROUND: At present the study of endospore germination is conducted using microbiological methods which are slow and yield data based on the means of large heterogeneous populations. Flow cytometry (FCM) offers the potential to rapidly quantify and identify germination and outgrowth events for large numbers of individual endospores. METHODS: Standard methods were employed to arrest the germination of Bacillus cereus endospores at defined stages. Endospores were then stained with SYTO 9 alone or carboxyfluorescein diacetate (CFDA) together with Hoechst 33342 and analysed using FCM. Comparisons were made between FCM as a method to measure germination rate and standard microbiological techniques. RESULTS: Germinating endospores displayed increases in permeability to SYTO 9 and hydrolysis of CFDA compared with controls. Statistically significant correlations were found between the standard plate count method and both FCM methods for measuring the percentage of germinating and outgrowing endospores up to 75 min after addition of germinant. CONCLUSIONS: Using FCM, the percentage of germinating or outgrowing endospores at various time points during germination and/or outgrowth can be quantified. FCM with CFDA/Hoechst 33342 staining may be used to estimate overall germination rate, whereas FCM with SYTO 9 staining may be used to quantify ungerminated, germinating and outgrowing endospores.

Starter strain related effects on the biochemical and sensory properties of Cheddar cheese.

A detailed investigation was undertaken to determine the effects of four single starter strains, Lactococcus lactis subsp. lactis 303, Lc. lactis subsp. cremoris HP, Lc. lactis subsp. cremoris AM2, and Lactobacillus helveticus DPC4571 on the proteolytic, lipolytic and sensory characteristics of Cheddar cheese. Cheeses produced using the highly autolytic starters 4571 and AM2 positively impacted on flavour development, whereas cheeses produced from the poorly autolytic starters 303 and HP developed off-flavours. Starter selection impacted significantly on the proteolytic and sensory characteristics of the resulting Cheddar cheeses. It appeared that the autolytic and/or lipolytic properties of starter strains also influenced lipolysis, however lipolysis appeared to be limited due to a possible lack of availability or access to suitable milk fat substrates over ripening. The impact of lipolysis on the sensory characteristics of Cheddar cheese was unclear, possibly due to minimal differences in the extent of lipolysis between the cheeses at the end of ripening. As anticipated seasonal milk supply influenced both proteolysis and lipolysis in Cheddar cheese. The contribution of non-starter lactic acid bacteria towards proteolysis and lipolysis over the first 8 months of Cheddar cheese ripening was negligible.

Starter bacteria are the prime agents of lipolysis in cheddar cheese.

To assess the contribution of starter lactic acid bacteria (LAB) to lipolysis in Cheddar cheese, the evolution of free fatty acids (FFAs) was monitored in Cheddar cheeses manufactured from pasteurized milks with or without starter. Starter-free cheeses were acidified by a combination of lactic acid and glucono-delta-lactone. Starter cultures were found to actively produce FFAs in the cheese vat, and mean levels of FFAs were significantly higher in starter cheeses over ripening. The contribution of nonstarter LAB toward lipolysis appears minimal, especially in starter-acidified cheeses. It is postulated that the moderate increases in FFAs in Cheddar cheese are primarily due to lack of access of esterase of LAB to suitable lipid substrate. The results of this study indicate that starter esterases are the primary contributors to lipolysis in Cheddar cheese made from good quality pasteurized milk.

Evidence of a relationship between autolysis of starter bacteria and lipolysis in cheddar cheese during ripening.

Cell viability, autolysis and lipolysis were studied in Cheddar cheese made using Lactococcus lactis subsp. cremoris AM2 or Lactococcus lactis subsp. cremoris HP. Cheddar cheese was made in triplicate over a 3 month period and ripened for 238 days at 8 degrees C. Cell viability in cheese was lower for AM2 (a non-bitter strain) than for strain HP (a bitter strain). Autolysis, monitored by the level of the intracellular marker enzyme, lactate dehydrogenase (EC 1.1.1.27) in cheese 'juice' extracted by hydraulic pressure, was much greater in the cheese made using AM2 than that made with HP. Lipolysis was determined by the increase during ripening of individual free fatty acids (FFA) from butyric (C4:0) to linolenic acid (C18:3) measured using a high performance liquid chromatographic technique. Levels of individual FFA from butyric (C4:0) to linolenic (C18:3) acids increased significantly (P<0.05) during ripening in cheeses made with either starter culture. Palmitic (C16:0) and oleic (C18:1) acids were the most abundant FFA throughout ripening in all cheeses. Levels of caprylic (C8:0), myristic (C14:0), palmitic (C16:0) and stearic (C18:0) acids were significantly higher (P<0.05) in cheeses manufactured with Lc. lactis subsp. cremoris AM2 than in cheeses manufactured with Lc. lactis subsp. cremoris HP. Differences in levels of lipolysis between strains was not due to differences in the specific lipolytic or esterolytic activities in cell free extracts of the strains as measured by activity on triolein (lipase) and p-nitrophenylbutyrate (esterase) substrates. Therefore, evidence is provided for a relationship between the extent of starter cell autolysis and the level of lipolysis during Cheddar cheese ripening.