Harmonization and standardization -...

técnicas de LABORATORIO 736 Nº 427 DICIEMBRE 2017

food controlfood control

Harmonization and standardization in food molecular microbiology:

Listeria monocytogenes and Salmonella spp. cases

David Rodríguez-LázaroÁrea de Microbiología. Departamento de Biotecnología y Ciencia de los Alimentos, Facultad de Ciencias, Universidad de [email protected] · @rodlazda

icrobial analysis is a prosperous, lucrative and growing industry, with a

global market of more than 2 billion analyses performed in 2014. Food microbial testing comprised more than 966 million analyses in 2013, with annual global revenues of $3.05 billion for the 2,350 food contract test labs worldwide. It is evident that the food industry requires the service of third party food contract labs to address its main demands: selectivity of the methodology, short time to delivery of fit for purpose final results, and reduction of the associated cost of analysis through efficient working practices. As a result, any alternative methodology, and in particular molecular microbial diagnostic methods must try to meet these objectives. For that two major

actions are needed: the harmonization of the route from the design and development of a molecular microbial diagnostic method to its final effective implementation in routine food labs, and global effort towards standardization of molecular methods. In this presentation, we review the current situation on harmonisation and standardization of molecular microbial diagnostic method, and the particular cases of Listeria monocytogenes and Salmonella spp. will be presented.

Introduction: The market of food microbiology diagnostics

Microbial analysis is a prosperous, lucrative and growing industry, with a global market of more than 2 billion

analyses in 2014 (Strategic consulting Inc, 2014a), with food and water testing areas comprising more than 70% (Strategic consulting Inc, 2014a). Food microbiology testing comprised more than 966 million analyses in 2013, with an annual increase in the total test volumes of 128% (StrategicConsulting Inc, 2013). Interestingly, the percentage of the testing volume devoted to foodborne pathogens has been historically much lower than that for index/indicator microorganisms, but the derived cost is, however, higher, and the increasing annual demand of those analysishasgrown(from13.7%in1998to 23.2% in 2013) (Strategic Consulting Inc, 2013).

The food industry prefers not to perform microbiology tests in situ by its own laboratories and personnel, but prefer

M

(La armonización y la estandarización en microbiología molecular alimentaria: los casos de Listeria monocytogenes y Salmonella spp)


food control

to submit them to third party contract labs, and this trend is ever increasing (annual increase 9.4% worldwide) (Strategic consulting Inc, 2013). As an example, 61% of the Salmonella testing from the food industry was performed in food contract labs in the USA in 2013 (from 37% in 2001) (Strategic consulting Inc, 2014b). Globally, it represents a market of 2,350 food contract test labs worldwide and annual global revenues of $3.05 billion in 2013 (a 5-year increase of 156.41%) (Strategic consulting Inc, 2013). Similarly, the availability of diagnostic tests and consumables for food microbiology testing based on different technologies has increased during the last two decades. The number of companies providing diagnostic test kits has grown more than 400% in this period and the number of kits and tests available in the market has also increased exponentially; diagnostic companies prepare diagnostics tests in advance of legal microbiological requirements. However, the high number of companies in the market and the financial pressures affecting third party testing laboratories are likely to result in a reduction of the growth (and even a decrease) in the number of diagnostic companies and related test kits.

Evolutions of food microbiology diagnostics: From Petri dishes to PCR

Microbiology diagnostics is a relatively young microbiological sub-discipline. We have to go back to the late 19th century to find the basis of the current established approach for microbiology testing (and particularly for food microbiology diagnostics), i.e. the isolation of presumptive strains on solidmedia. In1881, a seminalpaperdescribed the use of solid media for study of pathogenic microorganisms (Koch,1881).Sincethen,ithasbecomethe gold standard, particularly in food microbiology (bacteriology); each major foodborne bacterial pathogen

possesses an international standard (i.e. ISO standard) for its detection and identification based on a final step of isolation in particular solid media (e.g. the ISO standard 11290-1 for detection of Listeria monocytogenes, or the ISO standard 6579-1 for detection of Salmonella). However, this methodological procedure is tedious and time-consuming, laborious, prone to errors due to massive handling of different plates and bottles in different steps. It is furthermore unable to detect bacterial pathogens in a particular physiological state namely viable but not cultivable (VBNC), in which bacterial strains are still viable and therefore able to pose a subsequent risk for consumers, but they cannot grow in solid media. In addition, the isolation of a particular strain in a given medium specific for a class of pathogen only represents a presumptive result (detection) and must be completed with biochemical and/or serological tests (identification). This can result in a significant delay before a confirmed result can be provided.

One century after the Robert Koch’s seminal paper, these drawbacks could befinally fulfilled in themid-1980’sbya key development, the polymerase chain reaction (PCR) technique (Saiki et al.,1985,1988;Mulliset al.,1986).At first, this technique, although very promising was still difficult to implement in routine analysis as it required the addition of new DNA polymerase in each cycle and the PCR equipment was still quite rudimentary. The arrival of DNA polymerases from extremophilic microorganisms such as Thermus aquaticus (the bacterium from which the Taq DNA polymerase is obtained) and new developments in PCR equipment (particularly more efficient Peltier systems) generated a biotechnological revolution: PCR has been used in more than 365,000 scientific publications (search ‘PCR’ in pubmed.com), and has been applied in many different areas due to its versatility, specificity and sensitivity, and particularly for microorganism identification (Rodríguez-Lázaro et

al., 2007; Rodríguez-Lázaro, and Hernández, 2013). A new PCR development, real-time PCR (qPCR) (Heid et al., 1996), further spurred that revolution and continues to do so. It represents a significant advance in many molecular techniques involving nucleic acids analysis allowing the monitoring of the synthesis of new amplicon molecules during the PCR by fluorescence (i.e. in real time), and not only at the end of the reaction (Rodríguez-Lázaro and Hernandez, 2013). Major advantages of qPCR are the reduced risk of carry-over contamination due to its closed-tube format (no post-amplification handling), fast and easy to perform analysis, high precision and accuracy, excellent selectivity, significantly higher reliability and sensitivity of the results (down to 1 microbial cell or genome equivalent per reaction) and extremely wide dynamic range of quantification (Rodríguez-Lázaro and Hernandez, 2013). Since Heid and coworkers’ seminal paper published in 1996 (Heid et al., 1996), the number of publications where qPCR is used has increased nearly exponentially. Similarly, the number of available platforms has moved from a handful in the late 1990’s to around 60 currently (2016) provided by more than 15 biotechnology companies (http://cyclers.gene-quantification.info). Similarly, the international bodies for standardisation in food analysis (the International Organisation for Standardisation –ISO, and the European Committee for Normalisation –CEN) have launched several ad hoc expert groups for the development of ISO-EN standards based on PCR (e.g. the TAG3 group for development of PCR based standards for detection of pathogens in food, and TAG 4 for the development of ISO standards for detection of enteric viruses in food by PCR). Currently 10 ISO PCR-related standards have been publishedw w w. i s o . o r g / i s o / h o m e / s e a r c h .htm?qt=food+pcr+microbiology

Harmonisation in the design and development of a novel methodology, and global efforts for standardisation


food control

The International Union of Food Science and Technology (IUFOST) held the 13th World Congress of Food Science and Technology in Nantes in September, 2006, with a symposium entitled “Analytical Methodology in Food Safety: current status, lessons learned and future challenges” in which different aspects of food safety were discussed (bacteria, viruses, parasites, antibiotic resistance, etc). A review paper was subsequently published (Rodríguez-Lázaro et al., 2007) in which the most important conclusions were included, and one of the main areas of discussion was the current challenges. Several aspects were defined including the development of rational and easy-to-use strategies for pre-amplification treatment of the food samples, the design and application of analytical controls, the unambiguous determination of viable forms and the development of strategies for the quantitative use of real time PCR (Rodríguez-Lázaro et al., 2007). It is quite surprising that these challenges have not all been met after almost one decade. Many scientific publications are still being published trying to address one or several of these aspects, indicating that the problems are still an issue requiring resolution. In addition to the challenges previously described, there are other aspects that are very important in the food chain (both for the food industry and third party food contract labs). Between them; two are really important: the harmonisation in the design and development of a novel methodology, and global efforts for standardisation.

Over the last 25 years the research community has devoted extensive resources in harnessing the evolving knowledge of the genome to the development of technologies for rapid and sensitive detection of different foodborne microbial targets. A considerable body of information and techniques has been built up thereby, but the methods have very often not been transferred to deployment in routine analysis. Thus, there has been little effective return to society from

the resources which it has provided. This is largely because hitherto there has been no systematically defined route which development of a method must follow from conception by the method developer to implementation by the analytical community. In the first instance a clearly defined business plan, proving the societal need and the market for the method should be produced. Next, the minimal performance criteria, including cost efficiency, necessary to fulfil the intended purpose must be acknowledged. Subsequently, progress towards development must follow a traceable and recorded path. The new method must be demonstrably valid, i.e. perform at least as well as any established standard, and be repeatable and reproducible. Communication of the method should be done in a systematic way, with preplanned dissemination and training activities leading to its widespread recognition among the necessary stakeholders.

Over the last decade, molecular detection methods have been developed at a relatively high pace. For instance, quantitative real-time PCR has become very popular as a detection tool. However, the perceived ease of use of the method and the frequent lack of understanding of all important parameters in the workflow contributing to accurate and precise results – both among users, authors and reviewers- has resulted in an accumulation of unreliable reports in the scientific literature. Among the most compelling stories are the retraction of the Science breakthrough paper of the year 2005 (Hwang et al., 2005) due to irreproducible qPCR results, and the USA lawsuit resulting after a publication of a report on the alleged causal link between the MMR vaccine and autism with children (Uhlmann et al., 2002). It turned out that the conclusions were false, based on anomalies in the lab when conducting and analyzing the qPCR data. To improve the quality

and transparency of experiment design, data-analysis and reporting of results, the Minimal Information for publication of Quantitative PCR Experiments (MIQE) guidelines were established in 2009 (Bustin et al., 2009). While adoption of the guidelines is going relatively well, with more than 4,800 citations (GoogleScholar, October 2016), the guidelines are focused on the application of qPCR in biomedical research. For diagnostic applications in other fields (including foodborne pathogens), wider and more flexible guidance is required. In addition, the MIQE-proposed universal data format RDML (Real-time PCR Data Markup Language) (Lefever et al., 2009) to store, exchange and report raw qPCR data is not compatible with data coming from applications such as digital PCR, high-resolution melting, and next-generation sequencing.

The requirement for good laboratory practice, quality assurance programs and accreditation, in its turn requires the availability of standard methods. Research laboratories are continually developing novel molecular diagnostic methods, and hundreds of new tests have been published, but very few have actually been implemented in end-user laboratories. In large part, this is because their customers increasingly demand that only accredited standard methods are used for analysis of foodborne microorganisms. As regards analysis for pathogens, standard protocols have been available for many years for culture-based microbiological methods, but standards based on alternative methods for analysis of foods have been formulated only recently. Successful reproduction of results, in the hands of different personnel under different laboratory conditions with various batches of reagents, is an absolute prerequisite for adoption of a nucleic acid amplification-based detection method as a standard (Hoorfar and Cook, 2003; Malorny et al., 2003). It is desirable for end-users and reference laboratories to have access to open-formula, non-


food control

commercial and non-proprietary assays for which the targets, performance characteristics and validation criteria are known (Hoorfar and Cook, 2003). However, while many published methods present in-house validation data, inter-laboratory reproducibility data is largely lacking. Such data are necessary to demonstrate the robustness of tests.

From a foodborne pathogen perspective, the early international activities aimed at developing standardized nucleic acid-based methods for detection, determined their repeatability and reproducibility (Abdulmawjood et al., 2004; D’Agostino et al., 2004; Josefsen et al., 2004; Malorny et al., 2004); however they were not fully validated as alternative methods according to ISO 16140 (ISO, 2003) to demonstrate that their performance characteristics were at least the equivalent of the culture-based standard methods (D’Agostino and Rodriguez-Lazaro, 2009). This is what is required to be able to convince the potential end-users of these methods’ effectiveness as reliable and robust alternatives that can stand alongside the “gold standard” method, and to be able to realise their full potential as tools in the continual battle against the ever-present threat of foodborne pathogens (D’Agostino and Rodriguez-Lazaro, 2009). Progress however has been made: PCR-based methods for detection of Salmonella spp. (Löfström et al., 2009; Malorny et al., 2007) and Campylobacter spp. (Krause et al., 2006) in foods have been fully validated, and in the case of the Salmonella method become a national (DIN) standard. At present, detection of norovirus in foods is not represented by an international standard; however the CEN TC 275 / WG6 / TAG4 is developing methods to detect norovirus and hepatitis A virus in leafy green vegetables, shellfish, and soft fruit. The methods which contain real time reverse transcription-PCR assays, are designed to be quantitative, and incorporate sample process and

amplification controls; publication of the IS/TS preliminary standards was in 2013, and publication of the full standardsisscheduledfor2018.

The cases of Listeria Monocytogenes and Salmonella

The European Union granted a multiyear collaborative research project BaseLine (www.baselineeurope.eu) to analyse and develop sampling strategies to support the European policies in food safety with the final goal of improve quantitative risk analysis. One of the main purpose of BaseLine was the validation and harmonization of alternative molecular methods for detection of main foodborne pathogens such as Salmonella spp. and Listeria monocytogenes. In particular, BaseLine made an international effort to validate an alternative method based on an ISO-compatible pre-enrichment coupled to bacterial DNA extraction and real-time PCR detection of Salmonella spp. and Listeria monocytogenes in two of the major food contamination sources; pork meat and cheese.

The validation study

The Istituto Superiore di Sanità (ISS) (Italy) was the organizing laboratory and led the studies for both Salmonella spp. and Listeria monocytogenes. Thirteen laboratories from seven European countries participated in the trial for Salmonella spp.: the University of Bologna (Italy); the National Veterinary Institute (Norway); the Centro Nacional de Tecnologia y Seguridad Alimentaria (Spain); the National Food Chain Safety Office (Hungary); the University of Zagreb (Croatia); the Instituto Tecnológico Agrario de Castilla y León (Spain); the University of Copenhagen (Denmark); the French Agency for food, environmental and occupational health safety, Anses(France), the Istituto Zooprofilattico Sperimentale (IZS) delle Venezie (Italy); the IZS del Lazio e

Toscana (Italy); the IZS della Lombardia e dell’Emilia Romagna (Italy) and the IZS del Mezzogiorno (Italy). Twelve laboratories from six European countries participated in the trial for Listeria monocytogenes: the Centro Nacional de Tecnologia y Seguridad Alimentaria (CNTA) (Spain); Instituto Tecnológico Agrario de Castilla y León (Spain); the Istituto Zooprofilattico Sperimentale (IZS) dell’Abruzzo e Molise (Italy); the Istituto Zooprofilattico Sperimentale (IZS) del Lazio e Toscana (Italy); the Istituto Zooprofilattico Sperimentale (IZS) della Lombardia e dell’Emilia Romagna (Italy); the Istituto Zooprofilattico Sperimentale (IZS) delle Venezie (Italy); the National Food Chain Safety Office (Hungary); the Norwegian Veterinary Institute (NVI) (Norway); University of Bologna (UniBO) (Italy); the University of Copenhagen (Denmark); and the University of Zagreb(Croatia). Each participant was provided with a standard operating procedure (SOP) for performance of this trial.

The organizing laboratory provided to the participating laboratories, materials (pork meat and reference material as Lenticule® discs in the case of Salmonella spp. or cheese and Lyophilized bacteria in the case of L. monocytogenes) and all reagents to perform Real-Time PCR. The organizing laboratory provided the foodstuffs and the lyophilized strains in ready to use, coded containers, as well as all reagents to perform the Real-Time PCR assay by a courier service. This means that the laboratories performing the analysis were blind to the actual content of the pathogens in each sample.

The laboratories performed the artificial contaminationof8 food samplesusing8 blind coded lyophilized bacterialstrains for each level of contamination (M, L, and B). After the experimental contamination of food samples, a ten-fold dilution of each sample in BPW was performed for Salmonella or in Half Fraser for L. monocytogenes. Samples were homogenised for 90 s, and incubated at37°C±1°Cfor18hours±2hours.


food control

Samples were subsequently analysed following the two methodological alternatives: traditional culture method (ISO6579:2002/Corr. 1:2004 or ISO 11290-1:1996+Amd.1:2004 for L. monocytogenes) for Salmonella or ISO and the alternative methods (the specific Real-Time PCR-based methods).

All the laboratories did not show appreciable problems during the ring trial for Salmonella, but one of the laboratories was excluded for L. monocytogenes as it reported serious problems during the preparation of the trial materials leading to negative results from both ISO and alternative methods. In the case of Salmonella, all the results obtained from artificially contaminated samples showed a correct determination by both methodologies. However, results were differnt for L. monocytogenes: at a low level load of L. monocytogenes (10 CFU per 25 g of sample) and concomitant contamination of L. innocua (3 CFU per sample), 70.45% (62 out 88) of thesamples were considered as positive using the reference method ISO 11290-1:1996+Amd.1:2004, whereas 87.50%(77outof88)wereconsideredas positive using the real-time PCR-based method. As a result, the real-time PCR-based method detected significantly more samples positive than the reference method (22.22% more samples).

The classical performance parameters (the diagnostic specificity, diagnostic sensitivity, positive and negative predictive values, accordance and concordance values and the concordance odds ratio) showed excellent results for both foodborne pathogens (Delibato et al, 2014; Gianfranceschi et al., 2014). In addition, the relevant parameters defined in ISO 16140 for validation of alternative methods in food microbiology (relative accuracy, relative specificity, relative sensitivity, and false negative and positive ratios) also showed an excellent performance (Delibato et al, 2014; Gianfranceschi et al., 2014).

The results showed that Real-Time PCR methods not only has a robust detection limit of at least 10 CFU per 25 g of food sample, but also has a superior capacity for correctly determine the nature of sample tested (values of 100% for diagnostic specificity and sensitivity and for positive and negative predictive values) regardless the analysis is performed in different food samples or laboratories (values of 100% for accordance, concordance and concordance odd ratio). On top of that, the performance of the Real-Time PCR method is equal to that using the reference ISO method. This finding is demonstrated by the excellent results for the major parameters indicated in the ISO standard 16140:2003 for validation of alternative methods in food microbiology (values of 100% for relative accuracy, sensitivity and specificity, and 0% of false negative and positive ratios). Interestingly, in the L. monocytogenes study, the food samples were artificially contaminated with L. monocytogenes and co-contaminated with L. innocua to mimic as much as possible the real scenario found in food samples (Ryser, 1999; Gravani, R. 1999). Previous studies have highlighted the possibility of an overgrowth of L. monocytogenes by L. innocua, during selective enrichment, leading to high rates of false-negative results (Zitz et al., 2011; Besse et al., 2010; Oravcová et al., 2008). Animportant observation from the results of this inter-laboratory study was that the reference method was not completely reliable for the detection of L. monocytogenes in the presence of L. innocua. From a practical perspective, it implies the possibility of false negative results with serious consequences on public health since the non-compliant food lots will not be withdrawn from the market. In addition, false negative results can complicate the ability of public health investigators to trace-back the source of contamination, allowing the spread of contamination. As a result, there is a need for improving the reference method for L. monocytogenes detection or

alternatively shift to more sensitive methods, like the molecular method used in the current study. Interestingly, the L. monocytogenes detection by the real-time PCR method was affected to a lower extend in the same situation, which in the current study resulted in a more reliable detection of positive samples, than the reference method.

As the two steps for fully validation indicated in that ISO standard (in house validation –Rodríguez-Lázaro et al., 2015a;b- and the interlaboratory studies) have been successfully conducted, we can conclude that the Real-Time PCR method meets the requirements of a diagnostic PCR and has the potential to become a standardized method for the rapid detection of both pathogens in diagnostic laboratories. On top of that, the Real-Time PCR method is cost effective (2 EUR vs. 12 EUR for Salmonella and 3 EUR vs 15 EUR for L. monocytogenes) and time saving (23 hours vs. > 7 days for positive results for Salmonella and 27 hours vs 7 days for L. monocytogenes) providing satisfactory reproducibility when carried out by different laboratories with different Real-Time PCR platforms. However, due to the high cost of an inter-laboratory study, the trials were only performed for two food category (pork meat and cheese) among those listed in the ISO 16140. Other validation studies must be conducted for other food categories (e.g. poultry meat or molluscs) to corroborate the satisfactory results obtained in this interlaboratory investigation using those foods.

References

• Abdulmawjood,et al (2004). Toward an international standard for PCR-based detection of foodborne Escherichia coli O157: validation of the PCR-based method in a multicenter collaborative Trial. Journal of AOAC International 87:856-860.

• Besse et al. (2010) The overgrowth of Listeria monocytogenes by other Listeria spp. in food samples


food control

undergoing enrichment cultivation has a nutritional basis. International Journal of Food Microbiology. 136(3). 345-51.

• Bustin et al. (2009) The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin. Chem. 55:611-22.

• D’Agostino, M. and Rodríguez-Lázaro, D. (2009). Harmonisation and Validation of Methods in Food Safety - “Food-PCR” a case study. In: Global Issues in Food Science and Technology. Edited by: Gustavo Barbosa-Cánovas, Alan Mortimer, David Lineback, Walter Spiess, Ken Buckle and Paul Colonna. Elsevier. Pp 199-209.

• D’Agostino et al.(2004). A validated PCR-based method to detect Listeria monocytogenes using raw milk as a food model – towards an international standard. Journal of Food Protection 67: 1646 - 1655.

• Delibato et al. (2014) European validation of a Real-Time PCR method for detection of Salmonella spp. in pork meat. Int. J. Food Microbiol. 184:128-133.

• Gianfranceschiet al. (2014) European validation of a Real-Time PCR-based method for detection of Listeria monocytogenes in soft cheese. Int. J. Food Microbiol. 184:134-138.

• Gravani, R. 1999. Incidence andcontrol of Listeria in food-processing facilities. 657–709. In E. T. Ryser and E. H. Marth (ed.). Listeria, listeriosis, and food safety, 2nd ed. Marcel Dekker, Inc., New York, N.Y.

• Heid et al. (1996). Real time quantitative PCR. Genome Research, 6,986-994.

• Hoorfar, J., and Cook, N. (2003).Critical aspects in standardization of PCR. In Methods in Molecular Biology: PCR detection of microbial pathogens, K. Sachse and J. Frey, eds. (Totowa, USA: Humana Press), pp. 51–64.

• Hwang et al. (2005) Patient-specific embryonic stem cells derived from human SCNT blastocysts. Science. 308:1777-83.

• Josefsen et al. (2004). Validation of a PCR-based method for detection of foodborne thermotolerant Campylobacters in a multi-center collaborative trial. Applied and Environmental Microbiology, 70: 4379-4383.

• Koch (1881) Zur Untersuchungenvon pathogenen Organismen. Mittheilungen aus dem Kaiserlichen Gesungheitsanye,1,1-48

• Krause et al. (2006). Comparative, collaborative, and on-site validation of a TaqMan PCR method as a tool for certified production of fresh, Campylobacter-free chickens. Applied and Environmental Microbiology, 72: 5463-5468.

• Lefeveretal. (2009)RTPrimerDB: theportal for real-time PCR primers and probes. Nucleic Acids Res. 2009 37: D942-5.

• Löfström et al. (2009). Validation of a same-day real-time PCR method for screening of meat and carcass swabs for Salmonella. BMC Microbiol.9:85.

• Malorny et al. (2004). Multicenter collaborative trial validation of a PCR-based method for detection of Salmonella in chicken and pig samples. Journal of AOAC International, 87:861-866.

• Malornyet al. (2003). Standardization of diagnostic PCR for the detection of foodborne pathogens. International Journal of Food Microbiology,83:39–48.

• Malorny et al. (2007). Multicenter validation study of two blockcycler- and one capillary-based real-time PCR methods for the detection of Salmonella in milk powder. International Journal of Food Microbiology,117:211–218.

• Mulliset al.(1986).Specificenzymaticamplification of DNA in vitro: the polymerase chain reaction. Cold Spring Harbor Symp. Quant. Biol. 51, 263-73.

• Oravcováet al.(2008)Limitationinthedetection of Listeria monocytogenes in food in the presence of competing Listeria innocua. Journal of Applied Microbiology 104. 429–437.

• Rodríguez-Lázaroet al. (2007) Trends

in Analytical Methodology in Food Safety and Quality: Monitoring Microorganisms and Genetically Modified Organisms. Trends in Food Science and Technologies. 18. 306-319.

• Rodríguez-Lázaro, D. and Hernandez,M. (2013) Real-time PCR in Food Science: Introduction. Curr. Issues Mol. Biol.15:25-38.

• Ryser,E.T.1999.Foodbornelisteriosis.299–358. In E. T. Ryser and E. H.Marth (ed.), Listeria, listeriosis, and food safety, 2nd ed. Marcel Dekker, Inc., New York, N.Y.

• Saiki et al. (1985). Enzymaticamplification of beta-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230, 1350-1354.

• Saiki et al. (1988). Primer-directedenzymatic amplification of DNA with a thermostable DNA polymerase. Science239,487-491.

• Strategic consulting Inc (2013) Food Micro, 8th Edition. MicrobiologyTesting in the Global Food Industry.

• Strategic consulting Inc (2014a)Industrial Microbiology Market Review, 4th Edition: Global Review of Microbiology Testing in the Industrial Market.

• StrategicconsultingInc(2014b) Food Contract Lab Report: Global Food Quality and Safety Testing by Contract Laboratories.

• Uhlmannet al. (2002) Potential viral pathogenic mechanism for new variant inflammatory bowel disease. Mol Pathol. 55:84-90.

• Zitzet al. (2011). Reduced detectability of Listeria monocytogenes in the presence of Listeria innocua. Journal of Food Protection.74(8).1282-7.

abril 2017/ 43

J. M. Peinado, M.I de Silóniz, P. Wrent, E.M. Rivas, E. Gil de Prado, O. Esteban y J. F. Vera. Dpto. Microbiología III, Facultad de Biología, Universidad Complutense de Madrid.

En este artículo, los autores analizan el efecto de determinadas levaduras en la conservación y calidad de los alimentos, considerando las posibles alteraciones producidas en función de las tecnologías empleadas para su producción. Entre otros ejemplos, los investigadores han estudiado el comportamiento de las levaduras en la inoculación de cultivos iniciadores en elaborados cárnicos

Levaduras en los alimentos: ¿buenas amigas, peores enemigas?

Yeast in food: best friends or the worst enemies?In this article, the authors analyse the effect of certain yeasts on food preservation and quality considering the possible alterations that can occur based on the technologies used for their production. Among other examples, the research has studied the behaviour of yeast used to inoculate starter cultures in processed meat products

Está generalmente aceptado que uno de los pilares fundacionales de la Microbiología como cien-cia moderna fueron los estudios de Pasteur sobre la producción

y deterioro de la cerveza. La producción de alimentos había sido hasta entonces un proceso puramente empírico, basado en la consolidación de los aciertos y rechazo de los errores, que había dado lugar a unas reglas de fabricación ancestrales cuyo estricto cumplimento aseguraba, aunque no siempre, el éxito final. Fue mérito, en este caso de los

industriales cerveceros, no de los científicos, el que se les ocurriera que la ciencia podía ayudar a disminuir los fracasos. Pasteur, al mismo tiempo que mostró que el análisis de un problema industrial puede llevar a resol-ver polémicas científicas tan trascendentales como la generación espontánea, demostran-do que Francesco Redi tenía razón y “Omne vivum ex vivo” (“Toda vida sale de vida”), sentó las bases metodológicas para al análisis del deterioro de los alimentos asociándolo al proceso de fabricación. Como consecuencia empezó a quedar claro desde entonces, y

seguimos confirmándolo día a día, que los microorganismos solo hacen lo que han aprendido a hacer a lo largo de su evolución, desarrollando sus actividades metabólicas según las condiciones del ambiente en que se encuentran. La fermentación alcohólica de Saccharomyces, que es esencial para la fabri-cación de vino y cerveza, es negativa cuando ocurre en los yogures. La fermentación de las bacterias lácticas, esencial para la producción de yogur, es negativa si ocurre en vinos y cer-vezas. Una consecuencia importante de este análisis es que la ciencia microbiológica que

dossier derivados cárnicos - dossier derivados cárnicos - dossier derivados cárnicos - dossier derivados cárnicos -

44 /abril 2017

Tecnifood. La revista de la tecnología alimentaria dossier derivados cárnicos - dossier derivados cárnicos - dossier derivados cárnicos - dossier derivados cárnicos - - dossier derivados cárnicos - dossier derivados cárnicos -

En contraste con el área de la seguridad de los alimentos, donde hay enemigos declarados que deben estar siempre ausentes, en el área de la calidad lo que nos encontramos son levaduras que deben estar presentes, porque pueden ser beneficiosas, pero que mal escogidas y peor controladas, pueden resultar perjudiciales

Figura 1

Vesículas de gas en el interior del embu-tido provocadas por microorganismos

fermentativos

hay que desarrollar y aplicar a los procesos industriales microbianos, sea producción o deterioro, es exactamente la misma: la ecofi-siología microbiana. Con Pasteur, un nuevo campo de investigación con una orientación claramente aplicada, se había abierto: El des-cubrimiento y caracterización de los microor-ganismos responsables por las transformacio-nes, positivas o negativas, que ocurrían en procesos desarrollados artesanalmente.

El primer empresario en montar su pro-pio laboratorio de I+D fue Jacob Christian Jacobsen (1811-1887), fundador de la Carlsberg Brewery en 1847, y del Carlsberg Laboratory en 1875. Fue pionero y modelo paradigmático de empresario que cree en la ciencia. Siguiendo la estela de Pasteur, el laboratorio inició con Hansen, creador la primera colección de microorganismos, la práctica industrial de los cultivos iniciadores y al mismo tiempo en él se desarrollaron con-ceptos científicos universales como el de pH, descrito por el Dr. Sorensen en 1909. Sin embargo, las dificultades de aplicar el método científico a un proceso artesanal, debidas a las características esenciales de cada abordaje, pronto se hicieron notar. El método empírico había producido procesos basados en una red muy compleja de interrelaciones entre los diferentes factores intervinientes, que el abor-daje reduccionista propio de la ciencia era incapaz de abarcar en algunos casos. Como consecuencia, en los casos de redes más com-plejas, la aplicación directa de un resultado científico podía romper el equilibrio alcan-zado empíricamente y el proceso se alteraba negativamente. Para superar esas dificultades

es necesario cerrar el círculo artesanía-empi-rismo-ciencia con otra rama que incluye el análisis de los problemas planteados por la aplicación del resultado científico para que la tecnología los resuelva, de manera que se vuelva a producir un alimento con la calidad del artesano pero con las ventajas introduci-das por la tecnología.

La relevancia del análisis científico de las alteraciones producidas por los cambios tecnológicos en la fabricación de alimentos fue reconocida de manera explícita, en la década de los 90, con varios proyectos euro-peos financiados con ese objetivo. El enva-sado en atmósferas modificadas, las nuevas formulaciones de los alimentos, la ausencia de conservantes en los productos orgánicos, el control biológico de plagas en la agricul-tura y los nuevos cultivos iniciadores, son ejemplos de las nuevas oportunidades que las novedades tecnológicas ofrecen para su estudio por los microbiólogos de alimentos. En nuestro laboratorio hemos investigado algunos de ellos, especialmente los protago-nizados por levaduras.

Deterioro por producción de CO2 fermentativo

Las levaduras, y especialmente las especies pertenecientes a la familia Saccharomycetaceae, son las mayores productoras de gas entre los microorganismos. Ello se debe a varias carac-terísticas: todas tienen una gran capacidad de transporte de azúcares y producción de piru-vato y poseen alcohol deshidrogenasa que cataliza la formación de etanol y CO2, pero se diferencian en su capacidad respiratoria. S. cerevisiae reprime la síntesis de citocromos en presencia de concentraciones altas de glucosa, por lo que en esas condiciones toda la glucosa es fermentada. Zygosaccharomyces y Torulaspora no reprimen los citocromos pero tienen una capacidad respiratoria pequeña y la fermentación ocurre cuando el flujo glicolítico es mayor que esta(2). Finalmente hay un grupo muy diverso de especies, que incluye aquellas de los géneros Debaryomyces, Kluyveromyces, Pichia, Candida y Meyerozyma entre otros, con una capacidad respiratoria mayor, que es capaz de absorber todo el flujo glicolítico y, por tanto, solo fermentan en

ausencia de oxígeno. Curiosamente, en nues-tro laboratorio apenas una vez hemos encon-trado a S. cerevisiae como agente etiológico de alimentos alterados por hinchamiento y en el resto de los casos se trataba de especies pertenecientes a los otros dos grupos.

La inyección de salmuera en diversos tipos de productos cárnicos es una práctica industrial que parece aumentar su frecuen-cia. Por otra parte, también se va extendien-do la inoculación de cultivos iniciadores en embutidos, además de las bacterias lácticas (BAL), para mejorar las características orga-nolépticas. La levadura Debaryomyces hanse-nii, por sus características enzimáticas, es de las especies más recomendadas para este fin. Sin embargo la inyección de salmuera, si no está estéril, puede ser la vía de introducción de microorganismos fermentativos que creen vesículas de gas en el interior del embutido (Fig.1), así como también puede producir gas la incorporación de levaduras a la masa del embutido (imagen de apertura en página 43). Si la cepa puede crecer aeróbicamente utilizando el lactato producido por las BAL, se crea una simbiosis entre levaduras y bac-terias porque el consumo del ácido por la levadura hace subir el pH, lo que permite a las BAL seguir creciendo y produciendo más ácido. Esta subida del pH altera la correcta compactación y maduración de los embu-tidos. Por otra parte, la gran población de levaduras acumulada comienza a fermentar cuando se agota el oxígeno, con la acumu-lación de gas. Nuestros estudios fisiológicos han demostrado que la fermentación la reali-zan las levaduras Crabtree-, en la fase estacio-naria en la que todo el azúcar consumido es fermentado, en contraste con las Crabtree+, que la realizan durante el crecimiento para obtener energía.

Para la identificación rápida de Debaryomyces hansenii hemos diseñado un medio diferencial (Debaryomyces Differential Medium, DDM) basado en la detección de la actividad enzimática β-glucuronidasa. En este medio, entre 120 especies de bac-terias, levaduras y mohos ensayados, solo Debaryomyces hansenii produjo colonias vio-letas(7). El uso de este medio permitió iden-tificar como Candida cretensis a unas cepas con todas las características fisiológicas de D. hansenii, pero que no producían colo-nias violeta. De esta especie solo se conocía una cepa(8). El análisis de la región IGS en Debaryomyces hansenii fue de gran utilidad. Analizando su tamaño descubrimos que este dato permitía distinguir entre 21 especies diferentes, contaminantes habituales de ali-

abril 2017/ 45


mentos. Además el análisis RFLP permitió distinguir las distintas especies del género y también las dos variedades de la especie Debaryomyces hansenii, var. hansenii y var. fabrii. Estos datos apoyaban la transforma-ción de las variedades en especies indepen-dientes, como finalmente ha ocurrido(9,10). La experiencia acumulada con Debaryomyces hizo que se le encargara la redacción del capítulo correspondiente a este género en la “Encyclopedia of Food Microbiology” a la Dra Siloniz(12). De nuevo hay que resaltar que estos resultados, que demuestran que Debaryomyces hansenii puede producir CO2 fermentativo y por tanto es un peligro para la calidad de los alimentos fermentados, no indican que haya que renunciar a sus ventajas como productor de enzimas de interés. Sin embargo si indican que habrá que minimizar el riesgo de deterioro seleccionando cepas con baja o nula capacidad fermentativa en las condiciones de elaboración y conservación del alimento.

El envasado de alimentos en atmósferas modificadas con plásticos impermeables a los gases puede poner de manifiesto fermentacio-nes que sin envasado pasarían desapercibidas.

Ese fue el caso de unos higos envasados al vacío en el que identificamos como levadura responsable del hinchamiento a una cepa de Zygosaccharomyces bailii. Para hacer estudios de trazabilidad diseñamos un método de tipado basado en el análisis RFLP de la región IGS del rADN que no solo permitió locali-zar el origen de la contaminación sino que puso de manifiesto alguna de las dificultades ligadas a estas técnicas, como la incorrecta identificación de cepas en las colecciones de cultivos y la presencia de híbridos(11).

Más interesante y complejo resultó el caso de los yogures orgánicos contaminados con levaduras. En este problema se reunían inicialmente dos modificaciones al proceso tradicional: cambio en la formulación clási-ca por la adición de mermelada y la ausencia de conservantes por ser producto “orgáni-co”. Además, la mermelada era de fresas de cultivo orgánico, por lo que podían haber sido inoculadas en el campo con microorga-nismos para el control biológico. Los resul-tados iniciales del laboratorio de control eran paradójicos porque la contaminación por levaduras era muy abundante en todos los yogures, estuvieran hinchados o no, pero

no aparecía rastro de ellas en ninguno de los ingredientes. El problema consistía en la media o baja actividad de agua de las matri-ces analizadas. Las levaduras son contami-nantes habituales de este tipo de alimentos porque muchas de las especies son osmoto-lerantes y resisten la muerte por apoptosis provocada por las elevadas concentraciones de azúcar(6). El empleo de la metodología adecuada para el análisis microbiológico de alimentos con media y baja actividad de agua(1,4,5) reveló la presencia de levaduras también en la mermelada. La aplicación completa de los postulados de Koch, espe-

La inyección de salmuera, si no está estéril, puede ser la vía de introducción

de microorganismos fermentativos que creen

vesículas de gas en el interior del embutido

46 /abril 2017


cialmente la reproducción del efecto tras la reinoculación de la cepa aislada, permitió atribuir el hinchamiento a una levadura que fue identificada como Meyerozyma (antigua Pichia) guilliermondii. Los análisis de tipado, resultaron complejos porque inicialmente hubo que utilizar RFLPs de mitADN e IGS de rDNA, pero consiguie-ron demostrar que la cepa presente en la mermelada era la misma que estaba en los yogures(13). Actualmente hemos desarrolla-do otra metodología de tipado, basada en el análisis de microsatélites, más barata y eficiente(15). Pretendimos trazar el origen hasta los cultivos de fresa, pero no fuimos autorizados. Los estudios fisiológicos con la cepa responsable del deterioro nos per-mitieron identificar las características pro-pias de la cepa que facilitaban su potencial deteriorante. Era una cepa capaz de crecer en lactato, incluso a partir de inóculos muy bajos lo que explicaba su presencia en

todos los tipos de yogur, pero era incapaz de fermentar lactosa, lo que explicaba que solo en los yogures con azúcares fermenta-bles añadidos con la mermelada, hubiera producción de gas. Un aspecto novedoso lo constituyó la inhibición selectiva de la fermentación, pero no del crecimiento, a temperaturas bajas, lo que explicaba que el hinchamiento solo se detectase cuando se interrumpió la cadena de frío. Además formaba biofilms lo que facilitaba la con-taminación de las máquinas en la fábrica. Es interesante considerar que estas caracte-rísticas son precisamente las que se buscan en la selección de microorganismos para el control biológico en agricultura. Este no es un argumento en contra del control bio-lógico, simplemente indica que la inocu-lación de levaduras fermentativas implica un peligro cuyo riesgo hay que minimizar eliminando determinadas características en las cepas industriales.

Malos olores: producción de 1,3 Pentadieno por descarboxilación del sorbato

El sorbato es uno de los conservantes más ampliamente utilizados en la elaboración de alimentos. Su acción se basa en la acidifica-ción del pH intracelular, por lo que, para que sea eficaz, el pH externo tiene que ser inferior a su pK, ya que las formas protonadas son las únicas capaces de entrar en la célula. Esos protones son los que se disocian en el cito-plasma, por su pH neutro, y para mantener ese pH la célula tiene que expulsarlos a través de una bomba que gasta ATP (18,21). Es posi-ble que el anión también tenga alguna acción tóxica y por ello algunos autores piensan que su descarboxilación por algunas especies de mohos y levaduras, transformándolo en 1,3 pentadieno que se volatiliza, constituiría un mecanismo de destoxificación. Pero ese compuesto tan volátil tiene un desagradable

Publicaciones del grupo de levaduras de interés industrial (UCM) sobre ecofisiología y diagnóstico de levaduras deteriorantes de alimentos comentadas en este trabajo:

Levaduras deteriorantes de alimentos con baja o media actividad de agua

1.-Beuchat, L.R., Jung Y., Deak T., Kefler T., Golden D.A., Peinado J.M. Gonzalo P., de Siloniz M.I.

and Valderrama M.J. (1998) An interlaboratory study on the suitability of diluents and recovery media for

enumeration of Zygosaccharomyces rouxii in high sugar foods. J. Food Mycol. 1:117-130.

2.-Leyva J.S., Manrique M., Prats L., Loureiro-Días M.C. and Peinado J.M. (1999) Regulation of

fermentative CO2 production by the food spoilage yeast Zygosaccharomyces bailii. Enz. Microbial. Tecnol.

24:270-275.3.-de Silóniz M.I., Valderrama M.J. and Peinado

J. M (2000) A chromogenic medium for the detection of yeasts with beta-galactosidase and beta-glucosidase activi-ties from intermediate moisture foods. J. Food Protection,

63:651-654.4.-Beuchat L.R., Frändberg E., Deak T, Almazora

S.M., Chen J., Guerrero, S. López-Malo A., Ohlsson I., Olsen M., Peinado J. M., , Shnurer J., de Silóniz M.I., Tornai-Lehoczki T. (2001) Performance of mycological

media in enumerating desiccated food spoilage yeasts: an interlaboratory study. Int. J. Food Microbiol., 70:89-96.

5.-Marquina D., Llorente P., Santos A. and Peinado J.M. (2001) Characterization of the yeast

population in low water activity foods . Adv. Food Sci., 23:63-67.

6.-Silva R. D., Sotoca R., Johansson B., Ludovico P, Sansonetty F., Silva M. T., Peinado J. M. and Corte-Real M. (2005) Hyperosmotic stress induces metacaspa-

se- and mitochondria-dependent apoptosis. Molecular Microbiology 58:824-834.

Deterioro de Alimentos por producción de CO2 fermentativo: Técnicas de identificación y tipado de las

levaduras responsables.7.-Quirós, M., Wrent, P., Valderrama, M.

J., de Silóniz, M. I. and Peinado, J. M. (2004) β-glucuronidase based agar medium for the differential

detection of the yeast Debaryomyces hansenii from foods. J. Food Protection 68:802-804.

8.- Quirós M, Martorell P, Querol A, Barrio E, Peinado JM y de Silóniz MI. (2008). Four new

Candida cretensis strains isolated from Spanish fermented sausages (chorizo): Taxonomic and phylogenetic implica-

tions. FEMS Yeast Res 8: 485-491.9.-Quirós M, Martorell P, Valderrama MJ, Querol

A, Peinado JM y de Silóniz MI. (2006). PCR-RFLP analysis of the IGS region of rDNA: a useful tool for

the practical discrimination between species of the genus Debaryomyces. Anton Leeuw Int J G 90: 211-219.10.- Romero P, Patiño B, Quirós M, González-Jaén

T, Valderrama MJ, de Silóniz MI, Peinado JM. (2005). Differential detection of Debaryomyces hansenii isolated from intermediate-moisture foods by PCR-RFLP of the

IGS region of rDNA. FEMS Yeast Res 5: 455-461.11.-Wrent P, Rivas EM, Peinado JM y de Silóniz

MI. (2010). Strain typing of Zygosaccharomyces yeast species using a single molecular method based on the intergenic spacer region (IGS). Int J Food Microbiol

142: 89-96.12.-Wrent P, Rivas EM, Gil de Prado E, Peinado

JM y de Silóniz MI. (2014). Debaryomyces in Encyclopedia of Food Microbiology, vol 1. Ed. Batt CA

y Tortorello ML (eds). Elsevier Ltd, Academic Press, 563-570.

13.- Wrent P, Rivas E., Gil Prado E., Peinado J. M. and, de Silóniz. M.I.(2015) Assesment of the factor

contributing to the growth or spoilage of Meyerozyma guillermondii in organic yogurt: Comparison of methods for strain differentiation”. Microorganisms, 3:428-440.

14.-Wrent P., Rivas E., Gil Prado E., Peinado J. M. and . de Silóniz M.I. (2015) Development of species-specific primers for rapid identification of Debaryomyces

hansenii. Int. J. Food Microbiol., 193:109-113. P. Wrent, E. Rivas, J. M. Peinado and M.I. de Silóniz.

15.-P. Wrent, E. Rivas, J. M. Peinado and M.I. de Silóniz.(2016) “Development of an affordable typing

method for Meyerozyma guillermondii using microsate-llite markers Int. J. Food Microbiol. 217:1-6. 2016.

Ácidos orgánicos como conservantes de alimen-tos: Bases fisiológicas de su toxicidad, y producción

de 1-3 pentadieno (olor a petróleo) a partir de sorbato por levaduras

16.-Casas E., Valderrama M.J.,y Peinado J.M. (2004) Sorbate detoxification by spoilage yeasts isolated

from marzipan products. Food. Technol. Biotechnol. 37:87-91.

17.-Casas E., De Ancos B., Valderrama M.J., Cano P. y Peinado J.M. (2004) Pentadiene production from

potasium sorbate by osmotolerant yeasts. Int. J. Food Microbiol. 94:93-96.

18.-Leyva JS y Peinado JM. (2005). ATP requi-rements for benzoic acid tolerance in Zygosaccharomyces

bailii. J Appl Microbiol 98: 121-126.19.-Rivas E., Maldonado M., Diezma B, Wrent

P., Peinado J. M., de Silóniz M.I., Vergara G., García-Hierro J., Robla J I.and Barreiro P. (2015) Detection

of Biological CO2 and 1,3-Pentadiene Using Non-refrigerated Low-Cost MWIR Detectors. Food Anal.

Methods: DOI 10.1007/s12161-015-0320-6.

Microbiología Predictiva: Modelos matemáticos del crecimiento de las levaduras en medios líquidos y sólidos y

en presencia de conservantes.20.-Barandica, J.M., Santos, A., Marquina, D.,

Acosta, F.J., Peinado, J.M. (1999) AMathematical model for toxin accumulation by killer yeasts based on the yeast

population growth. J. Appl. Microbiol. 86:805-811.21.-Quintas C., Leyva J.S., Sotoca R., Loureiro-

Días M.C. and Peinado J. M. (2005) A model of the specific growth rate inhibition by weak acids in yeasts based on energy requirements. Int. J. Food Microbiol.

100:125-130.22.-Rivas E.M., Gil Prado E., Wrent P., Silóniz

M.I, Barreiro P,. Correa E. C., Conejero F., Murciano A., and Peinado J. M. (2014) A simple mathematical

model that describes the growth of the area and the num-ber of total and viable cells in yeast colonies. Letters Appl.

Microbiol . 08/2014; DOI: 10.1111/lam.1231423.- Gil de Prado E., Rivas E.M., Silóniz M.I.

Diezma B, Barreiro P. and Peinado J.M. (2014) Quantitative analysis of morphological changes in yeast

colonies growing on solid medium: The eccentricity and Fourier indexes. Yeast 08/2014; DOI: 10.1002/

yea.3036.

abril 2017/ 47


olor a keroseno y los alimentos con sorbato en los que esta reacción ocurre tienen que ser retirados. En nuestro laboratorio hemos analizado varios productos con este tipo de alteración, fundamentalmente en alimen-tos ricos en azúcar, en los que las cepas productoras pertenecían a las especies Debaryomyces hansenii y Zygosaccharomyces rouxii. Hemos encontrado que la capaci-dad de producción es una característica de cepa, no de la especie(16,17).

El objetivo investigador en este caso era encontrar un marcador genético específico, no de especie, sino de efecto deteriorante. Los genes implicados en la descrboxilación son al menos dos, PAD1 que codifica una enzima denominada decarboxilasa del ácido fenilacrilico, y FDC1 que codifica la decar-boxilasa del ácido ferulico. Actualmente se cree que ambas enzimas son necesarias para la descarboxilación ya que la primera reclu-taría y aportaría un cofactor (FMN) que es necesario para la acción descarboxiladora de la segunda. Nuestros resultados indican que PAD1 no sirve como marcador, porque se encuentra en todas las cepas de ambas especies, productoras o no, por lo que nos

estamos concentrando en el estudio del gen FDC1. Sin embargo existe un polimorfismo en el gen PAD1 de Debaryomyces hansenii, en contraste con la homogeneidad que presenta en Zygosaccharomyces rouxii, que podría ser una herramienta más en la carac-terización de esta especie.

Nuestro trabajo se ha extendido también al diseño de pruebas rápidas para la detección temprana de los productos que finalmente causan el deterioro. El desarrollo de una metodología para la detección cuantitativa de CO2 y 1,3 pentadieno ha sido un trabajo que hemos desarrollado en colaboración con el grupo Tagralia de la ETSIA de la Universidad Politécnica de Madrid y la empresa New Infrared Technologies. Se trata de un aparato que emite rayos infrarrojos en un espectro de frecuencia media, que no necesita refrige-ración y que permite medir la presencia de ambos gases de forma económica y fiable(19).

Las levaduras: ¿amigas o enemigas?

La respuesta a esta pregunta después de haber llegado hasta aquí, parece fácil: depende de cómo las tratemos. En contraste

con el área de la seguridad de los alimentos, donde hay enemigos declarados que deben estar siempre ausentes, en el área de la calidad lo que nos encontramos son levaduras que deben estar presentes, porque pueden ser beneficiosas, pero que mal escogidas y peor controladas pueden resultar perjudiciales. Como microorganismos que son, solo van a tener efectos macroscópicos, tanto buenos como malos, si la población alcanza un tama-ño crítico, que hay que determinar porque puede ser muy variable e incluso puede ser diferente para ambos efectos.

La modelización y el subsecuente control del crecimiento microbiano que permite la microbiología predictiva son por tanto de especial interés en este caso. Sin embargo, la estructura del alimento es un factor que ha sido poco considerado hasta ahora en esta área. La modelización del crecimiento sobre alimentos sólidos es una trabajo que también estamos desarrollando en nuestro grupo y que ya está dando sus frutos (20-23). De esta forma iremos consiguiendo el objetivo de cualquier microbiólogo industrial: que los microorganismos hagan lo que queremos y no hagan lo que no queremos.❑

70 /octubre 2017

Mª del Carmen Portillo Guisado (BIOTENOL - Biotecnologia Enológica) - Departamento de Bioquímica y Biotecnología - Universitat Rovira i Virgili

Las levaduras y bacterias lácticas y acéticas que pueden acompañar a la producción de los vinos, se convierten en enemigos de la calidad del producto final si no se tienen en consideración las mejores condiciones en todos los pasos del proceso, desde la propia calidad sanitaria de la uva, hasta las adecuadas circunstancias de almacenamiento o crianza del vino, pasando por el control de contenido de azúcares durante la fermentación. En este artículo, basado en un trabajo presentado en el XV Workshop sobre métodos rápidos y automatización en microbiología alimentaria (MRAMA), se analizan las técnicas para detectar los microorganismos alterantes en vinos y cavas

La transformación de mosto de uva en vino es un proceso eco-lógico y bioquímico complejo que implica el desarrollo secuen-cial de especies microbianas tales

como levaduras, bacterias lácticas y bacte-rias acéticas (Pretorius, 2000). Cualquiera de estos microorganismos es susceptible de producir bien efectos positivos en el vino

o cava o bien alteraciones no deseables dependiendo de la etapa de la que se trate y en qué cantidad se encuentren.

El número de levaduras en la uva, justo antes de la cosecha, varía de 103 a 106 células/ml (Romano et al., 2006) en fun-ción de las variedades de uvas, las condi-ciones climáticas, prácticas vitícolas, etapa de maduración, el daño físico (causada por

hongos, insectos y aves) y fungicidas apli-cados a los viñedos (Pretorius et al., 1999). Aunque el mosto de uva es relativamente completo en cuanto a nutrientes, el bajo pH y el alto contenido en azúcar determi-nan que solo unas pocas bacterias y espe-cies de levaduras puedan crecer. Además, la adición de dióxido de azufre, principal compuesto antioxidante y antimicrobiano,

Tecnifood. La revista de la tecnología alimentaria - dossier vinos - dossier vinos - dossier vinos - dossier vinos

XV Workshop

M R A M A

Alteraciones microbiológicas en vinos y cavas: quiénes las producen y cómo detectarlas

octubre 2017/ 71

impone una selección adicional, princi-palmente contra bacterias y levaduras con menor poder fermentativo. Otro factor importante es la restricción creada por las condiciones anaeróbicas una vez iniciada la fermentación. Al inicio de esta, las especies de levaduras predominantes perte-necen a las denominadas no-Saccharomyces de los géneros Hanseniaspora, Candida, Pichia, Metschnikowia, Kluyveromyces, Zygosaccharomyces, Torulaspora, Starmerella, Dekkera y Schizosaccharomyces. Por el con-trario, la población de la principal levadura vínica Saccharomyces cerevisiae en el mosto es muy baja. Las no-Saccharomyces prolife-ran en los primeros días de fermentación hasta poblaciones aproximadamente de 106-107 células/ml. Sin embargo, durante la fermentación alcohólica el aumento de la concentración de etanol, el agotamiento de nutrientes claves como el nitrógeno y el aumento constante de la temperatura de fermentación favorecen la rápida prolifera-ción de las cepas de S. cerevisae, llegando a alcanzar poblaciones entre 5x107-108 ufc/ml en detrimento de estas otras especies mayoritarias en el mosto de uva (Goddard, 2008; Salvadó et al, 2011). En ocasio-nes, algunas levaduras como Brettanomyces, Kluyveromyces, Schizosaccharomyces, Torulas pora y Zygosaccharomyces aparecen en fases finales de la fermentación alcohó-lica. Sin embargo, esta presencia suele ir asociada a fermentaciones problemáticas y que acaban en alteraciones de la calidad del vino final.

Así pues, durante la fermentación alco-hólica (FA), las levaduras son las verdaderas protagonistas (levaduras no-Saccharomyces al principio y Saccharomyces posterior-mente), mientras que la contribución de bacterias lácticas y acéticas a la fermenta-ción alcohólica es considerada nula y su desarrollo suele estar ligado a alteraciones. Sin embargo, a pesar de que no haya proliferación, sí que hay supervivencia de bacterias (recuentos de 102 a 104 ufc/ml en el mosto) y, después de la fermenta-ción alcohólica, hay un crecimiento de las poblaciones de bacterias lácticas supervi-vientes (hasta 107 ufc/ml) que llevarán a cabo la fermentación maloláctica (FML), que consiste en la transformación del ácido málico presente de forma natural en el vino en ácido láctico y CO2. Tan pronto como el ácido málico se transforma completa-mente a ácido láctico, esta población de bacterias comienza a declinar. No obstante, si el vino no está suficientemente sulfitado

después de la FML, estas bacterias perma-necen durante meses. Durante la FML, las alteraciones que pueden producirse en el vino dependen fundamentalmente de la existencia de sustratos capaces de ser meta-bolizados en ese momento por las bacterias lácticas, por levaduras residuales que hayan sobrevivido a la FA o por las bacterias acé-ticas. Un vino en el que se han consumido todos los azúcares durante la FA presenta una probabilidad de sufrir alteraciones muchísimo más bajas que aquellos en los que se ha producido una parada de fermen-tación o en los que ha quedado una cierta cantidad de azúcares residuales que no han podido ser consumidos por las levaduras. Las bacterias acéticas también pueden estar presentes durante la fermentación tanto alcohólica como maloláctica. Su presencia depende mayoritariamente de la calidad sanitaria de la uva de origen. Uvas podri-das, por exceso de lluvias, o contaminadas con Botrytis cinerea producen mostos con poblaciones muy elevadas de acéticas. En el mosto, las especies mayoritarias son Gluconobacter oxydans, Acetobacter aceti y A. pasteurianus, y, en menor medida, Gluconacetobacter liquefaciens y Ga. hanse-nii (González et al., 2005), que son las ais-ladas mayoritariamente también en la uva. Durante la fermentación alcohólica, como consecuencia de las fuertes condiciones de anaerobiosis impuestas por el metabolismo de las levaduras, y la alta dependencia del oxígeno que presentan las acéticas para su desarrollo, las posibilidades de pro-liferación son prácticamente nulas. Sin embargo, estas bacterias son mucho más resistentes al sulfuroso, bajo pH, etanol, etc., que las bacterias lácticas. De manera que al final de la fermentación alcohólica presentan poblaciones todavía elevadas, que pueden proliferar si las condiciones de almacenamiento o crianza del vino no son las adecuadas, fundamentalmente sulfitado correcto y almacenamiento en atmósferas anaerobias. Si proliferan las bacterias acéti-cas, el vino final se altera con presencia de ácidos acéticos producidos por el metabo-lismo de las mismas.

Por último, Brettanomyces bruxellensis ha sido caracterizada históricamente como la principal levadura contaminante res-ponsable de la formación de fenoles volá-tiles como el 4-etilfenol, 4-etilguaiacol y tetrahidropiridinas, que producen aromas desagradables y que alteran el vino, sobre todo en las etapas de la crianza del vino y embotellado donde las poblaciones de esta

levadura aumentan de manera lenta pero sin competencia, originando los principa-les riesgos.

En el caso del cava, parte de los pro-blemas descritos anteriormente no serían aplicables porque el mosto original sufre filtraciones que bajan la carga microbio-lógica inicial disminuyendo el riesgo de contaminación. Sin embargo, también se tiende a añadir poca cantidad o ninguna de sulfito, lo que puede dar problemas de aparición de bacterias lácticas y refermen-taciones en botellas.

Cómo detectar los microorganismos alterantes

El sector vinícola necesita técnicas rápi-das y prácticas para la monitorización y detección de microorganismos que causan el deterioro del vino o el cava en cualquiera de sus etapas de fabricación y almacenado, para evitar las consecuentes pérdidas eco-nómicas. Tanto levaduras como bacterias, pueden ser responsables de la contamina-ción del vino o cava y su fuente primordial suele ser el propio ambiente de la bodega.

A pesar de que, en muchas ocasiones, se ha podido establecer una relación direc-ta entre especies alterantes y alteraciones específicas del vino, esta no ha sido siempre la situación real, dándose deterioro del vino con poblaciones muy bajas de especies deteriorantes conocidas hasta el momento (incluso, en casos en los que no se detec-tan). Asimismo, no se conocen con pre-cisión las razones por las que una deter-minada población alterante se desarrolla en un momento determinado en una tina de fermentación, barrica o botella de cava mientras que en otras del mismo origen y con las mismas condiciones no manifiesta

Las técnicas más usuales a nivel de bodega

son la observación al microscopio, cultivo en placas de medio selectivo o general, análisis sensoriales,

cromatografía de gases, análisis de PCR y

luminómetros

dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos

72 /octubre 2017

Tecnifood. La revista de la tecnología alimentaria dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos

dicha población. Entre las razones plausi-bles se podría considerar que las comuni-dades microbianas en cada tina, barrica o botella han evolucionado de forma dife-rente y la interacción con otros microorga-nismos podría favorecer el desarrollo de las poblaciones alterantes. Este hecho dificulta la predicción de la posible contaminación y su control.

Tradicionalmente, la detección de los microorganismos que alteran el vino se ha realizado mediante técnicas dependientes de cultivo o análisis sensoriales. Sin embar-go, algunos microorganismos no son capa-ces de crecer en los medios empleados o tienen un crecimiento lento, lo cual alarga la detección. Por otro lado, al igual que los métodos sensoriales, los métodos basados en el cultivo de microorganismos suelen ser confirmativos, dificultando una acción de previsión. Recientemente, la utilización de técnicas moleculares basadas en el material genético supone una ventaja respecto a las técnicas clásicas basadas en el cultivo, pues no es necesario este para la detección y son más sensibles a la hora de detectar cierto microorganismo. No obstante, el material y conocimiento necesario para llevar a cabo las técnicas moleculares no está siempre al alcance de una bodega pequeña o mediana.

Las técnicas más usuales a nivel de bodega, en la actualidad, son la observa-ción al microscopio, cultivo en placas de medio selectivo o general, análisis senso-riales, cromatografía de gases, análisis de PCR y luminómetros. Estos últimos son equipos de bioluminiscencia basados en la detección del ATP (adenosine triphosphate), que es la molécula energética por excelen-cia y se encuentra en todos los organismos vivos, con lo que su cuantificación sería proporcional a la presencia de organis-mos con actividad metabólica. El análisis mediante luminómetros de las muestras tomadas de las superficies de las bodegas mediante hisopos o bastoncillos estériles sería una buena indicación de presencia o ausencia de microflora, siendo un método rápido para monitorizar la eficiencia en la

limpieza de equipos y superficies, aunque los microorganismos específicos quedarían sin identificar.

Las técnicas moleculares utilizadas para la detección y estudio de microorganismos en el vino incluyen análisis como la restric-ción del ADN mitocondrial (Martorell et al., 2006), PCR-RFLP (Dias et al., 2003; Esteve-Zarzoso et al., 1999), PCR-RAPD (Martorell et al., 2006), PCR con cebado-res específicos (Campolongo et al., 2010) y PCR anidada (Ibeas et al., 1996). Sin embargo, muchas de estas técnicas nece-sitan un paso de enriquecimiento previo para extraer el ADN de forma que son técnicas semi-cuantitativas. Por ello, se está tendiendo al uso de qPCR (PCR cuantita-tiva) para la detección específica y cuanti-ficación directa de microorganismos con-taminantes del vino como B. bruxellensis. Aunque estas técnicas reducen el tiempo de análisis con respecto a las técnicas de cul-tivo y detectan incluso las células VPNC, al usar el ADN, se detectan tanto células vivas como muertas, pudiendo dar lugar a una sobreestimación del número de leva-duras en la muestra (Andorrà et al., 2010). Gracias a la introducción del bromuro monoazódico de etidio (EMA) o bromuro moazódico de propidio (PMB) combi-nados con la qPCR se han podido detec-tar únicamente las células vivas (Andorrà et al., 2010). Sin embargo, es necesario optimizar la concentración de EMA para ensayos diferentes y la concentración de etanol afecta los resultados (Andorrà et al., 2010). El ARN es considerado como un buen indicador de viabilidad ya que se degrada más rápidamente que el ADN (Ivey y Phister, 2011) aunque también es dependiente del gen usado para la detec-ción. El problema principal de la qPCR, a partir tanto de ADN como de ARN, es que la qPCR sólo detecta y cuantifica los microorganismos diana para los cuales hayamos diseñado cebadores de amplifica-ción específicos y, en la actualidad, se des-conoce si dichos microorganismos son los únicos implicados en el deterioro del vino.

Una nueva técnica molecular que puede emplearse para el estudio de la diversidad microbiana es la secuenciación masiva o HTS (de sus siglas en inglés high throughput sequencing). Consiste en la ordenación de miles de secuencias por cada muestra tras la extracción directa de ácidos nucleicos de la matriz que se esté estudian-do, en nuestro caso el vino o cava. Existen diferentes tecnologías de secuenciación masiva (Shendure and Ji, 2008) cada una con sus ventajas y desventajas (Suzuki et al., 2011; Liu et al., 2012). El estudio de la composición microbiana se realiza de forma habitual mediante la amplificación de genes de interés taxonómico (normal-mente el gen ARNr 16S para bacterias y el gen de ARNr 18S o el ITS para hongos) y puede ofrecer la proporción de los distintos grupos taxonómicos dentro de un alimento mediante la secuenciación de estos genes y su comparación con las bases de datos de referencia que proporcionarán la identi-ficación de las distintas secuencias con lo que, además, tiene carácter cuantitativo.

En los últimos años la secuenciación masiva se ha aplicado en prácticamen-te todos los campos de investigación de microbiología incluidos estudios sobre ali-mentos aunque el coste de estos análisis y el requerimiento de habilidades específicas bioinformáticas aún limitan su aplicación industrial. Muchos de estos estudios tenían un marcado carácter de ecología microbia-na (revisado en Ercolini, 2013) y, recien-temente, se ha puesto de manifiesto la capacidad y potencial de estas técnicas de HTS para poder detectar contaminaciones en alimentos y su posible trazabilidad en el entorno en el cual se procesan dichos alimentos (De Filippis et al., 2013). Hay que recordar que en muchos alimentos existen microorganismos pertenecientes al mismo género y en estos casos los estudios de HTS basados en secuencias muy cortas a nivel de género no serían de utilidad para diferenciar entre dichas especies. En estos casos, para obtener información a nivel de especie habría que tener como diana frag-mentos más largos incluyendo más regio-nes variables de los genes taxonómicos o bien complementar la técnica de HTS con alguna técnica de tipificación de especies como la RFLP (por ejemplo, Bokulich et al., 2012).

La ventaja indudable que las técnicas HTS aportarían al estudio de la conta-minación microbiana del vino o cava es que, al obtener miles de secuencias para

Un vino en el que se han consumido todos los azúcares durante la fermentación alcohólica presenta una probabilidad de sufrir alteraciones muchísimo más baja que aquellos en los que se ha producido una parada de fermentación o en los que ha quedado una cierta cantidad de azúcares residuales que no han podido ser consumidos por las levaduras

octubre 2017/ 73

Tecnifood. La revista de la tecnología alimentaria dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos - dossier vinos

una única muestra, se puede tener una descripción detallada de los microorganis-mos presentes en el deterioro y el cambio poblacional previo a la proliferación de un determinado microorganismo responsable de la alteración, con lo que podría tener carácter predictivo. Además, las técnicas HTS pueden procesar cientos de muestras simultáneamente y se pueden realizar tanto a partir de ADN como de ARN, con lo que se podría tener información tanto de los microorganismos presentes, como los metabólicamente activos en el momento del deterioro, respectivamente. Por otro lado, mediante la metatranscriptómica (secuenciación masiva de todos los genes que se están transcribiendo en un determi-nado momento) se podría tener informa-ción de las interacciones metabólicas entre los distintos microorganismos implicados en el deterioro del vino. Todas estas téc-nicas basadas en secuenciación masiva, sin duda ofrecen una oportunidad para un estudio más detallado de los microbios responsables de la fermentación del vino y en su caso, de su posible deterioro.

Aunque las técnicas HTS no sean aplicables a corto plazo por las bodegas como rutina, debido a los elevados costes de inversión en la tecnología necesaria y el grado de conocimiento específico que se necesita para realizar dicho análisis e interpretación de datos, sí que se pueden desarrollar servicios a bodegas para que realicen análisis periódicos de poblaciones en vinos de crianza y así poder prevenir posibles contaminaciones microbianas.❑

Bibliografía• Andorrà, I., Esteve-Zarzoso, B., Guillamón, J.M., Mas, A., 2010. Determination of viable wine yeast using

DNA binding dyes and quantitative PCR. Int J Food Microbiol, 144: 257-262.• Bokulich, N.A., Joseph, C.M.L., Allen, G., Benson, A.K., Mills, D.A. 2012. Next-generation sequencing reveals

significant bacterial diversity of botrytized wine. PLoS ONE, 7: e36357. doi:10.1371/journal.pone.0036357.• Campolongo, S., Rantsiou, K., Giordano, M., Gerbi, V., Cocolin, L. 2010. Prevalence and biodiversity of

Brettanomyces bruxellensis in wine from Northwestern Italy. Am J Enol Vitic, 61(4): 486-491.• De Filippis, F., La Storia, A., Villani, F., Ercolini, D. 2013. Exploring the sources of beefsteaks contamination

by culture-independent high-throughput sequencing. PLoS ONE, 8: e70222.• Dias, L. Pereira-da-Silva, S., Tavares, M., Malfeito-Ferreira, M., Loureiro, V. 2003. Factors afecting the produc-

tion of 4ethylfenol by the yeast Dekkera bruxelliensis in enological conditions. Food Microbiol, 4: 377-384.• Ercolini, D. 2013. High-throughput sequencing and metagenomics: moving forward in the culture-independent

analysis of food microbial ecology. Appl Environ Microbiol, 79(10): 3148-3155.• Esteve-Zarzoso, B., Belloch, C., Uruburu, F. Querol, A. 1999. Identification of yeasts by RFLP analysis of the

5.8S rRNA gene and the two ribosomal internal transcribed spacers, Int J Syst Bacteriol, 49: 329-337.• Goddard, M.R. 2008. Quantifying the complexities of Saccharomyces cerevisiae’s ecosystem engineering via fer-

mentation. Ecology, 89: 2077-2082.• González, A., Hierro, N., Poblet, M., Rozès, N., Mas, A., Guillamón, J.M. 2005. Application of molecular

methods to demonstrate species and strain evolution of acetic acid bacteria population during wine production. Int J Food Microbiol, 102: 295-304.

Ibeas, J.I., Lozano, I., Perdigones, F., Jimenez, J. 1996. Detection of Dekkera-Brettanomyces strains in sherry by a nested PCR method. Appl Environ Microbiol, 62(3): 998-1003.

• Ivey, M.L., Phister, T. 2011. Detection and identification of microorganisms in wine: a review of molecular techniques. J Ind Microbiol Biotech, 38: 1619-1634.

• Liu, L., Li, Y., Li, S., Hu, N., He, Y., Pong, R., Lin, D., Lu, L., Law, M. 2012. Comparison of next-generation sequencing systems. J Biomed Biotechnol, 2012: 251364.

• Martorell, P., Barata, A., Malfeito-Ferreira, M., Fernández-Espinar, M.T., Loureiro, V., Querol, A. 2006. Molecular typing of the yeast species Dekkera bruxellensis and Pichia guilliermondii recovered from wine related

sources. Int J Food Microbiol, 106: 79-84.• Millet, V., Lonvaud-Funel, A. 2000. The viable but non-culturable state of wine micro-organisms during stora-

ge. Lett Appl Microbiol, 30: 136-141.• Pretorius, I.S., 2000. Tailoring wine yeast for the new millennium: novel approaches to the ancient art of wine-

making. Yeast, 16: 675-729.• Pretorius, I.S., van der Westhuizen, T.J., Augustyn, O.P.H. 1999. Yeast biodiversity in vineyards and wineries

and its importance to the South African wine industry. S Afr J Enol Vitic, 20: 61-70.• Romano, P., Capece, A., Jespersen, L. 2006. Taxonomic and ecological diversity of food and beverage yeasts. In:

Yeast in Food Beverages. Querol A & Fleet GH (eds). pp. 55-82. Springer, Berlin, Germany.• Salvadó, Z., Arroyo-López, F.N., Barrio, E., Querol, A., Guillamón, J.M. 2011. Quantifying the individual

effects of ethanol and temperature on the fitness advantage of Saccharomyces cerevisiae. Food Microbiol, 28: 1155-1161.

• Shendure, J., Ji, H. 2008. Next-generation DNA sequencing. Nat Biotechnol, 26: 1135-1145.• Suzuki, S., Ono, N., Furusawa, C., Ying, B.-W., Yomo, T. 2011. Comparison of sequence reads obtained from

three next-generation sequencing platforms. PLoS ONE, 6: e19534. doi:10.1371/journal.pone.0019534.

La importancia de identificar los organismos alterantes

En Corbion, frecuentemente determinamos la floraalterante de diferentes productos de nuestros clientes,principalmente productos cárnicos, para identificar lasprincipales especies que deterioran un grupo de produc-

tos en particular. Luego determinamos la sensibilidadde los principales organismos alterantes frente a va-rios antimicrobianos como el ácido láctico, el ácidoacético y otros. Con esta información podemos acon-sejar el mejor producto y nivel de uso a nuestros clien-tes. Tenemos una gran base de datos con las curvas derespuesta a diferentes dosis de los organismos alteran-tes frente a diversos antimicrobianos. A menudo ac-tualizamos esta base de datos con nuevas cepas y nue-vos antimicrobianos. Los datos de respuesta a la dosisse derivan de los experimentos de crecimiento de colo-nias de microorganismos en medios de cultivo y utili-zando máquinas de bio-análisis. El bio-análisis puedemedir simultáneamente 200 muestras de crecimientomicrobiano en una sola sesión.

También sometemos a los organismos alterantes másimportantes a diferentes niveles de pH, temperatura y ac-tividad de agua, y junto con las curvas de respuesta a ladosis formamos modelos predictivos conforme al con-cepto gamma. El concepto gamma supone que ningu-no de los parámetros diferentes (pH, temperatura, etc.)interactúan entre sí y cada cual tiene un efecto separa-

Análisis y control

68 eurocarneNº 255. Abril 2017

El empleo de los métodos clásicos de aislamientoe identificación de la flora alterante requiere

unos costes y un esfuerzo considerables. La metagenómica es una técnica alternativa

que permite hacer análisis económicos,completos y sin el sesgo que se produce

por la selección de cultivos.

Deterioro de la carne y los productos cárnicos

en la era genómica: análisis mediante

metagenómica y métodos de cultivos

para identificación de flora sospechosa

Olav Sliekers

Grupo de microbiología alimentaria y modelos predictivos,Corbion

Leuconostoc mesenteroides.

do en la tasa de crecimiento[1]. Porlo tanto, se pueden multiplicar losefectos, llevando así a modelos me-canísticos de crecimiento relativa-mente simples y transparentes.

Los modelos predictivos necesi-tan muchos datos y esfuerzo, y porello debemos tener la certeza dehacer esto solo para los organis-mos alterantes más importantes.Por consiguiente, la identificaciónde los organismos alterantes juegaun papel importante en la micro-biología alimentaria en Corbion.

Metagenómica frente a identificación clásica

Un alimento se considera dete-riorado cuando la flora llega a unrecuento de 107 [2].

Tradicionalmente la analítica serealiza diluyendo una muestra delalimento deteriorado y aislando losorganismos de las placas con la di-lución más alta que muestra colo-nias. Así, cada colonia que se des-arrolla en una placa muestra unorganismo importante. Sin embar-go, el alimento deteriorado siem-

pre contiene varias especies domi-nantes, y para lograr una buena re-seña de la flora, deben obtenersemuchos cultivos puros de coloniase identificarlas posteriormente (fi-gura 1).

Hay un sesgo introducido por eltipo de medio de cultivo que se uti-liza; por eso es una buena prácti-ca colocar las muestras de alimen-tos en placas usando medios deagar diferentes. Algunos organis-mos tienen dificultades para cre-cer en medios de agar. Otro sesgose introduce cuando se eligen di-ferentes colonias para su posterioraislamiento. Las colonias muy pe-queñas pueden ser tan importantescomo las colonias grandes y se de-ben elegir cuidadosamente variascolonias de formas, colores y ta-maños diferentes para su posterioraislamiento e identificación. EnCorbion, la identificación de cul-tivos puros se externaliza, y es unaempresa especializada la que de-termina la identidad basándose enuna parte de la secuencia de 16 SrDNA y la homología con otras se-cuencias en la base de datos SILVA

Figura 1. Reseña esquemática de la identificación clásica frente a la metagenómica

Iden�ficación A&

Iden�ficación B

Iden�ficación A Iden�ficación B

TSA, BHI, MEA, MRS

Corte para iden�ficarespecies: 97 %

16 S rDNA ampliconsIllumina Myseq (>400 bp)

16 S rDNAamplicons

107 UFC

mailto:[email protected]

http://www.chemital.es

rRNA. Aislar e identificar varias colonias es laboriosoy costoso, y además la cantidad de bacterias aisladas deuna muestra de alimento se ve limitada por los costes,en tanto un método estadísticamente apto requeriríamás aislados que lo que obtenemos actualmente pormuestra. No obstante, debido a la gran cantidad demuestras de productos cárnicos cocidos envasados alvacío deteriorados que se analizan en Corbion, se pue-den entresacar las 5 bacterias alterantes más impor-tantes (tabla 1).

No es de extrañar que determinar la flora alteran-te con métodos independientes del cultivo se hayahecho cada vez más popular. El desarrollo de la me-tagenómica tiene ventajas especialmente claras so-bre otros métodos independientes del cultivo comoDGGE. En Corbion utilizamos los métodos de se-cuenciación 16 S rDNA amplicon de empresas co-mo Baseclear y Quality Partner. Hay otras empresasque ofrecen este servicio en Europa. En otras regio-nes del mundo, es más difícil hallar este servicio. Ennuestro caso, la región V1-V3 del 16S rDNA de unamuestra alimentaria se amplifica y se secuencian losamplicons con Illumina Myseq. Este método reali-za entre 3.000 y un millón de lecturas dependiendo dela profundidad de la secuenciación (¡y del precio!).Debido a la cantidad relativamente alta de secuen-cias, se puede imaginar que el uso de la metagenómi-ca brinda así una mucho mejor reseña de la floracomparándolo con el aislamiento y la identificaciónclásicos, donde los límites por su coste no permitenmás de aproximadamente 10 organismos aislados pormuestra. La metagenómica también facilita muchoel análisis estadístico y cuantitativo de varias mues-tras[7].

Análisis y control


Figura 2. Especies predominantes en productos cárnicos envasados al vacío por regiones

L. sakei 40%L. curvatus 18%

Lc. mesenteroides 10%

Lc. mesenteroides 29%L. sakei 20%

L. plantarum 13%

L. sakei 18%Lc. carnosum 18%

Lc. mesenteroides 11% L. plantarum 31%Lc. mesenteroides 20%

Lc. carnosum 16%

Tabla 1. Organismos alterantes predominantes en más de 50 muestras analizadas de productos cárnicos cocidos envasados al vacío deteriorados

Recuento absoluto Porcentaje

Lactobacillus sakei 48 23%Leuconostoc mesenteroides 35 17%Leuconostoc carnosum 24 11%Lactobacillus plantarum 23 11%Lactobacillus curvatus 22 10%Porcentaje total 72%

71eurocarneNº 255. Abril 2017

Estudio del deterioro de los productoscárnicos cocidos envasados al vacío

Aunque los productos cárnicos cocidos como las sal-chichas y el jamón cocido se someten a una etapa decalentamiento a unos 72 °C, no son las bacterias for-madoras de esporas las que frecuentemente deterioranestos productos cárnicos. Se ha establecido correcta-mente que las bacterias lácticas (LAB) son importantesorganismos alterantes de los productos cárnicos coci-dos y los defectos organolépticos y el descenso del pHson factores fuertemente correlacionados con la canti-dad de bacterias lácticas[2, 3]. Además de los defectos decolor, sabor u olor, el crecimiento de las bacterias lácti-cas puede causar la formación de gas y la producción delimo[4]. Un nivel de 107 bacterias lácticas por gramo enproductos cárnicos cocidos puede considerarse deterio-ro[2]. Se han hallado bacterias lácticas como Lactoba-cillus sakei y Leuconostoc mesenteroides, pero tambiénespecies de Weisella y Carnobacterium[5, 6]. Sin embar-

go, los estudios hasta ahora no dan un panorama clarosobre las especies predominantes, ni si existe variaciónen las especies predominantes entre regiones del mun-do.

Entre 2007-2014, se han aislado más de 200 orga-nismos de más de 50 productos cárnicos cocidos quehabían sobrepasado la fecha de caducidad. Las mues-tras provenían de 11 países diferentes, de todas las re-giones del mundo (América del Norte, América delSur, Europa, Medio Oriente, África, Asia). Se aisla-ron las cepas de las placas con la dilución más alta dela muestra. La identidad de las cepas se determinó me-diante un análisis de secuencia 500 bp 16 S RNA.

El organismo encontrado más frecuentemente fueLactobacillus sakei, representando más del 20 % detodos los aislados (tabla 1). También se hallaron fre-cuentemente Lactobacillus plantarum, Leuconostocmesenteroides y Leuconostoc carnosum (11-17 % decada uno). Lactobacillus curvatus se ubicó en el últi-mo lugar de los primeros 5, con un 10 %.

Análisis y control

ANÁLISIS EN CONTINUO PARA LA INDUSTRIA CÁRNICAIncremente el rendimiento de su planta y simplifiquesus operaciones diarias. QVision analiza el nivel de grasa, proteínas, colágeno y humedad.Con QVision podrá ofrecer un producto de calidad constante y disfrutar de un importante ahorro para su empresa.

www.tomra.com/es/qvision [email protected]

CARNE PICADA,CARNE EN DADOS,RECORTES, DESHUESADOSPEQUEÑOS Y MUCHO MÁS

http://www.tomra.com/es/qvision


Las primeras 5 especies representaron el 70 % detodos los organismos aislados y los primeros 5 predo-minaron en la flora alterante de los productos cárni-cos cocidos en todas las regiones (figura 2).

Se aislaron otras especies. Ninguna de ellas superó el4 %. Solo 3 de más de 50 productos cárnicos cocidosno contuvieron ninguna de las primeras 5 especies pre-dominantes (pero estuvo presente en estos casos Lac-tobacillus paraplantarum, que está relacionada estre-chamente).

Se puede imaginar que es costoso aislar e identificarmás de 200 cepas, especialmente teniendo en cuentalos costes de las horas de trabajo. Por lo tanto, se empe-zó a usar la metagenómica en Corbion desde 2013. Du-rante los primeros experimentos se usó un método, rin-diendo más de 1 millón de lecturas por muestra despuésde la ampliación de la región V3-V5 del gen 16 S rDNA.Se analizaron también algunas de las muestras median-te aislamiento clásico de organismos, para verificar losresultados de la metagenómica. Entre las muestras hu-bo una muestra t=0 de una salchicha envasada al vacíoproveniente de Asia (< 100 CFU/g), una muestra dete-riorada del mismo lote guardada durante 11 días a 10 °C(condiciones en supermercados asiáticos) así como unamuestra deteriorada que se guardó intermitentementea 30 °C y -18 °C (condiciones en mercados al aire libreasiáticos). El análisis demostró claramente que la di-versidad microbiana inicial fue mucho mayor con t=0 encomparación con la muestra deteriorada final (figuras3a y 3b).

Las salchichas que fueron incubadas a un nivel cons-tante de temperatura (10 °C) demostraron menos di-versidad en comparación con las salchichas incubadas

Análisis y control


Figura 4. Desglose a nivel de especie de una salchichaasiática deteriorada incubada a temperatura constante, y los aislados obtenidos del mismo lote

Lc mesenteroides68%

Lc carnosum18%

Aislados

Leuconostoc carnosum

Leuconostoc mesenteroides

Microbacterium lac�cum

Figura 5. Desglose a nivel de especie de salchichas holandesas deterioradas, incubadas a 7°C

Lb sakei69%

Lc carnosum10%

Lb curvatus8%

Lb graminis5%

Lc pseudomesenteroides4%

Lb fruc�vorans2% Lc gasicomitatum

1%

Figura 3. Desglose de la población al nivel de género presente en una salchicha asiática (a) sin deterioro y (b) deteriorada después de la incubación a 10 ºC durante 11 días

(a) (b)LactobacillusLeuconostocWeissellaLactococcusStreptococcusClostridiaPseudomonasPhotobacteriumPropionibacteriumBacillusBrochothrixEnterobacteriaceaeOtros

a diversas temperaturas (cambios diarios desde -18 ºCa +30°C). Los resultados de metagenómica coincidie-ron bien con los aislamientos (figura 4).

Otras muestras, como las salchichas holandesas, tam-bién mostraron una flora alterante, lo que encaja biencon los primeros 5 organismos alterantes predominan-tes (figura 5). Las muestras analizadas por duplicadoindicaron casi los mismos resultados.

El análisis metagenómico subsiguiente de otras mues-tras de productos cárnicos cocidos envasados al vacíoreafirmó el predominio de los primeros 5 organismosalterantes.

Conclusiones del análisis del deterioro de

productos cárnicos cocidos envasados al vacío

Se pudo determinar la presencia clara de las prime-ras 5 especies predominantes asociadas con los pro-ductos cárnicos cocidos, Leuconostoc y Lactobacillus,predominantes en productos de todo el mundo. La me-tagenómica de los productos cárnicos cocidos fue unéxito reproducible cuyos resultados encajan bien con lasprimeras 5 especies predominantes obtenidas median-te aislamiento. La metagenónica reduce en gran me-dida los costes relacionados con los análisis del dete-rioro.

El deterioro de la carne fresca

La carne fresca sufre generalmente un deterioro rá-pido. Hemos aplicado la metagenómica a diversasmuestras deterioradas de carne fresca envasadas al va-cío obtenidas del supermercado y dos lotes separados

de carne picada de vacuno (8 muestras en total y am-bos lotes envasados al vacío, aire y atmósfera modifi-cada). Se amplificó la región V1-V3 del gen bacteria-no 16S rDNA , y se efectuó la secuenciación decolonias de microorganismos con la tecnología Illu-mina Myseq. El método aplicado obtuvo 2.000-3.000lecturas por muestra. Se calcularon los porcentajes di-vidiendo la cantidad de lecturas pertenecientes a unaunidad taxonómica operativa (UTO) por la cantidadtotal de lecturas de una muestra.

Tres paquetes de f iletes de redondo de vacunoenvasados al vacío provenientes de mataderos si-tuados en distintos países (Irlanda, Reino Unido,Uruguay) y un trozo de filete de cordero, estuvieroninvariablemente colonizados por Lactococcus pis-cium (56,8 % ± 9,6) y Leuconostoc inhae (15,9 ±6,8) (figura 6). En investigaciones anteriores no seencontró ninguna de estas dos especies en produc-tos cárnicos deteriorados. Ambas especies contri-buyeron a un 70 % o más de la flora total en cadamuestra.

En el caso de carne picada de vacuno deteriorada,parece que el tipo de envase tiene cierta importancia,pero es limitada (figura 7). Se realizó un análisis en 8muestras diferentes de 2 lotes diferentes. Durante elprocesamiento de los dos lotes, la carne fue envasadaal vacío, al aire o MAP (80% O2/20% CO2 y 65%O2/35% CO2). La carencia de diferencias reales en mi-crobioma entre los tipos de envases puede deberse aque las muestras de carne picada mostraron altos re-cuentos bacterianos desde el principio.

Aparte de cuatro especies diferentes de LAB pre-sentes sobre un 10 % en muestras diferentes (Leuconos-

Análisis y control


+34 962 871 [email protected]

Metagenómica aplicadaa la seguridad alimentaria

Identifícalos a todosAccede a un nivel superior

de conocimiento sobre las comunidades microbianas


toc gelidum y las especies relacionadas estrechamen-te de Leuconostoc gasicomitatum, Lactobacillus algi-dus y Lactococcus piscium), 6 de cada 8 muestras decarne picada de vacuno contenían al menos 20 % de fo-tobacterias. En 3 de cada 8 muestras, más del 70 % delas lecturas pertenecían a Photobacterium, resultandoser un alterante importante imprevisto de la carne pi-cada de vacuno.

La observación de que había especies de Photobac-terium presentes en grandes cantidades fue una sor-presa para nosotros en el momento en que se realizaronlos experimentos (finales de 2015). Sin embargo, yaen 2011, se mencionaba Photobacterium como organis-mo alterante de la carne fresca al usar PCR-DGGE pa-ra identificar la flora alterante de la carne de vacuno re-frigerada[8].

Recientemente, cada vez más artículos científicosdemuestran la presencia de Photobacterium en carnefresca, y todos los artículos tienen en común que se uti-lizaron métodos independientes en los cultivos[7, 8, 9, 10, 11].

Conclusiones de la investigación

sobre el deterioro de la carne fresca

Las bacterias lácticas halladas en la carne frescadeteriorada son diferentes de la flora hallada en losproductos cárnicos cocidos, y las especies de Photo-bacterium también son organismos alterantes impor-tantes.

Conclusión general de la investigaciónsobre el deterioro de la carne usandodiferentes métodos

Aunque el aislamiento y la identificación clásicosdan una buena reseña de la flora alterante, los costesy el esfuerzo resultan ser considerables. El aislamien-to clásico es, naturalmente, la única manera de obte-ner cultivos puros para una mejor caracterización yobtener curvas de respuesta a dosis y datos sobre cre-cimiento a diferentes valores de pH, temperatura yactividad de agua, y estos son necesarios para for-mar modelos predictivos de crecimiento. Los mode-los predictivos pueden automatizarse hasta cierto ni-vel y siempre necesitarán de ajuste manual. Por lotanto, la cantidad de cepas diferentes que puede usar-se es limitada, principalmente debido a la cantidadde horas que hay que dedicar a recopilar, procesar yconsolidar datos. Por ejemplo, no hay necesidad de200 cepas diferentes.

Por otro lado, la metagenómica permite hacer análi-sis económicos y completos de la flora alterante. Latécnica es independiente de los cultivos y por eso no tie-ne los sesgos que surgen de los cultivos y las selec-ciones. También queda claro que con la metagenómi-ca, se pueden detectar especies que antes no se hallabancon las técnicas dependientes del cultivo. De hecho,la metagenómica ha revolucionado recientemente nues-tra perspectiva sobre el deterioro de la carne fresca alencontrar Photobacterium como uno de los principalesagentes alterantes.

Corbion

Corbion es una empresa productora de ingredientesde origen biológico. Produce y vende a la industria ali-mentaria, entre otros, ácidos orgánicos como ácido lác-tico y sales lácticas. Se utilizan para mejorar el sabory regular la acidez, pero también para reforzar la segu-ridad y aumentar la vida útil de los productos. Muchosproductos presentes en el supermercado contienen es-

Análisis y control


La metagenómica permite hacer análisis económicos y completos

de la flora alterante, independiente de los cultivos y sin sesgos que surjan

de los cultivos y las selecciones

Figura 6. Desglose a nivel de especie de valores promedio a partir de 4 cortes de carne diferentes

Lactococcuspiscium

57%

Leuconostocinhae16%

Leuconostoc sp3%

Carnobacterium sp.1%

Lactobacillales19%

Otros (<1%)4%

tos ingredientes. Especialmente los fabrican-tes de productos cárnicos cocidos como salchi-chas y jamón cocido los utilizan para ralen-tizar el deterioro e incrementar la protecciónfrente al crecimiento del patógeno Listeriamonocytogenes.

Referencias

1.Biesta-Peters EG, Reij MW, ZwieteringMH, Gorris LG. (2011) Comparing non synergy gamma models and interaction mod-els to predict growth of emetic Bacillus cereusfor combinations of pH and water activity values. Appl Environ Microbiol. Aug15;77(16):5707-15.

2.Kreyenschmidt J., Hübner A., Beierle E.,Chonsch L., Scherer A., Petersen B. (2010)Determination of the shelf life of sliced cookedham based on the growth of lactic acid bacte-ria in different steps of the chain. J Appl Micro-biol. 108(2):510-20.

3.Nerbrink E., Borch E. (1993) Evaluation ofbacterial contamination at separate processingstages in emulsion sausage production. Int JFood Microbiol. 20(1):37-44.

4.Borch E., Kant-Muermans M.L., Blixt Y.(1996) Bacterial spoilage of meat and curedmeat products. Int J Food Microbiol 33(1):103-20.

5.Krockel L. (2013). The role of lactic acid bac-teria in safety and flavour development of meatand meat products, lactic acid bacteria - R & Dfor food, health and livestock purposes, Dr. J.Marcelino Kongo (Ed.), ISBN: 978-953-51-0955-6, InTech, DOI: 10.5772/51117.

6.Samelis J, Kakouri A, Georgiadou KG,Metaxopoulos J. (1998) Evaluation of the ex-tent and type of bacterial contamination at differentstages of processing of cooked ham. J Appl Microbiol.84(4):649-60.

7.Delhalle L, Korsak N, Taminiau B, Nezer C, BurteauS, Delcenserie V, Poullet JB, Daube G. (2016) Explor-ing the bacterial diversity of Belgian steak tartare us-ing metagenetics and quantitative real-time PCR analy-sis. J Food Prot. Feb; 79(2):220-9.

8.Pennacchia C, Ercolini D, Villani F. (2011) Spoilage-related microbiota associated with chilled beef stored inair or vacuum pack. Food Microbiol. Feb; 28(1):84-93.

9.Stoops J, Ruyters S, Busschaert P, Spaepen R, VerrethC, Claes J, Lievens B, Van Campenhout L. (2015)

Bacterial community dynamics during cold storage ofminced meat packaged under modified atmosphere andsupplemented with different preservatives. Food Micro-biol. Jun; 48 192-199.

10. Ferrocino I, Greppi A, La Storia A, Rantsiou K,Ercolini D, Cocolin L. (2015) Impact of nisin-activat-ed packaging on microbiota of beef burgers during stor-age. Appl Environ Microbiol. Nov 6; 82(2):549-59.

11. Nieminen TT, Dalgaard P, Björkroth J. (2016)Volatile organic compounds and Photobacterium phos-phoreum associated with spoilage of modified-atmos-phere-packaged raw pork. Int J Food Microbiol. Feb 2;218:86-95. e

Análisis y control


Figura 7. Desglose de la flora en carne picada de vacuno alterada en 4 paquetes diferentes en dos lotes separados

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

aire vacío MAP80 O2/20 CO2

MAP65 O2/35 CO2

Otras bacterias

Photobacterium phosphoreum/kishitanii

Photobacterium phosphoreum

Photobacterium kishitanii

Leuconostocaceae (familia)

Leuconostoc sp.

Leuconostoc gelidum

Leuconostoc gasicomitatum

Lactococcus piscium

Lactobacillus algidus

Lactobacillales (orden)

Carnobacterium sp.

Carnobacteriaceae (familia)

Bacteria (reino)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

aire vacío MAP 80/20 MAP 65/35

Otras bacterias

Pseudomonas taetrolens

Pseudomonas sp.

Pseudomonadaceae (familia)

Photobacterium sp.

Photobacterium kishitanii

Leuconostocaceae (familia)

Leuconostoc sp.

Leuconostoc inhae

Leuconostoc gelidum

Leuconostoc gasicomitatum

Lactococcus sp.

Lactococcus piscium

Lactobacillus algidus

Lactobacillales (orden)

Gammaproteobacteria (clase)


flow citometry

Flow Cytometry and Food Microbiology: challenges, opportunities and progress to date

Martin G. Wilkinson ([email protected])Department of Life Sciences, University of Limerick, Castletroy, Limerick, Ireland

icrobiology is a highly dynamic and important branch of research, quality control and product development for the food industry. In general

terms it can be divided into two main areas, (1) detection and monitoring of pathogens and spoilage microorganisms and (2) the usage of beneficial microorganisms for the control and production of fermented foods. For both of the above activities enumeration and identification of bacterial populations are routinely undertaken using a variety of methodologies including viable cell counting on selective and non-selective media, genetic detection of species specific markers using PCR and the use of emerging technologies such as flow cytometry (FCM) and its more advanced formats such as fluorescence activated cell sorting (FACS) and imaging flow cytometry. The scope of this paper will be to consider the main technical challenges, progress to date and future opportunities for the application of flow cytometry to Food Microbiology. In this regard the use of FCM for analysis of both specific pathogens/spoilage bacteria and beneficial Lactic Acid Bacteria (LAB) will be reviewed as case studies. The food industry globally spends much time and expense ensuring that a range of pathogens/spoilage bacteria are excluded as much as possible from their products. Achieving regulatory standards of food safety for consumers ranging from infants to the elderly population cohorts requires the food industry to employ microbiologists in quality control laboratories. This activity is designed to prevent consumer illnesses including those from ingestion from foods containing live bacterial pathogens such as Enterobacteriaceae, E. coli 0157:H7 and Staphylococcus aureus, or, from ingestion of emetic toxins from spore formers such as Bacillus cereus. All of the above species can generate a range of conditions from mild to serious and potentially life threatening. Conversely, a wide variety of fermented foods are consumed which are

rendered safe, nutritious and flavoursome by the addition of bacteria such as LAB which stabilise the products by colonisation, acidification and the release of enzymes which generate optimal flavour, texture and aroma. The list of products which rely on LAB activity is extensive and includes natural cheeses, yogurts, fermented milk drinks, fermented meats including salami, fermented vegetables such as pickles or sauerkraut and beverages such as wines and beers.

Limitations of current plate based techniques

Currently, except for a few notable exceptions the food industry is still using quality control analytical techniques based on plating of diluted samples onto media and counting the resultant colonies after incubation under defined conditions of time and temperature. In the case of certain pathogens such as S. aureus this process involves plating onto Baird-Parker agar and identifying typical colonies, which can take up to 48 hours. Moreover, further identification of coagulase positive suspects which can prolong the duration before a definitive result can be obtained for up to 4-5 days. Overall this methodology is dependent on a “plate and wait” approach whereby data on product safety is generated after a time lag and in the interim the product may require to be stored before a positive release is obtained. In addition, a cornerstone of this approach is the presence or absence of live cells or those capable of growth under highly selective conditions. However, what is becoming much clearer in modern microbiology is that cells especially in foods that have been processed are present in a multiplicity of physiological states including; live, damaged/permeabilised and dead with varying potential for outgrowth which may render them non-detectable under conventional plate based assays.

M


flow citometry

Flow Cytometry (FCM) for Food Microbiology: the basics

One emerging technology that may offer the food industry a new approach towards bacterial detection in terms of enumeration, identification and physiological profiling is flow cytometry (FCM). The principles of physical and engineering principles of FCM are very well explained by Shapiro (2015; 2003) and will not be dealt in detail here. However, in terms of microbiological applications FCM involves suspending bacterial cells within a moving liquid stream which is then interrogated by a laser and gathering of resultant light scattering properties in a Forward angle light scatter (Forward scatter or FSC) or Side angle light scatter (Side scatter or SSC); additionally, fluorescence arising from cells passing thorough the laser beam can be collected at various wavelengths. While FSC and SSC can be used to locate cells on a cytograph profile and give general information on cell size and granularity, the major information on individual cell physiology and structure is obtained by staining of cells with specific fluorescent dyes and collecting the degree of fluorescence from a single or multiple stains applied to cells. The major advantages/potential applications of FCM for food microbiology are: rapid assay times and data generation (1-2 min), high number of cells that can be analysed per sample (10,000 and upwards), minimal sample volume (from 5 ml), potential high throughput, multiplicity of stains available to examine various aspects of cell viability, structure and/or metabolism (multi-parametric), and less labour and space intensive compared with conventional plating techniques (Wilkinson, 2015).

FCM: general technical challenges

If the above sounds too good to be true then it can be balanced with a heavy dose of reality! FCM-based assays for cells in foods are still beset by a range of both minor and major technical and perhaps regulatory and mind-set obstacles. Firstly, cells have to be recovered from foods in samples which are essentially free of interfering “bacteria-sized” particles such as debris obtained after homogenisation and stomaching commonly used to release cells for conventional plate analysis. The use of food samples which contain particles which generate auto-fluorescence will interfere with detection of bacterial cells and therefore may require some specialised removal procedures prior to FCM. If analysis using FCM involves complex formulated food samples, such as prepared ready meals, the above problems may exist along with the presence of ingredients containing DNA/RNA (such as food flavourings) which will non-specifically bind with DNA-based cell viability dyes. In terms of detection limits most cytometers operate best when analysing 105-106 cells per sample so in general terms you need a good population of cells to get optimal analysis. This means that you either use concentrated samples containing target cells or a carry out a prior enrichment step to increase the low level of target cells to detectable levels. The latter step thus renders the FCM based assay as giving a qualitative “presence or absence” data output. A key drawback for pathogen analysis in foods is the inability of most of the stains used by FCM to enable a high degree of species-specific pathogen identification. However, the use of DNA based techniques such as fluorescent in situ hybridisation (FISH) when combined with cytometry (FLOW-FISH) or the use of mono/polyclonal antibodies has certainly made some progress. Overall, it is fair to say that FCM has not yet benefitted from the availability of cheap, reliable and sensitive antibodies with which to target and label particular a pathogen within a mixture of species. However, this paper will focus on the progress and application of antibody labelling of cells as a potential avenue for immuno-cytometric pathogen detection.

FCM assay development

To consider progress in FCM for Food Microbiology it is necessary to follow a typical assay development route then to compare and contrast it with conventional techniques. Firstly, in terms of release and recovery of cells in foods for FCM analysis the procedure follows quite common initial routes used for traditional plating analysis including: dilution of sample, mixing via stomaching or homogenisation and final production of a suspension of food particles and bacterial cells. Thereafter, the necessity for the use of specific clean up procedures for FCM analysis is encountered. These, and other procedures typically used in bacterial sample preparation methods have been reviewed in detail by Dwivedi and Jaykus (2011). Recovery of cells may involve an initial low speed centrifugation (3,000 x g for 10 min) to obtain a cell pellet, followed by resuspension

Figure 1. FCM and Potential Applications in Food Microbiology/

Food Fermentations (Source: Doolan and Wilkinson, 2016).


flow citometry

of the pellet in buffer and thereafter filtration through various particle exclusion membranes such as 15, 10, or 5 µm to reduce non-cellular particulates while a final centrifugation step may often be needed to concentrate cells. Thereafter, the cell pellet is either stained directly or may be antibody-labelled to detect a specific population followed by differential staining. If the population of target cells is below the sensitivity of the cytometer or fails to produce a signal well beyond background “noise” then a further enrichment step is often necessary to enable growth of detectable levels of cells for analysis by FCM. A single step method we use in our laboratory to both recover and concentrate LAB cells from semi-hard Cheddar cheese for FCM analysis involves a 3 hr extraction to express the aqueous phase (“juice”) from 300 g of grated cheese sample at 320 MPa of hydraulic pressure to yield a relatively clear solution which can be centrifuged and contains from 104-109 cells per ml (Wilkinson et al., 1994; Sheehan et al., 2005; Yanachkina et al., 2016). This particular sample is highly versatile and can be analysed simultaneously for FCM profiling alongside the determination of released peptidolytic enzymes and peptide/amino acid ripening products. As Cheddar cheese is a semi-solid having a moisture content of 38-40% it may well be possible to apply this system, or some modifications thereof, to other food types as a potential integrated cell recovery/concentration procedure. However, at this juncture it is reasonable to suggest that for FCM analysis additional sample preparation steps are required when compared to current ISO-accredited methodologies used for viable plate counting of pathogens and other microorganisms. However, method development is proceeding apace for FCM and it should be possible in the medium term to identify agreed validated procedures for sample preparation and cell extraction from various food groups. Regarding sample volume, FCM can certainly claim to be more flexible in this aspect compared with plate counting (traditionally requiring 1 ml for pour plating) as volumes from 5 ml up to 100 ml can easily be analysed in the cytometer based on initial cell numbers and flow rate manipulation. Analytical speed for FCM is quite amazing and tens of thousands of cells (known as “events” in FCM-speak) can be analysed and their data obtained within 2-3 minutes per sample. Contrasting that with conventional plate counting where a range of dilutions must be prepared, plated and then incubated, following which, those giving from 30 to 300 colonies per plate are selected as having statistically valid data.

Dyes/Stains useful for FCM in Food Microbiology

Assuming that a good particle free sample has been obtained with sufficient cell concentrations, what can FCM offer in terms of analysis? The methodology used for microbial analysis is based on fluorescent staining of cells, this can involve, at its simplest, a single stain for enumeration of the entire population or combinations of stains where multiple cellular characteristics are analysed. However, a wide range of stains are commercially

available and must be chosen on the basis of analytical data required, non-toxicity to the cell and suitability for detection by the laser in the particular cytometer along with other factors such as spectral over-lap with other stains (Leonard et al., 2016). Single viability stains, which are designed to bind to the DNA/RNA of the cell population such as SYTO9 or Thiazole Orange (TO), allow a general estimation of bacterial biomass based on the fact that these stains can enter all cells and bind to the nucleic acid of both live (intact), injured or dead cells. However, this data is of limited use for obvious reason, hence the general approach is to use combinations of stains to measure various aspects of cell viability and metabolism. The Live/Dead BacLight stain combination from Invitrogen (USA) is comprised of SYT09 and propidium iodide (PI)—a green and a red fluorescent dye combination that has been very useful in determination of cell viability and is based on the principle of displacement of the green SYTO9 dye from cells with damaged membranes by the red dye, PI. In general, a range of stains can be combined to give multi-parametric data on aspects of membrane integrity, intracellular enzyme activity (cFDA), membrane potential, (DiBAC4 (3)), intracellular pH (BCECF), and cell division (CFSE). It is beyond the scope of this paper to describe the attributes of the range of dyes available for food microbial applications but readers can refer to the chapters within the text of Wilkinson and other authors (2015) for a more detailed discussion on their merits. A key technical issue to remember when looking at dye combinations for differing microbial species/strains is that a preliminary check should be carried out to ensure the stains are not toxic to cells at dosages used in the particular application. Therefore a parallel experiment using the stains with cross checking of viability by plate counting should be carried out to establish correlation between methods and whether any bacterial strain or FCM-stain related toxicity is present.

Antibody Labelling and FCM analysis: progress, challenges and future opportunities

As mentioned earlier, the current available dye combinations are generally not indicative of the presence or absence of particular bacterial species within a mixed population. An avenue which is being developed is the use of antibodies to label specific

Figure 2. Flow Cytometry Profile arising from differential

staining using SYTO9/PI combinations showing live, dead and

damaged/permeabilised sub-populations.


flow citometry

bacteria and then either measure the fluorescence directly or amplify the primary antibody-bacterial cell complex using a secondary antibody which gives an enhanced signal suitable for collection by cytometry. One such application of this technology is Immunomagnetic Separation or IMS. This utilises magnetic coated antibodies against target cells which are specifically labelled within a liquid sample and then retained or captured on application of a magnetic field, resulting in the concentration of target cells from a mixed culture. Thereafter, these cells may then be directly subjected to FCM analysis. Hibi et al., (2006) reported the separation of Listeria monocytogenes from mixed cultures by IMS followed by FCM analysis over range of 104-108 CFU/ml. In general, antibody capture of target cells such as E. coli 0157:H7 or Listeria monocytogenes requires further enrichment to allow cell numbers to reach detectable levels using FCM (from 104 cfu/ml and upwards). I would like to use a number of published studies to illustrate the challenges, progress and opportunities for immuno-FCM in food pathogen analysis. Wilkes et al., (2012) described the development of methodology for the rapid and sensitive detection of E. coli 0157:H7 in spinach. These workers used two proprietary reagents, A or B, containing either FITC-conjugated polyclonal antibodies against the pathogen or detergents conditioning chemicals for enhanced epitope presentation along with PI stain for dead cells, respectively. In terms of sample preparation, this product is quite challenging as regards particle interference, matrix colour interference and subsequent recovery and detection of pathogen. Spinach (25 g) was spiked with 100 ml of either 5 or 50 cells and followed by a 4 hour enrichment step. In terms of sample preparation, liquid media added to the spinach was decanted and used as the starting sample. Thereafter, the use of buoyant gradient centrifugation essentially allowed a concentration of cells which were then removed, washed and filtered (5 mm) prior to antibody labelling and cytometric analysis. The use of a specialised cytometer having a wide cross section flow cell with a 130 nm resolution was reported by the authors as being superior for analysis of samples containing particulates likely to interfere with bacterial analysis. The approach used by these workers for FCM analysis was quite interesting and utilised a series of multi-dimensional gates beginning from the usual initial FSC and SSC plot, progressing thereafter to exclude PI positive dead cells and matrix particles to finally enable exclusive detection of live labelled E. coli 0157:H7. The performance of this qualitative or screening assay represented a Time to Result (TTR) of under 4 and a half hours, with a limit of detection of 1 viable cell in 25 g of sample. This excellent work represents tangible progress for immuno-FCM with the development of an assay suitable for generation of a “presence or absence” result for this pathogen well within a typical 8 hour production cycle. However, the inability to directly detect low pathogen numbers without an enrichment step along with the relatively detailed sample preparation procedure still falls short of a more desirable direct FCM based enumeration assay. Williams et al., (2015) outlined further progress on this assay by means of a Level 2 FDA approved validation process for detection of E. coli 0157:H7 in

raw spinach. The validation procedure involved 20 spiked and 20 non-spiked samples for analysis with a comparison of FCM and an FDA approved q-PCR test. Levels of 1-4 viable cells per 100 ml were inoculated into the test spinach samples with a 17.5 hour aging period before sampling. Subsequent preparation steps included the addition of a photobleaching agent, phloxine B, to reduce matrix colour interference and a 5 hour incubation to increase cell numbers to detectable levels. Using reagents A and B, detection was carried out as before. In this report another wide diameter flow cell cytometer was used but with a somewhat larger resolution of 170 nm. Overall performance of the FCM method was very favourable when compared with the approved q-PCR method. Sensitivity was similar between methods at 2-4 cells per 100 ml, TTR for the FCM method was ~9h while that for the PCR was ~51h, this FCM method can potentially deliver data within a processing shift while its throughput was estimated at 20 samples during this interval. The number of false negatives for FCM was 4 out of 10 and for the PCR method was 5 out of ten. This study represents further progress towards a commercial and accredited qualitative FCM immuno-assay for food pathogen detection. Importantly, it provides solid data regarding the comparative performance of immuno-FCM with accredited assays such as q-PCR and therefore builds confidence for the adoption of FCM assays by the food industry. The work of Subires et al., (2014) also provides a very good insight into the complexities of FCM assay development for particular food groups such as prepared pasta salads which contain a range of ingredients any, or all, of which may be difficult to remove from the FCM analyte. These authors also report on a method to detect E. coli 0157:H7. The preparatory steps used to recover cells and reduce interference included pulsification in 63 mm bags and centrifugal filtration. In this work, 103-107 cfu/g of live cells were inoculated into the unpasteurised salads which were stored at 4 °C for 2 weeks. Detection of the target pathogen was carried out using a polyclonal antibody conjugated to R-phycoerythrin (R-PE, yellow-orange fluorescence), while live/dead staining was carried out on the labelled complex using SYBR GREEN I and PI. In this study, the authors did not use an enrichment step rather a direct labelling of a filtered and re-suspended cell pellet was undertaken. Correlations were also undertaken with conventional plate counting of samples. A careful selection of gates using a good range of controls allowed discrimination of cells from food particles and between live or dead/damaged cells. Data generated indicated an LOD of 105 cfu/g attributed to particle interference despite the preparatory steps undertaken. A good correlation was found between FCM and plate count data at a particular antibody concentration, showing the necessity for extensive assay optimisation for immuno-FCM to perform optimally. The additional information that multi-parametric FCM can provide was well illustrated from the SYBR Green I and PI combination which revealed that initially most cells had sustained membrane damage but appeared to have recovered by day 14. This physiological insight is beyond plate counting, and some other techniques, but may allow a greater understanding of the fate of stressed cells in foods and to estimate their potential


flow citometry

for posing a latent threat to consumer safety from convenience foods. Using the above studies as a method to compare and contrast the ability of immuno-FCM to overcome the current issues with cell specific recovery and labelling it can be seen that good progress has been made towards the development of sensitive and rapid qualitative detection of particular pathogens in a limited range of foods. However, the direct species-specific enumeration of pathogens using sensitive quantitative immuno-FCM assays is still some way off and will rely on the development of novel methods to recover cells from foods in concentrations which can be detected by sensitive antibodies and using cytometers which are specially adapted for sensitivity at low event detection limits.

Spores: “the Final Frontier”

Spore forming bacteria are especially important to detect and quantify in the food industry, including Bacillus cereus in prepared consumer foods such as rice and cereal dishes. Bacillus cereus is ubiquitous in soil, and can enter the food chain at a very early stage such as at harvesting of cereals or into dairy products during milking. Generalised spore count monitoring in the food industry is often used to indicate the status of process hygiene and the likelihood of the associated presence of the highly dangerous pathogen Clostridium botulinum. Generally, testing for sporeformers in foods involves laborious methods initially in enrichment media, followed by growth on selective agars with confirmatory testing by molecular methods. Hence, FCM is being investigated as a potential method to detect and study spores and vegetative cells following outgrowth in foods. I will deal with progress made using FCM for the study of Bacillus cereus which was also reviewed by Cronin and Wilkinson (2009). In general, FCM has been successfully used to study a range of properties involving spores, their germination events,

outgrowth of vegetative cells and survival of vegetative cells within model food systems. This in turn has formed the basis for cytometric and biochemical evaluations of the effects of various food processing treatments on spores/vegetative cells. Cronin and Wilkinson (2008) showed that a number of heterogenous sub-populations were generated following exposure of B. cereus endospores subjected to simulated cooking temperatures and time regimes using FCM together with SYTO 9 ⁄ PI and CFDA⁄ Hoechst 33342. In terms of the development of direct FCM based assays for spores and/or vegetative cells, there is still much work to be done on aspects such as cell recovery from foods, rapid differentiation of spores from vegetative cells and of course the provision of sensitive, low cost specific antibody probes with which to label cells for FCM analysis. The opportunities and challenges for FCM especially immuno-FCM for spore/vegetative cell detection in the food industry are worth re-iterating. Any new FCM method should be more rapid than growth-based methods, with data gained within hours rather than days currently required for plate counts. Confirmatory tests could be eliminated should specific antibodies become available while useful multi-parametric physiological data could also be generated by fluorescent staining of the same sample.

Flow Cytometry and Lactic Acid Bacteria (LAB)

Lactic Acid Bacteria (LAB) are the main bacteria used as starter cultures in the fermentation of milk, cheeses, yogurts, wines, meats and vegetables. They are inoculated into food substrates such as cheese to produce lactic acid within a specific production time period. Thereafter, during ripening the LAB cells become non-viable and release their intracellular peptidase enzymes such as Pep X to ripen the cheese and generate the typical flavour for that variety. It has been known for many years that LAB strains especially Lactococcus. lactis subsp. cremoris or

Figure 3. Potential

applications of Cell

Sorting to Pathogen/

Target Cell detection in

Food Microbiology (Source:

Cronin, Alonso-Gómez and

Wilkinson, 2016)


flow citometry

Lc. lactis subsp. lactis differ in respect to viability and that Lc. cremoris strains generally die off faster to autolyse and release intracellular enzymes e.g Pep X, thus ripening the cheese in a balanced fashion without off-flavours . However, the relationship between low viability, autolysis and enzyme release has not been fully answered and a question that has arisen over recent years is whether differences in cell permeability influences intracellular enzyme release into the food matrix. Traditionally, LAB viability has been measured by a decrease in population recovered on selective media such as L-M17. Quantification of strain-related autolytic properties is currently undertaken by measurement of the activity of released intracellular marker enzymes of which Pep X and Lactate Dehydrogenase (LDH) are notable. However, the issue of demonstrating and quantifying strain related permeability properties is not well suited to the previous methods and here is where FCM analysis has proven to be very helpful. The general area of FCM and LAB has been reviewed by Doolan et al., (2015). The application of FCM methodology for LAB was reported by Bunthof et al., (2001) for pure cultures grown in broth media who showed that for LAB and other species labelling with c FDA was useful for determination of the fate of live cells before and after heat treatment. However, these workers did not achieve good discrimination of live/dead cells for Lactococcal strains using c FDA but this was improved for Lactobacillus strains. Overall, TOTO-1 staining gave good discrimination of LAB in either live or dead states based on the generation of a highly intense fluorescence signal from dead cells. Bunthof and Abee (2002) successfully applied this FCM double-staining methodology for the determination of viability of LAB including probiotic bacteria in milk, starter preparations and in commercial probiotic drinks In terms of determination of strain-related cell permeabilisation properties in higher solids dairy products such as Cheddar cheese, Sheehan et al., (2005) used a combination of SYTO9 and PI to determine the percentage live/intact, permeabilised/damaged and dead sub-populations in cheese manufactured using two lactococcal strains and demonstrated the existence of differing percentages of permeabilised cells between these strains. Interestingly, one strain became non-viable, highly permeabilised and also released intracellular Pep X (Lc. lactis AM2). The other strain was less permeabilised (Lc. lactis HP) but these permeabilised cells did not appear to release intracellular Pep X to any substantial degree and typically the resultant cheese lacks optimum flavour and can be bitter. Overall, FCM and LAB including cheese starters and probiotics have benefited significantly from FCM methodology which has allowed a unique insight into aspects of culture performance with economic and human health implications. It should be pointed out here that determination of LAB in fermentations in most cases is rendered less difficult than pathogen detection by virtue of the fact that sample populations from LAB fermentations are generally of the order of 106 cfu per ml or per gram

and upwards. Hence, samples are often diluted (rather than concentrated for pathogen detection) to achieve correct cell concentrations for FCM analysis, for which those of us who work on FCM for LAB analysis are very grateful!

Sorting out the issues!

The use of fluorescent activated cell sorting (FACS) commonly known as “cell sorting” represents a higher grade of FCM-based analytical resolving power for use in food microbiology. Briefly, cell sorters are modified cytometers with the ability to gate and physically sample the cells from a sub-population of interest. This resolving power arises from the ability of the cell sorter to enclose a single cell of interest within a droplet which is then given either a positive or negative charge and deflected by passage through a high voltage electrical field (~5,000 V) to be deposited onto an agar plate, test tube, or a microtitre plate well for further analysis (Muller and Nebe-von-Caron, 2010). Using this technology we can recover stained cells which have been exposed to various stressors and correlate their cytometric profile which may reflect their differing physiological states with their ability to recover and grow on various media. Therefore cell sorting can potentially simultaneously answer a number of questions regarding FCM; (1) what equivalence has FCM with plate counting for particular bacterial species? (2) do VBNC cells exist within a population? and, (3) can a rapid sorting-based multiplexing assay simultaneously discriminate and confirm the presence of different species within a single sample? (4) can sorting evaluate the effects of stressors (heating, cooling, etc) on survival and recovery of cells. In terms of some of these questions, Kennedy et al., (2011) examined the responses of the food pathogens E. coli, Listeria monocytogenes and Staphylococcus aureus when subjected to the effects of various stressors encountered by pathogens during food processing. The strains were analysed by FCM for viability using SYTO9/PI or for the presence/absence of a functioning membrane potential by staining with DiOC2 (3). Using FACS sorting of various sub-populations, these workers showed that extensively damaged cells (as per staining and FCM profiles) sorted onto various solid media were still capable of growth. Indeed, differing overall recovery rates were noted for the various pathogens on both selective and non-selective media and these recoveries were also affected by differing stressor treatments. This study clearly showed the potential for survival and outgrowth of damaged cells, it also demonstrated the heterogenous nature of the sub-populations generated within a culture following exposure to stressors and demonstrated the potential usefulness of cell sorting in gaining a novel profound insight into cell physiology and potential pathogenicity arising from persistence of damaged cells within foods. Cell sorting is still expensive and generally non-user friendly for applications within a routine food microbiology quality analysis laboratory however new cheaper sorters are becoming available and in the medium term could become lend themselves to specialised analysis of particular food pathogens.


flow citometry

Let it Flow! How the water quality sector is showing the path towards acceptance and accreditation of FCM methodologies for the Food Industry

In order to see what possibilities for FCM and Food Microbiology there may be, it is necessary to compare and contrast with an industry/sector in which FCM is becoming a standard quality and production monitoring technique. The provision of a simple reliable and rapid microbiological test for water quality has long been a goal of workers in this field and traditionally it involved an agar based heterotrophic plate count (HPC) which was of limited usefulness. Egli and Kotzsch (2015) describe in detail the development and widespread acceptance (including by national regulatory bodies) of an FCM based assay to determine total cell count (TCC) and the ratio of High Nucleic Acid to Low Nucleic Acid content of cells in the population of a drinking water sample. The clearly demonstrated unrivalled ability of FCM to rapidly generate a meaningful data set from water samples was the result of work done by the drinking water research group at Eawag in Switzerland and led initially to the uptake by laboratory practitioners and thereafter by the regulatory authorities such that validation and accreditation of FCM methods for water quality are now in place in Switzerland and work is now nearing completion for the use of specially developed cytometers for on/in line quality control which will provide continuous quality control data and enable immediate remedial action to be undertaken in response to problems. This, in my view, is the vision we can have for flow cytometry in the food industry, however many technical and regulatory challenges remain before we can reach this goal. Overall, much progress has been made and indeed a few FCM based analytical quality control systems have been commercialised so perhaps the “cytometry based future” may not be that far distant!

References

• Bunthof,C.J.andAbee,T. (2002)DevelopmentofaFlowCytometric Method to Analyze Subpopulations of Bacteria in Probiotic Products and Dairy Starters. Applied and Environmental Microbiology, 68, 2934–2942.

• Bunthof, C. J., K. Bloemen, P. Breeuwer, F. M. Romboutsand T. Abee. (2001). Flow Cytometric Assessment of Viability of Lactic Acid Bacteria. Applied and Environmental Microbiology. 67, 2326-2335.

• Cronin, U.P. and Wilkinson, M.G. (2008) Bacillus cereus endospores exhibit a heterogeneous response to heat-treatment and low temperature storage. Food Microbiology, 25, 235–243.

• Cronin,U.P.,andWilkinson,M.G.(2009).Thepotentialofflow cytometry in the study of Bacillus cereus. Journal of Applied Microbiology, 108, 1-16.

• Dwivedi,H.andJaykus,L-A.(2011).Detectionofpathogensin foods: the current state-of-the art and future directions. Critical Reviews in Microbiology, 37, 40-63.

• Doolan,I.A.,Wilkinson,M.GandHickey,D.K.(2015).TheApplication of Flow Cytometry to the Study of Lactic Acid Bacteria Fermentations. In: Flow Cytometry in Microbiology Technology and Applications. Caister Academic Press, UK. Ed. M.G. Wilkinson.

• Egli, T. and Kotsch, S. (2015). Flow Cytometry for RapidMicrobiological Analysis of Drinking Water: From Science to Practice - An unfinished story. In: Flow Cytometry in Microbiology Technology and Applications. Caister Academic Press, UK. Ed. M.G. Wilkinson.

• Hibi, K.,Abe,A.,Ohashi, E.,Mitsubayashi,K.,Ushio,H.,Hayashi, T., Ren, H., and Endo, H. (2006). Combination of immunomagnetic separation with flow cytometry for detection of Listeria monocytogenes. Analytica Chimica Acta, 573-574: 158-163.

• Leonard, L., Bourab Chibane, L., Ouled Bouhedda, B.,Degraeve, P. and Oulahal, N. (2016). Recent Developments in Multi-Parameter Flow Cytometry to Characterize Antimicrobial Treatments. Frontiers in Microbiology, 7, Article 1225, 1-16.

• Muller, S., and Nebe-von-Caron, G. (2010). Functionalsingle-cell-analyses: flow cytometry and cell sorting of microbial populations and communities. FEMS Microbiology Reviews, 34, 554-587.

• Sheehan, A., O’Loughlin, C., O’Cuinn, G., Fitzgerald,R.J., and Wilkinson, M.G. (2005). Cheddar cheese cooking temperature induces differential lactococcal cell permeabilization and autolytic responses as detected by flow cytometry: implications for intracellular enzyme accessibility. Journal of Applied Microbiology. 99, 1007–1018.

• Subires, A., Yuste, J., and Capellas, M. (2014). Flowcytometry immunodetection and membrane integrity assessment of Escherichia coli 0157:H7 in ready-to-eat pasta salad during refrigerated storage. International Journal of Food Microbiology, 168-169, 47-56.

• Wilkies,J.G.,Tucker,R.K.,Montgomery,J.A.,andCooper,W.M. (2012). Reduction of food matrix interference by a combination of sample preparation and multi-dimensional gating techniques to facilitate rapid, high sensitivity analysis for Escherichia coli serotype 0157 by flow cytometry. Food Microbiology, 30, 281-288.

• Wilkinson,M.G.,Guinee,T.P.,O’Callaghan,D.M.,andFox,P.F. (1994). Autolysis and proteolysis in different strains of starter bacteria during Cheddar cheese ripening. Journal of Dairy Research, 61, 249–262.

• Williams,A.J.,Cooper,W.M.,Summage-West,C.V.,Sims,L.M., Woodruff, R., Christmas, J., Moskal, T.J., Ramasaroop, S., Sutherland, J.B., Alusta, P., Wilkes, J.G and Buzatu, D.A. (2015). Level 2 validation of a flow cytometric method for detection of Escherichia coli 0157: H7 in raw spinach. International Journal of Food Microbiology, 215, 1-6.

• Yanachkina, P., McCarthy, C., Guinee, T., and Wilkinson,M.G. (2016) Effect of varying the salt and fat content

in Cheddar cheese on aspects of the performance of a

commercial starter culture preparation during ripening.

International Journal of Food Microbiology, 224, 7–15.

técnicas de LABORATORIO 666 Nº 426 NOVIEMBRE 2017

food control

1. Microbiological environmental monitoring and food safety

icrobiological monitoring programs play a crucial role in the verification of the effectiveness of implemented hygiene control measures as defined

in prerequisite and operational prerequisite programs as well as in HACCP plans.

The safety of foods can only be ensured through the implementation of effective preventive measures and cannot be achieved just by analyzing finished products. Testing of finished products for a pathogen is clearly not sufficient to guarantee their safety as low but significant levels of pathogens may not be detected even when sampling and testing large number of samples.

Figure 1. Statistical limitations of finished product testing

(from ICMSF 2002, Volume 7).

Microbiological monitoring plays an essential role to detect deviations such as presence of pathogens in processing environments or increasing levels of hygiene indicators as. This allows to rapidly take corrective actions to ensure the safety and compliance of finished products.

Microbiological monitoring programs are used to verify the effectiveness of the implemented hygiene control measures. They are composed of the following four elements:

•Rawmaterials•Environmentalsamples•Linesamples•Finishedproducts.

Environmental sampling aims at detecting as early as possible the presence of pathogens or deviations of hygiene indicator beyond defined limits in order to rapidly apply corrective actions. These samples may be subdivided according to a priority rating which reflects the impact on processing lines and product in case of contamination:

•Samplestakenfromexternalsurfacesofequipmentwherea direct contamination of the processing line or even of the product could easily occur.

•Samplestakenfromsamplingsitesfromareasfurtherawayfrom the processing lines or exposed products, where presence of a pathogen could easily lead to its presence in locations close to the product or processing line.

•Samplestakenfromsamplingsitesfromveryremoteareasand therefore no risk for processing lines and product. Results from such samples may provide information on specific niches or in case of investigations.

Sampling techniques for microbiological environmental monitoring in the food industry

David Tomás Fornés. Nestlé Research Center. Lausanne (Switzerland)[email protected]

M


food control

A correct selection of sampling sites and sampling tools is critical to assure the quality of the subsequent analysis and results.

2. Enviromental samples

Complete absence of pathogens and levels of hygiene indicators consistently below established limits are certainly the target to ensure the food safety of products. Such results may however also reflect a poor microbiological monitoring program, inappropriate sampling procedures, inappropriate conditions during transportation of samples, their handling and testing. Such results need therefore to be challenged and procedures scrutinized carefully to determine whether there are weaknesses and identify opportunities for improvement.

To be representative and useful, sampling must be done taking the following elements in consideration:

•Correct timing reflecting the worst case situation such asbuild up

•Correct timing to avoid a bias caused by cleaning and“hiding” the real situation

•Correct timing to avoid affecting results by residues ofcleaning agents or disinfectants

•Detailsonlocationswherebuildcantakeplace•Appropriatetoolstorecovercontaminatedresidues.

3. Surface sampling methods

The purpose of sampling of surfaces in food industry is to determine the presence of, or the number of, viable microbes on the surfaces of utensils, work surfaces and other equipment in contact with food to estimate the level of contamination during production, or the effectiveness of cleaning and disinfecting protocols.

Most typical surface sampling methods are the contact plates and the swab method, both included in the ISO 18593:2004. “Horizontal methods for sampling techniques from surfaces using contact plates and swabs”.

3.1. Contact agar method is only applicable to flat surfaces. A contact plate (or dipslide) filled with a suitable agar medium is pressed against the surface to be tested. After incubation, an estimate of the surface contamination is obtained by counting the number of developed colonies. This sampling method is not applicable to pathogen detection methods, when a pre-enrichment step is required.

Contact plates, also named RODAC (Replicate Organisms Direct Agar Contact) are plastic dishes with diameter 65 mm, filled with a controlled volume of agar medium (chosen according to the target microorganisms), especially made for

the sampling of surfaces. Dishes vary in diameter or area, according to the type of surface to be sampled (usually aprox. 25 cm2), considering the agar shall form a convex meniscus with the dish.

Dipslides are synthetic slide (7 cm2 to 10 cm2), one or both sides of which are covered with a layer of a solid growth medium (chosen according to the target microorganisms).

Surfaces are analysed by pressing the agar surface of the contact plate or the dipslide firmly and without any lateral movement against the flat and regular test surface. For a reproducible and optimal results, it is recommended (ISO 18593:2004) a contact time of 10 s and a pressure obtained with a mass of 500 g. These specific conditions can be obtained using specific devices like Count-Tact© (BioMerieux ref. 96300).

Example of RODAC plate (with central grid).

RODAC plate + Count-Tact© (from BioMérieux).

Dipslide (from 3M).

Figure 2. Direct contact agar sampling.

3.2. Swab method can be used for all types of surfaces. For the sampling of large surfaces (> 100 cm2), sterile cloths or sponges can be used. Using the swab method, a specified area of the surface to be examined is marked (e.g. using a template) and then wiped.


food control

The ideal sample collection material is one that (Ward, 2012 a):

• Isfreeofanytoxicsubstancesthatwouldcauseinjuryorbelethal to microorganisms after the surface sample has been collected and before the sample has been tested

• Issufficiently ruggedsothatthesurfacecanbevigorouslyscrubbed to disrupt and lift biofilm without disintegration of the collection material

• Iseffectiveatcollectingthemicroorganismsfromthesurface•Releases all of the microorganisms during a procedure

designed to count levels of microorganisms collected•Doesnotinterferewithadiagnostictestperformedonthe

sample and produce false positive or false negative results.

Figure 3. Different swabs and sampling devices (from Faille et

al, 2014).

Polyurethane foam Sponge Samplers (eg. Ref. EZ Reach™ Worldbioproducts) is designed to allow the user to aggressively sample an environmental surface to disrupt biofilms and lift strongly attached cells. It is less likely to fragment or tear during sampling and big areas can be sampled. Bacteria can survive on the polyurethane sponges up to 72 hours at refrigerated temperatures.

Biocide-free cellulose sponges are manufactured from natural materials such as wood pulp and vegetable fibers; some batch to batch variability in chemical and/or mechanical properties may occur. Cellulose sponges have been recommended by the USDA for carcass sampling.

Dacron fibers are usually applied to small areas with limited access and may fray or unwind during sample collection. In the study conducted by Botrugno et al., 2015, flocked tipped swabs (COPAN SRK™) showed an average recovery of 69,6% higher than classic dracon (viscose) swabs.

Faille et al., 2014 evaluated the interaction between swab materials versus sampling surfaces for biofilm recovery, showing there is not a single swab material giving a high recovery in all surfaces (see Table 1).

Table 1. Ability of microbial tests to detect biofilms on the

different materials.

Swabbing can be performed in dry conditions (usually to collect contamination in dry environments like milk powder, chocolate) or in wet conditions. Wet swabbing usually allows a better recovery in different surfaces compared with dry swabbing (see Figure 4).

Figure 4. Recovery in wet and dry conditions in different

surfaces. . (Data provided by Nestlé Product Technology Center,

Konolfingen)

After swabbing, the swab sticks are broken into a tube or bottle containing a sterile dilution fluid or neutralizing fluid and mixed by hand. If the surface is wiped with a sterile (damp) cloth or sponge, the sampling device is stored in a known volume of dilution liquid (e.g. 100 ml for 100 cm2).

In the laboratory the initial suspension and, if necessary, further decimal dilutions are used to determine the presence or the


food control

number of microorganisms using the procedures described in the methods for the enumeration or detection of the (groups of) microorganisms to be investigated.

In both cases, after sampling, the surface is cleaned and disinfected, if necessary, to avoid traces of nutrients resulting from the sampling procedure remaining on the sampled surface.

4. Neutralizers and diluents

A collection broth has two primary purposes (Ward et al., 2012 b):

•Toneutralizesanitizersthatmaybepresentonthesurfacethat is being sampled

•Tomaintaintheviabilityofthemicroorganismsafterasampleis collected and until the sample is processed by the laboratory.

In general, the base for neutralizing liquid is buffered peptone water, or peptone salt, or any other appropriate diluent (such as quarter-strength Ringer’s solution, phosphate buffer at pH 7,5, peptone solution at 1 g/l). However other diluents have been developed to improve transport and viability of bacteria.

In cases where residues of disinfectants are expected, appropriate neutralizers should be added to the dilution fluid and to the media used on the contact plates to prevent any inhibitory effect of the disinfectants on the growth of microorganisms. Disinfectants are generally formulated for a disinfection contact time of 5 to 15 minutes. Wait for a period in accordance with the disinfectant specification before investigating the surface with swabs or contact plates to assess the performance of the cleaning and disinfection program.

An appropriate neutralizer for all situations cannot be prescribed. A number of disinfectant neutralizers are recommended in EN 1276, EN 1650, EN 13697 and EN 13704. Also specific neutralizers have been developed allowing better recovery and also less interference with further steps for some specific methods (e.g. HiCap™ Neutralizing Broth for 3M petrifilm or Real Time PCR).

The components of a neutralizer which may be used in most situations can be prepared with a solution of peptone (1 g/l), sodium chloride (8,5 g/l), and the components included in the Table 2.

Table 2. Neutralizer which can be used in most situations (from

ISO 18593:2004).

Transport conditions can have a big influence in the final results and needs to be considered as a critical factor. Temperature, time and transport diluent combination can have a big impact on microorganisms stability during transport (see Figure 5).

Figure 5. Listeria growth in different buffers and temperatures

(From Bazako et al., 2007).

According 18593:2004, samples should be transported preferably within 4 h at 1 °C to 4 °C, in a way that no contamination can occur. The analysis must be started as soon as possible and not later than 24 hours.

Despite of that, commercial information provided by suppliers shows a broad range of transport times for combinations swabs/sponge materials of neutralizers. For most of the combinations, transport times up to 48-72 hours at refrigeration temperature are not affecting the viability of microorganisms for further analysis. For ambient temperature transport, in most of the cases there is not information available about the impact on microorganisms but, some suppliers recommend not to exceed 48-72 hours after sampling.

5. Laboratory analysis

Contact agar plates must be incubated according to the type of microorganisms to be enumerated (remember this sampling is not applicable to qualitative methods).


food control

For swabs in tubes with neutralizing/transport diluent, thoroughly mix the contents of tubes containing swabs using a mixer for 30 s, adjusting the speed so that the wall of the tube is wetted up to a height of 2 cm to 3 cm below the top. This step is critical to avoid the collected microorganisms don’t remain trapped in the tip of the swab instead of being easily released (Ismaïl et al., 2013). Enough diluent/neutralizing liquid must be used to guarantee the entire sample is immersed (e.g. for big samples like shocks, cloths or sponges).

Swabs + diluents/neutralizing liquids should be considered the initial suspension. For qualitative methods, the initial suspension must be incubated for pre-enrichment according the instructions described in the specific method.

For enumeration, an aliquot from the initial suspension must be plated. If high numbers of microorganisms are expected, prepare further decimal dilutions and proceed with the enumeration method.

Swabs can be also analysed by ATP measurement, allowing a rapid result (in minutes). However, these techniques are not targeting only microbiological contamination but all organic material present in the surface, giving a total result in Relative Luminiscence Units (RLU).

6. Calculation of results

6.1. Contact agar methods, results can be estimated easily by counting the total number of colonies in the surface of the agar. For an accurate enumeration, the effective surface of contact agar should be considered. For example, if the agar surface is Ø 55 mm, this corresponds to a surface of 23.75 cm2.

Also it is possible to count (and report) only the colonies present in the central grid, knowing each square is equal to 1 cm2 (total surface = 16 cm2).

6.2. “Swabs” methods, the number of colonies per surface (CFU/cm2) of surface NS using the formula:

Where

•N is the number of CFU in a plate (considering 1 mL ofdiluent has been poured)

•F is the amount, in mililitres, of diluent in the tube orhomogenizer bag

•Aisthesurfaceinvestigated,insquarecentimetres•disthedilutiontestedIf the area swabbed was not defined, calculate the CFU/sample N

SW using the formula:

Example:

A surface of 100 cm2 has been swabbed. The swab was diluted in 5 ml of BPW (initial suspension or 10-1). 1 mL from the initial suspension was plated in a Petri dish. After incubation, 12 colonies were detected in the plate.

The final count in CFU/cm2 is:

In case of qualitative methods when an enrichment step is included, the target microorganism must be reported as detected or not detected in the area swabbed, or per sample if the area is not known.

Results for environmental sampling are often presented as hygiene scores based on the presence of a pathogen or a number of colony-forming units (CFU) per sample or square centimeter present on a test surface. Usually these methods are not enough not quantitatively reliable or reproducible and results should only be used in a “trend analysis”. Control limits should be established for each specific use.

7. New technologies

Some new technologies have been developed to improve the environmental sampling in food industries. These methods are mainly focusing on Listeria detection in surfaces.

The prototype developed in the project BioliSME, supported by the European Commission within the Seventh Framework Programme, (FP7-SME-2011-286713) was able to improve the recovery from Listeria monocytogenes biofilm on stainless steel and polytetrafluoroethylene surfaces from 11% by conventional techniques (swabbing) to a 98 % by using a device based on air/water ablation (Gião et al., 2015). This device combined with a detection system based on biosensor


food control

technology allowed the detection of 103 cfu of Listeria monocytogenes in 10 minutes.

Other commercial method available is the Sample6 Detect/LTM Test developed in USA (www.sample6.com). The method allows the detection of Listeria spp. in 6 hours in stainless steel with a Probability Of Detection (POD) equivalent to the reference method (POD < 1). The detection technology applies next-generation synthetic-biology techniques (Lu et al., 2013) to enable bacterial pathogen detection from a swab or sponge. The system uses the inherent specificity of naturally occurring bacteriophage, that leads to the production of the light-producing enzyme luciferase. Since production of luciferase is an active biological process, Detect/LTM can detect only living cells.

In both cases further validation studies against the reference method will be needed to allow a full implementation and recognition by food companies and regulatory authorities.

Bibliography

•BazacoM.,EiferTJ.D.,WilliamsR.C.,Katharioubazaco S. 2007. Quantitative Recovery of Listeria monocytogenes and Select Salmonella Serotypes from Environmental Sample Media. Journal of AOAC international VOL. 90, No. 1, 2007.

•Faille C., Bénézech T., Lamour JB.,Hannequin M., Chéné C. 2014.Evaluation of sampling devices for monitoring surface hygiene in food environments. Poster Cambridge University.

•Gião M.S., Blanc S., Porta S.,Belenguer J. and Keevil C.W. 2015. Improved recovery of Listeria monocytogenes from stainless steel and polytetrafluoroethylene surfaces using air/water ablation. Journal of Applied Microbiology 119, 253—262.

•ICMSF 2002. Microorganisms in Foods 7: Microbiological Testing in

Food Safety Management. Kluwer Academic/Plenum Publishers.

•IsmaïlR.,AviatF.,MichelV.,LeBayonI.,Gay-PerretP.,KutnikM.,FédérighiM., 2013. Methods for Recovering Microorganisms from Solid Surfaces Used in the Food Industry: A Review of the Literature. Int. J. Environ. Res. Public Health, 10, 6169-6183.

•ISO 18593:2004. Microbiology offood and animal feeding stuffs — Horizontal methods for sampling techniques from surfaces using contact plates and swabs.

•LuT.K.,BowersJ.,KoerisM.S.,2013Advancing bacteriophage-based microbial diagnostics with synthetic biology Trends in Biotechnology Volume 31, Issue 6, June 2013, pages 325-327.

•Ward R., 2012. White Paper onCellulose and Polyurethane Sponges for Surface Sampling. WorldBioproducts.

•Ward, R., and Tchelak R., 2012.Collection Broths for Environmental Monitoring Programs. WorldBioproducts.

www.finescience.com

Harmonization and standardization -...

Documents

Transcript of Harmonization and standardization -...