Pesquisas sobre Listeria monocytogenes em indústrias de produtos lácteos no Brasil são escassas. Três laticínios (A, B e C) produtores de queijos do Estado de São Paulo foram monitorados para a presença de L. monocytogenes no período de outubro/2008 a setembro/2009. Foram realizadas 12 coletas, correspondentes a 12 lotes de queijo produzidos, sendo quatro de cada laticínio. Em cada laticínio, as visitas foram realizadas com intervalos de aproximadamente 2 meses entre cada uma. Foram analisadas 393 amostras, sendo 201 de superfícies com e sem contato com alimentos e 192 de alimentos (leite cru e pasteurizado e queijo) água e salmoura, para pesquisa de L. monocytogenes. As análises foram realizadas de acordo com o método do Food and Drug Administration (FDA). Os resultados confirmam a presença de Listeria spp nas instalações dos três laticínios. L. monocytogenes, L. innocua, L. seeligeri e L. welshimeri foram as espécies isoladas neste estudo. Especificamente a espécie L. monocytogenes não foi encontrada no laticínio A, entretanto, o microrganismo foi isolado de 12,5% das amostras do laticínio B e de 9,1% do laticínio C. L. monocytogenes não foi isolada do leite cru dos silos, do leite pasteurizado, da água e dos queijos Minas frescal, nos 3 laticínios. Porém, no laticínio C, L. monocytogenes foi isolada de amostras de queijo Prato que foram incluídas apenas na 4ª coleta deste laticínio, além de ter sido isolada de amostras de salmoura. As maiores prevalências de contaminação por L. monocytogenes ocorreram em superfícies sem contato com alimentos, sendo positivas 51,6% das amostras do laticínio B e 21,7% do laticínio C. Em ambos os laticínios a bactéria também foi isolada de superfícies com contato com alimentos. Os resultados fornecem informações detalhadas dos pontos prioritários para o desenvolvimento de estratégias de controle de L. monocytogenes em laticínios e mostram a importância de programas de monitoramento ambiental do patógeno, mesmo em pequenas indústrias. Os 85 isolados identificados como L. monocytogenes revelaram-se de quatro sorotipos: 1/2a, 1/2b, 1/2c e 4b, com predomínio do 4b, em ambos os laticínios, o que é preocupante para a saúde pública. Com base nos resultados de PFGE (perfis combinados ApaI e AscI), 40 perfis (pulsotipos) foram obtidos. Pulsotipos foram isolados repetidamente entre coletas nos laticínios B e C, sugerindo persistência de linhagens nos laticínios. Apesar dos laticínios serem distantes e independentes, um pulsotipo foi compartilhado entre ambos. O laticínio A apresentou contaminação por mais de um pulsotipo de L. seeligeri e houve isolamento repetido de um pulsotipo dessa espécie, entre as coletas, sugerindo adaptação da bactéria e necessidade de controle do gênero Listeria nessa indústria. A ocorrência de um mesmo pulsotipo de L. monocytogenes com sorotipos diferentes (1/2b e 4b) mostra que a sorotipagem deve acompanhar análises mais refinadas como as de natureza genotípica. / Listeria monocytogenes surveys in cheese manufacturing plants in Brazil are rare. Three cheese manufacturing plants (A, B and C) in São Paulo state were monitored for the presence of Listeria monocytogenes during the period of October/2008 - September/2009. Twelve samples surveys were taken corresponding to 12 cheese lots produced, four in each plant. In each cheese plant, the samples were taken at intervals of approximately 2 months. There were 393 samples analyzed, 201 from surfaces with and without contact with food and 192 of food (raw and pasteurized milk and cheese), water and brine, with the objective of searching for L. monocytogenes. The analyses were performed in accordance with Food and Drug Administration (FDA) method. The results confirmed the presence of Listeria spp in the facilities of three plants. L. monocytogenes, L. innocua, L. seeligeri and L. weshimeri were the species isolated in this study. Specifically the L. monocytogenes specie was not isolated from plant A. However, the microorganism was isolated in 12.5% of the samples from plant B and 9.1% from plant C. Listeria monocytogenes was not isolated from raw milk in storage tanks, pasteurized milk, water or Minas frescal cheese samples from the three plants. Nevertheless, in plant C, L. monocytogenes was isolated in Prato cheese that was included only in the 4th sampling survey and also from the brine samples. The major prevalence of contamination by L. monocytogenes occurred on surfaces without contact with food, with 51.6% of the samples positive from plant B and 21.7% from plant C. In both plants, the microorganism was also isolated from food contact surfaces. The results provide detailed information about the critical points for the development of L. monocytogenes control strategies in cheese processing plants and, moreover, show the relevance of sampling programs of the pathogen, even in small cheese processing plants. The 85 isolates identified as L. monocytogenes were classified in four serotypes: 1/2a, 1/2b, 1/2c and 4b, with 4b dominating in both cheese plants, which is of concern to human health. On the basis of PFGE results (combined profiles ApaI and AscI), 40 profiles (pulsotypes) were found. Pulsotypes were isolated repeatedly among sampling surveys in plants B and C, suggesting persistence of lineages in the plants. Despite these plants being distant and independent, one pulsotype was shared between them. Plant A presented contamination by more than one pulsotype of L. seeligeri and there was a repetitive isolation of one pulsotype of this specie among samplings, suggesting adaptation of the bacterium and the need for control of the Listeria genus in this plant. The occurrence of one single pulsotype of L. monocytogenes with different serotypes (1/2b and 4b) show that serotyping should follow more refined analyses as the ones of genotypic nature.

A legislação brasileira permite o uso de leite cru na fabricação de queijos curados se o período de cura for superior a 60 dias (a 5°C ou mais). Entretanto, não há evidência científica sólida de que durante a cura ocorre suficiente inativação de Brucella abortus, sob a perspectiva da segurança dos alimentos. Além disso, não há metodologia oficial para quantificação de brucelas em matriz alimentar. Desta forma este projeto propõe um protocolo para estudos da curva de decaimento de Brucella abortus durante a cura de queijos. Três partidas de queijo tipo parmesão foram feitas com leite tipo A pasteurizado, artificialmente contaminado com uma cepa pouco patogênica de Brucella abortus (1119-3). O queijo foi curado a 18°C e analisado a cada 3-4 dias até o 60º dia ou até que duas análises consecutivas fossem negativas. As amostras foram submetidas à pesquisa quantitativa de brucela em meio Farrel (36°C/3 dias). A média do Tempo de Redução Decimal (D18°C), ponderado pelas incertezas, foi de 4,31 ± 0,49 dias. Estes resultados são preliminares e é necessário realizar mais análises para gerar dados mais precisos sobre esse parâmetro, especialmente usando cepas de campo. O Tempo de Redução Decimal associado à cura do queijo é um parâmetro importante para os modelos de avaliação de risco da transmissão de brucelose pelo consumo de queijo. / Brazilian legislation allows the use of raw milk to produce ripened cheeses, when the ripening period is over 60 days (at 5°C or above). However, there is not solid scientific evidence that the ripening process inactivates enough amounts of Brucella abortus, from the food safety perspective. In addition, there is no official methodology for brucela counting in food matrix. So, this project proposes a protocol to study the decay curve of Brucella abortus during cheese ripening. Three batches of parmesan type cheese were made with pasteurized milk artificially contaminated with a less pathogenic strain of Brucella abortus (1119-3). The cheese was ripened at 18°C and analyzed each 3-4 days until 60th or until two consecutive negative results. The samples were submitted to quantitative culture for brucela in Farrel medium (36°C/3 days). The average of Decimal Reduction Time (D18°C), weighted by uncertainties, was 4,31 ± 0,49 days. This is a preliminary result and we need to run more samples to generate more accurate data about this parameter, specially using field strains. The Decimal Reduction Time associated with ripening step is an important data to input risk assessment models for cheese-born brucelosis.

O presente estudo teve por finalidade avaliar os efeitos da contagem de células somáticas (CCS) no leite cru (nos níveis de 100-200.000, 400-500.000 e >800.000 céls./mL) sobre as características físicas, químicas, microbiológicas, sensoriais e funcionais dos queijos Minas Frescal e tipo Mussarela. Utilizou-se um delineamento experimental em parcelas subdivididas em blocos, considerando-se a contagem de células somáticas como efeito principal, os dias de análise como subparcelas e os processamentos como blocos. Cada tipo de leite foi obtido da ordenha de animais previamente selecionados de acordo com o nível de células somáticas. As etapas de elaboração dos queijos incluíram a pasteurização do leite (65º C, 30 minutos), adição de cloreto de cálcio, fermento (para a Mussarela) e coalho, coagulação e obtenção do coágulo, dessoragem, salga na massa, filagem, moldagem, salga na salmoura (para a Mussarela), e embalagem dos produtos. Os queijos foram mantidos em B.O.D. a 4 ºC e avaliados nos dias 2, 9, 16, 23 e 30 após a fabricação. A seqüência de elaboração dos queijos Minas e tipo Mussarela foi repetida 5 e 3 vezes, respectivamente, para cada tratamento. As análises realizadas nos queijos foram: pH, acidez, percentuais de gordura, matéria seca (MS), cinzas, nitrogênio total (NT), nitrogênio não protéico (NNP), nitrogênio não caseinoso (NNC), índice de proteólise, ácidos graxos livres (AGL), textura, avaliação sensorial, propriedades funcionais (capacidade de derretimento a 107 ºC e percentual de óleo livre - somente no queijo tipo Mussarela), contagem de mesófilos, psicrotróficos e número mais provável a 35 e 45 ºC. No queijo Minas Frescal, não houve interação entre CCS e dias de armazenamentos nas avaliações físico-químicas, microbiológicas, funcionais e índice de lipólise, porém houve efeito significativo (P<0,05) para índices de proteólise e profundidade da proteólise. Já para mesófilos, psicrotróficos, acidez, matéria seca, firmeza e ácidos graxos livres houve efeito significativo (P<0,05) ao avaliar dias de armazenamento. A avaliação sensorial, no primeiro dia de análise, não apresentou diferença significativa nos atributos, porém o queijo de alta CCS (> 800.000 céls./mL) diferiu dos demais no dia 30, em que recebeu menor nota em todas as características. No queijo tipo Mussarela, houve interação entre CCS e dias de armazenamento apenas para a capacidade de derretimento, resultando em aumento do derretimento no decorrer do tempo. Para as características de pH, ácidos graxos livres, índices de proteólise, extensão e profundidade da proteólise, mastigabilidade e elasticidade houve efeito significativo (P<0,05) para dias de armazenamento. Entre os atributos avaliados na sensorial, apenas a aparência apresentou diferença entre tratamentos. O rendimento dos queijos, Minas frescal e tipo Mussarela, não foram influenciados pela quantidade de células somáticas dos leites. Os resultados do trabalho indicaram que o leite destinado à fabricação dos queijos Minas e tipo Mussarela deve apresentar CCS até 400-500.000 céls./mL, de maneira a evitar alterações na qualidade dos produtos ao longo do período de armazenagem. / The aim of the present study was to evaluate the effect of somatic cells counts (SCC) in raw milk (at levels of 100-200,000, 400-500,000 and 800,000 cells./mL) on physical, chemical, microbiological, sensorial and functional characteristics of Minas frescal and Mozzarella type cheeses. A completely randomized block design was used, considering SCC as the main effect, the days of analysis as sub parcels and the processing batches as the blocks. Each type of milk was obtained from cows previously selected according to its individual SCC. The manufacture of cheeses included: pasteurization of milk (65ºC, 30 minutes), addition of calcium chloride, starter culture (for Mozzarella) and rennet, coagulation and separation of the curd, whey drainage, salting (for Minas cheese), stretching of the curd, kneading and salting in brine (for Mozzarella), and packing the products. The cheeses were stored in a B.O.D. oven at 4ºC and evaluated on days 2, 9, 16 23 and 30 after the manufacture. The sequence of elaboration of the Minas frescal and Mozzarella cheeses was repeated 5 and 3 times, respectively, for each treatment. The analyses carried out in the cheeses were: pH, acidity, percentages of fat, dry matter (DM), ashes, total nitrogen (NT), non-protein-nitrogen (NPN), non-casein nitrogen (NCN), proteolysis, free fatty acids (FFA), texture, sensorial evaluation, functional properties (melting capacity the 107 ºC and percentage of free oil - only in the cheese Mozzarella), mesophile and psychrotrophic counts and the most probable number at 30 and 45ºC. For Minas frescal cheese, no interaction was found between SCC and days of storage when considering the data on chemical, physical, microbiological, functional and lipolysis index. However, a significant effect (P<0.05) was observed for proteolysis and depth of proteolysis. For mesophiles, psychrotrophics, acidity, dry matter, firmness and free fatty acids, there was a significant effect (P<0.05) for days of storage. The sensorial evaluation made on the first day of analysis did not present significant differences for all attributes. However, the Minas cheese made with high SCC (800,000 cells/mL) differed on day 30, when it received worse grades for all the attributes. For the Mozzarella cheese, an interaction between SCC and days of storage was observed only for the melting capacity, which resulted in an increment of the melting along the time of storage. The parameters of pH, free fatty acid, proteolysis, extension and depth of proteolysis, springiness and elasticity had significant effect (P<0.05) for days of storage. Amongst the attributes evaluated in the sensorial, only the appearance presented difference between treatments. The yield of Minas frescal and Mozzarella cheese was not influenced by the amount of somatic cells in the original milk. Results indicated that milk used for the manufacture of Minas frescal and Mozzarella cheeses should present SCC up to 400-500,000 cells/mL, in order to avoid quality changes in those products during storage.

A legislação brasileira permite o uso de leite cru na fabricação de queijos curados se o período de cura for superior a 60 dias (a 5°C ou mais). Entretanto, não há evidência científica sólida de que durante a cura ocorre suficiente inativação de Brucella abortus, sob a perspectiva da segurança dos alimentos. Além disso, não há metodologia oficial para quantificação de brucelas em matriz alimentar. Desta forma este projeto propõe um protocolo para estudos da curva de decaimento de Brucella abortus durante a cura de queijos. Três partidas de queijo tipo parmesão foram feitas com leite tipo A pasteurizado, artificialmente contaminado com uma cepa pouco patogênica de Brucella abortus (1119-3). O queijo foi curado a 18°C e analisado a cada 3-4 dias até o 60º dia ou até que duas análises consecutivas fossem negativas. As amostras foram submetidas à pesquisa quantitativa de brucela em meio Farrel (36°C/3 dias). A média do Tempo de Redução Decimal (D18°C), ponderado pelas incertezas, foi de 4,31 ± 0,49 dias. Estes resultados são preliminares e é necessário realizar mais análises para gerar dados mais precisos sobre esse parâmetro, especialmente usando cepas de campo. O Tempo de Redução Decimal associado à cura do queijo é um parâmetro importante para os modelos de avaliação de risco da transmissão de brucelose pelo consumo de queijo. / Brazilian legislation allows the use of raw milk to produce ripened cheeses, when the ripening period is over 60 days (at 5°C or above). However, there is not solid scientific evidence that the ripening process inactivates enough amounts of Brucella abortus, from the food safety perspective. In addition, there is no official methodology for brucela counting in food matrix. So, this project proposes a protocol to study the decay curve of Brucella abortus during cheese ripening. Three batches of parmesan type cheese were made with pasteurized milk artificially contaminated with a less pathogenic strain of Brucella abortus (1119-3). The cheese was ripened at 18°C and analyzed each 3-4 days until 60th or until two consecutive negative results. The samples were submitted to quantitative culture for brucela in Farrel medium (36°C/3 days). The average of Decimal Reduction Time (D18°C), weighted by uncertainties, was 4,31 ± 0,49 days. This is a preliminary result and we need to run more samples to generate more accurate data about this parameter, specially using field strains. The Decimal Reduction Time associated with ripening step is an important data to input risk assessment models for cheese-born brucelosis.

This thesis examined the efficacy of various sampling and detection methods used for environmental monitoring of Listeria species on wooden surfaces used for cheese aging. Government agencies including the Food and Drug Administration (FDA) and United States Department of Agriculture (USDA) recommend enrichment methods coupled with use of environmental sponges and swabs. Our study compared efficacy of sponge swabs manufactured by 3M™ and World Bioproducts. There is a lack of research validating the best performing swab type and enrichment method combination that is sensitive when used on rough, porous surfaces. The sensitivity of these environmental sampling tools and methods are critical considerations to effectively monitor the presence of Listeria species on wooden boards used during aging of artisan cheese. Seasoned spruce wooden shelves, cut into 100cm2 replicates, were spot inoculated with varying concentrations of Listeria species inocula, the Listeria species strains consisted of two L. monocytogenes strains and a Green Florescence Protein (GFP) expressing strain of L. innocua. The inoculated wooden surface was swabbed with three environmental sampling sponge/swab formats (World Bioproducts© EZ ReachTM environmental swabs (WBEZ) with HiCap (WBHC) and Dey-Engley (WBDE) neutralizing broths; and 3MTM environmental swabs (3MTM) with Dey-Engley neutralizing broth). Enumeration methods were used to determine the low target limits of detection. Once the low target concentrations were identified, five enrichment methods consisting of 3MTM Listeria Environmental Plate, FDA, Dual Enrichment, modified USDA, and modified FDA were challenged against low concentrations of Listeria species inocula (0.01 cfu/cm2, 0.1 cfu/cm2, 1 cfu/cm2) and the three environmental sponge swab formats. Performance of the swab formats was assessed by collection of naturally contaminated environmental samples (n=405) from dairy farm environments, swabbing where wooden surfaces existed, and analyzed using the most effective enrichment methods found from previous experiments. Lastly, the wooden surfaces and sponge swabs were observed under a Florescent Microscope using GFP L. innocua to visually determine how each sponge material of the 3M™ and World Bioproducts recovered the inocula. When wood surfaces were inoculated at high concentration levels of Listeria spp., all swab formats performed equally for detecting Listeria. Success of positive recovery at low concentrations was variable, where enrichment methods and swabs were not dependent on each other. The swab format that worked best for detecting low levels of Listeria species was the WBDE sponge swab. The WBDE swab also performed the best in dairy farm environmental sampling. The m-USDA enrichment method was found to be most effective in recovery and repair of low and potentially injured Listeria spp. Wooden surfaces are rough and porous and should be taken into consideration when creating an environmental sampling plan for these food contact surfaces. All swabs and methods performed with only slight variation, but the variation could be significant when monitoring wooden shelves with low level contamination of Listeria species. Artisan cheesemakers who use wooden shelves during the aging of their cheese, should ensure use of the most sensitive detection methods.

Raw milk cheese has unique flavor and textural characteristics not obtainable in cheese from pasteurized milk. Several specialty cheeses made from raw milk are marketed worldwide, especially in Europe. However, because of safety concerns, many countries have imposed stringent restrictions on production and sale of raw milk cheeses. The purpose of this thesis research was to use high pressure (HP) treatment as a novel alternative for conventional pasteurization so that raw milk quality cheese could be produced without compromising food safety. The specific objectives of this research were to evaluate (i) the effect of HP treatment of milk on its coagulation and gelation characteristics, (ii) the destruction kinetics of microorganism and enzymes in milk, (iii) cheese making characteristics of HP treated milk as compared to the raw, pasteurized and micro-filtered milk (controls) and, finally (iv) to evaluate ripening characteristics of cheddar cheese made from HP treated milk in comparison with the controls. / Three coagulation parameters of milk---lag time, mean coagulation rate, and inflexion time (time for reaching the point of maximum coagulation rate)---were evaluated as a function of pressure (200--400 MPa), temperature (3--21°C) and holding time (10--110 min) using a response surface methodology. In general, the lag time and inflexion time decreased while the mean coagulation rate increased with an increase in pressure, holding time or a decrease in temperature: The rennet gel characteristics were evaluated as gel strength (GS) and water-holding capacity (WHC). With a decrease in pressure level, temperature and holding time, there was a decrease in water-holding capacity and an increase in the gel-strength of the rennet curd. (Abstract shortened by UMI.)

The application of high pressure (HP) processing on raw milk cheese was investigated in order to assure safety and improve quality. Fresh raw milk cheese inoculated with contaminant, spoilage and pathogenic microorganisms (Escherichia coli K-12, E. coli O157:117 and Listeria monocytogenes), as well as natural micro-flora, were subjected to UP treatment at selected pressures (200--400 MPa) for various holding times (0 to 100 min), or number of pulses. HP destruction of microorganisms followed the dual effect destruction behavior involving a step change in the population due to a pressure pulse (pulse effect, PE) and a first order rate log-linear kinetics during the pressure hold. The pressure dependency of kinetic parameters followed the pressure destruction time (PDT) and Arrhenius type models. / The results suggest that high pressure treatment as a powerful tool for microbial control do not result in major change in raw milk cheese quality properties (color and texture). It would thus be an effective method of inactivation of spoilage bacteria and pathogens for ensuring safety and keeping the quality of raw milk cheese.

Curd syneresis, a critical step in cheesemaking, directly influences the quality of cheese. The syneresis process is empirically controlled in cheese manufacturing plants. A sensor technology for this step would improve process control and enhance cheese quality. A light backscatter sensor with a Large Field of View (LFV) was tested using a central composite design over a broad range of cheese process conditions including milk pH, calcium chloride addition level, milk fat to protein ratio, temperature, and a cutting time factor (β). The research objectives were to determine if the LFV sensor could monitor coagulation and syneresis steps and provide information for predicting pressed curd moisture. Another objective was to optimize cheese yield and quality. The LFV sensor was found to monitor coagulation and syneresis and provide light backscatter information for predicting curd moisture content. A model for relating final curd moisture content with light backscatter response was developed and tested. Models for predicting whey fat losses, pressed curd moisture, and cheese yield were successfully developed (R2>0.75) using the test factors as independent variables. This was the first attempt to develop a technology for controlling pressed curd moisture using a sensor to monitor the syneresis step.

Graduation date: 1993

Thesis (Ph.D.)--University of Western Sydney, 2007. / A thesis submitted to the University of Western Sydney, College of Health and Science, Centre for Plant and Food Science, in fulfilment of the requirements for the degree of Doctor of Philosophy. Includes bibliographical references.