Host specificity and virulence of the phytopathogenic bacteria Pseudomonas savastanoi

Instituto de Hortofruticultura Subtropical y Mediterránea “La Mayora” Universidad de Málaga –CSIC Año 2017 TESIS DOCTORAL Host specificity and virulence of the phytopathogenic bacteria Pseudomonas savastanoi Eloy Caballo Ponce Director: Cayo Ramos Rodríguez Programa de Doctorado: Biotecnología Avanzada TESIS DOCTORAL Eloy Caballo Ponce 2017 Memoria presentada por: Eloy Caballo Ponce para optar al grado de Doctor por la Universidad de Málaga Host specificity and virulence of the phytopathogenic bacteria Pseudomonas savastanoi Director: Cayo J. Ramos Rodríguez Catedrático. Área de Genética. Departamento de Biología Celular, Genética y Fisiología. Instituto de Hortofruticultura Subtropical y Mediterránea (IHSM) Universidad de Málaga – Consejo Superior de Investigaciones Científicas Málaga, 2016 AUTOR: Eloy Caballo Ponce http://orcid.org/0000-0003-0501-3321 EDITA: Publicaciones y Divulgación Científica. Universidad de Málaga Esta obra está bajo una licencia de Creative Commons Reconocimiento-NoComercialSinObraDerivada 4.0 Internacional: http://creativecommons.org/licenses/by-nc-nd/4.0/legalcode Cualquier parte de esta obra se puede reproducir sin autorización pero con el reconocimiento y atribución de los autores. No se puede hacer uso comercial de la obra y no se puede alterar, transformar o hacer obras derivadas. Esta Tesis Doctoral está depositada en el Repositorio Institucional de la Universidad de Málaga (RIUMA): riuma.uma.es COMITÉ EVALUADOR Presidente Dr. Jesús Murillo Martínez Departamento de Producción Agraria Universidad Pública de Navarra Secretario Dr. Francisco Manuel Cazorla López Departamento de Microbiología Universidad de Málaga Vocal Dra. Chiaraluce Moretti Departamento de Ciencias Agrarias, Alimentarias y Medioambientales Universidad de Perugia Suplentes Dr. Pablo Rodríguez Palenzuela Departamento de Biotecnología – Biología Vegetal Universidad Politécnica de Madrid Dr. Diego Romero Hinojosa Departamento de Microbiología Universidad de Málaga Dr. Roberto Buonaurio Departamento de Ciencias Agrarias, Alimentarias y Medioambientales Universidad de Perugia Área de Genética. Departamento de Biología Celular, Genética y Fisiología. Instituto de Hortofruticultura Subtropical y Mediterránea (IHSM) Universidad de Málaga – Consejo Superior de Investigaciones Científicas Dr. CAYO J. RAMOS RODRÍGUEZ, Catedrático del Área de Genética del Departamento de Biología Celular, Genética y Fisiología. INFORMA: Que ELOY CABALLO PONCE ha realizado en este Departamento y bajo mi dirección el trabajo titulado “Host specificity and virulence of the phytopathogen bacteria Pseudomonas savastanoi”, que constituye su memoria de Tesis Doctoral para aspirar al grado de Doctor con Mención Europea en Biotecnología Avanzada. Y para que así conste, y tenga los efectos que correspondan, en cumplimiento con la legislación vigente, extiendo el presente informe. En Málaga, a 14 de octubre de 2016. Fdo: Cayo J. Ramos Rodríguez A mi familia A Blanca Este trabajo ha sido financiado por los proyectos AGL2011-30343-C02-01 y AGL201453242-C2-1-R del Plan Nacional I+D del Ministerio de Economía y Competitividad (cofinanciado por FEDER) y por la Beca para la Formación de Personal Investigador (FPI) (2013-2016) Index Index Abbreviations.............................................................................................................................. 21 Resumen ..................................................................................................................................... 25 General introduction .................................................................................................................. 37 The Pseudomonas syringae complex ...................................................................................... 39 Pseudomonas savastanoi ........................................................................................................ 41 P. savastanoi virulence and adaptation factors ...................................................................... 46 Objectives ................................................................................................................................... 55 CHAPTER I: WHOP, a genomic region associated to woody hosts in the Pseudomonas syringae complex, contributes to the virulence and fitness of Pseudomonas savastanoi pv. savastanoi in olive plants .................................................................................................................................. 59 Introduction ............................................................................................................................ 61 Materials and methods ........................................................................................................... 64 Results ..................................................................................................................................... 73 Discussion ................................................................................................................................ 92 CHAPTER II: Comparative analysis of the quorum sensing system in Pseudomonas savastanoi pv. savastanoi NCPPB 3335 and other Pseudomonas syringae strains....................................... 97 Introduction ............................................................................................................................ 99 Material and methods ........................................................................................................... 101 Results ................................................................................................................................... 108 Discussion .............................................................................................................................. 120 CHAPTER III: Pseudomonas savastanoi as an emerging pathogen: the case of dipladenia. .... 123 Introduction .......................................................................................................................... 125 Materials and methods ......................................................................................................... 127 Results ................................................................................................................................... 134 Discussion .............................................................................................................................. 146 Concluding remarks ......................................................................................................... 149 Conclusions ..................................................................................................................... 155 Appendix ......................................................................................................................... 159 References ...................................................................................................................... 163 Abbreviations Abbreviations 23 AHL: acyl-homoserine lactone ANOVA: Analysis of variance Ap: Ampicillin ASAP: A systematic annotation package bv: biovar Blast: Basic local alignment search tool bp: Base pair(s) cCNA: Complementary DNA CECT: Colección española de cultivos tipo cfu: Colony forming units CI: Competitive index CK: Cytokinin Cm: Cloramphenicol CV: Crystal violet DNA: Deoxyribonucleic acid dpi: Days post-inoculation DW: Dry weight E. coli: Escherichia coli EPS: Exopolysaccharides GFP: Green fluorescent protein Gm: Gentamycin h: Hours HSL: homoserine lactone HPLC: High performance liquid chromatography HPLC-MS: High performance liquid chromatography-mass spectrometry HR: Hypersensitive response IAA: indole 3-acetic acid ICMP: International Collection of Microorganisms from Plants Kb: Kilobase KB: King B medium Km: Kanamycin LB: Luria Bertani medium M: Molar MEGA5: Molecular evolutionary genetic analysis mg: Milligrams min: Minutes ml: Millilitres mM: Millimolar mV: Millivolts M9A: M9 minimal medium with anthranilate NCPPB: National collection of plant pathogenic bacteria nm: nanometre OD: Optical density ORF: Open reading frame PCR: Polymerase chain reaction ppb: parts per billion PG: Phylogroup Psd: Pseudomonas savastanoi isolated from dipladenia Psf: Pseudomonas savastanoi pv. fraxini Psn: Pseudomonas savastanoi pv. nerii Psr: Pseudomonas savastanoi pv. retacarpa Pst: Pseudomonas syringae pv. tabaci Psv: Pseudomonas savastanoi pv. savastanoi Psy: Pseudomonas syringae pv. syringae RACE: Rapid amplification of cDNA ends RNA: Ribonucleic acid Rt: Retention time RT-PCR: Reverse transcription polymerase chain reaction Abbreviations 24 RT-qPCR: Reverse transcription quantitative polymerase chain reaction SOB: Super optimal broth medium STM: Signature-tagged mutagenesis T3SS: Type three secretion system UV: Ultraviolet WHOP: Woody host and Pseudomonas µg: Micrograms Resumen 31 la concentración de este compuesto pasadas 24 horas mediante cromatografía líquida de alta eficacia (HPLC). En el sobrenadante correspondiente a la cepa silvestre Psv NCPPB 3335 no se detectó antranilato, confirmando la capacidad de esta bacteria de degradarlo. Por el contrario, una cantidad de antranilato prácticamente igual a la del inicio del experimento se detectó en el medio correspondiente al mutante del gen antA. En la cepa mutante antA complementada con el operón antABC completo se detectó una pequeña cantidad de antranilato (aproximadamente el 10% de la cantidad inicial) y la acumulación de catecol en el medio, posiblemente debido a la generación de un cuello de botella en la asimilación de catecol en esta cepa. Para confirmar el papel del operón catBCA en el metabolismo de catecol se empleó una metodología análoga a la descrita anteriormente, con la excepción de suplementar el medio mínimo con catecol en lugar de antranilato. En el medio correspondiente a Psv NCPPB 3335 no se detectó catecol, indicando que esta cepa consumió completamente el catecol presente en el medio. Por el contrario, en el sobrenadante de una cepa mutante del gen catB se detectaron varios compuestos que no se pudieron identificar, pero que posiblemente estaban relacionados con el catetol al presentar un tempo de retención y un espectro de absorción en el ultravioleta (UV) muy similar al de este compuesto. Esta hipótesis se ve reforzada por el hecho de que la cepa mutante del gen catB complementada con el operón catBCA no acumuló en el sobrenadante estos compuestos posiblemente relacionados con el catecol. Relacionado con el operón catBCA, se ha descrito que la actividad dienolactona hidrolasa (posiblemente codificada por el gen dhoA) está involucrada en el metabolismo de catecoles halogenados en la cepa Pseudomonas B13 (Schlomann et al., 1990), en la que también participan las enzimas CatA y CatB. Para analizar la posible participación de los operones dhoAB y catBCA en el catabolismo del 4-clorocatechol en Psv NCPPB 3335, la cepa silvestre y sus mutantes ΔcatB y ΔdhoAB se expusieron a un medio mínimo suplementado con 4-clorocatecol durante 24 horas y el sobrenadante fue analizado por HPLC. No se detectó 4-clorocatecol en el sobrenadante de Psv NCPPB 3335 y el mutante del operón dhoAB, mientras que en el mutante catBCA la concentración del compuesto era cuatro veces inferior a la inicial. Por tanto, Psv NCPPB 3335 puede degradar 4-clorocatecol por la acción, en parte, de las enzimas del operón catBCA, pero de forma independiente del operón dhoAB. Finalmente, se determinó que el operón ipoABC presenta actividad oxigenasa sobre compuestos aromáticos. A esta conclusión se llegó gracias al hecho de que determinadas oxigenasas son capaces de oxidar la molécula de indol y, tras una serie de pasos que ocurren espontáneamente, conducir a la formación del compuesto azul índigo. En este sentido, el cultivo en un medio mínimo suplementado con indol de una cepa de Psv que expresa el operón ipoABC desde un plásmido multicopia resultó en la producción de indigo. Resumen 32 Debido a que la región WHOP se encuentra principalmente en patógenos órganos leñosos dentro del complejo P. syringae, resultó particularmente interesante analizar su papel en la interacción de Psv NCPPB 3335 con plantas de olivo. Para este fin se utilizaron dos modelos diferentes, plantas de olivo micropropagadas (no leñosas) y plantas de un año de edad (leñosas). Las plantas se inocularon con Psv NCPPB 3335 y cada una de sus cepas mutantes en los clusters de la región WHOP. Transcurridos 30 días desde la inoculación, no se observaron diferencias visuales en el tamaño de los tumores desarrollados en plantas micropropagadas entre la cepa silvestre y sus mutantes. De hecho, la cuantificación del volumen tumoral usando un escáner 3D confirmó que el volumen de los tumores generados por los mutantes no era significativamente diferente al de Psv NCPPB 3335. Sin embargo, sí se observaron diferencias en el tamaño del tumor generado en plantas leñosas 90 días después de la inoculación. La mayor reducción en el volumen de los tumores se obtuvo para el mutante ΔantA, 3.1 veces más pequeñas en comparación con la cepa silvestre. Además, las plantas inoculadas con los mutantes ΔcatB y ΔipoABC desarrollaron tumores que eran, respectivamente, 2.37 y 1.57 veces más pequeños que los producidos por Psv NCPPB 3335. Por tanto, los operones antABC, catBCA e ipoABC juegan un papel en la virulencia de Psv NCPPB 3335 en las plantas de olivo leñosas, pero no en plantas no leñosas. Se ha propuesto que durante la degradación de IAA pueden generarse catecol y antranilato (Leveau & Gerards, 2008). El antranilato es un precursor en la síntesis del triptófano, sustrato inicial en la ruta de la 3-indolacetamida para la biosíntesis de IAA. Teniendo en cuenta que la cepa mutante del operón antABC acumula antranilato, podría producirse un desequilibrio en la cantidad de triptófano y, por tanto, en la biosíntesis de IAA causando una disminución en la virulencia de la bacteria. También se llevaron a cabo infecciones mixtas de la Psv NCPPB 3335 con cada uno de los mutantes afectados en la región WHOP tanto en olivos leñosos como no leñosos. En plantas de olivo no leñosas, los valores del índice competitividad (CI por sus siglas en inglés) no fueron estadísticamente diferentes a uno, a excepción de la del mutante ΔipoABC (CI= 0.39±0.16). Sin embargo, cuando las inoculaciones mixtas se llevaron a cabo en olivos leñosos, los valores de CI obtenidos fueron estadísticamente menor que uno, excepto para el mutante ΔantA. De manera similar a los resultados descritos anteriormente para la virulencia de las cepas mutantes, los mutantes ΔcatB, ΔPSA3335_3206 y ΔdhoAB están afectados en su crecimiento competitivo exclusivamente en plantas de olivo leñosas. Aunque la estructura química de la madera no se puede definir con precisión para una especie determinada, se compone principalmente de los polisacáridos celulosa y hemicelulosa (65-75%) y lignina (18-35%) (Pettersen, 1984). La lignina es un polímero complejo sintetizado a partir de tres alcoholes hidroxicinamil (alcoholes p- Resumen 33 cumaril, coniferil y sinapil) (Boerjan et al., 2003; Ralph et al., 2004). Se sabe que la degradación de moléculas relacionadas con la lignina se canaliza vía protocatecuato y catecol. Mientras que la maquinaria de degradación catecol se limita a varios patovares dentro del complejo P. syringae, la degradación de protocatecuato en intermediarios del ciclo de Krebs es una característica ampliamente distribuida en las bacterias del complejo P. syringae. Por lo tanto, la presencia de la región WHOP podría permitir la degradación vía catecol de compuestos relacionados con la lignina y facilitar la colonización de órganos leñosos. El capítulo II de esta Tesis Doctoral se centró en el estudio del sistema de regulación en respuesta a la densidad celular, denominado quorum sensing, en Psv NCPPB 3335. Como se mencionó anteriormente, este sistema se compone principalmente de proteínas pertenecientes a las familias LuxI y LuxR y una molécula señal denominada acil-homoserinalactona (AHL) en bacterias Gram (-).; El análisis de 265 genomas bacterianos mostró que el número de proteínas LuxI y LuxR no siempre es coincide: además del par luxI/luxR canónico, muchos genomas también contienen genes luxR adicionales que no están asociados a un luxI (Case et al., 2008). Las proteínas codificadas por estos luxR, llamados luxR solos (Subramoni & Venturi, 2009), responden a las AHL producidas por la propia bacteria (Chugani et al., 2001) u otras bacterias que se encuentren en el mismo cercanas (Ahmer et al., 1998). Además, se ha descrito una subfamilia de luxR solos exclusivos de bacterias asociadas de plantas y que, posiblemente, responden señales producidas por la planta (Patel et al., 2013). El sistema de quorum sensing de P. syringae está bien caracterizado en P. syringae B728a (Quinones et al., 2004; Quinones et al., 2005), habiéndose identificado los genes regulados por luxR en esta cepa (Yu et al., 2014). Además, un reciente estudio transcriptómico identificó los genes regulados por quorum sensing en P. syringae pv. tabaci 6605 (Taguchi et al., 2015). Ambos estudios se han llevado a cabo en cepas patógenas de plantas herbáceas; sin embargo, se sabe muy poco de este sistema de regulación en bacterias del complejo P. syringae patógenas de huéspedes leñosos. P. syringae pv. actinidiae, patógeno de kiwi, no produce AHLs, pero sí responde a ellas gracias a la presencia de varios luxR solos en su genoma (Patel et al., 2014). Por otro lado, los genes pssI y pssR de Psv DAPP-PG 722, homólogos a luxI y luxR respectivamente, juegan un papel importante en la virulencia de esta bacteria (Hosni et al., 2011). Dado que P. syringae pv. actinidiae y P. savastanoi pv. savastanoi difieren en el la presencia/ausencia de luxI, se llevó a cabo la búsqueda de homólogos a los genes luxI y luxR en los genomas de 27 cepas pertenecientes a los filogrupos 1 a 4. Mientras todas las cepas del filogrupo 3 analizadas contienen un homólogo a luxI, la presencia de éste en las cepas analizadas de los filogrupos 2 y 3 es variable, incluso entre aislados diferentes del mismo patovar. Por otro lado, el número y tipo de homólogos a luxR también es Resumen 34 variable entre las cepas analizadas. Mientras que todas las cepas del filogrupo 3 analizadas contienen tres homólogos a luxR (dos de ellos luxR solos), salvo P. syringae pv. aesculi NCPPB 3681 que contiene un tercer luxR solo, el número y tipo de homólogo a luxR es diverso entre las bacterias del filogrupo 2 y varía ligeramente entre las cepas del filogrupo 1. El genoma de Psv NCPPB 3335 contiene el par luxI/luxR canónico junto con dos luxR solos, una organización que está conservada en otras cepas pertenecientes a los cuatro patovares de P. savastanoi cuyos genomas se han secuenciado recientemente (Moretti et al., 2014; Bartoli et al., 2015b; Thakur et al., 2016). Sin embargo, y aunque la secuencia de ambos pssI es 100% idéntica, Psv NCPPB 3335 y Psv DAPP-PG 722 producen diferentes tipos de AHL. Dado que en Psv DAPP-PG 722 el sistema de regulación por quorum sensing juega un papel fundamental en la virulencia de la bacteria, decidimos determinar los genes cuya expresión depende de la síntesis de AHLs en Psv NCPPB 3335, la cepa que usamos rutinariamente en nuestro laboratorio. Para ello, construimos un mutante por intercambio alélico en el gen pssI de Psv NCPPB 3335 e hicimos un análisis comparativo del transcriptoma de esta cepa con la silvestre, a partir del cual se concluyó que pocos genes muestran una expresión diferencial. Se seleccionaron los genes sobreexpresados/reprimidos 1.87 o más veces en el mutante con respecto a la cepa silvestre para su validación mediante RT-PCR cuantitativa (RT-qPCR). Se validó la expresión diferencial de tres de estos genes (pssR, pdhT y pdhQ), todos ellos sobreexpresados en el mutante ΔpssI. Por tanto, en las condiciones ensayadas, el regulón pssI está compuesto por tres genes, un resultado contrasta con el gran número de genes cuya expresión depende del respectivo homólogo a luxI en el patógeno de tabaco P. syringae pv. tabaci 6605 (Taguchi et al., 2015). Para comparar los sistemas de regulación de las dos cepas, se seleccionaron los seis genes más sobreexpresados/reprimidos en el mutante del homólogo a luxI en P. syringae pv. tabaci 6605 y se analizó su expresión mediante RT-qPCR en Psv NCPPB 3335. Ninguno de los genes examinados mostró la misma regulación dependiente de luxI observada P. syringae pv. tabaci 6605, dado que la transcripción de la mayoría estos genes es independiente de pssI en Psv NCPPB 3335. Estos resultados indican que, aunque ambas cepas producen el mismo tipo de molécula señal, la regulación por quorum sensing es diferente en Psv NCPPB 3335 y P. syringae pv. tabaci 6605. En P. syringae pv. syringae B728a LuxR promueve la transcripción de luxI al llegar a un umbral de concentración de AHLs (Quinones et al., 2004), un fenómeno común en otros sistemas de quorum sensing. Puesto que la regulación por quorum sensing en Psv NCPPB 3335 parece ser diferente a otras cepas del complejo P. syringae decidimos analizar si PssR promueve la expresión de pssI. Para ello, el promotor de pssI se fusionó al gen lacZ y se midió la actividad Resumen 35 β-galactosidasa a lo largo de una curva de crecimiento en Psv NCPPB 3335 y sus cepas mutantes de los genes pssI y pssR. En la cepa silvestre se observó un aumento de la actividad enzimática al aproximarse a la fase estacionaria de cultivo que también ocurre en las cepas mutantes de los genes pssI y pssR. Estos resultados sugieren, por tanto, que el aumento de la transcripción de pssI es independiente de la síntesis de AHLs y de PssR, lo cual coincide con lo observado por otros autores en Psv DAPP-PG 722 (Hosni et al., 2011). Además, otra diferencia importante entre los sistemas de quorum sensing entre Psv NCPPB 3335 y P. syringae pv. syringae B728a aparece al analizar la actividad β-galactosidasa asociada al promotor de pssI en P. syringae pv. syringae B728a. La actividad enzimática medida en B728a fue, por lo general, 60 veces más alta que en NCPPB 3335 bajo las mismas condiciones experimentales, lo cual refuerza la hipótesis de una regulación por quorum sensing diferente en bacterias muy relacionadas filogenéticamente. Curiosamente, se ha descrito que cepas de Pseudomonas aeruginosa aisladas de nichos diversos difieren en los genes regulados por quorum sensing (Chugani et al., 2012). Además de las diferencias observadas entre Psv NCPPB 3335 y otras cepas del complejo P. syringae, resultados generados en esta Tesis Doctoral también muestran diferencias entre aislados de Psv. Además de no producir el mismo tipo de AHL, la virulencia y la producción de exopolisacáridos están controlados por quorum sensing en Psv DAPP-PG 722, pero no en Psv NCPPB 3335. Finalmente, el capítulo III se ha centrado en la caracterización de varias cepas de P. savastanoi patógenas de dipladenia, las cuales producen manchas necróticas en las hojas y tallos además de tumores en éstos. El primer aislamiento de estas cepas se produjo en Estados Unidos en 2010 (Putnam et al., 2010) y se han descrito casos en Francia (Eltlbany et al., 2012), Alemania (Eltlbany et al., 2012) y Eslovenia (Pirc et al., 2014). Durante el desarrollo de esta Tesis Doctoral se detectaron plantas con los síntomas típicos de la necrosis bacteriana de la dipladenia al sur de España, procediéndose al aislamiento e identificación del agente causal (Caballo-Ponce & Ramos, 2016). En este capítulo se trabajó con una colección de aislados procedente de cada uno de los países donde se ha descrito la enfermedad. Anteriormente se ha sugerido que los aislados de P. savastanoi patógenos de dipladenia (denominados aquí Psd) podrían haberse originado a partir de plantaciones infectadas de adelfa en el sur de Francia (Eltlbany et al., 2012). En esta Tesis Doctoral se construyó un árbol filogenético basado en las secuencias parciales de cuatro genes housekeeping que confirmó la identificación de los Psd como P. savastanoi. Además, la agrupación de los Psd con cepas de P. savastanoi pv. nerii en una sub-rama dentro de la rama correspondiente a la especie P. savastanoi apoya la relación entre ambos. Para caracterizar el rango de huésped de los Psd se llevaron a cabo inoculaciones en olivo, adelfa, fresno, dipladenia y retama. Los resultados mostraron que todos los aislados de dipladenia son patógenos, además Resumen 36 de en dipladenia, en olivo y fresno, aunque la sintomatología generada en éste es variable. Ninguna de las cepas de los patovares savastanoi, fraxini y retacarpa inoculadas en dipladenia desarrollaron síntomas. Por el contrario, la sintomatología generada por las cepas del patovar nerii en dipladenia variaba desde la formación de un tumor a la ausencia de síntomas. Estos resultados indican que las cepas de P. savastanoi aisladas de dipladenia constituyen un nuevo patovar dentro de la especie P. savastanoi y apoyan la relación entre las cepas Psd y P. savastanoi pv. nerii. Se procedió a una caracterización a nivel molecular de los Psd en base al genotipo del gen iaaL y el perfil de plásmidos nativos. Para estudiar el genotipo del gen iaaL se empleó una técnica diseñada para distinguir entre diferentes alelos del gen empleando PCRRFLP (Matas et al., 2009). El patrón de bandas obtenido en todos los PSDs fue idéntico, lo que contrasta con la variabilidad observada entre cepas diferentes de P. savastanoi pv. savastanoi (Matas et al., 2009). De manera similar, el número y tamaño de los plásmidos nativos era muy parecido entre diferentes P. savastanoi aisladas de dipladenia, mientras que el perfil plasmídico en cepas de P. savastanoi pv. savastanoi varía de un aislado a otro (Perez-Martinez et al., 2008). Teniendo en cuenta que en el año 2010 se describió por primera vez la enfermedad y que los Psd son muy parecidos en las características moleculares analizadas en esta Tesis Doctoral se podría especular con que la generación de estas cepas ocurrió recientemente y no ha transcurrido el tiempo suficiente para originar variabilidad entre ellas. Dado que el contenido de plásmidos nativo es similar entre los diferentes PSDs y que el aislado P. savastanoi Ph3 contiene una copia plasmídica del gen iaaM (Eltlbany et al., 2012), decidimos generar una cepa curada de este plásmido siguiendo un método descrito previamente (Comai & Kosuge, 1980). Conseguimos curar la copia plasmídica del gen iaaM, pero no el plásmido que contiene dicha copia (llamado aquí pIAAM). Los resultados sugieren que se produjo una reorganización de los plásmidos que condujo a la curación de otro plásmido (denominado pPh3A) y no del pIAAM. En cualquier caso, la cepa obtenida, a la que llamamos CRpiaaM, mostró una atenuación severa de la virulencia y la supervivencia en plantas de dipladenia, además de la producción de IAA. Sin embargo, CRpiaaM produce cantidades pequeñas de IAA que aumentan cuando el medio se suplementa con triptófano. Este resultado refuerza la hipótesis propuesta por nuestro grupo de investigación de la existencia de una ruta de biosíntesis de IAA alternativa a la de la 3indolacetamida en cepas de P. savastanoi (Aragon et al., 2014) General introduction General introduction 39 1. The Pseudomonas syringae complex The Pseudomonas syringae complex includes a great number of Gram (-) phytopathogenic bacteria with a special significance from an economic, agricultural and scientific point of view. The resurgence of ancient plant diseases (as the bacterial speck of tomato caused by P. syringae pv. tomato) and the continuous outbreak of new infections worldwide (for instance, the bleeding canker of horse chestnut caused by P. syringae pv. aesculi) placed the P. syringae complex on the top of the ranking of plant-pathogenic bacteria (Mansfield et al., 2012). Strains from the P. syringae complex show high variability in their epiphytic survival and host range, including many economically important crops, woody plants and weeds, such as the model plant Arabidopsis thaliana. Moreover, P. syringae infections cause high diversity of symptoms, including spotting and necrosis on leaves, fruit rooting and knots and cankers on stems (Fatmi et al., 2008). Besides plant–associated bacteria, the P. syringae complex includes strains isolated from other diverse habitats such as snowmelt waters, irrigation canals or epilithic biofilms (Morris et al., 2007; Morris et al., 2010; Berge et al., 2014). Based on the host range, the P. syringae complex currently encompasses over 60 different pathovars and 10 species (Young, 2010; Parkinson et al., 2011), which have been classified following two diverse methodologies. On one hand, bacterial strains were sorted in nine genomospecies by DNA-DNA hybridizations, although some incongruence in the nomenclature appeared (Gardan et al., 1999). For instance, genomospecie 2 includes Pseudomonas savastanoi, Pseudomonas ficuserectae, Pseudomonas meliae and Pseudomonas amygdali type strains and, therefore, should be considered synonyms and the correct name should be P. amygdali, as it was the earliest recorded (Gardan et al., 1999). However, the scientific community refers to these strains as separate species. On the other hand, Multi Locus Sequence Typing (MLST) classified the strains of the P. syringae complex in phylogroups (PGs). Initially four different groups were identified (Sarkar & Guttman, 2004) that have been sequentially widened to 13 different phylogroups, which in turn split in several clades (Berge et al., 2014). The classification in phylogroups correlates to genomospecies (Fig. 1), although small differences are found (Parkinson et al., 2011; Berge et al., 2014). General introduction 40 General introduction 47 Figure 4. Typical symptoms of olive knot disease. White arrows in (A) mark knots induced on naturally infected plants. Symptoms induced by artificial inoculation of P. savastanoi pv. savastanoi NCPPB 3335 on (B) woody and (C) micropropagated (non-woody) olive plants. 3.1 Production of phytohormones The development of knots on olive and oleander plants by P. savastanoi strains depends on bacterial production of the phytohormones indole 3-acetic acid (IAA) and cytokinins (CKs) (Magie & Wilson, 1962; Smidt & Kosuge, 1978; Surico et al., 1985; Iacobellis et al., 1994; Aragon et al., 2014). The biosynthesis of IAA in P. savastanoi occurs via the indole 3-acetamide pathway (Fig. 5), where tryptophan is initially converted into indole 3-acetamide by a tryptophan 2monooxygenase (encoded by the iaaM gene). Indole 3-acetamide is further transformed into IAA by the enzyme indole 3-acetamide hydroxylase (encoded by the iaaH gene) (Magie et al., 1963; Kosuge et al., 1966). Both iaaM and iaaH are co-transcribed in an operon, generally located on plasmids in Psn strains or in the chromosome in Psv (Comai & Kosuge, 1980; Caponero et al., 1995; Perez-Martinez et al., 2008). Two chromosomal copies of the iaaMH operon (named iaaMH-1 and iaaMH-2) were found in the genome of Psv NCPPB 3335 (Rodriguez-Palenzuela et al., 2010); however, only iaaMH-1 contributes to the synthesis of IAA (Aragon et al., 2014). A mutation in the iaaMH-2 operon did not affect the virulence of Psv NCPPB 3335 on olive trees, whereas a mutant strain in iaaMH-1 and a double mutant affected in both iaaMH operons (named iaaMH-1.2) induced identical attenuation of knot size in comparison to the wild type strain (Aragon et al., 2014). Interestingly, small amounts of IAA are produced by the iaaMH-1 and the iaaMH-1.2 mutants of Psv NCPPB 3335 (Aragon et al., 2014), suggesting that an alternative pathway for the biosynthesis of this phytohormone should be present in P. savastanoi. Additionally, P. savastanoi strains carry the iaaL gene, which in Psn has been shown to encode a protein that conjugates the IAA to lysine, leading to a less active IAA derivative (Hutzinger & Kosuge, 1968). This gene has been demonstrated to be involved in the virulence of Psn; however, iaaL mutants of Psn strains have been reported either not to cause General introduction 48 gall symptoms on oleander (Glass & Kosuge, 1988) or to be hypervirulent (Cerboneschi et al., 2016). On the other hand, an iaaL mutant of P. syringae pv. tomato DC3000 has been reported to be hypovirulent in tomato plants (Castillo-Lizardo et al., 2015). IAA not only favours the formation of the knot, but also acts as a signal molecule. It was recently demonstrated that the addition of IAA to a culture of Psv NCPPB 3335 resulted in reduction of the expression of two type three secretion system (T3SS) genes, as well as in an increased transcription of a type VI secretion system gene (Aragon et al., 2014). Figure 5. Metabolism of the auxin indole 3-acetic acid in P. savastanoi strains. IaaM, tryptophan 2monooxigenase; IaaH, indole 3-acetamide hyrolase; IaaL, 3-indole-acetyl-ε-L-lysine synthase. Production of cytokinins is distributed among knot-forming phytopathogenic bacteria (Morris, 1986). Initially, an isopentenyl transferase enzyme converts adenosine monophosphate into isopentenyl adenine and isopentenyl adenosine. Later, both molecules are hydroxylated to generate trans-zeatin and trans-zybosilzeatin. Only the ptz gene (Pseudomonas trans-zeatin) has been described in P. savastanoi (Powell & Morris, 1986), which is located in both plasmids and chromosome of olive and oleander isolates (Powell & Morris, 1986; Iacobellis et al., 1994; PerezMartinez et al., 2008). Psv NCPPB 3335 carries a copy of the ptz gene in one of its three native plasmids (pPsv48A). Curing of pPsv48A led to an attenuation of the virulence in Psv NCPPB 3335, but did not affect the growth of the strain within the plant tissues (Bardaji et al., 2011). 3.2 Type III secretion system The T3SS is a molecular syringe that allows the translocation of proteins, named effectors, from the bacterial cytoplasm to the plant cell, where interference with cellular processes might help the pathogen to establish a disease (Büttner & He, 2009). Knot formation critically depends on the functionality of the T3SS (Sisto et al., 2004; Perez-Martinez et al., 2010; Matas et al., 2012). A Psv ITM 317 miniTn5-insertion mutant in the hrcC gene (coding for a structural element of the T3SS) was impaired in knot development. Similarly, a knockout mutant of the hrpA gene (coding for the component of the T3SS pilus) showed a drastic attenuation of the virulence on woody olive trees compared to the parental strain Psv NCPPB 3335 (Perez-Martinez et al., 2010). Not only structural elements of the T3SS play a role in the virulence of P. savastanoi: a miniTn5- General introduction 49 insertion mutant in the hrpR gene, coding for a positive regulator of the T3SS in P. syringae (Hutcheson et al., 2001), was considered non-virulent compared to the wild type Psv NCPPB 3335 (Matas et al., 2012). In addition, a Psv NCPPB 3335 derivative affected in the positive regulatory gene hrpL was described not to induce knot formation in olive (Matas et al., 2014). Neither the hrpA nor the hrpR GFP-tagged mutant strains of Psv NCPPB 3335 colonize the plant hypertrophied tissue generated during the infection nor create the internal cavities observed in the knots developed by Psv NCPPB 3335. (Perez-Martinez et al., 2010; Matas et al., 2012). T3SS effectors are important during different stages of pathogen-host interactions as they (i) promote the penetration and persistence of the pathogen in the host tissue; (ii) suppress the plant defence responses; and (iii) contribute to the access to nutrients, proliferation and growth of the pathogen (Gohre & Robatzek, 2008). Although the concrete function of the majority of T3SS effectors in the interference with plant defences remains unknown, the T3SS effector repertoire of phytopathogenic bacteria is one of the most relevant factors in determining the host range (Baltrus et al., 2011). The genome of Psv NCPPB 3335 contains 33 putative effectors of the T3SS; most of which are encoded in the chromosome and only two (hopAF1-1 and hopAO1) are plasmid-encoded (Bardaji et al., 2011; Matas et al., 2014). Recent studies from our laboratory have demonstrated the translocation of nine of these effectors through the T3SS and their interference with responses associated with plant immunity (Matas et al., 2014; Castaneda-Ojeda et al., 2016). 3.3 Metabolism of c-di-GMP The modified nucleotide cyclic-di-guanylate (c-di-GMP) is a relevant secondary messenger in bacteria that mainly regulates the transition between sessile and planktonic lifestyles. In particular, c-di-GMP controls swimming and swarming motilities, synthesis of the extracellular matrix components and exopolysaccharide production (Zogaj et al., 2001; Solano et al., 2002; Christen et al., 2006; Recouvreux et al., 2008; Pérez-Mendoza et al., 2011). This second messenger is synthesized from two molecules of guanosine triphosphate (GTP) by diguanylate cyclases and is degraded by phosphodiesterases. A characteristic GGDEF catalytic domain is found in diguanylate cyclases (Paul et al., 2004), whereas EAL (Christen et al., 2005; Tischler & Camilli, 2005) or HD-GYP (Ryan et al., 2007) are the typical domains of phosphodiesterases. The genome of Psv NCPPB 3335 contains 34 proteins with a GGDEF domain, of which 14 also encode for an EAL domain (Rodriguez-Palenzuela et al., 2010). Results from our laboratory demonstrated that a mini-Tn5 insertion mutant in a phosphodiesterase (BifA) exhibited a reduced virulence in comparison to Psv NCPPB 3335 (Matas et al., 2012). Recently, a deeper General introduction 50 analysis of the BifA protein confirmed its involvement in the virulence and fitness of Psv in olive plants, as well as in controlling swimming motility (Aragon et al., 2015a). The diguanylate cyclase DgcP is Pseudomonads-specific and has been shown to control biofilm formation, swimming motility and virulence not only in Psv NCPPB 3335, but also in Pseudomonas aeruginosa PAK (Aragon et al., 2015b). In addition, the DgcP-mediated modulation of c-di-GMP is related with the regulation of the type VI secretion system-related genes in both Psv NCPPB 3335 and P. aeruginosa PAK (Aragon et al., 2015b). 3.4 The quorum sensing system Bacteria modulate their behaviour in a cell density-dependent manner through a mechanism named quorum sensing. Gram (-) bacteria synthesize a signal molecule, commonly N-acyl-homoserine lactones (AHLs), by the action of a LuxI-family protein. AHLs diffuse through the plasma membrane and accumulates in the bacterial environment. Upon reaching a threshold concentration of AHLs (the quorum), a LuxR-family protein binds to its cognate AHL and modulate the expression of target genes (Fuqua et al., 2001; Whitehead et al., 2001) (Fig. 6). Several transcriptomic analysis have concluded that the presence of AHLs regulates the expression of a vast number of genes in plant-associated bacteria (Coutinho et al., 2013; Kim et al., 2013; Kim et al., 2014). Interestingly, a mutant in the luxI homolog of the tobacco pathogen P. syringae pv. tabaci, closely related to P. savastanoi, showed over 200 genes which expression differed compared to the wild type strain (Taguchi et al., 2015). The quorum sensing system controls traits related with the virulence and epiphytic fitness of phytopathogenic bacteria (Von Bodman et al., 2003). In this regard, this regulatory network is crucial for the virulence of Psv DAPP-PG 722 , as mutant strains in the luxI and luxR homologs induced the generation of knots significantly smaller than those generated in plants infected with the wild type strain (Hosni et al., 2011). However, genes which expression is under the control of the quorum sensing system in Psv remain to be elucidated. General introduction 51 Figure 6. The quorum sensing regulation at (A) low cell densities and (B) high cell densities. LuxI synthesizes the signal molecule (AHL) that binds to LuxR at high cell densities and modulate the expression of target genes. Picture adapted from Witehead et al. (2001). 3.5 Other virulence and adaptation factors Previous work in our laboratory following a signature-tagged mutagenesis (STM) strategy identified 58 genes necessary for the full virulence and growth of Psv on olive trees (Matas et al., 2012) (Fig. 7). Genes required for the synthesis of nine amino acids and three vitamins were identified, as well as genes coding for transporters of compounds present in the plant apoplast, such as citrate, sulphate and glutamate. Aside from the hrpR gene mentioned above (see section “Type III secretion system”), other factors identified include genes of the types II and IV secretion systems, as well as proteins taking part in stress tolerance or the synthesis of the cell wall, the phytohormone IAA and the second messenger c-di-GMP. Figure 7. Schematic representation of virulence-associated mechanisms in Pseudomonas savastanoi pv. savastanoi identified by signature-tagged mutagenesis. Functional categories are represented in different General introduction 52 colors: blue, secretion systems; green, cell-surface structures; orange, stress tolerance; gray, transporters; and red, others. OM, outer membrane; PG, peptidoglycan; IM, inner membrane; T3SS, type III secretion system; T4SS, type IV secretion system; MCP, methyl-accepting chemotaxis protein; GGDEF, GGDEF/EAL domain protein-coding gene; IAA, indole 3-acetic acid. Picture from Matas et al. (2012). The degradation of aromatic compounds is also crucial for the virulence and fitness of Psv NCPPB 3335 on olive knots. A Psv NCPPB 3335 catJ mutant, impaired in the final steps of the βketoadipate pathway of aromatic compounds degradation (Fig. 8), induced the generation of smaller knots compared with the wild type strain and was severely impaired in its competitive growth (Matas et al., 2012). Figure 8. Schematic representation of the β-Ketoadipate pathway for the bacterial degradation of aromatic compounds into intermediaries of the Krebs cycle. Genes coding for the enzymes of this pathway are shown next to the arrows. Picture adapted from Harwood & Parales (1996). General introduction 53 The comparison of the genome of Psv NCPPB 3335 with other P. syringae strains infecting herbaceous host unveiled a region of about 15 kb unique to Psv that encodes for genes likely related with the metabolism of aromatic compounds (Rodriguez-Palenzuela et al., 2010). This region, originally named VR8 and referred as the WHOP (from woody host and Pseudomonas) region in this PhD Thesis, is shared with other strains of the P. syringae complex isolated from woody hosts, and is absent in P. syringae strains pathogenic on non-woody hosts (Ramos et al., 2012) (Fig. 9). Among others, the catBCA operon, which is implicated in the degradation of catechol through the β-Ketoadipate pathway (Fig. 8), is located within the WHOP region (Fig. 9). Moreover, this region has been associated with the ability of P. syringae strains to colonize kiwifruit trees (Bartoli et al., 2015a). However, the role of this region in the virulence and adaptation of Psv to olive plants has not been addressed up to date. Figure 9. Schematic map of the WHOP region in the genomes of P. savastanoi pv. savastanoi NCPPB 3335 and other P. syringae strains which genome have been sequenced. (A) P. savastanoi pv. savastanoi NCPPB 3335, P. syringae pathovars aesculi strains 2250 and NCPPB 3681, morsprunorum MAFF 302280 and actinidiae MAFF 302091. (B) P. syringae pathovars tabaci ATCC 11528, mori 301020, phaseolicola 1448A, glycinea race 4, lachrymans MAFF 302278 and japonica MAFF 301072. (C) P. syringae pathovars syringae B728a, tomato DC3000 and oryzae 1_6. Picture adapted from Ramos et al. (2012). Objectives Chapter I 63 function of several genes located within this chromosomal region. Our results broaden the distribution of the WHOP region to other PG1 and PG3 members infecting woody hosts and show its partial conservation in only three PG2 pathovars. Furthermore, we demonstrate that several of the genes located within this region contribute to the fitness and virulence of this bacterium in woody olive plants but not in non-woody olive plants. Chapter I 64 MATERIALS AND METHODS Bacterial strains, media and growth conditions Bacterial strains used in this study are listed in Table 1. Psv and Escherichia coli were grown at 28 °C or 37 °C, respectively, in Luria-Bertani (LB) media (Miller, 1972), Super Optimal Broth (SOB) (Hanahan, 1983) or M9 minimal salts medium with an additional carbon source (Sambrook, 2001). Solid and liquid media were amended with the appropriate antibiotic when required. Antibiotic concentration used were: kanamycin (Km) 10 µg ml-1 for Psv and 50 µg ml-1 for E. coli; gentamycin (Gm) 10 µg ml-1; ampicillin (Ap) 400 µg ml-1 for Psv and 100 µg ml-1 for E. coli; and tetracycline 10 µg ml-1. aKm, kanamycin; Cm, chloramphenicol; Sp, spectinomycin; Tc, tetracycline Construction of bacterial strains and plasmids Plasmids and oligonucleotides used in this study are listed in Tables 2 and 3. For the generation of Psv NCPPB 3335 mutants, DNA fragments of approximately 1 kb, corresponding to the upstream and downstream flanking regions of the gene/operon to be deleted, were amplified in independent three rounds of polymerase chain reaction (PCR) using genomic DNA from NCPPB 3335 as a template and Expand High Fidelity polymerase (Roche Applied Science, Mannheim, Germany). Restriction sites for EcoRI (for the plasmid pECP1), BamHI (for the plasmid pIMC) or HindIII (for the plasmids pECP2, pECP3 and pECP4) were included in the primers as described previously (Matas et al., 2014). The resulting products, consisting of Table 1. Strains used in this study Strain Relevant characteristicsa Source P. savastanoi pv. savastanoi NCPPB 3335 Wild type strain (Perez-Martinez et al., 2007) ΔcatB catB (PSA3335_3199) deletion mutant (KmR) This work ΔantA antA (PSA3335_3204) deletion mutant (KmR) This work Δ06 PSA3335_3206 deletion mutant (KmR) This work ΔipoABC ipoABC operon deletion mutant (KmR) This work ΔdhoAB dhoAB operon deletion mutant (KmR) This work Escrherichia coli DH5α F-, ϕ80dlacZ M15, (lacZYA-argF) U169, deoR, recA1, endA, hsdR17 (rk – mk-), phoA, supE44, thi-1, gyrA96, relA1 (Hanahan, 1983) GM2929 F-, ara-14, leuB6, thi-1, tonA31, lacY1, tsx-78, galK2, galT22, glnV44, hisG4, rpsL136, xyl-5, mtl-1, dam13::Tn9, dcm-6, mcrB1, hsdR2, mcrA, recF143 (SpR CmR) (Palmer & Marinus, 1994) XL1-Blue hsdR17, supE44, recA1, endA1, gyrA46, thi, relA1, lac/ F´ [proAB+ lacZq M15::Tn10 (TcR)] (Bullock et al., 1987) Chapter I 65 upstream and downstream flanking regions separated by the one of the mentioned restriction sites, were cloned into pGEMT and sequenced to discard mutations. Next, the kanamycin resistance gene nptII was extracted by enzyme restriction from pGEM-T-KmFRTEcoRI, pGEMT-KmFRTHindIII or pGEM-T-KmFRTBamHI (Table 2) and cloned in the vectors described above to yield pECP1-Km, pIMC-Km, pECP2-Km, pECP3-Km and pECP4-Km, respectively. Then the kanamycin resistance cassette was cloned in an inverted position with respect to the transcription orientation of each cluster. All the plasmids generated for the construction of Psv NCPPB 3335 mutants were suicide vectors in Psv. Plasmids were transferred to NCPPB 3335 by electroporation (Perez-Martinez et al., 2007) and transformants were selected in LB-Km plates. Plasmids pECP1-Km, pIMC-Km, pECP2-Km, pECP3-Km, and pECP4-Km were used for the construction of ΔcatB, ΔantA, Δ06, ΔipoABC and ΔdhoAB, respectively. To select the allelic interchange (double recombination event) and discard plasmid integration (single recombination event), individual colonies were replicated onto LB-Ap plates and ApR colonies were discarded. Finally, Southern blot analyses were carried out to confirm integration in the correct position in the genome. Plasmids for the complementation of the mutants were generated as follows. The complete open reading frames of catBCA, antABC, ipoABC and dhoAB operons, and the PSA3335_3206 gene together with their corresponding promoter and transcriptional terminator regions were amplified by PCR using NCPPB 3335 genomic DNA as a template and Phusion DNA Polymerase (Thermo Scientific, MA, USA). Then, the fragments were cloned into pBluescript SK II (+) or pENTR/SD/D-TOPO (only for the antABC operon). After sequencing to discard mutations, the fragments were subcloned into pBBR1MCS-5, and the correct orientation was selected by PCR using primers hybridizing with the plasmid and the cloned fragment. Plasmids pBBR:catBCA, pBBR:antABC, pBBR:PSA3335_3206, pBBR:ipoABC and pBBR:dhoAB were used for the complementation of ΔcatB, ΔantA, Δ06, ΔipoABC and ΔdhoAB, respectively For the generation of promoter fusions to lacZ, a fragment of 274 bp upstream from the antA gene was amplified by PCR using NCPPB 3335 genomic DNA as a template and with primers which included the KpnI and BglII restrictions sites. The fragment was A/T cloned into pGEMT, the sequence verified, and then subcloned into pMP220 (Spaink et al., 1987). Chapter I 66 Table 2. Plasmids used in this study Name Descriptiona Source pGEM-T easy Cloning vector containing ori f1 and lacZ (ApR) (Promega, Madison, WI, USA) pBluescript II SK (+) Cloning vector containing ori f1 and lacZ (ApR) (Agilent Technologies, Santa Clara, CA, USA) pENTR/SD/D-TOPO Cloning vector containing pUC ori (KmR). (Invitrogen Corp, CA, USA) pBBR1MCS-5 Broad host-range cloning vector (GmR) (Kovach et al., 1995) pGEM-T-KmFRTEcoRI Contains KmR from pKD4 and EcoRI sites (ApR KmR) (Zumaquero et al., 2010) pGEM-T-KmFRTHindIII Contains KmR from pKD4 and HindIII sites (ApR KmR) (Aragon et al., 2014) pGEM-T-KmFRTBamHI Contains KmR from pKD4 and BamHI sites (ApR KmR) (Ortiz-Martin et al., 2010) pIMC pGEMT-derivative containing 1kb on each side of the antA (PSA3335_3204) gene from NCPPB 3335 (ApR). This work pIMC-Km pGEMT-derivative containing 1kb on each side of the antA (PSA3335_3204) gene from NCPPB 3335 interrupted by the kanamycin resistance gene nptII (ApR, KmR). This work pECP1 pGEMT-derivative containing 1kb on each side of the catB (PSA3335_3199) gene from NCPPB 3335 (ApR). This work pECP1-Km pGEMT-derivative containing 1kb on each side of the catB (PSA3335_3199) gene from NCPPB 3335 interrupted by the kanamycin resistance gene nptII (ApR, KmR). This work pECP2 pGEMT-derivative containing 1kb on each side of the PSA3335_3206 gene from NCPPB 3335 (ApR). This work pECP2-Km pGEMT-derivative containing 1kb on each side of the PSA3335_3206 gene from NCPPB 3335 interrupted by the kanamycin resistance gene nptII (ApR, KmR). This work pECP3 pGEMT-derivative containing 1kb on each side of the PSA3335_3207 to PSA3335_3209 operon from NCPPB 3335 (ApR). This work pECP3-Km pGEMT-derivative containing 1kb on each side of the PSA3335_3207 to PSA3335_3209 operon from NCPPB 3335 interrupted by the kanamycin resistance gene nptII (ApR, KmR). This work pECP4 pGEMT-derivative containing 1kb on each side of the PSA3335_3210 to PSA3335_3211 operon from NCPPB 3335 (ApR). This work pECP4-Km pGEMT-derivative containing 1kb on each side of the PSA3335_3210 to PSA3335_3211 operon from NCPPB 3335 interrupted by the kanamycin resistance gene nptII (ApR, KmR). This work pBBR:catBCA pBBR1MCS-5-derivative containing the Psv NCPPB 3335 catBCA and catBCA promoter region (263 bp) flanked by EcoRI restriction sites (GmR) This work Chapter I 67 aAp, ampicillin; Tc, tetracycline; Gm, gentamycin; Km, kanamycin Table 3. Oligonucleotides used in this study Name Sequence (5’3’) RT-PCR catA_R46 ACTGTTGCACATCGTCGG catB_F1012 CGAAGACATTCTGGTCGAGC catB_R221 GACAAGAAGCGGAGCGAAC antC_F874 CGGTTACATCGTCGATCACC antC_R224 TGGCAGCTCAGCACCTTG antB_F282 CACGCTGCACCTGATCAGC antB_R23 GGTACAGGAACTGCTCGATGC antA_F904 GCCTGGTGGAGGAATACG 00_R120 ATCGCGAGCGTAACCATG 01_F771 CTTCGCGACTGAGTATGCG 01_R352 GTTCGACAGGTAGCTGAGCTG 02_F378 CTGCTCGATGACGTGGTC 02_R251 GCACCGAGCACGTTGATC 03_F1059 GGAGCAGACCTTCGAGCAG 03_R185 CTGCAGCGAGGTGTTGAAC 04_R645 GGCTTCGTATCACCTGCTG 04_R118 CGTCAATGTCGAACAGCG 05_F1104 CAGCGAGATCGACTGGATG Construction of the mutants strains TA-antA-R-13-mut CCCTATAGTGAGTCGGATCCGTCATCAGCCTTGAACTG TA-antR-R879 ATGTTGTAGCCACTGTGC TD-antB-F6-mut GGATCCGACTCACTATAGGGTGAGGTCTGAACATGAGC Table 2. Plasmids used in this study (continuation) Name Descriptiona Source pBBR:antABC pBBR1MCS-5-derivative containing the Psv NCPPB 3335 antABC operon and its promoter region (241 bp) flanked by EcoRI restriction sites (GmR) This work pBBR: PSA3335_3206 pBBR1MCS-5-derivative containing the Psv NCPPB 3335 PSA3335_3206 gene and its promoter region (168 bp) flanked by HindIII restriction sites (GmR) This work pBBR:ipoABC pBBR1MCS-5-derivative containing the Psv NCPPB 3335 PSA3335_3207 to PSA3335_3209 operon (ipoABC) and its promoter region flanked by HindIII restriction sites (187 bp) (GmR) This work pBBR:dhoAB pBBR1MCS-5-derivative containing the Psv NCPPPB 3335 PSA3335_3210 to PSA3335_3211 operon (dhoAB) and its promoter region (381 bp) flanked by HindIII restriction sites (GmR) This work pMP220 Broad-host-range, low-copy-number promoter probe vector, IncP replicon, lacZ (TcR) (Spaink et al., 1987) pMP220-PantABC Contains a fragment of 274 bp corresponding with the antABC promoter from Psv NCPPB 3335 directionally cloned with KpnI and BglII restriction sites (TcR). This work Chapter I 68 Table 3. Oligonucleotides used in this study (continuation) Oligonucleotide Sequence (5’3’) TD-antC-R673 CGATCTGATCGAGCATGC TA-catB_R17 GAATTCGACTCACTATAGGGATCGTGATGGCACTCATGG antC_F92 GCTGGATGCTGCGTTACG TA-catB_R17 GAATTCGACTCACTATAGGGATCGTGATGGCACTCATGG antC_F92 GCTGGATGCTGCGTTACG TD-catB_F1144 CCCTATAGTGAGTCGAATTCGCCGGAGCACAAGACTAGC catA_R659 ACGTTGACCATGACGACCC TD-00_F1573 CCCTATAGTGAGTCAAGCTTGATCGTTAGCGGTTACGGC antR_R853 TGGAATTCGAACATTCGGC TA-00_R51 AAGCTTGACTCACTATAGGGGACCTTTGAGGTTGGTCCG 01_F28 GCAACTGATGTGTGTTTGCG TD-01_R753 CCCTATAGTGAGTCAAGCTTACTGGCTCAAGCAGATCACC 00_R773 GTCCGATCAGTTCATGCAGC TA-03_R416 AAGCTTGACTCACTATAGGGCAGCAGCACCAGATCATGG 04_R281 ACCTGCTGATCAACCTGGC TD-04_F673 CCCTATAGTGAGTCAAGCTTGCCAACGTCGTGACTTTCC 03_R855 CATACCCAACACCTGACGTCC TA-05_R208 AAGCTTGACTCACTATAGGGCTGTAACCCAGCTCGACGC benR_R507 TAGCTCGCGTTAGGATGGG Complementation of the mutant strains antC_R1360 ACTCATGGGAATTCTGGTCAGCGAAGAGCAGG catA_R1305 ACTCATGGGAATTCTGATCGTTCCTCACGCTCC antR_R13 CACCGAATTCCAAGTCGTTGGGCAAAACAG catB_R190 GAATTCGTTCACCTTGATGCTTTCCG antR_F349 CACCGGATCCGATCAGTTCCATCTCCAGATCG antR_R-1196 GGATCCCAGGTGGAGGGTTTGAGGAG antR_F900 ACTCATGGGGTACCACAGTGGCTACAACATCGCG 01_F871 ACTCATGGGGTACCGTTCGTCAACGCGTACTTCG 04_F720 ACTCATGGGGTACCAGCAGCCAGCTTTGTCACC 00_R126 ACTCATGGGGTACCCTCATCGCGAGCGTAACC benR_R60 ACTCATGGGGTACCGTTCTGCGCTGTAAACAGCG 03_R1029 ACTCATGGGGTACCGCCGATACCCTCGACAGC 5’ RACE antA_R68 CCGGGCAATGCGGTACACACCTTC catB_R-20 GGCACTCATGGGGGTTCCGTCTG antR_R34 GGTGCGCGGCATGAAGGCCATG RT-qPCR gyrA-R TTCCAGTCGTTACCCAGCTCG gyrA-F GACGAGCTGAAGCAGTCCTACC catB_F146 TGACGGCATTATCGGTATCG catB_R248 CGTTGAGGTTGGTGGCG Cloning of the antABC and antR promoters BglII-antR_R12 AGATCTAGTCGTTGGGCAAAACAGC KpnI-antA_R17 GGTACCTGCTGCCATTGGCTTACG BglII-antA_R17 AGATCTTGCTGCCATTGGCTTACG KpnI-antR_R12 GGTACCAGTCGTTGGGCAAAACAGC Chapter I 69 Distribution of the WHOP region across the Pseudomonas syringae complex Genome sequences of 96 strains belonging to the P. syringae complex, 49 (51%) of which correspond to strains isolated from woody hosts, were downloaded from GenBank and the presence of the WHOP region was analysed by blastn using Geneious software v7.1.13 (http://www.geneious.com) and the nucleotide sequence of the Psv NCPPB 3335 WHOP regions as template. Positive hits were determined using the “Grade” value, a percentage calculated by Geneious combining the query coverage, e-value and identity values for each hit with weights 0.5, 0.25 and 0.25, respectively. Genomes yielding a Grade value ≥ 90% were considered to encode a complete WHOP region. Genomes yielding pairwise alignments with > 70% identity to specific clusters, were considered to encode a partial WHOP region. RNA extraction and RT-PCR Psv NCPPB 3335 cells were grown in LB broth to OD600 2, pelleted and stored at -80 °C. Total RNA was then extracted using the TriPure Isolation Reagent (Roche Applied Science, Mannheim, Germany) according to the manufacturer’s instructions, except that the TriPure was preheated at 65 °C, the lysis step was performed at 65 °C, and 1-bromo-3-chloropropane (Molecular Research Center, Cincinnati, OH, U.S.A.) was used instead of chloroform as previously described (Matas et al., 2014). The RNA concentration was determined spectrophotometrically and its integrity was assessed by agarose gel electrophoresis. Total RNA was treated with a TURBO DNAfree™-Kit (Applied Biosystems; CA, USA) as detailed by the manufacturer’s instructions. Later the samples were tested for genomic contamination by PCR. DNA-free RNA was reverse transcribed using random hexamers included in the iScript™ cDNA synthesis kit (BioRad; CA, USA). 50 ng of cDNA were used as a template to amplify intergenic regions by PCR using GoTaq Polymerase (Promega; Madison, USA). PCR products were analysed by 1% agarose gel electrophoresis. Primers used for RT-PCR are listed in Table 3. Promoter activity and identification To measure antABC promoter activity two different β-galactosidase assays were performed. In a first experiment Psv carrying pMP220-PantABC were grown to OD600 0.5 in LB broth, cultures were then split into two different flasks and 1 mg/ml (7.3 mM) of anthranilate was added to one of them. β-galactosidase activity was measured 2, 4 and 19 hours postaddition as previously described (Miller, 1972). In a second experiment, bacteria were grown in Chapter I 70 LB or LB amended with 7.3 mM of anthranilate or 6-chloroantranilate and β-galactosidase activity was measured after 24 h. To map the transcription start point of antA, antR, catB and catA Psv NCPPB 3335 cells were grown on LB to OD600 0.5 and anthranilate was then added to a final concentration of 7.3 mM. After 4 hours bacteria were pelleted and stored at -80 °C for further RNA extraction. SMARTer RACE kit (Clontech, Mountain View, CA, USA) was used according to the manufacturer’s instructions. RT-qPCR assays Psv and the ΔantA mutant were grown in LB to OD600 0.5 and anthranilate was then added to a final concentration of 7.3 mM. After 4h bacteria were pelleted, frozen in liquid nitrogen and RNA was subsequently extracted. RNA extraction and cDNA synthesis were carried out as described above. Bacteria grown in unamended LB were used as a control. The primer efficiency tests, reverse transcription quantitative PCR (RT-qPCR) and confirmation of the specificity of the amplification reactions were performed as described previously (Vargas et al., 2011). The relative transcript abundance was calculated using the ΔΔ cycle-threshold (Ct) method (Livak & Schmittgen, 2001). Transcriptional data were normalized to the housekeeping gene gyrA and are presented as the fold change of the expression in LB amended with anthranilate relative to the expression in unamended LB. The relative expression ratio was calculated as the difference in qPCR threshold cycles (ΔCt = Ctgene of interest – CtgyrA). One PCR cycle represents a twofold difference in template abundance; therefore, fold-change values were calculated as 2-ΔΔCt as previously described (Pfaffl, 2001; Rotenberg et al., 2006). Identification and quantification of aromatic compounds by high performance liquid chromatography Separation and identification of aromatic compounds of interest and related compounds were performed by high performance liquid chromatography (HPLC) using a HP series 1050 chromatograph (Hewlett-Packard, Waldbronn, Germany) equipped with a diode array UV-Vis detector and a 5-µm C18RP column (Novapak C18, 150×3.9 mm, Waters S.A., Barcelona). Compounds were identified by comparison of their respective UV absorption spectra and retention times with commercial pure standards (Sigma Aldrich, San Luis, MO, USA). Chapter I 71 To detect anthranilate, 6-chloroanthranilate and catechol, samples were run for 20 minutes at a flow rate of 1 ml/minute with the detector set at 210 and 230 nm. Eluents used were 0.1% ortho-phosphoric acid in deionized water and methanol in a gradient of concentrations: 5% to 25% methanol for 15 minutes, 25% to 5% methanol for 1 minute and a final step for column reequilibration at 5% methanol for 5 minutes. Under these conditions, the retention times (Rt) for anthranilate, catechol and 6-chloroanthranilate were 9.72, 7.61 and 12.9 minutes, respectively. For the functional analysis of the antABC operon, bacteria were grown overnight in LB broth, washed twice in M9 medium and resuspended in M9 minimal medium amended with 5 mM succinate and 500 µM anthranilate to an approximate final concentration of 109 cfu/ml. After 24h of incubation at 28 °C cells were removed by centrifugation, the supernatant was then filtered (pore diameter 0.22 µm) and analysed by HPLC. An identical procedure was carried out for the analysis of 6-chloroanthranilate transformation, except that 500 µM of this compound was added to the M9 minimal media instead of anthranilate. For the functional analysis of the catBCA operon, the same procedure was followed, with the exception that 7.3 mM of anthranilate was added to the overnight culture of LB broth and 300 µM of catechol to the M9 minimal media instead of anthranilate. As a control, cell-free media were subjected to the same procedure. For the quantification of anthranilate and catechol in the media, calibration curves with pure standards were performed in which the concentration was compared to the area of each peak. For 4-chlorocatechol detection, an isocratic protocol consisting of 30% methanol and 70% deionized water with 0.1% ortho-phosphoric acid was used. The flow rate was 1 ml/minute during 15 minutes and the detector was set at 210 and 230 nm. The retention time of 4chlorocatechol was 5.05 minutes. Samples were prepared as described above for the analysis of the catBCA function with the exception that 300 µM 4-chlorocatechol was added to the M9 minimal media instead of catechol. For indigo detection, the HPLC protocol consisted of an isocratic method with a mixture of 90% acetonitrile and 10% deionized water and a flow rate of 1 ml/min. The running time was 5 minutes and the detector was set to 290 nm. Cells were grown overnight in LB broth, washed twice, resuspended in M9 minimal media amended with 5mM succinate and 500 µM indole and incubated at 28 °C until the culture became blue. When formed, the blue compound was partially purified by centrifugation of 2 ml of the culture. A blue precipitate formed along the inner side of the tube and was then grazed with a scraper, solubilized in acetonitrile and transferred to a new tube without disrupting the cell in the pellet. The partially purified blue Chapter I 72 precipitate was filtered and analysed by HPLC where it was compared to pure standard of indigo (Sigma, San Luis, MO, USA). Plant infection and isolation of bacteria from knots Olive plants were micropropagated, rooted and maintained as previously described (Rodriguez-Moreno et al., 2008). Micropropagated (non-woody) plants were infected in a stem wound with a bacterial suspension (≈5.103 cfu) and incubated in a growth chamber for 30 days as described previously (Rodriguez-Moreno et al., 2008). The morphology of the knots developed was observed with a stereoscopic microscope (Leica MZ FLIII; Leica Microsystems, Wetzlar, Germany). Bacteria were recovered from the knots using a mortar and pestle containing sterile 10 mM MgCl2. Serial dilutions were plated on LB plates or LB supplemented with the corresponding antibiotic and population sizes were determined from three different knots. A minimum of three representative knots were 3D scanned and the knot size determined using the Neftabb Basic 5.2 software. Statistical analyses were performed by analysis of variance (ANOVA) test (α=0.05). The pathogenicity of Psv was also analysed on 1-year old (woody) olive explants as previously described (Penyalver et al., 2006; Perez-Martinez et al., 2007; Matas et al., 2012). Morphological changes scored at 90 dpi were captured with a high-resolution camera Canon D6200 (Canon Corporation, Tokyo, Japan). The knot volume was calculated from a minimum of three representative knots as described previously (Moretti et al., 2008; Hosni et al., 2011). Statistical analyses were performed by analysis of variance (ANOVA) using the software SigmaPlot 12.0. Competitive assays between the wild type and mutant strains were carried out in both woody and non-woody plants as described previously (Rodriguez-Moreno et al., 2008; Matas et al., 2012). Bacteria were mixed in a 1:1 ratio and 5.103 cells were place on each inoculation point. Chapter I 79 Figure 2. Phylogenetic analysis of the catBCA operon in the P. syringae complex. The tree was constructed using MEGA5 (Tamura et al., 2011) with the maximum-likelihood method (Jones et al., 1992) .The length of the sequence used for phylogenetic analysis was 2412 bp. P. aeruginosa PAO1 and P. resinovorans NBRC 106553 are included as outsiders. PG, phylogroup The WHOP region is organized in four operons and two independently transcribed genes Fourteen ORFs are encoded in the WHOP region (Fig. 1A, Table 5). PSA3335_3199 to PSA3335_3201 and PSA3335_3202 to PSA3335_3204 encode, respectively, homologs of the catBCA and antABC operons found in other pseudomonads (Houghton et al., 1995; Nojiri et al., 2002; Maeda et al., 2003; Urata et al., 2004; Li et al., 2010). The antABC operon participates in the conversion of anthranilate into catechol, which is further transformed into β-ketoadipate enol-lactone by the catBCA-encoded activity. PSA3335_3205 (antR) and PSA3335_3213 (benR), which are the only genes oriented in the opposite direction to all others (Fig. 1A), encoded the positive transcriptional regulators of the antABC and benABCD operons, respectively. The benABCD operon is involved in the degradation of benzoate to catechol in Pseudomonas putida (Jeffrey et al., 1992), however, this operon is not present in the genome of Psv NCPPB 3335. Annotation of the PSA3335_3207 to PSA3335_3211 genes yielded functions likely to be involved also in the metabolism of phenolic compounds (Rodriguez-Palenzuela et al., 2010). A search for protein domains using MyHits web tool (http://myhits.isb-sib.ch/) revealed that PSA3335_3206 (annotated as an aerotaxis receptor) might play a role in chemotaxis, since it encodes a protein containing a methyl-accepting domain and a bacterial chemotaxis sensory transducer domain. No specific domain was obtained for PSA3335_3207. The product encoded by this ORF annotates as a protein involved in the meta-pathway of phenol degradation but without a described function. The products of PSA3335_3208 and PSA3335_3209 contain two structural motifs characteristic of iron-sulphur proteins (2Fe-2S and 4Fe-4S centres) and an NADP-dependent oxidoreductase domain, respectively. The product of PSA3335_3211 (named dhoA) carries a dienelactone hydrolase family domain. This family of enzymes has been involved in the metabolism of halogenated catechols in Pseudomonas knackmusii, previously known as Pseudomonas sp. strain B13 (Schlomann et al., 1990). PSA3335_3210, the product of which is annotated as a short chain alcohol dehydrogenase gene (hereafter named dhoB), encodes for a protein showing a KR domain, characteristic of bacterial polyketide synthases and involved in the reductive modification of the ß-carbonyl centres in the growing polyketide chain. Chapter I 80 aAccession number corresponding to the genome of P. savastanoi pv. savastanoi NCPPB 3335. bUN - Unnamed The organization of all genes encoded in the WHOP region into different operons was analysed by retrotranscriptase-polymerase chain reaction (RT-PCR) in Psv NCPPB 3335. Primers targeting sequential ORFs were used to amplify intergenic regions, except for those of the catBCA operon where a region spanning from PSA3335_3199 (catA) to PSA3335_3201 (catB) was amplified (Fig. 1A). Positive controls using genomic DNA as a template were used for all PCRs. As shown in Fig. 1D, a DNA fragment of identical size to those of the corresponding intergenic regions was observed for antC-antB and antB-antC, confirming their organisation as an operon of antA, antB and antC. A similar result was obtained for the region spanning from catB to catA, demonstrating the organisation of catB, catC and catA in operon. No amplification product was obtained for the intergenic region located between the catBCA and the antABC operons, showing that the transcription of antABC and catBCA are independent. Regarding the intergenic regions of PSA3335_3207-PSA3335_3208 and PSA3335_3208-PSA3335_3209, amplification products of the expected sizes were observed in both cases, while no amplification was observed for PSA3335_3206-PSA3335_3207 and PSA3335_3209-PSA3335_3210. These Table 5. Genes located within the WHOP region Accession numbera Demonstrated function Gene b Annotated gene product Catechol metabolism PSA3335_3199 catA Catechol 1,2-dioxygenase PSA3335_3200 catC Muconolactone isomerase PSA3335_3201 catB Muconate cycloisomerase Anthranilate metabolism PSA3335_3202 antC Anthranilate dioxygenase reductase component PSA3335_3203 antB Anthranilate dioxygenase beta subunit PSA3335_3204 antA Anthranilate dioxygenase alpha subunit PSA3335_3205 antR antABC regulatory protein Indigo-producing oxygenase PSA3335_3207 ipoC Involved in meta pathway of phenol degradation PSA3335_3208 ipoB Nitrilotriacetate monooxygenase component B PSA3335_3209 ipoA Putative oxygenase subunit Not determined PSA3335_3206 UN Aerotaxis receptor PSA3335_3210 dhoB Short chain alcohol dehydrogenase PSA3335_3211 dhoA Dienelactone hydrolase PSA3335_3213 benR Positive regulator of the benABCD operon Chapter I 81 results indicate that the transcription of the PSA3335_3206 gene is independent. However, PSA3335_3207 to PSA3335_3209 co-transcribe in an operon, here named the ipoABC operon (indigo producing operon), in relation to the ability of these genes to transform indole into the blue pigment indigo (see below). The intergenic region between PSA3335_3211 (dhoA) and PSA3335_3210 (dhoB) was also amplified, suggesting their organisation into an operon (hereafter called the dhoAB operon). No amplification products corresponding to the intergenic regions located immediately upstream or downstream from this operon were obtained. In summary, the results show that the genes encoded in the WHOP regions are organised into four operons (catBCA, antABC, ipoABC and dhoAB) and three independently transcribed genes (antR, PSA3335_3206 and benR). Transcriptional analysis of the antABC and the catBCA operons The transcriptional start site of the catBCA and antABC operons and that of the antR gene were determined by 5’-RACE and localized, respectively, 16, 31 and 56 nucleotides upstream from their corresponding start codons (Fig. 3). Alignment of the Psv NCPPB 3335 -10 and -35 boxes of the catBCA and antABC promoters revealed partial similarity with the same sequences of other pseudomonads. Figure 3. Determination of the transcription start points of (A) antA, (B) antR and (C) catB genes by 5’- RACE. Translational start codons are represented in capital letters. The proposed -10 and -35 boxes are shown in bold and the proposed -10 and -35 boxes in other pseudomonads are represented below. PAO1; P. aeuruginosa PAO1; MB101; Pseudomonas fluorescens MB101; CA10; P. resinovorans CA10; PRS2000, P. putida PRS2000; Pf-5, P. fluorescens Pf-5. Asterisks indicate conserved nucleotides. Dash indicate a gap to obtain maximal homology. Chapter I 82 In other pseudomonads, the antABC and catBCA operons are involved in the degradation of anthranilate through the β-ketoadipate pathway (Fig. 4A). Anthranilate is known to be an inducer of the antABC operon in P. resinovorans (Urata et al., 2004), P. aeruginosa (Kim et al., 2012) and P. fluorescens. To analyse whether anthranilate also promotes the expression of antABC in Psv NCPPB 3335, a region expanding 274 bp upstream the antA gene was amplified and cloned into pMP220. The resulting plasmid (pMP220-PantABC) was transformed into wild type Psv NCPPB 3335. Psv cells were grown in LB medium to mid-log phase (OD600nm=0.5), the culture was then divided into two and anthranilate was added to one. As a control, Psv cells harbouring the ‘empty’ vector (pMP220) were used. After different periods of time, expression of the antABC promoter (PantABC) was determined by β-galactosidase activity. Four hours after the addition, corresponding to late log phase cells (OD600nm=0.9), the activity of PantABC was eightfold higher in anthranilate-amended medium (Fig. 4B). However, at 19 hours post-addition, corresponding to stationary-grown cells (OD600nm= 2.3), a 27.6 fold increase of PantABC expression in the amended medium was observed, confirming anthranilate as an inducer of the transcription of antABC also in Psv NCPPB 3335. On the other hand, the higher activity of PantABC observed in Psv cells grown to stationary phase in comparison with exponentially-grown cells suggest that the antABC operon might be involved in secondary metabolic pathway(s), which activate during the stationary phase. Other inducers of the expression of antABC have been reported in pseudomonads. In P. fluorescens MB101, 6-chloroanthranilate is a gratuitous inducer, promoting the expression of the operon without being degraded (Retallack et al., 2006). Activation of PantABC by 6chloroanthranilate was also analysed in Psv NCPPB 3335 cells harbouring pMP220:PantABC. Bacterial cells were grown on LB medium or LB supplemented with anthranilate or 6chloroanthranilate at the same concentration (7.3 mM), and β-galactosidase was measured 24 hours after the addition. As shown in Fig. 4C, both anthranilate and 6-chloroanthranilate induced the activity of PantABC to the same extent in comparison with the unamended medium, suggesting that transcription of the antABC operon might also be induced by other anthranilate-derived compounds. Expression of the catBCA operon was analysed in both the wild type Psv NCPPB 3335 and its ΔantA mutant. For this purpose, Psv cells were grown in LB medium and LB amended with 7.3 mM anthranilate, and the expression of the catB gene was measured by RT-qPCR. Expression of catB in Psv NCPPB 3335 was 232 fold higher in the amended medium compared to LB (Fig. 4D). However, expression of the catB gene in the ΔantA mutant was similar in both media, suggesting that in this strain the antABC operon is essential for the activation of the catBCA Chapter I 83 operon by anthranilate. Thus, and although the protein activator of the catBCA operon (CatR) is not encoded in the genome of Psv NCPPB 3335, catechol (the product of antABC activity), and perhaps other catechol-related compounds, might act as activators for catBCA expression. Figure 4. Transcriptional analysis of the antABC and catBCA operons. (A) Degradation pathways of anthranilate and catechol by proteins encoded in the antABC and the catBCA operons. (B) β-galactosidase activity of PantABC in Psv NCPPB 3335 cells grown in LB amended or not with anthranilate. Lines represent the OD600nm of the cultures. (C) β-galactosidase activity of PantABC in Psv NCPPB 3335 cells grown for 24 hours in LB and LB amended with anthranilate or 6-chloroanthranilate. Bars in Figures 4B and 4C correspond to the average Miller Units ± the standard deviation from the average of three different assays. (D) Quantification of the transcription of the catB gene by RT-qPCR in Psv NCPPB 3335 and its derivative ΔantA mutant. The fold change was calculated after normalization using the gyrA gene as an internal control. Fold change refers to the expression in LB amended with anthranilate relative to the expression in non-amended LB 4 hours after the addition of anthranilate. Bars represent the means of three replicates ± the standard deviation. Involvement of the antABC operon in the degradation of anthranilate The antABC and catBCA operons act sequentially in the degradation of anthranilate to βketoadipate enol-lactone (Fig. 4A), which is further degraded into the Krebs cycle (Harwood & Parales, 1996). To confirm the activity of the antABC operon in Psv NCPPB 3335, a ΔantA mutant and its complemented strain harbouring pBRR:antABC (Table 2) were constructed. Psv cells grown overnight on LB medium were transferred to M9 medium containing 500 µM of Chapter I 84 anthranilate (M9A). After 24 hours, the filtered supernatants were analysed by HPLC. A single peak corresponding to anthranilate (Rt = 9.72 min) was observed in the chromatogram corresponding to uninoculated M9 plus anthranilate (Fig. 5A), whereas this peak was not present in the supernatant of M9A inoculated with the wild type strain Psv NCPPB 3335 (Fig. 5B), suggesting the total disappearance of the compound under the conditions tested. On the contrary, anthranilate hardly disappeared in the supernatant of M9A inoculated with the ΔantA mutant as an anthranilate peak corresponding to a concentration of 472 µM remained (Fig. 5C) which is only slightly lower than the initial concentration of 500 µM in M9A. When M9A was inoculated with the ΔantA mutant complemented with the antABC operon in a plasmid (pBBR:antABC), two peaks with Rt 7.8 min and 9.9 min were detected corresponding to catechol at a concentration of 229 µM and anthranilate at 66 µM, respectively. Thus, total degradation of the compound did not take place in the culture of the complemented strain. All these results confirm the implication of the Psv NCPPB 3335 antABC operon in the degradation of anthranilate to catechol. Since 6-chloroanthranilate also promotes the transcription of the antABC operon (Fig. 4C), we analysed the ability of Psv NCPPB 3335 to degrade this compound. However, no degradation of the compound was observed in the M9 medium amended with 6chloroanthranilate inoculated with Psv NCPPB 3335 (data not shown). However, no degradation of the compound was observed in the M9 medium amended with 6-chloroanthranilate inoculated with Psv NCPPB 3335 (data not shown). Chapter I 85 Figure 5. Functional analysis of the antABC and catBCA operons. HPLC chromatograms of filtered supernatants after 24h exposure to M9 minimal media supplemented with (A-D) 500 µM anthranilate or (E-H) 300 µM catechol. Simultaneous detection at 210 nm (blue lines) and 230 nm (red lines) was performed. (A) Control cell-free medium supplemented with anthranilate; (B) anthranilate-amended medium (M9A) inoculated with Psv NCPPB 3335; (C) M9A with the ΔantA mutant; (D) M9A with the ΔantA mutant complemented with pBBR:antABC; (E) Control cell-free medium supplemented with catechol; (F) catechol-amended medium with Psv NCPPB 3335; (G) catechol-amended medium with the ΔcatB mutant; (H) catechol-amended medium with the ΔcatB mutant complemented with pBBR:catBCA. The concentrations of the compounds were determined using a standard curve of anthranilate or catechol (5500 µM). The intensity of the signal is measured as millivolts (mV). Involvement of the catBCA operon in the degradation of catechol and halogenated catechols To confirm the involvement of the Psv NCPPB 3335 catBCA operon in the metabolism of catechol, a ΔcatB mutant was constructed. This mutation should have a polar effect over the transcription of the catC and catA genes located downstream from catB. A complemented ΔcatB strain carrying the complete catBCA operon from Psv NCPPB 3335 in a plasmid, strain ΔcatBpBBR:catBCA, was also constructed. Psv cells were grown overnight in LB broth amended with anthranilate 7.3 mM to induce the expression of the catBCA operon, collected by centrifugation and transferred to M9 media containing 300 µM catechol. After 24 hours, the filtered supernatant was analysed by HPLC. As shown in Figure 5E, the chromatogram corresponding to Chapter I 86 the cell-free medium showed a peak at Rt= 7.61 min, corresponding to non-modified catechol. On the contrary, the chromatogram of the wild type Psv NCPPB 3335 (Fig. 5F) showed no peak at this Rt, indicating that catechol was completely removed from the medium. However, several other peaks corresponding with the accumulation of unidentified compounds were observed in this chromatogram, some of which showed retention times between 1 and 4 minutes and a single additional peak at Rt = 12 min. Two large peaks at Rt= 8.01 and 8.66 min appeared in the chromatogram of the ΔcatB strain (Fig. 5G). The UV absorption spectra of these peaks resembled that of catechol, suggesting that several catechol-related compounds accumulated in the culture supernatant of this strain. This hypothesis is reinforced by the fact that the chromatogram of the complemented ΔcatB strain (ΔcatB-pBBR:catBCA) did not display these peaks (Fig. 5H). Since the catBCA operon is unique in the genome of Psv NCPPB 3335, it can be speculated that other enzyme(s) of this strain can unspecifically transform catechol under the conditions tested. Dienelactone hydrolases have been reported to be involved in the metabolism of halogenated catechols in Pseudomonas sp. strain B13 (Schlomann et al., 1990). In this pathway, halogenated catechols are initially attacked by catechol 1,2-dioxygenase (CatA), followed by muconate cycloisomerase (CatB) activities. To address the possible involvement of the dhoAB and catBCA operons in the catabolism of 4-chlorocatechol in Psv NCPPB 3335, the wild type strain and the ΔcatB and ΔdhoAB mutants were cultured in M9 media supplemented with 4chlorocatechol. A peak corresponding to 4-chlorocatechol (Rt= 5.05 min) was observed in the cell-free culture (Fig. 6A). However, 4-chlorocatechol was not detected in the chromatogram corresponding to the supernatant of the wild type strain or its ΔdhoAB mutant (Fig. 6B, 6C), indicating the ability of the strain to degrade 4-chlorocatechol through a pathway independent of dhoAB. In addition, other peaks corresponding to unidentified compounds at Rt (4.45 and 4.85 min) close to that of 4-chlorocatechol were observed in these chromatograms, suggesting the accumulation of 4-chlorocatechol-related compounds in the culture medium of these two strains. In contrast, and although the 4-chlorocatechol was detected in the supernatant of the ΔcatB mutant (Fig. 6D), the calculated area under the peak for this compound was approximately four fold lower than that of the cell-free medium (Fig. 6A), suggesting that partial degradation of this compound is dependent on the catBCA operon. Chapter I 87 Figure 6. 4-Chlorocatechol transformation by the P. savastanoi pv. savastanoi NCPPB 3335 catBCA operon. HPLC chromatograms of filtered supernatants after 24h exposure to M9 minimal medium supplemented with 300 µM 4-chlorocatechol. Simultaneous detection at 210 nm (blue lines) and 230 nm (red lines) was performed. (A) Control cell-free medium; (B) Psv NCPPB 3335; (C) ΔdhoAB mutant; (D) ΔcatB mutant. Involvement of the ipoABC operon in the oxidation of indole into indigo To analyse the function of the ipoABC operon, which genes were annotated as putative components of an oxygenase involved in the metabolism of aromatic compounds (Table 5), we constructed a plasmid harbouring the whole operon under the control of its own promoter region (plasmid pBBR:ipoABC, Table 2). Several authors have reported that heterologous expression of certain oxygenases in E. coli leads to the production of the blue pigment indigo (Ensley et al., 1983; Woo et al., 2000; Choi et al., 2003; Doukyu et al., 2003; van Hellemond et al., 2007), which is synthesized via the oxidation of indole (Fig. 7A). E. coli DH5α cells harbouring pBBR:ipoABC were grown in M9 medium (colourless) amended with 0.5 mM indole. After 24 h of incubation at 37 °C, the culture turned blue, whereas the culture of the untransformed strain remained colourless. Furthermore, a Psv NCPPB 3335 ΔipoABC mutant and its complemented strain harbouring the plasmid pBBR:ipoABC (strain ΔipoABC-pBBR:ipoABC) were constructed and incubated in M9 minimal medium amended with 0.5 mM indole. While cultures of this medium inoculated with the wild type strain Psv NCPPB 3335 or its ΔipoABC mutant remained colourless, the culture of the complemented strain turned blue and accumulated insoluble blue precipitates (Fig. 7B). HPLC analysis of these blue precipitates showed a single peak at Rt= 1.7 min (Fig. 7C) with an UV spectrum identical to that previously reported for indigo (Olvera Vargas Chapter I 88 et al., 2010). These results confirm the involvement of the ipoABC operon in the transformation of indole into indigo, suggesting the codification of an oxygenase able to use aromatic compounds as substrates. Figure 7. Functional analysis of the ipoABC operon. (A) Proposed pathway for the biosynthesis of indigo by toluene dioxygenase (Ensley et al., 1983). Picture adapted from Woo et al. (2000) (B) Appearance of M9 minimal media supplemented with 0.5 mM indole after growth of Psv NCPPB 3335, its ΔipoABC mutant and the complemented ΔipoABC strain expressing the whole ipoABC operon on a plasmid (ΔipoABC-pBBR:ipoABC). (C) HPLC chromatogram of the blue compound partially purified from the supernatant of the ΔipoABC-pBBR:ipoABC culture. Role of the WHOP region in virulence and in planta competitive fitness of Psv NCPPB 3335 Due to the exclusivity of the WHOP region to strains of the P. syringae complex infecting the woody organs of woody hosts (Table 4), it resulted particularly interesting to analyse its role in the interaction between Psv NCPPB 3335 and olive plants. For this purpose we used two different models, micropropagated olive explants (non-woody plants) (Fig. 8A) and one year-old woody olive plants (Fig. 8B). Plants were inoculated with the wild type Psv NCPPB 3335 or each of the derivative mutants affected in the WHOP region. After 30 days, no visual differences were observed in the knot sizes developed on micropropagated plants among the strains tested (Fig. 8A). In fact, quantification of the volume of the knots generated using a 3D scanner confirmed no significant differences among the samples (Fig. 8C). However, clear differences could be Chapter I 95 lacking the “catechol operon” were unable to grow endophytically in kiwifruit trees, whereas other strains that carry this region could do so (Bartoli et al., 2015a). Similarly, the Psv NCPPB 3335 ΔcatBCA, ΔdhoAB and Δ06 mutants were less competent than the wild type strain on woody olive plants, but not on microprogated plants (Fig. 9), a finding that reinforces the hypothesis that the WHOP region contributes to the adaptation of bacteria from the P. syringae complex to woody niches. Although the chemical structure of wood cannot be defined precisely for a given tree specie; wood is composed mainly of the polysaccharides cellulose and hemicellulose (65-75%,) and lignin (18-35%) (Pettersen, 1984). Lignin is a complex organic network synthesized from three hydroxycinnamyl alcohols (p-coumaryl, coniferyl and sinapyl alcohols) (Boerjan et al., 2003; Ralph et al., 2004). It is well-known that lignin-related monomers are funnelled into protocatechuate and catechol prior to degradation through the βketoadipate pathway, which is widely distributed in soil bacteria and fungi (Harwood & Parales, 1996). While the catechol degradation machinery is restricted to several pathovars within the P. syringae complex (Table 4), the degradation of protocatechuate into Krebs cycle intermediaries is a feature widely distributed among P. syringae pathovars. Thus, the WHOP-encoded activities might complement the protocatechuate catabolic pathway to allow degradation of additional lignin-related compounds in bacteria that colonise woody organs. For example, cinnamate, an unsubstituted lignin-related compound, is metabolized via catechol (Andreoni et al., 1991). Thus, the WHOP region could help bacteria to utilise lignin precursor compounds as carbon/nitrogen sources. Another putative advantage conferred to bacteria by the WHOPencoded activities could be the modification of these compounds to decrease their toxicity. In fact, catechol and guaiacol, the latter a type of lignin subunit derived from coniferyl alcohol, are highly toxic for P. savastanoi (Capasso et al., 1995). Some other published works support the linkage between lignin and catechol-degrading activities. For example, the amendment of a soil with wood sawdust favoured the development of bacteria carrying the catA gene (catechol 1,2dioxygenase), in comparison with a non-amended soil (Cebron et al., 2015). In addition, strains possessing high lignocellulose degrading activity were able to mineralise catechol partially (Vetrovsky et al., 2014). Unlike the ΔcatBCA, ΔdhoAB and Δ06 mutants, which are impaired in their fitness exclusively on woody plants; the Psv NCPPB 3335 ΔipoABC operon mutant displayed a slower growth rate in both woody and non-woody plants in comparison to the wild type strain. This finding suggests that this impairment is not directly related to compounds that may act as wood precursors. Phenolic compounds are secondary metabolites produced by plants upon pathogen attack (Poiatti et al., 2009). Perhaps, the ipoABC operon is involved in the degradation of such defence compounds, which are produced by both non-woody and woody plants, to decrease their toxicity. However, further work is needed to confirm this hypothesis. Chapter I 96 In summary, we show that besides PG1 and PG3 strains infecting woody organs of woody hosts, a complete WHOP region is also encoded in P. syringae pv. ciccaronei and P. syringae pv. actinidifoliorum, which cause leaf spots in woody hosts. Partial conservation of the WHOP occurs in only a few PG2 strains. Furthermore, several clusters located within the WHOP region play a role in the adaptation of Psv NCPPB 3335 to woody hosts, but not to non-woody plants, supporting the role of the WHOP region in the adaptation of other P. syringae pathovars to fruit trees and other woody hosts. We confirmed the function of the antABC and the catBCA operons and identified an oxygenase activity for the ipoABC operon. However, further work is necessary to determine the environmental or plant-derived substrate(s) used by the enzymes encoded in the WHOP region, as well as to elucidate the activity of the dhoAB operon and the PSA3335_3206 gene. This work has been accepted for publication the 22nd of December 2016 in Molecular PlantMicrobe Interactions, a journal of the American Phytopathological Society (APS) (Caballo-Ponce et al., 2016). Chapter II Comparative analysis of the quorum sensing system in Pseudomonas savastanoi pv. savastanoi NCPPB 3335 and other Pseudomonas syringae strains. Chapter II 99 INTRODUCTION Many bacteria assess their local population via the secretion of small signalling molecules in a process named quorum sensing. The signal molecules, commonly acylated homoserine lactones in Gram (-) bacteria (AHLs) are synthesized by a LuxI-family protein and diffuse through the bacterial membrane. The increase of a bacterial population brings to accumulation of AHLs in the environment and, upon reaching a threshold concentration (the quorum), are recognized inside the bacterial cells by a LuxR-family transcriptional regulator that controls the transcription of target genes (Bassler, 1999; Fuqua et al., 2001; Miller & Bassler, 2001; Whitehead et al., 2001). A canonical quorum sensing system is here defined as that composed of an adjacent luxI/luxR pair. However, the number of LuxIand LuxR-type proteins is not always equal for a given bacterial species. In addition to the paired luxI/luxR genes, the analysis of a set of 265 protobacterial genomes showed that a number of genomes also contain a luxR gene, generally named luxR solos (Subramoni & Venturi, 2009), which are not encoded in the vicinity of luxI (Case et al., 2008). LuxR solos are able to respond to endogenously produced AHLs (Chugani et al., 2001) or to the AHLs produced by neighbouring bacteria (Ahmer et al., 1998). Interestingly, a sub-family of LuxR solos unique to plant-associated bacteria, which respond to plant signals and are encoded in the proximity of the pip gene (encoding for a proline iminopeptidase) has been described (Patel et al., 2013). The quorum sensing system of Pseudomonas aeruginosa is the best characterized among Pseudomonads and consist of two interconnected luxI/luxR pairs (Chugani et al., 2001; Heurlier et al., 2006; Willcox et al., 2008; Lee & Zhang, 2015). In P. aeruginosa PAO1 these two LuxI homologs govern the transcription of over 300 genes (Schuster et al., 2003; Wagner et al., 2003). Plant-associated Pseudomonads also interact via the quorum sensing system (Loh et al., 2002; Von Bodman et al., 2003), although some strains, including several Pseudomonas syringae, do not produce detectable amounts of AHLs (Dumenyo et al., 1998). For instance, the genome of P. syringae pv. actinidiae does not encode for a luxI homolog, but carry several luxR homologs that allow for “eavesdropping” of the signals produced by other bacteria (Patel et al., 2014). Many efforts have been dedicated to understand the quorum sensing system in the P. syringae complex, mainly using P. syringae pathovars syringae and tabaci as models. In P. syringae pv. syringae (hereafter Psy) B728a this system has been shown to regulate alginate production and motility, as well as bacterial virulence measured as the incidence of lesions caused on leaves. Moreover, AHL-deficient mutants did not provoke tissue maceration of bean pods, typical symptoms of wild type infections (Quinones et al., 2005). Two different transcriptomic analyses identified quorum sensing-controlled genes in P. syringae strains. AhlR, a LuxR homolog in P. syringae B728a, activates a small number of genes that, in addition, are located on the vicinity Chapter II 100 of the luxI/luxR pair (Yu et al., 2014). In contrast, a luxI mutant of P. syringae pv. tabaci (hereafter Pst) 6605 showed a larger number of genes with an altered transcription level in comparison to the wild type strain (Taguchi et al., 2015). Despite the attention paid to the quorum sensing system of P. syringae strains pathogenic on non-woody host, very little research focused on the quorum sensing systems of woody-host pathogens of the P. syringae complex. In fact, up to our best knowledge, only two published works examined this regulation circuit in this type of bacteria. One of these studies focused on the intriguing regulation of the non AHL-producing kiwifruit pathogen P. syringae pv. actinidiae (Patel et al., 2013), whereas the other explored the quorum sensing of the olive tree pathogen Pseudomonas savastanoi pv. savastanoi (Psv) (Hosni et al., 2011). Bacteria-bacteria communication via quorum sensing critically affects the virulence and exopolysaccharides (EPS) production of Psv DAPP-PG 722. Moreover, this strain shares the same signal molecules with Erwinia toletana DAPP-PG 735, a bacterium commonly associated to Psv within olive knots (Rojas et al., 2004; Hosni et al., 2011; Passos da Silva et al., 2014). The genome of Psv NCPPB 3335, reference Psv strain in our laboratory, harbours a canonical luxI/luxR pair (RodriguezPalenzuela et al., 2010), which is identical to that of Psv DAPP-PG 722, although its quorum sensing system has not been investigated up to date. In this chapter we examined the quorum sensing system of Psv NCPPB 3335. The first aim of this study was to determine differences in the presence of the luxI/luxR homologs among a collection of 27 strains of the P. syringae complex belonging to phylogroups (PGs) 1 to 4. Secondly, to identify genes controlled by the amount of AHL in Psv NCPPB 3335, as well as to analyse the similarities and differences among the quorum sensing regulons of NCPPB 3335 and those of other P. syringae strains. Chapter II 101 MATERIAL AND METHODS Bacterial strains, media and growth conditions Bacterial strains used in this study are listed in Table 1. P. savastanoi pv. savastanoi and Escherichia coli were grown at 28 °C or 37 °C, respectively, in Luria-Bertani (LB) medium (Miller, 1972) and Super Optimal Broth (SOB) (Hanahan, 1983). Solid and liquid media were amended when required with the appropriate antibiotic. Antibiotic concentration used were: kanamycin (Km) 10 µg ml-1 for P. savastanoi and 50 µg ml-1 for E. coli, gentamycin (Gm) 10 µg ml-1, ampicillin (Ap) 400 µg ml-1 for P. savastanoi and 100 µg ml-1 for E. coli; and tetracycline 10 µg ml-1. Table 1. Strains used in this study Strain Relevant characteristicsa Source Pseudomonas savastanoi pv. savastanoi NCPPB 3335 Wild type strain (Perez-Martinez et al., 2007) ΔpssI Deletion pssI (PSA3335_1621) mutant (KmR) This work ΔpssR Deletion pssR (PSA3335_1622) mutant (KmR) This work Pseudomonas syringae pv. syringae B728a Wild type strain (Loper & Lindow, 1987) Escherichia coli DH5α F-, ϕ80dlacZ M15, (lacZYA-argF) U169, deoR, recA1, endA, hsdR17 (rk – mk-), phoA, supE44, thi-1, gyrA96, relA1 (Hanahan, 1983) GM2929 F-, ara-14, leuB6, thi-1, tonA31, lacY1, tsx-78, galK2, galT22, glnV44, hisG4, rpsL136, xyl-5, mtl-1, dam13::Tn9, dcm-6, mcrB1, hsdR2, mcrA, recF143 (SpR CmR) (Palmer & Marinus, 1994) JB52 Harbours pJBA130 and is used as a sensor for acylhomoserine lactone production (Andersen et al., 2001) aKm, kanamycin Construction of bacterial strains and plasmids Plasmids and oligonucleotides used in this study are listed in Tables 2 and 3 approximately. For the generation of P. savastanoi pv. savastanoi NCPPB3335 mutants, DNA fragments of approximately 1 kb corresponding to the upstream and downstream flanking regions of the gene to be deleted were amplified in three rounds of polymerase chain reaction (PCR) using genomic DNA from NCPPB 3335 as a template and Expand High Fidelity polymerase (Roche Applied Science, Mannheim, Germany). Restriction sites for HindIII were included in the primers as previously described (Matas et al., 2014). The resulting products, consisting on upstream and downstream flanking regions separated by the HindIII restriction site, were cloned into pGEM-T easy to yield pECP10 and pECP11 (Table 2), and sequenced to discard mutations. Later, the kanamycin resistance gene nptII was extracted by enzyme restriction from pGEM-T-KmFRT- Chapter II 102 HindIII (Aragon et al., 2014) and cloned into the plasmids mentioned above to generate pECP10Km and pECP11-Km (Table 2). All plasmids generated for the construction of Psv NCPPB 3335 mutants were suicide vectors in P. savastanoi. Plasmids were transferred to NCPPB 3335 by electroporation (Perez-Martinez et al., 2007) and transformants were selected in LB-Km plates. Plasmids pECP10-Km and pECP11-Km were used for the construction of ΔpssI and ΔpssR mutants respectively. To select the allelic interchange (double recombination event) and discard plasmid integration (single recombination event), individual colonies were replicated into LB-Ap plates and ApR colonies were discarded. Finally, Southern blot analyses were carried out to confirm single integration in the correct position in Psv genome. Plasmids for the complementation of the mutant strains were generated as follows. The complete open reading frames of pssI and pssR genes and their corresponding promoter and transcriptional terminator regions were amplified by PCR using NCPPB 3335 genomic DNA as a template and Expand High Fidelity polymerase (Roche Applied Science, Mannheim, Germany) and cloned into pGEM-T easy (Promega, Madison, WI, USA). After sequencing to discard mutations, the fragments were directionally subcloned into pBBR:MCS5. Plasmids pBBR:pssI and pBBR:pssR were used for the complementation of ΔpssI and ΔpssR respectively. Table 2. Plasmids used in this study Name Descriptiona Source pGEM-T easy Cloning vector containing ori f1 and lacZ (ApR) (Promega, Madison, WI, USA) pBBR1:MCS5 Broad host-range cloning vector (GmR) (Kovach et al., 1995) pJBA130 pME6031-luxR-PluxI-RBSII-gfpmut3*-T0-T1 (Tcr) (Andersen et al., 2001) pGEM-TKmFRTHindIII Contains KmR from pKD4 and HindIII sites (ApR KmR) (Aragon et al., 2014) pECP10 pGEMT-derivative containing 1kb on each side of the pssI (PSA3335_1621) gene from NCPPB 3335 (ApR) This work pECP10-Km pGEMT-derivative containing 1kb on each side of the pssI (PSA3335_1621) gene from NCPPB 3335 interrupted by the kanamycin resistance gene nptII (ApR, KmR) This work pECP11 pGEMT-derivative containing 1kb on each side of the pssR (PSA3335_1622) gene from NCPPB 3335 (ApR) This work pECP11-Km pGEMT-derivative containing 1kb on each side of the pssR (PSA3335_1622) gene from NCPPB 3335 interrupted by the kanamycin resistance gene nptII (ApR, KmR) This work pBBR:pssI pBBR1:MCS5-derivative containing the Psv NCPPB 3335 pssI and its promoter region (352 bp) flanked by EcoRI and XbaI restriction sites (GmR) This work pBBR:pssR pBBR1:MCS5-derivative containing the Psv NCPPB 3335 pssR and its promoter region (435 bp) flanked by EcoRI and XbaI restriction sites (GmR) This work pMP220 Broad-host-range, low-copy-number promoter probe vector, IncP replicon, lacZ (TcR) (Spaink et al., 1987) Chapter II 103 Table 2. Plasmids used in this study (continuation) Name Descriptiona Source pMP220-PpssR Contains a fragment of 338 bp corresponding with the pssR promoter from Psv NCPPB 3335 cloned with KpnI and BglII restriction sites (TcR). This work pMP220-PpssI Contains a fragment of 352 bp corresponding with the pssI promoter from Psv NCPPB 3335 cloned with KpnI and BglII restriction sites (TcR). This work pMP220-PahlI Harbours a fragment of 344 bp corresponding with the ahlI promoter from P. syringae pv. syringae B728a cloned with KpnI and BglII restriction sites (TcR). This work aAp, ampicillin; Gm, gentamycin; Km, kanamycin; Tc, tetracycline; Gm, gentamycin. Promoter activity assays For the generation of promoter fusions to β-galactosidase, fragments of 338, 352 and 344 bp upstream the open reading frame of Psv NCPPB 3335 pssI and pssR genes and Psy B728a ahlI gene, respectively, were amplified by PCR using genomic DNA as a template. KpnI and BglII restrictions sites were included in the primers. The amplicons were independently cloned into pGEM-T easy, sequenced to discard mutations and subcloned into pMP220. Finally, plasmids were introduced in bacteria by electroporation. For the measurement of pssR promoter activity, the Psv NCPPB 3335 and its derivative pssI mutant strains were grown in LB broth at an initial OD600nm of 0.2 and the β-galactosidase activity was monitored along the growth curve. For pssI promoter activity, Psv NCPPB 3335 and its derivative pssI and pssR mutant were grown in King’s B (KB) (King et al., 1954) and LB media at an initial OD600nm of 0.3 and samples for β-galactosidase activity were taken after 4, 6, 8 and 10 hours. An identical procedure was followed for the pssI promoter activity in P. syringae B728a, with the exception of an initial OD600nm of 0.1. βgalactosidase activity was quantified as previously described (Miller, 1972). Production and identification of acyl-homoserine lactones The production of AHLs in Psv NCPP 3335, ΔpssI and ΔpssR mutants was analysed as previously described (Andersen et al., 2001). The indicated strains were incubated for 24 hours at 28ºC on a LB plate together with E. coli JB52pJBA130 (sensor strain) and were visualized in a Leica MZ FLIII stereoscopic fluorescence microscope equipped with a 100 W mercury lamp and a GFP2 filter (excitation 480/40 nm; emission 510LP nm). For the identification of the AHLs produced, 100 ml of an overnight culture in LB broth of Psv NCPPB 3335, the pssI mutant and its complemented strain carrying pBBR:pssI (Table 2) were centrifuged to remove bacteria. The supernatant of each culture was filtered (pore diameter 0.45 µm) and mixed with the same Chapter II 104 volume of 0.1% acetic acid (v/v) in ethyl acetate and shaken for 30 minutes (extraction step). The organic phase was recovered prior and the extraction step was repeated. The organic phases were put alltogether and dried. The AHLs produced by each strain were identified by high performance liquid chromatography coupled to mass spectrometry (HPLC-MS) as described previously (Coutinho et al., 2013). RNAseq experiment Three independent cultures of Psv NCPPB 3335 and ΔpssI were incubated at 28ºC in LB to early stationary phase (OD600nm= 2) and RNA was then extracted with the RiboPure™ RNA Purification Kit (Ambion, Thermo Scientific, MA, USA). Total RNA was treated with a TURBO DNAfree™-Kit (Applied Biosystems; CA, USA) as detailed by the manufacturer’s instructions. DNAfree RNA samples were subjected to depletion of the ribosomal RNA using the Ribo-Zero Magnetic Kit specific for gram-negative bacteria (Epicentre, Madison, WI, USA). rRNA-depleted RNA were processed and sequenced as detailed previously (Coutinho et al., 2015). RNA sequencing was performed by IGA Technology Services Srl (Udine, Italy). Bowtie2 software was used to map the raw reads to the Psv NCPPB 3335 genome (Langmead & Salzberg, 2012) and the reads in genes were counted with Granges (Lawrence et al., 2013). Finally, analysis of differentially expressed genes was carried out with DESeq2 (Love et al., 2014). Genes with a log2(fold change) over 0.9 or under -0.9 were selected for validation by reverse transcription quantitative PCR (RT-qPCR) as detailed later. Biofilm, EPS, motility and virulence assays Biofilm formation by Psv NCPPB 3335, ΔpssI and ΔpssR was analysed on glass surfaces using a static microcosm assay, as detailed previously (Pliego et al., 2008). Crystal violet (CV) was used to quantify biofilm formation (O'Toole & Kolter, 1998). For exopolysaccharide detection, a drop containing 106 cfu was placed on LB, KB, MM (Hosni et al., 2011) and MGY+Sorbitol 0.6 M (Quinones et al., 2005) agar plates containing 50 µg ml-1 of Congo Red and incubated at 25ºC for two days. The results were captured with a high-resolution camera Canon D6200 (Canon Corporation, Tokyo, Japan). Swimming motility assays were performed in AB minimal medium supplemented with 10mM citrate and containing 0.3% agar (Huber et al., 2001). The centre of a Petri dish was inoculated with 2 µl of a bacterial suspension containing 108 cfu/ml and incubated at 25 °C for 96 hours. Swarming motility was examined as described before (Huber et al., 2001; Quinones et al., 2005). A drop 2 µl of bacterial suspension containing 107 cfu/ml was inoculated on the surface of a Petri dish containing KB or AB media and 0.4% agar and incubated at 25 °C for 96 hours. Olive plants were micropropagated and inoculated as detailed previously Chapter II 111 Identification of the Psv NCPPB 3335 quorum sensing regulon A whole-genome transcriptional comparative analysis of wild type Psv NCPPB 3335 and its derivative ΔpssI mutant was performed, a strategy previously used by several authors to establish the quorum sensing regulon in different plant-associated bacteria (Kim et al., 2013; Kim et al., 2014; Coutinho et al., 2015; Taguchi et al., 2015). The strains were grown in LB broth to late-log phase (OD600nm= 2), RNA was then extracted and sequenced as described in material and methods. Three independent RNA samples where sequenced for each strain. The results of the comparative transcriptomic analysis yielded a small number of genes differentially expressed between the strains. Table 4 shows the genes with a log2(fold change) over 0.9 or under -0.9; meaning a 1.87-fold upregulation/downregulation, respectively, in the ΔpssI mutant compared to the wild type strain. Only 10 genes fitted this criterion: five were upregulated and five were downregulated in the ΔpssI background compared to the wild type NCPPB 3335 (Table 4). These results contrast with the high number of genes showing expression dependency on luxI in other bacteria grown under the same conditions (Kim et al., 2013; Kim et al., 2014). In addition, a recent report described 236 genes with a log2(fold change) over 0.9 or under -0.9 in a comparative transcriptome analysis between Pst 6605 and its derivative ΔluxI mutant. However, in this case bacterial cells where transferred to MMMF medium after growth in LB before RNA extraction (Taguchi et al., 2015). To validate the results obtained from these RNAseq experiments, the expression of all genes showing a 1.87-fold upor down-regulation in the ΔpssI mutant (Table 4) was analysed by RT-qPCR. Significant upregulation in the ΔpssI mutant was found only for three genes (PSA3335_1621, PSA3335_1623 and PSA3335_1624) (Table 4). These genes encoded for the quorum sensing transcriptional regulator pssR (PSA3335_1621), a pyruvate dehydrogenase E1 component beta subunit (PSA3335_1622, pdhT) and a pyruvate dehydrogenase E1 component (PSA3335_1624, pdhQ) (Table 5). On the other hand, none of the five genes identified as downregulated by RNAseq analysis were validated by RT-qPCR (Table 4). As a control, the expression of pssI was analysed by RT-qPCR in Psv NCPPB 3335 and the ΔpssI mutant. As expected, expression of this gene was higher in the wild type strain. Thus, after validation of the RNAseq data by RT-qPCR, the pssI regulon of Psv NCPPB 3335 was restricted to three genes (pssR, pdhT and pdhQ) under the conditions tested, two of which (pdhT and pdhQ) are controlled by the pssR homolog in Psy B728a (Yu et al., 2014). Similarly, expression of pssR in Pst 6605 has been demonstrated to be dependent on pssI (Taguchi et al., 2015). Chapter II 112 aUpregulated or downregulated genes in the ΔpssI mutant according to RNAseq data. bUN, unnamed cThe log2 (fold change) obtained in the RNAseq and RT-qPCR experiments are represented. The fold change refers to the ratio of the average expression obtained in the ΔpssI mutant versus its wild type strain of three biological replicates. In the RNAseq analysis, genes with a log2 (fold change) over 0.9 or under -0.9 are shown. Genes validated by RT-qPCR are underlined (T-Student; p <0.05). Quorum sensing regulation in Psv NCPPB 3335 differs from that of other P. syringae strains Up to date, only two published works delved into the quorum sensing regulons of P. syringae strains (Psy B728a and Pst 6605). In Psy B728a, a mutant of the ahlR gene (encoding a PssR homolog) was analysed in comparison to the wild type strain under several environmental conditions, i.e. synthetic media, epiphytically and endophytically grown cells. Interestingly, under all these conditions AhlR was shown to control a small number of genes, which are all located in a cluster surrounding the ahlI/ahlR pair (hereafter named the SyrQS cluster) (Fig. 3A) (Yu et al., 2014). Among other conditions, in this work Psy B728a cells were grown in synthetic medium amended with the AHL produced by this strain (C6-3-oxo HSL), in order to enhance differential expression of genes between Psy B728a and its ΔahlR mutant. As mentioned before, genes controlled by PssI in Psv NCPPB 3335 (pdhT and pdhQ) are regulated by AhlR (the PssR homolog) in Psy B728a. For comparison with the results obtained in Psy B728a, Psv NCPPB 3335 and its ΔpssR mutant were cultured in LB medium amended with C6-3-oxo HSL, and the involvement of pssR in the regulation of the SyrQS cluster was analysed by RT-qPCR. Enhanced expression of paoA, paoB, paoC, paoD and pssI was observed in the mutant compared to the wild type. On the contrary, all these genes were downregulated in the ΔahlR of Psy B72a (Fig. 3B) (Yu et al., 2014). Moreover, in Psv NCPPB 3335, PssR is not involved in the control of paoE, pdhT, qrpR and pdhQ under the conditions tested. As expected, the expression of pssR also differed in the ΔpssR mutant compared to Psv NCPPB 3335 (not shown). Table 4. Genes regulated by pssI in Psv NCPPB 3335 Locus taga Geneb Gene product RNAseqc RT-qPCRc Upregulated PSA3335_1622 pdhT Pyruvate dehydrogenase E1 component, beta subunit 3.27 3.6 PSA3335_1624 pdhQ Pyruvate dehydrogenase E1 component 2.97 2.32 PSA3335_1621 pssR LuxR transcriptional regulator 1.44 3.95 PSA3335_2315 UN Putative hydrocarbon oxygenase 1.33 -0.13 PSA3335_4742 UN Urocanate hydratase 1.01 -1.56 Downregulated PSA3335_1620 pssI Homoserine lactone synthase -3.41 -8.01 PSA3335_4623 UN Copper chaperone -1.07 -0.82 PSA3335_2048 UN Hypothetical protein -0.99 -0.42 PSA3335_0454 mcdE Malonate decarboxylase delta subunit -0.94 -0.13 PSA3335_2054 UN Hypothetical protein -0.93 0.67 PSA3335_4121 UN Pectin lyase precursor -0.92 0.52 Chapter II 113 A comparative transcriptome analysis between the tobacco pathogen Pst 6605 and its derivative pssI mutant showed that 236 genes were upregulated/downregulated in the pssI mutant (Taguchi et al., 2015). In order to compare this pattern of expression with that of Psv NCPPB 3335, six of the genes showing high downregulation/upregulation in Pst 6605, as well as the pssR gene, were investigated in the Psv NCPPB 3335 ΔpssI mutant in comparison with the wild type strain. A completely different regulation pattern was observed between Psv NCPPB 3335 and Pst 6605. Transcription of most of the genes analysed resulted to be independent on pssI in Psv NCPPB 3335. On the other hand, and in agreement with the RNAseq results obtained in this study, but opposite to Pst 6605, transcription of pssR was upregulated in the ΔpssI mutant of Psv NCPPB 3335 (Table 4). All the findings described above support the existence of a distinct regulation mediated by quorum sensing between Psv NCPPB 3335 and other strains of the P. syringae complex. Table 5. Genes controlled by quorum sensing in P. syringae strains analysed in this study Locus taga Geneb Predicted function Strain reportedc PSA3335_1615 paoE Rieske iron-sulfur domain-containing protein Psy, Pst PSA3335_1616 paoD Hypothetical protein Psy, Pst PSA3335_1617 paoC Dihydrodipicolinate synthase Psy, Pst PSA3335_1618 paoB Hypothetical protein Psy, Pst PSA3335_1619 paoA Aspartate/tyrosine/aromatic aminotransferase Psy, Pst PSA3335_1620 pssI Homoserine lactone synthase Psy, Pst PSA3335_1621 pssR LuxR transcriptional regulator PsyR Psy, Pst PSA3335_1622 pdhT Pyruvate dehydrogenase E1 component beta subunit Psy PSA3335_1623 qrpR Transcriptional regulator, MarR family Psy, Pst PSA3335_1624 pdhQ Pyruvate dehydrogenase E1 component Pst PSA3335_0003 UN hypothetical protein Pst AER-0003118 mexF C-terminal fragment of a multidrug efflux transporter Pst PSA3335_2264 mexE Membrane fusion protein of RND family multidrug efflux pump Pst PSA3335_3741 trmD tRNA (Guanine37-N1) -methyltransferase Pst PSA3335_3135 romA Outer membrane protein romA Pst PSA3335_4922 UN sodium/alanine transporter Pst aLocus tags correspond to the Psv NCPPB 3335 genome, except for the mexF gene, which locus tag is not available. AER-0003118 indicate the AER code of the gene in the ASAP database (https://asap.genetics.wisc.edu/asap/logon.php). bUN, unnamed. cPsy, genes controlled by AhlR in P. syringae pv. syringae B728a; Pst genes controlled by PsyI in Pseudomonas syringae pv. tabaci 6605. Chapter II 114 Figure 3. Comparative analysis of quorum sensing regulated genes in P. syringae strains. (A) Scheme of the gene cluster (SyrQS cluster) controlled by the LuxR homolog AhlR in Psy B728a (Yu et al., 2014). Gene names are shown under the arrows. (B) RT-qPCR analysis of the SyrQS cluster in Psv NCPPB 3335 and its ΔpssR grown in LB medium. Results are compared with those obtained for Psy B728a cells grown in HMM medium (Yu et al., 2014). (C) RT-qPCR analysis of selected genes in Psv NCPPB 3335 and its ΔpssI mutant incubated in MMMF medium. Genes were selected according to their altered expression pattern in a Pst 6605 pssI mutant grown under the same conditions (Taguchi et al., 2015). For Psv NCPPB 3335, bars represent the log2 of the average expression (three biological replicates) obtained for its (B) ΔpssR or (C) ΔpssI mutants relative to the wild type strain. Genes showing significant differences (T-Student, P<0.05) in their expression levels between Psv NCPPB 3335 and its respective mutants strains are indicated with an asterisk. Data corresponding to gene expression in (B) Psy B728a and (C) Pst 6605 were extracted from Yu et al. (2014) and Taguchi et al. (2015), respectively. Expression analysis of pssI and pssR in P. savastanoi NCPPB 3335 While in V. fischeri LuxR controls its own transcription in an AHL-dependent manner (Dunlap & Ray, 1989; Shadel & Baldwin, 1991), transcription of ahlR in Psy B728a was shown to be independent on the cell concentration (Quinones et al., 2004). Taking into account this dissimilar expression of luxR homologs and that pssR expression was shown to be pssIdependent in Psv NCPPB 3335 according to our RNAseq experiments, we investigated the transcription of pssR in both Psv NCPPB 3335 and Psy B728a cells grown in LB. For this purpose, the promoter region of the Psv NCPPB 3335 pssR gene was fused to the lacZ gene (Table 2) and β-galactosidase activity was examined. As shown in Figure 4A, from 6 hours to the end of the experiment, expression of this gene was higher in the ΔpssI strain than in Psv NCPPB 3335. In addition, we observed that expression of the promoter was higher during late exponential phase and decreased as the bacteria reached the stationary phase. These results demonstrate that Chapter II 115 transcription of pssR is not constant throughout the growth curve in either Psv NCPPB 3335 or its pssI mutant, and suggest that a mechanism independent of the presence of AHLs controls the expression of pssR. A paradigm of the quorum sensing regulation is that, upon reaching a threshold concentration of AHLs, LuxR binds to its cognate AHL and promotes transcription of luxI leading to a positive feedback of the system (Whitehead et al., 2001; Von Bodman et al., 2003). Since all the results mentioned above suggest that Psv NCPPB 3335 possesses a particular quorum sensing regulation, the promoter region of pssI was fused to lacZ and β-galactosidase activity was monitored at different points during growth of Psv NCPPB 3335 and its ΔpssI and ΔpssR mutants in KB medium. In the wild type strain, expression of pssI was similar at 4 and 6 hours, but it doubled at 8 hours and remained at this level 10 hours after incubation (Fig. 4B). Furthermore, expression of the pssI promoter displayed an identical behaviour in the ΔpssI and ΔpssR mutant backgrounds. Similar results were obtained in LB medium (not shown). These results demonstrate that transcription of pssI is activated in stationary phase and is independent on PssI and PssR. In fact, pssI transcription was suggested to be independent on PssR in Psv DAPP-PG 722 (Hosni et al., 2011). Transcription of the Psv NCPPB 3335 pssI promoter was also analysed in Psy B728a (Fig. 4C). As observed for Psv strains, an increase in the transcription of the pssI promoter was observed at 8 hours and remained at the same level after 10 hours (Fig. 4C). However, a stronger transcription of the pssI promoter was detected in Psy B728a compared to Psv NCPPB 3335 (Fig. 4B, 4C). For example, at 6 hours the β-galactosidase activity measured in Psv NCPPB 3335 was 57 Miller Units, whereas in Psy B728a it was 2421 Miller Units. Similar results were obtained when the expression of the ahlI promoter was analysed in both bacteria (data not shown). In summary, although Psy B728a and Psv NCPPB 3335 shared the same activation behaviour of the pssI promoter, its overall expression was remarkably higher in Psy than in Psv, suggesting that regulation of the expression of luxI homologs is not identical in both bacteria. Chapter II 116 Figure 4. Expression of the pssR and pssI promoters from Psv NCPPB 3335. The ß-galactosidase activity from (A), the pssR (PpssR::’lacZ) and (B, C), the pssI (PpssI::’lacZ) promoters was determined in (A) Psv NCPPB 3335 and its ΔpssI mutant, (B) Psv NCPPB 3335 and its ΔpssI and ΔpssR mutants and, (C) Psy B728a. Lines and bars correspond, respectively, to the average OD600nm and the mean β-galactosidase activity obtained from three different cultures; error bars correspond to the standard deviation from the average. pMPP220, promoterless empty vector used as control. Phenotypes associated to quorum sensing in other P. syringae strains are not dependent on AHL production in Psv NCPPB 3335 EPS production, virulence, motility and biofilm formation, phenotypes all controlled by quorum sensing in other plant-pathogenic bacteria (Von Bodman et al., 2003), were analysed in Psv NCPPB 3335. EPS production was tested in different media: LB, KB, minimal mannitol (MM) medium and MGY+Sorbitol. As shown in Figure 5, NCPPB 3335 and its derivative quorum sensing mutants (ΔpssI and ΔpssR) showed an identical appearance in all media tested, suggesting that quorum sensing does not control EPS production in this strain. Furthermore, complementation of the ΔpssI and ΔpssR mutants with pssI and pssR, respectively, cloned in a multicopy plasmid, also yielded identical phenotypes. EPS production deficiency of Psv DAPP-PG 722 pssI and pssR mutants have been reported both in KB and MM media (Hosni et al., 2011), although we could not reproduce these observations under our laboratory conditions. Also related with EPS production, a double mutant of the paired luxI/luxR homologs in Psy B728a showed a reduced mucoid phenotype in comparison to the wild type strain growing in MGY+Sorbitol (Quinones et al., 2005), a medium promoting alginate biosynthesis in Psy FF5 (Penaloza-Vazquez et al., 1997). Chapter II 117 However, this phenotype was neither observed for the Psv NCPPB 3335 ΔpssI and ΔpssR mutants. Figure 5. EPS production of Psv NCPPB 3335 quorum sensing mutants. Bacteria were inoculated in the indicated media, all amended with 50 µg ml-1 of Congo Red, and incubated for 24 hours at 28 °C. Psv, Pseudomonas savastanoi pv. savastanoi NCPPB 3335; ΔpssI, knockout mutant of the NCPPB 3335 luxI homolog; ΔpssI-pBBR:pssI, NCPPB 3335 ΔpssI expressing its pssI gene from a multicopy plasmid; ΔpssR, knockout mutant of the NCPPB 3335 luxR1 homolog; ΔpssR-pBBR:pssR, NCPPB 3335 ΔpssR expressing its pssR gene from a multicopy plasmid. Swimming and swarming motilities were also analysed. A drop of a bacterial suspension was inoculated in the centre of a Petri dish containing 0.3% agar AB minimal medium (Huber et al., 2001) and the diameter of the swimming halos were recorded after 96 hours. No differences in the swimming ability of Psv NCPPB 3335 and its derivative ΔpssI and ΔpssR mutants were found (Fig. 6A). Swarming motility was evaluated on 0.4% agar AB and KB media. For this purpose, a drop of a bacterial suspension was placed on top of the agar and swarming motility was analysed daily during 4 days. Although NCPPB 3335 and its quorum sensing mutants did not swarm on AB minimal medium, the three strains swarmed equally in KB (not shown). These results are in agreement with those previously reported in Psv DAPP-PG 722 (Hosni et al., 2011). It should be noticed that the ability of NCPPB 3335 to swarm has not been reported before. Chapter II 118 We also analysed the biofilm formation ability of Psv NCPPB 3335 and its ΔpssI and ΔpssR mutants on a glass surface. Figure 6B shows the quantification of biofilm formation measured after staining with crystal violet. No differences among the strains analysed were found. In summary, phenotypes typically regulated by quorum sensing in plant-pathogenic bacteria such as EPS production, motility and biofilm formation are not controlled by this system in Psv NCPPB 3335 under the conditions tested. Figure 6. Swimming motility and biofilm formation of Psv NCPPB 3335 quorum sensing mutants. (A) Swimming motility of the indicated strains in AB minimal media amended with 10 mM citrate. The average diameter of the swimming halo is represented. Error bars represent the standard deviation from the average. (B) Quantification of biofilm formation by the indicated strains over glass slides using crystal violet (CV). Bars represent the average OD595nm of three independent replicates and the error bars represent the standard deviation from the average. Hosni and collaborators (2011) found that virulence of Psv DAPP-PG 722 in olive plants was severely reduced in their pssI and pssR mutants compared to the wild type strain. The ΔpssI and ΔpssR mutants of Psv NCPPB 3335 and their respective complemented strains were inoculated in micropropagated and woody olive plants. However, significant differences in knot development among the strains tested were not found either in non-woody (micropropagated) or woody olive plants (Fig. 7). It is important to mention that virulence attenuation of the Psv DAPP-PG722 ΔpssI and ΔpssR mutants was neither observed under the conditions tested in this study. Chapter II 119 Figure 7. Virulence assay of the Psv NCPPB 3335 quorum sensing mutant on olive plants. Knots induced by the indicated strains on (A) non-woody and (B) woody olive plants at 30 dpi and 90 dpi, respectively. Knot volume (mm3) induced by the indicated strains in (C) non-woody and (D) woody olive plants. Bars represent the mean of a minimum of three knots and error bars represent the standard deviation from the average. Chapter II 120 DISCUSSION Research on the quorum sensing system within the Pseudomonas genus has mainly focused in the widespread bacterium P. aeruginosa. Although this regulatory network has also been characterized in several P. syringae, most efforts have been devoted to P. syringae strains infecting non-woody plants. Up to our knowledge, only two works took a close look to the quorum sensing system of these kind of strains and, in addition, analysed two different regulatory networks: the AHL producing system of Psv (Hosni et al., 2011) and the non-AHL producing system of P. syringae pv. actinidiae (Patel et al., 2014). The intriguing quorum sensing circuit of the later, which lacks a luxI homolog, motivated our analysis of the presence of luxI and luxR homologs within a collection of 27 P. syringae strains of PG1 to PG4. Substantial differences were observed among the bacteria analysed: all type of LuxI LuxR genotypes were found, including several LuxI+ LuxRand LuxILuxR+ strains, as well a LuxILuxRstrain (P. syringae pv. aptata DSM 50252). Furthermore, the number and type of luxR homologs greatly vary among different strains, although all of them, with the exception of the P. syringae pv. aptata DSM 50252, carry at least one LuxR solo predictably responding to AHLs. Conversely to other plantassociated Pseudomonads encoding two different luxI homologs (Zhang & Pierson, 2001; Mattiuzzo et al., 2011), only one luxI gene was found in all LuxI+ P. syringae strains analysed (Fig. 1). Similar luxI luxR content was observed in all PG3 strains examined, although this finding does not imply an identical quorum sensing regulation. For instance, several strains of P. syringae pv. phaseolicola and P. syringae pv. glycinea contain a luxI homolog, but do not produce detectable amounts of AHLs (Dumenyo et al., 1998). Another example is the type of signal molecule(s) synthesized by different strains of PG3 sharing the same genomic content: whereas Psy B728a produce 3-oxo-C6-HSL (Cha et al., 1998; Quinones et al., 2004), Pst strains synthesize mainly C6HSL and, in lower amounts, 3-oxo-C6-HSL (Taguchi et al., 2006). Similarly, we found that wild type Psv NCPPB 3335 produces exclusively C6-HSL, whereas Psv DAPP-PG 722 synthesizes 3-oxoC6and 3-oxo-C8-HSLs (Hosni et al., 2011). This particularly interesting result suggests that, in the P. syringae complex, the quorum sensing system may differ not only at the pathovar level but also among strains belonging to the same pathovar. Taking into account that the luxI homologs encoded by both Psv isolates are 100% identical, some other factor(s) might be responsible for the generation of different signal molecules. Nevertheless, overexpression of pssI in the ΔpssI mutant of Psv NCPPB 3335 also yielded 3-oxo-C6and 3-oxo-C8-HSLs, suggesting that a different expression level of this gene between these two strains might explain differences in AHLs production. AHLs are synthesized by LuxI using S-adenosylmethionine and an acyl group, provided by an acyl-carrier protein (ACP), as substrates (Whitehead et al., 2001). During AHLs synthesis, every LuxI protein shows higher affinity for a particular type of acyl-ACP (Schaefer et Chapter III 127 MATERIALS AND METHODS Bacterial strains, media and growth conditions Bacterial strains and plasmids used in this study are listed in Tables 1 and 2 respectively. P. savastanoi and Escherichia coli were grown at 28 °C and 37 °C respectively in Luria-Bertani (LB) medium (Miller, 1972) and Super Optimal Broth (SOB) (Hanahan, 1983). When required, media were supplemented with the appropriate antibiotic at the following concentration: kanamycin (Km) 10 µg ml-1 for P. savastanoi and 50 µg ml-1 for E. coli; gentamycin (Gm) 10 µg ml-1; nitrofurantoin 20 µg ml-1; and cycloheximide 100 µg ml-1. aPlasmid-cured derivative of P. savastanoi Ph3 lacking the iaaM gene (see Appendix 1) bCECT, Colección Española de Cultivos Tipo (Spanish Collection of Type Cultures). cNot available Table 1. Pseudomonas strains used in this study Strain Origin Host of isolation Source Pseudomonas savastanoi Ph3 France Mandevilla sanderi (Eltlbany et al., 2012) MSI13L Spain Mandevilla spp. (Caballo-Ponce & Ramos, 2016) MSI14S Spain Mandevilla spp. (Caballo-Ponce & Ramos, 2016) MSI14L Spain Mandevilla spp. (Caballo-Ponce & Ramos, 2016) Ph5 Germany Mandevilla sanderi (Eltlbany et al., 2012) 1397 EEUU Mandevilla splendens (Putnam et al., 2010) NIBZ1413 Slovenia Mandevilla sanderi (Pirc et al., 2014) CRpiaaMa This study pv. savastanoi NCPPB 3335 France Olea europaea (Perez-Martinez et al., 2007) pv. savastanoi CFBP 1670 Italy Olea europaea (Perez-Martinez et al., 2007) pv. savastanoi PseNe107 Nepal Olea europaea (Balestra et al., 2009) pv. nerii ITM 519 Italy Nerium oleander (Surico et al., 1985) pv. nerii ESC23 Italy Nerium oleander (Tegli et al., 2011) pv. nerii CFBP 5067 Spain Nerium oleander (Janse, 1991) pv. fraxini NCPPB 1464 UK Fraxinus excelsior (Janse, 1991) pv. fraxini NCPPB 1006 UK Fraxinus excelsior (Janse, 1991) pv. fraxini CFBP 5062 Netherlands Fraxinus excelsior (Janse, 1991) pv. retacarpa ICMP 16945 Spain Retama sphaerocarpa CECTb Pseudomonas aeruginosa PAO1 Australia Wound (Holloway, 1955) Pseudomonas syringae pv. tomato DC300 NAc Solanum lycopersicum (Cuppels, 1986) Chapter III 128 Table 2. Plasmid used in this study Name Descriptiona Source pLRM1-GFP pBBR1-MCS5::PA1/O4/O3-RBSII-GFPmut3*-T0-T1(GmR) (Rodriguez-Moreno et al., 2009) pEXP-hopA1 pCPP5040 expressing PSA3335_5065-HA tag (GmR) (Matas et al., 2014) pEXP-hopBK1 pCPP5040 expressing PSA3335_2068-HA tag (GmR) (Matas et al., 2014) pEXP-hopAO1 pCPP5040 expressing PSA3335_0875-HA tag (GmR) (Castaneda-Ojeda, 2014) pAME8 pBBR1-MCS4 expressing avrRpt2 from nptII and lacZ promoters (KmR) (Macho et al., 2009) aKm, kanamycin; Gm, gentamycin LOPAT test Levan production, oxidase, pectolysis, arginine dihydrolase and HR in tobacco leaves (LOPAT test) were evaluated for primary characterization of Psd (Lelliott et al., 1966). For levan production, single colonies were streaked into solid PVF-I media (Surico & Lavernicocca, 1989) and incubated for 48h at 28 °C. Cytochrome oxidase activity was analysed by adding one drop of BactiDrop OXIDASE (Thermo Scientific, Rockford, IL, U.S.A.) on bacteria placed on a sterile filter paper. For potato pectolysis activity, potato slices (7-8 mm thick) were washed with distilled H2O and alcohol-flamed prior to placing into Petri dishes containing a moistened and sterile filter paper. Bacterial suspensions were placed into a depression cut in the slices and incubated at 25 °C for 24 hours. Arginine dihydrolase activity was analysed on a semisolid media described by Thornley (Thornley, 1960). Tubes containing 5 ml of media were spike-inoculated, covered with a 1 centimetre thick paraffin oil layer and incubated at 28 °C for 5 days. For HR in tobacco leaves bacterial suspensions containing approximately 108 cfu/ml were infiltrated in the abaxial side using a needle-less syringe. Nicotiana tabacum cvs. Newdel and Xanthi were used and symptoms were recorded 48 hours post-infiltration with a high-resolution camera Canon D6200 (Canon Corporation, Tokyo, Japan). Different strains were used as controls: for levan production, P. syringae pv. tomato DC3000 (positive control) and E. coli DH5α (negative control); P. aeruginosa PAO1 for positive cytochrome oxidase and arginine dihydrolase activities; Erwinia carotovora pv. atroseptica SCRI1043 for positive potato pectolytic activity; and Psv NCPPB 3335 for positive HR in tobacco leaves. For the heterologous expression of the T3E avrRpt2, hopA1, hopBK1 and hopAO1 P. savastanoi Ph3 cells were transformed by electroporation (Perez-Martinez et al., 2007) with plasmids pAME8, pEXP-hopA1, pEXP-hopBK1 or pEXP-hopAO1, respectively (Table 2), transformants were selected in LB agar plates containing the appropriate antibiotic and single colonies were verified by polymerase chain reaction (PCR). Chapter III 129 DNA techniques For strains which genome sequences are not available gyrB, rpoD, gapA and rpoA housekeeping genes partial sequences were amplified by PCR using GoTaq Flexi DNA polymerase (Promega, Madison, WI, USA) and primers listed in Table 3. PCR fragments generated were sequenced (StabVida, Caparica, Portugal) and the sequences generated were submitted to GenBank (see accession numbers in Table 4). The three was constructed using MEGA5 (Tamura et al., 2011) with the maximum likelihood, minimum evolution and neighbourjoining methods. 16s rDNA partial sequences of MSI13L, MSI14S and MSI14L were amplified by PCR using GoTaq Flexi DNA polymerase (Promega, Madison, WI, USA) and primers 16s_F and 16s_R (Table 3). PCR fragments generated were sequenced (StabVida, Caparica, Portugal) and the sequences were submitted to GenBank (accession numbers KP761689, KP743983 and KP743984). Polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis of the iaaL gene was carried out as described previously (Matas et al., 2009). An internal 454 bp fragment of iaaL was amplified by PCR and digested with the restriction enzyme HaeIII (Takara Biotech Cor.; Kaohsiung, Taiwan). DNA fragments were separated by 3% agarose gel electrophoresis pre-stained with ethidium bromide. Plasmid minipreparations of Psd were made following a quick protocol (Zhou et al., 1990) with some modifications to minimize the isolation of chromosomal DNA (Murillo & Keen, 1994). Plasmids were separated by 0.8% agarose gel electrophoresis, transferred to a nylon membrane (GE Healthcare, Little Chalfont, UK) and hybridized with a dioxygenin-labelled iaaM probe. The iaaM probe was synthesized by PCR using iaaM_F615 and iaaM_R1090 primers (Table 3) and PCR DIG labelling mix (Roche, Applied Science, Mannheim, Germany) following the manufacturer’s indications. Table 3. Primers used in this study Name Sequence (5’3’) Partial sequencing of housekeeping genes gyrB_F351 GTGGTCGCGACCTTGTGC gyrB_R920 AAGTATCCGGCGGCTTG rpoD_F383 CGCCAAACGTATCGAAGAAG rpoD_R1055 GCTATTTTCAGGCCGGTTTC gapA_F220 GACCGTCAATGGTGACCG gapA_R931 GCCCATTCGTTGTCGTACC rpoA_F22 ATGCAGATTTCGGTAAATGAGT rpoA_R350 GGGTTAACGATCTCGACATC 16s_F GCTCAGATTGAACGCTGGCG 16s_R GCTACCTTGTTACGACTTCACCCC Chapter III 130 Table 3. Primers used in this study (continuation) Name Sequence (5’3’) PCR-based method for the detection of dipladenia isolates repA-F1 AGCTTCAAGAYCAGGGMAA repA-R2 ARRTCCATCARYCGGTCRAA Identification of type III secretion system effectors hopA1_F91 TAAATCCGAGGGTCAGTTGG hopA1_R228 TTGAGCACTTGGGCTATCTG hopBL1_ F1332 TCCAGCTTCGACCTTCAAC hopBL1_ R1514 AACGGCAGATTGTCCACAT hopBL2_ F1134 ACAGGGACGAAGGCTTGGA hopBL2_ R1241 CGGAGACATGGTGGATGGA hopAA1_F18 CACCATGCACATCAACCGATCC hopAA1_R552 TCATGGATGCCTGTCCGG hopBK1_F66 CGAGCACAGCAACTCCTTG hopBK1_R195 TGGTGTACCGTCCAGGTAG hopAF1-F GCGGATGCATTGAAACAGCT hopAF1-R AATAACAGCGCCCAGCAGAGT hopAF1-2F CACTGCGGAGGCATTGAAAA hopAF1-2R AGAATAACGGCGCCAAGCA hopAO1_F TGGTTGTGAACGCATTTCCTC hopAO1_R GCATGAGATTCTTCGCGCA hopAO2_F TGCGTTTGATGGTGACCGA hopAO2_R ACCGCAATGGATATGTACCCG PCR-RFLP of the iaaL gene iaaL-F-221 GGCACCAGCGGCAACATCAA iaaL-R-696 CGCCCTCGGAACTGCCATAC Synthesis of the iaaM probe iaaM_F615 GCATGGCAGGGATGGC iaaM_R1090 CGCGGACAGCACGAGC Quantification of indole 3-acetic acid production The production of indole 3-acetic acid (IAA) was detected using Salkowski reagent (1.8% 0.5M FeCl3 in 37% sulphuric acid). Culture supernatant were mixed with Salkowski reagent (1:2 v/v) and incubated at room temperature under dark conditions. After 20 minutes, IAA-producing bacteria developed a pink colour. For the quantification of IAA, bacterial cultures were incubated at 28 °C in King’s B (KB) broth (King et al., 1954) or KB amended with 2.5 mM tryptophan at an initial OD600nm of 0.1. After 24 hours, 5 ml of culture were centrifuged, the supernatant was collected and acidified to pH 2.5-3 with 1M HCl. Then, it was extracted with ethyl acetate (1:1 v/v) and, after incubation at room temperature in a shaker for 30 minutes (extraction step), the organic phase was recovered. The extraction step was repeated two more times, the organic phases of the three subsequent extractions were put together and evaporated at room Chapter III 131 temperature. The pellet was dissolved in 1 ml methanol-water (10:90 v/v) and analysed by High Performance Liquid Chromatography coupled to Mass Spectrometry (HPLC-MS) as detailed recently (Aragon et al., 2014). IAA amounts were normalized to the dry weight (DW) of 5 ml pellet, which was obtained by complete drying at 80 °C for, at least, two hours. Extractions of IAA were performed in triplicate. Plant infections Plants were pre-treated with 300 g/hl of Bordeaux mixture (stock containing 20% CuSO4). After 3 weeks, plants were washed with 70% ethanol and air dried prior to inoculation. Olea europaea (olive), Nerium oleander (oleander), Fraxinus excelsior (ash) and Mandevilla spp. (dipladenia) were wounded along the stem with a scalpel and approximately 106 cfu were placed per wound. For Retama sphaerocarpa (Spanish broom) around 106 bacteria were inoculated with a needle coupled to a syringe. Each inoculation point was wrapped with parafilm (Bemis, Neenah, WI, USA) for 7 days. Plants were kept in a greenhouse for 3 months under natural photoperiod (15 hours light/9 hours dark) at room temperature (≈26 °C day/18 °C night). Symptoms developed were captured with a high-resolution camera Canon D6200 (Canon Corporation, Tokyo, Japan). For the isolation of P. savastanoi from dipladenia, knots were macerated in 10 mM MgCl2 with a mortar and a pestle, serial diluted and plated into LB containing nitrofurantoin 20 µg ml-1 and cycloheximide 100 µg ml-1. To fulfil Koch’s postulates, identification of dipladenia isolates was carried out using the repA PCR-based method (Eltlbany et al., 2012). To visualize bacterial infections in dipladenia, P. savastanoi Ph3 and CRpiaaM were tagged with pLRM1-GFP plasmid and plants were examined with a epifluorescence microscopy as previously described (Rodriguez-Moreno et al., 2009). Psd Ph3 on dipladenia bacteria were transformed by electroporation (Perez-Martinez et al., 2007) with plasmid pLRM1-GFP (Table 2) and transformants were selected as green colonies under UV light. To visualize bacterial infection on dipladenia in real-time, plants were inoculated as described above and symptomatic areas were examined at 23 days post-inoculation directly with a stereoscopic fluorescence microscope (Leica MZ FLIII) equipped with a 100 W mercury lamp and a GFP2 filter (excitation 480/40 nm; emission 510LP nm). Images were captured with a high-resolution digital camera (Nikon DXM 1200). Chapter III 132 Table 4. Locus tag/accession numbers for the gene sequences used to construct the phylogenetic tree shown in Figure 1 Strains gyrB rpoD gapA rpoA Pseudomonas savastanoi isolated from dipladenia Ph3 KX296761a KP730688a KX353842a KX379606a 13L KX296762a KP730689a KX353843a KX379607a Ph5 KX296763a KX369597a KX353844a KX379608a 1397 KX296764a KX369598a KX353845a KX379609a NIBZ1413 KX296765a KX369599a KX353846a KX379610a P.savastanoi pv. savastanoi NCPPB 3335 PSA3335_0009 PSA3335_0701 PSA3335_1269 PSA3335_4553 CFBP 1670 ALO58_04531 ALO58_02277 ALO58_00431 ALO58_04894 DAPP-PG 722 GS14_RS0118475 GS14_RS0120510 GS14_RS0113260 GS14_RS0115670 PseNe107 PSAVPseNe107_RS21055 PSAVPseNe107_RS21455 PSAVPseNe107_RS06270 PSAVPseNe107_RS20530 P. savastanoi pv. nerii ESC23 KX296766a KX369600a KX353847a KX379611a CFBP 5067 ALO61_04256 ALO61_03410 ALO61_04841 ALO61_03136 P. savastanoi pv. fraxini NCPPB 1006 KX296767a KX369601a KX353848a KX379612a CFBP 5062 ALP79_200380 ALP79_04657 ALP79_02304 ALP79_00291 P. savastanoi pv. retacarpa ICMP 16945 ALO49_04129b ALO49_200246 ALO49_03177 ALO49_00201 P. syringae pv. actinidae M302091 PSYAC_05930 PSYAC_04173 PSYAC_00605 PSYAC_19988 pv. aesculi NCPPB 3681 PsyrpaN_010100024288 PsyrpaN_010100021637 PsyrpaN_010100017259 PsyrpaN_010100001555 pv. glycinea race 4 Pgy4_23473 Pgy4_22411 Pgy4_05892 Pgy4_14629 pv. japonica M301072 PSYJA_14042 PSYJA_15502 PSYJA_03099 PSYJA_19971 pv. lachrymans M302278 PLA106_16879 PLA106_16494 PLA106_04632 PLA106_19276 pv. mori 301020 PSYMO_16078 PSYMO_21228 PSYMO_17513 PSYMO_25203 pv. morsprunorum M302280 PSYMP_13789 PSYMP_21479 PSYMP_02311 PSYMP_24736 Chapter III 133 aAccession numbers at GenBank for the sequences generated in this study bSequence cured manually Table 4. Locus tag/accession numbers for the gene sequences used to construct the phylogenetic tree shown in Figure 1 (continuation). Strains gyrB rpoD gapA rpoA pv. oryzae 1_6 Psyrpo1_010100022716 Psyrpo1_010100026871 Psyrpo1_010100037909 Psyrpo1_010100001310 pv. phaseolicola 1448A PSPPH_0004 PSPPH_0619 PSPPH_1176 PSPPH_4567 pv. syringae B728a Psyr_0004 Psyr_4641 Psyr_1108 Psyr_4524 pv. tabaci ATCC11528 PSYTB_19816 PSYTB_18734 PSYTB_04965 PSYTB_22761 pv. tomato DC3000 PSPTO_0004 PSPTO_0537 PSPTO_1287 PSPTO_0651 Pseudomonas aeruginosa PAO1 PA0004 PA0576 PA3195 PA4238 Chapter III 134 RESULTS First Report of Dipladenia (Mandevilla spp.) Leaf and Stem Spot Caused by Pseudomonas savastanoi in Spain The evergreen climber dipladenia (Mandevilla spp.), native to South and Central America, has lately become widespread as an ornamental plant around the world. In May 2013 and in November 2014, dipladenia plants showing necrotic lesions surrounded by a chlorotic halo on leaves and stems were detected in the south of Spain (Fig. 1A-B). To isolate the causal agent, symptomatic leaves and stems were homogenized by mechanical disruption into 10 mM MgCl2, and serial dilutions were plated. Circular, smooth, flat, cream-colored colonies were isolated on LB containing 20 µg/ml nitrofurantoin and 100 µg/ml cycloheximide. Three single isolates, two from leaves (MSI13L, 2013 and MSI14L, 2014) and one from a stem (MSI14S, 2014), were obtained and identified as Pseudomonas savastanoi on the basis of phenotypic and molecular tests. In agreement with Pirc et al. (2014), the three isolates were weakly fluorescent on KB under UV light and showed a negative reaction in all phenotypes included in the LOPAT test (levan, oxidase, arginine dihydrolase, pectinolytic activity and tobacco hypersensitivity). Genomic DNA was extracted from the isolates, and DNA fragments corresponding to the rpoD and the 16s ribosomal RNA genes were amplified (Parkinson et al. 2011) and sequenced. The obtained sequences were submitted to GenBank (accession numbers KP730689, KP739953 and KP739954 for the rpoD gene; KP761689, KP743983 and KP743984 for the 16s ribosomal RNA gene). All sequences showed 100% identity and coverage with those corresponding to P. savastanoi Ph3 (JX678983 and KP730688), a strain isolated from Mandevilla sanderi (Eltlbany et al. 2012). Additionally, PCR amplicons of the expected size were obtained from all isolates using a repA-based method (1,100 bp) directed specifically to P. savastanoi isolates from Mandevilla (Eltlbany et al. 2012). Pathogenicity was examined on both leaves and stems of dipladenia. Bacterial suspensions (≈1x108 cfu/ml in 10 mM MgCl2) were prepared and ≈200 µl were infiltrated into the abaxial side of the leaves using a needless syringe. Stems were wounded with a sterile scalpel and inoculated with bacteria (≈2x105 cfu per wound). Negative control plants were inoculated with 10 mM MgCl2. Three to five plants (4 months old), inoculated in three leaves or three stem points, were used for each strain and treatment. Plants were grown with a 16/8 h light/dark photoperiod at 24/18 °C day/night. Necrotic symptoms were observed on inoculated leaves and stems at seven (Fig. 1C) and 21 (Fig. 1D) days post-inoculation (dpi), respectively. Infected stems also showed a slight swelling, which increased over time and developed into a necrotic overgrowth at 60 dpi (Fig. 1E). Control plants did not develop any symptoms. Bacterial cells were re-isolated from symptomatic leaves and stems and identified as Chapter III 135 P. savastanoi by the repA-based detection method described above (Fig. 2). Thus, Koch’s postulates were fulfilled for the three Spanish isolates used in this study. After identification of dipladenia leaf and stem spot caused by P. savastanoi in France (Eltlbany et al. 2012), Germany (Eltlbany et al. 2012), USA (Putnam et al. 2010) and Slovenia (Pirc et al. 2014), this study shows the spread of this emergent disease into Spain. This work has been recently published in Plant Disease, a journal of the American Phytopathological Society (APS) (Caballo-Ponce & Ramos, 2016). Strain MSI13L was included in a collection of Psd strains used for the characterization tests described below. Apart from the Spanish Psd isolate, other Psd strains from France (Ph3), Germany (Ph5), the USA (1397) and Slovenia (NIBZ1413) were included in this collection (Table 1). Figure 1. Symptomatology incited by P. savastanoi in dipladenia. Symptoms caused by P. savastanoi on naturally infected dipladenia (A) leaves and (B) stems from which strains MSI13L and MSI14S were respectively isolated. Symptoms induced on the (C) leaves and (D, E) stems of dipladenia plants (C) 7, (D) 21 and (E) 60 days after artificial inoculation with strain MSI13L. Figure 2. Identification of P. savastanoi strains isolated from dipladenia plants using a PCR method based on the repA gene (Eltlbany et al., 2012). (L) Molecular weight DNA marker (1 kb plus DNA ladder, Chapter III 136 Invitrogen); P. savastanoi strains (1) Ph3; (2) MSI13L; (3) MSI14S; (4) MSI14L and (5) P. savastanoi pv. savastanoi NCPPB 3335. P. savastanoi strains isolated from dipladenia encode a functional T3SS but are unable to elicit the HR in tobacco leaves Biochemical features included in the LOPAT test (Levan, Oxidase, Pectolysis, Arginine dihydrolase and hypersensitive response in Tobacco leaves) have been traditionally used for the identification of bacteria from the P. syringae complex, which also include P. savastanoi. Most P. syringae strains are L+O-P-A-T+. However, over 80% of P. savastanoi strains are negative for levan, oxidase, and arginine dihydrolase (L-O-A-), are able to trigger an HR on tobacco leaves (T+) but show a variable pectolysis phenotype on potato (P+/-) (Schaad et al., 2001). Interestingly, the LOPAT profile of Psd strains isolated in Spain and Slovenia has been shown to be negative for every LOPAT feature (L-O-P-A-T-) (Pirc et al., 2014; Caballo-Ponce & Ramos, 2016). To further analyse this observation, five Psd strains (Ph3, MSI13L, Ph5, 1397 and NIBZ1413; Table 1) isolated in different countries were selected and their LOPAT profiles were established. As shown for the Slovenian and Spanish isolates, all five strains resulted to be L-OP-A-T-. Figure 3A shows the results obtained for the HR test in tobacco leaves in comparison to Psv NCPPB 3335. Taking into account that the HR response is a reaction triggered by the plant immune system after recognition of type III secretion system (T3SS) effectors (T3E) secreted by the pathogen into the plant cell, we analysed the functionality of the T3SS in the model Psd strain Ph3, isolated from M. sanderi in France (Eltlbany et al., 2012). Psd Ph3 was transformed with plasmid pAME8 (Table 2) expressing the T3E avrRpt2 from P. syringae pv. tomato 1065 (Macho et al., 2009), an effector known to elicit a visible HR in N. tabacum cv. Xanthi (Mudgett & Staskawicz, 1999). As shown in Figure 3B, tobacco leaves infected with the transformed strain Ph3 (pAME8) developed an eye visible cell death typical of the HR 48h after inoculation, demonstrating the functionality of the T3SS in this strain. Translocation of nine T3E through the Psv NCPPB 3335 T3SS has recently been reported (Matas et al., 2014; Castañeda-Ojeda et al., 2016). Codification of these nine T3E in the genome of Psd Ph3 was analysed by PCR using primers designed according to the sequences of their respective genes in the genome of Psv NCPPB 3335. Whereas no amplicons were obtained for hopAO1, hopA1 and hopBK1, fragments of the expected sizes were visualized for hopAA1, hopBL1, hopBL2, hopAF1, hopAF1-2 and hopAO2 (data not shown). P. savastanoi Ph3 was transformed with plasmids expressing the Psv NCPPB 3335 T3E hopAO1, hopA1 or hopBK1 (Table Chapter III 143 Figure 7. Symptoms generated by P. savastanoi strains isolated from diverse hosts on dipladenia at 90 dpi. Psf strains, symptoms induced by P. savastanoi pv. fraxini isolates NCPPB 1006, NCPPB 1464 and CFBP 5062; Psv strains, symptoms induced by P. savastanoi pv. savastanoi strains NCPPB 3335, PseNe107 and CFBP 1670; Psd Ph3; symptoms induced by the dipladenia isolate P. savastanoi strain Ph3. Symptoms generated by P. savastanoi pv. nerii (Psn) were strain-dependent. A plasmid-cured derivative of P. savastanoi Ph3 lacking the iaaM gene is unable to induce tumour formation in dipladenia P. savastanoi synthesizes the auxin IAA through the indole 3-acetamide pathway, where tryptophan is firstly transformed by the IaaM protein to indole 3-acetamide, that is further converted into IAA by IaaH. Psd Ph3 and Psd Ph5 contain a plasmid-encoded copy of the iaaM gene (Eltlbany et al., 2012), although its role in the production of IAA and the virulence of Psd has not been reported. Using a method developed for curing P. savastanoi plasmids encoding a copy of the iaaM gene (Comai & Kosuge, 1980), a P. savastanoi Ph3 plasmid-cured derivative lacking the iaaM gene (CRpiaaM) was isolated (see Appendix 1). To analyse whether production of IAA was impaired in CRpiaaM in comparison to the wild type strain, the supernatant of bacterial cultures amended or not with 2.5 mM tryptophan were analysed by HPLC-MS as described in the materials and methods section. In addition, all five selected Psd strains were included in these assays. As shown in Figure 8A, all five wild type Psd strains analysed produced IAA in KB broth, and the production was enhanced when tryptophan was supplied exogenously. However, production of IAA by Psd CRpiaaM was reduced more than 10 times in comparison with Psd Ph3. These results are in agreement with the loss of the iaaM gene in the CRpiaaM strain (see Appendix 1). Taking into account that CRpiaaM still produces a small amount of IAA, which is also increased in the presence of tryptophan (Fig. 8A), it could be possible that a second copy of the iaaMH operon is encoded in the chromosome of this strain. Psd CRpiaaM was inoculated on dipladenia stems to analyse its ability to cause an infection. After two weeks, plants inoculated with the wild type strain Psd Ph3 developed symptoms, Chapter III 144 consisting on a light green halo around the inoculation point, but these symptoms were not observed in the plants infected with CRpiaaM. Whereas no symptoms were generated on the plants inoculated with CRpiaaM at 60 days post-inoculation, Psd Ph3 induced the development of necrotic overgrowths (Fig. 8B). Additionally, approximately 108 cfu of P. savastanoi Ph3 were isolated per knot, but less than 102 cfu were recovered from the inoculation points of plants infected with CRpiaaM. All these results suggest that IAA production might be crucial for bacterial survival and for the induction of symptoms by Psd on dipladenia plants. These results are in agreement with the attenuation of the virulence observed for Psv and Psn mutants impaired in the production of this phytohormone (Smidt & Kosuge, 1978; Surico et al., 1985; Iacobellis et al., 1994; Aragon et al., 2014). Figure 8. IAA production and virulence of P. savastanoi strain Ph3 and its mutant derivative cured of the plasmid-encoded iaaM gene. (A) Quantification of IAA produced by the indicated P. savastanoi strains isolated from dipladenia. Bacteria were grown in KB (blue bars) or KB amended with 2.5 mM tryptophan (red bars) for 24 hours at 28 °C. Data are represented as part per billion (ppb) of IAA per milligram of dry weight of biomass (DW). Bars represent the mean of three biological replicates, error bars correspond to the standard deviation. (B) Symptoms induced by P. savastanoi Ph3 and its plasmid-cured derivative lacking the iaaM gene (CRpiaaM) on dipladenia stems at 60 days post-inoculation. P. savastanoi Ph3 can produce a systemic infection that leads to plant death Secondary symptoms, which consist on the generation of overgrowths on the stems in sites distant to the points of inoculation and the appearance of chlorotic and necrotic areas on noninoculated leaves, were occasionally observed in dipladenia plants inoculated with Psd Ph3, suggesting that the pathogen might spread systemically through the plant. To address the possible migration, Psd Ph3 was transformed with pLRM1-GFP, a plasmid expressing the green fluorescent protein (GFP) from a constitutive promoter (Rodriguez-Moreno et al., 2009), and inoculated on dipladenia stems. After 23 days, secondary symptoms were clearly visualized (Fig. 9A) and GFP fluorescence was then examined using an epifluorescence microscope. GFP fluorescence was clearly visible under UV light in symptomatic non-inoculated leaves (Fig. 9B) Chapter III 145 and petioles (Fig. 9C), whereas no fluorescence was detected in the leaves (Fig. 9D) and petioles (Fig. 9E) of a non-inoculated plant, showing the migration of P. savastanoi Ph3 through the plant. Secondary symptoms were more evident with time and led to the dieback of the plant after 90 days (Fig. 9F). Figure 9. Systemic infection caused by P. savastanoi Ph3 on dipladenia plants. (A) Development of secondary symptoms induced at 23 days post-inoculation by a GFP-tagged derivative of P. savastanoi Ph3. The asterisk indicate the inoculation point. White boxes correspond to the leaf and petiole areas which epifluorescence images are shown in (B) and (C), respectively. (D) and (E) epifluorescence images of a leaf and petiole from a non-inoculated dipladenia plant (negative control). The red background is due to chlorophyll fluorescence. (F) Bacterial wilt occasionally observed in dipladenia plants infected with P. savastanoi Ph3 at 90 days post-inoculation. Chapter III 146 DISCUSSION P. savastanoi is a prevalent pathogen of dipladenia, although only one study delved into the genotypic characterization of Psd strains isolated in Germany and France (Eltlbany et al., 2012). Dipladenia plants showing leaf and stem spots were detected during this PhD Thesis in Spain and the causative agent was identified as P. savastanoi (Caballo-Ponce & Ramos, 2016). In agreement with results reported for Slovenian isolates (Pirc et al., 2014), here we show that P. savastanoi strains isolated from dipladenia in Spain are unable to trigger an HR on tobacco leaves. Results exposed in this Chapter broaden this observation to all Psd strains examined here and discarded the possibility that Psd strain Ph3 encoded a non-functional T3SS. Therefore, it is possible that (i) Psd strains lack a T3E recognized by the plant immune system leading to HR elicitation in the tobacco cultivar used in this study, or (ii) Psd strains encode a T3E that supress plant immune responses and prevent induction of the HR on N. tabacum leaves. Additional work is necessary to determine the T3E repertoire in Psd and to identify the T3SS effector(s) responsible for this phenotype. An epidemiological marker established before by our laboratory (Matas et al., 2009) was used to examine five Psd strains isolated in different countries. No variability was found among these Psd strains, which is in agreement with previous data that revealed similarities between BOX-PCR fingerprints of French and German isolates (Eltlbany et al., 2012). In addition, Psd strains isolated in France and Germany showed identical plasmid restriction hybridization patterns using a repA probe (Eltlbany et al., 2012), which might indicate that these strains harbour the same plasmid content. However, results in this Chapter showed that Psd Ph3 and Psd Ph5 differ in their native plasmid content, although the analysed French, Slovenian and Spanish isolates shared the same plasmid profile. Therefore, plasmid purification and separation in agarose gel seems to be more resolutive than digestion followed by hybridization with a repA probe to identify differences in the plasmid content of the strains. Gathering all this information and considering that the first report of dipladenia spots caused by P. savastanoi date back to the late 2000s (Putnam et al., 2010), it is possible that this disease has emerged recently, not allowing the generation of substantial genetic diversity among different Psd isolates. The phylogenetic analysis shown in this Chapter, constructed with the partial sequence of four housekeeping genes, revealed that all Psd are phylogenetically related to Psn strains. These results are in agreement with those obtained by analysis of the 16s rDNA (Eltlbany et al., 2012) and rpoD partial sequences (Pirc et al., 2014). Another finding supporting the proximity of Psn and Psd strains emerged from the cross-pathogenicity tests: two out of three Psn strains were able to induce knots/excrescences on dipladenia. Curiously, the most virulent Psn isolate on Chapter III 147 dipladenia was CFBP 5067, which did not show a 100% phylogenetic association with the other Psd strains analysed. In contrast, Psn ESC23 was not pathogenic on this host. It is well described that the T3E repertoire highly vary in the P. syringae complex from one strain to another and this diversity might explain differences in host range (Baltrus et al., 2011). Thus, the T3E pool of Psn CFBP 5067 might be different to that of Psn ESC23 and permit the pathogenicity on dipladenia. In relation to this, it is important to mention that oleander and dipladenia belong to the Apocynaceae family, what might have facilitated adaptation of the pathogen from one host to the other. Thus, placing dipladenia nurseries close to oleander plantations should be avoided to prevent the generation new outbreaks of the disease. Cross-pathogenicity test showed a unique host range for Psd strains compared to those of the well established P. savastanoi pathovars. Moreover, the host range of all Psd strains here analysed is common, which contrast with the strain-dependent pathogenicity observed for P. savastanoi strains isolated from myrtle (Cinelli et al., 2014) and with the results here obtained for Psn strains in dipladenia. Symptoms generated by Psd on oleander and olive (Table 5) also differ from the observations mentioned by Eltlbany et al. (2012), but not published. Among other factors, it is well known that the age of the plant and the cultivar contribute to the generation of symptoms after artificial P. savastanoi inoculations (Penyalver et al., 2006). Therefore, the observed differences could be attributed to the selection of distinct plant cultivars and/or to the age of the plants. A Psd Ph3 plasmid-cured derivative lacking the iaaM gene (CRpiaaM, see Appendix 1) did not develop a knot on dipladenia (Figure 8B), which is consistent with the role of this gene in the virulence of P. savastanoi (Smidt & Kosuge, 1978; Surico et al., 1985; Iacobellis et al., 1994; Aragon et al., 2014). Approximately 108 cfu were isolated from the knots generated by wild type Psd Ph3, whereas very little or no cfu were recovered from the sites inoculated with CRpiaaM, indicating that this strain is impaired on its survival in dipladenia. Consistent with this observation is that a Psv NCPPB 3335 mutant strain in the iaaMH-1 operon is outcompeted by the wild type strain on olive plants (Aragon et al., 2014). Curiously, CRpiaaM still produces a small amount of IAA, perhaps due to the presence of a chromosome-encoded copy of the iaaMH operon. In this regard, codification of a second iaaMH operon (iaaMH-2) has been reported in Psv NCPPB 3335; however, this operon is not functional and does not contribute to the synthesis of IAA nor to the virulence of the strain on olive (Aragon et al., 2014). As also suggested for Psv NCPPB 3335, a different pathway for the biosynthesis of IAA might be encoded in the genome of Psd Ph3. As IAA production by CRpiaaM is also enhanced by tryptophan, an IAA pathway different from that dependent on the iaaMH operon might be active in this strain. Three Chapter III 148 additional tryptophan-dependent pathways (indole 3-acetonitrile, indole 3-pyruvate and tryptamine) have been proposed for bacterial production of IAA (Spaepen et al., 2007). However, all sequenced P. savastanoi strains lack genes required to complete each of these three pathways. Considering the later information, two situations are possible: (i) that none of these biosynthetic pathways is active in P. savastanoi, or (ii) that other P. savastanoi enzymes complement the function of the missing proteins. Results in this Chapter demonstrate that Psd Ph3 is able to migrate in dipladenia and induce the development of necrotic lesions several centimetres away from the point of inoculation, a phenotype previously observed for Psv and Psn strains (Wilson, 1935; Wilson & Magie, 1964; Penyalver et al., 2006). Previously, it was reported the association of Psv cells with xylem vessels (Marchi et al., 2009; Rodriguez-Moreno et al., 2009; Maldonado-González et al., 2013) and of Psn with laticifers (Wilson & Magie, 1964). Since dipladenia and oleander belong to the Apocynaceae family and share latex production, Psn and Psd might use a common mechanism(s) for their dissemination through the plant. When occurring, the systemic spreading of P. savastanoi in dipladenia brings the wilting of the plant, a phenomenon not reported before in dipladenia plants infected with P. savastanoi. The systemic infection caused by Psd on dipladenia must be considered by nurseries to control the disease and prevent its dissemination. Implementation of control methods including plant removal and incineration of infected plants might help preventing dissemination of the pathogen. In summary, results shown in this Chapter suggest that the P. savastanoi strains pathogenic to dipladenia might have emerged recently. Evidences for the proximity of Psn and Psd strains from a phylogenetic and pathogenicity perspectives are provided. However, the unique host range of Psd isolates compared with those of the four well-established P. savastanoi pathovars suggest that these strains might constitute a novel pathovar of this specie. Wilting of dipladenia due to systemic infection of Psd was observed, which should be considered in dipladenia management to prevent pathogen dissemination. Concluding remarks Concluding remarks 151 The study of Pseudomonas savastanoi interactions with its hosts has been addressed from diverse perspectives in this PhD Thesis. P. savastanoi pv. savastanoi (Psv) has been established as a model for the study of bacterial molecular interactions with a woody host (Ramos et al., 2012); in this regard considerable advances have been achieved over the last years in unveiling the factors contributing to the virulence and adaptation of Psv to olive (Perez-Martinez et al., 2010; Bardaji et al., 2011; Hosni et al., 2011; Matas et al., 2012; Aragon et al., 2014; Matas et al., 2014; Aragon et al., 2015a; Aragon et al., 2015b). A genomic region probably involved in the metabolism of phenolic compounds has been shown to be unique to strains from the P. syringae complex isolated from woody hosts (Rodriguez-Palenzuela et al., 2010; Ramos et al., 2012; Nowell et al., 2016). Chapter I of this PhD Thesis broaden the distribution of this region to many other P. syringae strains which genomes have been recently sequenced (Moretti et al., 2014; Bartoli et al., 2015b; Thakur et al., 2016) and provides the first evidence of the WHOP region (from woody host and Pseudomonas) contribution to virulence and fitness of Psv NCPPB 3335 into woody olive plants (Figure 1). The identification and functional characterization of virulence factors capture the scientific interest. However, compared to herbaceous plant pathogens, little attention is currently paid to bacterial virulence factors of woody host pathogens. In this sense, Chapter I confirmed the aromatic compounds catabolism as a novel virulence factor of Psv NCPPB 3335 that play a role in bacteria-woody host interactions and particularly enhance the knowledge of bacterial interactions with a woody host. In addition, Chapter II focused on the quorum sensing system of Psv NCPPB 3335, a well-defined virulence factor for Psv DAPP-PG 722 (Hosni et al., 2011), and show differences with the quorum sensing systems of two P. syringae strains. Comparative genomic analyses of strains from the P. syringae complex will allow future identification of novel pathovar-specific elements that might define the host range of particular strains. In this sense, the identification of P. savastanoi strains causing stem knots and leaf spots on dipladenia was reported during progression of this PhD Thesis (Eltlbany et al., 2012; Pirc et al., 2014). Evidences of the relationship between P. savastanoi strains isolated from dipladenia (Psd) and oleander (P. savastanoi pv. nerii, Psn) are provided in Chapter III, although their host range is different. Sequencing of the genome of a Psd strain and comparative analysis with the pathotype Psn strain ICMP 16943, which genome sequence has been recently published (Thakur et al., 2016), could explain the differences found in the host range of these bacteria. Furthermore, the type III secretion system (T3SS) effectors (T3E) repertoire has been proposed to play a key role in determining the host range (Baltrus et al., 2011). Variability in the T3E repertoire of P. savastanoi Ph3 and Psv NCPPB 3335 might be responsible for the different plant Concluding remarks 152 responses observed after infiltration of these strains in tobacco leaves. A RNAseq-based strategy similar to that used in Chapter II could be used to find genes regulated by HrpL, a transcription factor that promotes the transcription of the T3SS and its T3E (Lindeberg et al., 2005). Unlike the LB medium used for the RNAseq in Chapter II, a minimal medium that mimics the plant apoplast and promote T3E expression, as the hrp medium (Huynh et al., 1989), should be used for the identification of the HrpL regulon. Since Psv NCPPB 3335 is unable to grow in hrp medium and, hence, do not reach the quorum, it could not be used in this PhD Thesis for the identification of genes controlled by PssI. It is important to remark here that many Psv virulence genes are not expressed in LB and its pssI-dependency could not be examined. Nonetheless, RNA sequencing data shown in Chapter II might also serve to map the transcription start points and to identify the active promoters of Psv NCPPB 3335 under the conditions tested. P. savastanoi strains pathogenic on dipladenia (Psd) in Spain were isolated during progression of this PhD Thesis (Chapter III) (Caballo-Ponce & Ramos, 2016), which constitute a severe problem given that Spain is one of the main European producers of dipladenia. Sequencing of a Psd strain genome would also help to identify putative targets to design control strategies against the pathogen. Dissemination of Psd strain Ph3 through dipladenia has been observed (Chapter III, Figure 9), which occasionally led to the bacterial wilting of the plant and contrast with the consequences of systemic Psv strains infections in one-year old woody olive plants (Penyalver et al., 2006). Bacterial wilting of dipladenia was observed in four-month old non-woody plants, which is the plant format chosen for its commercialisation. Data not shown in this PhD Thesis revealed high sensitivity of Psd strains to cupric compounds; therefore, chemical treatments to prevent the disease should be carried out, as well as incinerating the infected plants to prevent the dissemination of the pathogen. Additionally, identification of dipladenia cultivars more resistant to P. savastanoi infections is a future challenge for breeders. All these practices should be integrated for an efficient control of P. savastanoi infections in dipladenia. Sequencing of the Psv NCPPB 3335 genome (Rodriguez-Palenzuela et al., 2010) opened a vast repertoire of putative virulence factors in this bacterial pathogen, several of which were later confirmed using a functional genomic strategy (Matas et al., 2012). Although the role in virulence of the indole 3-acetic acid (Smidt & Kosuge, 1978; Surico et al., 1985; Iacobellis et al., 1994) and cytokinins (Iacobellis et al., 1994) was reported more than 30 years ago, during the last few years several studies offered new insights into the mode of action of these phytohormones in Psv (Bardaji et al., 2011; Aragon et al., 2014). In addition, several other mechanisms have been recently studied in detail, i.e. the T3SS and the T3E repertoire of Psv