Fast, Affordable and multiplexed foodborne pathogen detection on miniaturized devices

TESE DE DOUTORAMENTO FAST, AFFORDABLE AND MULTIPLEXED FOODBORNE PATHOGEN DETECTION ON MINIATURIZED DEVICES Sarah Gaspar Ferreira Azinheiro ESCOLA DE DOUTORAMENTO INTERNACIONAL DA UNIVERSIDADE DE SANTIAGO DE COMPOSTELA PROGRAMA DE DOUTORAMENTO EN INNOVACIÓN EN SEGURIDADE E TECNOLOXÍA ALIMENTARIAS SANTIAGO DE COMPOSTELA 2022 DECLARACIÓN DO AUTOR/A DA TESE D./Dna. Sarah Gaspar Ferreira Azinheiro Título da tese: Fast, Affordable and multiplexed foodborne pathogen detection on miniaturized devices Presento a miña tese, seguindo o procedemento axeitado ao Regulamento, e declaro que: 1) A tese abarca os resultados da elaboración do meu traballo. 2) De ser o caso, na tese faise referencia ás colaboracións que tivo este traballo. 3) Confirmo que a tese non incorre en ningún tipo de plaxio doutros autores nin de traballos presentados por min para a obtención doutros títulos. 4) A tese é a versión definitiva presentada para a súa defensa e coincide a versión impresa coa presentada en formato electrónico E comprométome a presentar o Compromiso Documental de Supervisión no caso de que o orixinal non estea na Escola. En Braga, 18 de Maio de 2022. Sinatura electrónica AUTORIZACIÓN DEL DIRECTOR / TUTOR DE LA TESIS [Fast, Affordable and multiplexed foodborne pathogen detection on miniaturized devices] D. Jorge Barros Velázquez INFORMAN: Que la presente tesis, se corresponde con el trabajo realizado por Dª. Sarah Gaspar Ferreira Azinheiro, bajo mi dirección/tutorización, y a utorizo su presentación , considerando que reúne l os r equisitos exigidos en el R eglamento de Estudios de Doctorado de la USC, y que como director de esta no incurre en las causas de abstención establecidas en la Ley 40/2015. De acuerdo con lo indicado en el Reglamento de Estudios de Doctorado, declara también que la presente tesis doctoral es idónea para ser defendida en base a la modalidad de Monográfica con reproducción de publicaciones, en los que la participación del doctoranda fue decisiva para su elaboración y las publicaciones se ajustan al Plan de Investigación. En Santiago de Compostela, 18 de Maio de 2022 AUTORIZACIÓN DEL DIRECTOR / TUTOR DE LA TESIS [Fast, Affordable and multiplexed foodborne pathogen detection on miniaturized devices] Dª. Marta Prado Rodríguez INFORMAN: Que la presente tesis, se corresponde con el trabajo realizado por Dª. Sarah Gaspar Ferreira Azinheiro, bajo mi dirección/tutorización, y a utorizo su presentación , considerando que reúne l os r equisitos exigidos en el R eglamento de Estudios de Doctorado de la USC, y que como director de esta no incurre en las causas de abstención establecidas en la Ley 40/2015. De acuerdo con lo indicado en el Reglamento de Estudios de Doctorado, declara también que la presente tesis doctoral es idónea para ser defendida en base a la modalidad de Monográfica con reproducción de publicaciones, en los que la participación del doctoranda fue decisiva para su elaboración y las publicaciones se ajustan al Plan de Investigación. En Braga, 18 de Maio de 2022 iv v TABLE OF CONTENTS RESUMO ............................................................................................................................................ ix ABSTRACT ....................................................................................................................................... xv LIST OF PUBLICATIONS .............................................................................................................. xxi LIST OF ABBREVIATIONS ......................................................................................................... xxiii Chapter 1. INTRODUCTION ......................................................................................................... 1 1.1 The importance of food safety in food supply chain ......................................................... 3 1.2 Problematic pathogens ....................................................................................................... 4 1.3 Characteristic of the most problematic pathogens ............................................................. 5 1.3.1 Salmonella spp. ............................................................................................................. 5 1.3.2 Shiga Toxin-producing E. coli (STEC) ........................................................................ 6 1.3.3 L. monocytogenes ......................................................................................................... 8 1.4 Regulation in the food industry ....................................................................................... 10 1.5 Detection methodology .................................................................................................... 10 1.5.1 Gold standards culture based analysis ........................................................................ 10 1.5.2 Commercial methods available for improved analysis ............................................... 13 1.5.3 Alternative sample pre-treatment strategies ............................................................... 15 1.5.4 Polymerase Chain Reaction (PCR) and real-time PCR (qPCR) ................................. 17 1.5.5 Alternative isothermal amplification .......................................................................... 18 1.6 Devices for DNA amplification ....................................................................................... 21 1.6.1 Equipment in the market ............................................................................................. 21 1.6.2 Devices under study .................................................................................................... 22 Chapter 2. OBJECTIVES .............................................................................................................. 25 2.1 Main Objectives ............................................................................................................... 27 2.1.1 Specific Objectives ..................................................................................................... 27 2.2 Thesis structure ................................................................................................................ 28 Chapter 3. METHODOLOGY ...................................................................................................... 31 3.1 Introduction ...................................................................................................................... 33 3.2 Reference bacteria strains used ........................................................................................ 33 3.3 DNA extraction ................................................................................................................ 34 3.3.1 Pure culture ................................................................................................................. 34 3.3.2 Complex food matrixes ............................................................................................... 35 3.4 Protocols for improved sample pre-treatment ................................................................. 36 vi 3.4.1 Comparison of different media for standard 24 h enrichment .................................... 36 3.4.2 Concentration of bacteria ............................................................................................ 38 3.4.3 Protocols for time reduction ....................................................................................... 43 3.5 Evaluation of the DNA amplification alternatives .......................................................... 45 3.5.1 Primers design ............................................................................................................ 45 3.5.2 Real-time amplification .............................................................................................. 50 3.5.3 RPA combined with naked-eye detection ................................................................... 53 3.5.4 LAMP reaction coupled with naked-eye detection..................................................... 55 3.6 Optimized selected methodology ..................................................................................... 57 3.7 Miniaturized Devices ....................................................................................................... 58 3.7.1 Microfluidic device for capture and concentration of bacteria ................................... 58 3.7.2 PDMS channels prototype .......................................................................................... 59 3.7.3 Milled channels prototype .......................................................................................... 60 3.8 DNA amplification evaluation ......................................................................................... 62 3.9 Evaluation of the developed methodologies .................................................................... 63 3.9.1 Evaluation with complex food matrixes ..................................................................... 64 3.9.2 LoD determination ...................................................................................................... 66 3.9.3 Fitness for purpose ...................................................................................................... 66 3.10 Mesophilic bacteria analysis in food samples .................................................................. 68 3.10.1 Enumeration of microorganisms ................................................................................. 68 3.10.2 Long-read next generation DNA sequencing ............................................................. 68 Chapter 4. RESULTS – SAMPLE PRE-TREATMENT ALTERNATIVES ............................... 71 4.1 Introduction ...................................................................................................................... 73 4.2 Improvement of the standard enrichment step ................................................................. 73 4.2.1 L. monocytogenes growth in simplex enrichment ...................................................... 74 4.2.2 Growth of the three targets in co-culture .................................................................... 77 4.3 Concentration of bacteria ................................................................................................. 80 4.3.1 Immunomagnetic separation (IMS) ............................................................................ 80 4.3.2 PDMS sponge ............................................................................................................. 84 4.4 Time reduction ................................................................................................................. 88 4.4.1 PAA ............................................................................................................................ 88 4.4.2 Matrix lysis ................................................................................................................. 91 4.4.3 Short enrichment ......................................................................................................... 92 4.5 Comparison of sample pre-treatment approaches ............................................................ 94 4.6 Conclusions ...................................................................................................................... 96 4.7 Publication of the results .................................................................................................. 97 vii Chapter 5. RESULTS - DNA AMPLIFICATION APPROACHES ........................................... 101 5.1 Introduction .................................................................................................................... 103 5.2 Real-time amplification ................................................................................................. 103 5.2.1 SYBR-qPCR for multiplex detection ....................................................................... 104 5.2.2 Probe-qPCR .............................................................................................................. 110 5.2.3 qLAMP, qRPA and comparison between real-time techniques ............................... 115 5.3 RPA combined with naked-eye detection ...................................................................... 119 5.3.1 RPALateral flow (RPA-LF) ................................................................................... 119 5.3.2 RPA with SYBR Green (RPA-SYBR) ..................................................................... 125 5.4 LAMP combined with naked-eye detection .................................................................. 128 5.4.1 Turbidity ................................................................................................................... 128 5.4.2 Naked-eye detection approach by combination of MUA and AuNPs ...................... 132 5.4.3 Colorimetric mastermix ............................................................................................ 134 5.5 Comparison of the DNA amplification and detection approaches ................................ 137 5.6 Conclusions .................................................................................................................... 138 5.7 Publication of the results ................................................................................................ 139 Chapter 6. RESULTS – SELECTED METHODOLOGY AND INTEGRATION ON MINIATURIZED DEVICE ............................................................................................................. 143 6.1 Introduction .................................................................................................................... 145 6.2 Selected methodology .................................................................................................... 145 6.2.1 Optimization ............................................................................................................. 145 6.2.2 Evaluation of the colorimetric LAMP reaction ........................................................ 153 6.2.3 Evaluation of the full methodology .......................................................................... 157 6.2.4 Analysis of mesophilic bacteria ................................................................................ 161 6.3 Integration on the amplification protocols on miniaturized devices .............................. 163 6.3.1 Milled channels prototype ........................................................................................ 163 6.3.2 PDMS channels prototype ........................................................................................ 166 6.3.3 Comparison between thermocycler and miniaturized devices ................................. 168 6.3.4 Evaluation of the method integration on the milled channels prototype .................. 169 6.4 Conclusions .................................................................................................................... 170 Chapter 7. FINAL CONCLUSIONS AND FUTURE WORK ................................................... 175 7.1 Conclusions .................................................................................................................... 177 7.1.1 Major conclusion ...................................................................................................... 177 7.1.2 Specific conclusions ................................................................................................. 177 7.2 Future Work ................................................................................................................... 179 BIBLIOGRAPHY ............................................................................................................................ 181 viii LIST OF FIGURES ......................................................................................................................... 197 LIST OF TABLES ........................................................................................................................... 203 LIST OF EQUATIONS ................................................................................................................... 205 APPENDIX ...................................................................................................................................... 207 ix RESUMO As intoxicacións alimentarias son un problema de saúde pública mundial que afecta non só aos países en desenvolvemento, senón tamén aos países desenvolvidos [1]. A Organización Mundial da Saúde (OMS) destacou que 1 de cada 10 persoas en todo o mundo enfermará debido ao consumo de alimentos contaminados [2]. Diferentes axentes patóxenos son os responsables destas enfermidades de transmisión alimentaria que causan hospitalización e morte, entre eles Salmonella spp., E. coli produtora de toxina Shiga (STEC) e L. monocytogenes son moi problemáticos para a industria alimentaria. Os dous primeiros, presentan o maior número de casos de hospitalización informados polas autoridades europeas, mentres que L. monocytogenes segue mostrando unha maior gravidade e taxa de mortalidade de todos os patóxenos de transmisión alimentaria monitorizados [3]. A pesar do esforzo por mellorar os estándares internacionais de seguridade alimentaria, seguen xurdindo novos riscos na cadea de subministración de alimentos [4] e as dificultades para rastrexar as fontes dos brotes aumentan o risco de que haxa máis casos de infección. Existe unha necesidade urxente de métodos máis sensibles e rápidos para detectar microorganismos patóxenos nos produtos alimenticios, para evitar posibles enfermidades e mortes. Os métodos tradicionais para detectar patóxenos transmitidos por alimentos baséanse en cultivos, que son laboriosos, levan moito tempo e requiren persoal de laboratorio capacitado [5]. Estas metodoloxías amplían a análise ata cinco días, sendo non sostibles para produtos de curta vida útil e non encaixando na intensa produción actual. Ademais, a maior demanda de metodoloxías que permitan detectar máis dun patóxeno ao mesmo tempo non se pode alcanzar coas técnicas tradicionais. Nos últimos anos xurdiron novos métodos baseados no ADN ou na análise de proteínas co obxectivo de superar algúns destes inconvenientes [6]. Así, dispositivos como os micro sistemas de análise total (µTAS) representan unha verdadeira vantaxe, aumentando a velocidade de análise e a sensibilidade, diminuíndo ademais o consumo de reactivos, o risco de contaminación cruzada das mostras e abre a posibilidade dunha automatización total [7,8]. Probouse a integración de técnicas de amplificación de ADN en dispositivos miniaturizados, non só para a detección de patóxenos, senón tamén para a identificación de trastornos xenéticos e enfermidades infecciosas [9]. A amplificación enzimática en dispositivos microfluídicos realízase principalmente mediante PCR, o que require un control de temperatura e un aumento rápido da temperatura, aumentando a complexidade da implementación e aumentando o custo do instrumento. Nos últimos anos xurdiron novas técnicas alternativas de amplificación de ADN co obxectivo de ofrecer solucións analíticas a algúns dos inconvenientes do método de referencia para a amplificación in vitro, entre eles as técnicas de amplificación isotérmica de ADN son especialmente interesantes para fins de miniaturización. Debido á sinxeleza do control de temperatura para a amplificación de secuencias, a amplificación isotérmica pódese implementar facilmente en microchips simples sen complicados controles térmicos e/ou fluídos [10]. x O obxectivo desta tese foi desenvolver unha metodoloxía mellorada para a detección múltiple de varios patóxenos transmitidos por alimentos baseada na detección de ADN, e a súa integración nun dispositivo miniaturizado. Para acadar este obxectivo abordáronse os diferentes pasos da análise, incluíndo o pretratamento da mostra, a amplificación do ADN e a visualización dos resultados, nos que se avaliaron varios enfoques para a escolla da mellor opción para reducir o tempo de análise, reducir o custo e permitir a detección a simple vista. Para acadar a sensibilidade requirida da análise para os patóxenos transmitidos por alimentos, é esencial o enriquecemento da mostra. Neste momento, as metodoloxías máis rápidas aínda precisan de 18-48 h de tempo de enriquecemento, sen melloras significativas, sendo este punto o principal pescozo de botella na análise microbiolóxica de alimentos á hora de reducir o tempo de análise. É fundamental mellorar o xeito no que se trata a mostra, e por iso neste proxecto realizouse a optimización do pretratamento da mostra para permitir unha análise máis rápida. Avaliáronse diferentes enfoques para unha recuperación simultánea eficiente de bacterias patóxenas, incluíndo a optimización do medio no enriquecemento estándar, a concentración das bacterias e as estratexias de redución de tempo. Probouse a influencia de diferentes medios selectivos ou non selectivos, no crecemento das bacterias diana, así como a suplementación con varios compostos. Valorouse o mellor medio para a detección simultánea de Salmonella spp., E. coli O157 e L. monocytogenes. Debido a que L. monocytogenes presenta unha menor taxa de crecemento, en comparación coas outras dúas bacterias, a optimización do medio centrouse na mellora da concentración deste patóxeno. O medio xeral, Tryptic Soy Broth (TSB), mostrou ser o medio que permitiu a maior redución da fase de latencia, o que reduciu o tempo de análise coa maior concentración bacteriana obtida. Ademais, probáronse varios suplementos e avaliouse o seu rendemento para mellorar o paso de enriquecemento, incluíndo celobiosa, extracto de levadura, piruvato de sodio, sangue de cabalo hemolisado (LHB), suplemento de crecemento de Campylobacter e suplemento selectivo do Fraser a media concentración. Entre estas, a celobiosa foi a única que permitiu un aumento da concentración final de L. monocytogenes despois de 24 h, sen mostrar cambios na concentración das outras dúas dianas. Non obstante, non se viu ningunha mellora na fase de latencia con este composto. Para aumentar a sensibilidade da metodoloxía, avaliáronse diferentes enfoques para concentrar as bacterias no pretratamento da mostra. A primeira metodoloxía consistiu nunha Separación Inmunomagnética (IMS), onde se avaliaron catro anticorpos comerciais diferentes, específicos de L. monocytogenes, comparando a súa pureza e especificidade. O enfoque seleccionado consistiu na funcionalización de nanosferas magnéticas e na súa utilización para analizar mostras de alimentos. O enfoque IMS permitiu un LoD de 9,7 ufc/ 25 g, cando se combina cun enriquecemento selectivo en Half Fraser (HF) durante 24 h e análise de qPCR. A segunda alternativa implicou o uso dun dispositivo microfluídico onde se funcionalizou unha esponxa de polidimetilsiloxano (PDMS) 3D con ligandos específicos para reter as bacterias. Utilizouse un ligando bacteriano inespecífico, a proteína ApoH, para a captura múltiple e un anticorpo antiL. monocytogenes específico (seleccionado anteriormente no enfoque IMS), para unha captura dirixida. Ambos enfoques avaliáronse con cultivos bacterianos puros, e tamén para analizar superficies de aceiro inoxidable contaminado. O ligando inespecífico, a proteína ApoH, permitiu acadar unha eficiencia de captura lixeiramente maior, e a captura múltiple de bacterias dianas, sendo unha vantaxe en comparación cos ligandos específicos, que requirían un anticorpo para cada patóxeno. Non obstante, os ligandos universais non se poden usar en matrices complexas cun alto número de microorganismos de fondo, xa que uniranse aleatoriamente a outros microorganismos. Cando se analizaron as mostras de superficie combinando a esponxa PDMS coa análise de qPCR sen enriquecemento previo, o LoD só está limitado pola propia técnica de amplificación do ADN ou polo proceso de mostraxe. Ambas opcións, IMS e a esponxa funcionalizada mostraron un alto rendemento e permiten unha detección xi fiable para diferentes aplicacións. Unha vantaxe da captura das células bacterianas, é a súa separación do resto da solución de enriquecemento, o que permite eliminar posibles compostos inhibidores da reacción de amplificación do ADN. Non obstante, o uso destes ligandos específicos aumenta o custo da análise, xa que os diferentes ligandos, anticorpos ou outros seguen sendo caros. Co obxectivo de reducir o tempo de análise, probáronse tres alternativas: Ensaio de amplificación de fagos (PAA), lise da matriz e enriquecemento curto. O PAA consiste nunha estratexia indirecta para detectar un microorganismo mediante a detección dun bacteriófago específico que infectará a bacteria diana. Coa replicación máis rápida do fago, obtivemos unha maior concentración do microorganism diana en menos tempo en comparación coa detección directa. Este enfoque aplicouse para a detección de S. Enteritidis en mostras de alimentos, permitindo realizar a análise por qPCR en só 10 h coa mesma sensibilidade que as metodoloxías convencionais, acadando unha LoD de 8 ufc/ 25 g. O enfoque de lise da matriz depende dun xeito completamente diferente de tratar a mostra. En lugar de tomar só unha pequena alícuota da mostra enriquecida, como nas análises tradicionais, a mostra enteira é tratada para degradar o máximo posible a matriz e recuperar as bacterias para posterior extracción de ADN. Con certa similitude, o enriquecemento curto consiste nun reducido tempo de incubación en medio de cultivo, e recuperación de todo o líquido da mostra a procesar, para permitir a separación da gran maioría dos microorganismos da mostra do alimento. Ambas metodoloxías foron avaliadas para a recuperación e detección de L. monoctogenes en simplex mostrando unha redución do tempo de análise a 5 h e 7 h, con lise da matriz e enriquecemento curto, respectivamente. A lise da matriz presenta un LoD excesivamente alto (1,1 x 105 ufc/ 25 g) polo que non é apto para a industria alimentaria, sendo descartado como posible opción para a metodoloxía final. O enriquecemento curto mostrou unha sensibilidade moito maior, cunha LoD de 8,6 ufc/ 25 g, debido ao paso de cultivo realizado antes do tratamento da mostra, polo que o uso desta metodoloxía para a detección de E. coli O157 en formato simple, e a simultánea detección de E. coli O157 e Salmonella spp. logrouse con resultados similares (LoD = 3-4 ufc/ 25 g), demostrando a posibilidade de ser utilizado na detección múltiple. Despois de avaliar as diferentes alternativas para o tratamento da mostra, a metodoloxía mostrou os resultados máis prometedores para permitir unha detección múltiple máis rápida e económica de L. monocytogenes, E. coli O157 e Salmonella spp. probouse como o enriquecemento curto. Como parte da estratexia avaliáronse diferentes alternativas de amplificación e detección de ADN. As técnicas de amplificación isotérmica demostraron varias vantaxes sobre a PCR/qPCR, como ser realizadas a temperatura constante sen necesidade de equipos complexos, como termocicladores. Esta característica foi interesante para o desenvolvemento do proxecto xa que simplifica a análise, reduce o custo e permite unha integración máis sinxela deste paso nun sistema miniaturizado, permitindo ademais un menor consumo de enerxía. Ademais, pódense empregar facilmente diferentes alternativas para a detección a simple vista cando se combinan cunha técnica de amplificación isotérmica. Por este motivo, avaliáronse dúas técnicas de amplificación isotérmica, LAMP e RPA, para detectar os patóxenos diana en formato simple ou múltiple, é obter un resultado visual sinxelo. O primeiro paso foi comprender o rendemento da análise tradicional de qPCR, utilizando un colorante intercalante (SYBR-qPCR) e unha sonda de hidrólise (Probe-qPCR), e comparalos con LAMP e RPA. A súa especificidade, sensibilidade e precisión con diferentes dianas xenéticas avaliáronse con cultivos puros e mostras de alimentos contaminados. O SYBR-qPCR foi un enfoque máis económico xa que non precisa de sondas específicas, non obstante o deseño do ensaio é máis complexo, así como a análise dos resultados cando se require a detección múltiple. Por outra banda, a Probe-qPCR ofreceu unha maior especificidade debido á implementación da sonda. Ambas metodoloxías permitiron realizar detección múltiple con sensibilidades similares. En canto á avaliación da amplificación isotérmica, tanto LAMP como RPA, deron como resultado unha sensibilidade inferior á dos enfoques qPCR, cando se examinou a sensibilidade analítica mediante cultivo puro. Non obstante, cando se xii analizaron mostras contaminados combinadas cun enriquecemento en dous pasos, en caldo mTA10 e Full Fraser (FF), conseguiuse un LoD similar (1,4 ufc/ 25 g). Tamén se observou un rendemento lixeiramente superior de RPA, que proporciona resultados máis rápidos en comparación con LAMP. Probáronse tamén diferentes alternativas para a visualización dos resultados dun xeito máis sinxelo para cada técnica de amplificación isotérmica. Para RPA, avaliouse a carga da reacción de amplificación nunha banda de fluxo lateral (LF) e logrouse unha alta sensibilidade e especificidade con este método para a detección de L. monocytogenes en mostras de superficie, mostrando un LoD de 18,2 ufc/ cm2 despois dun enriquecemento de 24 h en o caldo, ONE broth. A desvantaxe do RPALF foi o aumento do custo da análise debido á necesidade do uso de cebadores modificados e unha sonda, xunto coas tiras LF. Como alternativa, pódense engadir diferentes compostos á reacción, antes ou despois da amplificación, para conseguir un resultado detectable a simple vista en solución. Neste sentido, avaliouse a adición de SYBR Green, un colorante intercalante de dsDNA, despois da reacción RPA, para permitir a visualización da fluorescencia en mostras positivas cando se expoñan a luz UV. A detección de E. coli O157 en mostras de alimentos foi posible e conseguimos unha LoD95 de 19 ufc/ 25 g cando se combina cun pretratamento baseado nun enriquecemento curto. Ademais deste enfoque que presenta un resultado prometedor, o feito de que SYBR Green se una a calquera dsADN presente na mostra, fai que a análise de mostras complexas e a detección múltiplex sexan limitadas, debido á presenza do ruido de fluorescencia de fondo obtido en mostras negativas, o que dificulta para identificar resultados positivos. Para a visualización dos resultados da amplificación LAMP probáronse tres enfoques, incluíndo a turbidez e o cambio de cor mediante o uso de nanopartículas de ouro (AuNPs) ou cunha mastermix colorimétrica comercial. A turbidez na reacción LAMP conséguese pola produción do subproduto insoluble, pirofosfato de magnesio, cando se produce a amplificación. Este enfoque utilizouse para detectar diferentes serovares de Salmonella spp., dirixíndose a catro dianas xenéticas, o que provocou diferenzas no rendemento da análise, cando se probaron mostras de alimentos despois un enriquecemento de 24 horas en mTA10. A turbidez pódese controlar en tempo real mediante un turbidímetro ou como detección de punto final mediante a visualización a simple vista dos resultados. Non obstante, a observación a simple vista debe realizarse nun ángulo de luz específico, o que fai que a análise sexa máis subxectiva e complexa. Os AuNPs foron probados para diferentes aplicacións e grazas ás posibilidades de modificar as súas propiedades ópticas en función do estado de agregación, pódese conseguir a detección a simple vista. A funcionalización do AuNP con ácido 11-mercaptoundecanoico (MUA) permite controlar a agregación das nanopartículas cando se orixina o produto de amplificación, provocando diferenzas de cores. Neste sentido, as mostras positivas serán vermellas, mentres que as negativas mostrarán unha cor morada. Este enfoque probouse para a detección de Salmonella spp. despois dun enriquecemento de 24 h en BPW. O paso de amplificación realizouse nun dispositivo microfluídico e o produto da amplificación mesturouse co MUA-AUNP, producindo un claro cambio de cor. Outra alternativa para orixinar un cambio de cor na reacción de amplificación é a adición dun colorante sensible ao pH, para detectar a amplificación LAMP. Este enfoque probouse para a detección de Salmonella spp. para comprender as posibilidades na metodoloxía final, neste caso o uso dos cebadores Loop, e o tempo de amplificación houbo que optimizar para evitar falsos positivos e obter resultados fiables. Probáronse mostras enriquecidas durante 24 h, presentando un LoD95 de 2,1 ufc/ 25 g, ademais avaliouse a combinación cun enriquecemento curto de 6 h mostrando unha total concordancia coa metodoloxía qPCR. Despois dunha análise en profundidade das diferentes opcións avaliadas para realizar a amplificación do ADN e para permitir un enfoque simplificado para a visualización a simple vista dos resultados, a metodoloxía que permite a detección máis sinxela, rendible e sensible parece ser a LAMP colorimétrica. A desvantaxe do uso da metodoloxía MUA-AuNP relacionouse coa xiii manipulación do produto de amplificación despois da reacción, que pode provocar contaminacións cruzadas e falsos positivos. Co mastermix colorimétrico de NEB superouse este inconveniente, proporcionando unha solución vantaxosa. Por este motivo, a metodoloxía final deseñouse para combinar a metodoloxía de enriquecemento curto para o crecemento dos tres patóxenos, coa LAMP colorimétrica empregando a mastermix comercial. Para obter unha detección múltiple de Salmonella spp. E. coli O157 e L. monocytogenes, o enriquecemento curto foi optimizado para permitir o crecemento de todos eles nun único paso de enriquecemento permitindo resultados moi sensibles. Despois de probar medios xerais e selectivos para este fin, o enriquecemento en TSB durante 7 h foi o que deu mellores resultados para a detección de L. monocytogenes, xa que este foi o patóxeno máis problemático pola súa menor taxa de crecemento. Avaliouse a metodoloxía completa analizando diferentes tipos de mostras de leite (UHT, Fresco e cru), acadando un LoD aceptable, cunha detección máis sensible de Salmonella spp. e E. coli O157 cunha LoD95 de 1,6 ufc/ 25 mL. Para L. monocytogenes foi necesaria unha maior concentración de bacterias para permitir unha detección fiable, obtendo unha LoD95 de 79 ufc/ 25 mL. Observouse unha clara influencia da microflora natural para a detección de L. monocytogenes, cun aumento asociado da LoD con maiores recontos mesófilos. A caracterización mesófila tamén se realizou mediante a análise de secuenciación MinION onde se identificaron principalmente bacterias acidolácticas. Despois da avaliación completa do método, probouse a integración do paso de amplificación en dispositivos miniaturizados e comparouse cos resultados anteriores. Comparouse primeiro o rendemento de dous sistemas distintos analizando cultivos puros, para determinar a mellor opción a implementar na metodoloxía final. Un dos dispositivos consistía nun dispositivo integrado que combinaba o control de temperatura e un soporte flexible que permitía o uso de tubos de silicona con diferentes volumes. A outra alternativa consistía nun dispositivo microfluídico con 8 microcanles que se podía colocar nunha incubadora de laboratorio convencional para realizar a reacción. Conseguiuse unha clara diferenciación entre resultados positivos e negativos, cando se integrou a LAMP colorimétrica en ambos os dispositivos, non obstante observouse unha diminución da sensibilidade analítica. O dispositivo de microcanles mostrou os peores resultados, cunha escasa sensibilidade analítica, sendo mesmo imposible detectar L. monocytogenes nas concentracións máis altas ensaiadas. Por este motivo, este dispositivo foi descartado para a seguinte análise. Seleccionouse, e avaliouse, o sistema de quentamento integrado para a análise de mostras de leite. A metodoloxía final obtivo un LoD maior en comparación co método termociclador, como se esperaba polos resultados da sensibilidade analítica. Obtívose unha LoD95 de 15, 17 e 141 cfu/ 25 mL para Salmonella spp. E. coli O157 e L. monocytogenes, respectivamente. Coas mostras analizadas conseguiuse unha total concordancia coa metodoloxía qPCR. En conclusión, a metodoloxía desenvolvida neste proxecto permitiu a detección múltiple de Salmonella spp. E. coli O157 e L. monocytogenes, reducindo o tempo de análise a só 9 h fronte a 7 días cando se realiza mediante técnicas tradicionais baseadas en cultivos e proporcionando detección a simple vista. Obtívose unha sensibilidade e especificidade similares cando se comparou co método de referencia, qPCR, sen necesidade dunha instrumentación complexa como un termociclador en tempo real. O uso do enriquecemento curto como pretratamento da mostra permitiu conseguir unha importante redución do tempo de resposta, sen aumentar o custo, o que fai a análise máis accesible para todos. Ademais, a posibilidade de integrar o paso de amplificación nun dispositivo miniaturizado abre a porta ao desenvolvemento de sistemas Point-of-Care (POC), de tamaño reducido e coa posibilidade de automatizar a análise a realizar en configuracións descentralizadas. xx xxi LIST OF PUBLICATIONS Publication 1 - Combination of Recombinase Polymerase Amplification with SYBR Green I for Naked-Eye, Same-Day Detection of Escherichia coli O157:H7 in Ground Meat ....................... 209 Publication 2 - Short Pre-Enrichment and Modified Matrix Lysis. A Comparative Study towards Same-Day Detection of Listeria monocytogenes .......................................................................... 210 Publication 3 - Application of Recombinase Polymerase Amplification with Lateral Flow for a Naked-Eye Detection of Listeria monocytogenes on Food Processing Surfaces .......................... 211 Publication 4 - Multiplex Detection of Salmonella spp., E. coli O157 and L. monocytogenes by qPCR Melt Curve Analysis in Spiked Infant Formula ............................................................................. 212 Publication 5 - A Smart Microfluidic Platform for Rapid Multiplexed Detection of Foodborne Pathogens .............................................................................................................................. 213 Publication 6 - Evaluation and Implementation of Commercial Antibodies for Improved Nanoparticle-Based Immunomagnetic Separation and Real-Time PCR for Faster Detection of Listeria monocytogenes .............................................................................................................................. 214 Publication 7 - Comparative Study of Multiplex Real-Time Recombinase Polymerase Amplification and ISO 11290-1 Methods for the Detection of Listeria monocytogenes in Dairy Products…… 215 Publication 8 - Application of Short Pre-Enrichment, and Double Chemistry Real-Time PCR, Combining Fluorescent Probes and an Intercalating Dye, for Same-Day Detection and Confirmation of Salmonella spp. and Escherichia coli O157 in Ground Beef and Chicken Samples……… .... 216 Publication 9 - Specific Detection of Viable Salmonella Enteritidis by Phage Amplification Combined with qPCR (PAA-qPCR) in Spiked Chicken Meat Samples ……… .......................... 217 Publication 10 - Optimized Sample Treatment, Combined with Real-Time PCR, for Same-Day Detection of E. coli O157 in Ground Beef and Leafy Greens ……… .......................................... 218 Publication 11 - Rapid and Sensitive Detection of Viable Listeria monocytogenes in Food Products by a Filtration-Based Protocol and qPCR ……… ......................................................................... 219 Publication 12 - Evaluation of Different Genetic Targets for Salmonella Enterica serovar Enteriditis and Typhimurium, Using Loop-Mediated Isothermal AMPlification for Detection in Food Samples .............................................................................................................................. 220 Publication 13 - Combination of Microfluidic Loop-Mediated Isothermal Amplification with Gold Nanoparticles for Rapid Detection of Salmonella spp. in Food Samples ……… ......................... 221 Publication 14 - Development and Evaluation of Loop-Mediated Isothermal Amplification, and Recombinase Polymerase Amplification Methodologies, for the Detection of Listeria monocytogenes in Ready-to-Eat Food Samples ……… ......................................................................................... 222 xxii xxiii LIST OF ABBREVIATIONS Abbreviation Definition µmax Maximum Specific Growth Ab Antibody AC Relative Accuracy ALOA Agar Listeria Ottavani & Agosti ApoH Apolipoprotein H AuNP Gold Nanoparticles B3 Backward Outer Primer BAM Bacteriological Analytical Manual BHI Brain Heart Infusion BIP Backward Inner Primer BLAST Basic Local Alignment Search Tool BLEB Buffered Listeria Enrichment Broth BPW Buffered Peptone Water BSA Bovine Serum Albumin CDC Center for Disease Control and Prevention CE Capture Efficiency CECT Spanish Type Culture Collection cfu Colony Forming Unit cIAC Competitive IAC CNC Computer-Numerical-Control Cq Cycle of Quantification DAEC Diffusely Adherent E. coli DAS Strand Displacement Amplification DMSO Dimethyl Sulfoxide DNA Deoxyribonucleic Acid dsDNA Double-Stranded DNA EAEC Enteroaggregative E. coli EC European Commission ECDC European Centre for Disease Prevention and Control EDTA Ethylenediaminetetraacetic Acid EFSA European Food Safety Authority EIEC Enteroinvasive E. coli ELISA Enzyme-Linked Immunosorbent Assay EPEC Enteropathogenic E. coli xxiv ETEC Enterotoxigenic E. coli EU European Union F3 Forward Outer Primer FDA Food and Drug Administration FF Full Fraser FIP Forward Inner Primer FN False Negative FP False Positive FSIS Food Safety and Inspection Service GuSCN Guanidine Thiocyanate HACCP Hazard Analysis for Critical Control Point HC Haemorrhagic Colitis HF Half Fraser HRP Horseradish Peroxidase HUS Haemolytic-Uraemic Syndrome IAC Internal Amplification Control IgG Immunoglobulin G IMS Immunomagnetic Separation ISSO International Organization for Standardization k Cohen’s Kappa Index LAMP Loop-Mediated Isothermal Amplification LB primer Loop Primer B LB Luria-Bertani LEB Listeria Enrichment Broth LF Lateral Flow LF primer Loop Primer F LHB Laked Horse Blood LoD Limit of Detection MKTTn Muller-Kauffmann Tetrathionate-Novobiocin Broth MNP Magnetic Nanospheres MOPS 3-(N-Morpholino)Propanesulfonic Acid MPC Magnetic Particle Concentrator mTSBn Modified TSB with Novobiocin MUA 11-Mercaptoundecanoic Acid NA Negative Agreement NASBA Nucleic Acid Sequence-Based Amplification NB Nutrient Broth NC-IAC Non-Competitive IAC ND Negative Deviation NMEC Neonatal Meningitis-Associated E. coli NPV Negative Predicted Value OD Optical Density xxv ODmax Maximum Optical Density ON Overnight ONE broth Oxoid Novel Enrichment PA Positive Agreement PAA Phage Amplification Assay PB Phosphate Buffer PBS Phosphate Buffered Saline PCR Polymerase Chain Reaction PD Positive Deviation PDMS Poly(Dimethylsiloxane) PMMA Poly(Methyl Methacrylate) POC Point-of-Care PoD Probability Of Detection PPV Positive Predicted Value PSR Polymerase Spiral Reaction qLAMP Real-time LAMP qPCR Real-time Polymerase Chain Reaction qRPA Real-time RPA RCA Rolling Circle Amplification RFU Relative Fluorescence Units RNA Ribonucleic Acid RPA Recombinase Polymerase Amplification RT Room Temperature RTE Ready to Eat RT-qPCR Real-Time Reverse Transcription PCR RVS Rappaport-Vassiliadis-Soya Broth SDS-PAGE Sodium Dodecyl Sulphate–Polyacrylamide Gel Electrophoresis SE Relative Sensitivity SEPEC Sepsis-Causing E. coli SP Relative Specificity SSB Single-Strand DNA Binding Protein ssDNA Single-Stranded DNA ssRNA Single-Stranded RNA STEC Shiga Toxin-Producing E. coli Stx Shiga Toxin TEC Thermoelectric Cooler TP True Positive TSA Tryptic Soy Agar TSB Tryptic Soy Broth TSYEA Tryptic Soy Yeast Extract Agar UDG Uracil-DNA Glycosylase UHT Ultra-High Temperature xxvi UPEC Uropathogenic E. coli USDA United States Department of Agriculture UV Ultra-Violet VBNC Viable But Non-Culturable Cells WDCM World Data Centre for Microorganisms XLD Xylose Lysine Desoxycholate Agar YE Yest Extract λ Lag Time 1 CHAPTER 1. INTRODUCTION 2 Chapter 1. Introduction 3 1 INTRODUCTION 1.1 THE IMPORTANCE OF FOOD SAFETY IN FOOD SUPPLY CHAIN With the growth of the human population worldwide the demand for food has risen, and with this, an intensification of the production, and large-scale food processing and distribution systems. The complexity of the food supply chain leads to higher possibilities for contamination of the food products and additionally, the time for food products to reach the consumer is relatively short, particularly for fresh products, which makes the time to detect possible contaminations very short. Additionally, the increased globalization in the recent years have also affect the food production system, making foodborne disease easily spreading all around the world. Every year, nearly one in 10 people around the world falls ill after eating contaminated food [2]. Besides foodborne illness are more frequent in developing countries, due to the lack of food safety regulation and lower access to optimal hygienic conditions of the population, developed countries also are affected and suffer with this situation. The Centres for Disease Control and Prevention (CDC) estimates 47.8 million people get sick, 127,839 are hospitalized, and 3,037 die from foodborne diseases every year in the United States [11,12]. In Europe, more than 23 million people fall ill from eating contaminated food every year, resulting in 4,654 deaths and more than 400,000 disabilityadjusted life years [13]. The major cause of all these hospitalizations and death are enteric disease, known as intestinal illnesses, caused by different types of microorganisms such bacteria, parasites and even virus. The most common symptoms are related to intestinal disorders, such as diarrhoea, but can have much serious consequences and even lead to long-term effects and death [14,15]. Even if no illness is reported, the recall, due to an identified threat in a food product, can cause important economic lost for the company and originate high food waste. These cases may harm the responsible company not only economically but also tarnish its reputation, by leading to low consumer trust For all this reasons, important resources have been put in place to control and avoid the contamination of the food value chain with this foodborne pathogens. Monitoring needs to be done from farm to fork in the whole food value chain, as the hazard can be at any point, from the production of the raw material, passing by the processing industry until the transportation and retail of food product. Safety measures and good practices have been implemented in all stages of the supply chain, with the Codex Alimentarius [16] guidelines and the application of Hazard Analysis for Critical Control Point (HACCP). Since the implementation of these standards, more specific regulation were created by different organization for the food industry, in order to avoid and mitigate contamination. Even with the regulation in place and the strong vigilance existing nowadays, the number of outbreaks related with foodborne pathogens did not shown any decrease in the last years, as reported by European Food Safety Authority (EFSA) and European Centre for Disease Prevention and Control (ECDC) Figure 1.1. The linkage between cases of illness reported in a specific outbreak is difficult to identify, and the source of contamination can remain undetected, leading to more infections. This highlights the important to identify the contaminated food product before reaching the consumer. SARAH AZINHEIRO 10 1.4 REGULATION IN THE FOOD INDUSTRY For the protection of public health, the legislation stipulates maximum levels of defined pathogens. In USA, two regulatory authorities, Food and Drug Administration (FDA) and United States Department of Agriculture (USDA) are responsible for the establishment of this legislations [55], while in Europe the European Comission (EC) state this criteria in the Regulation 2073/2005 [56] and its different amendment. This microbiological criteria vary according to the pathogen itself, and the category of the food. Different regulatory authorities can also state distinct limits, showing variations between countries. Regarding L. monocytogenes a zero tolerance in USA is considered for any RTE food [54], while in the European Union the regulation is not so strict, due to the fact that this pathogen is ubiquitous in the environment, and a concentration of <100 cfu/ g is not enough to cause illness [57]. For this reason, up to 100 cfu/ g are allowed in Europe in foods that do not allow its growth during self life. However, the absence of L. monocytogenes is sustained in RTE food which support its growth during self-life, and also in infant products and for special medical purposes [56,58]. Implying the need for the food producers to ensure the failure of the pathogen to grow in the specific food product. Similar regulations are provided by both USA and European regulations, regarding the presence of Salmonella in food products. In the Regulation 2073/2005 strict criteria have been specified for minced meat, meat preparations and meat products intended to be eaten raw or cooked, where the absence in 25 g or in 10 g, respectively, need to be complied. The processing of these types of meat product provides an opportunity for the pathogens present on the carcase surface or in the close environment to be spread into the product. Although the foodstuff will be cooked, the bacteria may not be destroyed in the centre of the product, leading to unsafe consumption [59]. Being eggs an important source of contamination, the same criteria of absence of Salmonella, need to be followed also in egg products and RTE product containing raw egg, excluding products where the manufacturing process or the composition of the product will help to minimise the Salmonella risk. Some other RTE food, as raw milk products, milk and whey powder, cooked shellfish, sprouted seeds, fruits and vegetables are also included in this criteria due to the possibility of cross-contamination. Even though STEC were identified as foodborne pathogens back in 1982, it was only recently when specific legislation was put into place for certain types of foods. After the occurrence of several outbreaks linked to vegetables, an amendment to the EU No. 2073/2005 (No. 1441/2007, [60]) was put into place and specified the absence in 25 g of sprouts. However, in Europe, meat from ruminants is out of the regulation due to the lack of data available with STEC-contaminated food [61]. On the contrary, in the regulation from the USA, this type of product was already included [62], and also encourages importers companies to comply with the same criteria. The failure to meet these criteria, implies the recall of the product from the market in order to avoid, or at least reduce, the risk for the consumer’s health. For this reason detection methodologies for the target pathogens need to be reliable, accurate and as fast as possible. 1.5 DETECTION METHODOLOGY 1.5.1 Gold standards culture based analysis The use of culture media is the classical methodology for the detection of foodborne pathogens. Their sensitivity has shown to be high, allowing reliable results and high cost-effective performance, as the resources needed for this approach are relatively affordable. The media used can be of three different categories: general, selective and differential [63]. The General medium are mostly Chapter 1. Introduction 11 employed to perform a pre-enrichment were no selective compounds are added to inhibit the nontargeted microorganism, but allowing the growth of the target microorganism. This can also be used for the recovery of stressed bacteria in some type of samples. Selective media contain inhibitory agents, such as antibiotics, which constrain the growth of the other non-target bacteria. The concentration of these additives must be carefully evaluated, as interference in the growth of the target microorganism may occur. Differential media make easier the identification of the bacteria of interest as they will develop distinct characteristics in comparison to other microorganisms which may also grow, for instance, the presence of chromogenic compounds will produce a specific colour change. The ISO methodologies for the detection of foodborne pathogen have been focus in this type of approach for recovering, growth, isolation and identification of a specific microorganism. The EN ISO 6579-1 specify a horizontal methodology for the detection of Salmonella spp. in the food supply chain, not only in products intended for human consumption, but also in animal feeding and environmental samples. The recent update in 2017 also included testing in milk and milk product, as well as the testing in animal faeces, dust and boot socks with the objective of preventing crosscontamination and detect the contamination source [64], the procedure is depicted in Figure 1.4. As it may be observed, it follows the standard approach in the sense of a first pre-enrichment in a general medium, BPW, followed by a selective enrichment in two different media, RVS and MKTTn. After the enrichment both broths are plated on two selective and differential media, which can be XLD and a second medium which is open for the laboratories to choose, and finally, if typical colonies are observed they are purified on a general agar medium, typically Nutrient Agar (NA), for subsequent biochemical identification, and serological analysis if needed.. The full methodology takes a total of 6 days, being extremely laborious, lengthy and time-consuming. Figure 1.4. Scheme of horizontal method for the detection and serotyping of Salmonella (ISO 6579-1:2017) SARAH AZINHEIRO 12 For the detection of E. coli O157 a slightly different approach is followed in order to differentiate this specific serogroup from all other E. coli. Attending to the ISO 16654:2001 (Figure 1.5), after a selective enrichment an immunomagnetic separation (IMS) is performed taking advantage of specific antibodies to recover this pathogen. The captured bacteria are then plated on two different solid media, and the subsequent steps are as those for Salmonella spp., re-isolation of typical colonies, followed by biochemical and serological analysis. This approach delivers the final results in 5 days, and by the use of the IMS the process is less laborious than the protocol for the detection of Salmonella spp. Figure 1.5.Scheme of horizontal method for the detection of E. coli O157 (ISO/TS 16654:2001) The detection of L. monocytogenes needs more time than the protocols previously presented, taking the full analysis 7 days, following the procedure described in the ISO 11290-1:2017 (Figure 1.6). The slower growth of this bacterium influences the time of analysis, and the need for the use of selective medium also delays its growth. One semi-selective enrichment, in HF, followed by a selective enrichment, in FF, are needed (24 h each), and then platting on two solid media, ALOA and second selected by the laboratory, for up to 48 h. The culture-based approaches make the analysis for the detection of foodborne pathogens very laborious and lengthy and the time of analysis is also not compatible with the intense demands of current food production system existing nowadays. These fact highlight that the food industry is in need of better methodologies. Chapter 1. Introduction 13 Figure 1.6. Scheme of the horizontal method for the detection and enumeration of Listeria monocytogenes and of Listeria spp. (ISO 11290-1:2017) 1.5.2 Commercial methods available for improved analysis To cover the needs of the food industry, new products are being introduced into the market to reduce hands-on time, increase sensitivity or to allow the detection of several pathogens in multiplex. To assist the culture-based methodologies, new type of solid media have been developed, as Petrifilm (3M), Dryplate (MICROKIT, Spain) or SimPlate (IDEXX Laboratories, USA) (Figure 1.7 A, B, C), with the advantage of being ready–to-use, and removing agar preparation. Regarding the colony counting for microbial concentration determination, can also be improved by other technologies, like Spiral Plates counter (Figure 1.7 D) which allows automatization of the process. Bacterial quantification by Most Probable Number can also be automatized as exemplified by the TEMPO system (BioMérieux, France) (Figure 1.7 E) avoiding the usual tedious preparation. All of them have the aim to provide a faster and simpler analysis, however the culture-based methodologies continue to require a long period of time to reach the results. Different molecular alternatives have emerged, being immunological methodologies well established in the market, with different lateral flow products (Figure 1.7 F) which are based on colloidal gold immunoassay strips, and are sold by different companies (Biocontrol, Merck, Neogen among others) for different pathogens [65]. The major drawback of this approach falls on its sensitivity, being a good methodology for preliminary analysis, but always with the need to be complemented with a more sensitive detection. Automated system for immunological detection have also been developed and commercialized, as VIDAS® from BioMérieux (France) (Figure 1.7 G). SARAH AZINHEIRO 14 Figure 1.7. Examples of product in the market developed to improve the food analysis and detection of pathogens. (A) SimPlate, Easy-to-count wells with change of colour for bacteria quantification; (B) Dryplate, ready to use plates with dehydrated media; (C) Petrifilm, ready selective culture system in a disk for fast enumeration; (D) Spiral Plater, automatic and standardized plating of a sample generating decimal dilution on a single plate; (E) TEMPO, fully automated enumeration system, using the most probable number method; (F) Lateral Flow , immunochromatographic test, based on gold-labelled antibodies. (G) VIDAS, automated benchtop immunoanalyzer, based on the Enzyme Linked Fluorescent Assay (ELFA) technology. Chapter 1. Introduction 15 1.5.3 Alternative sample pre-treatment strategies Most of the efforts for reducing the time of analysis without affecting the reliability and sensitivity of the traditional methodologies in food analysis have been focused in molecular techniques. However, sample pre-treatment continues to be the bottleneck of food analysis, as the enrichment continues to be a crucial step to increase the concentration of the targeted bacteria to obtain a sensitive detection, particularly when the target pathogen tends to be present in low concentration. Additionally, the current legislation requires in some cases to ensure the absence of the pathogens in the food product and to fulfil this requirement the detection methodology needs to be able to detect 1 cfu/ 25 g of samples. The turnaround time of the molecular approaches is not limited by the technology itself but by the need to always perform a pre-enrichment step of at least 16h to 24h. Extensive selection of media have been formulated with this aim, and alternatives are constantly appearing in the market to allow faster growth or reduce the natural microflora present in the sample, with the addition of proper selective agents. However, alternatives to significantly reduce the time needed for the sample pre-treatment are not available in the market. Different approaches have been more recently studied to overcome this fact with alternatives to concentrate the bacteria, reduce or substitute the enrichment time. Improvements in the ISO protocols were already implemented, with the inclusion of the Immunomagnetic Separation (IMS) for the detection of E. coli O157 [34] as previously described, to capture the specific bacteria cells, and separate them from the remaining microorganism present, to allow a specific detection with downstream culture-based methodology. IMS consist in the binding of a specific antibody to a magnetic particle to allow its recovery by magnetic force, which has been studied to be use for the detection of several pathogen to shorten the enrichment or even replace it completely thanks to the possibility of concentrating the microorganisms of interest [66–68]. Similar approaches involving the use of specific ligands, as antibodies in a solid phase have emerge with the same objective [69,70], and having the advatage to concentrate the bacteria, separating them from the food matrix and enrichment solution, allowing the removal of inhibiting compounds of the DNA amplification reaction or other molecular approaches [71]. Other alternatives for an indirect detection of the pathogen, as Bacteriophage amplification assays (PAA), have been reported to decrease the analysis time [72]. With the addition of a specific phage which will infect only living cells, the method takes advantage of its faster replication compared to the bacteria. Besides reducing the time of analysis, this methodology also allow to detect only viable pathogen cells and has been combined with qPCR for the detection of plant and human pathogens [73,74]. The matrix lysis approach developed by Rossmanith et al., [75], and further optimized later [76] relies in a completely different strategy, where the enrichment step is totally removed. The method consists on the degradation and solubilisation of the whole sample in order to obtain a sufficiently small pellet capable to be fully processed in the DNA extraction step for later DNA amplification analysis. However the authors were only capable to perform the analysis in 6.25 to 12.5 g or mL of sample, as the bigger quantity of starting sample, the bigger the pellet recovered will be and this makes it harder to process. This novel approach showed to be able to detect different foodborne pathogens, such as S. aureus, Salmonella spp. and L. monocytogenes in several food matrixes ([77– 79]). With the same purpose to decrease the time spent in the sample pre-treatment, a different procedure, known as short enrichment, has been developed combining a reduced incubation in an enrichment medium, with the degradation of the food debris recovered from the liquid portion, and the recovery of the bacteria by centrifugal steps. Fachmann et al. use this methodology allowing the detection of Salmonella spp. in meat samples in a short enrichment of 3 h [80]. SARAH AZINHEIRO 16 Overall different strategy could be optimized to reduce the time of sample pre-treatment and allow a sensitive detection of pathogens in food commodities. Some of them, presented in Figure 1.8 were evaluate in this project in order to understand their advantage and limitations. Figure 1.8. Graphical representation of the alternative methodologies to improve the sample pre-treatment tested in this project. (A) For the Immunomagnetic separation (IMS) the magnetic beads were functionalized with a specific antibody for the detection of L. monocytogenes. (B) The solid phase capturing approach was tested using a PDMS sponge inserted in a microfluidic device: for the specific detection of L. monocytognes when the sponge was functionalized with a specific antibody; or for multiplex detection of L. monocytogenes and Salmonella spp. by a non-targeted detection using a universal ligand, ApoH protein. (C) The indirect detection using bacteriophage was also tested for the specific detection of S. Enteritidis. (D) The matrix lysis approach was first developed by Rossmanith et al., [75] for smaller samples size and in this work the detection of L. monocytogenes in 25 g of samples was evaluated. (E) Finally the short enrichment reported by Fachmann et al [80] approach was tested for the detection of L. monocytogenes and E. coli O157 in simplex and a multiplex detection of E. coli O157 and Salmonella spp.. Chapter 1. Introduction 17 1.5.4 Polymerase Chain Reaction (PCR) and real-time PCR (qPCR) Trying to overcome the problems found in the traditional culture-based methodologies, molecular approaches, such as those based on nucleic acids have been studied. Amplification methods due to their ability to increase the concentration of DNA are especially important for the detection of microorganisms in very low numbers. PCR is the gold standard amplification technique. The technique consist on the denaturisation of the double-stranded DNA (dsDNA) at high temperature (95 ºC), then the temperature is decreased to enable the annealing of the primers (56 ºC) and finally the temperature is increased again (72 ºC) to proceed with the extension of the primers using as template the complementary sequence. This last two step can be combined in one single step at an intermediate temperature (60-65 ºC). For the conventional PCR, also called end-point PCR, the results need to be visualized after reaction in an agarose gel after electrophoretic separation. The PCR reaction can also be performed in real-time by the monitoring of a fluorescent dye, not requiring additional manipulations of the amplification product. In real-time PCR (qPCR), the fluorescence obtained is plotted against the quantification cycle (Cq), when the florescence amount is higher than the background. Thanks to this, an inverse correlation can be established between the Cq and the concentration of DNA loaded, serving as a relative quantification of the target. The fluorescence signal can be generated by two different chemistries. The first approach consists on the binding of a DNA intercalating dye to the dsDNA molecules generated during amplification, being SYBR Green the most commonly used for this purpose. One of the inconvenients of this method is that these dyes will bind to any dsDNA present in the reaction, and if non-specific amplification are originated that can lead to false positive results. However this effect can be avoided by the analysis of melting curves, which allow to differentiate between the different amplification products. The melting analysis consists in the assessment of the dissociation characteristics of the amplicon fragment generated, being the value of melting temperature (Tm) obtained when 50% of the DNA molecules are single-stranded (ssDNA). The melting temperature is influenced not only by the length of the DNA fragment, but also by the guanine-cytosine (GC) content, which will allow to discriminate between the specific and non-specific amplicons presents in the reaction [81]. Another way to obtain the fluorescent signal tracked in the real-time approach is the use of duallabelled probes, also called hydrolysis or TaqManTM probes, which are complementary to the amplicon, increasing the specificity of the reaction. This probe have a fluorophore attached to the 5’ and a quencher to the 3’ which absorbs the fluorescent signal when the probe is intact. However, over the amplification the probe is hybridised to its target site in the annealing step, and in the extension step, the polymerase will cleave the probe due its 5’-> 3’ exonuclease activity, separating the fluorophore from quencher and allowing the detection of the signal. qPCR are now routinely used for the detection of pathogens in different type of samples in the clinical field. Also in the food industry the ISO regulation already integrated this technology in different standards, being the first one, ISO 13136:2012 for the detections of STEC and determination of O157, O111, O26, O103 and O145 serogroups, and more recently the ISO 15216:2019 for the detection of hepatitis A virus and norovirus by RT-qPCR. Different guidelines have been set in order to develop a PCR/qPCR methodology and evaluate its performance and validate its implementation in the food industry [82–84]. An important reason for the use of DNA-based techniques in the detection of pathogens is the decrease in the analysis time. By the ISO 13136:2012 the STEC identification can be obtained in only 3 days (Figure 1.9), comparing with the 5 days with the previous protocol. Additionally to the protocols already implemented in the ISO regulation, different PCR kits have been validated according with ISO 16140, in order to be used for the detection of other foodborne SARAH AZINHEIRO 18 pathogens, as BAX® System (Hygiena), BACGene (Eurofins), SureTect™ (Thermo Scientific) among others. Figure 1.9. Scheme of the horizontal method for the detection of Shiga Toxin-producing Escherichia coli (STEC) and the determination of O157, O111, O26, O103 and O145 serogroups (ISO/TS 13136:2012). In this methodology the sample pre-treatment was also improved, specifying distinct enrichment depending on the type of samples analysed, being mTSB supplemented with Novobiocin or acriflavin where high background microflora is normally present, and BPW, a general medium, when stressed bacteria could be present. 1.5.5 Alternative isothermal amplification As PCR needs complex equipment to perform fast, and accurate changes of temperature, the development of isothermal amplification techniques have arisen interest. These allowed to reduce the cost of the analysis and made easier their integration in portable platforms, as well as to reduce energy consumption. Many different techniques, performing isothermal amplification, have been described in the literature, being the first one Nucleic acid sequence-based amplification (NASBA) developed in 1991, which allowed to amplify single-stranded RNA (ssRNA). Since its development, other options have emerged not only for RNA but also for DNA amplification with diverse enzymatic mechanisms such as Loop-mediated isothermal amplification (LAMP), Recombinase polymerase amplification (RPA), Rolling circle amplification (RCA), Strand displacement amplification (SDA), Polymerase spiral reaction (PSR) among others [85]. In this project, RPA and LAMP will be the alternative amplification techniques tested due to their advantage and possibilities for naked-eye visualization of the results. 1.5.5.1 Loop-mediated isothermal amplification (LAMP) Loop-mediated isothermal amplification (LAMP) is the most studied technique for isothermal DNA amplification [86]. It was originally described in 2000 [87] attracting attention due to its potential to substitute the traditional PCR analysis, by the rapidity and accuracy of the reaction Chapter 1. Introduction 19 achieved in constant temperature (60-65 ºC), with results between 30 min to 1 h . The amplification is performed thanks to a DNA polymerase with strand-displacement activity and 4 to 6 primers (outer, inner and loop primers). To begin with, the formation of loops at the end of the specific sequence need to occur to allow the exponential cycling amplification and elongation. In this sense, first the 2 inner primers, FIP and BIP, which are composed by two different region recognition (F2c-F1 and B1B2c), hybridize with the complementary target in the F2c/B2c region, leaving the F1/B1 part of the primer free. The sequence is then extended by the polymerase. This product is displaced by the syntheses of a second strand, initiated by the outer primers (F3/B3). The ends of the first product are now free do form a loop with the hybridization of the F1/B1 region, called dumbbell structure. In this format the sequence contain multiple site to initiate the syntheses in the open loop. The amplification proceed from these multiple sites, where the products grows and forms long concatamers. The loops primers (LF/LB) are optional and accelerate the reaction, by providing additional starting points for the polymerase, Figure 1.10 illustrate the LAMP reaction. Different ways to visualize the results can be achieved by this technique, as presented in section 1.5.5.3. Figure 1.10. Loop-mediated isothermal amplification (LAMP) reaction (image used with licence from JOHN WILEY AND SONS [88]). 26 Chapter 2. Objectives 27 2 OBJECTIVES 2.1 MAIN OBJECTIVES The main objective of the present work was the development of a new analytical method to enable the detection and identification of three of the most problematic foodborne pathogenic bacteria, Salmonella spp., E. coli O157 and L. monocytogenes in food samples reducing the time of analysis and evaluating the possibility of its integration in miniaturized devices. 2.1.1 Specific Objectives To fulfil this main objective, the specific objectives were  To develop, optimize and test and improved sample pre-treatment strategy for optimal recovery of the three foodborne pathogens of interest in multiplex, allowing a reduction in the time spent in this critical part of the analysis.  To develop, optimize and compare highly specific isothermal amplification methods, LAMP and RPA, to obtain a sensitive analysis detection and allowing a naked-eye detection of the three pathogens of interest.  To develop an optimized methodology based in the results obtained with the previous approaches tested and to integrate the DNA amplification step in a miniaturized device.  To evaluate the novel protocol and device against traditional detection methods. SARAH AZINHEIRO 28 2.2 THESIS STRUCTURE The thesis was divided in seven chapters.  Chapter 1 correspond to a general introduction, where the importance of the detection of foodborne pathogen was state and the different methodology use in the food industry for this purpose were presented. Beside traditional techniques (culture-based), a brief explanation of DNA amplification by PCR/qPCR and isothermal amplification (RPA and LAMP) was included, and as well as the different product and devices available in the market;  Chapter 2 state the objective of the project;  Chapter 3 describe the Methodology employed to perform the project; Results and Discussion was separated in three chapters (sample pre-treatment approaches, DNA amplification/ detection methods and Final methodology):  Chapter 4: results of the evaluation of different approaches to improve the Sample pretreatment;  Chapter 5: the assessment of results testing several alternatives for the DNA amplification, in order to obtained a isothermal amplification allowing a naked-eye detection;  Chapter 6: the optimization and validation of the final methodology, and the integration on miniaturized device, testing two different systems;  Chapter 7: summarize the final conclusions of the work presented and future work is proposed to complement the results obtain and improve the methodology for the detection of foodborne pathogens in food samples. Chapter 2. Objectives 29 SARAH AZINHEIRO 30 31 CHAPTER 3. METHODOLOGY 32 Chapter 3. Methodology 33 3 METHODOLOGY 3.1 INTRODUCTION In order to develop a new methodology for a faster, but reliable detection, the different steps of the analysis were optimized as defined in the objective of the project, involving the improvement of sample pre-treatment, DNA amplification and optical detection. To improve the sample pre-treatment different approaches were evaluated to concentrate the bacteria or reduce the time of analysis. First, the analysis of a standard 24 h enrichment was tested with several selective and non-selective media to enhance the growth and competiveness between the targeted microorganisms. The concentration of bacteria to increase the sensitivity of the methodology was evaluated by Immunomagnetic separation (IMS) and secondly using a functionalized solid phase to capture the bacteria. Finally, protocols for the reduction of the enrichment time were developed and tested, including an indirect detection using bacteriophages and other two approaches involving the degradation of the matrix and a shorter enrichment. The second step, consisted on the evaluation of different DNA amplification techniques with the aim to obtain a methodology which would allow naked-eye detection and could be easily integrated in a miniaturized device. We focused in the study of two main isothermal amplification techniques, RPA and LAMP, which were first compared against qPCR implementing two different detection chemistries (SYBR-qPCR and Probe-qPCR). Two naked-eye detection strategies were evaluated for RPA (RPA-LF and RPA-SYBR), and three other for LAMP (Turbidity, MUA-AuNP and colorimetric mastermix). All these approaches were used for the analysis of different samples and targeting different pathogens, namely L. monocytogenes, Salmonella spp. or E. coli O157, in order to compare their performance. After the final methodology was selected, the ability to detect these three microorganisms was evaluated, as well as the integration of the DNA amplification part in a miniaturized device. This chapter describes the protocols and reagents used to perform this complete analysis and optimization to reach the final methodology. 3.2 REFERENCE BACTERIA STRAINS USED To perform all experiments developed in this project, reference strains acquired from the Spanish Type Culture Collection (CECT), presented in Table 3.1, were used, including Listeria monocytogenes serovar 4b (WDCM 00021), Escherichia coli serotype O157:H7 (WDCM 00014) and Salmonella enterica serovar Typhimurium (WDCM 00031). The code provided for these strains correspond to the reference from the World Data Centre for Microorganisms (WDCM). A non-toxicogenic strain of E. coli O157:H7 was selected for safety reasons. A wild strain of Salmonella enterica serovar Enteritidis (S1400) from poultry, was an exception, belonging to the private collection of the University of Bristol. SARAH AZINHEIRO 34 Table 3.1. List of reference strains use in the approaches tested Bacteria species WDCM reference Acquired from L. monocytogenes WDCM 00021 CECT E. coli O157 WDCM 00014 CECT Salmonella enterica serovar Typhimurium WDCM 00031 CECT Salmonella enterica serovar Enteritidis -* S1400 (UB) WDCM: World Data Centre for Microorganisms; CECT: Spanish Type Culture Collection; UB University of Bristol * Wild strain obtained from UB private collection. Do not have a WDCM reference All the spiking experiments were performed with fresh cultures, which were prepared by inoculating a single colony of the corresponding bacterium in 4 mL of a general media: Buffered Peptone Water (BPW, Biokar diagnostics S.A., France) or Nutrient Broth (NB, Biokar Diagnostics S.A., France). The culture was incubated overnight (ON) at 37 °C. Concentration of bacteria used in each assay, was calculated by preparing ten-fold serial dilutions of the initial culture in the same general media and two dilution were plated in duplicated on Tryptic Soy Agar (TSA) (Biokar Diagnostics S.A., France) for E. coli O157:H7 and Salmonella spp., and on Tryptic Soy Yeast Extract Agar (TSYEA, Biokar Diagnostics S.A., France) for L. monocytogenes. The plates were incubate ON, at 37 °C, and resulting colonies counted. The following formula was used to obtained the starting bacterial concentration, as specified in the FDA/BAM standard method [100] : 𝑁 = ∑𝐶 [(1×𝑛1)+(0.1×𝑛2)]×(𝑑) Eq. 1 N: plate counts in a sample C: The total number of colonies on the plates N1: The number of colonies on the plates of the first proper dilution degree; N2: The number of colonies on the plates of the number of colonies on the plates of the second proper dilution degree; D: Dilution Factor (the first dilution degree). 3.3 DNA EXTRACTION 3.3.1 Pure culture To obtain the DNA extract from pure culture to be used in the following experiments a simple thermal lysis was performed. Briefly, 1 mL of a ON cultures obtained as described in section 3.2, was first centrifuged at 16,000 × g for 5 min to concentrate the bacteria, the supernatant was removed, the pellet resuspended in 1 mL of TE 1X (10 mM Tris-HCl, 1mM EDTA (Sigma–Aldrich, USA, pH 7.5) and centrifuged again in the same conditions. Again, the supernatant was discarded, the pellet was resuspended in 300 µL of TE 1X. This new bacterial suspension was incubated for 15 min at 99 °C, with constant agitation (1400 rpm), to lyse the cells in a Thermomixer comfort (Eppendorf AG, Germany). Chapter 3. Methodology 35 Finally, the thermally lysed bacteria were centrifuged at 16,000 × g for 5 min at 4 °C, to separate the DNA (supernatant) from the other cell debris (pellet). The supernatant was transferred to a clean tube, and stored at −20 °C until needed. This methodology was performed to obtained DNA standard from the targeted pathogens, but also from other strains to evaluate the inclusivity and exclusivity of the amplification techniques. 3.3.2 Complex food matrixes For the DNA extraction from food matrixes, two different approaches were performed depending on the type of bacteria to be detected, unless otherwise specified. In this sense, the extraction of DNA from E. coli O157 and Salmonella spp., which are Gram-negative bacteria, was performed by thermal lysis. For L. monocytogenes, this step needed the addition of an enzymatic mixture in order to cause disruption of the cell wall, due to the strong peptidoglycan barrier, characteristic of Gram-positive bacteria. Before starting with the DNA extraction process, the samples were submitted to a washing step. Briefly, 1 mL of the pre-enriched sample was centrifuged at 380 × g, for 2 min to pellet any large food particles, and the supernatant was transferred to a new tube, which was centrifuged at 16,000 × g for 5 min. The supernatant was discarded, and the pellet was resuspended in PBS. The suspension was centrifuged again under the same conditions. The supernatant was discarded again, and the pellet was processed for thermal lysis or enzymatic lysis as described in section 3.3.2.1 and 3.3.2.2. 3.3.2.1 Thermal lysis The thermal lysis consisted in the resuspension of the pellet in 300 μL of 6 % Chelex®100 (w/v. Bio-Rad Laboratories, Inc., USA) and incubated at 56 °C for 15 min. The addition of 25 μL of Proteinase K (10 mg/ mL, Macherey-Nagel, Germany) was also included in this first incubation step for the RPALF approaches (section 3.5.3.1.3) when surface samples were analysed. The lysis is then performed by heating the samples at 99 °C for 10 min. Both incubation steps were performed with constant agitation in a Thermomixer comfort (Eppendorf AG, Germany). Finally, the samples were centrifuged at 16000 × g at 4 °C for 5 min and the supernatants were transferred to new, clean, and sterile tubes. 3.3.2.2 Enzymatic lysis The DNA extraction from food samples spiked with L. monocytogenes was performed based on the Lysis-GuSCN method described by Kawasaki et al., [101] and modified by Garrido et al., [102]. The pellet obtained from section 3.3.2 was treated with 200 µL of an enzymatic solution containing 1 mg/ mL of achromopeptidase (Sigma–Aldrich, USA) and 20 mg/ mL of lysozyme (Sigma–Aldrich, USA) in TE 2X with 1.2% of Triton X-100 (Sigma–Aldrich, USA). The samples were incubated for 30-60 min at 37 °C, with constant agitation in a Thermomixer comfort (1400 rpm). After the incubation, 300 µL of a solution containing 4 M of Guanidine isothiocyanate (Sigma–Aldrich, USA) and 1 % of Tween 20 (Sigma–Aldrich, USA) were added, and 400 µL of this solution were transfer to 400 µL of 100 % isopropanol (Sigma–Aldrich, USA), and centrifuged for 10 min at 16,000 × g. The pellet was rinsed with 1 mL of 75 % isopropanol, resuspended in 160 µL of sterile Milli-Q water, and incubated at 70 °C, for 3 min. The DNA was separated from any remaining debris by a 5 min centrifugation at 16,000 × g and 4 °C, and supernatant use for DNA analysis. SARAH AZINHEIRO 42 modification, the sponges were functionalized with either 5 μg/ mL of ApoH protein (ApoHTechnologies, France), for a non-specific bacteria targeting or with 10 μg/ mL of antiL. monocytogenes antibody (Ab) (MAB8953, Abnova, Taiwan), previously evaluated and used in the IMS approach described above, for specific L. monocytogenes targeting. Two hundred microliters of the solution were added to a tube with the sponge, vortexed vigorously and incubated ON at 4 °C to allow the binding of the protein or antibody to the sponge. The sponge was then washed three times with PBS, and stored at 4 °C until use. 3.4.2.2.3 Bacteria concentration using pure cultures To evaluate the capacity of the device with the PDMS sponge to concentrate the bacteria, the capture efficiency was calculated. For this, a pure bacterial solution of L. monocytogenes and Salmonella spp. separated and in co-culture, were flown through the sponge, with a controlled flow rate of 10 μL/ min. A washing step with 800 μL of PBS at the same flow rate was performed to recover all bacterial cells unbound to the sponge. The outlet solution was collected and used for culture plate counting to confirm the capture efficiency in the sponge. To determine the concentration of the noncaptured bacteria, serial dilutions were made in PBS and plated on COMPASS and Xylose Lysine Desoxycholate Agar (XLD, Biokar diagnostics S.A., France) for the isolation of both pathogens separately. The capture efficiency was calculated using the following equation: CE sponge device (%)=Nt − Ne Nt x 100 Eq. 4 Where Nt is the number of bacterial cells in the sample introduced in the device and Ne is the number of uncaptured bacterial cells, recovered from the device outlet. Concentrations between 103-105 cfu of each microorganism, in pure or in mix cultures, were passed through the device containing the sponge functionalized with ApoH protein and the anti-Listeria antibody, not only to determine the capture efficiency but also to define the limit of detection by qPCR using this approach. For the Limit of Detection (LoD) evaluation, after the sample solution passed through the device, DNA extraction from the sponge was performed by enzymatic lysis as described in section 3.3.2.2 with incubation of 30 min for downstream qPCR analysis, as described in section 3.5.2.2 in Table 3.8. 3.4.2.2.4 Evaluation of spiked surface sample To ensure the reliability of the results, the detection of L. monocytogenes and Salmonella spp. were tested spiking stainless steel surfaces in order to test the applicability of the developed methodology in the evaluation of cleaning procedures in the food industry. To contaminate the surfaces, an ON culture was diluted and 105 cfu of a bacterial mixture was spread on the surface and it was allowed to dry at RT. The bacteria were recovered with a cotton swab pre-moisturised in PBS with 0.01% of Tween 80, and re-suspended in 2 mL of PBS by vortexing. One mL of the solution was passed through the device as specified in section 3.4.2.2.3, and the sponge treated for downstream DNA extraction and qPCR. The capture efficiency was also determine for these surface samples. Chapter 3. Methodology 43 3.4.3 Protocols for time reduction 3.4.3.1 Phage Amplification Assay (PAA) Another approach tested to improve the sample pre-treatment, is the phage amplification assay (PAA), as an indirect detection strategy. This approach was used for the detection of S. Enteritidis using the Salmonella phage vB_SenS_PVP-SE2 (GenBank accession no. MF431252.1), previously named ϕ38, isolated by Sillankorva et al., [113]. 3.4.3.1.1 PAA optimization For the determination of the phage stock concentration, ten-fold serial dilutions of the phage prepared in SM buffer (100mM NaCl, 50mM Tris, 8000mM MgSO4·7H2O, pH 7.5) were performed, and a mixture containing 5 mL of molten semi-solid Luria-Bertani (LB) (7.5 g/ L agar), 100 μL of an ON culture of S. Enteritidis and 100 μL of the corresponding phage dilution were poured on solid LB. These plates were incubated at 37 °C, ON. This methodology is based in the detection of the phage DNA, being limited by the ability of the amplification technique to detect it. For that reason, the first step was to establish the concentration to be use to infect the S. Enteritidis in the samples, which originate a positive result by qPCR. Ten-fold dilutions of the pure virus in BPW were directly analysed by qPCR, as describe in Table 3.8 of section 3.5.2.2. The phage stock (dilution 0) was diluted 1:2 and this was used as the highest concentration. The minimum time of enrichment to reach detectable levels of phage was also determined in spiked samples. For this, after a pre-incubation to grow Salmonella cells and allow a higher quantity of cells to be infected by the virus, the time to let the phage multiply inside the cells was analysed. To performed this evaluation, samples were treated as describe in section 3.4.3.1.2 and aliquots of 1 mL taken after different times of the second enrichment, 3h and 6h, and compare with samples without this incubation (T0). 3.4.3.1.2 Evaluation with complex food matrixes Raw chicken breast samples were contaminated with four different contamination level: <10, 10-102, 102-103, >103 cfu/ 25g. The analysis was performed as following: 25 g were weighed and 225 mL of 37 °C pre-warmed BPW were added, the matrix was homogenized for 30 s in a Stomacher 400 Circulator (Seward Limited, UK); then 1 mL of the appropriate dilution of S. Enteritidis, prepared as mentioned above, was added and homogenized again for 30 sec. These samples were first incubated for 3 h at 37 °C with agitation (120 rpm). After this initial incubation step, 103-104 pfu/ mL vB_SenS_PVP-SE2 phages were added (concentration selected after evaluation of pure phage LoD) and the matrix was re-incubated to perform the second enrichment. Confirmation of the presence of S. Enteritidis in spiked food samples after pre-enrichment, was performed by streaking the pre-enriched samples on XLD. The plates were incubated at 37 °C overnight. A different thermal lysis DNA extraction protocol was performed, were no washing step was made, and 1mL of sample was directly heated to lyse the bacteriophage as describe in section 3.3.1. Once finished, the samples were centrifuged at 4000 × g for 10 min and 4 °C. SARAH AZINHEIRO 44 To verify that the proposed methodology only detected viable bacteria, 6 additional samples were spiked with dead S. Enteritidis. Non-viable bacteria were obtained by an autoclave step at 121 °C for 30 min to completely inactivate the bacteria. Once the treatment was completed, 103 105 and 107 cfu/ mL of dead bacteria were added to the corresponding food sample, and processed as described above. The bacteria were plated before autoclaved to determine the concentration of dead cells. 3.4.3.2 Matrix lysis In a molecular analysis, the standard procedure is to do an enrichment and take a small aliquot, normally 1 mL to be tested. In the Matrix lysis methodology the whole food sample was treated to degrade the tissue and concentrate the bacteria in a small pellet. The protocol was based on the one described by Rossmanith et al. [75] with some modifications in order to process 25 g of sample instead of the 6.25-12.5 g described by the authors, and in this study, this approach was developed for the recovery of L. monocytogenes from the sample. Thereby, 25 g of sample were added to a stomacher bag with filter, and homogenize with 40 mL of sucrose buffer (0.25 M sucrose, 1 mM EDTA, 0.05 M Tris, pH 7.6), for 30 s in a Stomacher 400 Circulator [76]. The liquid part was recovered and transferred to a clean 50 mL tube. Additional 10 mL of the sucrose buffer were added to the bag, homogenized for an additional 10 s, and the extra liquid was added to the same 50 mL tube, to reach 40-45 mL. The tubes were centrifuged at 8960 × g for 10 min, the supernatant was decanted and the pellet, containing bacteria and food debris, were resuspended in PBS to which 5 mL of the protease buffer (1/100 dilution in PBS of Alcalase and Neutrase, Novozymes, USA) were added. The samples were incubated horizontally at 37 ºC for 30 min with constant agitation (200 rpm), then the tubes were centrifuged again under the same conditions and once more the supernatant was decanted and the new pellet was resuspended in the lysis buffer as specified by different authors [76,114,115] containing 8 M urea, 1 M MgCl2, 50 mM Tricine and 0.35 % of a surfactant mixture with a hydrophilic/ lipophilic balance equivalent to that of Lutensol AO-07 [76]. The mixture was incubated once more at 37 ºC for 30 min with constant agitation, and was followed by a new centrifugation and the supernatant was discarded. The pellet was resuspended in washing buffer containing PBS with 0.35 % of a surfactant mixture as above mentioned, incubated 37 ºC for 30 min with constant agitation, and centrifuged again. Finally, the pellet was recovered in 1.5 mL of PBS, transferred to a clean tube and centrifuged at 11000 × g for 5 min. The resulting pellet was used for DNA extraction as described above in section 3.3.2.2, reducing the incubation with the enzymatic solution to 30 min and qPCR process as mentioned in Table 3.8 of section 3.5.2.2. 3.4.3.3 Short enrichment Another approach to reduce the time of analysis, is based on the performance of a shorter enrichment and recovery of the total liquid from the enrichment for downstream treatment to remove the remaining food debris and obtained the bacteria cells pellet. This will allow to reach a lower LoD, thanks to this step of bacteria growth, when compared with the matrix lysis. The methodology was tested for L. monocytogenes and E. coli O157 in simplex, and for E. coli O157 and Salmonella spp. in multiplex. The modifications between the protocols used for each one is present in Table 3.4, as the enrichment conditions, type of samples analyzed, centrifugation speed and time, and type of DNA extraction applied. Chapter 3. Methodology 45 Table 3.4. Short enrichment conditions Pathogen tested Type of samples Enrichment media Centrifugation conditions DNA extraction used E. coli O157 Ground beef Leafy greens mTSBn (3h) 4700 × g, 5 min Chelex®100 E. coli O157 Salmonella spp. Ground beef Chicken breast BPW + 0.4 % Tween 80 (3h) 8960 × g, 10 min L. monocytogenes Smoked salmon TSB (5h) Lysis-GuSCN method mTSBn correspond to the commercial modified TSB supplemented with 20 mg/ L of novobiocin. 3.4.3.3.1 Protocol optimization and sample treatment Starting with L. monocytogenes detection, two aspect of the protocol were optimized, the volume of broth used to dilute the sample and proceed with the enrichment (25 mL vs 50 mL), and whether or not to perform the incubation under constant shacking (200 rpm). To perform this evaluation, the samples were spiked with 102-103 cfu/ mL, an aliquot was taken at T0, and a second one after 4 h of incubation (T4) under each condition. This aliquots were ten-fold serially diluted, plated on TSYEA and incubated overnight at 37 ºC, to determine the bacterial increase. After this condition sets, the samples were analysed with the following established protocol. Twenty five grams were weighted in a stomacher bag with filter (< 250 μm) and the corresponding bacterial concentration was added, along with 25 mL of enrichment medium mentioned in Table 3.4 pre-warmed at 37 °C. The matrix was homogenized for 30 s in a Stomacher 400 Circulator (Seward Limited, UK) or hand-massaged in the case of the multiplex detection, due to the type of sample analysed. The samples were incubated at 37 °C with constant agitation (200 rpm) for the corresponding time. After incubation, the whole liquid was recovered and transferred to a conical 50 mL tube. The tube was centrifuged at 8960 x g for 10 min and the supernatant was discarded, the pellet resuspended in 45 mL of protease buffer, as described above in section 3.4.3.2, and was incubated horizontally at 37 °C for 10 min at 200 rpm. After digestion, the samples were centrifuged again in the same conditions. Once more the supernatant was discarded, and the pellet was resuspended in washing buffer, as described in section 3.4.3.2 followed by a new centrifugation step. Finally, the new pellet was resuspended in 1.5 mL of washing buffer, transferred to a clean tube and centrifuged for 5 min at 11000 × g, and the pellet used for corresponding DNA extraction depending on the pathogen targeted, and downstream qPCR analysis as detailed in Table 3.8 of section 3.5.2.2. 3.5 EVALUATION OF THE DNA AMPLIFICATION ALTERNATIVES 3.5.1 Primers design For all amplification approaches specific primers had to be designed in conserved regions, in order to allow a specific and reliable detection of the pathogens. To ensure that the primers are targeting a common region within the gene of interest a consensus sequence was generated after the alignment of the target sequences with CLC Sequence Viewer (C L C Bio-Qiagen 2016). This consensus sequence was used as a template for the primer and probe design. For qPCR and RPA approaches the free online SARAH AZINHEIRO 46 software Primer3 [116] was used, and Primer Explorer V4 for LAMP primers (https://primerexplorer.jp/e/index.html). Different parameters were evaluated to confirm the primer performance, as the GC content, melting temperature (Tm), the probability to form secondary structures and primer dimers. Different genetic targets were tested and the choice of the target gene was made based in previously published studies describing their specificity to discriminate among strains. All primers were evaluated after being designed in terms of specificity by BLAST analysis. In order to develop the amplification techniques for the detection of L. monocytogenes, three different genetic targets were tested, actA, hlyA, plcA, which belongs to a same gene cluster, named Listeria Pathogenicity Island 1 (LIPI-1)[117]. The actA is a virulence-associated gene, coding for a protein involved in the actin filament assembly and shows a high discriminatory power between L. monocytogenes strains and subtyping [29], furthermore, it has already been used in different qPCR approaches [118,119]. The hlyA gene has also been applied in a variety of qPCR [75,120–122], the gene encodes for the hemolysin listeriolysin, a major virulence factors involved in host-pathogen interactions [123]. plcA, coding for the phosphatidylinositol-specific phospholipase C, is also a virulence factor, being less reported for the detection of L. monocytogenes by qPCR [124]. This last one was only used for the LAMP reaction. To detect E. coli O157, the gene rfbE, which encodes for the “O” antigen, was targeted. This gene has been highly used for the development of PCR assays to detect this pathogen, and demonstrated its good performance [125–127]. For the detection of Salmonella spp. three different genes, fimA, ttr and invA, were evaluated and used in different methodologies. The fimA is identified as one of the major fimbrial subunit genes of Salmonella spp., wherein several qPCR targeted it because of its discriminative capacity with other species [128–130]. The gene operon implicated in tetrathionate respiration (ttrRSBCA) with the same purpose was tested for the detection of this pathogen by qPCR, and also proved its specificity in previous studies [131–133]. Another genetic target, which encoded the invasion protein gene (invA), was used in LAMP reactions for real-time and naked-eye approaches development, and as fimA and ttr, this gene was targeted also by different authors [134,135]. In a different approach, primers for the detection of specific serovars of Salmonella species were designed. safA gene was targeted for the specific identification of S. Enteritidis. This gene encodes the major subunit of S. enterica atypical fimbriae, involved as a virulence factor in the host-restricted colonization of the porcine ileum, [136] and STM4497, coding for a putative cytoplasmic protein, for S. Typhimurium [137]. For the qPCR approach two different Internal Amplification Controls, one competitive (cIAC) and another non-competitive (NC-IAC), were developed to detect reaction inhibition. Both use a similar sequence for the amplification only varying in the 3’ and 5’ end with the addition of the respective primers. DNA sequence of the IAC was designed using http://usersbirc.au.dk/biopv/php/fabox/random_sequence_generator.php to generate a random DNA fragment and this sequence was subsequently used as template for primer and probe design as previously described. The cIAC was developed to be amplified with the same pair of hly primers by qPCR, having the simultaneous amplification of both targets (hly and cIAC). For this reason the cIAC can only be used in the detection of L. monocytogenes. The NC-IAC has specific primers to amplify the IAC fragment, and offer a wide range of possibilities, and the capacity to be used for the detection of other targeted pathogens. The details about the sequence of the these primers and probes, used for the different methodologies tested, qPCR, RPA and LAMP is provided in Table 3.5, Table 3.6, Table 3.7, and their specificity was Chapter 3. Methodology 47 verified in silico with BLAST® (Basic Local Alignment Search Tool, https://blast.ncbi.nlm.nih.gov/Blast.cgi?CMD=Web&PAGE_TYPE =BlastHome). All primers and probes, as well as the IAC DNA fragment, were purchased from Integrated DNA Technologies Inc. (IDT, Belgium), except the RPA probes (hly-exo-P, and hly-RPA-LF ), which were ordered from Eurogentec (Eurogentec, Belgium). Table 3.5. Primer used in qPCR reaction Target microorganism Target gene Sequence 5’-3’ Concentration used L. monocytogenes hly F GCAACAAACTGAAGCAAAGGAT 200 nM hly R CGATTGGCGTCTTAGGACTTGC hly P FAM-CATGGCACC-ZEN-ACCAGCATCTCCG-IABkFQ 150 nM actA F TTAAGACTTGCTTTGCCAGAGAC 200 nM actA R GGTGGTGGAAATTCGAATGAGC actA P CY5-AATGCTCCT -TAOGCTACATCGGAACCGA-IAbRQSp 150nM Salmonella spp. ttr F GGCTAATTTAACCCGTCGTCAG 200 nM ttr R GTTTCGCCACATCACGGTAGC ttr P NED-AAGTCGGTCTCGCCGTCGGTG-MGBNFQ 150 nM fimA F CACTAAATCCGCCGATCAAACG 100 nM fimA R TTCAGGACGATGGAGAAAGGC E. coli O157 rfbE F TCAACAGTCTTGTACAAGTCCAC 200 nM rfbE R ACTGGCCTTGTTTCGATGAG rfbE P FAM-AC TAG GAC C-ZEN-G CAG AGG AAA GAG AGG AAIABkFQ 150 nM NC-IAC NC-IAC F TTAAGACTTGCTTTGCCAGAGAC 100 nM NC-IAC R GGTGGTGGAAATTCGAATGAGC IAC P YY-AGT GGC GGT -ZENGAC ACT GTT GAC CTIABkFQ Salmonella spp. (PAA) RBP F CCGAACAACAGTCTCACCGA 100 nM RBP R CTACAATTTTACCGGCG GCG RBP P FAMAACAACAAG-ZEN-GCGCGCCCGTACGA-3IABkFQ 150 nM NC-IAC correspond to the Non-Competitive Internal Amplification Control which allow the identification of amplification inhibition, to avoid false-negative results; The probe for the IAC can be used for both competitive (cIAC) and non-competitive (NC-IAC); FAM, CY5 NED, YY (Yakima Yellow), MGBNFQ (Minor Groove Binder nonfluorescent quencher), IABkFQ (Iowa Black®FQ), IAbRQSp (Iowa Black RQ quencher), ZEN (secondary, internal quencher) are trademarks from IDT. SARAH AZINHEIRO 48 Table 3.6. Primer used in RPA reaction for L. monocytogenes detection Target microorganism Target gene Sequence 5’-3’ qRPA hly-RPA-F TTACACTTATATTAGTTAGTCTACCAATTGCG hly-RPA-R TCCAATCCTTGTATATACTTATCGATTTCATC hly-PCR-F GCAACAAACTGAAGCAAAGGAT hly PCR-R CGATTGGCGTCTTAGGACTTGC hly-exo-probe TCTGCATTCAATAAAGAAAATTCAATTTCATCZATGGCACCACCAGCATC RPA-LF hly-RPA-F TTACACTTATATTAGTTAGTCTACCAATTGCG hly-RPA-R Biotin-TCCAATCCTTGTATATACTTATCGATTTCATC hly-PCR-F GCAACAAACTGAAGCAAAGGAT hly-PCR-R Biotin-CGATTGGCGTCTTAGGACTTGC hlyLF-probe FAM-TCTGCATTCAATAAAGAAAATTCAATTTCATCTHF-ATGGCACCACCAGCATCSpC3 hly-P3F and hly-P3R are the same primer sequence used in qPCR methodology. For the RPA-LF assay, the reverse was modified with a biotin on 5’.*”Z” indicates the position of the THF. Table 3.7. Primer used in LAMP reaction Target microorganism Target gene Primer Sequence 5’-3’ Concentration used L. monocytogenes plcA FIP GCAGCGCTCTCTATACCAGGTACAttttAATGTCCATG TTATGTCTCCGTTA 1000 BIP AGGTTTGTTGTGTCAGGTAGAGCGttttCGCTTAATA ACTGGAATAAGCCAA F3 TGTGTTTGAGCTAGTGGTTTGG 200 B3 CCCATTAGGCGGAAAAGCATAT LB CATCCATTGTTTTGTAGTTACAGAG 500 Salmonella spp. InvA HK-FIP GACGACTGGTA CTGATCGATAGTTTTTCAACGTTTCCTGCGG 700nM HK-BIP CCGG TGAAATTATCGCCACACAAAACCCACCGCCAGG HK-F3 GG CGATATTGGTGTTTATGGGG 100nM HK-B3 AACGATAAACTGGACC ACGG HK-LF GACGAAAGAGCGTGGTAATTAAC 50nM HK-LB GGGCAATTCGTTATTGGCGATAG Chapter 3. Methodology 49 Continuation Table 1.7 S. Typhimurium STM4497 STM-FIP ACC TGC AGC TCA TTC TGA GCA G-TCA AAA ACA ACG GCT CCG G 400 STM-BIP GAA AAG GAC CAC AAG TTC GCG C-TCA GTG AGC ATG TCG ACG AT STM-F3 AGC CGC ATT AGC GAA GAG 100 STM-B3 GCG GTC AAA TAA CCC ACG T STM-LF TCA AAA ATC CAG AAC CCA ATC TCA 200 typh Typh FIP TGC TGC TGT GCT TAT TAC TTT GTA AGT ATT TGT TCA CTT TTT ACC CCT 1600 Typh BIP GAT GCG CAG TGC CTA TTA AAC CTT AAG GCA ACG TAT CCT CTC Typh F3 CAT CGT TGC GCA ATA GCT 400 Typh B3 GTT TTT CAA CAC CAT TTT TCA AC S. Enteritidis safA SEN-FIP AGC CCA CAG TGA GTA TCG TG-CGC TGC TGG TAG TGC ATG G 600 SEN-BIP CAG AGG TCA TGG CGC GCA AAT-GGC ATT GGT ATC AAA GGT GA SEN-F3 GTT GCT AAC ACG ACA CTG GAC 100 SEN-B3 GTG GGA TAT TCT GAG CCC CTA T SEN-LB GTG GAA TGG GAG GAG CTG GT 300 SEN-F3 GTT GCT AAC ACG ACA CTG GAC Sdf I Sdf-FIP CAT GCT CGC TGC ACA AAA G C-GAG AGG CGG TTT GAT GTG 800 Sdf-BIP CTG GAA AGC CTC TTT ATA TAG CTC A-TGA TAT ACT CCC TGA ATC TGA GA Sdf-F3 GGG AGG AGC TTT AGC CAA 200 Sdf-B3 ATG GTG AGC AGA CAA CAG Sdf-LF GCC TAA AAA ATC AGT GAC GAA CCA A 400 Sdf-LB CTG ACC TCT AAG CCG GTC AAT G E. coli O157 rfbE rfbE-FIB TGCCAATATTGCCTATGTACAGCTAttttGACAAAACA CTTTATGACCGTTG - rfbE-BIP GGATGACAAATATCTGCGCTGCTATttttTCAGCAATT TCACGTTTTCGT rfbE-F3 GGTGGAATGGTTGTCACGAA rfbE-F3 GTGGACTTGTACAAGACTGTTGAT rfbE-LB AGGATTAGCCCAGTTAGAACAAGC The final methodology was performed using the plcA,invA and rfbE set of primers with the concentration mentioned in section 3.6. SARAH AZINHEIRO 50 3.5.2 Real-time amplification The analysis were performed in a StepOne Plus™ RealTime PCR system with StepOne™ Software v2.2.2, or in a QuantStudio 5 Real-Time PCR System with the QuantStudio™ Design and Analysis Software v1.4.3. Both were obtained from the same supplier (Applied Biosystems™, USA) 3.5.2.1 SYBR-qPCR for multiplex detection Different intercalating dyes, such as SYBR Green, have been extensively used to monitor the amplification curve, but also to allow the visualization of melting curves of the amplicons generated. In this way, the identification, and differentiation, between different targets in a same sample is possible. In order to evaluate this detection methodology, a SYBR Green multiplex qPCR methodology was developed for the simultaneous identification of Salmonella spp., L. monocytogenes and E. coli O157 in food samples, by melting curve analysis. Three sets of primers were designed as described in M&M and optimized to allow the detection of the three targets. The genetic targets chosen were actA, fimA and rfbE for L. monocytogenes, Salmonella spp. and E. coli O157, respectively. The primers were designed to have different melting temperatures in order to discriminate between the pathogens based on their Tm. To do so, the melting curve of the fragments obtained by the sets of primers designed were predicted and analysed using uMELT online software (https://www.dna.utah. edu/umelt/umelt.html), to ensure the distinction between the melting peak generate. The theoretical peak pattern was correlated with the one experimentally determined over the analysis of the samples. NC-IAC was also introduced in the reaction to ensure a reliable result and avoid false negative results due to reaction inhibition. After optimization, the qPCR reaction was performed in final volume of 25 µL with 3 µL of template. The reaction was carried with a primer concentrations of 900 nM for actA and 100 nM for fimA rfbE and NC-IAC, 1 µL of NC-IAC DNA (926 copies/ µL), and 15 µL of PowerUp. SYBR® Green Master Mix (Applied Biosystems™, USA). The thermal profile consisted on a first step of UDG treatment at 50 °C for 2 min (avoid carryover contamination), followed by a hot-start activation of the polymerase at 95 °C during for 2 min, and 40 cycles of 95 °C for 15 s and 63 °C for 1 min. The melting curve stage was performed by heating at 95 °C for 15 s, cooling to 70 °C for 1 min, and increasing back to 95 °C with continuous increments of 0.015 °C/ s. Once reached the final temperature it was kept for 15 s. Infant milk formula was selected as reference food matrix for the evaluation of the developed methodology. Twenty-five mL of milk sample were inoculated with different contamination levels of each target microorganism, and diluted 1:10 with 225 mL of mTA10-MOPS (as describe in section 3.4.1.1). The matrix was homogenized in a Stomacher 400 Circulator, at 230 rpm, for 30 s. The preenrichment was performed at 35 °C for 24 h and after incubation 1 mL was used to perform the DNA extraction as described in section 3.3.2.2, by enzymatic lysis with an incubation of 30 min. 3.5.2.2 qPCR with hydrolysis probe (Probe-qPCR) The major advantage of the hydrolysis probes approach is that they increase the specificity of the assay, and do not require post-PCR analysis such as melt curve. The Probe-qPCR also simplify the detection of several targets at the same time in qPCR reaction, due to the possibility of multiple combinations between different sets of primers, without the designed complexity demonstrate by the SYBR-qPCR. Chapter 3. Methodology 51 This approach was evaluated for the detection of L. monocytogenes, Salmonella spp. and E. coli O157 in simplex, but the primers designed could be combined in case of the need for multiplex detection. The primers and probes sequences and concentrations used are presented in Table 3.5 All pre-treatment approaches were evaluated using this amplification technique as described in Table 3.8. Table 3.8. qPCR conditions for each pre-treatment approach evaluation Methodology Bacteria Target MMix Thermal profile Reaction Template Vf IMS L. monocytogenes hlyA Maxima Probe/ROX qPCR Master Mix (Thermo Fisher Scientific Inc., Waltham, MA, USA) 50 °C, 2 min 95 °C, 10 min 40 cycles of 95 °C, 15 sec 63 °C, 60 sec 2µL 20uL PAA Salmonella spp. ttr Matrix Lysis L. monocytogenes hlyA TaqMan®Fast Advanced Master Mix supplier (Applied Biosystems™, Foster City, CA, USA) 50 °C, 2 min 95 °C, 2 min 40 cycles of 95 °C, 1 sec 63 °C, 20 sec 5µL Short enrichment L. monocytogenes hlyA E. coli O157 Salmonella spp. rfbE ttr 3µL E. coli O157 rfbE Sponge device L. monocytogenes Salmonella spp. actA ttr The thermal profile was based on the manufacturer standard protocol adapting the annealing/extension step to the optimized temperature. 3.5.2.2.1 L. monocytogenes For L. monocytogenes detection by Probe-qPCR two different primers were tested. actA, previously used in the SYBR-qPCR approach and hlyA another genetic target. Regarding hly, the reaction using these primers and probes were combined with the amplification of the cIAC, as mentioned in the section 3.5.1 section. The optimization of this assay was mainly focused in the determination of the appropriate concentration of cIAC DNA to be added, in order to avoid interference with the amplification of hly, thus a range of concentrations between 20000 to 1000 copies per reaction were tested. Once the optimal concentration of cIAC was determined and no interference with the amplification of hly was confirmed, the evaluation of the qPCR reaction was performed as detailed below. The reactions were performed in 20 µL including 10 µL of Maxima Probe/ROX qPCR Master Mix (Thermo Fisher Scientific Inc., USA), 200 nM and 150 nM of hly primers and probe respectively, 2000 copies of cIAC DNA and 100 nM of its probe. Two µL of template were added per reaction. The thermal profile selected consisted in 2 min at 50 °C for UracilDNA Glycosylase (UDG) treatment, followed by 10 min at 95 °C hot-start polymerase activation, and 40 cycles of denaturation at 95 °C for 15 s and annealingextension at 63 °C for 30 s. SARAH AZINHEIRO 58 The DNA amplification was performed by colorimetric LAMP to enable naked-eye detection, and allow for a more economic analysis. L. monocytogenes, Salmonella spp. and E. coli O157 were detected targeting the plcA, invA and rfbE genes, respectively, and the primers used are described in Table 3.7, wherein the loop primer used for Salmonella detection was the LF. The reaction was performed in a final volume of 25 µL, 15 µL of mastermix and supplemented with GuHCl as mentioned in section 3.5.4.3, and the primers concentration added was the recommended by the mastermix manufacturer (NEB), 1600 nM FIB/ BIP, 200nM F3/ B3 and 400 nM of the loop primers, with the exception of plcA LF which was used in a higher concentration, 600 nM. For the LAMP targeting Salmonella spp. and E. coli O157, a mixture of the primers 10X concentrated was prepared in order to be use in the different experiments, and adding 3 µL of DNA template. However for L. monocytogenes the volume of the template had to be increased for higher sensitivity, to 6 µL, and the primer mix was concentrated 20X in order to fit all reagent in the reaction. The amplification was performed at 65 °C, for 30 min for the two Gram-negative bacteria, and 1 h to detect L. monocytogenes. The newly developed methodology was tested performing the DNA amplification in a thermocycler, but also integrated in the two alternative, miniaturized amplification devices described below, in sections section 3.7.2 and section 3.7.3 The methodology was evaluated analysing different milk samples (UHT, fresh and raw) spiked with different combinations of L. monocytogenes, S. typhimurium and E. coli O157 with several contamination levels ranging from 6 x 102 to 1 cfu/ 25 g of sample. The results were confirmed by ProbeqPCR, and also plating in selective media, COMPASS Listeria, CHROMagar™ Salmonella Plus and CHROMagar™ E. coli O157. 3.7 MINIATURIZED DEVICES The aim of this project was to develop an improved methodology for the detection of pathogens in food samples, and its integration in a miniaturized device to allow an automated and portable analysis. For this last part, three different devices previously designed and fabricated at INL were tested. One of the devices was evaluated to improve the sample pre-treatment, allowing the concentration of bacteria as mentioned in section 3.4.2.2 and the two other were tested for the integration of the isothermal amplification approaches. 3.7.1 Microfluidic device for capture and concentration of bacteria To improve the sample pre-treatment, a microfluidic device was developed with an embedded functionalized PDMS sponge to capture and concentrate the bacteria [145]. This device was composed by two parts made of PMMA with the same dimensions (25 mm × 25 mm × 10 mm) and assembled with 4 screws at each corner. A chamber in each part of the device was created to allocate the sponge, with an internal diameter of 7 mm with 5 mm depth in the top and 1 mm depth in the bottom part. An O-ring was also introduced in order to properly seal the chamber. The schematic representation and real device are presented in Figure 3.1. The device was developed with 4 inlet/outlet, but only 2 were used in this study to allow the entrance of the sample and recovery of the liquid after passing by the device as exemplified in Figure 3.1 D. The device also included an electrode for electrochemical analysis, not needed for the concentration of the bacteria. The device was designed using a 3D CAD software (SolidWorks, Dassault Systèmes Corp., Waltham, MA) and then transferred to Art-Cam software to generate the G-Code for high speed milling. Chapter 3. Methodology 59 A computer numerical control (CNC) machine (FlexiCAM, Germany) performed the milling process using 1 mm and 3 mm End-mill tips (DIXI, Switzerland). Figure 3.1. Microfluidic device with PDMS sponge for bacteria concentration. (A) Design in SolidWorks, displaying the different components. (B) The two parts of the device separated and image of the PDMS sponge. (C) Device assembled and ready to use. (D) Schematic representation of the flow passing through the device with the sponge. 3.7.2 PDMS channels prototype 3.7.2.1 Device design and fabrication A microfluidic device composed by 8 capillarity-driven microchannels was designed to perform the amplification reaction at constant temperature (Figure 3.2). The final dimensions of the device were 4 mm (thickness) × 76 mm (length) × 26 mm (width), and each channel presented a geometry of 40 mm (length) × 800 µm (depth) × 600 µm (width) allowing the loading of a maximum reaction volume of 20 µL, and presenting one inlet and one outlet. For the fabrication of this device, a mold was designed by AutoCAD software and fabricated in poly(methyl methacrylate) (PMMA) using a computer-numerical-control (CNC) miller (FlexiCAM Viper 606). After the mold completed, the PDMS replica was produced, mixing the base with the curing agent, from the Sylgard 184 silicone elastomer kit, in a proportion of 10:1 and centrifuge 6 min at 3000 SARAH AZINHEIRO 60 x g to remove the air bubbles. The PDMS was then poured onto the PMMA mold and placed under vacuum for 20 min, to ensure that the last bubbles were eliminated. Finally the replica was cured in an incubator at 65 °C for 1 h. After cooling, the replica was removed from the mold and the bonding to a glass slide was performed using oxygen plasma to seal the microfluidic channels. Figure 3.2. Microfluidic PDMS device with 8 capillarity-driven microchannels to performed DNA amplification (PDMS channels prototype). (A) Design in AutoCad and (B) PDMS replica after fabrication 3.7.2.2 DNA amplification experiment After the bonding was performed, the device was ready to be use. The amplification reaction was first prepared in 0.2 mL PCR tubes in a final volume of 25 µL, and 20 µL were loaded into the channels. Finally, the inlet and outlet were covered with another glass slide with the same dimensions and held together with a clamp to avoid evaporation during incubation. The amplification was then performed in conventional laboratory incubator as described in section 3.5.4.2. 3.7.3 Milled channels prototype The second device used to perform the DNA amplification included an integrated heating system where silicone tubing loaded with the reaction could be placed, and therefore not requiring additional instrumentation. 3.7.3.1 Prototype design, fabrication and temperature control Figure 3.3 illustrates the prototype including different units: heating elements, temperature control, fan for cooling, electrical wires, and a display. A Peltier module allowed to maintain a constant temperature and the combination with a standard passive aluminium heat sinks created a uniform temperature distribution. A platform and holder with milled channels to maintain the tubes close to the Chapter 3. Methodology 61 aluminium part, and decrease variations of temperature during the incubation, was fabricated in PMMA using a laser cutter Widlaser LS1390 Plus (Widinovations, Portugal). This holder present different possibilities to accommodate tubes with different lengths. The volume capacity of the device is flexible, between 10 to 30 µL, depending on the length of the tube selected. Figure 3.3. Miniaturized device using silicon tubing in integrated heating system for DNA amplification (Milled channels prototype). (A) Schematic representation of the Peltier module with standard passive aluminum heat sinks, and PMMA platform and holder. (B) Tubing chamber. (C) PMMA platform and holder. (D) Completing device including the heating block where the amplification reactions are placed, the TEC controller and the display. SARAH AZINHEIRO 62 To control the temperature a thermoelectric cooler (TEC) controller, TEC-1091 (Meerstetter Engineering GmbH, Switzerland), was used for high precision and stability and the monitoring of the temperature was achieved by a precision platinum sensor, connected to the high-performance closed-loop Peltier controller. A DPY-1113 TEC status display (Meerstetter Engineering GmbH, Switzerland), was connected to the device to allow the visualization of the temperature and current parameters. The miniaturized device did not need to be connected to a computer to work, being the parameters set before experiment with the TEC Service Software (Meerstetter Engineering GmbH, Switzerland), and only when some modification was required, the connection was established again by a standard USB connection port to a PC running Windows. 3.7.3.2 DNA amplification experiment After the setup of the equipment was completed, the device was pre-heated until the desired temperature was reached. The amplification reaction was prepared as for the previous device, first in 0.2 mL PCR tubes and then the whole volume was transferred to a standard silicone tubing with inner and outer diameters of 1.02 mm and 2.16 mm, respectively. The tubing with the reactions were then introduced in the device, and the holder placed over them. The incubation was performed during the desired time depending on the targeted pathogen and, when finished, the tubes were removed in order to be observed and photographed. 3.8 DNA AMPLIFICATION EVALUATION To evaluate the performance of the different amplification approaches the efficiency and dynamic range of the assay, as well as the inclusivity/ exclusivity was determined, in order to understand which could be the best option to integrate a micro-device for the detection of the three target microorganisms. This parameters were tested for the different sets of primers designed for each target. For the evaluation of the dynamic range and analytical sensitivity, two different approaches were followed, the first allowed to determine the lowest DNA concentration needed to have a positive result by the amplification technique. To accomplish this, ten-fold serial dilutions of the corresponding DNA extract were made, and used as the template in the amplification reaction. The extracts were obtained from the target microorganism after thermal lysis as described in section 3.3.1. For some of the amplification approaches, the lowest concentration of bacteria capable to give a positive result was also determined. For this second one, ten-fold serial dilutions of an ON pure culture were done and the cell lysates were prepared from each one of the dilutions. Through these results, the efficiency of the reaction was also evaluated for the real-time techniques performed in a real-time qPCR equipment, and also for the LAMP with turbidity detected by the Loopamp Realtime Turbidimeter LA-500. This parameter was calculated using the following equation [146]: 𝐸 =101/𝑠𝑙𝑜𝑝𝑒 −1 Eq. 5 where “E” is the efficiency, and the slope was calculated by linear regression. Chapter 3. Methodology 63 Values between 0.9 and 1.1, -3.1 and -3.6, and >0.99 of the efficiency, slope and coefficient of correlation (R2), respectively, are related with a good performance of the reaction. The inclusivity and exclusivity of the amplification reactions were evaluated, in order to confirm empirically the specificity observed by the BLAST analysis when the primers were designed. For this purpose, an ON culture was performed as mentioned in section 3.2 from each strain to be tested, and the DNA extraction achieved by thermal lysis as described in section 3.3.1. This DNA extracts were then used as template in the amplification reactions. 3.9 EVALUATION OF THE DEVELOPED METHODOLOGIES The methodologies tested to reduce the enrichment time and to optimize the DNA amplification step were evaluated for the detection of one, or more, foodborne pathogens in complex food matrixes. This evaluation was accomplished by the determination of several parameters. First the Limit of Detection (LoD) was established and then the analysis of the fitness for purpose performed as presented in Table 3.9. Table 3.9. Methodologies employed for the valuation of the developed protocols Type of approach Methodology Bacteria LoD Fitness for purpose Sample pre-treatment IMS L. monocytogenes 90% positives ISO PDMS Sponge device L. monocytogenes Salmonella spp. Analytical sensitivity* PAA S. Enteritidis 90% ISO Matrix lysis L. monocytogenes PoDLoDu ISO Short enrichment L. monocytogenes PoDLoD ISO E. coli O157 PoDLoD ISO E. coli O157 Salmonella spp. 90% ISO DNA amplification SYBR-qPCR L. monocytogenes E. coli O157 Salmonella spp. PoDLoD ISO Real-time comparison L. monocytogenes 90% ISO RPA-LF L. monocytogenes PoDLoD Nordval RPA-SYBR E. coli O157 PoDLoD ISO LAMP turbidity S. Enteritidis S. Tiphimurium Analytical sensitivity* ISO LAMP MUA-AuNP Salmonella spp. 90% ISO LAMP colorimetric Salmonella spp. PoDLoD Nordval The LoD was obtained by two different approaches: when at leat 90% of the samples were positives; and by the PoDLoD analysis; The fitness for purpose was evaluate by two different methodologies: ISO refers to ISO 16140-2003criteria [147]; and NordVal refers to NordVal International Protocol [148]; * The fitness for purpose of the methodology was only evaluate with pure cultures, not with food samples. SARAH AZINHEIRO 64 3.9.1 Evaluation with complex food matrixes All approaches, for samples pre-treatment and DNA amplification were evaluated analysing different types of matrixes, and targeting different pathogens. A summary of the analysis performed is presented in Table 3.10 and Table 3.11. The conditions for the new methodology are detailed in Table 3.12. All contamination of the samples were confirmed by a culture-based methodology, plating the enrichment on a selective medium in order to isolate the target pathogen, and allow the growth of typical colonies and identification of the bacteria present. With this analysis the LoD was determined and the parameters for the fitness for purpose calculated. Table 3.10. Samples analysis for the evaluation of sample pre-treatment approaches Approach Pathogen tested Type of samples Enrichment conditions DNA extraction DNA amplification Confirmation IMS L. monocytogenes Chicken breast Hard and fresh cheese Fish HF (24 h) Enzymatic lysis Probe-qPCR Plating in selective media Sponge device L. monocytogenes Salmonella spp. Surfaces - Matrix lysis L. monocytogenes Smoked salmon - Short enrichment L. monocytogenes Smoked salmon TSB (5 h) E. coli O157 Ground beef Leafy greens mTSBn (3 h) Chelex®100 E. coli O157 Salmonella spp. Ground beef Chicken breast BPW + 0.4 % Tween 80 (3 h) mTSBn + Chr O157 RVS + XLD or TBX PAA S. Enteritidis Chicken breast BPW (3+6 h) Plating in selective media For the confirmation of the short enrichment for the multiplex detection of E. coli O157 and Salmonella spp a selective second enrichment was performes in mTSBn and RVS, respectively, before plating in a selective agar. The other confirmations were performed plating directly in the respective selective medium. COMPASS or PALCOM for L. monocytogenes, XLD or CHROMagar™ Salmonella Plus for Salmonella spp. and Tryptone Bile X-Glucuronide Agar (TBX) or CHROMagar™ E. coli O157 for E. coli O157 Chapter 3. Methodology 65 Table 3.11. Samples analysis for the evaluation of DNA amplification approaches Approach Pathogen tested Type of samples Enrichment conditions Sample Pre-treatment DNA extraction Confirmation SYBR-qPCR L. monocytogenes E. coli O157 Salmonella spp. UHT milk mTA10 (35 °C, 24 h) Enzymatic lysis Plating in selective media Real-time comparison L. monocytogenes Smoked salmon mTA10 (35 °C, 24 h) + FF (35 °C, 24 h) - RPA-LF L. monocytogenes Surfaces ONE broth (30 °C, 24 h) FF (24 h48 h) + COMPASS RPA-SYBR E. coli O157 Ground beef mTSBn (3 h) Sort enrichment Chelex®100 Plating in selective media LAMP turbidity S. Enteritidis S. Tiphimurium Egg product Chicken Turkey mTA10 (37 °C, 18 h24 h) - LAMP MUA-AuNP Salmonella spp. BPW + (37 °C, 18 h24 h) - LAMP colorimetric Salmonella spp. UHT milk mTA10 (35 °C, 24 h) TSB (6 h) Sort enrichment Enzymatic lysis For the confirmation of the RPA-LF a second enrichment in Full Fraser (FF) was performed and only then plated in a selective medium (COMPASS). The other confirmations were performed plating directly in the respective selective medium. COMPASS or PALCOM for L. monocytogenes, XLD or CHROMagar™ Salmonella Plus for Salmonella spp. and CHROMagar™ E. coli O157 Table 3.12. Optimized condition for the selected methodology Pathogen detected Type of samples Samples pre-treatment DNA amplification Confirmation Approach Enrichment conditions DNA extraction Approach Genetic targets qPCR Culture L. monocytogenes E. coli O157 Salmonella spp. Milk (UHT, Fresh, Raw) Short enrichment TBS (7 h) Enzymatic lysis LAMP colorimetric plcA, rfbE, invA ProbeqPCR Plating in selective media Confirmations were performed plating directly the enrichment in the respective selective medium. COMPASS for L. monocytogenes, CHROMagar™ Salmonella Plus. and CHROMagar™ E. coli O157 SARAH AZINHEIRO 66 3.9.2 LoD determination To determine the minimum concentration of targeted bacteria providing a positive result in 25 g of food sample, two different approaches were followed. In the first, the LoD was identified as the lowest, spiked bacterial concentration which could be reliably detected. To determine this parameter 10 samples were inoculated with less than 10 cfu of the targeted bacteria, and after analysis (enrichment, IMS, DNA extraction and qPCR) at least 9 had to be positive (90%). The other approach to determine the LoD was performed analysing several samples inoculated with different contamination levels until obtained negative results. The LoD50 and LoD95 were calculated using PoDLoD calculation program, version 9, which also allowed to predict the Probability of Detection (PoD). This approach provided more robust results. Extra, non-inoculated samples were also analysed to assure absence of the pathogen in the original matrix. 3.9.3 Fitness for purpose To assess if the different methodologies developed were suitable to be used in the food industry and allowed reliable results, with high sensitivity, different parameters were calculated following the criteria of ISO 16140-2003 [147] and NordVal International Protocol [148]. The samples analysed were classified as positive and negative agreement (PA/ NA), and Positive and Negative deviation (PD/ ND) comparing the result obtained after analysis, with the expected results or a reference methodology. ND are the number of samples expected positive with a negative result, and PD, are the number of samples expected negative with a positive result. In the NordVal regulation, three other concepts are identified when a confirmation to the reference method is applied. ND are False Negative (FN) when the reference method is confirmed positive. PD becomes True Positive (TP) when the alternative method is confirmed positive, and False Positive (FP) when the confirmation was negative Using these data, the relative sensitivity, specificity, accuracy (SE/ SP/ AC) and the Cohen’s kappa, or , were calculated according to ISO and Nordval criteria, and the positive and negative predictive values (PPV/ NPV) following previous studies [149,150]. SE was defined as the percentage of positive samples giving a correct positive result. 𝑆𝐸 =𝑃𝐴 (𝑃𝐴+𝑁𝐷) ×100 (ISO) Eq. 6 𝑆𝐸 =𝑃𝐴+𝑇𝑃 (𝑃𝐴+𝐹𝑁) ×100 (NordVal) Eq. 7 SP was defined as the percentage of negative samples giving a correct negative result. 𝑆𝑃 =𝑁𝐴 (𝑃𝐷+𝑁𝐴) ×100 (ISO) Eq. 8 Chapter 3. Methodology 67 𝑆𝑃 =𝑁𝐴 (𝐹𝑃+𝑁𝐴) ×100 (NordVal) Eq. 9 AC is defined as the degree of correspondence between the response obtained by the expected result and the method on identical samples. P0 𝐴𝐶 =(𝑃𝐴+𝑁𝐴) 𝑁 ×100 (ISO) Eq. 10 𝑝0=𝐴𝐶 =(𝑃𝐴+𝑁𝐴 +𝑇𝑃) 𝑁 ×100 (NordVal) Eq. 11 N= total number of samples analysed. PPV and NPV are measures of the performance of the method by giving the probability of a sample being really positive or negative when the method shows a positive or negative result. 𝑃𝑃𝑉 = 𝑃𝐴 (𝑃𝐴+𝑃𝐷) ×100 Eq. 12 𝑁𝑃𝑉 = 𝑁𝐴 (𝑁𝐴+𝑁𝐷) ×100 Eq. 13 Finally, the index kappa (  ) of concordance shows the degree of concordance between the method and the expected result. The following equation refers to the calculation of the  by ISO standard:  = 2 x (PA x NA) − (ND x PD) (PA + PD) x (PD + NA) + (PA + ND) x (ND + NA) Eq. 14 To calculate the  following the NordVal regulation another parameter need to be calculated, the expected frequency of agreement, the expected accuracy, or repeatability by chance (𝑝𝑒). The  was then obtained using the 𝑝𝑒 and the previously calculated accuracy (𝑝0) 𝑝𝑒= (PA x FN) × (TP+NA+FP)+(PA+TP)×(FN+NA+FP) 𝑁2 Eq. 15  = 𝑝0− 𝑝𝑒 1−𝑝𝑒 Eq. 16 When the kappa value is between 0.81 and 1.00 the results are interpreted as “nearly complete concordance”, showing a very good performance of the methodology [148]. For the new methodologies to fit the purpose a result between this values is required. SARAH AZINHEIRO 74 monocytogenes is observed as the most challenging from the bacteria targeted, with a slower growth rate and for this reason the attention was focus in the improvement of the enrichment step for this pathogen, evaluated in simplex and in co-culture. 4.2.1 L. monocytogenes growth in simplex enrichment A first evaluation of the effect of different media in the growth of L. monocytogenes was accomplished to understand which could be the best option to include as an alternative enrichment medium in the final methodology to be developed. For this purpose, the growth of the pathogen was followed in a microplate reader during 24 h, measuring the absorbance at 600 nm each 30 min, and kinetic curves were obtained plotting the OD versus the time of measurement. The results are presented in Figure 4.1. After the data obtained and modelled by the respective equation, the main parameters were extracted and are summarized in Table 4.1. The experimental data showed an excellent correlation, with adjusted correlation coefficients (adjusted R2) higher than 0.99. Table 4.1. Evaluation of the L. monocytogenes kinetic growth in different medium formulations. OD600 max µmax λ 30 ºC HF 2.334 ± 0.036 0.900 ± 0.061 16.3 ± 0.3 LEB - - - BLEB - - - ONE 2.708 ± 0.030 0.636 ± 0.012 14.3 ± 0.2 30 ºC mTA10 0.288 ± 0.004 0.074 ± 0.004 14.2 ± 0.2 mTA10 MOPS 0.261 ± 0.013 0.078 ± 0.003 14.0 ± 0.2 mTA10 MOPS+Glu 0.328 ± 0.010 0.088± 0.001 14.8 ± 0.1 35 ºC mTA10 0.242 ± 0.003 0.089 ± 0.017 12.4 ± 0.1 mTA10 MOPS 0.202 ± 0.006 0.095 ± 0.003 13.1 ± 0.2 mTA10 MOPS+Glu 0.292 ± 0.002 0.085 ± 0.004 13.4 ± 0.1 37 ºC mTA10 MOPS 0.468 ±0.033 0.115 ± 0.033 11.0 ± 1.0 0.25g/ L 0.693 ± 0.046 0.243 ± 0.021 12.0 ± 0.8 5g/ L 0.382 ± 0.033 0.135 ± 0.008 11.3 ± 0.1 20g/ L 0.990 ± 0.003 0.120 ± 0.009 11.7± 0.2 37 ºC TSB 0.944 ± 0.025 0.215 ± 0.006 9.9 ± 0.2 TSB+YE 1.085 ± 0.040 0.290 ± 0.020 12.4 ± 0.2 TSB+YE+SP 1.082 ± 0.020 0.266 ± 0.034 11.7 ± 0.6 BHI 1.006 ± 0.028 0.171 ± 0.006 11.2 ± 0.2 BHI+YE 1.040 ± 0.016 0.189 ± 0.012 11.4 ± 0.3 BHI+YE+SP 1.100 ± 0.018 0.193 ± 0.007 11.3 ± 0.3 OD600 max correspond to the maximum optical density; µmax represent the maximum specific growth rate; λ the lag time in hours. Results were given by the model. Chapter 4. Results – Sample pre-treatment alternatives 75 Figure 4.1. Evaluation of the growth of L. monocytogenes within 24 h in different selective and general media. (A) Selective media (HF, LEB, BLEB, ONE) at 30 °C. (B-D) Variation of mTA10 medium. MOPS buffering and 5 g/L glucose addition at 30 °C (B) and 35 °C (C) and supplementation with cellobiose (0.25, 5, 20 g/ L) at 37 °C (D). (E) TSB and BHI, with the addition of yeast extract (6 g/ L) and sodium pyruvate (1 g/ L) were also evaluated at 37 °C. (F) Comparison of the best media evaluated. SARAH AZINHEIRO 76 The standard protocols, from both ISO and FDA regulation authorities, specify several selective media for the enrichment of samples with the objective to identifying L. monocytogenes. These media allows optimal conditions for the specific bacteria to grow and, most importantly, inhibits the development of interfering microorganisms that may delay the growth of the target microorganism. With this in mind, several selective media, normally used in the reference protocols, were selected and tested, between them HF, LEB and BLEB, including also a commercial selective medium, ONE broth, with the same objective (Figure 4.1 A). LEB and BLEB showed an important delay in growth, and the model was not capable to calculate the parameters. However, the BLEB medium seemed to allow a lower lag phase than LEB, as can be observed in Figure 4.1 A. Comparing the two other media, ONE broth showed a shorter lag phage (λ = 14.3 ± 0.2) and also showed a higher maximum optical density (OD600 max = 2.708 ± 0.03) compared to HF (λ = 16.3 ± 0.3; OD600 max = 2.334 ± 0.036). The selective media were tested at an incubation temperature of 30 °C, as recommended by suppliers and the regulation. ONE broth showed better performance not only to achieve higher L. monocytogenes concentration, but also to inhibit the growth of naturally present microorganisms as already reported [151]. The use of selective media has its advantage when complex samples are analysed, however may not always be the best choice, not even when the aim is to target a single microorganism. In samples where stressed or injured bacteria may be present, the selective media can interfere with the recovering process of such cells [152], being a preenrichment in non-selective medium an advantage in these situations. Additionally, in some cases, some competitive microorganisms can present a faster growth in selective media than L. monocytogenes, as reported for L. innocua [153,154]. Finally, if a multiplex enrichment is the aim of the methodology, selective medium will not be able to allow the best growth of the three targets and for this reason is not compatible with all approaches. The TA10, commercial name of the original No. 17 broth, was already improved twice, first by Omiccioli et al., [155] removing the dextrose, and later then adapted by Garrido et al., [156] by changing the amount of buffering salts used, increasing the final pH of the broth. However, when used in high concentration, phosphate salts were reported to have a toxic effect in bacteria [157]. Previous studies have described the use of MOPS to recover stressed L. monocytogenes cells and can be use in higher concentration to stabilized pH during enrichment, the substitution of the KH2PO4 and Na2HPO4 by MOPS was evaluated [157,158]. No statistically significant differences (p >0.05) in lag phase, maximum OD600, growth rate between the use of the two different buffers were observed (Figure 4.1 B and C). The addition of 0.5 g/ L of glucose was evaluated in the new broth containing MOPS and an increase in the maximum concentration of bacteria was achieved obtaining a value of 0.328 ± 0.010 and 0.292 ± 0.002 when growth performed at 30 ºC and 35 ºC respectively. However, regarding the other parameters, µmax and λ no statistical differences were observed when glucose was added. Three different temperatures were also analysed for the different formulation, and the results showed that even though a 1h reduction was observed between the enrichment at 35 °C compared to 30 °C, no statistical difference were observed in all parameters evaluated (Table 4.1). Additional the mTA10 MOPS was tested with the incubation performed a 37 °C and the OD 600 presented significantly higher value (0.468 ±0.033) when compared with 30 °C (0.288 ± 0.004) and 35°C (0.242 ± 0.003). With this results in mind, modifications to this medium were tested with the higher temperature (37 °C). Different concentration of cellobiose (0.25, 5, 20 g/ L) were tested in mTA10 with MOPS (Figure 4.1 D) and the data obtained from the growth studies, are in agreement with the results previously reported, showing a significantly higher growth rate when this carbohydrate is added to the medium [108,109], even when lower concentrations (0.25 g/ L) of cellobiose were used. The addition of 20 g/ L showed the same growth rate and lag phase as the control without cellobiose, however during the Chapter 4. Results – Sample pre-treatment alternatives 77 stationary phase the bacteria concentration did not stabilized as observed in the other kinetics and a slower growth continued, allowing to have the highest concentration after 24h, comparing with all other concentrations of cellobiose tested. The results can possibly be explained, by the capacity of L. monocytogenes to use cellobiose as an alternative carbon source [159,160], but continue to have other preferential compounds, more efficiently metabolized as glucose-6-phosphate [159,161], being cellobiose the choice when the other sources are exhausted. By this, the growth of the bacteria can be extended by the metabolization of cellobiose in a later stage. Two additional general media, TSB and BHI, were analysed by the same methodology in their original formulation and supplemented with yeast extract (YE) and sodium pyruvate (SP). The results are represented in Figure 4.1 E. The shortest lag phase was obtained with native TSB (9.9 ± 0.2). The supplementation of the medium with yeast extract (YE) and/ or sodium pyruvate (SP) obtained results similar to those of mTA10. No significant effect was observed on the other parameters evaluated. Regarding BHI, all variations obtained a similar result in all parameters, between 1-1.1, 0.17-0.19 and 11.2-11.4 for the OD600 max, µmax, and λ, respectively. For the detection of L. monocytogenes in a methodology where the enrichment is performed in 24 h, the best option may pass by the use of mTA10 broth supplemented with 0.25 g/ L or 20 g/ L of cellobiose, as both are able to improve the growth of this pathogens and may allow to recover stressed cells (Figure 4.1 F). This medium can also gave some advantage regarding the competition existing with other target and interfering bacteria presented in the sample. On the other hand, since the objective was to reduce the enrichment time to have a same day detection, the medium needed to reduce the lag phase, as the bacteria will never reach the stationary phase, being the final concentration of bacteria after 24 h not relevant for this approach. For this end, TSB is the medium which gave the best results and have the ability to achieve higher L. monocytogenes concentration in less time Figure 4.1 F. 4.2.2 Growth of the three targets in co-culture After studying the growth of L. monocytogenes in simplex to identify which medium could provide the best conditions to enhance the final concentration obtained, different variations of mTA10 MOPS were tested for the growth in co-culture of L. monocytogenes, Salmonella spp. and E. coli O157. Laked Horse Blood (LHB), Campylobacter Growth Supplement and Half Fraser Selective Supplement were chosen to evaluate their effect alone and combined in the growth of the three pathogens (Figure 4.2). The results presented in Figure 4.2 B show that LHB and Campylobacter Supplement combined improve the growth of Salmonella spp., but not in a statistically significant manner, the plate counts were higher with a small standard deviation. No major effects, promoting or inhibiting, were observed on the plate counts of E. coli O157 Figure 4.2 C. However, none of them were able to improve the growth of L. monocytogenes, actually had a detrimental effect even though it was not statistically significant. Figure 4.2 A shows lower concentration of this bacterium after 24 h of incubation in mTA10MOPS supplemented with the different compounds in co-culture. LHB seems to originate the lower concentrations, and its combination with any of the other supplements increase this effect, being the plate counts for these lower than 6 log cfu/ mL. Having in consideration the results obtained by the kinetic analysis, mTA10 modified with MOPS buffer and 0.25 g/ L cellobiose was similarly evaluated to enhance the growth of L. monocytogenes in co-culture for a multiplex enrichment in 24 h, the results are presented in Figure 4.2 D. A statistically significant difference was observed in the growth of L. monocytogenes, increasing 2.3 log cfu/ mL the SARAH AZINHEIRO 78 final concentration obtained. The results are in concordance with the kinetic assay, confirming the improvement in the growth of L. monocytogenes when cellobiose was added to the medium. The final concentration of Salmonella and E. coli O157 did not show any difference with the implementation of cellobiose, which benefit the detection of L. monocytogenes, improving its competiveness. Chapter 4. Results – Sample pre-treatment alternatives 79 Figure 4.2. Optimization of mTA10 MOPS medium. (A-C) Plating results of mix cultures of L. monocytogenes, S. Typhimurium and E. coli O157 after 24 h of incubation in mTA10 MOPS with the different supplements: 4 mL/L Campylobacter Growth Supplement (Campylobacter suppl.) (Oxoid, UK), 225 µL/225 mL Half Fraser Supplement (Selective suppl.) (Biokar Diagnostics S.A., France), 50 mL/L Laked Horse Blood (LHB) (Oxoid, UK), and as well the combinations of these compounds simultaneously. (D) Co-culture results of the supplementation with 0.25 g/ L of cellobiose for the three pathogens. SARAH AZINHEIRO 80 4.3 CONCENTRATION OF BACTERIA 4.3.1 Immunomagnetic separation (IMS) The first step performed to develop the IMS methodology was the evaluation of different commercial antibodies in terms of purity and specificity in order to choose the best option to functionalize the magnetic beads to concentrate L. monocytogenes. 4.3.1.1 Commercial antibodies evaluation To screen the purity of the antibodies SDS-PAGE was performed and the resulting gel is presented in Figure 4.3. The results shows that all antibodies presented two similar main bands, one about 20–25 kDa and another one at 55 kDa, corresponding to the light and heavy chains of the Abs. No additional bands are present in the two monoclonal antibodies (MA1-20271 and MAB8953), however the two polyclonal antibodies contain other contaminant proteins showing a band with a molecular weight around 66 kDa, which can correspond to BSA. Another protein is present in the goat pAb (MD05-0329) with higher molecular weight between 70–100 kDa, which could also represent some intact antibody, due to incomplete reduction by Laemmli buffer. Figure 4.3. SDS-PAGE results obtained for the evaluation of the purity of the evaluated Ab. Three concentrations of each Ab were loaded: 100, 10 and 1 µg/ mL Chapter 4. Results – Sample pre-treatment alternatives 81 To evaluate the specificity of the antibodies to bind to L. monocytogenes, an indirect ELISA was performed with other microorganisms, including S. Typhimurium and L. innocua. As shown in Figure 4.4, cross-reactivity against the non-specific targets was exhibited by both polyclonal antibodies. The monoclonal antibody MA1-20271, in addition to cross-reactivity resulted in a very low signal for L. monocytogenes. The only antibody which presented acceptable results in terms of specificity and signal intensity was Ab MAB8953, with a high signal for the target pathogen and only a residual cross-reactivity for the other. If cross-reactivity with other bacteria is present the recovering of the target can be compromised, as they will compete for the antibody binding, reducing the capture efficiency, and consequently the limit of detection of the methods being tested [162]. As expected, all the antibodies tested were relatively pure, but exhibited great differences in terms of specificity. The mAb MAB8953 showed the highest specific for L. monocytogenes, and thus selected for nanoparticle functionalization. Figure 4.4. Specificity of the antibodies tested for IMS. PA1-7230 and MD-05-0329 are polyclonal antibodies, while MA120271 and MAB8953 are monoclonal. Signal intensity is presented in Relative Luminescence Units (RLU) SARAH AZINHEIRO 82 4.3.1.2 Capture efficiency The optimization of the IMS protocol was made evaluating the capture efficiency (CE) of different alternatives to the standard protocol recommender by the MNPs supplier. The quantification results obtained by qPCR are presented in Figure 4.5, expressed as log cfu/ mL, and the CE was calculated based on these results. Performing Three bead washing steps, prior to further analysis, provided lower CE values than only one washing step, The direct analysis of the MNPs provides a CE of 95 % (Cb = 3.8; C0 = 4.0), while extensive washing, as indicated in the MNPs protocol, reduces this value to 75 % (Cb = 3.6; C0 = 4.8), which will compromise the sensitivity of the methodology. For this reason, no extensive MNPs washing was included in the final protocol. The influence of the washing step in the capture of the target bacteria was also reported by other authors [163,164], showing a decrease of sensitivity when too many washing steps were performed. The use of IMS reported in other studies also obtained similarly high CE values [165]. Figure 4.5. Quantification performed by qPCR for the determination of the CE. In green are represented the results for the standard IMS protocol, including three washing steps with PBS (dark green for the direct quantification and light green after the MNPs treatment). The blue points indicate the optimized IMS, with direct separation from sample enrichment, without washing (dark blue for the direct quantification and light blue after the MNPs treatment) 4.3.1.3 Evaluation with food matrixes The optimized protocol was tested and evaluated in spiked samples from different types of food product and the results obtained by qPCR analysis are presented in Table 4.2. From this total of 42 samples, 10 samples of fresh cheese samples, were spiked with a lower bacterial concentration (9.7 cfu/ 25 g) where the LoD was established, as describe in Chapter 3, section 3.9.2. Additionally it was possible Chapter 4. Results – Sample pre-treatment alternatives 83 to detect the presence of L. monocytogenes in samples with a lower concentration ranging between 4.6 to 7.1 cfu/ 25 g, showing the ability of the methodology to reach higher sensitivity. All the remaining samples inoculated with higher concentration obtained 100 % positive results and the remaining noninoculated samples were negative, presenting a k index of 1.0. Table 4.2. Samples analysed by IMS Type of food N Inoculation level (cfu/ 25 g) IMSqPCR Plate confirmation Milk 2 - - - 1 < 10 + + 1 10-102 + + 1 102 - 103 + + Hard cheese 1 - - - 4 < 10 + + 1 10-102 + + 2 102 - 103 + + 1 > 103 + + Anchovy 3 - - - 1 < 10 + + 1 10-102 + + Chicken 4* 10-102 + + 2* < 10 + + 2 < 10 + + 2 10-102 + + 2 102 - 103 + + Fresh cheese (LoD) 10 < 10 + + 1 - - - * These samples were not originally inoculated, but were naturally contaminated with L. monocytogenes Over this study, it was observed that two batches of chicken samples, which included 6 non-inoculated samples, presented a positive result. Further analysis of the enriched samples following the ISO 11290 method confirmed that these samples were naturally contaminated with L. monocytogenes. These samples were also analysed according to the 11290-2:2017, in triplicate to determine the concentration of L. monocytogenes, which was determined to be lower than 10 cfu/ 25 g for one batch, and between 10 and 100 cfu/ 25 g for the other. With total concordance between the results obtained by IMS-qPCR methodology and the culture-based confirmation, the performance parameter resulted on a value of 100 % and a kappa index of 1.00, which demonstrate the reliability of the developed methodology. Comparable results were obtained in other studies using IMS combined with a DNA amplification methods, not only PCR/ qPCR based methods but also isothermal approaches [135,163,164]. SARAH AZINHEIRO 90 concentration no positive result was observed, confirming the detection of only viable Salmonella cells. With the analysis of samples spiked with viable bacteria, only one ND was observed in samples with a concentration close to the LoD (9 cfu/ 25 g), resulting in a SE, SP and AC of 96.6 %, 100 % and 97.6 % respectively. The PPV and NPV obtained values of 100 % and 92.3 % respectively, representing results in high degree of concordance with the expected result, obtaining a  of 0.94 (Table 4.7). Table 4.6. Samples analysed by PAA methodology Spiking N Inoculation level (cfu/ 25g) PAAqPCR Live Bacteria 10* 8 + 3 5/10/0 +/+/- 5 5.2/9.8/9.4/8.6**×10 + 5 5.2/9.8/9.4/8.6×102 + 2 9.8×103 + 2 9.8×104 + 2 9.8×105 + 6 - - Non-viable Bacteria 2 8.0×107 - 2 8.0×105 - 2 8.0×103 - * Samples used to determine the LoD; ** Two samples were spiked with this concentration. Table 4.7. Evaluation of PAA methodology N PA PD NA ND SE SP AC PPV NPV κ LoD (cfu/ 25 g) 41 28 0 12 1 96.6 100 97.6 100 92.3 0.94 8 ND spiked with 9 cfu/ 25g; Chicken breast samples were used for the evaluation. A LoD of 8 cfu/ 25 g was obtained, resulting from the analysing of 10 samples spiked with a concentration <10 cfu/ 25 g, but lower concentration, down to 5 cfu/ 25 g could also be detected. In conclusion, the developed method enables the detection of S. Enteritidis cells with high sensitivity in a total analysis time of 10 h, and has the capacity to specifically detect viable S. Enteritidis. This represents a significant time reduction with respect to the standard culture-based, and other molecular biology-based methods [142,167,177], which need between 20 to 48 h to have the results. Chapter 4. Results – Sample pre-treatment alternatives 91 Due to the high confidence of the results obtained, this method can be suitable for the implementation on routine laboratories. Further optimization can be done to allow a faster analysis, as the selection of bacteriophage with shorter latent periods, or for a multiplex detection a phage cocktail can be developed [178]. 4.4.2 Matrix lysis The matrix lysis approach was developed by Rossmanith et al., [75], and optimized later [76]. Due to quantity of food debris recovered by this protocol, the authors were only capable to perform a reliable analysis until 12.5 g or mL of sample. The European regulation advise the sampling of 25 g from the foodstuff [179], and for this reason, to be able to process a higher sample size, the original protocol was modified. Instead of taking all the sample, after homogenisation with the sucrose buffer, the mixture was recovered through the filter of a stomacher bag with a pore size of <250 µm, working as a barrier for most of the larger food debris, reducing the pellet obtained. This step was introduced as it was observed that the protease treatment, and lysis buffer, were not enough to reduce the pellet recovered to a size suitable for direct DNA extraction. The modified methodology tested for 25 g of sample provided a LoD95 of 1.1 x 105 cfu/ 25g, and for the 16 samples above this concentration, the analysis performed correctly, with no deviation observed and provided a κ of 1.0 (Table 4.8). From 14 samples spiked below the LoD, it was observed that 7 of them gave a positive result even at a concentration of 8 x 103 cfu/ 25 g. However, the original protocol obtained a lower LoD of 7.3 cfu/ mL for the analysis in milk samples [75], but also had similar values in other types of liquid samples (7.8 cfu/ mL) when analysed 12.5 mL. It was observed a slight increase when 6.25 g of solid foodstuff was process (15.4 cfu/ g) [76], showing the influence of the type of matrix analysed. The increase of the samples from 6.25 g to 25 g could be the reason of the difference in the LoD, as the proportion of matrix recovered compared with the bacteria is much higher and can interfere with the qPCR reaction. Other studies have shown the application of this approach for different purposes, such as the recovery of Mycobacteria from animal tissues [180], detection of Staphylococcus aureus in milk [78] and cheese [181], among others. This methodology still needs further optimization to reach a lower LoD, down to 100 cfu/ g, to be apply in the food industry for the detection of L. monocytogenes, in foods which do not support its growth. Table 4.8. Evaluation of Matrix lysis approach N PA PD NA ND SE SP AC PPV NPV  LOD95 (cfu/ 25 g) 16 13 0 3 0 100 100 100 100 100 1.00 1.1 x 105 Salmon samples were used to evaluate the methodology. SARAH AZINHEIRO 92 4.4.3 Short enrichment 4.4.3.1 Protocol optimization After the results obtained by the matrix lysis methodology, showing not being adequate for the detection of L. monocytogenes in food samples due to the high LoD value obtained, an alternative approach was tested based on the protocol developed by Fachmann et al. [80] for the detection of Salmonella in fresh meat. In their study a short enrichment associated with a matrix degradation was performed. To apply this approach for the detection of L. monocytogenes different aspects of the protocol were optimized. Fist the volume of enrichment medium to be added to the sample was tested, and no significant difference was observed between the use of 25 mL and 50 mL of TSB to growth L. monocytogenes in simplex, with a variation of 1.08 ± 0.12 and 1.07 ± 0.26 log cfu/ g, respectively, however 25 mL showed a smaller deviation between replicates (Figure 4.9). The final protocol was for this reason performed with 25 mL, which also make the analysis less expensive and the smaller volume simplify the following step. The need to use agitation during the enrichment was evaluated in the same way, resulting in an statistically significant increase of bacteria growth when constant shaking at 200 rpm is used (variation of 1.40 ± 0.02 log cfu/ g), compared with static incubation (variation of 1.02 ± 0.16 log cfu/ g) (Figure 4.9). Even if this optimization was performed with an enrichment time of 4 h, the results obtained by qPCR when the samples were analysed in these conditions, were not consistent, not allowing to achieve a proper LoD. The enrichment required a longer incubation time to provide reliable results, therefore the following sample analysis was accomplish with the optimized protocol, where 25 mL of TSB were added to the sample and incubated with constant shacking at 37 ºC for 5 h. Figure 4.9. Evaluation of different conditions for short enrichment optimization. Bacterial growth variation between T0 and T4 (4 h incubation). * indicates statistically significant differences Chapter 4. Results – Sample pre-treatment alternatives 93 4.4.3.2 Evaluation with food matrixes The methodology was tested for the detection of L. monocytogenes and a similar approach was also tested for E. coli O157 and the multiplex detection of E. coli O157 and Salmonella spp. with the enrichment condition specified in Table 3.4 of Chapter 3, section 3.4.3.3. The results obtained for the analysis of different samples is presented in Table 4.9, including the parameters to evaluate the fitness for purpose, but also the LoD achieved for each one of them. For all pathogens, it was possible to detect a concentration of targeted bacteria lower than 10 cfu/ 25 g, being 8.6 cfu/ 25 g for L. monocytogenes in simplex and between 3 - 4 cfu/ 25 g for the E. coli O157 in simplex and multiplex detection of E. coli O157 and Salmonella spp.. Comparing with the previous study in which the method was based, Salmonella and E. coli O157 obtained a lower value, as the authors observed a LoD50 of 8.8 cfu/ 25 g. However L. monocytogenes showed higher values, which could be related with the target itself as L. monocytogenes has a slower growth rate and the DNA extraction of this pathogen present more difficulties than for the other targets. Another cause for this higher value could the type of matrix analysed. The sample type could make the recovery of the bacteria difficult, due to several factors. The size of the debris generated and the fat content could block the filter of the stomacher bag not allowing the passage of all bacterial cells, as may also interfere with the washing steps and protease treatment [80]. Table 4.9. Short enrichment evaluation and comparison Bacteria Food sample N PA PD NA ND SE SP AC PPV NPV  LoD95 (cfu/ 25 g) L. monocytogenes Salmon 17 12 0 5 0 100 100 100 100 100 1.00 8.6 E. coli O157 Ground beef 19 14 0 4 1 93 100 95 100 80 0.85 3.9 Leafy green 20 12 0 8 0 100 100 100 100 100 1.00 3.3 Combined 39 26 0 12 1 96 100 97 100 92 0.94 3.6 Salmonella spp. E. coli O157 (Multiplex)* Ground beef and Chicken breast 39 30 0 8 1 97 100 97 100 89 0.92 3.4 39 30 0 8 1 97 100 97 100 89 0.92 3.7 78 60 0 16 2 97 100 97 100 89 0.92 3.6 The “combined” results indicate the values obtained for the all methodology considering both type of samples (ground beef and leafy green) tested; * Results presented are for the detection of Salmonella spp. E. coli, and combined results in that order. On the other hand, some specific components may not be washed in the pre-treatment and inhibit the qPCR reaction [182], leading to a False negative result, and compromising the reliability of the method. The leafy green samples demonstrate this effect, as the use of a specific plant kit (NucleoSpin® Plant, Macherey-Nagel, Germany) to purify these samples after the DNA extraction was required, and the DNA extract obtained needed to be additionally diluted ½ to obtain positive results. Several studies already reported the inhibition of qPCR by plant component [183,184], and the need to adequate the protocol to overcome this effect. Another aspect that can interfere with the detection of the targeted pathogen is the presence of interfering microorganism in high levels in certain types of samples [151] With a  index of the method for the detection of different pathogens, with values ranging between 0.85 and 1.00, the short enrichment methodology show a real advantage to be used in the food industry, allowing a same day detection with an enrichment of 3 or 5 hour depending on the pathogen to be SARAH AZINHEIRO 94 detected. An important fact to be mentioned is the relevance of proper defrost of the samples before the enrichment step and to add the medium pre-warmed at 37 ºC, to allow the appropriate growth of the bacteria. A ND was observed in the simplex E. coli O157 detection. It was related with a sample where the ON defrosting was not properly performed. This could lead to decrease the temperature of the starting enrichment, delaying the growth of the bacteria that was already spiked in a low concentration (5.1 cfu/ 25 g). When the methodology was evaluated for the multiplex detection of E. coli O157 and Salmonella spp, the recovery and detection of stressed bacteria were also tested. Before spiking the samples, both bacteria went through a heat treatment at 60 ºC for 10 min with constant agitation (1000 rpm) and diluted in PBS to contaminate 8 samples at a final concentration lower than 10 cfu/ 25 g. Half of these samples were stored at 4-8º C during 24 h and the other half during 48 h and then the protocol for the short enrichment was applied. Regarding Salmonella, the detection of 3 of 4 spiked samples was possible, regardless of the treatment. The same results were observed for E. coli O157 when the samples were refrigerated for 24 h. However when this period is extended to 48 h only 2 of the 4 samples gave a positive result. The influence of non-viable pathogens, was also determined, in order to understand if the methodology could allow to specifically detect viable bacteria. For this evaluation, the approach for the multiplex detection was applied in 8 samples where different concentration, ranging from103 cfu/ 25 g to 106 cfu/ 25 g of non-viable bacteria were spiked from both targets. The LoD was calculated, and showed that it was necessary at least 3.9 x 105 cfu/ 25 g and 2.9 x 105 cfu/ 25 g of dead Salmonella spp. and E. coli O157 respectively, to originate false positives associated with the presence of DNA from the non-viable pathogen cells. This methodology showed promising results, allowing to perform a same-day detection, and with further optimization could be applied to different food commodities. The type of matrix will be the principal factor that influence the performance of the method, and variation in the protocol may be needed for the methodology to fit in a wide range of foodstuffs. The pore size of the stomacher bags, the homogenisation of the sample with the enrichment medium (manual or automatic), and the protease used for the degradation of food debris, are some of the conditions to be improved for better recovering of the bacteria and a cleaner DNA extract [80]. 4.5 COMPARISON OF SAMPLE PRE-TREATMENT APPROACHES After the analysis of the four different approaches to improve the sample pre-treatment (Table 4.10), all of them presented some advantages when compared with the traditional methodologies based on culture techniques. Even if the IMS did not present a significant difference in terms of time of the analysis comparing with different qPCR approach developed with the same objective [121,167,185], the concentration of the target bacteria allowed to separate them from the rest of the sample and enrichment media, removing most inhibitors present in this mixture. With the analysis of different type of foodstuff, the presented IMS, demonstrated to be suitable for the analysis of a wide variety of matrix, as milk, cheese, anchovies and chicken. Chapter 4. Results – Sample pre-treatment alternatives 95 Table 4.10. Comparison of the different approaches Methodology Analysis time LoD (cfu/ 25g) Cost Advantage Disadvantage IMS 26 h (L. monocytogenes) 9.7 +8€ Elimination of inhibitors Still need long enrichment PDMS Sponge > 5h a (L. monocytogenes Salmonella spp.) - b * Elimination of inhibitors Automatize analysis Not suitable for food with large debris PAA 10h (S. Enteritidis) 8 -** Viable bacteria differentiation Need to growth bacteria before Matrix lysis < 5 h (L. monocytogenes) 1.1 x 105 + 4€ More bacteria recovered Inhibitors remaining Short enrichment 5h (Salmonella and E. coli O157) 7h (L. monocytogenes) 3-4 8.6 + 2€ The cost column refers to the approximation in the cost increase per sample, performing the respective pretreatment methodology, without enrichment or DNA extraction steps. Only including reagents and specific compounds required; a The analysis time do not include enrichment step b The LoD was not determined in complex samples. For surface samples without enrichment a bacteria concentration of 105 was detected. * The cost will depend in the fabrication price of the device to allocate the sponge; ** No additional cost, only the purchase of the bacteriophage at the beginning that can be replicate posteriorly. In a very similar way, the use of PDMS sponge to concentrate the bacteria allowed to wash the sample recovered, from PCR inhibitors. In this case, the approach was evaluate for detection of pathogens on surface samples, which many times contain cleaning agents, known to interfere with PCR reaction, leading to false negative results [186]. The multiplex detection by IMS was already reported [170,187] but required the use of a specific antibodies for each target pathogen, making the analysis much more expensive. Universal ligands as ApoH protein have the advantage to bind to any bacteria, virus or fungi allowing an easier and cheaper multiplex recovery, however its use is incompatible with most complex matrix with a high concentration of background microorganisms, being suitable only for samples subjected to sterilization process as pasteurization or more intense heat treatment [188], which have a very low initial natural microorganism content. These two methodologies are the unique from the approaches tested, which allowed the purification from inhibitors compounds. Replacing the washing step normally performed before DNA extraction by the concentration, we also simplified the process and reduced the sample manipulation, improving the sensitivity of the analysis. The automatization of the concentration process was demonstrated by the integration onto a micro-device, which simplified and reduced the hands-on time for the sample pretreatment. The use of specific bacteriophages to improve the sensitivity of the analysis present some advantages, such as the specific detection of viable pathogens in 10 h. The need for the differentiation between viable and non-viable bacteria is still controversial in the food industry. Most of the companies will want to know if the bacteria are still alive to determine the safeness of their product. However if the DNA of the pathogen was found in the product, it means that the pathogen was present at some point, and maybe the sample taken does not have any viable pathogens, but it does not assure that the rest of the production line is safe [189]. Some studies have combined the use IMS to recover the target bacteria from the enrichment medium, and allow removal of inhibitors, with the bacteriophages infection, to increase the signal [190,191], enhancing a more sensitive and reliable detection. Both approaches, PAA and IMS, are the most complex SARAH AZINHEIRO 96 to be developed. The PAA requires to understand the dynamics between the bacteriophage and its host, and for this both elements need to be well characterized. With the need to find a specific antibody and bacteriophage for each target to be detected and make sure to do not have interference among them. Also the use of these extra components in the protocol makes the analysis more expensive. In terms of time of analysis, the methodologies which present a significant improvement were the matrix lysis and the short enrichment, allowing a same-day detection. This could be a real advantage in the food processing chain taking into account the intense production existing nowadays and the fast delivery of the products. However the matrix lysis was discarded, as the results of the LoD obtained were too high to suite the limit required by the legislation. Regarding the short enrichment approach, a LoD < 10 cfu/ 25g was achieved for all targeted pathogens, showing the possibility of application in the industry, The only limitation of this approach could be the interference of the matrix in qPCR reaction, due to the presence of high levels of inhibitors, which required extra measures in certain types of samples. The combination of the short enrichment with the IMS could be an option to overcome this limitation, however the cost of the analysis will consequently increase, as purchase of specific antibody could be expensive. For these reasons, the short enrichment was chosen as the most adequate option for the development of a faster, and economic methodology to detect in multiplex L. monocytogenes, E. coli O157, and Salmonella spp. The evaluation of a wide range of media were tested, having in attention the reduction of the lag phase, growth rate and maximum concentration of L. monocytogenes achieved, as this was the pathogen with more restrained growth, when compared with E. coli O157 and Salmonella spp.. Even if mTA10 supplemented with cellobiose showed promising results to be used in a 24 h enrichment, the most important factor for the medium to be integrated in the short enrichment protocol, was to begin the exponential phase as soon as possible, and TSB provided a significantly shorter lag phase, around 10 hours, demonstrating to be the most suitable from all media tested for this type of approach. 4.6 CONCLUSIONS After intense analysis of different approaches to improve the specificity, sensitivity, and time of the analysis, all of them demonstrate several advantages regarding the existing, and implemented, methodologies used in the food industry. Standard enrichment  By the analysis of different media variation, as expected general media allow to improve the growth of L. monocytogenes when compared with the selective ones, being TSB the one which was able to have higher reduction in the lag phase  The cellobiose was the only supplementation allowing a significantly improvement in the final concentration of L. monocytogenes obtained, without having any effect in growth of the Salmonella spp. and E. coli O157. This improves the competiveness of this Gram-positive bacteria. Concentration of the bacteria  The IMS allow to concentrate the bacteria from the enrichment medium, achieving a sensitive detection with the advantage of the elimination of possible inhibitors. Chapter 4. Results – Sample pre-treatment alternatives 97  Regarding the functionalized PDMS sponge, with the use of the non-specific ApoH ligand, higher capture efficiency was achieved comparing with the use of specific antibody. Time reduction  The PAA approach using bacteriophage allowed to reduce the time of the analysis to a total of 10 h for the detection of S. Enteritidis.  The analysis by matrix lysis can be faster, but due to the very high limit of detection obtained, this methodology was not consider suitable to be applied in regular testing.  The shot enrichment approach allows a fast same day detection, and achieved a very low limit of detection for the three pathogens targeted. The presented study had as objective the development of a fast, affordable and multiplexed foodborne pathogen detection, and with this in mind, the short enrichment was chosen as appropriate approach for the detection of L. monocytogenes, E. coli O157 and Salmonella spp. The reason for this choice, is the significant time reduction, as the experiments performed for the detection of L. monocytogenes in simplex demonstrated a decrease from 7 days by the ISO method [103,192], or 26 h for alternative qPCR methods [122,193] to just 6-7 h. Another aspect to have in consideration is the simplicity of the protocol, without the need a specific compounds, as antibodies used for the IMS approach and bacteriophages for the PAA, which allow a low-cost but sensitive detection. 4.7 PUBLICATION OF THE RESULTS The results presented in this chapter were published in the following scientific articles: Azinheiro, S., Carvalho, J., Fuciños, P., Pastrana, L., Prado, M., & Garrido-Maestu, A. (2022). Short pre-enrichment and modified matrix lysis. A comparative study towards same-day detection of Listeria monocytogenes. LWT, 154, 112900. https://doi.org/10.1016/J.LWT.2021.112900 Azinheiro, S., Kant, K., Shahbazi, M. A., Garrido-Maestu, A., Prado, M., & Dieguez, L. (2020). A smart microfluidic platform for rapid multiplexed detection of foodborne pathogens. Food Control, 114, 107242. https://doi.org/10.1016/j.foodcont.2020.107242 Garrido-Maestu, A., Azinheiro, S., Carvalho, J., Espiña, B., & Prado, M. (2020). Evaluation and implementation of commercial antibodies for improved nanoparticle-based immunomagnetic separation and real-time PCR for faster detection of Listeria monocytogenes. Journal of Food Science and Technology, 1–9. https://doi.org/10.1007/s13197-020-04450-1 Garrido-Maestu, A., Azinheiro, S., Roumani, F., Carvalho, J., & Prado, M. (2020). Application of Short Pre-enrichment, and Double Chemistry Real-Time PCR, Combining Fluorescent Probes and an Intercalating Dye, for Same-Day Detection and Confirmation of Salmonella spp. and Escherichia coli O157 in Ground Beef and Chicken Samples. Frontiers in Microbiology, 11. https://doi.org/10.3389/fmicb.2020.591041 Garrido-Maestu, A., Fuciños, P., Azinheiro, S., Carvalho, C., Carvalho, J., & Prado, M. (2019). Specific detection of viable Salmonella Enteritidis by phage amplification combined with qPCR (PAA-qPCR) in spiked chicken meat samples. Food Control, 99, 79–83. SARAH AZINHEIRO 98 https://doi.org/10.1016/J.FOODCONT.2018.12.038 Garrido-Maestu, A., Azinheiro, S., Carvalho, J., Fuciños, P., & Prado, M. (2019). Optimized sample treatment, combined with real-time PCR, for same-day detection of E. coli O157 in ground beef and leafy greens. Food Control, 106790. https://doi.org/10.1016/j.foodcont.2019.106790 Chapter 4. Results – Sample pre-treatment alternatives 99 SARAH AZINHEIRO 106 Figure 5.2. Agarose gel presenting the amplicons originated by multiplex SYBR-qPCR. The amplification using as template a DNA extract from a pure culture of L. monocytogenes (Lm), E. coli O157 (O157), S. Typhimurium (Salm). A negative control with water as template (NTC) showed the amplification of the NC-IAC and a positive control (PC) when a mixture of DNA from three pathogens is loaded. The theoretical amplicon size was 89, 142, 105 and 200 bp for hlyA, rfbE, fimA and NC-IAC, respectively, being consistent with the results obtained. 5.2.1.2 Evaluation of the SYBR-qPCR reaction To evaluate the performance of the optimized qPCR method with actA, fimA, rfbE and NC-IAC primers, the inclusivity/exclusivity and efficiency were evaluated. A total of 45 pure cultures from different strains were tested to evaluate the inclusivity/exclusivity of the qPCR and the results are presented in Table 5.1. The inclusivity of the multiplex qPCR was evaluated with, 13 Salmonella spp., 18 L. monocytogenes and 1 E. coli O157, presenting all the strains the expected melting peak. Regarding the exclusivity, 13 other bacteria were evaluated, including two other Listeria species and two E. coli strains. As expected, all non-target microorganism amplified with a Cq values of 35.36 ± 0.81 with a melting peak of 82.78 ± 0.09, specific for NC-IAC amplification. The amplification efficiency of the qPCR was also evaluated, in simplex and multiplex. The lowest DNA concentration detected for L. monocytogenes, Salmonella spp. and E. coli O157 was 1.4, 1.6, 1.9 pg/ µL respectively, see Figure 5.3 A, B and C. Regarding the multiplex detection of the three targets simultaneously, the qPCR reaction showed a LoD 10 times higher of 11 pg/ µL when a mixture of the three DNA extract from each bacteria was loaded (Figure 5.3 D). This result could be due to the fact that the peaks for the other targets are more predominant than the one for E. coli O157, causing some interference in its detection when a lower range of DNA concentration is present Figure 5.4 D. However, it is still possible to detect and identify correctly L. monocytogenes and Salmonella spp. even when present at a 10 times lower DNA concentration. The amplification efficiency was calculated after plotting the standard curves, and the determined values were 98.2 %, 93.2 %, 92.6 % for L. monocytogenes, Salmonella spp. and E. coli O157 respectively; and 91.4 % for the multiplex format. These results are shown in Figure 5.3 and the values are between the previously reported acceptable limits (90–110 %) [35]. Chapter 5. Results – DNA amplification approaches 107 Table 5.1. List of bacteria strain used to evaluate the inclusivity/exclusivity of the primers and SYBR-qPCR Bacteria species Source N fimA actA rfbE IAC Salmonella spp. (AMC 28, 60, 82, 84, 90, 96, 198. 200, 238, 253, 255, 260, 261, UB, WDCM 00031) 15 + - - -* L. monocytogenes WDCM 00021, Mollusk, chestnut, chicken 16 - + - -* L. ivanovii WDCM 00018 1 - - - + L. innocua WDCM 00017, CECT 5376, 4030, 1325, 1141, 2110 6 - - - + S. aureus WDCM 00034, 00033 2 - - - + Staphylococcus coagulase + Proficiency test 1 - - - + C. coli University of Minho 1 - - - + E. coli WDCM 00013, 00012 2 - - - + E. coli O157 WDCM 00014 1 - - + - N: number of strains; IAC amplification allow to prove that any inhibition is present when the absence of amplification is observed in the rest of the targets; * The absence of IAC amplification is due to the amplification of at least one of the targeted bacteria. Figure 5.3. SYBR-qPCR reaction evaluation. The efficiency, dynamic range and coefficient of determination was obtained in simplex for E. coli O157 (A), Salmonella spp. (B) and L. monocytogenes (C) and multiplex (D). Standard for each situation were obtained by three replicates of ten-fold serial dilutions of a DNA extract from each pathogen, and a mixture of the three extracts for multiplex experiment. SARAH AZINHEIRO 108 The influence of the DNA concentration in the melting peak is presented in Figure 5.4, for simplex and multiplex detection as reported above, showing the decrease of the peak intensity with the decrease of DNA concentration in the qPCR reaction. Figure 5.4. Melting analysis from SYBR-qPCR efficiency. In simplex for E. coli O157 (A), Salmonella spp. (B) and L. monocytogenes (C) and multiplex (D) reaction. 5.2.1.3 Evaluation with food matrixes The methodology was evaluated in a panel of 44 samples spiked at different contamination levels. The results are summarized in Table 5.2 for the four targets. All negative samples for one or more pathogens were correctly identified by qPCR, and in the total absence of the targeted pathogen, a Cq higher than 34 was observed, originate by the amplification of the NC-IAC, presenting only its respective melting peak. All positive samples were detected with the correct pathogen identification and none PD were observed, allowing a relative specificity, sensitivity and accuracy, of 100 % and a Cohen’s  of 1. Chapter 5. Results – DNA amplification approaches 109 Table 5.2. Spiked samples to evaluate SYBR-qPCR approach Type of sample Contamination level (cfu/ 25 g of sample)* N SYBR-qPCR L. monocytogenes S. Typhimurium E. coli O157 actA fimA rfbE IAC Infant Milk (LoD) 16.6 7.4 5.5 6 + + +(2ND) - 10.2 5.5 5 6 + + + - 5.1 2.8 2.5 6 + + + - 2.2 1.1 0.8 6 + (2ND) + (3ND) + - 2.3 0.9 0.145 6 + + +(3ND) - Infant Milk - - - 4 - - - + 9 6 4 2 + + + - 9.1 13 8 1 + + + - 9 6 1 + + - - 6 4 1 - + + - 9 4 1 + - + - 91 1.3 x 103 1 + + - - 9 8.0 x 102 1 + - + - 1.3 x 104 8.0 x 103 1 - + + - 9.1 x 103 1.3 x 104 8.0 x 103 1 + + + - N: number of samples; ND: Negative deviation; * The contamination level correspond to concentration of bacteria inoculate before enrichment and was obtained by results of the plating in TSA for S. Typhimurium and E. coli O157 and TSYEA for L. monocytogenes; The ND obtained were all below the LoD, and for this reason were not consider as so for the methodology evaluation. The determination of LoD was done using PoDLoD analysis. A total 30 samples were tested with five different levels of contamination. The LoD50 was calculated to be 0.1, 0.5 and 0.6 cfu/ 25 g and the LoD95 0.6, 2.1 and 2.6 cfu/ 25 g for E. coli O157, Salmonella spp. and L. monocytogenes, respectively. The LoD of the multiplex detection was also evaluated, being the LoD50 0.4 and the LoD95 1.7 cfu/ 25 g. In the current study, the primer concentration for actA had to be increased, with respect to the other targets in order to improve the amplitude of the peak, and so the LoD. This could be due to the lower concentration of L. monocytogenes after the enrichment, or to the preferential binding of SYBR Green to specific DNA fragments, previously it has been reported a preference of SYBR Green for G+C rich sequences [194,195]. The G-C content of the fragment generated in this studies were 40.1, 54.3 and 47.2% for rfbE, fimA and actA respectively, what would agree with the fact that fimA had a bigger peak with a lower primer concentration. The detection of the different microorganisms, spiked at different concentration levels, in the same sample, was successfully accomplished. Simultaneous detection has advantages in terms of cost savings and shorter time to result. Overall the methodology proved to be reliable and sensitive, as 100% of the results obtained were in concordance with the expected results for all evaluated parameters, and had comparable results to other studies using qPCR with probes for multiplex detection [196–198]. SARAH AZINHEIRO 110 5.2.2 Probe-qPCR qPCR with hydrolysis probes has been extensively used for a broad range of applications. This approach has the advantage of being easily multiplexed, without the difficulties of the primer design presented in the intercalating dye qPCR. Different sets of primers were evaluated using this technique, targeting specific genetic targets of the three pathogens. For L. monocytogenes the comparison between actA primers, used in the SYBR-qPCR approach, and hly primers combined with their respective probe, was performed. Regarding Salmonella spp. and E. coli O157, ttr and rfbE genes were chosen. The efficiency and inclusivity/ exclusivity of this primers/probes were analysed in simplex or multiplex with an IAC. 5.2.2.1 L. monocytogenes When targeting hly, a cIAC was implemented. This approach has the advantage of using the same primers to amplify both targets, the internal amplification control sequence and the specific target sequence. For this reason the first step was to optimize the concentration of the IAC to avoid interference in the amplification of the main target. It was observed that the addition of 100 copies/ µL (2000 copies per reaction) of cIAC DNA, provided optimal results, as lower concentrations presented higher replicate deviations as well as a final Cq value excessively high (>33 cycles), these results are depicted in Figure 5.5 A. After the conditions of the reaction were optimized, it was confirmed that the IAC DNA did not interfere with the amplification of hly, as no preferential amplification over the samples was seen. The same was observed for the interaction between the different probes and targets, as no amplification was originated when the hly probe was used with IAC DNA, or when the IAC probe was added with DNA from L. monocytogenes. These results confirmed the appropriate design of the reaction and implementation of the IAC. The efficiency of the qPCR reaction was evaluated only implementing hly, and compared with the results obtained in multiplex when both targets, hly and IAC, were co-amplified. An efficiency of 92 % and 90 % in simplex and multiplex (co-amplification of IAC) format was obtained, respectively. In both situations, the R2 of the equation was higher than 0.99 and covered 6 consecutive dilutions from 120 ng/ µL to 1.2 pg/ µL, as shown in Figure 5.5 B and C. This experiment also allowed to determine the lowest concentration of DNA providing positive amplification of hly. Both formats needed a minimum concentration of 1.2 pg/ µL from L. monocytogenes total DNA to have a positive result. It was observed that when high concentrations of pathogen DNA was amplified, presenting a low Cq value for hly, the amplification of IAC could be delayed or even completely absent. However as the IAC was integrated in the reaction as control to identify false negatives results when no amplification is detected for hly, therefore the absence of IAC amplification when hly amplify was not considered problematic. For the detection of L. monocytogenes targeting actA gene, a NC-IAC was tested as a different approach. The reaction efficiency for this combination is represented in Figure 5.5 D, showing a value of 93.4 %, slightly higher than the results with the hly multiplex method detailed above, but once more between the acceptable values. The lowest concentration of bacterial DNA detected was 0.2 pg/ µL, 6 times lower than in the case of the hly reaction. Chapter 5. Results – DNA amplification approaches 111 Figure 5.5. Probe-qPCR optimization and evaluation for the detection of L. monocytogenes using different genetic targets, hly and actA. (A) Concentration optimization of competitive IAC to be used with hly primers. The specified quantities are expressed in copies/ mL. (B) and (C) Efficiency of hly detection in simplex and multiplex format with simultaneous detection of both targets (hly with 100 copies/µL of cIAC DNA), respectively. (D) qPCR efficiency targeting actA and NC-IAC simultaneously. Efficiency curves were obtained by three replicates of ten-fold serial dilutions of a pure DNA extract. Beyond the efficiency and dynamic range, the inclusivity/ exclusivity of both primer sets were evaluated, testing the methods with the different bacterial stains presented in Table 5.3. The optimized qPCR protocol using actA and hly combined with their respective IACs, provided positive results exclusively with the 16 strains of L. monocytogenes. The other 8 Listeria spp., and extra 22 bacterial stains, were all negative, confirming the optimal specificity of this genetic targets. SARAH AZINHEIRO 112 Table 5.3. Inclusivity and exclusivity evaluation of Probe-qPCR Bacterium Source N hly actA rfbE ttr E. coli O157:H7 WDCM 00014, clinical isolated (AMC) 2 + E. coli O157:H7 In silico PCR 5 + E. coli WDCM 00013, 11 Mollusk, Cow stool (AMC 275), Intestinal biopsy (LSP-389-99), Sea water (AMC 176) 15 -* -* - -* E. coli In silico PCR 60 - L. monocytogenes WDCM00021, Mollusk, chestnut, chicken 16 + + - -** L. seeligeri CECT 917 1 - - - L. ivanovii WDCM00018 1 - - - - L. innocua WDCM00017, CECT 5376, 4030; CUP 1141, 1325, 2110 6 - - - - C. coli UM 1 - - - - C. freundii CECT 401 - E. faecalis WDCM 00009 - S. sonnei CECT 413 - S. aureus WDCM 00034, 00033 - Proteus spp. Mollusk (AMC 178) 1 - Salmonella spp. (AMC 28, 60, 82, 84, 90, 96, 198. 200, 238, 253, 255, 260, 261, UB, WDCM 00031) 15 - - - + Evaluation of the inclusivity and exclusivity of the qPCR reaction using rfbE primers in simplex. All E. coli O157 strains were correctly identified and all non-target bacteria were not detected in the reaction; N: number of strains; * only tested against WDCM 00013; ** refers to results obtained with the strains from mollusk source. Both primer sets performed well, showing high specificity and sensitivity for the detection of L. monocytogenes, with lowest concentration of DNA detected by actA. In both cases no interference with the IAC was observed when a multiplex reaction was performed. The gene actA has been extensively used to detect L. monocytogenes by qPCR [193,199,200], as well as hly [115,120,121,201]. The competitive format of the IAC is recommended when only one target is detected in order to minimize the chances of undesired interactions among several primers [202], however in a multiplex reaction targeting different pathogens a NC-IAC is more convenient to test the overall reaction inhibition, without the competition for the same primers [203]. The two sets of primers, and respective probes, were both used to evaluate several of the pre-treatment approaches as mentioned in Chapter I, proving their performance to detect L. monocytogenes. 5.2.2.2 E. coli O157 For the detection of E. coli O157, rfbE was the genetic target chosen. As performed for actA and hlyA, these primers were evaluated in terms of efficiency and specificity. Chapter 5. Results – DNA amplification approaches 113 Both simplex and multiplex, including a NC-IAC, approaches were evaluate for the detection of tenfold serial dilution of of E. coli O157 DNA (Figure 5.6 A and B) and with dilution of the bacterial culture followed by DNA extraction (Figure 5.6 C and D). The analysis of the dilutions of the pure E. coli O157 DNA, showed no impact when the NC-IAC was included in the qPCR assay, as there were no major differences between the results obtained from the simplex and multiplex assays, 99.4 and 100.7 % respectively. Concentrations of DNA ranging from 252 ng/ μL to 0.0252 pg/ μL produced a reproducible amplification, being 0.0252 pg/ μL the lowest DNA concentration which could be detected. Similar efficiency was obtained when DNA extracted from the diluted pure cultures of the pathogen were tested as template. In simplex and multiplex, the efficiency calculated was 101.3 and 103.8 % respectively. And as for the other approach 8 orders of magnitude, from 1.7× 108 to 17 cfu/ mL were covered by the dynamic range and the lowest bacterial concentration to give a reliable amplification was 17 cfu/ mL. Figure 5.6. Evaluation of the Probe-qPCR reaction targeting rfbE in simplex or multiplex with simultaneous amplification of NC-IAC. (A) and (B) represent the ten-fold dilutions of pure bacterial DNA, and (C) and (D) ten-fold dilutions of bacterial culture followed by DNA extraction from each of the dilutions. The specificity of the set of primers was not only tested in vitro, performing the qPCR reaction using pure DNA extracts of 2 E. coli O157 and 15 nonO157 strains, but also in silico, where an online software was used (http://insilico.ehu.es/PCR/) to test 5 E. coli O157 and 60 nonO157. Additionally, the cross- SARAH AZINHEIRO 114 reactivity against a panel of 41 non-target microorganisms was also evaluated , which included 15 Salmonella spp., 16 L. monocytogenes, 8 Listeria spp., 1 Proteus spp. and 1 Campylobacter coli. All targeted strains were well identified and no amplification was originated by the non-targeted microorganisms, confirming the specificity of the reaction, as presented in Table 5.3 The multiplex qPCR protocol using this primers and the NC-IAC, was used for the detection of E. coli O157 to evaluate the short enrichment approach, described in Chapter 4, section 4.4.3, allowing a LOD95 3.6 cfu/25 g and a  of 0.94 [204]. 5.2.2.3 Salmonella spp. For the detection of Salmonella spp. by Probe-qPCR, ttr was selected as genetic target, and the efficiency and specificity of the reaction were evaluated for the simplex and multiplex detection when the NC-IAC was also implemented. A dynamic range of five ten-fold serial dilutions was obtained, being possible to detect Salmonella DNA between 193 ng/ mL and 0.0193 ng/ mL (Figure 5.7). This shows lower detection levels that the ones obtained for hly, actA and rfbE which ranged from 0.2 to 0.02 pg/ mL. Additionally a lower efficiency value was observed with this set of primers, obtaining a value of 84% and 89 %, for the simplex and multiplex approach, respectively. In order to test the specificity of this set of primers and probe, 20 non-Salmonella strains, covering 10 different species, including other bacteria frequently found in food samples, such as E. coli or S. aureus, were analysed (Table 5.3). All the 15 Salmonella strains were correctly identified, while the nontargets obtained negative results. The multiplexing of these primers and probe with the ones targeting actA and NC-IAC were used for the detection of the respective pathogens using the PDMS sponge for bacterial concentration described in Chapter 4, section 4.3.2. High sensitivity and specificity with a LoD of 103 cfu/ mL was achieved, similar to other qPCR approach using hydrolysis probes [205], while previous similar studies reported a higher LoD (104 cfu/ mL) [206]. Figure 5.7. Efficiency and dynamic range of the Probe-qPCR targeting ttr gene. (A) Simplex detection (B) Multiplex detection, implementing NC-IAC. Efficiency curve obtained by three replicates of ten-fold dilutions of pure bacterial DNA. Chapter 5. Results – DNA amplification approaches 115 5.2.3 qLAMP, qRPA and comparison between real-time techniques A comparison between qPCR, qRPA and qLAMP approach was performed for the detection of L. monocytogenes. Primers were designed for the LAMP assay targeting plcA gene, and the F3/ B3 were used to perform the qPCR assays using intercalating SYBR-qPCR, while the primers hly designed for qPCR were used for the Probe-qPCR methodology, and combined with the Exo probe for RPA amplification. All approaches were evaluated assessing the dynamic range, detection probability, inclusivity/ exclusivity and also the detection of the pathogen in spiked samples. 5.2.3.1 Evaluation of the different amplification reactions Concentrations between 120 ng/ μL and 0.000012 ng/ μL in 9 replicates were tested by each approach to evaluate the dynamic range. The results are presented in Figure 5.8 A. In both isothermal amplification techniques, qLAMP and qRPA, a concentration down to 0.12 ng/ μL was detected with a 100 % probability. However, the traditional qPCR methodology achieved a lower minimum concentration of DNA with the same 100 % of detection probability, being 0.012 ng/ μL and 0.0012 ng/ μL for the SYBR-qPCR and Probe-qPCR respectively (Figure 5.8 B). While previous studies presented comparable results, showing a decrease in sensitivity using isothermal approaches [207,208], others reported similar or even higher detection sensitivity in LAMP and RPA over standard qPCR methodologies [134,209]. On the other hand, the fastest methodology showed to be qRPA with positive amplification after only 4 min with the highest concentration tested, while qLAMP required 15 min in order to enable a positive result with the same concentration. Comparatively, the qPCR methodologies needed more time than qRPA to achieve similar results. The specificity of qPCR and qRPA using the same primers, as well as the SYBR-qPCR with F3/ B3 and qLAMP was compared and results are detailed in Table 5.4. The qLAMP primers F3/ B3 were first used to confirm their specificity in silico by qPCR. All Listeria spp. from the data bank were tested, and included 43 strains (1 L. innocua, 1 L. ivanovii,1 L. seeligeri,1 L. welshimeri and 39 L. monocytogenes), and only L. monocytogenes showed amplification. The in vitro tests were equally performed with all primers, confirming 100 % specificity when tested against 16 L. monocytogenes strains, and 30 nontarget strains, including 6 L. innocua, 1 L. seeligeri and 1 L. ivanovii, without any cross-reactivity. SARAH AZINHEIRO 218 PUBLICATION 10 Garrido-Maestu, A. 1, Azinheiro, S. 1, Carvalho, J. 1, Fuciños, P. 2, & Prado, M. 1 (2019). Optimized sample treatment, combined with real-time PCR, for same-day detection of E. coli O157 in ground beef and leafy greens. Food Control, 106790. 1 Department of Life Sciences, Nano4Food Unit, Food Quality and Safety Research Group, Av. Mestre José Veiga s/n, 4715-330, Braga, Portugal 2 Department of Life Sciences, Nano4Food Unit, Food Processing Research Group, Av. Mestre José Veiga s/n, 4715-330, Braga, Portugal DOI: https://doi.org/10.1016/j.foodcont.2019.106790 Journal: Food Control Year: 2019 Impact Factor (2020): FOOD SCIENCE & TECHNOLOGY - SCIE 5.548 Q1 (19/143) Contribution of Azinheiro S. (CRediT taxonomy): Performed the experiments, Writing - review & editing Journal Authorization for publication in the PhD thesis: Publisher Elsevier https://www.elsevier.com/about/policies/copyright/permissions (Accessed on 11/05/2022) Appendix 219 PUBLICATION 11 Garrido-Maestu, A. 1, Azinheiro, S. 1, Carvalho, J. 1, & Prado, M. 1 (2018). Rapid and sensitive detection of viable Listeria monocytogenes in food products by a filtration-based protocol and qPCR. Food Microbiology, 73, 254–263. 1 International Iberian Nanotechnology Laboratory, Av. Mestre Jos_e Veiga s/n, 4715-330, Braga, Portugal DOI: https://doi.org/10.1016/j.fm.2018.02.004 Journal: Food Microbiology Year: 2018 Impact Factor (2020): MICROBIOLOGY - SCIE 5.516 Q1 ( 29/136) Contribution of Azinheiro S. (CRediT taxonomy): Performed the experiments, Writing - review & editing Journal Authorization for publication in the PhD thesis: Publisher Elsevier https://www.elsevier.com/about/policies/copyright/permissions (Accessed on 11/05/2022) SARAH AZINHEIRO 220 PUBLICATION 12 Azinheiro, S. 1, Carvalho, J. 1, Prado, M. 1, & Garrido-Maestu, A. 1 (2018). Evaluation of Different Genetic Targets for Salmonella enterica Serovar Enteriditis and Typhimurium, Using Loop-Mediated Isothermal AMPlification for Detection in Food Samples. Frontiers in Sustainable Food Systems, 2(February), 1–8. 1 Department of Life Sciences, Nano4Food – Food Quality and Safety Research Group, International Iberian Nanotechnology Laboratory, Braga, Portugal DOI: https://doi.org/10.3389/fsufs.2018.00005 Journal: Frontiers in Sustainable Food Systems Year: 2018 Impact Factor (2020): FOOD SCIENCE & TECHNOLOGY - ESCI 0.81 (JCI) Q2 ( 53/163) Contribution of Azinheiro S. (CRediT taxonomy): Performed the experiments and helped writing the manuscript. Journal Authorization for publication in the PhD thesis: Open Access Appendix 221 PUBLICATION 13 Garrido-Maestu, A. 1, Azinheiro, S. 1, Carvalho, J. 1, Abalde-Cela, S. 1, Carbó-Argibay, E. 1, Diéguez, L. 1, Prado, M. 1 (2017). Combination of Microfluidic Loop-Mediated Isothermal Amplification with Gold Nanoparticles for Rapid Detection of Salmonella spp. in Food Samples. Frontiers in Microbiology, 8, 2159. 1 International Iberian Nanotechnology Laboratory, Braga, Portugal DOI: https://doi.org/10.3389/fmicb.2017.02159 Journal: Frontiers in Microbiology Year: 2017 Impact Factor (2020): MICROBIOLOGY - SCIE 5.640 Q1 ( 28/136) Contribution of Azinheiro S. (CRediT taxonomy): Performed the DNA amplification experiments and AuNPs evaluation Journal Authorization for publication in the PhD thesis: Open Access SARAH AZINHEIRO 222 PUBLICATION 14 Garrido-Maestu, A. 1, Azinheiro, S. 1, Carvalho, J. 1, Fuciños, P. 1, & Prado, M. 1 (2017). Development and evaluation of Loop-mediated isothermal amplification, and Recombinase Polymerase Amplification methodologies, for the detection of Listeria monocytogenes in ready-to-eat food samples. Food Control, 86. 1 Department of Life Sciences, International Iberian Nanotechnology Laboratory, Av. Mestre Jose´ Veiga s/n, 4715-330 Braga, Portugal DOI: https://doi.org/10.1016/j.foodcont.2017.11.006 Journal: Food Control Year: 2017 Impact Factor (2020): FOOD SCIENCE & TECHNOLOGY - SCIE 5.548 Q1 (19/143) Contribution of Azinheiro S. (CRediT taxonomy): Performed the experiments, Writing - review & editing Journal Authorization for publication in the PhD thesis: Publisher Elsevier https://www.elsevier.com/about/policies/copyright/permissions (Accessed on 11/05/2022) Appendix 223 APPENDIX II – COPYRIGHT LICENSE AGREEMENTS AND PERMISSIONS Publication Reference Copyright License Agreement Publication 6 SPRINGER NATURE LICENSE Figure Reference Copyright License Agreement/ Author Permission Figure 1 [3] EFSA Legal notice Figure 2 Table 1 Table 2 Table 3 Figure 8 [88] JOHN WILEY AND SONS LICENSE (5306460306893) Figure 9 SARAH AZINHEIRO 224 5/12/22, 8:20 PM Legal notice | EFSA https://www.efsa.europa.eu/en/legalnotice?fbclid=IwAR0iP9pQPOalJX_WLDF33hewUOReejRaxpCjkfMt_eOju8HvUZPkQ2omdYQ#copyright-noti… 1/4 Legal notice The information on the EFSA website is subject to the following: a disclaimer; a linking policy; a copyright notice; a statement on confidential information. Disclaimer The European Food Safety Authority (EFSA) maintains this website to enhance public access to information about its activities and initiatives. Our goal is to keep this information timely and accurate. If errors are brought to our attention, we will try to correct them as soon as possible. EFSA accepts no responsibility or liability (including, but not limited to, any direct or consequential loss or damage that might occur to you and/or any other third party) arising out of, or in connection with the information on this site, including information relating to the documents it publishes. EFSA cannot be held liable for errors, inaccuracies or inconsistencies with regard to texts available on its website being different from the scientific and administrative documents officially adopted, issued or endorsed by its responsible governance bodies. In case of disagreement between the adopted text and the document made publicly available on EFSA’s website, the adopted text will always prevail. Information on the EFSA website: reflects the information and state of knowledge available on the date it was adopted or endorsed; is not necessarily comprehensive, complete, accurate or up to date; is not to be interpreted as professional or legal advice (if you need specific advice, you should always consult a suitably qualified professional). It is our goal to minimise disruption caused by technical errors. However, some data or information on our website may have been created or structured in files or formats that are not error-free, and we cannot guarantee that our service will not be interrupted or otherwise affected 5/12/22, 8:20 PM Legal notice | EFSA https://www.efsa.europa.eu/en/legalnotice?fbclid=IwAR0iP9pQPOalJX_WLDF33hewUOReejRaxpCjkfMt_eOju8HvUZPkQ2omdYQ#copyright-noti… 2/4 by such problems. EFSA accepts no responsibility with regard to such problems incurred as a result of using this site or any linked external sites. This disclaimer is not intended to limit EFSA’s liability in contravention of any requirements laid down in applicable national law nor to exclude its liability for matters which may not be excluded under such law. Linking policy EFSA encourages organisations and individuals to create links to its website. Linking should not: be used in a defamatory context; be displayed in a manner which suggests that EFSA endorses any commercial product or service, scientific or administrative output, or the information or text in the context of which it is published; confuse users of the other website as to the source of the content. EFSA should be fully acknowledged as a source or author of the relevant documents; be displayed alongside advertising. Some EFSA pages link to other websites maintained by external parties, over which EFSA has no control and for which EFSA is not responsible. Linking to another website should not be taken as an endorsement by EFSA. Users clicking and visiting another site are subject to the policies applicable to the new site. Copyright notice Copyright © for the entire content of this website unless otherwise stated: European Food Safety Authority – EFSA. Re-use is authorised, provided that EFSA is acknowledged as the source of the material. For individual documents, the general principle of re-use outlined above may be subject to specific conditions as indicated in individual copyright notices contained therein. Users are advised to refer to these individual copyright notices. The EFSA logo is the exclusive property of the European Food Safety Authority – EFSA and is registered under Article 6ter of the Paris Convention for the protection of industrial property. Its use is prohibited without the prior written permission of EFSA. Statement on confidential information In accordance with its founding legislation, Regulation 178/2002 (“the General Food Law”), as amended by the implementation of Regulation (EC) 1381/2019 (“the Transparency Regulation”) on 27.03.2021, EFSA shall carry out its activities in the area of food law with a high level of transparency. This entails that for all mandates and applications received by the Authority 5/12/22, 8:20 PM Legal notice | EFSA https://www.efsa.europa.eu/en/legalnotice?fbclid=IwAR0iP9pQPOalJX_WLDF33hewUOReejRaxpCjkfMt_eOju8HvUZPkQ2omdYQ#copyright-noti… 3/4 on/after 27.03.2021, to which the Transparency Regulation applies, EFSA shall make publicly available on its website information listed in Article 38(1) of the General Food Law or applicable Union law unless it is considered confidential pursuant to Articles 39-39e of the General Food Law or applicable Union law. In accordance with Article 39(2) of the General Food Law, upon the request of an applicant confidential treatment may be given to items of information where the disclosure is demonstrated by the applicant to potentially harm its interests to a significant degree. Confidentiality requests may be submitted only with respect to items of information expressly set out in the closed positive list contained in Article 39(2)(a) to (d) and Article 39(3) of the General Food law as listed in the Appendix of EFSA’s Practical Arrangements of transparency and confidentiality (“EFSA’s Practical Arrangements”). The provisions of Article 39(2)(a) to (d) and of Article 39(3) are without prejudice to the provisions contained in Article 39(4)(a) and Article 39(4)(b) which relate to the disclosure of information where urgent action is essential to protect human health, animal health or the environment, such as in emergency situations, or to information which forms part of conclusions of scientific outputs, including scientific opinions, delivered by the Authority and which relate to foreseeable effects on human health, animal health or the environment. Pursuant to Article 38(1a), the disclosure of a) scientific data, studies and other information supporting applications, including supplementary information supplied by applicants as well as other scientific data and information supporting requests from the European Parliament, the Commission and the Member States for a scientific output, including a scientific opinion, b) information on which scientific outputs of the Authority, including scientific opinions, are based is without prejudice to without prejudice to: 1. existing rules concerning intellectual property rights which set out limitations on certain uses of the disclosed documents or their content. 2. provisions protecting the investment made by innovators in gathering the information and data supporting relevant applications for authorisations (“data exclusivity rules”). Subject to the terms and conditions of this Legal Notice, and subject to other binding limitations provided for under applicable law, the information, documents and data made available on the website of the Authority, may be downloaded and printed, totally or in part, provided that the user acknowledges the applicable conditions and, in particular, that EFSA accepts no responsibility and/or liability for any use made of the information, documents or data. In relation to mandates and applications received by the Authority prior to 27.03.2021 the transparency and confidentiality provisions contained in Article 38 and Article 39(1) of the General Food Law in its pre Transparency Regulation version continue to apply. The reproduction, distribution or further use of information, documents and data posted or otherwise made available on EFSA’s website, the OpenEFSA portal or on websites linked to EFSA may be subject to protection under intellectual property rights regulations or other applicable law, and their utilisation without obtaining the prior permission from the right(s)holders(s) of the i i f i d d d i h i l h i h f h i 5/12/22, 12:24 PM RightsLink Printable License https://s100.copyright.com/AppDispatchServlet 6/6 Other Terms and Conditions: v1.10 Last updated September 2015 Questions? customercar[email protected] or +1-855-239-3415 (toll free in the US) or +1-978-646-2777. 5/10/22, 3:00 PM RightsLink Printable License https://s100.copyright.com/AppDispatchServlet 1/5 SPRINGER NATURE LICENSE TERMS AND CONDITIONS May 10, 2022 This Agreement between University of Santiago de compostela -- Sarah Azinheiro ("You") and Springer Nature ("Springer Nature") consists of your license details and the terms and conditions provided by Springer Nature and Copyright Clearance Center. License Number 5305380165949 License date May 10, 2022 Licensed Content Publisher Springer Nature Licensed Content Publication Journal of Food Science and Technology Licensed Content Title Evaluation and implementation of commercial antibodies for improved nanoparticle-based immunomagnetic separation and realtime PCR for faster detection of Listeria monocytogenes Licensed Content Author Alejandro Garrido-Maestu et al Licensed Content Date Apr 24, 2020 Type of Use Thesis/Dissertation Requestor type academic/university or research institute Format print and electronic Portion full article/chapter Will you be translating? no 5/10/22, 3:00 PM RightsLink Printable License https://s100.copyright.com/AppDispatchServlet 2/5 Circulation/distribution 20000 - 49999 Author of this Springer Nature content yes Title Msc Institution name University of Santiago de Compostela Expected presentation date Aug 2022 Requestor Location University of Santiago de compostela Praza do Obradoiro Santiago de compostela, 15705 Spain Attn: University of Santiago de compostela Total 0.00 USD Terms and Conditions Springer Nature Customer Service Centre GmbH Terms and Conditions This agreement sets out the terms and conditions of the licence (the Licence) between you and Springer Nature Customer Service Centre GmbH (the Licensor). By clicking 'accept' and completing the transaction for the material (Licensed Material), you also confirm your acceptance of these terms and conditions. 1. Grant of License 1. 1. The Licensor grants you a personal, non-exclusive, non-transferable, world-wide licence to reproduce the Licensed Material for the purpose specified in your order only. Licences are granted for the specific use requested in the order and for no other use, subject to the conditions below. 1. 2. The Licensor warrants that it has, to the best of its knowledge, the rights to license reuse of the Licensed Material. However, you should ensure that the material you are requesting is original to the Licensor and does not carry the copyright of another entity (as credited in the published version). 1. 3. If the credit line on any part of the material you have requested indicates that it was reprinted or adapted with permission from another source, then you should also seek permission from that source to reuse the material. 2. Scope of Licence 5/10/22, 3:00 PM RightsLink Printable License https://s100.copyright.com/AppDispatchServlet 3/5 2. 1. You may only use the Licensed Content in the manner and to the extent permitted by these Ts&Cs and any applicable laws. 2. 2. A separate licence may be required for any additional use of the Licensed Material, e.g. where a licence has been purchased for print only use, separate permission must be obtained for electronic re-use. Similarly, a licence is only valid in the language selected and does not apply for editions in other languages unless additional translation rights have been granted separately in the licence. Any content owned by third parties are expressly excluded from the licence. 2. 3. Similarly, rights for additional components such as custom editions and derivatives require additional permission and may be subject to an additional fee. Please apply to [email protected]/[email protected] for these rights. 2. 4. Where permission has been granted free of charge for material in print, permission may also be granted for any electronic version of that work, provided that the material is incidental to your work as a whole and that the electronic version is essentially equivalent to, or substitutes for, the print version. 2. 5. An alternative scope of licence may apply to signatories of the STM Permissions Guidelines, as amended from time to time. 3. Duration of Licence 3. 1. A licence for is valid from the date of purchase ('Licence Date') at the end of the relevant period in the below table: Scope of Licence Duration of Licence Post on a website 12 months Presentations 12 months Books and journals Lifetime of the edition in the language purchased 4. Acknowledgement 4. 1. The Licensor's permission must be acknowledged next to the Licenced Material in print. In electronic form, this acknowledgement must be visible at the same time as the figures/tables/illustrations or abstract, and must be hyperlinked to the journal/book's homepage. Our required acknowledgement format is in the Appendix below. 5. Restrictions on use 5. 1. Use of the Licensed Material may be permitted for incidental promotional use and minor editing privileges e.g. minor adaptations of single figures, changes of format, colour and/or style where the adaptation is credited as set out in Appendix 1 below. Any other changes including but not limited to, cropping, adapting, omitting material that affect the meaning, intention or moral rights of the author are strictly prohibited. 5. 2. You must not use any Licensed Material as part of any design or trademark. 5/10/22, 3:00 PM RightsLink Printable License https://s100.copyright.com/AppDispatchServlet 4/5 5. 3. Licensed Material may be used in Open Access Publications (OAP) before publication by Springer Nature, but any Licensed Material must be removed from OAP sites prior to final publication. 6. Ownership of Rights 6. 1. Licensed Material remains the property of either Licensor or the relevant third party and any rights not explicitly granted herein are expressly reserved. 7. Warranty IN NO EVENT SHALL LICENSOR BE LIABLE TO YOU OR ANY OTHER PARTY OR ANY OTHER PERSON OR FOR ANY SPECIAL, CONSEQUENTIAL, INCIDENTAL OR INDIRECT DAMAGES, HOWEVER CAUSED, ARISING OUT OF OR IN CONNECTION WITH THE DOWNLOADING, VIEWING OR USE OF THE MATERIALS REGARDLESS OF THE FORM OF ACTION, WHETHER FOR BREACH OF CONTRACT, BREACH OF WARRANTY, TORT, NEGLIGENCE, INFRINGEMENT OR OTHERWISE (INCLUDING, WITHOUT LIMITATION, DAMAGES BASED ON LOSS OF PROFITS, DATA, FILES, USE, BUSINESS OPPORTUNITY OR CLAIMS OF THIRD PARTIES), AND WHETHER OR NOT THE PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. THIS LIMITATION SHALL APPLY NOTWITHSTANDING ANY FAILURE OF ESSENTIAL PURPOSE OF ANY LIMITED REMEDY PROVIDED HEREIN. 8. Limitations 8. 1. BOOKS ONLY:Where 'reuse in a dissertation/thesis' has been selected the following terms apply: Print rights of the final author's accepted manuscript (for clarity, NOT the published version) for up to 100 copies, electronic rights for use only on a personal website or institutional repository as defined by the Sherpa guideline (www.sherpa.ac.uk/romeo/). 8. 2. For content reuse requests that qualify for permission under the STM Permissions Guidelines, which may be updated from time to time, the STM Permissions Guidelines supersede the terms and conditions contained in this licence. 9. Termination and Cancellation 9. 1. Licences will expire after the period shown in Clause 3 (above). 9. 2. Licensee reserves the right to terminate the Licence in the event that payment is not received in full or if there has been a breach of this agreement by you. Appendix 1 — Acknowledgements: For Journal Content: Reprinted by permission from [the Licensor]: [Journal Publisher (e.g. 5/10/22, 3:00 PM RightsLink Printable License https://s100.copyright.com/AppDispatchServlet 5/5 Nature/Springer/Palgrave)] [JOURNAL NAME] [REFERENCE CITATION (Article name, Author(s) Name), [COPYRIGHT] (year of publication) For Advance Online Publication papers: Reprinted by permission from [the Licensor]: [Journal Publisher (e.g. Nature/Springer/Palgrave)] [JOURNAL NAME] [REFERENCE CITATION (Article name, Author(s) Name), [COPYRIGHT] (year of publication), advance online publication, day month year (doi: 10.1038/sj.[JOURNAL ACRONYM].) For Adaptations/Translations: Adapted/Translated by permission from [the Licensor]: [Journal Publisher (e.g. Nature/Springer/Palgrave)] [JOURNAL NAME] [REFERENCE CITATION (Article name, Author(s) Name), [COPYRIGHT] (year of publication) Note: For any republication from the British Journal of Cancer, the following credit line style applies: Reprinted/adapted/translated by permission from [the Licensor]: on behalf of Cancer Research UK: : [Journal Publisher (e.g. Nature/Springer/Palgrave)] [JOURNAL NAME] [REFERENCE CITATION (Article name, Author(s) Name), [COPYRIGHT] (year of publication) For Advance Online Publication papers: Reprinted by permission from The [the Licensor]: on behalf of Cancer Research UK: [Journal Publisher (e.g. Nature/Springer/Palgrave)] [JOURNAL NAME] [REFERENCE CITATION (Article name, Author(s) Name), [COPYRIGHT] (year of publication), advance online publication, day month year (doi: 10.1038/sj. [JOURNAL ACRONYM]) For Book content: Reprinted/adapted by permission from [the Licensor]: [Book Publisher (e.g. Palgrave Macmillan, Springer etc) [Book Title] by [Book author(s)] [COPYRIGHT] (year of publication) Other Conditions: Version 1.3 Questions? customercar[email protected] or +1-855-239-3415 (toll free in the US) or +1-978-646-2777.