A mis aitas A Joseba

La vida es una obra de teatro que no permite ensayos; por eso canta, ríe, baila, llora y vive intensamente cada momento de tu vida antes que el telón baje y la obra termine sin aplausos.

Charles Chaplin

Agradecimientos

He leído un proverbio Masai que dice que Si quieres ir rápido camina solo, si quieres llegar lejos ve acompañado. Nada de lo que he conseguido hubiera sido posible sin todos y cada uno de los que me habéis acompañado y ayudado a construir este camino.

Por eso, quiero agradecer, en primer lugar, a todo el Departamento de Microbiología y

Parasitología de la Universidad de Navarra por hacerme sentir como en casa. Sois una gran familia, gracias por acogerme con los brazos abiertos desde el día en que llegué.

Gracias a las doctoras Maite Iriarte y Raquel Conde por la excelente dirección de este trabajo, el apoyo y el cariño que me habéis dado durante estos años. Raquel, gracias por estar en todo momento disponible para resolver mis dudas, por alegrarte de las buenas noticias (incluso a veces más que yo) y hacerme ver que las malas eran menos malas. Porque sin ti nada de esto hubiera sido posible. Por involucrarte siempre con ilusión y saber transmitirme ganas de aprender. Maite, por la paciencia, por enseñarme a escribir entendiendo el porqué de las cosas y hacer que las ideas cobrasen sentido; por tu implicación en esta recta final de la tesis. Gracias también al Doctor Ignacio Moriyón, por ser el alma del grupo Brucella, gracias por tus consejos y ayuda, por enseñarnos a mirar con otros ojos. Al Doctor Ignacio López-Goñi, por las correcciones de esta tesis; por valorar mi trabajo.

Gracias también al resto de profesores y doctores del Departamento: Guillermo Martínez de Tejada, Carlos Gamazo, Paul Nguewa, Anabel Vitas y David González por el apoyo, el cariño y vuestra sonrisa.

Gracias a la doctora Begoña Alonso, por tu gracia y salero, por enseñarnos a transmitir pasión en las explicaciones. A María Orbe, la esencia de Micro; por ser la principal responsable de que en este departamento se respire esa atmósfera tan especial que te hace sentir como en casa. Tu eficacia y eficiencia son asombrosas y es un gusto trabajar contigo.

Por hacer las cosas de buen humor y siempre con la mejor sonrisa. Gracias a Rosario, por tu cariño, tu buen humor y por cuidar de nosotros como una madre.

A mis compañeros de batallas durante el día a día. Gracias de corazón a Alberto, por tu buen humor, tu gusto por las cosas bien hechas, por tu PACIENCIA y tu disposición para ayudarme en todo momento, aun siendo una CANSINA con mayúsculas. A Amaia, gracias por enseñarme prácticamente todo lo que se. Por ayudarme a madurar tanto en el laboratorio como fuera de él. Por saber exactamente (a veces incluso mejor que yo) qué estoy pensando y como me siento en cada momento. Gracias porque juntas conseguimos uno de nuestros objetivos pendientes. De ti he heredado el gusto por las cosas bien hechas y he aprendido a quedarme con lo bueno de las cosas y saber valorar lo que tenemos. Gracias por dejarme compartir contigo las aventuras que están por venir. Celia, por tu buen humor y por traer la alegría a nuestras vidas junto con Olivia. A Leti, gracias por cederme un porcentaje de tu amistad (sea cual sea), por esas interesantes conversaciones que hacían que trabajar en el P3 fuera apetecible incluso los viernes por la tarde. Por tus “leticiadas”, por tu sentido del humor y por hacerme reír. A Miriam, los malos momentos son menos malos contigo aquí. Gracias por hacerme reír con tus nombres y exageraciones de maña. A Bea, por tu alegría, tu dulzura y por contagiar ganas de aprender y de hacer planes. Por hacerme ver que, si se quiere, se puede. Gracias por tu cariño cada día. A Yadira, porque siempre consigues sacarnos una sonrisa a todos. Por tu cariño, tu buen humor y por saber entenderme. A Lara, porque desde el día que llegaste supe que íbamos a llevarnos bien. Por compartir todos mis momentos como si fueran los tuyos propios. Porque sé que siempre puedo contar contigo. Tu cariño siempre me ha ayudado en los momentos más difíciles.

Pepe, gracias por traer tu gracia malagueña a nuestras vidas. Eres un grande, gracias por haber nacido en viernes ;-). A Melibea, María Fernández y Maite Loperena, las nuevas generaciones que habéis sabido calar hondo en esta familia. Maite, nueva chica Brucella;

Ezkaroz (y Garralda…) tienen una buena representante en la ciencia y en la música. Hawraa, thanks for coming, for your smile. I hope to try again lebanese food and dance. A los ya doctores Ana Camacho, Pedro, Raquel Ferrer, Estrella, Andrés y Sergio, quienes, cerca o lejos, seguís formando parte de esto. Ana, gracias por ser mi “abuela” favorita, por tus consejos y por cuidar siempre de mí. Pedro, por tu alegría y buen humor. Tenías razón; al final conseguí sacarle más punta al núcleo del LPS. Raquel, por los viejos tiempos, los momentos compartidos en campana, y porque con buena música y buena compañía todo sale mejor. A Andrés, por tu humor irónico que siempre me ha hecho reír. Gracias por tu

ayuda y tus consejos. GRACIAS a Estrella, por ser mi mejor mitad. Una “mijilla” de mí se fue contigo a Granada (o a Málaga), y siempre voy a echarte de menos, a ti y a tus abrazos. Sé que allí donde vayas todo te va a ir bien. Todos necesitamos una Estrella en nuestras vidas, y yo tuve esa suerte. Juntas los experimentos siempre fueron más amenos, aunque no siempre saliesen como queríamos… A Sergio, por compartir conmigo esta experiencia de principio a fin. Gracias por derrochar simpatía y por estar siempre a mi lado (literalmente).

Gracias también a Marijo, por saber escuchar, por tu creatividad y tu paciencia. A Mammar, por tu buen humor, venir siempre con una sonrisa en la cara y tu buena disposición para aprender. Da gusto trabajar contigo. Gaby, la “tica”; por enseñarme que los experimentos, cuando se resisten, es porque la naturaleza no quiere revelar sus más profundos secretos. Ahora en Marsella estoy segura de que será más fácil irnos de pintxos alguna tarde…

A todos los integrantes del Departamento de Histología, por cederme con cariño un sitio luminoso y tranquilo donde escribir la tesis.

A mis amigas, Oihane, Raquel, Leyre, Mayte, Cris y Raquel, por estar SIEMPRE y en todos los momentos a mi lado. Por los viernes desestresantes, los viajes, y por todo lo que nos queda por vivir juntas. Por conseguir restar importancia a los problemas y que las alegrías se multipliquen cuando estoy con vosotras.

A la Pampis, mejor ejemplo de hermana. Gracias por animarme desde el principio, por compartir alegrías y penas; por los sueños que aún nos quedan... A Rosa, gracias por tu alegría, por ser tan “como yo” y María, por tu risa, por nuestros planes, “asante”. Gracias a las dos por formar parte de mis primeros recuerdos.

A las biólogas, Maite, Mirian, Teresa, Bea, Ruth y Raquel, gracias por los maravillosos años de carrera. Sin vosotras todo hubiera sido mucho más difícil. Nunca pensé que contar palomas a -5ºC pudiese ser tan divertido. En especial a Miren, gracias por aparecer y por quedarte. Por tu sentido del humor y tus ganas de aprender. Aunque siempre te lo digo… de mayor quiero ser como tú.

A mis aitas, porque sin vosotros nada de esto hubiera sido posible. Gracias por estar siempre conmigo, por creer y confiar en mí; por animarme y apoyarme en todas mis decisiones. A Joseba, por cuidarme, entender y escuchar con paciencia; por ayudarme a llevar con buen humor los malos momentos. Por sacar lo mejor de mí y hacerme el camino más fácil. Por el pasado, el presente y el futuro. Al resto de mi familia, por el cariño y vuestro apoyo en todo momento.

Y por último, quiero agradecer a la Asociación de Amigos de la Universidad de Navarra la beca otorgada para la realización de esta tesis.

CONTENTS

Abstract … … … … … … … … … … … … … … … … … … … … … … … … … 1 Resumen … … … … … … … … … … … … … … … … … … … … … … … … … 3

General introduction … … … … … … … … … … … … … … … … … … … … … 7 Brucella and brucellosis … … … … … … … … … … … … … … … … … … … 9

Diagnosis and treatment … … … … … … … … … … … … … … … … … … … 12

Vaccination … … … … … … … … … … … … … … … … … … … … … … … … 13

Brucella virulence … … … … … … … … … … … … … … … … … … … … … 14

Brucella outer membrane and innate immunity … … … … … … … … … … … 15

Brucella lipopolysaccaride … … … … … … … … … … … … … … … … … … 17

References … … … … … … … … … … … … … … … … … … … … … … … … 21

Objectives and approach … … … … … … … … … … … … … … … … … … … … 31

Capítulo 1. Identificación de dos genes potencialmente implicados en la síntesis de diaminoglucosa, un componente esencial del lípido A de Brucella … … … … … … … … … … … … … … … … … … … … … 37

Abreviaturas … … … … … … … … … … … … … … … … … … … … … … … … 39

Resumen … … … … … … … … … … … … … … … … … … … … … … … … … 41

Introducción … … … … … … … … … … … … … … … … … … … … … … … … 43

Material y Métodos … … … … … … … … … … … … … … … … … … … … … 49

Resultados … … … … … … … … … … … … … … … … … … … … … … … … 57

Discusión … … … … … … … … … … … … … … … … … … … … … … … … 71

Referencias … … … … … … … … … … … … … … … … … … … … … … … … 75

Material suplementario … … … … … … … … … … … … … … … … … … … 81

Chapter 2. WadD, a new Brucella lipopolysaccharide core glycosyltranfserase identified by genomic search and phenotypic characterization … … … … … … … … … … … … … … … … … … 91

Abbreviations … … … … … … … … … … … … … … … … … … … … … … … 93

Abstract … … … … … … … … … … … … … … … … … … … … … … … … … 97

Introduction … … … … … … … … … … … … … … … … … … … … … … … … 99

Experimental procedures … … … … … … … … … … … … … … … … … … … 105

Results … … … … … … … … … … … … … … … … … … … … … … … … … 123

Discussion and Future directions … … … … … … … … … … … … … … … … 143

References … … … … … … … … … … … … … … … … … … … … … … … … 149

Supplemental material … … … … … … … … … … … … … … … … … … … 157

Annex … … … … … … … … … … … … … … … … … … … … … … … … … 197

Chapter 3. Towards a new vaccine against Brucella ovis infection … … … … 205

Abbreviations … … … … … … … … … … … … … … … … … … … … … … … 207

Abstract … … … … … … … … … … … … … … … … … … … … … … … … … 209

Introduction … … … … … … … … … … … … … … … … … … … … … … … 211

Experimental procedures … … … … … … … … … … … … … … … … … … 217

Results and discussion … … … … … … … … … … … … … … … … … … … 225

References … … … … … … … … … … … … … … … … … … … … … … … … 233

Supplemental material … … … … … … … … … … … … … … … … … … … 239

General discussion … … … … … … … … … … … … … … … … … … … … … … 249

Conclusions … … … … … … … … … … … … … … … … … … … … … … … … 261

Appendix … … … … … … … … … … … … … … … … … … … … … … … … … 267

Abstract

ABSTRACT

Brucellosis is a zoonotic disease caused by Brucella. Its lipopolysaccharide (LPS) is a modified Pathogen-Associated Molecular Pattern (PAMP) that plays a major role in virulence since impairs normal recognition by the innate immune system, and delays the Th1-mediated immune response, allowing the bacteria to reach a safe replicative niche. The core and lipid A LPS sections play a crucial role in this strategy.

In contrast to most gram-negative bacteria, Brucella lipid A presents a diaminoglucose disaccharide backbone, but its biosynthetic pathway remains unknown. We have identified in its genome the orthologues of gnnA and gnnB, responsible for the synthesis of diaminoglucose in other bacteria. Following a classical protocol, proven to be successful for many years, we were unable to construct mutants in any of these genes and we concluded that they are probably essential for outer membrane stability and Brucella viability. This was confirmed in parallel by a group of collaborators using a Tn-seq that enables straightforward saturating transposon mutagenesis of Brucella and the identification of genes that strongly contribute to the bacterium fitness. Although we were able to purify the proteins encoded by both genes, the assays to prove their enzymatic activity were not conclusive.

The core region of Brucella LPS also contributes to its ability to escape from innate immune system and is crucial for virulence. Mutants in glycosyltransferases involved in the synthesis of the core lateral branch not linked to the O-polysaccharide (O-PS) are attenuated, induce a significantly stronger immune response, and are good vaccine candidates against brucellosis. The chemical structure of the Brucella LPS core, recently elucidated, suggests that, in addition to the already identified WadB and WadC, other glycosyltransferases could be implicated in its synthesis. To clarify the genetics of core synthesis is thus crucial for the development of Brucella vaccines. In this work, we analysed B. abortus genome, a species that presents smooth LPS, to find new genes encoding putative glycosyltransferases involved in LPS synthesis. We constructed mutants in a total of 12 identified genes and analysed their LPS structure. Among them, 11 were not implicated in the synthesis of a complete LPS. Moreover, a

1

Abstract mucR mutant in B. abortus also presented a LPS similar to that of the parental strain. This is in accordance with the fact that the mucR-reguladed hypothetical glycosyltransferases were neither essential for the synthesis of a complete LPS, nor for interaction with elements of innate immunity or virulence in Brucella. Interestingly, mutant in BAB1_0953 (renamed wadD) lost reactivity against the antibodies that recognize the core section, but kept the O-PS. This suggests that WadD is a new glycosyltransferase adding one or more sugars to the core ramification of Brucella LPS that is not linked to the O-PS. wadD mutants were more sensitive than the parental strain to components of the innate immune system and in vivo studies suggest that WadD plays a role in chronic stages of infection.

Since mutants in genes involved in the synthesis of the core lateral branch are attenuated, induce a stronger immune response and protect against brucellosis, modification of these lateral branch has been proposed as a new strategy for the development of brucellosis vaccines. In this work, we have applied this strategy to modify a B. ovis genetically engineered strain able to grow in atmospheric conditions for the development of a B. ovis specific vaccine.

We have also clarified an open question and demonstrated that Open Reading Frame (ORF) BMEI0999 // BAB1_0998, situated immediately upstream the O-PS genes wboA and wboB is not required for the synthesis of a smooth LPS in B. melitensis or B. abortus.

2

Resumen

RESUMEN

La brucelosis es una enfermedad zoonótica causada por la especie Brucella. Su lipopolisacárido (LPS) es un PAMP (Patrón Molecular Asociado a Patógenos) modificado que juega un importante papel en la virulencia, ya que impide el reconocimiento por parte del sistema inmune innato y retrasa la respuesta mediada por Th1, permitiendo a la bacteria alcanzar su nicho replicativo. En esta estrategia, el núcleo y el lípido A del LPS juegan un papel crucial.

Al contrario de lo que sucede en la mayoría de las bacterias gram-negativas, el lípido A de Brucella presenta un esqueleto disacarídico de diaminoglucosa en lugar de glucosamina, pero la ruta implicada en su biosíntesis se desconoce. En este trabajo hemos identificado ortólogos de gnnA y gnnB en el genoma de Brucella, los genes responsables de la síntesis de diaminoglucosa en otras bacterias. Siguiendo el protocolo clásico de mutagénesis, previamente utilizado con éxito, no fue posible construir mutantes en ninguna de estas dos ORF (del inglés, Marco Abierto de Lectura) y concluimos que estos genes son probablemente esenciales para la estabilidad de la membrana externa y la viabilidad de Brucella. Esto fue confirmado en paralelo por un grupo de colaboradores mediante un estudio de mutagénesis por saturación con un transposón, en el cual han identificado los genes esenciales para Brucella. Pese a que las proteínas codificadas por ambos genes fueron purificadas con éxito, los ensayos para probar su actividad enzimática no fueron concluyentes.

La estructura del núcleo del LPS de Brucella también contribuye a su habilidad para escapar del sistema inmune innato y es crucial para la virulencia. Mutantes en glicosiltransferasas implicadas en la síntesis de la rama lateral del núcleo están atenuados, inducen una mayor respuesta inmune y son buenos candidatos vacunales frente a la brucelosis. La estructura química del núcleo del LPS de Brucella, recientemente elucidada, sugiere que, además de las ya identificadas WadB y WadC, otras glicosiltransferasas podrían estar implicadas en su síntesis. Por tanto, clarificar la genética de la síntesis del núcleo es crucial para el desarrollo de vacunas frente a la brucelosis. En el presente trabajo analizamos el genoma de B. abortus en busca de genes que codifican hipotéticas glicosiltransferasas implicadas en la síntesis del LPS.

3

Resumen

Construimos mutantes en un total de 12 ORF, y analizamos la estructura de sus LPS. Once de ellos no fueron necesarios para la síntesis de un LPS completo. De acuerdo con nuestros resultados, un mutante mucR en B. abortus también presentó un LPS similar al de la cepa parental. Esto concuerda con el hecho de que las hipotéticas glicosiltransferasas reguladas por mucR tampoco fuesen esenciales para la síntesis de un LPS completo ni para su interacción con elementos del sistema inmune innato o la virulencia de Brucella. Curiosamente, un mutante en BAB1_0953 (renombrada wadD) perdió la reactividad frente a anticuerpos que reconocen el núcleo, pero mantuvo la cadena O. Esto sugiere que WadD es una nueva glicosiltransferasa que añade uno o más azúcares a la ramificación del núcleo del LPS de Brucella que no está anclada a la cadena O. Mutantes en wadD fueron más sensibles que la cepa parental a componentes del sistema inmune innato y estudios in vivo sugieren que WadD juega un papel en estadios tardíos de la infección.

Dado que mutantes en genes de síntesis de la rama lateral del núcleo están atenuados, inducen una mayor respuesta inmune y protegen frente a la infección por Brucella, la modificación de esta rama lateral se ha propuesto como una nueva estrategia para el desarrollo de vacunas frente a la brucelosis. En este trabajo hemos aplicado esta estrategia para modificar una cepa de B. ovis capaz de crecer en condiciones atmosféricas para el desarrollo de una vacuna específica frente a la enfermedad causada por esta especie.

También hemos clarificado una cuestión abierta y demostrado que las ORF BMEI0999 // BAB1_0998, situadas inmediatamente corriente arriba de dos genes de síntesis de la cadena O (wboA y wboB), no son esenciales para la síntesis de un LPS liso en B. melitensis ni B. abortus.

4

General Introduction

General introduction

Abbreviations used in Geneal Introduction: BCV Brucella Containing Vacuole; CD14 Cluster of Differentiation 14; FAO Food and Agriculture Organization (of the United Nations); IL Interleukin; ILRI International Livestock Research Institute; Kdo 3-deoxy-D-manno-2-octulosonic acid; LPS Lipopolysaccharide; OIE World Organisation for Animal Health; OL Ornithine Lipids; O-PS O- polysaccharide; PAMP Pathogen-Associated Molecular Pattern; RNA Ribonucleic Acid; spp. species; TLR Toll-Like Receptor; TNF-α Tumour Necrosis Factor-alpha; WHO World Health Organization.

7

General introduction

Brucella and brucellosis

Brucella is a highly infectious gram-negative bacterium responsible for brucellosis. The first Brucella spp. was isolated by David Bruce in 1887 at the time of his service in Malta. The analysis of the gene of the 16S ribosomal RNA, some conserved genes, the protein profile, the immunological cross reactivity and the cell envelope components, place Brucella in the α-2 subdivision of the class Proteobacteria (Figure I.1). The α-2 subdivision includes several microorganisms living in close association with animal (e.g., Rickettsia, Bartonella, Ochrobactrum) or plant (e.g. Agrobacterium, Rhizobium) eukaryotic cells (Barquero-Calvo et al., 2007; Moreno et al., 1990; Rasool et al., 1992).

Figure I.1. Relationship (based on the 16S rRNA sequences) between Brucella and other relevant bacteria from the α-2 Proteobacteria class.

The genus Brucella comprises several nominal species (http://www.bacterio.net/- allnamesac.html). They form a core group that includes the so-named “classical” species (B. abortus, B. melitensis, B. suis, B. canis, B. ovis and B. neotomae) as well as the “non-classical” B. ceti, B. pinnipedialis, B. microti, B. papionis and B. vulpis, more recent species isolated from different mammals. The spp. in this core group are separated from several early-diverging brucellae including B. inopinata strain BO1 and B. inopinata-like strain BO2, a novel strain isolated from rodents in Australia, and several strains isolated from amphibians (Al Dahouk et al., 2017; Scholz et al., 2010;

9

General introduction

Soler-Lloréns et al., 2016; Tiller et al., 2010). Brucellae are classified according to host preferences and some of the species are organized in biotypes or biovars conforming to colorant sensitivity or serological type (OIE, 2016): B. melitensis affects goats and sheep, B. abortus infects cattle and B. suis domestic and feral swine and wild-boar (biovars 1, 2 and 3), reindeer (biovar 4), hares (biovar 2) and several rodent species (biovar 5) (Alton and Forsyth, 1996; Kulakov et al., 2010). B. ovis infects sheep, B. canis dogs, B. neotomae rodents and B. pinnipedialis and B. ceti pinnipeds and cetaceans respectively (Foster et al., 2007). B. microti has been isolated from small rodents, red foxes and directly from soil (Hubálek et al., 2007; Scholz et al., 2008). B. inopinata BO1 was isolated from a breast implant infection (Scholz et al., 2010). Most recently, B. papionis (isolated from baboons) and B. vulpis (from red foxes), were described (Hofer et al., 2016; Whatmore et al., 2014). However, not all Brucella species are strictly host- specific and some species can cross barriers naturally (Bricker, 2004; Moreno, 2014).

Brucellosis is a chronic disease, recognized as a significant public health challenge, with major economic and financial burdens in countries where it remains endemic (Corbel, 1997). It is a zoonosis; a transmissible disease between animals (domestic and wildlife) and humans. Many zoonoses are endemic in developing countries and common in poor and marginalized populations, reflecting the strong association among them, family poverty and livestock keeping. The Food and Agriculture Organization of the United Nations (FAO) estimates that more than 600 million people worldwide depend on livestock and represent up to 70% of the population in the most marginal areas. In a recent publication, the International Livestock Research Institute (ILRI, 2012) places Brucella as one of the top zoonotic diseases. Brucellosis is spread all around the world mainly because infected animals are widely used for meat, milk, hair, wool or for carrying burdens. Moreover, they are generally in parts of the world where animal and/or human health services are scarce or non-existent (Corbel et al., 2006; Moreno, 2014; Moreno and Moriyón, 2002). Although brucellosis has been eradicated in Europe and North America, is still an important problem especially in the Mediterranean countries of Europe, north and east of Africa, the Middle East, south and central Asia and Central and South America, although it frequently goes unreported (Ariza, 1999; Corbel et al., 2006). Studies that try to define the incidence of

10

General introduction animal brucellosis worldwide suggest that the number of domestic animals susceptible to the disease is estimated in 4,000,000,000 (Moreno, 2014; Moreno and Moriyón, 2006).

Animal brucellosis is a herd or flock problem as it is spread primarily by ingestion of contaminated material. There are different livestock production systems worldwide, from the landless (total confinement) intensive systems of dairy cattle, to the extensive husbandry of mixed species grazing with very low animal concentrations per unit area. Hence, the type of system will affect the spreading of brucellosis infection within and between herds and flocks. Infection is usually chronic in animals that shed the organisms in uterine discharges following abortion and subsequent parturition, and in the colostrum and milk (Blasco et al., 2016; Moreno and Moriyón, 2006). In sexually mature animals, infection by Brucella localizes in the reproductive organs and it is often followed by abortion in the female pregnant animals and epididymitis and orchitis in the male and subsequent delayed or permanent infertility. Infection may lead to several complications: stillborn, swollen testicles, arthritis or local abscess (Blasco et al., 2016). Cattle infection by B. abortus causes hygromas, abscesses and lowered milk production due to premature births. Infections are highly contagious because of close contact between animals in the same flock. Animal-to-animal transmission occurs as a result of the large number of organisms shed in the environment (OIE, 2016). Venereal infections can also occur, but this is mainly seen with B. suis infections. In most livestock, parturition is seasonal, and knowledge of these patterns is important in determining probable times for the occurrence of abortions.

Human brucellosis is considered an important neglected zoonosis distributed worldwide. The most frequent means of acquisition of brucellosis are ingestion of unpasteurized milk products from infected animals and direct contact with abortion materials or animal secretions. Brucella spp. can survive in solid surfaces, soil, dust, farm slurry animal waste, aborted fetuses, meat, dairy products and water sources may also be contaminated by recently aborted animals (Ariza, 1999; Corbel, 2006). Occupational exposure to brucellosis includes farmers, animal attendants, shepherds,

11

General introduction veterinarians, inseminators or laboratory staff. Infection may occur by inhalation, conjunctival contamination, accidental ingestion, skin contact and accidental self- inoculation with live vaccines (as B. abortus S19 and B. melitensis Rev1 are not entirely avirulent for humans, see below). Person-to-person transmission of brucellosis is extremely rare and laboratory workers processing contaminated samples run a high risk of infection. In general, B. melitensis produces the most severe disease but B. abortus and B. suis may also cause human infection. The World Health Organization (WHO) defines human brucellosis as “an illness characterized by acute or insidious onset, continued, intermittent or irregular fever of variable duration, profuse sweating, particularly at night, fatigue, anorexia, weight loss, headache, arthralgia, and generalized aching” (OIE, 2016). It is a systemic infection that can involve any organ or tissue. Infection may persist and progress to a chronically incapacitating and debilitating disease with severe complications such as lymphadenopathy (found in 15% of the cases), hepatosplenomegaly (30-50%) or osteoarticular problems including bursitis, tenosynovitis, sacroiliitis and spondylitis, which is the major osteoarticular focal disease, causing severe pain, functional limitation and significant sequelae. Orchitis and epididymitis are the main genitourinary complications in men and neurobrucellosis is rarely reported (Ariza, 1999; Corbel, 1997; Corbel et al., 2006; Moreno and Moriyón, 2006). The unspecific symptoms of human disease make it difficult to report the real number of human brucellosis but it has been estimated that there are half a million new cases every year, most of them located in the poorest rural areas of the world (Dean et al., 2012; Ducrotoy et al., 2016). A review of the studies available concludes that brucellosis is in all likelihood a major problem in low-income countries of Africa and Asia (McDermott et al., 2013, Dean et al., 2012). Moreover, in those areas, millions of people are at risk as animal infection is not controlled and heat treatment for milk is not routinely applied, hence transmission to humans occurs frequently.

Diagnosis and treatment

Diagnosis of brucellosis in animal and humans is based on the isolation of the microorganisms and serological tests (Ariza, 1999; Corbel et al., 2006; Díaz et al., 2011;

12

General introduction

Ducrotoy et al., 2016; Moreno and Moriyón, 2006). Treatment of human brucellosis requires up to six weeks of combined antibiotics. If untreated, it can be lethal in 1–5% cases (Ariza, 1999; Ariza et al., 2007; Solera, 2010).

Vaccination

Animal vaccination is critical to control brucellosis since animals are the main reservoir of the disease. Moreover, it is also the best way to eradicate human brucellosis since there is no safe human vaccine. In developing countries and endemic areas, mass vaccination is the only way to control the disease (Blasco, 1997; Blasco et al., 2016). There are three available live attenuated vaccines against brucellosis: B. melitensis Rev1 is the most extensively used vaccine for the prevention of brucellosis in sheep and goats. It is administered to lambs between 3 and 5 months as a single subcutaneous dose. Rev1 often causes abortion and excretion in milk in pregnant animals and implies a high risk of causing human infection. Vaccination with Rev1 should be performed when the animals are not pregnant or during the lambing season (Blasco, 1997). Indeed, Rev1 is the best available vaccine against B. melitensis and B. ovis infection (OIE, 2016). However, in areas free of B. melitensis infection, vaccination with Rev1 is banned, thus becoming B. ovis infection an emerging problem (OIE, 2016). All these issues will be discussed in more detail in Chapter 3. B. abortus strain 19 vaccine (S19) is widely used for the prevention of cattle brucellosis. It is administered subcutaneously to female calves between 3 and 6 months. It can also be administered to adult cattle with the risk that some animals may abort and excrete the vaccine strain in the milk. Finally, B. abortus strain RB51 has become the official vaccine for prevention of brucellosis in cattle in several countries. However, it induces severe placentitis and placental infection in vaccinated cattle, and it has been reported that there is excretion in milk (Moriyón et al., 2004). The protection conferred by B. abortus RB51 against B. melitensis infection in cattle is unknown. In summary, any of these vaccines provide complete protection, they are virulent for humans and interfere with the serological diagnosis of brucellosis since vaccination may lead to long-term persistence of antibodies. Moreover, Rev1 is resistant to streptomycin, an antibiotic of choice to treat brucellosis (Elberg and Faunce, 1956). To solve the serological

13

General introduction interference problem, vaccines must be applied by the conjunctival route early in life (Blasco et al., 2016; Moriyón et al., 2004; OIE, 2016).

Besides, no specific vaccines are available against B. suis and B. ovis. We have addressed B. ovis vaccination in more detail in Chapter 3.

Brucella virulence

Brucella lacks the classical bacterial pathogenicity factors (Moreno and Moriyón, 2002) and the principal virulence determinants are those that allow Brucella to resist the action of the innate immune response, invade the target cell, resist intracellular killing, reach and reorganize a replicative niche and adapt their metabolism to the available nutrients (Barbier et al., 2018; Barquero-Calvo et al., 2007; Celli, 2015; Gorvel and Moreno, 2002; Guzmán-Verri et al., 2001; Ronneau et al., 2014; Zúñiga-Ripa et al., 2014). Brucella virulence is tightly regulated and among the most significant control systems, BvrR/BvrS regulates the components of the outer membrane or the metabolism (Sola-Landa et al., 1998; Viadas et al., 2010), VjbR modulates the expression of genes encoding functions consistent with an intracellular adaptation strategy for survival during the initial stages of the host cell infection and cell surface structures (Delrue et al., 2005; Kleinman et al., 2017) and MucR controls the synthesis of the lipopolysaccharide (LPS), motility, quorum sensing, iron homeostasis, genome plasticity and cell envelope integrity (Caswell et al., 2013; Mirabella et al., 2013). The role of MucR and BvrR/BvrS in regulation of Brucella LPS synthesis will be discussed in more detail in Chapter 2.

Brucella does not have most of the classical Pathogen-Associated Molecular Pattern (PAMPs) present in gram-negative bacteria: does not produce functional flagella, toxins, pili or capsule, and some of the classical components of the outer membrane such as lipids or LPS, although present, are structurally modified PAMPs (see below).

Thus, Brucella is considered a silent parasite that is not well-recognized by the elements of the innate immune system, escaping prompt detection during initial stages of infection (Barquero-Calvo et al., 2007). This allows the bacteria to reach a

14

General introduction safe intracellular niche (Lamontagne et al., 2007; Moreno and Moriyón, 2002; Moriyón and Berman, 1982; Moriyón and López-Goñi, 1998).

Brucella then acts as a facultative intracellular parasite and penetrates the target cells where follows a sophisticated intracellular trafficking that ensures their survival, proliferation and persistence within the host. It controls the conversion of the Brucella- containing vacuole (BCV) from an endosomal (eBCV) to a replicative (rBCV) vacuole derived from the endoplasmic reticulum, and finally the biogenesis of an autophagic (aBCV) vacuole (Starr et al., 2008; Celli 2015), where it multiplies and inhibits apoptosis without affecting host cell division (Chaves-Olarte et al. 2002; Gross et al. 2000). Intracellular trafficking depends on type IV translocated Brucella´s effector proteins to the host cell that enable its arrival to the replicative niche (Boschiroli et al., 2002; Delrue et al., 2001; O’Callaghan et al., 1999; Sieira et al., 2000). Moreover, Brucella injects active effectors such as BtpA and BtpB through a Type IV secretion system. Both effector proteins contain a TIR domain that interferes with TLR signaling by directly interacting with MyD88 (Chaudhary et al., 2012; Cirl et al., 2008; Salcedo et al., 2008, 2013) and contribute to the control of dendritic cell (DC) activation during infection.

Brucella also avoids fusion with lysosomes by producing cyclic glucans that interact with intravacuolar lipid rafts (Arellano-Reynoso et al., 2005). Subsequently, bacteria infect myeloid cells, including macrophages in the spleen and liver, and persist within granulomatous lesions, or infect and proliferate within placental trophoblasts in pregnant animals (Atluri et al. 2011).

Brucella outer membrane and innate immunity

Brucella outer membrane plays an important role in virulence and acts as a protective barrier against large molecules and hydrophobic compounds (Figure I.2). Apart from lipoproteins, it contains characteristic lipids and a non-classical LPS, that in Brucella has specific modifications (see below and Chapters 1 and 2) (Lamontagne et al., 2007; Moreno and Moriyón, 2002; Moriyón and Berman, 1982; Moriyón and López-Goñi, 1998).

15

General introduction

Figure I.2. Schematic representation of the Brucella cell envelope. It shows the inner and outer bilayers separated by the periplasm, which contains peptidoglycan. The outer leaflet of the outer membrane contains the LPS, which is anchored to the membrane by lipid A. The inner leaflet of the outer membrane and also the entire inner membrane are based on lipids (PC: phosphatidylcholine, PE: phosphatidylethanolamine, PG: phosphatidylglycerol, CL: cardiolipin and OL: ornithine lipids).

Brucella outer membrane presents the free-lipids existing in most gram-negative bacteria (i.e., cardiolipin, phosphatidylglycerol and phosphatidylethanolamine) and in addition phosphatidylcholine and amino lipids (Moreno and Moriyón, 2002; Moriyón et al., 1987; Thiele and Schwinn, 1973). Phosphatidylcholine is an eukaryotic-type phospholipid required for Brucella virulence (Comerci et al., 2006; Conde-Álvarez et al., 2006). Among the amino lipids, ornithine lipids (OL) are the only ones carefully analysed in Brucella and, unlike their counterpart in Bordetella, do not trigger the release of IL-6 or TNF-α and seem to be dispensable for Brucella virulence (Palacios- Chaves et al., 2011). It has been reported that in some bacteria, OL can be hydroxylated and this is related to their biological activity (Geiger et al., 2010). Only B. microti and B. vulpis have a complete olsC, coding a functional acyl-chain hydroxylase and, indeed, hydroxylated ornithine lipids have been detected in B. microti (Conde- Álvarez et al., 2018; Palacios-Chaves et al., 2011; Vences-Guzmán et al., 2011) (see Chapter 1 and Appendix). The lack of a functional olsC in B. abortus is in keeping with the hypothesis that B. abortus OL represent dispensable ancestral structures that could have been eliminated during evolution and adaptation to the host (Palacios- Chaves et al., 2011).

16

General introduction

Brucella lipopolysaccharide

The LPS of most gram-negative bacteria (also known as endotoxin) is the major component in the outer membrane and is essential for establishing a barrier only permeable to low molecular weight and/or hydrophilic molecules (Nikaido, 2003). The LPS is composed of a lipidic portion (lipid A), responsible for toxicity, bound to a polysaccharide fraction that includes the core section and the O-polysaccharide (O-PS) and constitutes a PAMP. The lipid A of typical gram-negative bacteria such Escherichia coli or Salmonella is a glucosamine disaccharide phosphorylated at C1´ and C4’, carrying four ester- and amide-linked 3-hydroxy-myristoil groups, two of which bear lauric and myristic acyl chains in acyloxyacyl linkages. Fatty acids commonly found in lipid A include caproic (C6), lauric (C12), myristic (C14), palmitic (C16) and stearic (C18) acids. The core oligosaccharide is typically rich in acid sugars and phosphate. In Salmonella species, where LPS has been well studied, the core polysaccharide consists of 2-keto 3-deoxyoctulosonic (Kdo), various seven-carbon sugars, glucose, galactose and N-acetylglucosamine. The core oligosaccharide is attached to the lipid A through Kdo and, along with the phosphates in the latter, they create a densely charged area that is essential to maintain the permeability barrier function of the outer membrane (Nikaido, 2003; Raetz et al., 2007; Raetz and Whitfield, 2002). Connected to the core is the O-PS, which comprises galactose, glucose, rhamnose and mannose as well as one or more dideoxy-hexoses, connected in four- or five-membered sequences, which form a long chain when repeated. The O-PS plays an important role in virulence in gram-negative bacteria.

The LPS is recognized by the receptor complex TLR4-MD2-CD14. The lipid A phosphate groups interact with positively charged residues in a hydrophobic groove in MD2, into which the acyl chains fit (Gruber et al., 2004; Ohto et al., 2007). The inner core also contributes to the interaction with the TLR4-MD2 complex. Subsequent signaling to the nucleus of the host cell triggers potent proinflammatory response. The negatively charged complex formed by Kdo residues and phosphates is also in part responsible for the recognition of the inner core by polycationic bactericidal peptides and the C1q complement, a component of innate immunity.

17

General introduction

Brucella LPS differs from that of other gram-negative bacteria; it is considered a modified PAMP and an important virulence factor (Lapaque et al., 2005). The O-PS of smooth type Brucella LPS (S-LPS) is an unbranched homopolymer of α(1-2) or α(1-3) N- formyl-perosamine (N-formylated 4-amino, 4,6-dideoxyglucose) usually with an average chain length of 96 to 100 glycosyl subunits (Bundle et al., 1989). The relative frequency of α (1-2) or α (1-3) linkages produces three epitope structures of O-PS: biotype 1 B. abortus carries A epitope [at least five blocks of N-formyl-perosamine α (1-2) linked], biotype 1 B. melitensis carries M epitope [four α (1-2) linked and one α (1-3) linked] and both carry C epitope [at least four α (1-2) linked] (Díaz et al., 1968). This O-PS plays a relevant role in virulence, since it has been repeatedly reported that mutants lacking this part of the LPS (i.e. producing a rough (R) type-LPS) are attenuated (González et al., 2008; Mancilla et al., 2012; Monreal et al., 2003). In addition, the O-PS characteristic of smooth brucellae like B. abortus, B. melitensis or B. suis confers serum and complement resistance, a property not uncommon in the O-PS of gram-negative pathogens, and also modulates the entry into cells (Gorvel and Moreno, 2002).

The low endotoxicity of Brucella LPS and its poor detection by the innate immune system are related to main structural differences located in the lipid A and the core sections (Conde-Álvarez et al., 2012; Lapaque et al., 2006b). The lipid A is composed of a disaccharide of diaminoglucose and the primary 3-hydroxylated acyl chains (all amide-linked) are substituted in acyloxyacyl by very long chain fatty acids (up to 28 or 30 carbons in length) (as will be described in more detail in Chapter 1). The core section is a branched structure and lacks acidic sugars, heptoses or phosphate groups, and the number of negative charges both in the core and lipid A is very low when compared to the classical LPS of other gram-negative bacteria (Conde-Álvarez et al., 2012; Fontana et al., 2016; Iriarte et al., 2004; Moriyón et al., 2004) (this LPS section will be discussed in more detail in Chapter 2).

These differences are important for Brucella virulence since its LPS is poorly recognized by the CD14-MD2-TLR4 receptor complex and this results in the production of a significantly lower amount of cytokines and a very low stimulation of cells that participate in the innate immune response, including polymorphonuclear leukocytes

18

General introduction and antigen presenting cells (Conde-Álvarez et al., 2012; Rasool et al., 1992; Zhao et al., 2017). It is also involved in a deficient activation of the complement cascade and in resistance to polycationic peptides (Freer et al., 1995; Martínez de Tejada et al., 1995; Moreno et al., 1981). Moreover, the LPS is also required to reach the replicative niche (Gorvel and Moreno, 2002), is resistant to degradation by murine macrophages and remains in their membrane forming aggregates with cellular receptors (Forestier et al., 1999; Lapaque et al., 2006a).

All the available information highlights the relevance of LPS in Brucella virulence and modulation of the immune response. Since the genes and genetic pathways involved in its synthesis have not been completely elucidated, the main objective of this work has been to clarify some aspects that are still unknown. We have tried to identify and characterize the genes involved in the synthesis of the particular disaccharide that forms Brucella lipid A (Chapter 1), we have performed a new search for ORF coding hypothetical glycosyltransferases and studied their role in the synthesis of the LPS (Chapter 2). Interestingly, the fact that genetic modifications in the core section of Brucella LPS result in a different modulation of the immune response, opens new doors to design of new vaccines against brucellosis (Conde-Álvarez et al., 2013; Zhao et al., 2017). Thus, we have applied this knowledge to the synthesis of new vaccine candidates against B. ovis (Chapter 3). Furthermore, we have also investigated an ORF situated immediately upstream wboA-wboB, two genes required for the synthesis of a S-LPS. Previous work raised questions about a putative role of this ORF in O-PS synthesis (Annex in Chapter 2).

19

General introduction References

REFERENCES

Al Dahouk, S., Köhler, S., Occhialini, A., Jiménez de Bagüés, M. P., Hammerl, J. A., Eisenberg, T., et al. (2017). Brucella spp. of amphibians comprise genomically diverse motile strains competent for replication in macrophages and survival in mammalian hosts. Sci. Rep. 7, 44420. doi:10.1038/srep44420. Alton, G. G., and Forsyth, J. R. L. (1996). Brucella. University of Texas Medical Branch at Galveston. Arellano-Reynoso, B., Lapaque, N., Salcedo, S., Briones, G., Ciocchini, A. E., Ugalde, R., et al. (2005). Cyclic beta-1,2-glucan is a Brucella virulence factor required for intracellular survival. Nat. Immunol. 6, 618–25. doi:10.1038/ni1202. Ariza, J. (1999). Brucellosis: an update. The perspective from the Mediterranean basin. Rev.Med.Microbiol. 10, 125–135. Ariza, J., Bosilkovski, M., Cascio, A., Colmenero, J. D., Corbel, M. J., Falagas, M. E., et al. (2007). Perspectives for the treatment of brucellosis in the 21st century: The Ioannina recommendations. PLoS Med. 4, 1872–1878. doi:10.1371/journal.pmed.0040317. Atluri, V. L., Xavier, M. N., De Jong, M. F., Den Hartigh, A. B., and Tsolis, R. M. (2011). Interactions of the human pathogenic Brucella species with their hosts. Annu. Rev. Microbiol 65, 523–41. doi:10.1146/annurev-micro-090110-102905. Barbier, T., Zúñiga-Ripa, A., Moussa, S., Plovier, H., Sternon, J. F., Lázaro-Antón, L., et al. (2018). Brucella central carbon metabolism: an update. Crit. Rev. Microbiol. 44, 182–211. doi:10.1080/1040841X.2017.1332002. Barquero-Calvo, E., Chaves-Olarte, E., Weiss, D. S., Guzmán-Verri, C., Chacón-Díaz, C., Rucavado, A., et al. (2007). Brucella abortus uses a stealthy strategy to avoid activation of the innate immune system during the onset of infection. PLoS One 2, e631. doi:10.1371/journal.pone.0000631. Blasco, J. M. (1997). A review of the use of B. melitensis Rev 1 vaccine in adult sheep and goats. Prev. Vet. Med. 31, 275–283. doi:10.1016/S0167-5877(96)01110-5. Blasco, J. M., Moreno, E., and Moriyón, I. (2016). “Brucellosis vaccines and vaccine candidates.,” in Veterinary Vaccines for Developing Countries., ed. A. Metwally, S and Viljoen, G J and El Idrissi (Rome: FAO), 1–66. Boschiroli, M. L., Ouahrani-Bettache, S., Foulongne, V., Michaux-Charachon, S., Bourg, G., Allardet-Servent, A., et al. (2002). Type IV secretion and Brucella virulence. Vet. Microbiol. 90, 341–348. doi:10.1016/S0378-1135(02)00219-5. Bricker, B. J. (2004). “Molecular diagnostics of animal Brucellosis: A review of PCR- based assays and approaches.” in Brucella. Molecular and Cellular Biology (Wymondham, England: Horizon Bioscience), 25–52. Bundle, D. R., Cherwonogrodzky, J. W., Gidney, M. A., Meikle, P. J., Perry, M. B., and Peters, T. (1989). Definition of Brucella A and M epitopes by monoclonal typing

21

General introduction References

reagents and synthetic oligosaccharides. Infect. Immun. 57, 2829–36. Caswell, C. C., Elhassanny, a. E. M., Planchin, E. E., Roux, C. M., Weeks-Gorospe, J. N., Ficht, T. a., et al. (2013). Diverse genetic regulon of the virulence-associated transcriptional regulator MucR in Brucella abortus 2308. Infect. Immun. 81, 1040– 1051. doi:10.1128/IAI.01097-12. Celli, J. (2015). The changing nature of the Brucella-containing vacuole. Cell. Microbiol. 17, 951–8. doi:10.1111/cmi.12452. Chaudhary, A., Ganguly, K., Cabantous, S., Waldo, G. S., Micheva-Viteva, S. N., Nag, K., et al. (2012). The Brucella TIR-like protein TcpB interacts with the death domain of MyD88. Biochem. Biophys. Res. Commun. 417, 299–304. doi:10.1016/j.bbrc.2011.11.104. Chaves-Olarte, E., Guzmán-Verri, C., Méresse, S., Desjardins, M., Pizarro-Cerdá, J., Badilla, J., et al. (2002). Activation of Rho and Rab GTPases dissociates Brucella abortus internalization from intracellular trafficking. Cell. Microbiol. 4, 663–676. doi:10.1046/j.1462-5822.2002.00221.x. Cirl, C., Wieser, A., Yadav, M., Duerr, S., Schubert, S., Fischer, H., et al. (2008). Subversion of Toll-like receptor signaling by a unique family of bacterial Toll/interleukin-1 receptor domain–containing proteins. Nat. Med. 14, 399–406. doi:10.1038/nm1734. Comerci, D. J., Altabe, S., de Mendoza, D., and Ugalde, R. A. (2006). Brucella abortus synthesizes phosphatidylcholine from choline provided by the host. J. Bacteriol. 188, 1929–34. doi:10.1128/JB.188.5.1929-1934.2006. Conde-Álvarez, R., Arce-Gorvel, V., Gil-Ramírez, Y., Iriarte, M., Grilló, M. J., Gorvel, J. P., et al. (2013). Lipopolysaccharide as a target for brucellosis vaccine design. Microb. Pathog. 58, 29–34. doi:10.1016/j.micpath.2012.11.011. Conde-Álvarez, R., Arce-Gorvel, V., Iriarte, M., Manček-Keber, M., Barquero-Calvo, E., Palacios-Chaves, L., et al. (2012). The lipopolysaccharide core of Brucella abortus acts as a shield against innate immunity recognition. PLoS Pathog. 8, e1002675. doi:10.1371/journal.ppat.1002675. Conde-Álvarez, R., Grilló, M. J., Salcedo, S. P., de Miguel, M. J., Fugier, E., Gorvel, J. P., et al. (2006). Synthesis of phosphatidylcholine, a typical eukaryotic phospholipid, is necessary for full virulence of the intracellular bacterial parasite Brucella abortus. Cell. Microbiol. 8, 1322–1335. doi:10.1111/j.1462-5822.2006.00712.x. Conde-Álvarez, R., Palacios-Chaves, L., Gil-Ramírez, Y., Salvador-Bescós, M., Bárcena- Varela, M., Aragón-Aranda, B., et al. (2018). Identification of lptA, lpxE, and lpxO, Three Genes Involved in the Remodeling of Brucella Cell Envelope. Front. Microbiol. 8, 2657. doi:10.3389/fmicb.2017.02657. Corbel, M. J. (1997). Brucellosis: An overview. Emerg. Infect. Dis. 3, 213–221. doi:10.3201/eid0302.970219. Corbel, M. J., Alton, G., Banai, M., Díaz, R., Dranovskaia, E., Elberg, S., et al. (2006). Brucellosis in humans and animals. Who, 1–102. doi:10.2105/AJPH.30.3.299.

22

General introduction References

Dean, A. S., Crump, L., Greter, H., Schelling, E., and Zinsstag, J. (2012). Global burden of human brucellosis: a systematic review of disease frequency. PLoS Negl. Trop. Dis. 6, e1865. doi:10.1371/journal.pntd.0001865. Delrue, R.-M., Deschamps, C., Leonard, S., Nijskens, C., Danese, I., Schaus, J.-M., et al. (2005). A quorum-sensing regulator controls expression of both the type IV secretion system and the flagellar apparatus of Brucella melitensis. Cell. Microbiol. 7, 1151–1161. doi:10.1111/j.1462-5822.2005.00543.x. Delrue, R.-M., Martínez-Lorenzo, M., Lestrate, P., Danese, I., Bielarz, V., Mertens, P., et al. (2001). Identification of Brucella spp. genes involved in intracellular trafficking. Cell. Microbiol. 3, 487–497. doi:10.1046/j.1462-5822.2001.00131.x. Díaz, R., Casanova, A., Ariza, J., Moriyón, I., and Vinetz, J. M. (2011). The rose bengal test in human Brucellosis: A neglected test for the diagnosis of a neglected disease. PLoS Negl Trop Dis 5. doi:10.1371/journal.pntd.0000950. Diaz, R., Jones, L. M., Leong, D., and Wilson, J. B. (1968). Surface antigens of smooth brucellae. J. Bacteriol. 96, 893–901. Ducrotoy, M. J., Conde-Álvarez, R., Blasco, J. M., and Moriyón, I. (2016). A review of the basis of the immunological diagnosis of ruminant brucellosis. Vet. Immunol. Immunopathol. 171, 81–102. doi:10.1016/j.vetimm.2016.02.002. Elberg, Sanford S, and Kenneth Faunce. (1956). Immunization against Brucella Infection Vi. Immunity conferred on goats by a nondependent mutant from a streptomycin- dependent mutant strain of Brucella melitensis. Journal of Bacteriology 73. Fontana, C., Conde-Álvarez, R., Ståhle, J., Holst, O., Iriarte, M., Zhao, Y., et al. (2016). Structural studies of lipopolysaccharide defective mutants from Brucella melitensis identify a core oligosaccharide critical in virulence. J. Biol. Chem. 291, 7727–7741. doi:10.1074/jbc.M115.701540. Forestier, C., Moreno, E., Méresse, S., Phalipon, A., Olive, D., Sansonetti, P., et al. (1999). Interaction of Brucella abortus lipopolysaccharide with major histocompatibility complex class II molecules in B lymphocytes. Infect. Immun. 67, 4048–54. Foster, G., Osterman, B. S., Godfroid, J., Jacques, I., and Cloeckaert, A. (2007). Brucella ceti sp. nov. and Brucella pinnipedialis sp. nov. for Brucella strains with cetaceans and seals as their preferred hosts. Int. J. Syst. Evol. Microbiol. 57, 2688–2693. doi:10.1099/ijs.0.65269-0. Freer, E., Rojas, N., Weintraub, A., Lindberg, A. A., and Moreno, E. (1995). Heterogeneity of Brucella abortus lipopolysaccharides. Res. MicrobioL Pads 146, 569–578. Geiger, O., González-Silva, N., López-Lara, I. M., and Sohlenkamp, C. (2010). Amino acid-containing membrane lipids in bacteria. Prog. Lipid Res. 49, 46–60. doi:10.1016/j.plipres.2009.08.002. González, D., Grilló, M.-J. M., De Miguel, M. M.-J., Ali, T., Arce-Gorvel, V., Delrue, R. R.- M., et al. (2008). Brucellosis vaccines: assessment of Brucella melitensis

23

General introduction References

lipopolysaccharide rough mutants defective in core and O-polysaccharide synthesis and export. PLoS One 3, e2760. doi:10.1371/journal.pone.0002760. Gorvel, J. P., and Moreno, E. (2002). Brucella intracellular life: from invasion to intracellular replication. Vet. Microbiol. 90, 281–97. Gross, A., Terraza, A., Ouahrani-Bettache, S., Liautard, J. P., and Dornand, J. (2000). In vitro Brucella suis infection prevents the programmed cell death of human monocytic cells. Infect. Immun. 68, 342–51. Gruber, A., Mancek, M., Wagner, H., Kirschning, C. J., and Jerala, R. (2004). Structural model of MD-2 and functional role of its basic amino acid clusters involved in cellular lipopolysaccharide recognition. J. Biol. Chem. 279, 28475–82. doi:10.1074/jbc.M400993200. Guzmán-Verri, C., Chaves-Olarte, E., von Eichel-Streiber, C., López-Goñi, I., Thelestam, M., Arvidson, S., et al. (2001). GTPases of the Rho subfamily are required for Brucella abortus internalization in nonprofessional phagocytes: direct activation of Cdc42. J. Biol. Chem. 276, 44435–43. doi:10.1074/jbc.M105606200. Hofer, E., Hammerl, J. A., Zygmunt, M. S., Cloeckaert, A., Koylass, M., Whatmore, A. M., et al. (2016). Brucella vulpis sp. nov., a novel Brucella species isolated from mandibular lymph nodes of red foxes (Vulpes vulpes) in Austria. Int. J. Syst. Evol. Microbiol. 66, 2090–2098. doi:10.1099/ijsem.0.000998. Hubálek, Z., Scholz, H. C., Sedlácek, I., Melzer, F., Sanogo, Y. O., and Nesvadbová, J. (2007). Brucellosis of the common vole (Microtus arvalis). Vector Borne Zoonotic Dis. 7, 679–87. doi:10.1089/vbz.2007.0143. Iriarte, M., González, D., Delrue, R.-M., Monreal, D., Conde, R., López-Goñi, I., et al. (2004). “Brucella lipopolysaccharide: Structure, biosynthesis and genetics,” in Brucella. Molecular and Cellular Biology (Wymondham, England: Horizon Bioscience), 159–192. Kleinman, C. L., Sycz, G., Bonomi, H. R., Rodríguez, R. M., Zorreguieta, A., and Sieira, R. (2017). ChIP-seq analysis of the LuxR-type regulator VjbR reveals novel insights into the Brucella virulence gene expression network. Nucleic Acids Res. 45, 5757– 5769. doi:10.1093/nar/gkx165. Kulakov, I. K., Erdenebaator, J., Tsirelson, L. E., Tolmacheva, T. A., Zheludkov, M. M., and Korenberg, E. I. (2010). Molecular genetics typing of Brucella circulating in several provinces of Mongolia. Zh. Mikrobiol. Epidemiol. Immunobiol., 17–22. Lamontagne, J., Butler, H., Chaves-Olarte, E., Hunter, J., Schirm, M., Paquet, C., et al. (2007). Extensive cell envelope modulation is associated with virulence in Brucella abortus. J. Proteome Res. 6, 1519–1529. doi:10.1021/pr060636a. Lapaque, N., Forquet, F., de Chastellier, C., Mishal, Z., Jolly, G., Moreno, E., et al. (2006a). Characterization of Brucella abortus lipopolysaccharide macrodomains as mega rafts. Cell. Microbiol. 8, 197–206. doi:10.1111/j.1462-5822.2005.00609.x.

24

General introduction References

Lapaque, N., Moriyón, I., Moreno, E., and Gorvel, J.-P. (2005). Brucella lipopolysaccharide acts as a virulence factor. Curr. Opin. Microbiol. 8, 60–66. doi:10.1016/j.mib.2004.12.003. Lapaque, N., Takeuchi, O., Corrales, F., Akira, S., Moriyón, I., Howard, J. C., et al. (2006b). Differential inductions of TNF-alpha and IGTP, IIGP by structurally diverse classic and non-classic lipopolysaccharides. Cell. Microbiol. 8, 401–413. doi:10.1111/j.1462-5822.2005.00629.x. Mancilla, M., Marín, C. M., Blasco, J. M., Zárraga, A. M., López-Goñi, I., and Moriyón, I. (2012). Spontaneous excision of the O-polysaccharide wbkA glycosyltranferase gene is a cause of dissociation of smooth to rough Brucella colonies. J. Bacteriol. 194, 1860–7. doi:10.1128/JB.06561-11. Martínez de Tejada, G., Pizarro-Cerdá, J., Moreno, E., and Moriyón, I. (1995). The outer membranes of Brucella spp. are resistant to bactericidal cationic peptides. Infect. Immun. 63, 3054. McDermott, J., Grace, D., and Zinsstag, J. (2013). Economics of brucellosis impact and control in low-income countries. Rev. Sci. Tech. 32, 249–61. Mirabella, A., Terwagne, M., Zygmunt, M. S., Cloeckaert, A., De Bolle, X., and Letesson, J. J. (2013). Brucella melitensis MucR, an orthologue of Sinorhizobium meliloti MucR, is involved in resistance to oxidative, detergent, and saline stresses and cell envelope modifications. J. Bacteriol. 195, 453–65. doi:10.1128/JB.01336-12. Monreal, D., Grilló, M. J., González, D., Marín, C. M., De Miguel, M. J., López-Goñi, I., et al. (2003). Characterization of Brucella abortus O-polysaccharide and core lipopolysaccharide mutants and demonstration that a complete core is required for rough vaccines to be efficient against Brucella abortus and Brucella ovis in the mouse model. Infect. Immun. 71, 3261–71. doi:10.1128/iai.71.6.3261-3271.2003. Moreno, E. (2014). Retrospective and prospective perspectives on zoonotic brucellosis. Front. Microbiol. 5, 213. doi:10.3389/fmicb.2014.00213. Moreno, E., Berman, D. T., and Boettcher, L. A. (1981). Biological activities of Brucella abortus lipopolysaccharides. Infect. Immun. 31, 362–370. Moreno, E., and Moriyón, I. (2002). Brucella melitensis: a nasty bug with hidden credentials for virulence. Proc. Natl. Acad. Sci. U. S. A. 99, 1–3. doi:10.1073/pnas.022622699. Moreno, E., and Moriyón, I. (2006). “The genus Brucella,” in The Prokaryoes. Moreno, E., Stackebrandt, E., Dorsch, M., Wolters, J., Busch, M., and Mayer, H. (1990). Brucella abortus 16S rRNA and lipid A reveal a phylogenetic relationship with members of the alpha-2 subdivision of the class Proteobacteria. J. Bacteriol. 172, 3569–76. Moriyón, I., and Berman, D. T. (1982). Effects of nonionic, ionic, and dipolar ionic detergents and EDTA on the Brucella cell envelope. J. Bacteriol. 152, 822–8.

25

General introduction References

Moriyón, I., Gamazo, C., and Díaz, R. (1987). Properties of the outer membrane of Brucella. Ann. Inst. Pasteur. Microbiol. 138, 89–91. Moriyón, I., Grillo, M. J., Monreal, D., Gonzalez, D., Marin, C., López-Goñi, I., et al. (2004). Rough vaccines in animal brucellosis: Structural and genetic basis and present status. Vet. Res. 35, 1–38. doi:10.1051/vetres:2003037. Moriyón, I., and López-Goñi, I. (1998). Structure and properties of the outer membranes of Brucella abortus and Brucella melitensis. Int. Microbiol. 1, 19–26. Nikaido, H. (2003). Molecular basis of bacterial outer membrane permeability revisited. Microbiol. Mol. Biol. Rev. 67, 593–656. doi:10.1128/MMBR.67.4.593- 656.2003. O’Callaghan, D., Cazevieille, C., Allardet-Servent, A., Boschiroli, M. L., Bourg, G., Foulongne, V., et al. (1999). A homologue of the Agrobacterium tumefaciens VirB and Bordetella pertussis Ptl type IV secretion systems is essential for intracellular survival of Brucella suis. Mol. Microbiol. 33, 1210–20. Ohto, U., Fukase, K., Miyake, K., and Satow, Y. (2007). Crystal structures of human MD- 2 and its complex with antiendotoxic lipid IVa. Science (80-. ). 316, 1632–1634. doi:10.1126/science.1139111. OIE (2016). “OIE: Chapter 2.1.4. Brucellosis (Brucella abortus, B. melitensis and B. suis) (infection with B. abortus, B. melitensis and B. suis),” in Manual of diagnostic tests and vaccines for terrestrial animals. Palacios-Chaves, L., Conde-Álvarez, R., Gil-Ramírez, Y., Zúñiga-Ripa, A., Barquero-Calvo, E., Chacón-Díaz, C., et al. (2011). Brucella abortus ornithine lipids are dispensable outer membrane components devoid of a marked pathogen-associated molecular pattern. PLoS One 6, e16030. doi:10.1371/journal.pone.0016030. Raetz, C. R. H., Reynolds, C. M., Trent, M. S., and Bishop, R. E. (2007). Lipid A modification systems in gram-negative bacteria. Annu. Rev. Biochem. 76, 295– 329. doi:10.1146/annurev.biochem.76.010307.145803. Raetz, C. R. H., and Whitfield, C. (2002). Lipopolysaccharide endotoxins. Annu. Rev. Biochem. 71, 635–700. doi:10.1146/annurev.biochem.71.110601.135414. Rasool, O., Freer, E., Moreno, E., and Jarstrand, C. (1992). Effect of Brucella abortus lipopolysaccharide on oxidative metabolism and lysozyme release by human neutrophils. Infect. Immun. 60, 1699–702. Ronneau, S., Moussa, S., Barbier, T., Conde-Álvarez, R., Zúñiga-Ripa, A., Moriyón, I., et al. (2014). Brucella , nitrogen and virulence. Crit. Rev. Microbiol. 42, 1–19. doi:10.3109/1040841X.2014.962480. Salcedo, S. P., Marchesini, M. I., Degos, C., Terwagne, M., Von Bargen, K., Lepidi, H., et al. (2013). BtpB, a novel Brucella TIR-containing effector protein with immune modulatory functions. Front. Cell. Infect. Microbiol. 3, 28. doi:10.3389/fcimb.2013.00028.

26

General introduction References

Salcedo, S. P., Marchesini, M. I., Lelouard, H., Fugier, E., Jolly, G., Balor, S., et al. (2008). Brucella control of dendritic cell maturation is dependent on the TIR-containing protein Btp1. PLoS Pathog. 4, e21. doi:10.1371/journal.ppat.0040021. Scholz, H. C., Hubalek, Z., Nesvadbova, J., Tomaso, H., Vergnaud, G., Le Flèche, P., et al. (2008). Isolation of Brucella microti from soil. Emerg. Infect. Dis. 14, 1316–7. doi:10.3201/eid1408.080286. Scholz, H. C., Nockler, K., Gollner, C., Bahn, P., Vergnaud, G., Tomaso, H., et al. (2010). Brucella inopinata sp. nov., isolated from a breast implant infection. Int. J. Syst. Evol. Microbiol. 60, 801–808. doi:10.1099/ijs.0.011148-0. Sieira, R., Comerci, D. J., Sánchez, D. O., and Ugalde, R. A. (2000). A homologue of an operon required for DNA transfer in Agrobacterium is required in Brucella abortus for virulence and intracellular multiplication. J. Bacteriol. 182, 4849–55. Sola-Landa, A., Pizarro-Cerda, J., Grillo, M.-J., Moreno, E., Moriyón, I., Blasco, J.-M., et al. (1998). A two-component regulatory system playing a critical role in plant pathogens and endosymbionts is present in Brucella abortus and controls cell invasion and virulence. Mol. Microbiol. 29, 125–138. doi:10.1046/j.1365- 2958.1998.00913.x. Soler-Lloréns, P. F., Quance, C. R., Lawhon, S. D., Stuber, T. P., Edwards, J. F., Ficht, T. A., et al. (2016). A Brucella spp. isolate from a Pac-Man frog (Ceratophrys ornata) reveals characteristics departing from classical Brucellae. Front. Cell. Infect. Microbiol. 6, 116. doi:10.3389/fcimb.2016.00116. Solera, J. (2010). Update on brucellosis: Therapeutic challenges. Int. J. Antimicrob. Agents 36, 18–20. doi:10.1016/j.ijantimicag.2010.06.015. Thiele, O. W., and Schwinn, G. (1973). The free lipids of Brucella melitensis and Bordetella pertussis. Eur. J. Biochem. 34, 333–344. doi:10.1111/j.1432- 1033.1973.tb02764.x. Vences-Guzmán, M. Á., Guan, Z., Ormeño-Orrillo, E., González-Silva, N., López-Lara, I. M., Martínez-Romero, E., et al. (2011). Hydroxylated ornithine lipids increase stress tolerance in Rhizobium tropici CIAT899. Mol. Microbiol. 79, 1496–514. doi:10.1111/j.1365-2958.2011.07535.x. Viadas, C., Rodríguez, M. C., Sangari, F. J., Gorvel, J.-P., García-Lobo, J. M., and López- Goñi, I. (2010). Transcriptome analysis of the Brucella abortus BvrR/BvrS two- component regulatory system. PLoS One 5, e10216. doi:10.1371/journal.pone.0010216. Whatmore, A. M., Davison, N., Cloeckaert, A., Al Dahouk, S., Zygmunt, M. S., Brew, S. D., et al. (2014). Brucella papionis sp. nov., isolated from baboons (Papio spp.). Int. J. Syst. Evol. Microbiol. 64, 4120–8. doi:10.1099/ijs.0.065482-0. Zhao, Y., Hanniffy, S., Arce-Gorvel, V., Conde-Álvarez, R., Oh, S., Moriyón, I., et al. (2017). Immunomodulatory properties of Brucella melitensis lipopolysaccharide determinants on mouse dendritic cells in vitro and in vivo. Virulence, 0. doi:10.1080/21505594.2017.1386831.

27

General introduction References

Zúñiga-Ripa, A., Barbier, T., Conde-Álvarez, R., Martínez-Gómez, E., Palacios-Chaves, L., Gil-Ramírez, Y., et al. (2014). Brucella abortus depends on pyruvate phosphate dikinase and malic enzyme but not on Fbp and GlpX fructose-1,6-bisphosphatases for full virulence in laboratory models. J. Bacteriol. 196, 3045–57. doi:10.1128/JB.01663-14.

28

Objectives and approach

Objectives and approach

31

Objectives and approach

OBJECTIVES AND APPROACH

Brucella LPS is a modified PAMP critical for virulence since it is not well recognized by the principal elements of the innate immune response opening a time window that allows the bacteria to reach a safe intracellular niche. Brucella lipid A presents a diaminoglucose disaccharide, but the genes responsible for its synthesis remain unknown. Moreover, in the last years, some of the genes involved in the synthesis of the Brucella LPS core section have been identified. The recently elucidated chemical structure of this section shows that it is a branched structure. Mutants defective in the synthesis of the core lateral branch in smooth-type LPS spp. as B. abortus or B. melitensis trigger a stronger immune response and protect against brucellosis. This opens new perspectives for the design of more effective vaccines, an important need since vaccination is the only way to control brucellosis. Besides, there is not a specie- specific vaccine against B. ovis PA, a Brucella spp. that causes serious problems in livestock all over the world. The design of genetically modified Brucella vaccines requires a better understanding of the genes involved in LPS synthesis and some aspects about the genetics of the core-lipid A region remain still unknown.

Taking into account these ideas, our aim has been to contribute to the understanding of the genes involved in the synthesis of Brucella LPS core and lipid A sections and their role in the interaction with elements of the innate immune response and virulence. We have also applied this knowledge to the design of new B. ovis PA genetically modified vaccine candidates against brucellosis caused by this Brucella spp.

Accordingly, the specific objectives presented in this thesis have been:

1. To identify the genes involved in the synthesis of the diaminoglucose backbone of Brucella lipid A.

2. To identify ORFs coding for putative glycosyltransferases in Brucella genome and to study their role in LPS core synthesis.

3. To construct and characterize the LPS of mutants in those genes and study the consequences of those mutations in the interaction with elements of innate immunity and virulence.

33

Objectives and approach

4. To study the effect of mutation of genes involved in the synthesis of the LPS core lateral branch in a genetically modified B. ovis PA strain able to grow in atmospheric conditions.

5. To investigate the role of the ORF situated immediately upstream wboA-wboB in O- polysaccharide synthesis.

34

Capítulo 1

Capítulo 1

Identificación de dos genes potencialmente implicados en la síntesis de diaminoglucosa, un componente esencial del lípido A de Brucella

37

Capítulo 1 Abreviaturas

Abreviaturas utilizadas en el Capítulo 1

ADN Ácido desoxirribonucleico ARN Ácido ribonucleico Amp Ampicilina BSA Sero Albúmina Bovina CD (del inglés) cluster of differentiation CIMA Centro de Investigación Médica Aplicada cm centímetros EDTA Ácido Etilendiaminotetraacético EMBL Laboratorio Europeo de Biología Molecular GlcNac N-acetil glucosamina HCl Ácido Clorhídrico HPTLC Cromatografía en Capa Fina de Alta Resolución

H2O2 Peróxido de Hidrógeno IPTG Isopropil-β-D-1-tiogalactopiranósido kDa kilodaltons kdo Ácido 2-keto-3-deoxioctanato KEGG Enciclopedia de Genes y Genomas de Kioto Km Kanamicina LB Luria Bertani L-Glu Ácido glutámico LPS Lipopolisacárido M Molar mA miliamperios mg miligramos ml mililitros mM miliMolar NAD+ Nicotinamida Adenina Dinucleótido (forma oxidada) NADH Nicotinamida Adenina Dinucleótido (forma reducida) NaCl Cloruro sódico

NaH2PO4 Fosfato de disodio Nal Ácido nalidíxico NCBI Centro Nacional para la Información Biotecnológica NGlcNac N-acetil diaminoglucosa Ni níquel nm nanómetros ORF Marco Abierto de Lectura PAGE Electroforesis en Gel de Poliacrilamida pb pares de bases PBS Tampón fosfato salino

39

Capítulo 1 Abreviaturas

PCR Reacción en Cadena de la Polimerasa psi libra por pulgada cuadrada PVDF Polifluoruro de vinilideno rpm revoluciones por minuto SDS Dodecilsulfato sódico SSGCID Seattle Structural Genomics Center for Infectious Disease TLR Receptor tipo Toll Tn-seq (del inglés) transposon sequencing TSA Agar Soja Tripticasa TSB Caldo Soja Tripticasa UDP Uridina Difosfato V voltios vol volumen μl microlitros μm micrómetros μM microMolar

40

Capítulo 1 Resumen

RESUMEN

El lipopolisacárido (LPS) es una molécula clave en la virulencia de Brucella. Algunas bacterias pueden modificar su lípido A, sección interna del LPS, para evitar su reconocimiento por el sistema inmune innato. El lípido A de Brucella está formado por un disacárido de diaminoglucosa, a diferencia de la mayoría de bacterias gram negativas cuyo lípido A está formado por glucosamina. Brucella posee genes necesarios para la síntesis de N-acetil-glucosamina, el precursor de la glucosamina del lípido A en la vía de síntesis clásica; sin embargo, se desconoce la vía de síntesis de N- acetil-diaminoglucosa. En bacterias filogenéticamente lejanas a Brucella, la conversión de N-acetil-glucosamina a N-acetil-diaminoglucosa está mediada por las enzimas GnnA y GnnB, una oxidorreductasa y una aminotransferasa, codificadas por genes que se localizan en un mismo operón. El análisis del genoma de Brucella reveló la presencia de dos ORFs (BAB1_1617 y BAB1_1616) que podrían codificar gnnA y gnnB, respectivamente. Para probar si estos genes eran los responsables de la síntesis de N- acetil-diaminoglucosa en Brucella, se clonaron y se expresaron las proteínas codificadas por ellos. Se realizó un ensayo de actividad enzimática para identificar, mediante cromatografía en capa fina, la formación de N-acetil-diaminoglucosa a partir de N-acetil-glucosamina. Sin embargo, no fuimos capaces de detectar la formación de dicho producto de reacción. Igualmente, con el fin de estudiar si la diaminoglucosa del lípido A es esencial en la biología de esta bacteria, se intentaron mutar ambos genes. Tras numerosos intentos, no se pudieron obtener mutantes, lo que sugiere que ambos genes son esenciales para Brucella. Este hecho apunta que la diaminoglucosa no puede ser sustituida por glucosamina en el lípido A de Brucella y en consecuencia que, tal y como existe, la estructura del lípido A es fundamental para la estabilidad de la membrana externa y la viabilidad de la bacteria.

41

Capítulo 1 Introducción

INTRODUCCIÓN

El reconocimiento del lipopolisacárido por las células del huésped se produce a través del complejo receptor TLR4-MD-2-CD14, presente en muchos tipos de células, incluidos los macrófagos y las células dendríticas. En este proceso, además de la región del núcleo (ver Capítulo 2), el lípido A juega un papel importante ya que sus grupos cargados negativamente interaccionan con el anillo hidrofóbico de MD-2 (Gruber et al., 2004). La interacción con MD-2 activa una cascada de señalización que desencadena la respuesta pro-inflamatoria. El lípido A juega también un papel importante en la interacción con los péptidos catiónicos y el complemento presente en el suero (Gil-Ramírez 2011; Moreno et al. 1990, Conde-Álvarez et al., 2018). Para facilitar el escape de la respuesta inmune, algunos patógenos como Brucella han desarrollado mecanismos para modificar esta región del LPS (Ernst et al., 2001; Gruber et al., 2004).

El lípido A de la mayoría de las enterobacterias es un disacárido de N-acetil- glucosamina (GlcNAc) fosforilado en las posiciones C1´ y C4’, al que se unen cadenas acilo y acil-oxiacilo de C12, C14 y C16 mediante enlaces éster y amida (Raetz and Whitfield, 2002). En la mayoría de las bacterias gram-negativas, esta estructura se sintetiza siguiendo una vía clásica, que implica nueve enzimas (Figura 1.1). En E. coli, la vía clásica comienza con la acilación de GlcNAc mediada por LpxA, seguida de la desacetilación de UDP-3-O-(acil)-GlcNAc llevada a cabo por LpxC y una segunda acilación para formar UDP-2,3-diacil-GlcN catalizada por LpxD (Figura 1.1). A continuación, LpxH añade un grupo pirofosfato a la molécula de UDP-2,3-diacil-GlcN y LpxB forma el disacárido uniendo mediante enlace β (1-6) una de estas moléculas con

UDP-2,3-diacil-GlcN. Este disacárido (ya tetra-acilado) se llama lípido IVA, y es fosforilado a continuación por LpxK. Posteriormente, uno o dos residuos de Kdo son incorporados por la enzima bifuncional WaaA (KdtA). El último paso de la síntesis del Kdo2-lípido A es la adición de cadenas acil al esqueleto disacarídico, mediada por LpxL y LpxM (Iriarte et al., 2004). Además, algunas bacterias presentan enzimas que modifican el lípido A una vez sintetizado (Raetz et al. 2007; 2009).

43

Capítulo 1 Introducción

Figura 1.1. Representación de la vía clásica de síntesis lípido A en Escherichia coli. LpxA añade un grupo acilo a la glucosamina, que es a continuación desacetilada por LpxC y LpxD. LpxH añade un grupo pirofosfato y LpxB forma el disacárido tetra-acilado, que recibe el nombre de Lípido IVA. LpxK fosforila este Lípido IVA y posteriormente, uno o dos residuos de Kdo son incorporados por KdtA. Por último, LpxL y LpxM añaden las cadenas acilo. Adaptado de Raetz et al. 2007.

Como ya se ha comentado en la Introducción General, el disacárido que forma el esqueleto del lípido A de Brucella está formado por dos residuos de N-acetil- diaminoglucosa (2,3-diamino-2,3-dideoxi-D-glucosa, NGlcNac) unidos por enlace β(1-6) en lugar de N-acetil-glucosamina (Qureshi et al., 1994; Rojas et al., 1994), sustituido con ácidos grasos, algunos de ellos de cadena excepcionalmente larga (Moreno et al. 1990; Velasco et al. 2000; Introducción General). Presumiblemente, este cambio sucede porque la síntesis comienza con la acilación de diaminoglucosa (UDP-NGlcNAc) en vez de glucosamina (UDP-GlcNAc), como ocurre por ejemplo en Acidothiobacillus o Campylobacter (Sweet et al., 2004; van Mourik et al., 2010) (Figura 1.2).

44

Capítulo 1 Introducción

Figura 1.2. Representación de la vía de síntesis lípido A en B. abortus. Las ORF indicadas son hipotéticas y están basadas en la homología de secuencia con las descritas para E. coli (LpxA, LpxC, LpxB, LpxK y KdtA), R. leguminoarum (LpxXL) y A. ferrooxidans (GnnA y GnnB).

La comparación del genoma de Brucella con el de otras enterobacterias ha ayudado a identificar los genes que podrían estar implicados en la síntesis de la vía conservada del lípido A (Iriarte et al., 2004). Los homólogos en B. abortus de LpxA, LpxC y LpxD serían BAB1_1173, BAB1_1443 y BAB1_1175 respectivamente. Sin embargo, Brucella no presenta un homólogo de LpxH, pero conserva LpxB (BAB1_1171). El disacárido de diaminoglucosa unida a cuatro ácidos grasos saturados (C16:0 a C18:0) e hidroxilados (3-OH-C12:0 a 29-OH-C30:0) de cadena muy larga (lípido

IVA) sería fosforilado a continuación por LpxK (BAB2_0210) y los residuos de Kdo incorporados por la enzima bifuncional WaaA, KdtA (BAB2_0209). El genoma de Brucella no contiene homólogos de LpxL y LpxM; sin embargo, presenta proteínas con alta homología a la proteína transportadora de grupos acilo, AcpXL (BAB1_0874) y a la acil-transferasa LpxXL (BAB1_0870), que incorporan ácidos grasos de cadena muy larga (de C28 a C32) al lípido A de Rhizobium (Basu et al., 2002; Iriarte et al., 2004). Esto concuerda con la presencia de ácidos grasos similares en el LPS de Brucella (Ferguson et al., 2004; Moreno et al., 1990; Velasco et al., 2000). Estos ácidos grasos también contribuirían a la baja endotoxicidad del lípido A de esta bacteria, ya que impiden el correcto reconocimiento del LPS por TLR4-MD-2 (Lapaque et al. 2005; Conde-Álvarez

45

Capítulo 1 Introducción

2008), como ocurre también en Ochrobactrum, Rhizobium y Legionella (Basu et al., 2002; Lapaque et al., 2006; Zähringer et al., 1995). La estructura del lípido A de Brucella está regulada por BvrR/BvrS. Los mutantes en este regulador presentan un grado menor de acilación y un aumento en la fluidez de las cadenas acilo que afectan a la unión de péptidos catiónicos como la polimixina, pero no a la inducción de citoquinas (Manterola et al., 2005).

Para contribuir a la evasión del sistema inmune, además de los ácidos grasos de cadena excepcionalmente larga, Brucella incorpora otras modificaciones al lípido A. Uno de los mecanismos más utilizados por las bacterias con este fin es la reducción de la carga negativa global mediante la incorporación de arabinosamina y/o etanolamina- fosfato al lípido A (Raetz et al. 2009; Needham and Trent 2013). En Brucella no se ha detectado arabinosamina (Freer et al., 1995; Moreno et al., 1990); sin embargo, el LPS de B. melitensis 16M sí presenta etanolamina-fosfato unida probablemente al grupo fosfato C4’ (Casabuono et al., 2017; Conde-Álvarez et al., 2018; Gil-Ramírez, 2011; Apéndice). La transferencia de este azúcar cargado positivamente es llevada a cabo por LptA (BMEI0118), homóloga a EptA de E. coli (Gil-Ramírez, 2011; Raetz et al., 2007). Curiosamente, la ORF correspondiente de B. abortus contiene un cambio de nucleótido que conllevaría a la síntesis de una proteína truncada y, de hecho, no se ha detectado etanolamina-fosfato en su LPS (Conde-Álvarez et al. 2018; Moreno et al. 1990; Apéndice). La presencia de etanolamina-fosfato en el lípido A de B. melitensis aumenta la resistencia a los péptidos catiónicos y a la acción bactericida del suero normal (Gil-Ramírez 2011; Moreno et al. 1990, Conde-Álvarez et al., 2018; Apéndice). Por otra parte, algunas bacterias próximas a Brucella producen fosfatasas específicas que actúan sobre el lípido A, eliminando el fosfato en posición C1´ (LpxE) o en posición C4’ (LpxF) del disacárido (Needham and Trent, 2013; Raetz et al., 2007). Brucella tiene un gen homólogo (BMEI1212/BAB1_0761) a lpxE de Rhizobium (Iriarte et al. 2004; Karbarz et al. 2003; Karbarz, Six, and Raetz 2009; Conde-Álvarez et al. 2018; Apéndice), pero en los análisis del LPS de un mutante lpxE en B. abortus se siguen detectando formas mono y bis-fosforiladas, lo cual parece indicar que BAB_0761 no actúa sobre el lípido A de Brucella, sino probablemente sobre otro lípido de membrana (Introducción General; Conde-Álvarez et al. 2018; Apéndice), y que la fosfatasa que da

46

Capítulo 1 Introducción lugar a las formas mono-fosforiladas sigue sin ser identificada. La contribución de LpxE al aumento de la resistencia a la polimixina sería compatible con que esta enzima está actuando sobre otro lípido de membrana distinto de los presentes en el lípido A (Conde-Álvarez et al. 2018; Apéndice). La presencia de un lípido A con etanolamina- fosfato no afecta a la interacción de Brucella con las células dendríticas ni tiene un impacto significativo en su inmunogenicidad o en la virulencia en el modelo murino (Conde-Álvarez et al. 2018; Apéndice) en las condiciones estudiadas hasta la fecha. Finalmente, algunas bacterias modifican el perfil del lípido A hidroxilando sus cadenas acilo (Gibbons et al., 2000; Llobet et al., 2015) mediante LpxO, una dioxigenasa Fe2+/α-cetoglutarato-dependiente. En la mayoría de las especies de Brucella, el ortólogo de lpxO tiene una mutación que lleva a la expresión de una proteína truncada que carece del consenso de la familia de aspartil/asparaginil β- hidroxilasas a la que pertenece LpxO, y por lo tanto no es funcional. Estos resultados están de acuerdo con análisis previos que no detectaron ácidos grasos hidroxilados en el lípido A de B. abortus (Velasco et al., 2000). El ortólogo completo de LpxO, únicamente presente en B. microti y B. vulpis, actúa hidroxilando los lípidos de ornitina y no sobre el lípido A y por ello, lpxO en Brucella se ha denominado olsC (Conde- Álvarez et al. 2018; Introducción General; Apéndice).

El hecho de que la base de la estructura del lípido A de Brucella sea un disacárido de N-acetil-diaminoglucosa (Qureshi et al., 1994; Rojas et al., 1994), presente también en otras bacterias como Leptospira, Mesorhizobium, Acidothiobacillus o Campylobacter, (Russa et al., 1995; van Mourik et al. 2010; Sweet, Ribeiro, and Raetz 2004; Boon Hinckley et al. 2005), pero no en la mayoría de los parientes filogenéticos más cercanos a Brucella, podría ser significativo. Sin embargo, los genes implicados en la síntesis de este disacárido en Brucella no han sido identificados. En las bacterias que poseen este disacárido, la conversión de N-acetil-glucosamina a N-acetil- diaminoglucosa está mediada por GnnA y GnnB. GnnA actúa como una oxidorreductasa que cataliza la transferencia de electrones desde la glucosamina al NAD+, de forma que el alcohol (3CH-OH) en el carbono 3 de la glucosamina se oxida a aldehído (3C=O), dando lugar a una cetona intermedia. GnnB es una transaminasa que actúa sobre esta cetona sustituyendo el grupo aldehído (3C=O) por un grupo amino

47

Capítulo 1 Introducción

(CH-NH2) que toma del ácido glutámico (L-Glu), formándose así diaminoglucosa (Sweet et al., 2004). En este capítulo hemos identificado los genes que hipotéticamente codifican gnnA y gnnB en el genoma de Brucella.

48

Capítulo 1 Material y métodos

MATERIAL Y MÉTODOS

Cepas bacterianas y medios de cultivo. Las bacterias se crecieron en caldo Luria Bertani (LB, Pronadisa) o en placas de LB con agar bacteriológico (Pronadisa) a 37ºC. Cuando fue necesario, los medios se suplementaron con 100 µg/ml de ampicilina (Amp, Sigma Aldrich), 25 µg/ml de ácido nalidíxico (Nal, Sigma Aldrich) o 50 µg/ml de kanamicina (Km, Sigma Aldrich). Las cepas fueron conservadas en viales de leche descremada a -80ºC (Scharlau). Las cepas y plásmidos utilizados en este capítulo se detallan en las Tablas 1.1 y 1.2 respectivamente. El trabajo con Brucella se desarrolló en el laboratorio de Nivel 3 de Bioseguridad del Departamento de Microbiología y Parasitología de la Universidad de Navarra. Los ensayos de radiactividad fueron llevados a cabo en las instalaciones del Departamento de Ciencias de la Alimentación y Fisiología de la Universidad de Navarra.

Manipulación y análisis del material genético. Los cebadores para las PCR se diseñaron utilizando la herramienta Primer 3 input (http://bioinfo.ut.ee/primer3-0.4.0/) y fueron sintetizados por Sigma-Aldrich. El ADN plasmídico se extrajo utilizando el kit Qiaprep spin Miniprep (Qiagen GmbH, Hilden, Germany). La secuenciación de ADN se llevó a cabo por el Servicio de Secuenciación de CIMA (Centro de Investigación Médica Aplicada, Pamplona, España). Las secuencias de ADN fueron también obtenidas de KEGG (Kyoto Encyclopedia of Genes and Genomes) en http://www.genome.jp/kegg/. Para el análisis de las secuencias consenso se utilizó la herramienta Interpro (https://www.ebi.ac.uk/interpro/) del EMBL-EBI (Instituto Europeo de Bioinformática). Para la búsqueda de homologías entre ADN o proteínas se utilizó el programa NCBI (National Center for Biotechnology Information; U.S. National Library of Medicine) en http://www.ncbi.nlm.nih.gov, y el alineamiento de secuencias se realizó mediante la herramienta Clustal Omega del EMBL-EBI en http://www.ebi.ac.uk/Tools/msa/clustalo/. La modelización de estructuras proteicas se llevó a cabo utilizando el programa RaptorX (http://raptorx.uchicago.edu/).

49

Capítulo 1 Material y métodos

Tabla 1.1. Cepas bacterianas. Cepa Características Referencia Brucella abortus

B. abortus 2308 NalR Cepa salvaje, biotipo 1 virulento, LPS liso, Sangari and Mutante espontáneo de la cepa 2308 NalR. Agüero, 1991

Escherichia coli

Stellar F –, endA1, supE44, thi-1, recA1, relA1, gyrA96 phoA, Clontech Φ80d lacZΔ M15, Δ(lacZYA-argF) U169 Δ(mrr- hsdRMS-mcrBC), ΔmcrA, λ–

S17-λpir Cepa conjugante con el plásmido RP4 insertado Simon et al., en el cromosoma (Tpr Smr recA thi hsdRM+, lambda pir 1983 fago lisogen RP4::2-Tc::Mu::Km Tn7)

BL21(DE3) F– ompT hsdSB (rB– mB–) gal dcm (DE3) Novagen

Tabla 1.2. Plásmidos.

Plásmido Características Referencia pCR2.1 Vector de clonación, KmR Invitrogen

pJQKm Vector suicida KmR SacS derivado del pJQ200KS. Scupham and Triplett, 1997

pET21a (+) Vector de expresión con promotor T7, AmpR, lacZ, Novagen origen de replicación f1 y pBR322, secuencia de histidinas C-terminal.

pMSB-26 Fragmento XbaI de 585 pb de ADN cromosómico Este trabajo de B. abortus que contiene el alelo delecionado en fase (∆28-321) de BAB1_1616 (gnnB), clonado en los sitios correspondientes del pJQKm mediante el sistema de clonaje InFusion.

pMSB-35 Fragmento XbaI de 435 pb de ADN cromosómico Este trabajo de B. abortus que contiene el alelo delecionado en fase (∆26-299) de BAB1_1617 (gnnA), clonado en los sitios correspondientes del pJQKm mediante el sistema de clonaje InFusion.

pMSB-39 Fragmento NdeI y XhoI de 954 pb de ADN cromosómico Este trabajo de B. abortus que contiene la ORF BAB1_1617 (gnnA) completa, clonada en los sitios correspondientes del pET21a mediante el sistema de clonaje InFusion. Marcado con 6 histidinas en el extremo C-terminal.

pMSB-40 Fragmento NdeI y XhoI de 1122 pb de ADN cromosómico Este trabajo de B. abortus que contiene la ORF BAB1_1616 (gnnB) completa, clonada en los sitios correspondientes del pET21a mediante el sistema de clonaje InFusion. Marcado con 6 histidinas en el extremo C-terminal.

50

Capítulo 1 Material y métodos

Construcción de mutantes en gnnA y gnnB. Los plásmidos suicidas para la construcción de mutantes por deleción en fase se obtuvieron mediante PCR de solapamiento utilizando hervidos de ADN genómico de B. abortus 2308. Los cebadores se diseñaron en base a la secuencia disponible de los genes correspondientes en B. abortus 2308. Para la construcción del plásmido suicida de gnnA (Figura S1.1 en Material suplementario, Capítulo 1), primero se generaron dos fragmentos de PCR utilizando los cebadores gnnA-F1 (5`-TGGCGGCCGCTCTAGCCAATCGGCGTACAAATCGT-3`) y gnnA-R2 (5`- CCGAGACTCTTCAGGGTACG-3`), que amplificaron un fragmento de 224 pares de bases (pb) incluyendo los codones 1 al 25 de la ORF BAB1_1617, así como 150 pb corriente arriba del codón de inicio de BAB1_1617. Los cebadores gnnA-F3 (5`- CGTACCCTGAAGAGTCTCGGGCAAGGAAGCGATCAGTGTC-3`) y gnnA-R4 (5`- ATCCACTAGTTCTAGATATATCTGGCCCGCACGAA-3`) se utilizaron para amplificar un fragmento de 211 pb que incluía los codones 300 al 318 de BAB1_1617 además de 149 pb corriente abajo de su codón de parada. Los sitios de corte de la enzima de restricción XbaI se encuentran subrayados. Ambos fragmentos fueron ligados mediante PCR de solapamiento utilizando los cebadores F1 y R4 para la amplificación y las regiones complementarias entre los cebadores R2 y F3 para el solapamiento. Mediante el sistema In-Fusion HD Cloning Kit (Clontech), el fragmento F1-R4 se clonó en el vector suicida pJQK (Scupham and Triplett, 1997), cortado con la enzima de restricción XbaI, para generar el plásmido pMSB-35 que contenía el gen gnnA con una deleción de la región central.

La construcción del plásmido suicida del gen gnnB se llevó a cabo de manera similar (Figura S1.2), utilizando los cebadores gnnB-F1 (5`- TGGCGGCCGCTCTAGGCAAGCAGCGTAAAATTGTGTAA-3`) y gnnB-R2 (5`- TTCCGCAACAACCTTGGAAA -3`), que amplificaron un fragmento de 229 pb incluyendo los codones 1 al 27 de la ORF BAB1_1616, así como 148 pb corriente arriba del codón de inicio de BAB1_1616. Los cebadores gnnB-F3 (5´- TTTCCAAGGTTGTTGCGGAACAGACGGCCTATAAGCACTAT-3`) y gnnB-R4 (5`- ATCCACTAGTTCTAGGCCGAAACCGAACTTTCCAA-3`) amplificaron un fragmento de 356 pb incluyendo los codones 322 al 374 así como 194 pb corriente abajo del codón de

51

Capítulo 1 Material y métodos parada de la ORF BAB1_1616. Los sitios de corte de la enzima de restricción XbaI se encuentran subrayados. Ambos fragmentos se ligaron utilizando los cebadores F1 y R4 para la amplificación y R2 y F3 para el solapamiento y el fragmento resultante se clonó en el pJQK, para generar el plásmido pMSB-26 que contenía el gen gnnB con una deleción de la región central.

Los plásmidos pMSB-35 y pMSB-26 se secuenciaron para asegurar que se mantenía el marco de lectura y se transformaron en E. coli Stellar competentes (Clontech) y posteriormente en E. coli S17 λpir (Simon et al., 1983). A continuación, se introdujeron en B. abortus 2308 por conjugación. El primer evento de recombinación se produjo tras la integración del vector en el cromosoma y se seleccionó mediante la resistencia a Nal, inherente a la cepa de B. abortus 2308, y a Km, que aporta el vector pJQK. La segunda recombinación para llevar a la escisión del plásmido se seleccionó por la resistencia a Nal y sacarosa y la sensibilidad a la Km. La diferencia entre los recombinantes que hubieran conservado el gen salvaje y los que eran portadores del gen mutado se realizó por PCR con los cebadores F1 y R4 correspondientes. Así, en el caso de gnnA, se esperaba amplificar un fragmento de 435 pb en las cepas mutantes y de 1256 pb en las cepas revertientes al estado salvaje. La mutación gnnA resultaría en la pérdida del 86,4% de la ORF. En el caso del gen gnnB, se hubiera esperado un fragmento de 585 pb en el mutante y de 1467 pb en las cepas revertientes al estado salvaje. La mutación de gnnB supondría una pérdida del 78,8% de la ORF BAB1_1616. Como se explica en el apartado de Resultados, no se consiguió obtener ningún mutante en la segunda recombinación.

Construcción de vectores de expresión de BAB1_1617 y BAB1_1616. Los vectores que expresan los genes codificados por BAB1_1617 y BAB1_1616 (gnnA y gnnB respectivamente) se construyeron sobre el plásmido de expresión pET21a mediante el sistema de clonaje In-Fusion. Para construir el plásmido que expresaba GnnA con una cola de seis histidinas en el extremo carboxi-terminal, se utilizaron los cebadores gnnA-Fw (5´- AAGGAGATATACATAATGGCACCTCGTATCGCGG-3´) en el extremo amino-terminal y gnnA-Rv (5´-GGTGGTGGTGCTCGAAGGCCGGATGCCGCACCG-3´) en el extremo carboxi-

52

Capítulo 1 Material y métodos terminal. En el caso del plásmido que expresa GnnB, los cebadores gnnB-Fw (5`- AAGGAGATATACATAATGCAGTTCATTGATCTTGGA-3`) y gnnB-Rv (5´- GGTGGTGGTGCTCGAAGGCCTTCTTGCCGTGGAA-3´) fueron utilizados. Los sitios de corte de las enzimas de restricción NdeI y XhoI se encuentran subrayados. Ambos insertos fueron ligados en el vector pET21a, cortado con las enzimas de restricción NdeI y XhoI. Los plásmidos generados se designaron como pMSB-39 (pET21a-gnnA) y pMSB-40 (pET21a-gnnB) y fueron confirmados por PCR y secuenciación. Dichos plásmidos se transformaron en E. coli Stellar (Clontech) y posteriormente en la cepa de expresión E. coli BL21(DE3).

Preparación y caracterización de extractos de E. coli con plásmidos portadores de BAB1_1617 y BAB1_1616. Para la obtención de extractos celulares que contenían los plásmidos de expresión de GnnA y GnnB, las E. coli BL21(DE3)-pMSB-39 o E. coli BL21(DE3)-pMSB-40 se cultivaron a 37°C en 20 ml de caldo LB con 100 µg/ml de ampicilina durante una noche con agitación constante de 120 revoluciones por minuto (rpm). Al día siguiente, se refrescaron 5 ml de cada cultivo en 500 ml de LB con ampicilina y cuando alcanzaron una densidad óptica de 0,6 (A600nm) se indujo la expresión de los genes gnnA o gnnB mediante 1mM de isopropil-β-D-1-tiogalactopiranósido (IPTG) durante 2 horas a 37ºC. Se centrifugaron a 4000xG a 4ºC durante 10 minutos y los sedimentos se resuspendieron en 10 ml de solución tamponada fría por cada 50 ml del cultivo original. Se utilizaron dos soluciones tamponadas distintas para analizar cuál proporcionaba mayor rendimiento. La primera (Tampón A), contenía fosfato potásico (100 mM) a pH 7.5, cloruro sódico (200 mM) y glicerol (20%) y la segunda (Tampón B), Tris-HCl (10 mM) a pH 8, imidazol (10mM), EDTA de potasio (10mM) y lisozima (0,5 mg/ml). En cualquier caso, los extractos se volvieron a centrifugar en las mismas condiciones y los sedimentos se resuspendieron en 10 ml del mismo tampón. Para evitar que las proteínas se desnaturalizasen, además de trabajar en frío, se añadió un inhibidor de proteasas (cOmplete, EDTA-free, Roche).

53

Capítulo 1 Material y métodos

Análisis de la expresión de proteínas mediante SDS-PAGE y Western-Blot. Las muestras se mezclaron 1:1 con Sample buffer 2X (Bio-Rad), se hirvieron durante 10 minutos y analizaron en geles de poliacrilamida del 12% (Criterion XT Precast Gels, BioRad) (37.5:1 acrylamide/methylene-bisacrylamide ratio) en buffer XT MOPS (BioRad). Quince microlitos de cada muestra se corrieron a 200 V y amperaje constante durante 60 minutos. Las proteínas se tiñeron con azul de Coomassie (Coomassie Brilliant Blue R-250, BioRad). Para realizar el Western Blot, los geles fueron transferidos a membranas de PVDF (Whatman, Schleicher & Schuell, WESTRAN S.; tamaño de poro 0,2 μm) en buffer de transferencia (0,025M Tris, 0,192M glicina, 20% (vol/vol) metanol a pH 8,3). La transferencia se realizó a voltaje constante de 8 voltios y 200 mA durante 30 minutos en una célula de transferencia semi-seca (Bio-Rad). Las membranas fueron bloqueadas durante una noche con leche descremada al 5% en Tampón Fosfato Salino (PBS) con un 0,05% de Tween 20 a 4ºC. A continuación, se incubaron durante una noche a temperatura ambiente con un anticuerpo monoclonal anti-histidina (MA1-135 6x-His Epitope Tag Antibody, 4E3D10H2/E3, Thermo Scientific) a una dilución 1:1000 en PBS-Tween. Tras los lavados, se incubaron durante 5 horas con el correspondiente anticuerpo secundario conjugado con peroxidasa diluido 1:500.

Se revelaron utilizando 4-cloro-1-naftol-H2O2.

Purificación de proteínas. GnnA y GnnB se purificaron a partir de los extractos obtenidos de 500 ml de cultivos de E. coli BL21(DE3)-pMSB-39 y BL21(DE3)-pMSB-40 respectivamente. Las muestras fueron sometidas a rotura mecánica mediante dos pases en la Prensa de French (SIM AMINCO) a 1400 psi. Para descartar los restos celulares que no eran de nuestro interés, las muestras se centrifugaron a 10000xG durante 20 minutos a 4ºC separándose así el sobrenadante, que contiene las proteínas solubles entre las que se encuentran GnnA y GnnB, del precipitado, que además de restos celulares, también contiene proteínas asociadas a membranas. A continuación, se cargaron 4 ml de cada lisado en columnas de afinidad de polipropileno (Qiagen) con 1 ml de resina Ni-NTA Agarosa (Qiagen). La mezcla se agitó a 200 rpm durante 1 hora a 4ºC. La columna se lavó dos veces con 4 ml de tampón de lavado (50 mM NaH2PO4, 300 mM NaCl, 20 mM imidazol) a pH 8 y el producto final se

54

Capítulo 1 Material y métodos

recogió en distintas fracciones de 0,5 ml de tampón de elución (50 mM NaH2PO4, 300 mM NaCl, 250 mM imidazol) a pH 8. Dichas fracciones se analizaron mediante SDS- PAGE al 12% y tinción con azul de Coomassie. La cuantificación de las proteínas extraídas se realizó mediante la técnica de Lowry (Lowry et al., 1951; Markwell et al., 1978).

Ensayo de actividad enzimática. La actividad de GnnA y GnnB se analizó a partir de los extractos o de las proteínas purificadas monitorizando la conversión de N-acetil-glucosamina (UDP-GlcNAc) a N- acetil-diaminoglucosa (UDP-NGlcNAc). La reacción se llevó a cabo en 10 µl que contenían 50 mM HEPES pH 8,0, y 0,5 mg/ml (para extractos) o 10 µg/ml (para proteínas purificadas) de cada enzima. Como sustrato, se añadió 80 µM (para extractos) o 40 µM (para proteínas purificadas) de 14C UDP-N-acetil-D-glucosamina (American Radiolabeled Chemicals, Inc.) y como cofactores 1 mM NAD+ (Roche), y 100 mM (para extractos) o 10mM (para proteínas purificadas) de glutamato sódico (CODEX, Carlo Erba). En el ensayo con proteínas purificadas se añadieron además 0,25 mg/ml de BSA (Sigma Aldrich). Los ensayos se llevaron a cabo a 30ºC o 37ºC y la formación del producto de la reacción (UDP-NGlcNAc) se midió a las 4, 6 y 8 horas. Para ello, se cargaron 2 µl de la mezcla en placas de HPTLC Silica gel 60 (hojas de aluminio, 20x20 cm, Merck Millipore), se dejaron secar al aire libre y la carrera se realizó sumergiéndolas en metanol seco (PanReac, AppliChem) durante 2 horas. Las placas se colocaron sobre películas fotográficas para la detección de isótopos emisores de β- y γ- (Amersham Hyperfilm MP High performance autoradiography film 18x24cm, GE Healthcare) y se midió el resultado a distintos tiempos de exposición. En la cromatografía en capa fina, los componentes de la mezcla se separan al pasar a través de la fase estacionaria (placas de HPTLC Silica gel) mediante el flujo ascendente de una fase móvil (metanol seco). Los componentes se separan en función de su afinidad con la fase estacionaria. Así, las sustancias con mayor afinidad quedan retenidas y las de menor afinidad son arrastradas por la fase móvil hacia la parte superior. De este modo, la diaminoglucosa, producto de la reacción de GnnA y GnnB, es una molécula con menos cargas negativas que la glucosamina, por lo que presentará menor afinidad a la fase estacionaria y migrará más que el sustrato de la reacción.

55

Capítulo 1 Resultados

RESULTADOS

Identificación de los genes hipotéticamente implicados en la síntesis del disacárido de diaminoglucosa en el lípido A de B. abortus. Los análisis químicos del LPS han demostrado que Brucella posee un lípido A constituido por un esqueleto de diaminoglucosa (NGlcN) (Qureshi et al., 1994; Rojas et al., 1994). Cuando se analizó la presencia en el genoma de Brucella de genes homólogos a los implicados en la síntesis de diaminoglucosa en Leptospira, Acidothiobacillus o Campylobacter, se identificaron dos genes contiguos, BAB1_1617 y BAB1_1616, que presentaban un alto porcentaje de homología con los genes que codifican GnnA y GnnB en dichas especies (Figuras 1.3 y 1.4) Concretamente, BAB1_1616 comparte con los genes de estas especies el domino conservado, que es una lisina (K) en posición 191 (Figura 1.4). Además, ambos genes se encuentran conservados en las principales especies de Brucella y están presentes en sus parientes filogenéticos más cercanos como Ochrobactrum o Rhizobium (Tabla 1.3). Las ORF de B. abortus se encuentran en el cromosoma I separadas por 160 pares de bases y mantienen la anotación de oxidorreductasa (BAB1_1617) y aminotransferasa (BAB1_1616). Además, comparten la organización genómica descrita para gnnA y gnnB en Campylobacter (Sweet et al., 2004; van Mourik et al., 2010). Como se puede observar en la Figura 1.5, se trata de una región importante del genoma de Brucella ya que contiene genes del ciclo celular (ctrA, chpT y pdhS) y otros hipotéticamente relacionados con el LPS (ver Capítulo 2). Aunque en A. ferrooxidans las ORF forman un operón con lpxA y lpxB, también implicadas en la síntesis del lípido A, en Brucella, al igual que en Campylobacter jejuni, no es el caso (Figura S1.3, Material suplementario).

57

Capítulo 1 Resultados

Brucella --MAPRIAVLGCGYWGGNHIRTLKS--LGALQAVSDSNAEKAERFASEFNVPSIPVEELF Campylobacter ----MKIGIIGLGKMGQNHLNELSKNSHFKLNALFDLCKNPNLNIFD---DIFYDDLDKF Leptospira MTERVKLGVIGTGHMGQYHVNVAKTLNDATLVGIYDSDLERAKQIAEKHKTLAFSTIDEL Acidithiobacillus -MIHMRTGVIGVGHLGRFHAQKYAA--ISQLAGVFDENAERAAEVAAELRCRAFPSVDAL : .::* * * * . * .: * : .. : :

Brucella IHPDIDGIVLALPPQLHAQYAMEAVKNGKDVLVEKPIALDIPVALAEVEAARENGRVFMV Campylobacter LNQNNDIIIIATPTNSHLTIAKKVFKQCKCVLIEKPLALNLKEIDEISNLAKEYSIKVGV Leptospira I-SKTDAVVIAVPTFLHHEIAKKALEANKHVLVEKPIAETTNQAKELVKIAGEKNLVLLV Acidithiobacillus L-AEVDAVSIVTPTSTHFAVAEVAMQAGVHCLIEKPFTLDTEEADALIGMAQERHLVLAI : . * : :. * * * ..: *:***:: * * . :

Brucella GHVLRFHPAFEKLLDMVQSGELGDIRYVHSHRVGLG-KFHTEFDALWDFAPHDLSMILAI Campylobacter GFCERFNPAVLALKK---ELENEEIISINIQRFSPYPQRISDVGILQDLAVHDLDLLCFL Leptospira GHVERFNGAVLELSK----IVKD-PLLIESRRLAPSNSRIKDVGVVLDMMIHDIDIVLNL Acidithiobacillus GHIKRVHPAIQYLRQ---AGFGA-PRYLEAERLAPFKPRSLDIDVIMDLMIHDLDLTLLL *. *.: *. * . :. .*.. :.. : *: **:.: :

Brucella TGEEPNVVRGEGVAILDHLNDFAHLHMEFPSGIRGHLFASRLNAYRERRLSVTGTKGMAV Campylobacter SKQEITKTNLLKKYTQDQTRESESIILCGLEKCIASIHQSWNSTQKLRKIHLITKNHFYE Leptospira VNSPVKKLSASGKKVSSDHEDVANVLLEFENGCLASITASRATQAKIRTLNITQKDVYIM Acidithiobacillus TGAEPVDVRAVGVAAVTDKADMATAWMTLNNGTVANLAASRVVREPARRMRIFWQDRYAS . : : . . : * * : : .

Brucella FDDGEAWERKLALYRHEVWR---ENDRWAFKS-ADPVYIQTEEGMPLTRELQHFMHCIET Campylobacter AN---LNDFSL------LKDG------NFIELTQQSPLFSEHEALLKLIDN Leptospira LD---FTDQEIELHRQATSDILLLSEEIKYRQESIVEKIFVHKDNPLKQEHEHFIRCIRK Acidithiobacillus VD---FLNNTLHIYHRGAGTVPGIPGV---RD----EAVDLAKRDALAAEIEDFLNAIAA : : : : : * * : ::. *

Brucella RETPRTDGKEAISVLRILTE------GTVRHPA Campylobacter QPNHLASTS------DAYKVQEILERFA---- Leptospira EIEPIVNRNSDVSTLEIAYQILSEIHGTSNK-- Acidithiobacillus HRPVFCDGVAGRRVLAAALQVRVAVEAFLQR-- . . :

Figura 1.3. Alineamiento de la secuencia de aminoácidos codificada por gnnA en las distintas especies que presentan un disacárido de diaminoglucosa en el lípido A. Leptospira interrogans serovar Linhai (código GenBank LIL_10526), Campylobacter jejuni subespecie jejuni (NCTC 11168 = ATCC 700819; código GenBank Cj0504c), Brucella abortus 2308 (código GenBank BAB1_1617) y Acidithiobacillus ferrooxidans (ATCC 23270; código GenBank AFE_1457). Los asteriscos indican secuencias que presentan el mismo aminoácido conservado; dos puntos marcan un cambio de aminoácido por otro con propiedades similares y un punto indica un cambio por otro aminoácido con distintas propiedades.

58

Capítulo 1 Resultados

Brucella ------MQFIDLGAQRARIENRLNAAISKVVAEGRYILGPEVAEFEKKLGEYLGVEHVIA Campylobacter ------MNFINLQAQYLAYKDEINAEIESVLSSSSFIGGAKLNEFEQNLAHFLGVKHAIG Leptospira ---MINIPYINLVEQWKNEREELLPVLDSVLGSGQYIGGQEVAKFEEDVAKFCGVKYAVA Acidithiobacillus MTQNTAIPMVDLRAHFAPLRDEILTGIGKILDDASFILGNQGRALEAEVAGLSGVAHGVG : ::* : .:.: : .:: .. :* * : :* .:. ** : :.

Brucella CANGTDALQMPLMTRGIGPGHAVFVPSFTFAATAEVVALVGAEPVFVDVDPDSYNMNVEQ Campylobacter CSSGTSALYLALRVLDIGKDDEVIVPSFTFIATAEVVALVGAKPVFADINLSNYNLDFKA Leptospira LNSGTDALVCGLLELGIQPGDEVITPPNSFIASTASIVHIKAKPVFVDVGED-QNLDPEK Acidithiobacillus CASGTDALMLALRALEIGPGDEVIVPTFTFIATAEAVLYVGATPVFVDVDDRFYAMTIAG .**.** * * .. *:.* :* *:: : : * ***.*:. :

Brucella LEAAIAATIKEGRLEPKAIIPVDLFGLAASYNRITAIAEREGLFIIEDAAQSIGGKRDNV Campylobacter VQKAITS------KTKAVIAVSMFGQMSDLRALEEILKDKNITLIEDGAQSFGASFKGE Leptospira LEKAITS------KTKAIMPVHLTGRVAAMNEIMRIADKYSIPVIEDAAQSIGSKYDGK Acidithiobacillus IEAAITP------RTKAIIPVHLYGLPADMPGIMALAQKHGLRVIEDCAQAIGAQINGQ :: **: . **:: * : * : : : . .: :*** **::*.. ..

Brucella MCGAFGHVGATSFYPAKPLGCYGDGGAMFTNDAELADTLRSVLFHGKGETQYDNVRIGIN Campylobacter KSCSIAKISCTSFFPSKPLGAYGDGGAIFCHDDEIAKKIRILLNHGQTQ-RYKHEFIGIN Leptospira FSGSIGKIGCFSTHPLKNLNACGDGGFLTTNDEKIYNSVSRVRNHGLID-RNTVGEFGFV Acidithiobacillus GVGSFGDIGCFSFFPSKNLGAAGDGGMVVTADAELERKLRGLRNHGSWQ-TYHHDVLGYN ::..:.. * .* * *.. **** : * :: .: : ** : :*

Brucella SRLDTIQAAVLLEKLAILEDEMEARDRIARRYNEALKD-VVKVPELPAGNRSAWAQYSIE Campylobacter GRLDTLQAAILNVKLKYLEKELDKRQKLAQIYNANLKN--CQIPQINPNAFSAYAQYSVL Leptospira TRMDAIQAAILNFRLTRLPQVIEKRRQNAQLYKTLLDKRNVFIPEDKPLEFNTYHTFVIQ Acidithiobacillus SRLDEMQAVILRAEFPHLAAYNDGRRRAAGWYAEHLVGLDLQLPEAPAGYHHVFHQFTIQ *:* :**.:* .: * : * : * * * :*: .: : :

Brucella SENRDGLKAQLQAEGIPSVIYYVKPLHLQTAYKHYSVAPGGLPVSESLPSRILSLPMHPY Campylobacter VEDRASVLRKFEKANIPYAIHYPTPLHKQPCFSEFS--NLELKNSEYASEHILSLPFSPF Leptospira VERRDELKDYLFSQGIETSIHYPIPIHLQPASRNLGYKQGDFPITEKQAGRILTLPIHQY Acidithiobacillus LNARDAVKTALHAEGIASAIYYPIPGHQQKMFAHQA--QTHCPVAEHLAERVLSLPMFPE : * : : .* *:* * * * . . :* ::*:**:

Brucella LSEADQDKIIGVIRG-FHGKKA Campylobacter LSEEEQEQVICIFKD------Leptospira LKDVDLHKIAQAVNRFYES--- Acidithiobacillus LREEQIARIATVIRRTLHG--- * : : :: ..

Figura 1.4. Alineamiento de la secuencia de aminoácidos codificada por gnnB en las distintas especies que presentan un disacárido de diaminoglucosa en el lípido A. Leptospira interrogans serovar Linhai (código GenBank LIL_12320), Campylobacter jejuni subespecie jejuni (NCTC 11168 = ATCC 700819; código GenBank Cj0505c), Brucella abortus 2308 (código GenBank BAB1_1616) y Acidithiobacillus ferrooxidans (ATCC 23270; código GenBank AFE_1458). Todas comparten el dominio conservado lisina (K) de la proteína (sombreada en gris). Los asteriscos indican secuencias que presentan el mismo aminoácido conservado; dos puntos marcan un cambio de aminoácido por otro con propiedades similares y un punto indica un cambio por otro aminoácido con otras propiedades.

59

Capítulo 1 Resultados

Tabla 1.3. ORF de los genes que hipotéticamente codifican GnnA y GnnB en las especies que presentan un disacárido de diaminoglucosa en el lípido A y su porcentaje de homología con respecto a B. abortus y los parientes filogenéticos más cercanos a Brucella. B. abortus (B. melitensis biovar Abortus 2308), B. melitensis bv. 1 16M, B. ovis ATCC 25840, B. suis ATCC 23445, Rhizobium tropici CIAT899; Ochrobactrum anthropi ATCC 49188; Agrobacterium tumefaciens Ach5_43020; Acidithiobacillus ferrooxidans ATCC 23270, Campylobacter jejuni subespecie jejuni NCTC 11168 = ATCC 700819 y Leptospira interrogans serovar Linhai.

Gen gnnA gnnB Especie ORF % ORF % B. abortus BAB1_1617 BAB1_1616 B. melitensis BMEI0420 99 BMEI0421 99 B. ovis BOV_1545 99 BOV_1544 100 B. suis BSUIS_A1657 99 BSUIS_A1656 99 O. anthropi Oant_1564 88 Oant_1565 94 R. tropici PC07910 29 PB00870 26 A. tumefaciens Ach5_37920 25 Ach5_43020 39 A. ferrooxidans AFE_1457 33 AFE_1458 42 C. jejuni Cj0504c 26 Cj0505c 40 L. interrogans LIL_10526 26 LIL_12320 37

60

Capítulo 1 Resultados

Figura 1.5. Entorno genético de las ORF BAB1_1617 y BAB1_1616, hipotéticos gnnA y gnnB respectivamente en B. abortus. Ambos genes se encuentran en el cromosoma I de Brucella, rodeados por genes como ctrA, chpT y pdhS, relacionados con el ciclo celular (en color azul) o BAB1_1620, cuya función hipotética se ha asignado al LPS (Capítulo 2). Las flechas en color rosa indican genes que codifican para ARN de transferencia de distintos aminoácidos.

La estructura tridimensional de BAB1_1616 fue realizada por el Seattle Structural Genomics Center for Infectious Disease (SSGCID) y el análisis detallado de la secuencia de aminoácidos en la base de datos Interpro, confirmó que se trataba de una “DegT/DnrJ/EryC1/StrS aminotransferasa (Q2YRB9)”, homóloga a la superfamilia de las transferasas dependientes de piridoxal-fosfato. La proteína presenta los dominios conservados en los ortólogos de Leptospira, Campylobacter y Acidithiobacillus: el centro activo donde se encuentra el sitio catalítico cd00616 que incluye una lisina en posición 191 (Figura 1.6A); un sitio de unión al piridoxal 5'-fosfato y un bolsillo de unión inhibidor-cofactor.

Los estudios de cristalización y estructura tridimensional de la proteína codificada por BAB1_1617 se están llevando a cabo actualmente en el SSGCID. El análisis de la secuencia de aminoácidos codificada por BAB1_1617 (GnnA) en Interpro, muestra que se trata de una proteína que posee un dominio oxidorreductasa en el extremo N- terminal (aminoácidos 4 al 118) de unión al NAD(P) (Figura 1.6B). La modelización de la estructura de la proteína se llevó a cabo utilizando el programa RaptorX. Dicho

61

Capítulo 1 Resultados programa modelizó la estructura de BAB1_1617 basándose en el modelo 3ezy.1, una hipotética deshidrogenasa (TM_0414) de Thermotoga marítima (Ramagopal, U.A. et al., resultados no publicados).

Por todo esto, decidimos denominar a BAB1_1617 y BAB1_1616 gnnA y gnnB respectivamente.

A B

Figura 1.6. A. Proteína expresada por BAB1_1616 (gnnB), cristalizada por el SSGCID. Se trata de un homodímero de dos subunidades. En la imagen se representa con el piridoxal unido. B. Modelo propuesto para la proteína expresada por BAB1_1617 (gnnA), (http://raptorx.uchicago.edu/).

Los genes gnnA y gnnB son esenciales para la estabilidad de la membrana externa y la viabilidad de Brucella. Con el fin de estudiar el papel de BAB1_1616 y BAB1_1617 en Brucella, se intentaron construir mutantes por deleción en ambos genes, siguiendo el mismo procedimiento utilizado anteriormente con éxito para la obtención de mutantes por deleción en numerosos genes de Brucella (Barbier et al., 2014; Conde-Álvarez et al., 2006; Gil-Ramírez et al., 2014; Palacios-Chaves et al., 2011, 2012; Soler-Lloréns et al., 2014; Zúñiga-Ripa et al., 2014). Sin embargo, después de analizar más de 500 clones, no se obtuvo ningún mutante en ninguna de estas ORF.

62

Capítulo 1 Resultados

Obtención de proteínas recombinantes GnnA y GnnB y análisis de su actividad enzimática. En A. ferrooxidans, la conversión de UDP-N-acetil-glucosamina (UDP-GlcNAc) en UDP-N-acetil-diaminoglucosa (UDP-NGlcNAc) se puede llevar a cabo in vitro tanto con extractos celulares como con las proteínas GnnA y GnnB purificadas, utilizando como sustrato N-acetil-glucosamina marcada con 14C y como cofactores NAD+ y ácido glutámico (ver más adelante) (Sweet et al., 2004) Para analizar si BAB1_1617 y BAB1_1616 codifican las enzimas responsables de la conversión de N-acetil-glucosamina a N-acetil-diaminoglucosa en Brucella, nos planteamos analizar la actividad enzimática de las proteínas codificadas por ambos genes. Con este fin, ambas ORF se clonaron en el plásmido de expresión pET-21a para obtener pMSB-39 y pMSB-40 que codifican GnnA y GnnB respectivamente, fusionadas con una secuencia de histidinas. A continuación, se indujo su expresión con IPTG y los extractos celulares se analizaron mediante SDS-PAGE, tinción con azul de Coomassie y Western-Blot con un anticuerpo monoclonal anti-histidina. Como se puede observar en la Figura 1.7, E. coli BLB21(DE3)-pMSB-39 y pMSB-40 expresaron proteínas con pesos moleculares de 36,84 y 41,63 kDa, compatibles con los esperados para GnnA y GnnB respectivamente.

63

Capítulo 1 Resultados

Figura 1.7. Expresión de las proteínas GnnA y GnnB a partir de los plásmidos pMSB-39 y pMSB-40 respectivamente. A. Tinción con azul de Coomassie B. Western Blot con anticuerpo monoclonal anti- histidina. Pesos moleculares: GnnA: 36,84 kDa; GnnB: 41,63 kDa.

Seguidamente, se realizó un ensayo de actividad utilizando dichos extractos y siguiendo el modelo propuesto por Sweet y colaboradores (Sweet et al., 2004), utilizando como sustrato N-acetil-glucosamina marcada con 14C (14C-UDP-GlcNac) y como cofactores NAD+ y glutamato sódico (Figura 1.8).

Figura 1.8. Representación del ensayo enzimático propuesto para la conversión de glucosamina en diaminoglucosa llevada a cabo por las enzimas GnnA y GnnB en B. abortus. Se utiliza como sustrato N- acetil-glucosamina marcada con 14C (14C-UDP-GlcNAc), como cofactores nicotinamida adenina dinucleótido en su forma oxidada (NAD+) y ácido glutámico (L-Glu). El producto de la reacción es N- acetil-diaminoglucosa (UDP-NGlcNAc).

64

Capítulo 1 Resultados

El ensayo se realizó a distintas temperaturas (30º y 37ºC), durante distintos tiempos (4, 6 y 8 horas) y se visualizó mediante cromatografía en capa fina. Como controles se utilizaron el sustrato 14C-UDP-GlcNac comercial y una mezcla de los componentes de la reacción, sustituyendo los extractos por HEPES. Como era de esperar, ambos controles migraron a la misma altura, correspondiente a la 14C-UDP- GlcNac. Si en los extractos que contenían GnnA y GnnB de Brucella, se hubiera formado N-acetil-diaminoglucosa (UDP-NGlcNac), ésta debería migrar por encima del sustrato, ya que la primera tiene más cargas positivas; sin embargo, ese no fue el caso y, además, el producto que se detectó fue similar al que se obtuvo con el extracto que contenía el vector pET21a vacío (Figura 1.9).

Figura 1.9. Cromatografía en capa fina tras el ensayo de actividad con los extractos de E. coli que expresan GnnA y GnnB a partir de pMSB-39 y pMSB-40. Imagen representativa de las distintas condiciones probadas. En esta figura se muestra el ensayo realizado a 30ºC durante 4 horas. Se cargaron 2 µl del producto de reacción en una placa de Silica Gel que se expuso a una radiografía de detección de isótopos radiactivos durante 40 horas.

Puesto que los extractos completos podían contenían otras proteínas que podrían interferir en el ensayo (Figura 1.7), se purificaron GnnA y GnnB mediante una cromatografía de afinidad. Ambas proteínas se obtuvieron por elución con imidazol a distintas concentraciones y se seleccionó el eluído E2 puesto que presentaba buenas

65

Capítulo 1 Resultados concentraciones de proteínas (98% en el caso de GnnA y 72% en el de GnnB) (Figura 1.10) (Tabla 1.4). En ambos casos se apreciaron bandas por debajo del peso molecular esperado que podrían ser productos de degradación de la proteína original y bandas que, por su tamaño, podrían corresponder a formas diméricas.

A B

SP M L E1 E2 E3 E4 SP M L E1 E2 E3 E4

110 110 80 80 60 60 50 50 40 40

30 30 20 20

Figura 1.10. Purificación de las proteínas GnnA (36,84 kDa) (A) y GnnB (41,63 kDa) (B). S.P. Sin purificar; M. Matriz; L. Lavados; E. Eluído. Tinción con azul de Coomassie de SDS-PAGE (12%).

Tabla 1.4. Concentración de proteínas obtenida tras la purificación con diferentes soluciones tampón.

Concentración (mg/ml) Solución tampón GnnA GnnB A 0,98 0,72 B 0,72 0,22

A continuación, se realizó el ensayo de actividad descrito anteriormente utilizando las proteínas GnnA y GnnB purificadas en nuestro laboratorio y se observó que el producto de la reacción con ambas enzimas seguía migrando menos que el sustrato (14C-UDP-GlcNac). Como control del ensayo se incluyó una mezcla de reacción que sólo contenía la enzima GnnB purificada, que también migró menos que el sustrato (Figura 1.11).

66

Capítulo 1 Resultados

Figura 1.11. Cromatografía en capa fina tras el ensayo de actividad con las proteínas GnnA y GnnB purificadas. Imagen representativa de las distintas condiciones probadas. En esta figura se muestra el ensayo realizado a 30ºC durante 4 horas. Se cargaron 2 µl del producto de reacción en una placa de Silica Gel que se expuso a una radiografía de detección de isótopos radiactivos durante 3 días.

Por otra parte, disponíamos en nuestro laboratorio de muestras de GnnA y GnnB cristalizadas por el SSGCID, las cuales también analizamos mediante SDS-PAGE, tinción con azul de Coomassie y Western-Blot (Figura 1.12).

67

Capítulo 1 Resultados

Figura 1.12. Análisis de las proteínas cristalizadas por el SSGCID. A. Tinción con azul de Coomassie de SDS-Page (12%). B. Western Blot con anticuerpo monoclonal anti-histidina. Pesos moleculares: GnnA: 36,84 kDa; GnnB: 41,63 kDa.

Utilizando estas proteínas repetimos el ensayo de actividad. En la cromatografía en capa fina se observó que el producto de reacción que contenía GnnA y GnnB migraba a la misma altura que el sustrato (14C-UDP-GlcNac), pero no fuimos capaces de visualizar la formación de producto (UDP-NGlcNac). Como controles de la reacción se añadieron HEPES y también las proteínas purificadas anteriormente en nuestro laboratorio, las cuales volvieron a formar un producto que migró por debajo del sustrato (Figura 1.13).

Por lo tanto, no se pudo concluir si las ORF BAB1_1617 y BAB1_1616 codifican las proteínas necesarias para la formación de N-acetil-diaminoglucosa en Brucella.

68

Capítulo 1 Resultados

Figura 1.13. Cromatografía en capa fina tras el ensayo de actividad con las proteínas cristalizadas por el SSGCID. Imagen representativa de las distintas condiciones probadas. En esta figura se muestra el ensayo realizado a 37ºC durante 4 horas. Se cargaron 2 µl del producto de reacción en una placa de Silica Gel que se expuso a una radiografía de detección de isótopos radiactivos durante 2 días.

69

Capítulo 1 Discusión

DISCUSIÓN

Los resultados presentados sugieren que los genes responsables de la síntesis del esqueleto de diaminoglucosa en el lípido A de Brucella (gnnA y gnnB), aunque no están anotados, son codificados por BAB1_1617 y BAB1_1616 respectivamente. Ambos mantienen la anotación de oxidorreductasa y transaminasa respectivamente y la estructura tridimensional sugerida para ambas proteínas lo corrobora. La región del cromosoma I en la que se encuentran se trata de una región importante del genoma de Brucella ya que contiene genes del ciclo celular y otros hipotéticamente relacionados con el LPS. Además, el hecho de que no se hayan podido conseguir mutantes en gnnA y gnnB, aunque no lo demuestra directamente, apoya la hipótesis de que podrían estar implicados en la síntesis del lípido A que, como parte estructural del LPS, es fundamental para la estabilidad de la membrana externa y la viabilidad de la bacteria y por lo tanto serían esenciales para Brucella. En la misma línea, es también significativo el hecho de que los intentos de construir mutantes en acpXL o lpxXL, que por homología con otros parientes próximos a Brucella, deberían añadir los ácidos grasos de cadena larga al disacárido de diaminoglucosa del lípido A en Brucella, tampoco han tenido éxito (Conde-Álvarez et al., resultados no publicados), y esto podría indicar que el conjunto formado por dichos ácidos grasos y el disacárido de diaminoglucosa del lípido A son críticos para la estabilidad de la membrana de Brucella. Esta hipótesis se apoya también en un estudio de mutagénesis paralelo en el que se utilizó la metodología Tn-Seq (van Opijnen and Camilli, 2010) basada en el transposón Tn5 y que conlleva la saturación del genoma de B. abortus con una inserción del transposón en un punto cada 3,5 pb de media. Las ORF donde el transposón no se inserta se consideran esenciales para la viabilidad de la bacteria. En dicho estudio, se obtuvo una librería de 3.106 mutantes en B. abortus de entre los posibles 3419 genes anotados, pero ninguno de ellos tenía el transposón insertado en BAB1_1616 o BAB1_1617 (Sternon, 2017; De Bolle, comunicación personal), lo cual apoya de nuevo que estas ORF son esenciales para Brucella.

71

Capítulo 1 Discusión

Por todos estos motivos, se podría afirmar que gnnA y gnnB, dos genes hasta ahora desconocidos en Brucella, están codificados por BAB1_1616 y BAB1_1617 respectivamente. No obstante, para intentar demostrar la hipótesis de que estos genes son esenciales para la viabilidad de Brucella, nos planteamos varias alternativas. En primer lugar, la construcción de un mutante doble en ambos genes, técnica desarrollada con éxito en casos anteriores. Así, no se pudo obtener un mutante simple por deleción en bvrR o en bvrS que codifican un importante regulador de la virulencia de Brucella (Sola-Landa et al., 1998), pero sí un doble mutante por deleción que implica ambos genes (Conde-Álvarez, resultados no publicados). Otra alternativa sería utilizar la técnica de mutagénesis dirigida para cambiar la lisina (K) en posición 191 de GnnB, considerada el residuo catalítico específico de aminotransferasas del piridoxal. Dicha lisina forma un enlace de almidina con este cofactor, que actúa como un sumidero de electrones para estabilizar los intermediarios carbaniónicos formados en las reacciones que involucran a compuestos aminados. Además, se podría realizar un ensayo de complementación cruzada con los genes de Campylobacter; es decir, intentar mutagenizar gnnA y gnnB en una cepa de Brucella en la que se expresen los genes de Campylobacter, o bien complementar las mutaciones en Campylobacter con los genes codificados en el genoma de Brucella. Por otro lado, hay que tener en cuenta que tanto los extractos como con las proteínas purificadas utilizadas en los ensayos enzimáticos tienen fusionada una cola de 6 histidinas en el extremo carboxi-terminal que podría estar afectando a su configuración tridimensional y/o a su función enzimática. Además, habría que verificar que el ensayo de la actividad se realiza correctamente repitiéndolo en paralelo con proteínas GnnA y GnnB purificadas de otras bacterias que tienen un lípido A constituido por diaminoglucosa y cuya función ha podido ser claramente establecida. Como ya se ha comentado en los resultados obtenidos tras realizar el ensayo de actividad con los extractos completos, estos pueden contener otras proteínas que estén interfiriendo en el ensayo. En el gel de SDS-PAGE y Western Blot de dichos extractos se aprecian bandas por debajo del peso molecular esperado que podrían ser productos de degradación de la proteína original y bandas que, por su tamaño, podrían corresponder a formas dímericas.

72

Capítulo 1 Discusión

Asimismo, y como se ha comentado más arriba, la estructura tridimensional de BAB1_1616 (GnnB), sugiere que es una transferasa dependiente de piridoxal-fosfato. Las reacciones de transaminación generalmente requieren dicho cofactor y trabajos previos en otras bacterias han demostrado la actividad de GnnA y GnnB suplementando el medio de expresión de GnnB con piridoxina (Sweet et al., 2004). Aunque en estos trabajos la adición de piridoxina tuvo como objeto facilitar la expresión de la enzima, quizá también contribuyó en la catalización de la reacción. En nuestros ensayos no hemos suplementado el medio con piridoxina, y esto podría influir en la falta de actividad enzimática que observamos.

Otra de las alternativas posibles para confirmar que los extractos y/o las proteínas utilizadas funcionan correctamente y que el ensayo de actividad se está llevando a cabo de manera adecuada, sería medir la formación de NADH. Puesto que GnnA actuaría como una oxidorreductasa, catalizaría la reducción del NAD+ a NADH tomando los electrones del sustrato glucosamina. El pico de absorción en el ultravioleta del NAD+ se encuentra en una longitud de onda de 259 nanómetros (nm) y el del NADH a 339 nm. Esta diferencia en el espectro de absorción entre la forma oxidada y la reducida posibilitaría la medición de la conversión en el ensayo enzimático, midiendo la absorción a 340 nm.

Por último, podríamos expresar las proteínas GnnA y GnnB de Brucella en una cepa de E. coli que ha sido modificada genéticamente para expresar un único tipo de lípido A (Needham et al., 2013). En este contexto, la expresión de GnnA-GnnB permitiría interpretar más fácilmente las modificaciones del lípido A mediante la técnica de Maldi-Toff.

73

Capítulo 1 Referencias

REFERENCIAS

Barbier, T., Collard, F., Zúñiga-Ripa, A., Moriyón, I., Godard, T., Becker, J., et al. (2014). Erythritol feeds the pentose phosphate pathway via three new isomerases leading to D-erythrose-4-phosphate in Brucella. Proc. Natl. Acad. Sci. U. S. A. 111, 17815– 20. doi:10.1073/pnas.1414622111. Basu, S. S., Karbarz, M. J., and Raetz, C. R. H. (2002). Expression cloning and characterization of the C28 acyltransferase of lipid A biosynthesis in Rhizobium leguminosarum. J. Biol. Chem. 277, 28959–71. doi:10.1074/jbc.M204525200. Boon Hinckley, M., Reynolds, C. M., Ribeiro, A. A., McGrath, S. C., Cotter, R. J., Lauw, F. N., et al. (2005). A Leptospira interrogans enzyme with similarity to yeast Ste14p that methylates the 1-phosphate group of lipid A. J. Biol. Chem. 280, 30214–24. doi:10.1074/jbc.M506103200. Casabuono, A. C., Czibener, C., Del Giudice, M. G., Valguarnera, E., Ugalde, J. E., and Couto, A. S. (2017). New features in the lipid A structure of Brucella suis and Brucella abortus lipopolysaccharide. J. Am. Soc. Mass Spectrom. doi:10.1007/s13361-017-1805-x. Conde-Álvarez, R. (2008). A study on the role of phosphatidylcholine and inner lipopolysacharide sections in Brucella virulence. (PhD thesis). Universidad de Navarra. Conde-Álvarez, R., Grilló, M. J., Salcedo, S. P., de Miguel, M. J., Fugier, E., Gorvel, J. P., et al. (2006). Synthesis of phosphatidylcholine, a typical eukaryotic phospholipid, is necessary for full virulence of the intracellular bacterial parasite Brucella abortus. Cell. Microbiol. 8, 1322–1335. doi:10.1111/j.1462-5822.2006.00712.x. Conde-Álvarez, R., Palacios-Chaves, L., Gil-Ramírez, Y., Salvador-Bescós, M., Bárcena- Varela, M., Aragón-Aranda, B., et al. (2018). Identification of lptA, lpxE, and lpxO, three genes involved in the remodeling of Brucella cell envelope. Front. Microbiol. 8, 2657. doi:10.3389/fmicb.2017.02657. Ernst, R. K., Guina, T., and Miller, S. I. (2001). Salmonella typhimurium outer membrane remodeling: role in resistance to host innate immunity. Microbes Infect. 3, 1327– 34. doi:10.1016/S1286-4579(01)01494-0. Ferguson, G. P., Datta, A., Baumgartner, J., Roop, R. M., Carlson, R. W., Walker, G. C., et al. (2004). Similarity to peroxisomal-membrane protein family reveals that Sinorhizobium and Brucella BacA affect lipid-A fatty acids. Proc. Natl. Acad. Sci. U. S. A. 101, 5012–7. doi:10.1073/pnas.0307137101. Freer, E., Rojas, N., Weintraub, A., Lindberg, A. A., and Moreno, E. (1995). Heterogeneity of Brucella abortus lipopolysaccharides. Res. MicrobioL Pads 146, 569–578.

75

Capítulo 1 Referencias

Gibbons, H. S., Lin, S., Cotter, R. J., and Raetz, C. R. (2000). Oxygen requirement for the biosynthesis of the S-2-hydroxymyristate moiety in Salmonella typhimurium lipid A. Function of LpxO, A new Fe2+/alpha-ketoglutarate-dependent dioxygenase homologue. J. Biol. Chem. 275, 32940–9. doi:10.1074/jbc.M005779200. Gil-Ramírez, Y. (2011). Papel de una etanolamina-fosfato transferasa y una glicosil transferasa en la síntesis del lipopolisacárido de Brucella y utilización de una acetil transferasa para la modificación epitópica de la cadena O. (PhD thesis). Universidad de Navarra. Gil-Ramírez, Y., Conde-Álvarez, R., Palacios-Chaves, L., Zúñiga-Ripa, A., Grilló, M.-J., Arce-Gorvel, V., et al. (2014). The identification of wadB, a new glycosyltransferase gene, confirms the branched structure and the role in virulence of the lipopolysaccharide core of Brucella abortus. Microb. Pathog. 73, 53–9. doi:10.1016/j.micpath.2014.06.002. Gruber, A., Mancek, M., Wagner, H., Kirschning, C. J., and Jerala, R. (2004). Structural model of MD-2 and functional role of its basic amino acid clusters involved in cellular lipopolysaccharide recognition. J. Biol. Chem. 279, 28475–82. doi:10.1074/jbc.M400993200. Iriarte, M., González, D., Delrue, R.-M., Monreal, D., Conde, R., López-Goñi, I., et al. (2004). “Brucella Lipopolysaccharide: Structure, biosynthesis and genetics,” in Brucella. Molecular and Cellular Biology (Wymondham, England: Horizon Bioscience), 159–192. Karbarz, M. J., Kalb, S. R., Cotter, R. J., and Raetz, C. R. H. (2003). Expression cloning and biochemical characterization of a Rhizobium leguminosarum lipid A 1- phosphatase. J. Biol. Chem. 278, 39269–79. doi:10.1074/jbc.M305830200. Karbarz, M. J., Six, D. A., and Raetz, C. R. H. (2009). Purification and characterization of the lipid A 1-phosphatase LpxE of Rhizobium leguminosarum. J. Biol. Chem. 284, 414–25. doi:10.1074/jbc.M808390200. Lapaque, N., Moriyón, I., Moreno, E., and Gorvel, J.-P. (2005). Brucella lipopolysaccharide acts as a virulence factor. Curr. Opin. Microbiol. 8, 60–66. doi:10.1016/j.mib.2004.12.003. Lapaque, N., Takeuchi, O., Corrales, F., Akira, S., Moriyón, I., Howard, J. C., et al. (2006). Differential inductions of TNF-alpha and IGTP, IIGP by structurally diverse classic and non-classic lipopolysaccharides. Cell. Microbiol. 8, 401–413. doi:10.1111/j.1462-5822.2005.00629.x. Llobet, E., Martínez-Moliner, V., Moranta, D., Dahlström, K. M., Regueiro, V., Tomás, A., et al. (2015). Deciphering tissue-induced Klebsiella pneumoniae lipid A structure. Proc. Natl. Acad. Sci. U. S. A. 112, E6369-78. doi:10.1073/pnas.1508820112. Lowry, O. H., ROSEBROUGH, N. J., Lewis Farr, A., and Randall, R. J. (1951). Protein measurement with the Folin phenol reagent. Readings 193, 265–275. doi:10.1016/0304-3894(92)87011-4.

76

Capítulo 1 Referencias

Manterola, L., Moriyón, I., Moreno, E., Sola-Landa, A., Weiss, D. S., Koch, M. H. J., et al. (2005). The lipopolysaccharide of Brucella abortus BvrS/BvrR mutants contains lipid A modifications and has higher affinity for bactericidal cationic peptides. J. Bacteriol. 187, 5631–9. doi:10.1128/JB.187.16.5631-5639.2005. Markwell, M. A. K., Haas, S. M., Bieber, L. L., and Tolbert, N. E. (1978). A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal. Biochem. 87, 206–210. doi:10.1016/0003- 2697(78)90586-9. Moreno, E., Stackebrandt, E., Dorsch, M., Wolters, J., Busch, M., and Mayer, H. (1990). Brucella abortus 16S rRNA and lipid A reveal a phylogenetic relationship with members of the alpha-2 subdivision of the class Proteobacteria. J. Bacteriol. 172, 3569–76. Needham, B. D., Carroll, S. M., Giles, D. K., Georgiou, G., Whiteley, M., and Trent, M. S. (2013). Modulating the innate immune response by combinatorial engineering of endotoxin. Proc. Natl. Acad. Sci. U. S. A. 110, 1464–9. doi:10.1073/pnas.1218080110. Needham, B. D., and Trent, M. S. (2013). Fortifying the barrier: the impact of lipid A remodelling on bacterial pathogenesis. Nat. Rev. Microbiol. 11, 467–81. doi:10.1038/nrmicro3047. Palacios-Chaves, L., Conde-Álvarez, R., Gil-Ramírez, Y., Zúñiga-Ripa, A., Barquero-Calvo, E., Chacón-Díaz, C., et al. (2011). Brucella abortus ornithine lipids are dispensable outer membrane components devoid of a marked pathogen-associated molecular pattern. PLoS One 6, e16030. doi:10.1371/journal.pone.0016030. Palacios-Chaves, L., Zúñiga-Ripa, A., Gutierrez, A., Gil-Ramírez, Y., Conde-Álvarez, R., Moriyón, I., et al. (2012). Identification and functional analysis of the cyclopropane fatty acid synthase of Brucella abortus. Microbiology 158, 1037– 1044. doi:10.1099/mic.0.055897-0. Qureshi, N., Takayama, K., Seydel, U., Wang, R., Cotter, R. J., Agrawal, P. K., et al. (1994). Structural analysis of the lipid A derived from the lipopolysaccharide of Brucella abortus. J. Endotoxin Res. 1, 137–148. doi:10.1177/096805199400100303. Raetz, C. R. H., Reynolds, C. M., Trent, M. S., and Bishop, R. E. (2007). Lipid A modification systems in gram-negative bacteria. Annu. Rev. Biochem. 76, 295– 329. doi:10.1146/annurev.biochem.76.010307.145803. Raetz, C. R. H., and Whitfield, C. (2002). Lipopolysaccharide endotoxins. Annu. Rev. Biochem. 71, 635–700. doi:10.1146/annurev.biochem.71.110601.135414. Rojas, N., Freer, E., Weintraub, A., Ramirez, M., Lind, S., and Moreno, E. (1994). Immunochemical identification of Brucella abortus lipopolysaccharide epitopes. Clinical and diagnostic laboratory immunology. 1, 206–213. Sangari, F., and Agüero, J. (1991). Mutagenesis of Brucella abortus: comparative efficiency of three transposon delivery systems. Microb. Pathog. 11, 443–6.

77

Capítulo 1 Referencias

Scupham, A. J., and Triplett, E. W. (1997). Isolation and characterization of the UDP- glucose 4′-epimerase-encoding gene, galE, from Brucella abortus 2308. Gene 202, 53–59. doi:10.1016/S0378-1119(97)00453-8. Simon, R., Priefer, U., and Pühler, A. (1983). A broad host range mobilization system for in vivo genetic engineering: Transposon mutagenesis in gram negative bacteria. Bio/Technology 1, 784–791. doi:10.1038/nbt1183-784. Sola-Landa, A., Pizarro-Cerda, J., Grillo, M.-J., Moreno, E., Moriyón, I., Blasco, J.-M., et al. (1998). A two-component regulatory system playing a critical role in plant pathogens and endosymbionts is present in Brucella abortus and controls cell invasion and virulence. Mol. Microbiol. 29, 125–138. doi:10.1046/j.1365- 2958.1998.00913.x. Soler-Lloréns, P., Gil-Ramírez, Y., Zabalza-Baranguá, A., Iriarte, M., Conde-Álvarez, R., Zúñiga-Ripa, A., et al. (2014). Mutants in the lipopolysaccharide of Brucella ovis are attenuated and protect against B. ovis infection in mice. Vet. Res. 45, 72. doi:10.1186/s13567-014-0072-0. Sternon, J.-F. (2017). Tn-seq on the pathogen Brucella abortus uncovers essential functions for culture and critical pathways for macrophages infection including pyrimidines biosynthesis. (PhD thesis). Université de Namur. Sweet, C. R., Ribeiro, A. A., and Raetz, C. R. H. (2004). Oxidation and transamination of the 3″-position of UDP- N -acetylglucosamine by enzymes from Acidithiobacillus ferrooxidans. J. Biol. Chem. 279, 25400–25410. doi:10.1074/jbc.M400596200. van Mourik, A., Steeghs, L., van Laar, J., Meiring, H. D., Hamstra, H.-J., van Putten, J. P. M., et al. (2010). Altered linkage of hydroxyacyl chains in lipid A of Campylobacter jejuni reduces TLR4 activation and antimicrobial resistance. J. Biol. Chem. 285, 15828–36. doi:10.1074/jbc.M110.102061. van Opijnen, T., and Camilli, A. (2010). Genome-wide fitness and genetic interactions determined by Tn-seq, a high-throughput massively parallel sequencing method for microorganisms. Curr. Protoc. Microbiol. Chapter 1, Unit1E.3. doi:10.1002/9780471729259.mc01e03s19. Velasco, J., Bengoechea, J. A., Brandenburg, K., Lindner, B., Seydel, U., González, D., et al. (2000). Brucella abortus and its closest phylogenetic relative, Ochrobactrum spp., differ in outer membrane permeability and cationic peptide resistance. Infect. Immun. 68, 3210–3218. doi:10.1128/IAI.68.6.3210-3218.2000. Zähringer, U., Knirel, Y. A., Lindner, B., Helbig, J. H., Sonesson, A., Marre, R., et al. (1995). The lipopolysaccharide of Legionella pneumophila serogroup 1 (strain Philadelphia 1): chemical structure and biological significance. Prog. Clin. Biol. Res. 392, 113–39. Zúñiga-Ripa, A., Barbier, T., Conde-Álvarez, R., Martínez-Gómez, E., Palacios-Chaves, L., Gil-Ramírez, Y., et al. (2014). Brucella abortus depends on pyruvate phosphate dikinase and malic enzyme but not on Fbp and GlpX fructose-1,6-bisphosphatases for full virulence in laboratory models. J. Bacteriol. 196, 3045–57. doi:10.1128/JB.01663-14.

78

Capítulo 1 Material suplementario

Material suplementario

81

Capítulo 1 Material suplementario

5` TGGCGGCCGCTCTAG gnnA-F1 3` Inicio BAB1_1618 CCAATCGGCGTACAAATCGTTGTTCCTGCA

AGGAATGGCGGCGGTATATGTGCGGAACGAGACCCTGTCAAGCGTTCTTCGCATAGGTTT Fin BAB1_1618 CATCTTGTTTTCTATCATTTGCACCGATATGGAACTTGGCATCAGCAAAGGCTAAGTTTC

5` gnnA-R2 Inicio gnnA CCGAGA 1 ATGGCACCTCGTATCGCGGTTTTGGGTTGCGGCTATTGGGGCGGCAATCACATTCGTACC M A P R I A V L G C G Y W G G N H I R T 20 3` CTCTTCAGGGTACG 61 CTGAAGAGTCTCGGCGCCCTTCAGGCAGTTTCAGACTCCAACGCGGAAAAAGCGGAACGA L K S L G A L Q A V S D S N A E K A E R 40

121 TTCGCTTCCGAATTCAACGTTCCCAGCATTCCAGTCGAGGAACTCTTCATCCATCCGGAT F A S E F N V P S I P V E E L F I H P D 60

181 ATTGATGGCATCGTGCTGGCGCTGCCGCCGCAGCTTCATGCGCAATATGCCATGGAAGCG I D G I V L A L P P Q L H A Q Y A M E A 80

241 GTGAAGAACGGCAAGGATGTTCTGGTTGAAAAGCCGATTGCGCTCGATATTCCCGTAGCG V K N G K D V L V E K P I A L D I P V A 100

301 CTTGCCGAAGTTGAGGCCGCGCGCGAAAATGGCCGCGTGTTCATGGTCGGCCATGTGCTC L A E V E A A R E N G R V F M V G H V L 120

361 CGCTTCCATCCGGCTTTTGAAAAGCTTCTCGACATGGTGCAAAGCGGCGAACTTGGCGAT R F H P A F E K L L D M V Q S G E L G D 140

421 ATCCGCTATGTGCATTCGCACCGCGTCGGGCTTGGCAAGTTTCATACCGAATTCGATGCG I R Y V H S H R V G L G K F H T E F D A 160

481 CTATGGGATTTCGCGCCGCACGATCTTTCGATGATCCTTGCCATCACCGGCGAGGAGCCG L W D F A P H D L S M I L A I T G E E P 180

541 AACGTGGTGCGCGGCGAAGGCGTTGCTATTCTCGATCATCTGAACGATTTCGCGCATCTT N V V R G E G V A I L D H L N D F A H L 200

601 CACATGGAATTTCCGAGCGGAATTCGCGGCCATCTTTTTGCTTCGCGTCTGAACGCTTAT H M E F P S G I R G H L F A S R L N A Y 220

661 CGCGAACGTCGCCTCAGTGTCACCGGCACCAAGGGCATGGCTGTTTTCGACGATGGCGAA R E R R L S V T G T K G M A V F D D G E 240

721 GCCTGGGAGCGGAAGCTTGCCCTTTACAGGCATGAGGTCTGGCGGGAAAACGACCGCTGG A W E R K L A L Y R H E V W R E N D R W 260

781 GCTTTCAAGTCCGCCGATCCGGTCTATATACAAACCGAAGAAGGTATGCCGCTGACGCGC A F K S A D P V Y I Q T E E G M P L T R 280

83

Capítulo 1 Material suplementario

5´ gnnA-R2 3´ CGTACCCTGAAGAGTCTCGG

841 GAATTGCAGCATTTCATGCATTGCATCGAAACGCGTGAAACGCCGCGAACCGACGGCAAG E L Q H F M H C I E T R E T P R T D G K 300 5` gnnA-F3 3` Fin gnnA 901 GAAGCGATCAGTGTCCTGCGCATTCTCACCGAAGGCACGGTGCGGCATCCGGCCTGAAGC E A I S V L R I L T E G T V R H P A * 318

AGCAGTGCGCAAGCAGCGTAAAATTGTGTAAAGCGGTGTTACAGCACTCGCAAAACTTTG

AACGCTCCGGCCATGGCAGGGCGGGGCGCAAGCGAATAAATAGATGGCCGAGGATTCGCA

CCGGTTTTCGTGCGGGCCAGATATAT ATATATCTGGCCCGCACGAA gnnA-R4 3`

ATCCACTAGTTCTAG 5`

Figura S1.1. Secuencia de BAB1_1617 (gnnA) y sus regiones corriente arriba y abajo. Los codones de inicio y final están marcados en verde y rojo respectivamente. Las letras en azul indican nucleótidos intergénicos. Los oligonucleótidos utilizados para la mutagénesis están en negrita. Los oligonucleótidos F1 y R4 portan en su extremo 5` una secuencia complementaria al vector pJQK para poder ser clonados mediante el sistema InFusion. Los sitios de corte de la enzima de restricción XbaI se encuentran subrayados. Los caracteres en gris indican los aminoácidos delecionados en el mutante gnnA.

84

Capítulo 1 Material suplementario

5´ TGGCGGCCGCTCTAG gnnB-F1 3` GCAAGCAGCGTAAAATTGTGTAAAGCGG

TGTTACAGCACTCGCAAAACTTTGAACGCTCCGGCCATGGCAGGGCGGGGCGCAAGCGAA

TAAATAGATGGCCGAGGATTCGCACCGGTTTTCGTGCGGGCCAGATATATTGGAGCCAAC

Inicio gnnB 1 ATGCAGTTCATTGATCTTGGAGCGCAGCGCGCGCGTATCGAAAATCGTCTCAATGCCGCC M Q F I D L G A Q R A R I E N R L N A A 20

5´ gnnB-R2 3´ AAAGGTTCCAACAACGCCTT 61 ATTTCCAAGGTTGTTGCGGAAGGCCGTTATATTCTTGGGCCGGAAGTGGCTGAATTTGAA I S K V V A E G R Y I L G P E V A E F E 40

121 AAAAAGCTTGGCGAATATCTGGGTGTGGAGCATGTCATCGCCTGTGCCAATGGCACTGAC K K L G E Y L G V E H V I A C A N G T D 60

181 GCCTTGCAGATGCCCCTGATGACACGCGGCATCGGGCCGGGCCATGCGGTTTTTGTTCCG A L Q M P L M T R G I G P G H A V F V P 80

241 TCCTTCACCTTTGCAGCGACCGCTGAGGTTGTTGCGCTCGTCGGTGCAGAGCCGGTATTC S F T F A A T A E V V A L V G A E P V F 100

301 GTTGATGTCGATCCAGATAGCTACAACATGAATGTCGAGCAGCTCGAAGCTGCCATTGCC V D V D P D S Y N M N V E Q L E A A I A 120

361 GCCACCATCAAGGAAGGTCGTCTGGAGCCGAAGGCGATCATTCCGGTCGATCTTTTTGGC A T I K E G R L E P K A I I P V D L F G 140

421 CTTGCCGCCTCTTACAATCGCATCACCGCCATTGCCGAGCGCGAAGGCCTGTTCATCATC L A A S Y N R I T A I A E R E G L F I I 160

481 GAGGATGCGGCCCAGTCCATCGGCGGCAAGCGCGACAATGTCATGTGCGGCGCTTTCGGC E D A A Q S I G G K R D N V M C G A F G 180

541 CATGTGGGCGCAACCAGCTTCTATCCCGCCAAGCCGCTTGGCTGCTACGGTGACGGTGGC H V G A T S F Y P A K P L G C Y G D G G 200

601 GCCATGTTCACCAATGATGCAGAATTGGCAGATACACTGCGCTCCGTGCTGTTCCACGGC A M F T N D A E L A D T L R S V L F H G 220

661 AAGGGCGAGACGCAGTATGACAATGTCCGCATCGGCATCAATTCTCGCCTGGATACGATA K G E T Q Y D N V R I G I N S R L D T I 240

721 CAGGCGGCTGTTCTTCTGGAAAAGCTTGCCATTCTCGAAGATGAGATGGAAGCCCGCGAC Q A A V L L E K L A I L E D E M E A R D 260

781 CGCATTGCAAGGCGCTATAATGAGGCGCTGAAGGATGTGGTGAAGGTGCCGGAACTTCCA R I A R R Y N E A L K D V V K V P E L P 280

841 GCCGGCAACCGCTCGGCCTGGGCGCAATATTCTATCGAAAGCGAGAACCGCGACGGCTTG A G N R S A W A Q Y S I E S E N R D G L 300

85

Capítulo 1 Material suplementario

901 AAGGCACAACTTCAGGCAGAAGGTATTCCGTCCGTCATCTATTATGTGAAGCCTCTGCAC K A Q L Q A E G I P S V I Y Y V K P L H 320

5´ gnnB-R2 3´ TTTCCAAGGTTGTTGCGGAA

5´ gnnB-F3 3´ 961 TTGCAGACGGCCTATAAGCACTATTCCGTTGCCCCCGGTGGCCTGCCGGTTTCGGAAAGC L Q T A Y K H Y S V A P G G L P V S E S 340

1021 CTGCCATCACGCATTCTGAGCCTGCCGATGCATCCTTACCTTTCGGAAGCTGACCAGGAC L P S R I L S L P M H P Y L S E A D Q D 360 Fin gnnB 1081 AAGATTATCGGCGTGATCCGTGGCTTCCACGGCAAGAAGGCCTGAATTGGCTGAATTATA K I I G V I R G F H G K K A * 374

TGAAGTGATCAAAACCCGGCGAAAGCCGGGTTTTGTTTATGCTCGCAAATTCAGTTGCCA

GGATGCACTCAGCCGATCATGGCGCGGCGCTGGTTGGCCGCGCGTTCCATGACTTCCCGC

GGCGCCTTGACGCCATCGCGCTCTTCCAGTGCTTCGGCCTTGGAAAGTTCGGTTTCGGC AACCTTTCAAGCCAAAGCCG gnnB-R4 3` ATCCACTAGTTCTAG 5´

Figura S1.2. Secuencia de BAB1_1616 (gnnB) y sus regiones corriente arriba y abajo. Los codones de inicio y final están marcados en verde y rojo respectivamente. Las letras en azul indican nucleótidos intergénicos. Los oligonucleótidos utilizados para la mutagénesis están en negrita. Los oligonucleótidos F1 y R4 portan en su extremo 5` una secuencia complementaria al vector pJQK para poder ser clonados mediante el sistema InFusion. Los sitios de corte de la enzima de restricción XbaI se encuentran subrayados. Los caracteres en gris indican los aminoácidos delecionados en el mutante gnnB.

86

Capítulo 1 Material suplementario

A. ferrooxidans: lpxA-gnnA-gnnB-lpxB B. melitensis (cromosoma I): gnnA-gnnB, separados de lpxA-prot.*-lpxB B. melitensis biovar suis: gnnA-gnnB, separados de lpxA-prot.*-lpxB C. jejuni: gnnB-gnnA L. interrogans: ninguno M. loti: lpxA-prot.*-lpxB

* prot.: proteína de función desconocida

Figura S1.3. Organización genómica en torno a la región que comprende los genes gnnA y gnnB en Campylobacter jejuni y Brucella abortus (panel superior). gnnA y gnnB de Campylobacter jejuni subespecie jejuni (NCTC 11168 = ATCC 700819; códigos GenBank Cj0504c y Cj0505c respectivamente). Organización de gnnA y gnnB en otras especies (Acidithiobacillus ferrooxidans, Brucella melitensis, Brucella melitensis biovar suis, Campylobacter jejuni, Leptospira interrogans y Mesorhizobium loti) según el modelo predicho por Sweet y colaboradores (Sweet et al., 2004) (panel inferior).

87

Chapter 2

Chapter 2 WadD, a new Brucella lipopolysaccharide core glycosyltranfserase identified by genomic search and phenotypic characterization 1, 2

1 The studies presented from BAB1_1620 were performed during Yolanda Gil-Ramírez PhD studies. 2 The results presented in this chapter have been accepted for publication in Frontiers in Microbiology (manuscript ID 409514) (see Appendix).

91

Chapter 2 Abbreviations

Abbreviations used in Chapter 2 aa aminoacids ADP Adenosine diphosphate amp ampicillin Ba Brucella abortus bp base pairs BHI Brain Heart Infusion BLAST Basic Local Alignment Search Tool BSA Bovine Serum Albumin BSL-3 Biosafety Level 3 Cazy Carbohydrate-Active Enzymes CD14 Cluster of Differentiation 14 CFU Colony Forming Units CIMA Centro de Investigación Médica Aplicada cm chloramphenicol DNA Deoxyribonucleic acid ELISA Enzyme Linked Immunosorbent Assay EMBL-EBI European Molecular Biology Laboratory - European Bioinformatics Institute EMBOSS European Molecular Biology Open Software Suite g grams GDP Guanosine diphosphate glc glucose glcN glucosamine glcN3N4P 2,3-diaminoglucose-4-phosphate glcN3N1P 2,3-diaminoglucose-1-phosphate GT Glycosyltransferase HCl chlorhydric acid IFN-γ Interferon-gamma IL Interleukin INRA French National Institute for Agricultural Research IP Intraperitoneal kb kilobase Kdo 3-deoxy-D-manno-2-octulosonic acid KEGG Kyoto Encyclopedia of Genes and Genomes km kanamycin LOS Lipooligosaccharide LPS Lipopolysaccharide M Molar mA milliamps man mannose 93

Chapter 2 Abbreviations

MIC Minimal Inhibitory Concentration mg milligrams ml millilitres mM milliMolar MoAbs Monoclonal Antibodies n sample size (statistics) nal nalidixic acid nm nanometres NMR Nuclear Magnetic Resonance nt nucleotides O.D. Optical Density OMP Outer Membrane Protein O-PS O-polysaccharide ORF Open Reading Frame PAGE Polyacrylamide Gel Electrophoresis PAMP Pathogen-Associated Molecular Pattern PBS Phosphate-Buffered Saline PCR Polymerase Chain Reaction PI Post-infection pmx polymixin B PRR Pattern Recognition Receptor PVDF Polyvinylidene fluoride R Rough RD from Spanish Real Decreto R-Brucella Rough Brucella R-LPS Rough Lipopolysaccharide RNA Ribonucleic acid rpm revolutions per minute sac sucrose SD Standard Deviation SDS Sodium Dodecyl Sulphate S-Brucella Smooth Brucella S-LPS Smooth lipopolysaccharide spp. species TLR4 Toll-Like Receptor 4 TNF-α Tumour Necrosis Factor-alpha TSA Tryptic Soy Agar TSB Tryptic soy broth UDP Uridine diphosphate V volts vol volume

94

Chapter 2 Abbreviations

μl microlitres μm micrometres Δ deletion :: insertion

95

Chapter 2 Abstract

ABSTRACT

Brucellosis, an infectious disease caused by Brucella, is one of the most extended bacterial zoonosis in the world and an important cause of economic losses and human suffering. The lipopolysaccharide (LPS) of Brucella plays a major role in virulence as impairs normal recognition by the innate immune system and delays the immune response. The LPS core is a branched structure involved in resistance to complement and polycationic peptides and mutants in glycosyltransferases required for the synthesis of the lateral branch not linked to the O-polysaccharide (O-PS) are attenuated and have been proposed as good vaccine candidates against brucellosis. For this reason, the complete understanding of the genes involved in the synthesis of this LPS section is of particular interest. The chemical structure of the Brucella LPS core suggests that, in addition to the already identified WadB and WadC glycosyltransferases, others could be implicated in the synthesis of this lateral branch. To clarify this point, we identified and constructed mutants in 11 ORFs encoding putative glycosyltransferases in B. abortus.

Four of them, regulated by the virulence regulator MucR, involved in LPS synthesis, or BvrR/BvrS, implicated in the synthesis of surface components, were not required for the synthesis of a complete LPS, interaction with polycationic peptides and/or complement or virulence. Among the other ORFs, six seemed not to be required for synthesis of the core LPS since mutants in these ORFs kept the O-PS and reacted as the wild type with polyclonal sera. Interestingly, mutant in ORF BAB1_0953 (renamed wadD) lost reactivity against antibodies that recognize the core section while kept the O-PS. This suggests that WadD is a new glycosyltransferase adding one or more sugars to the core lateral branch. WadD mutants were more sensitive than the parental strain to components of the innate immune system and played a role in chronic stages of infection. These results corroborate and extend previous work indicating that the Brucella LPS core is a branched structure that constitutes a steric impairment preventing the elements of the innate immune system to fight against Brucella.

97

Chapter 2 Introduction

INTRODUCTION

Brucella LPS diverges from the classical LPS of other gram-negative bacteria, and this accounts for its very low endotoxicity and its important role in virulence. Brucella LPS diverges from that of other gram-negative bacteria, mainly in the lipid A (see Chapter 1) and core sections, that account for its very low endotoxicity and its important role in virulence (Lapaque et al., 2005 and General Introduction). In contrast to what happens in most gram-negative bacteria, the core section of Brucella LPS is a branched structure that lacks negatively charged groups, galacturonate, heptoses or orthophosphate, common in other α-2 Proteobacteria. The core LPS of B. melitensis (Figure 2.1) is a branched oligosaccharide built of lipid A- linked 3-deoxy-D-manno-2-octulosonic acid (Kdo), glucose, 2-amino-2,6-dideoxy-D- glucose (quinovosamine), mannose, and 2-amino-2-deoxy-D-glucose (glucosamine [GlcN]) (Conde-Álvarez et al., 2012; Fontana et al., 2016; Gil-Ramírez et al., 2014; Iriarte et al., 2004; Kubler-Kielb and Vinogradov, 2013). This structure accounts for the Brucella LPS core overlapping epitopes (Rojas et al., 1994) an inner one comprising the Kdo residues plus the glucose bridging KdoII with the O-PS and an outer epitope encompassing the mannose and GlcN residues (Fontana et al., 2016; González et al., 2008; Iriarte et al., 2004). Accordingly, the reactivity with R-LPS-specific monoclonal antibodies (MoAbs) strongly suggests that the structure elucidated for B. melitensis is conserved in the classical species (Bowden et al., 1995; Zygmunt et al., 2012). Moreover, availability of the corresponding structure of several mutants has also allowed assigning genes that upon mutation generate LPSs that lack (i.e. rough [R] LPS) or carry O-PS (Figure 2.1).

99

Chapter 2 Introduction

Figure 2.1. Schematic representation of Brucella LPS core section (adapted from Fontana et al., 2016). Kdo (3-deoxy-d-manno-octulosonic acid), Glc (glucose), Man (mannose), GlcN (glucosamine).

The structure of the Brucella core LPS is directly related to its virulence and resistance to the innate immune system. The absence of negative charges helps to explain Brucella marked resistance to bactericidal peptides (Martínez de Tejada et al., 1995) and the branched structure of the core itself creates a steric impairment preventing the binding of elements of the innate immune system and is crucial for Brucella virulence (Conde-Álvarez et al., 2012; Fontana et al., 2016; Gil-Ramírez et al., 2014; Soler-Lloréns et al., 2014). The first step in core biosynthesis (Figure 2.1) is the addition of Kdo to the lipid A disaccharide backbone by the Kdo transferase WaaA (KdtA), and, as expected, all

100

Chapter 2 Introduction

Brucella genomes carry a waaA homologue (Iriarte et al., 2004; Raetz and Whitfield, 2002). When the lipid A is fully acylated, the rest of core sugars are assembled. Genomic comparison with other enterobacteria indicates that Brucella contains two enzymes for Kdo formation: BMEI0850 in B. melitensis (or its orthologue BAB1_1156 in B. abortus) would encode the Kdo-8P-synthase that condenses arabinose-5P and phosphoenolpyruvate to yield Kdo-8P (homologous to KdsA), and BMEI1904 (BAB1_0035) would be the orthologue of KdsB, the Kdo-cytidyl-transferase (Goldman and Kohlbrenner, 1985). The product of BMEII1029 (BAB2_0209) would mediate transfer of the two Kdo residues to the lipid A (Iriarte et al., 2004). The genes involved in the synthesis of the core region that links the O-PS (pgm, wadA, manBcore and manCcore) have been identified by mutagenesis and chemical analysis. A mutation in any of them results in the loss of the complete O-PS and rough (R) LPSs can be classified in three phenotypes according to its apparent molecular mass. Gene pgm (BMEI1886/BAB1_0055) encodes a phosphoglucomutase necessary for the synthesis of ADP-glucose, UDP-glucose and UDP-galactose, the donors of glucose or galactose. Mutants in pgm lack O-PS and have a truncated core (R2- phenotype) with an electrophoretic mobility intermediate between that of the per mutants, that lack the O-PS but maintain a complete core (R1 phenotype), and that of the mutants with a completely truncated core (R3 phenotype) (Ugalde et al., 2000). Gene wadA (BMEI1326/BAB1_0639) corresponds to the enzyme linking KdoII and glucose. Mutants in wadA produce a rough LPS that lacks the inner epitope of the core and maintains the outer core epitope (González et al., 2008; Monreal et al., 2003). manBcore and manCcore (BMEII0899/BAB2_0855 and BMEII0900/BAB2_0856 respectively), are responsible for GDP-mannose synthesis (Allen et al., 1998; Monreal et al., 2003). Mutants in manBcore present a deep rough R3-phenotype with a core composed of the pentasaccharide β-D-Glcp-(14)-α-Kdop-(24)-α-Kdop-(26)-β-D-

GlcpN3N4P-(16)-α-D-GlcpN3N1P (Fontana et al., 2016). Up to now, two genes encoding glycosyltransferases (wadC and wadB) have been shown to be involved in the synthesis of the core lateral branch that forms the pentasaccharide β-D-GlcpN-(16)-β-D-GlcpN-(14)-[β-D-GlcpN-(16)]-β-D-GlcpN-

(13)-α-D-Manp-(15) in B. melitensis (Conde-Álvarez et al., 2012; Fontana et al., 2016; Gil-Ramírez et al., 2014) (Figure 2.1). wadC (BMEI0509/BAB1_1522) encodes the

101

Chapter 2 Introduction mannosyltransferase adding a mannose to KdoI, that is the depart of the lateral branch (Conde-Álvarez et al., 2012; Fontana et al., 2016; Kubler-Kielb and Vinogradov, 2013) and wadB (BMEI1602/BAB1_0351) encodes a glucosaminyltransferase involved in the assembly of the GlcN branch (Conde-Álvarez et al., 2012; Fontana et al., 2016; Gil- Ramírez et al., 2014; González et al., 2008; Iriarte et al., 2004; Kubler-Kielb and Vinogradov, 2013). Consistent with the anti-R-LPS MoAbs reactivity, these genes are highly conserved in the classical Brucella species (Bowden et al., 1995; Zygmunt et al., 2012). However, since most but not all glycosyltransferases involved in LPS synthesis are monofunctional (Raetz and Whitfield, 2002), it remains to be determined whether glucosaminyltransferases other than WadB are required for the synthesis of the GlcN branch. Based on the complete structure of the core and the phenotype of mutants in wadB and wadC, it is postulated that the lack of acidic groups other than the two Kdo and lipid A phosphates and the mannose-GlcN branch account for the role of Brucella core in virulence. By virtue of the density of amino groups and close position to the inner core and lipid A, the GlcN tetrasaccharide both neutralizes and sterically protects those inner anionic groups, thereby hampering binding of bactericidal peptides and PRRs (Pattern Recognition Receptors) such as the activators of the antibody- independent classical complement pathway and MD2, the TLR4 co-receptor. Accordingly, core defects bolster proinflammatory responses causing an activation of innate immunity earlier than that of the wild-type, thereby generating attenuation (Conde-Álvarez et al., 2012; Fontana et al., 2016; Gil-Ramírez et al., 2014; Soler-Lloréns et al., 2014). Also, although both wadB and wadC mutants maintain an intact O-PS, attenuation is more severe for the latter (Conde-Álvarez et al., 2012; Fontana et al., 2016; Gil-Ramírez et al., 2014; Iriarte et al., 2004; Kubler-Kielb and Vinogradov, 2013) strongly suggesting a correlation between the extent of core damage and the intensity of the immunoactivation that brings about attenuation. A complete elucidation of the genetics of Brucella LPS core could confirm such a correlation and, since LPS core mutants represent a tool for developing a new generation of brucellosis vaccines (Conde-Álvarez et al., 2013; Zhao et al., 2017), also provide a graded array of possibilities. With these possibilities in mind, we investigated all Brucella genes annotated as glycosyltransferases for their possible involvement in LPS core synthesis

102

Chapter 2 Introduction and relevant biological effects.

Moreover, it has also been reported that mutants in the general virulence regulator MucR (both in B. abortus and B. melitensis) have a defect in the core LPS, but the glycosyltransferases responsible for this defect have not been identified (Caswell et al., 2013; Mirabella et al., 2013).

In this work, we made a genomic search to continue the study of glycosyltransferases involved in the synthesis of the Brucella core.

103

Chapter 2 Experimental procedures Results

EXPERIMENTAL PROCEDURES

Bacterial strains and growth conditions. The bacterial strains, plasmids and oligonucleotides used are listed in Tables 2.1, 2.2 and Table S2.1 respectively. All bacteria were grown either on tryptic soy agar (TSA, Pronadisa) plates or in tryptic soy (TSB, Scharlau) or Mueller-Hinton (Becton Dickinson, Difco) broths at 37ºC. Where indicated, growth media were supplemented with kanamycin (Km) at 50 mg/ml, nalidixic acid (Nal) at 25 mg/ml, ampicillin (Amp) at 100 mg/ml and/or 5% sucrose. Bacterial growth rates were determined at 37ºC in Mueller- Hinton broth using a Bioscreen C apparatus (Lab Systems). All strains were stored in skim milk at –80ºC. Work with Brucella was performed at the Biosafety Level 3 (BSL-3) laboratory facilities of the Universidad de Navarra and “Centro de Investigación Médica Aplicada” (CIMA), Universidad de Navarra, Spain.

DNA manipulations and analyses. Sequence data were obtained from Kyoto Encyclopedia of Genes and Genomes (KEGG1). Searches for DNA and protein homologies between Brucella species and other α-proteobacteria such as Ochrobactrum, Rhizobium or Agrobacterium were carried out using KEGG, Basic Local Alignment Sequence Tool (BLAST2) and Clustal Omega3 from the European Molecular Biology Laboratory-European Bioinformatics Institute (EMBL- EBI4). New glycosyltransferase identification was supported by Carbohydrate-Active enZymes database (CAZy5). Primers were designed using Primer 3 input6 and synthesized by Sigma-Aldrich. Plasmid DNA was extracted with Qiaprep spin Miniprep (Qiagen GmbH). When needed, DNA was purified from agarose gels using Qiack Gel extraction kit (Qiagen) and sequenced by the Servicio de Secuenciación of CIMA. 1 http://www.genome.jp/kegg/ 2 http://blast.ncbi.nlm.nih.gov/Blast.cgi 3 http://www.ebi.ac.uk/Tools/msa/clustalo 4 http://www.ebi.ac.uk/ 5 http://www.cazy.org 6 http://bioinfo.ut.ee/primer3-0.4.0/

105

Chapter 2 Experimental procedures Results

Table 2.1. Bacterial strains.

Strain Characteristics Reference Brucella abortus Ba-parental B. abortus wild type, virulent biotipe 1, smooth Sangari and LPS, NalR spontaneous mutant of strain 2308 Agüero, 1991

Ba::pSKoriT-BAB1_0326 Ba - parental insertion mutant in BAB1_0326::261-383 This work

Ba::pSKoriT-BAB1_0417 Ba-parental insertion mutant in BAB1_0417::37-134 This work

Ba::pSKoriT-BAB1_0953 Ba - parental insertion mutant in BAB1_0953::112-230 This work

Ba::pJQK-BAB1_0932 Ba-parental insertion mutant in BAB1_0932::354-493 Thi s work

Ba::pJQK-BAB1_0114 Ba-parental insertion mutant in BAB1_0114::248-407 This work

Ba::pJQK-BAB2_0693 Ba-parental insertion mutant in BAB2_0693::249-386 This work

Ba::pJQK-BAB1_0607 Ba-parental insertion mutant in BAB1_0607::278-427 This work

Ba::pJQK-BAB1_1465 Ba-parental insertion mutant in BAB1_1465:72-190 This work

Ba∆BAB2_0133 Ba-parental deletion mutant in BAB2_0133∆38-299 This work

Ba∆BAB2_0134 Ba-parental deletion mutant in BAB2_0134∆23-237 This work

Ba∆BAB2_0135 Ba-parental deletion mutant in BAB2_0135∆40-441 This work

Ba∆BAB2_0105 Ba-parental deletion mutant in BAB2_0105∆33-307 This work

Ba∆BAB1_1620 Ba-parental deletion mutant in BAB1_1620∆23-241 This work

Ba∆wadD Ba -parental deletion mutant in BAB1_0953∆50-281 This work

Ba∆wadD::Tn7-PwadD Ba∆wadD complemented strain with miniTn7T This work KmR harbouring BAB1_0953 complete ORF with its own promoter.

Ba∆wadB Ba-parental LPS core mutant Gil-Ramírez et al., 2014

Ba∆wadC Ba-parental LPS core mutant Conde- Álvarez et al., 2012

Ba∆per Ba-parental O-PS mutant Martínez- Gómez et al., 2018

BamucR Ba-parental deletion mutant in BAB1_0954∆12-141. This work

106

Chapter 2 Experimental procedures Results

Strain Characteristics Reference

BamucR Ba-parental double deletion mutant in This work BAB1_0954∆12-141 and BAB2_0133∆38-299.

BawboX Ba-parental deletion mutant in BAB1_0998∆18-451 This work

Brucella melitensis Bme-parental B. melitensis wild type, smooth LPS, González et NalR spontaneous mutant of strain 16M. al., 2008

BmewboX Bme-parental deletion mutant in BMEI0999∆18-451 This work

Escherichia coli TOP10 F – lac/q Tn 10 (Tetr) mcrA ∆(mrr-hsdRMS- Invitrogen mcrBC) 80lacZ∆M15.∆lacX74 recA1alaD139 ∆(ara-leu)7697 galU galK rpsL endA1 nupG

Stellar F–, endA1, supE44, thi-1, recA1, relA1, gyrA96 Clontech phoA, Φ80d lacZΔ M15, Δ(lacZYA-argF) U169 Δ(mrr-hsdRMS-mcrBC), ΔmcrA, λ–

DH5α F- Φ80lacZΔM15 Δ(lacZYA-argF) U169 recA1 Invit rogen end A1 hsdR17(rk-, mk+) phoA supE44 thi-1 gyrA96 relA1 λ-

β2150 thrB1004 pro thi strA hsdS lacZDM15(F` lacZDM15 Dehio and laclq traD36 proA+ proB+)DdapA::erm (Ermr) pir Meyer, 1997

S17 λ pir Mating strain with plasmid RP4 inserted Simon et al., into the chromosome 1983

PIR1 F-∆lac169 rpoS(Am) robA1 creC510 hsdR514 Invitrogen endA recA1 uidA(∆MluI)::pir-116

SM10 λpir-pTNS2 Plasmid encoding the information for Tn7 Choi et al., transposition in the right place. AmpR 100 μg/ml 2005

HB101- pRK2013 pRK2013 is a helper plasmid for conjugation. Choi et al., KmR 35μg/ml 2005

107

Chapter 2 Experimental procedures Results

Table 2.2. Plasmids. Plasmid Characteristics Reference pCR2.1 Cloning vector, KmR Invitrogen pJQKm Derivate of pJQ200KS+. Suicide vector, KmR SacS Scupham and Triplett, 1997 pSKoriT Derivative of pBluescript. Suicide vector, KmR SacS Addgene pUC18 R6KT Broad host-range mini-Tn7 vector Llobet et al., 2009 mini Tn7T KmR pMSB-1 1014 bp of B. abortus chromosomal DNA This work containing the BAB2_0133 deletion allele (∆38-299), generated by PCR and cloned into pCR2.1. pMSB-2 BamHI/XbaI fragment from pMSB-1 cloned into the This work corresponding sites of pJQKm. pMSB-5 295 bp of B. abortus chromosomal DNA This work containing the BAB1_0417 internal fragment, generated by PCR and cloned into pCR2.1. pMSB-6 BamHI/XbaI fragment from pMSB-5 cloned into the This work corresponding sites of pSKoriT. pMSB-7 368 bp of B. abortus chromosomal DNA This work containing the BAB1_0326 internal fragment, generated by PCR and cloned into pCR2.1. pMSB-8 353 bp of B. abortus chromosomal DNA This work containing the BAB2_0105 internal fragment, generated by PCR and cloned into pCR2.1. pMSB-9 BamHI/XbaI fragment from pMSB-8 cloned into the This work corresponding sites of pSKoriT. pMSB-10 BamHI/XbaI fragment from pMSB-7 cloned into the This work corresponding sites of pSKoriT. pMSB-11 358 bp of B. abortus chromosomal DNA This work containing the BAB1_0953 internal fragment, generated by PCR and cloned into pCR2.1. pMSB-12 BamHI/XbaI fragment from pMSB-11 cloned into This work the corresponding sites of pSKoriT. pMSB-16 416 bp of B. abortus chromosomal DNA containing This work the BAB2_0693 internal fragment, generated by PCR and cloned into pCR2.1. pMSB-17 481 bp of B. abortus chromosomal DNA containing This work the BAB1_0114 internal fragment, generated by PCR and cloned into pCR2.1.

108

Chapter 2 Experimental procedures Results

Plasmid Characteristics Reference pMSB-18 357 bp of B. abortus chromosomal DNA containing This work the BAB1_1465 internal fragment, generated by PCR and cloned into pCR2.1. pMSB-19 449 bp of B. abortus chromosomal DNA containing This work the BAB1_0607 internal fragment, generated by PCR and cloned into pCR2.1. pMSB-20 352 bp of B. abortus chromosomal DNA containing This work the BAB2_0105 deletion allele (∆33-307), generated by PCR and cloned into pCR2.1. pMSB-21 BamHI/XbaI fragment from pMSB-19 cloned into the This work corresponding sites of pJQKm. pMSB-22 BamHI/XbaI fragment from pMSB-18 cloned into the This work corresponding sites of pJQKm. pMSB-23 566 bp of B. abortus chromosomal DNA containing This work the BAB1_0998 deletion allele (∆18-451), generated by PCR and cloned into pCR2.1. pMSB-24 BamHI/XbaI fragment from pMSB-23 cloned into the This work corresponding sites of pJQKm. pMSB-25 BamHI/XbaI fragment from pMSB-16 cloned into the This work corresponding sites of pJQKm. pMSB-27 BamHI/XbaI fragment from pMSB-20 cloned into the This work corresponding sites of pJQKm. pMSB-28 BamHI/XbaI fragment from pMSB-17 cloned into the This work corresponding sites of pJQKm. pMSB-29 420 bp of B. abortus chromosomal DNA containing This work the BAB1_0932 internal fragment, generated by PCR and cloned into pCR2.1. pMSB-30 BamHI/XbaI fragment from pMSB-29 cloned into the This work corresponding sites of pJQKm. pMSB-34 XbaI fragment of 479 bp from B. abortus This work chromosomal DNA containing the BAB1_0953 deletion allele (∆50-281) cloned into the corresponding sites of pJQKm by InFusion HD Cloning System. pMSB-37 XbaI fragment of 472 bp from B. abortus This work chromosomal DNA containing the BAB2_0135 deletion allele (∆40-441) cloned into the corresponding sites of pJQKm by InFusion HD Cloning System. pMSB-44 EcoRI fragment of 1771 bp from B. abortus chromosomal This work DNA containing the BAB1_0953 complete allele and its own promoter, cloned into the corresponding sites of pUC18 R6KT miniTn7T KmR by InFusion HD Cloning System.

109

Chapter 2 Experimental procedures Results

Plasmid Characteristics Reference pC3029 In-frame deletion of mucR (∆12-141) plus 1 kb of each flanking Caswell et al., 2013 region in pNPTS138. pYRI-16 885 bp of B. abortus chromosomal DNA containing Gil-Ramírez, 2011 the BAB1_1620 deletion allele (∆24-240), generated by PCR and cloned into pCR2.1. pYRI-17 BamHI/XbaI fragment containing the BAB1_1620 deletion Gil-Ramírez, 2011 allele (Δ24-240) cloned into the corresponding sites of pJQKm. pYRI-19 BamHI/XbaI fragment containing the BAB2_0134 deletion This work allele (Δ23-237) cloned into the corresponding sites of pJQKm.

Construction of mutants. ORFs BAB2_0133, BAB2_0134, BAB2_0135, BAB2_0105 and BAB1_1620 were mutagenized by in frame non-polar deletion. For the construction of BaΔBAB2_0133 mutant, we first generated two PCR fragments: oligonucleotides BAB2_0133-F1 (5'- GCGTTGGACAAGTTGAGGTT-3') and BAB2_0133-R2 (5´-CATAGCGGTCGGTTAAATGC- 3´) (Table S2.1) were used to amplify a 572 base pairs (bp) fragment including codons 1 to 38 of BAB2_0133, as well as 458 bp upstream of the BAB2_0133 start codon (Figure S2.1 in Supplemental material, Chapter 2). Oligonucleotides BAB2_0133-F3 (5´- GTATCGCCAGCCAATTTACGTCCGTATTGGAAGCCAAGAA-3´) and BAB2_0133-R4 (5´- CAGTAACAAAAGGCCGCTAT-3´) were used to amplify a 442 bp fragment including codons 299 to 326 of BAB2_0133 and 355 bp downstream of the BAB2_0133 stop codon. Both fragments were ligated by overlapping PCR using oligonucleotides F1 and R4 for amplification, and the complementary regions between R2 and F3 for overlapping. The resulting fragment, containing the BAB2_0133 deletion allele, was cloned into pCR2.1 (Invitrogen), to generate plasmid pMSB-01, sequenced to ensure the maintenance of the reading frame, subsequently subcloned into the BamHI and the XbaI sites of the suicide plasmid pJQK (Scupham and Triplett, 1997) and transformed into competent E. coli S17 λpir (Simon et al., 1983). The resulting suicide pJQK-derived plasmid was introduced into B. abortus 2308 by conjugation. The first recombination event (integration of the suicide vector in the chromosome) was selected by Nal and Km resistance, and the second recombination (excision of the

110

Chapter 2 Experimental procedures Results mutator plasmid leading to construction of the mutant by allelic exchange), was selected by Nal and sucrose resistance and Km sensitivity. The resulting colonies were screened by PCR with primers F1 and R4 which amplified a fragment of 1014 bp in the mutant and 1794 bp in the sibling strain that keeps the wild type gene. Primers BAB2_0133-F1 and BAB2_0133-R5 (5´-AAGACCCAGTAGTTAGCACT- 3´) amplified a fragment of 919 bp only in the wild type strain. The mutation generated results in the loss of the 80% of the ORF. The BaΔBAB2_0134 mutant was constructed following a similar strategy, in collaboration with Gil-Ramírez: oligonucleotides BAB2_0134-F1 (5´- TGCGGGGTACCAGTTATTACA-3´) and BAB2_0134-R2 (5´-TAGGACAGCGACGAGCTT-3´) amplified a 405 bp fragment including codons 1 to 22 of BAB2_0134, as well as 339 bp upstream of the BAB2_0134 start codon. Oligonucleotides BAB2_0134-F3 (5´- AAGCTCGTCGCTGTCCTAATCGAAGCCCGAATCCTG-3´) and BAB2_0134-R4 (5´- GGACGAATACGGCAGAGACA-3´) amplified a 432 bp fragment including codons 238 to 249 of BAB2_0134 and 393 bp downstream of the BAB2_0134 stop codon (Figure S2.2). The resulting colonies were screened by PCR with primers F1 and R4 which amplified a fragment of 837 bp in the mutant and 1482 bp in the sibling strain which keeps the wild type gene. Primers BAB2_0134-F1 and BAB2_0134-R5 (5´- AGTGCTGCCCTGACAAAAAT-3´) amplified a fragment of 878 bp only in the wild type strain. The mutation generated results in the loss of the 87% of the ORF and the mutant was called BaΔBAB2_0134.

BaΔBAB2_0135 mutant was constructed following the same procedure and using oligonucleotides BAB2_0135-F1 (5´- TGGCGGCCGCTCTAGAACACCGGACTGCCTGATAA - 3´) and BAB2_0135-R2 (5´- CGGGCAATTTCGGCATAG -3´) that amplified a 240 bp fragment including codons 1 to 40 of BAB2_0135, as well as 120 bp upstream of the BAB2_0135 start codon, and oligonucleotides BAB2_0135-F3 (5´- CTATGCCGAAATTGCCCGCCGGTTTGGAAATGCGGTCAA -3´) and BAB2_0135-R4 (5´- ATCCACTAGTTCTAGTTATGTAGCCGCCACCGTTT -3´) that amplified a 232 bp fragment including codons 441 to 478 of BAB2_0135 and 115 bp downstream of the BAB2_0135 stop codon (Figure S2.3). The resulting colonies were screened by PCR with primers F1 and R4 that amplified a fragment of 472 bp in the mutant and 1672 bp in the sibling

111

Chapter 2 Experimental procedures Results strain that keeps the wild type gene. Primers BAB2_0135-F1 and BAB2_0135-R5 (5´- CGATTGCCAGTCCCAGAAAG -3´) amplified a fragment of 628 bp only in the wild type strain. The mutation generated results in the loss of the 84% of the ORF.

For the construction of BaΔBAB2_0105 mutant oligonucleotides BAB2_0105-F1 (5´- GCGTGTTCTACAGCCATGAA -3´) and BAB2_0105-R2 (5´- CCGCCGAAATGTAGGAAGTG -3´) amplified a 198 bp fragment including codons 1 to 33 of BAB2_0105, as well as 99 bp upstream of the BAB2_0105 start codon. Oligonucleotides BAB2_0105-F3 (5´-

CACTTCCTACATTTCGGCGGTATGTTGGATTGGGACGGGT -3´) and BAB2_0105-R4 (5´- GCCGAATATGACGCTTGCTA -3´) amplified a 154 bp fragment including codons 307 to 330 of BAB2_0105 and 79 bp downstream of the BAB2_0105 stop codon (Figure S2.4). The resulting colonies were screened by PCR with primers F1 and R4 which amplified a fragment of 352 bp in the mutant and 1171 bp in the sibling strain which keeps the wild type gene. Primers BAB2_0105-F1 and BAB2_0105-R5 (5´- CAAAGACCGGATATTGCGGG -3´) amplified a fragment of 550 bp only in the wild type strain. The mutation results in the loss of the 83% of the ORF.

BaΔBAB1_1620 mutant was constructed using oligonucleotides 1620-F1 (5´- GTACGCGGTCGTAGCTCAGT-3´) and 1620-R2 (5´-CTCAAACTGAGACGCCATGA-3´), that amplified a 475 bp fragment including codon 1 to 23 of BAB1_1620 as well as 406 bp upstream of the ORF start codon. Oligonucleotides 1620-F3 (5´- TCATGGCGTCTCAGTTTGAGATAGCCAACGTCACCAAAACA-3´) and 1620-R4 (5´- CTCTGCAATTCTTGCGATCA-3´) were used to amplify a 410 bp fragment including codons 241 to 261 of the BAB1_1620 ORF and 347 bp downstream of the BAB1_1620 stop codon. Both fragments were ligated, cloned into pCR2.1 to generate plasmid pYRI- 16, and subcloned into the suicide pJQK (pYRI-17). After conjugation with B. abortus, the resulting colonies were screened by PCR with primers 1620-F1 and 1620-R4 which amplified a 885 bp fragment in the mutant and 1536 bp in the parental strain. The mutation generated results in the loss of the 83% of the BAB1_1620 ORF. BaΔmucR mutant was constructed using the suicide plasmid pC3029 of Caswell and collaborators, made with oligonucleotides mucR-F1 (5´- ACAATGTTATCGCCCACCAT -3´) and mucR-R2 (5´- GGTGCTTTCGTCGTTCGTT -3´) to amplify a 333 bp fragment including

112

Chapter 2 Experimental procedures Results codons 1 to 11 of BAB1_0594, as well as 300 bp upstream of the BAB1_0594 start codon. Oligonucleotides mucR-F3 (5´- AACGAACGACGAAAGCACCTGATTCTTCAGCGAGTGAATCACG -3´) and mucR-R4 (5´- ACCGGAAATCAGTTCAGTGG -3´) amplified a 68 bp fragment including codon 142 of BAB1_0594 (Figure S2.6) (Caswell et al., 2013). The resulting colonies were screened by PCR with primers F1 and R4 which amplified a fragment of 401 bp in the mutant and 794 bp in the wild type strain. Primers mucR-F5 (5´- AACGAACGACGAAAGCACC -3´) and mucR-R6 (5´- GCTCAGGCGTCATGTTGTAA -3´) amplified a fragment of 296 bp only in the wild type strain. The mutation generated results in the loss of the 92.3% of the ORF and the mutant was called BaΔmucR.

The double mutant BaΔmucRΔBAB2_0133 was constructed by deletion of BAB2_0133 over the mutant BaΔmucR.

The BaΔwboX mutant (see Annex Chapter 2) was constructed following the same strategy. Oligonucleotides wboX-F1 (5´- CTAAGAAACCCACACCCTGC -3´) and wboX-R2 (5´- AATGACGCGGTTATCAGGGA -3´) amplified a 204 bp fragment including codons 1 to 17 of BAB1_0998, as well as 153 bp upstream of the start codon. Oligonucleotides wboX-F3 (5´- TCCCTGATAACCGCGTCATTCTCAATTCATCCATCGACGG -3´) and wboX-R4 (5´- ACGCTTGTCCCTCCTGTAAA -3´) amplified a 362 bp fragment including codons 452 to 515 and 170 bp downstream of the ORF stop codon (Figure S2.7). The resulting colonies were screened by PCR with primers F1 and R4 which amplified a fragment of 566 bp in the mutant and 1868 bp in the sibling strain which keeps the wild type gene. Primers wboX-F1 and wboX-R5 (5´- AAACGACCAATAGCACCGAC -3´) amplified a fragment of 941 bp only in the wild type strain. The mutation generated results in the loss of the 84.5% of the ORF. For the construction of the BmeΔwboX mutant, the same suicide vector was used.

The rest of the ORFs were mutagenized by recombination and gene disruption using a suicides vector pJQK or pSKoriT (Tibor et al., 2002) carrying an internal fragment of the ORF. For the construction of Ba::pJQK-BAB1_0114 mutants, we generated a PCR fragment using oligonucleotides BAB1_0114-F1 (5'- TCAACAAATCGGCCAAGGAC -3')

113

Chapter 2 Experimental procedures Results and BAB1_0114-R2 (5´- GTCACGCGGTCAAACTGG - 3´) which amplified a 481 bp fragment containing the region that codes for aminoacids 248 to 407 (Figure S2.8). The fragment was cloned into pCR2.1, to generate plasmid pMSB-17, sequenced and subcloned into the BamHI and the XbaI sites of the suicide plasmid pJQK to obtain pMSB-28, and then transformed into competent E. coli S17 and transferred into B. abortus 2308 by conjugation. The integration of the suicide vector and disruption of the target gene was selected by Nal and Km resistance and by PCR combining

BAB1_0114-F3 (5´- CCTATATTCCCCAGGCCGTT - 3´) with M13 Forward (5'- CTGGCCGTCGTTTTAC -3') or with M13 Reverse (5'- CAGGAAACAGCTATGAC-3'). These last two primers hybridize in the suicide vector inserted in the chromosome. BAB1_0114-F3 and M13 Forward amplified a fragment of 881 bp only in the mutant strain. Following the same strategy, we constructed the rest of insertion mutants: Mutant Ba::pSKoriT-BAB1_0417 was made using oligonucleotides BAB1_0417-F1 (5'- TGATCGACCATGGCTCGG-3') and BAB1_0417-R2 (5´- TCAAGCCTGACCAGAAGCC - 3´) which amplified a 295 bp fragment of BAB1_0417 (codon 37 to 134) (Figure S2.9). The fragment was first cloned in pCR2.1 (pMSB-05), subcloned into the suicide plasmid pSKoriT (pMSB-06), and transferred into B. abortus 2308 by conjugation. Primers M13

Reverse and BAB1_0417-F3 (5´- CTGTTTCCCGACCAGCTTG - 3´) amplified a fragment of 649 bp only in the mutant. Oligonucleotides BAB2_0693-F1 (5'- CACTGCAAGCCGGTTACAAT -3') and BAB2_0693-R2 (5´- TGCAACGAAATTCTGTCCGG - 3´) were used for the construction of Ba::pJQK-BAB2_0693 mutant. F1 and R2 amplified a fragment of 416 bp (codons 249 to 386) (Figure S2.11). We generated plasmid pMSB-16, subsequently subcloned into the suicide plasmid pJQK (pMSB-24), and conjugated into B. abortus 2308. Primers

M13 Forward and BAB2_0693-F3 (5´- ACGAGCGCTATGATTTCGTC - 3´) amplified a fragment of 684 bp only in the mutant. For the construction of Ba::pJQK-BAB1_0932 mutants, we used oligonucleotides BAB1_0932-F1 (5'- GCCGTCGTCCTGAATGTTAC -3') and BAB1_0932-R2 (5´- GCCATTATCCAGTGCAGCC - 3´) which amplified a 420 bp fragment of BAB1_0932 (codons 354 to 493) (Figure S2.13). We generated plasmid pMSB-28, subsequently subcloned into the suicide plasmid pJQK (pMSB-29), and conjugated into B. abortus 2308. The resulting Nal-Km resistant colonies were screened by PCR. Primers M13

114

Chapter 2 Experimental procedures Results

Reverse and BAB1_0932-F3 (5´- GGCCGAGAATGGCTATATCA - 3´) amplified a fragment of 915 bp only in the mutant. Mutant Ba::pSKoriT-BAB1_0326 was made using oligonucleotides BAB1_0326-F1 (5'- GCACTCAACCGGCTCAATTG -3') and BAB1_0326-R2 (5´- AGCACCGCATATTCAAAGGC - 3´) which amplified a 368 bp fragment of BAB1_0326 (codons 261 to 383) that was cloned into pCR2.1 to obtain pMSB-07 (Figure S2.14). The fragment was then subcloned into the suicide pSKoriT (pMSB-10), and conjugated into B. abortus 2308. The resulting Nal-Km resistant colonies were screened by PCR. Primers M13 Reverse and BAB1_0326-F3 (5´- ATGTTGCCATGTCGCTGTTT - 3´) amplified a fragment of 678 bp only in the mutant strain. Construction of Ba::pJQK-BAB1_0607 mutant was made using oligonucleotides BAB1_0607-F1 (5'- GCCAATGTCGTTCTCTCCAA -3') and BAB1_0607-R2 (5´- CTTGGTGTCAGCCCCTTTTC - 3´) which amplified a 449 bp fragment of BAB1_0607 (codons 278 to 427) (Figure S2.15). We generated the pCR2.1-derived plasmid pMSB- 19, and then subcloned into the suicide plasmid pJQK (pMSB-21), and conjugated into B. abortus 2308. Primers M13 Forward and BAB1_0607-F3 (5´- TTCTTTCCAATGAGCGCACC - 3´) amplified a fragment of 800 bp only in the mutant. For the construction of Ba::pJQK-BAB1_1465 mutants, we used oligonucleotides BAB1_1465-F1 (5'- GGCACGGACGTCTCAAAATA -3') and BAB1_1465-R2 (5´- CTCGACTGCTTGCAGGAAAG - 3´) which amplified a 357 base pairs fragment of BAB1_1465 (codons 72 to 190) (Figure S2.10). We generated plasmid pMSB-18, subsequently subcloned into the suicide plasmid pSKoriT (pMSB-22), and conjugated into B. abortus 2308. The resulting colonies were screened by PCR with primers F1 and R2 which amplified a fragment of 357 bp. Primers M13 Reverse and BAB1_1465-F3 (5´- CCAGCAGCACGATCCTTG- 3´) amplified a fragment of 691 bp only in the mutant strain. We constructed two different mutants in ORF BAB1_0953 (wadD). The first, Ba::pSKoriT-BAB1_0953, carried the suicide vector inserted in the gene and was obtained with oligonucleotides BAB1_0953-F1 (5'- ACTTTTCGCCGAGCAACAAA -3') and BAB1_0953-R2 (5´- AGGCACGGTTTCATAGACGA - 3´) which amplified a 358 bp fragment of BAB1_0953 (codon 112 to 230) (Figure S2.12). We generated plasmid pMSB-11, subsequently subcloned into the suicide plasmid pSKoriT (pMSB-12), and

115

Chapter 2 Experimental procedures Results conjugated into B. abortus 2308. Primers M13 Forward and BAB1_0953-F3 (5´- GCTGGCTTCATGAAATCCGT - 3´) amplified a fragment of 612 bp in mutant.

We also constructed a non-polar wadD mutant (Ba wadD) by in frame deletion. Oligonucleotides wadD-F1 (5'- TCTATAATGAGAGGCGGCTTTT -3') and wadD-R2 (5´- AGAAGTGCTGGTCCTGTTGT - 3´) were used to amplify a 304 (bp) fragment including codons 1 to 50 of BAB1_0953, as well as 154 bp upstream of the BAB1_0953 start codon. Oligonucleotides wadD-F3 (5´- ACAACAGGACCAGCACTTCTATCCTCACCCTGCCATTCAA -3´) and wadD-R4 (5´- CTGGTACTAGACGCCCTGTT -3´) were used to amplify a 175 bp fragment including codons 281 to 324 of BAB1_0953 and 43 bp downstream of the BAB1_0953 stop codon (Figure S2.5). Both fragments were ligated by overlapping PCR using oligonucleotides F1 and R4 for amplification, and the complementary regions between R2 and F3 for overlapping. The resulting fragment, containing the BAB1_0953 deletion allele, was cloned directly into pJQK by the InFusion technique to generate pMSB-34. This suicide vector was sequenced to ensure the maintenance of the reading frame and transferred into B. abortus 2308 by conjugation. The resulting colonies were screened by PCR with primers F1 and R4 that amplified a fragment of 479 bp in the mutant and 1169 bp in the sibling strain which keeps the wild type gene. Primers wadD-F1 and wadD-R5 (5´- AGGCACGGTTTCATAGACGA -3´) amplified a fragment of 844 bp only in the wild type strain. The mutation generated results in the loss of the 71% of the ORF.

Complementation of wadD mutants. For complementation experiments, we performed a stable insertion of the miniTn7 transposon into the chromosome of BaΔwadD (Choi and Schweizer, 2006). Por this purpose, we first generated a PCR product using oligonucleotides Tn7-wadD-F1 (5`- CGGGCTGCAGGAATTGCGATTCCTTTGTGCCAGAT-3`) and Tn7-wadD-R2 (5`- GCTTCTCGAGGAATTATCATCGCCGCATTGAAGAC-3`), which amplified a 1771 bp fragment including codons 1 to 323 of BAB1_0953 together with 481 bp upstream of the ORF start codon including the putative wadD promoter and 318 bp downstream the ORF stop codon (Figure S2.16). This PCR product was cloned into the corresponding sites of the linearized pUC18 R6KT miniTn7T KmR vector (Llobet et al.,

116

Chapter 2 Experimental procedures Results

2009) to generate plasmid pMSB-44. The plasmid was sequenced to ensure the maintenance of the reading frame transformed into E. coli S17 and transferred to BaΔwadD mutant by tetra-parental conjugation between E. coli S17 (pMSB44), E. coli SM10 λpir (pTNS2) and E. coli HB101 (pRK2013). The conjugants harbouring pMSB-44 were selected by plating the mating mixture onto TSA-Nal-Km plates that were incubated at 37ºC for 4 days. To confirm that the transposon was inserted between genes glmS and recG (Choi and Schweizer, 2006), we performed PCR using different oligonucleotides: Tn7F (5`- TGGCTAAAGCAAACTCTTCATTT – 3`) and Tn7R (5`- GCGGATTTGTCCTACTCAGG – 3`) allowed to confirm that the Tn7 was inserted, oligonucleotides Glms_B (5`- GTCCTTATGGGAACGGACGT – 3`) and PTn7-R (5`- CACAGCATAACTGGACTGATT -3`) confirmed that the transposon was inserted immediately after the gene glmS, and RecG-R (5`- TATATTCTGGCGAGCGATCC – 3`) and PTn7-L (5`- ATTAGCTTACGACGCTACACCC – 3`) confirmed that the transposon was inserted before the gene recG (Choi and Schweizer, 2006). The resulting strain was named BaΔwadD::Tn7-PwadD.

Crystal violet exclusion test. To study if the mutants had smooth or rough phenotype, 5 ml of a crystal violet solution at 0.1 mg/ml in distilled water were used to cover isolated colonies on TSA plates for 20 seconds. Smooth colonies (with bacteria that conserve an intact LPS) excluded crystal violet and looked white, whereas rough colonies (with bacteria that have lost the LPS O-PS) capture crystal violet and looked violet.

LPS extraction and characterization. LPS was extracted by the proteinase-K sodium dodecyl sulphate (SDS) protocol (Dubray and Limet, 1987; Garin-Bastuji et al., 1990) with some modifications. Bacteria grown overnight in 10 ml of TSB were killed with 0.5% phenol during 3 days in agitation at 37ºC. After that, samples were weighed and pipetted into small polycarbonate cap tubes and then suspended by ultrasounds in 2% SDS-60 mM Tris-HCl buffer (pH 6.8) at a concentration of 0.5g (wet weight) of bacteria per 10 ml of buffer. Samples were then heated at 100ºC for 10 minutes, and lysates were cooled to 55ºC. This treatment was followed by digestion with 60μl of proteinase-K at 2.5 mg/ml in HCl-Tris per ml of

117

Chapter 2 Experimental procedures Results sample (Merck KGaA) for 3 hours at 55ºC, and overnight incubation at 20ºC. Afterwards, they were centrifuged at 20,000 x g for 30 minutes at room temperature, and the LPS was precipitated from the supernatant by addition of 3 volumes of methanol containing 1% sodium acetate-saturated methanol at -20ºC. After 60 minutes, the precipitate was harvested by centrifugation at 5,000 x g for 15 minutes at 4ºC and resuspended by sonication in 10 ml of distilled water. After a second methanol precipitation and centrifugation, the pellets were resuspended by sonication in 2-3 ml of 60 mM HCl-Tris (pH 6.8) and left at 37ºC. Then samples were treated with 20 μl/ml of RNase and DNase stock solutions at 0.5 mg/ml in HCl-Tris (MP Biomedicals and Sigma-Aldrich, respectively) at 37ºC for 30 minutes. Subsequently, the LPS was treated again with 5 μl/ml of proteinase K at 2.5 mg/ml in HCl-Tris, at 55ºC for 3 hours and then, at room temperature overnight. After a third methanol precipitation in the same conditions described above, the pellet containing LPS was recovered in 1 ml of distilled water and frozen at -20ºC.

SDS-PAGE and Western blots. Samples were mixed 1:1 with Sample buffer 2X (Bio-Rad), heated at 100ºC for 10 minutes and analysed in Tris-HCl-glycine-12, 15 or 18% polyacrylamide gels (37.5:1 acrylamide/methylene-bisacrylamide ratio). 15 μl of each sample were run at 30 mA constant current for 140 minutes. Finally, LPS molecules were revealed by the periodate-alkaline silver method (Tsai and Frasch, 1982). For Western blot, gels were electro-transferred onto PVDF sheets (Whatman, Schleicher & Schuell, WESTRAN S.; 0.2 μm pore size) in a transfer buffer (pH 8.3) containing 0.025M Tris, 0.192M glycine, and 20% (vol/vol) methanol. Transfer was performed at a constant voltage of 8V and 200mA for 30 minutes in a Trans-Blot Semi- Dry Transfer Cell (Bio-Rad). After transfer, membranes were blocked overnight with 3% skim milk in PBS containing 0.05% Tween 20, and next washed with the same buffer. They were then incubated overnight at room temperature with immune sera diluted 1:500 in the same solution. After washing, the corresponding peroxidase-conjugated secondary antibody was added, and incubation continued for at least 4 hours at room temperature. Bound immunoglobulins were detected using 4-chloro-1-naphthol-H2O2. Antibodies used in

118

Chapter 2 Experimental procedures Results this study were monoclonal A68/24D08/G09 and A68/24G12/A08 (from the ascitic fluid of an infected mouse), which recognize core epitopes (Bowden et al., 1995) and a polyclonal serum from a rabbit infected with 109 colony forming unit (CFU) of B. melitensis 16M and bled at day 45.

Enzyme Linked Immunosorbent Assay (ELISA). Reactivity of Brucella mutants was tested in 96-well flat-bottomed microtiter plates (NUNC MaxiSorp, Thermo Scientific, Waltham, Massachusetts, U.S.A) coated with a suspension of heat-killed (65°C, 60 min) Brucella cells (A600nm of 1.0) in classic Phosphate-Buffered Saline (PBS) pH 7.2 by triplicate. Plates were then washed with PBS-Tween and incubated with MoAbs anti OMP 10/16/19 (A76/10D03/H02, A76/08C03/G03, A68/04G01/C06 and A68/08E07/B11), anti O-PS (05D4, 04F9, 18H08, 2E11, 12B12, 12G12 and 07F09) or anti R-LPS (A68/24G12/A08, A68/24D08/G09, A68/03F03/D05 and A68/10A06/B11) (Bowden et al., 1995; Cloeckaert et al., 1990, 1993).

Growth curves. To study the growth rates of each strain and to analyse if any of the mutants had growth deficiencies, cells were inoculated into 10 ml of TSB in a 50-ml flask and incubated at 37ºC with orbital agitation for 18 hours. The number of bacteria was determined by optical density (O.D.) at 600nm in a spectrophotometer. Bacteria were harvested by centrifugation at 13,000 rpm for 10 minutes and resuspended in 1ml of Mueller-Hinton medium at an O.D. 600nm of 0.1. Starting with this O.D., cells were cultivated in a Bioscreen C (Lab Systems) with continuous shaking in a working volume of 200 μl/well. Temperature was controlled at 37ºC and O.D. at 420-580nm. Absorbance values of cell suspensions were automatically read at regular intervals of 0.5 hours, over a 120 hours’ period. All experiments were performed in triplicate. Tubes contained only the culture medium (Mueller-Hinton) were used as a sterility control.

119

Chapter 2 Experimental procedures Results

Sensitivity to polycationic bactericidal peptides. The minimal inhibitory concentration (MIC) of polymyxin B and poly-L-ornithine (both from Sigma-Aldrich), that affect bacterial growth were determined in Mueller- Hinton medium. Exponentially growing bacteria were adjusted to an O.D. equivalent to 1 of the McFarland scale, and exposed to serial dilutions of the bactericidal peptides. The MIC were determined after 2 days of incubation at 37ºC. As controls we used a well without antibiotic and another one without bacteria. Experiments were performed in duplicate.

Sensitivity to the bactericidal action of nonimmune serum. Exponentially growing bacteria were adjusted to 104 colony forming unit/ml in saline and dispensed in duplicate in microtiter plates (30μl/well) containing 60μl of new-born bovine serum. After 90 minutes of incubation at 37ºC with gentle agitation, complement action was blocked by adding brain heart infusion (BHI) broth (150μl/well). After mixing the BHI broth with the bacterial suspension, 75μl were plated by triplicate on TSA plates. 5 days after incubation at 37ºC, results were expressed as the percentage of CFU recovered with respect to control samples where new-born bovine serum was substituted by PBS.

Virulence in mice. Seven-week-old female BALB/c mice (ENVIGO, Harlan) were lodged in cages with water and food ad libitum for 2, 8 or 12 weeks. The animal handling and procedures were in accordance with the current Spanish and European legislation (RD 1201/2005; directive 14 86/609/EEC, respectively), supervised by the Animal Welfare Committee of the Institution (Protocol number R102/2007). Virulence assays with mutant in BAB1_1620 were carried out in the Unidad de Sanidad Animal of the Centro de Investigación y Tecnología Agroalimentaria (CITA) de Aragón during Yolanda Gil- Ramírez PhD studies (Gil-Ramírez, 2011). 18 groups of 5 mice each were inoculated with BaΔwadD, BaΔBAB2_0133, BaΔBAB2_0134, BaΔBAB2_0135, BaΔBAB1_1620 or Ba-parental. Inocula were prepared in sterile PBS and each mouse was administered intraperitoneally approximately with 5 x 104 CFU in 0.1 ml. To assess the exact dose retrospectively, dilutions of each inoculum were plated by triplicate on TSA plates.

120

Chapter 2 Experimental procedures Results

Spleen CFU in infected mice were counted at 2, 8 and 12 weeks after inoculation for BaΔwadD, BaΔBAB1_1620 and Ba-parental and at 2 and 8 weeks post-infection for BaΔBAB2_0133, BaΔBAB2_0134 and BaΔBAB2_0135. The spleens were weighed and homogenized in 9 volumes of PBS and serial ten-fold dilutions were accomplished and plated by triplicate on TSA plates. After 5 days of incubation at 37ºC the colonies were checked by crystal violet exclusion test and PCR. The data were normalized by logarithmic transformation and the mean log CFU/spleen values and the standard deviations were calculated.

Statistical analysis. Statistical significance for sensitivity to normal serum was evaluated with one-way ANOVA followed by Dunnett´s multiple comparisons test (∗∗∗∗p < 0.0001). For virulence analysis, statistical significance between the parental strain and the wadD mutant was evaluated using t-Student independent samples test (∗p < 0.05).

121

Results

Chapter 2 Results Results

RESULTS

Screening for putative LPS core glycosyltransferases. A bioinformatic search in the Carbohydrate-Active Enzymes database CAZy (www.cazy.org) revealed 23 ORFs in the genome of B. abortus 2308 that could code for glycosyltransferases. We excluded from further analysis BAB1_0108-cgs, which is involved in cyclic glucan synthesis (Briones et al., 2001), BAB1_1786-mtgA and BAB1_1450-murG, both related to peptidoglycan synthesis, BAB1_1171-lpxB, probably implicated in lipid A formation (Iriarte et al., 2004) and BAB1_0553-wbkA, BAB1_0563- wbkE; BAB1_1000-wboA and BAB1_1000-wboB, 4 genes that belong to the O-PS synthesis route (see Annex Chapter 2) (Godfroid et al., 2000; González et al., 2008; McQuiston et al., 1999). Similarly, 4 ORFs correspond to those glycosyltransferases already known to be involved in the synthesis of the LPS core: BAB1_0639-wadA (Monreal et al., 2003), BAB1_0351-wadB (Gil-Ramírez et al., 2014), BAB1_1522-wadC (Conde-Álvarez et al., 2012) and BAB2_0209-waaA (Iriarte et al., 2004) (see Introduction of this Chapter) (Table S2.2). The remaining 11 ORFs and data on their presence in other Brucella species and genetic location are in Table 2.3 and Figure 2.2. Of these, 7 (BAB1_0953, BAB2_0105, BAB2_0133, BAB2_0135, BAB1_1620, BAB1_0607 and BAB1_0932) were highly conserved in all Brucella spp., but 4 (BAB2_0693, BAB1_0417, BAB1_0114 and BAB1_0326) presented significant differences when compared to B. abortus sequences, mainly due to frameshifts generating shorter proteins (Table 2.3).

123

Chapter 2 Results

Table 2.3. Hypothetical glycosyltransferases identified in B. abortus and their corresponding orthologues in other Brucella spp.

Specie ORF and presence

B. abortus BAB1_1620 BAB1_0326 BAB2_0133 BAB2_0135 BAB1_0953 BAB2_0693 BAB1_0607 BAB1_0114 BAB1_0932 BAB1_0417 BAB2_0105 (260 aa1) (630 aa) (326 aa) (478 aa) (323 aa) (614 aa) (718 aa) (763 aa) (819 aa) (168 aa) (330 aa) B. melitensis ✓2 ✓ ✓ ✓4 ✓ ✓5 ✓ ✓7 ✓ ✓ ✓ B. suis bv. 1 ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ B. suis bv. 2 ✓ ✓ ✓ ✓ ✓  ✓ ✓ ✓ ✓ ✓ B. canis ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓  ✓ B. ovis ✓ ✓3 ✓ ✓4 ✓  ✓6 ✓7 ✓ ✓ ✓8 B. microti ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ B. pinnipedialis ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ 1 aa: aminoacids ✓ Ortholog present (when the symbol is alone, the ortholog is 100% identical to B. abortus sequence).  Frameshift leading to a premature stop. 2 8 aminoacids longer (N-terminal) than B. abortus. 3 57 aminoacids shorter (N-terminal) than B. abortus. 4 1 aminoacid shorter (N-terminal) than B. abortus. 5 142 aminoacids shorter (N-terminal) than B. abortus. 6 6 aminoacids longer (N-terminal) than B. abortus. 7 33 aminoacids shorter (N-terminal) than B. abortus. 8 21 aminoacids longer (N-terminal) than B. abortus.

124

Chapter 2 Results

125

Chapter 2 Results

Figure 2.2. Schematic representation of the not previously described ORFs encoding hypothetical glycosyltransferases in B. abortus 2308. Presence in other Brucella species, genetic regions, family to which they belong to and regulation. B. melitensis 16M; B. suis 1330 (biovar 1); B. suis ATCC 23445 (biovar 2); B. canis ATCC 23365; B. ovis ATCC 25840; B. microti CCM 4915 and B. pinnipedialis B2/94.

The hypothetical glycosyltransferases controlled by BvrR/BvrS or MucR are not required for the synthesis of a complete core LPS or for virulence. Perusal of the literature revealed some information on 5 of those 11 putative glycosyltransferases. BAB1_1620 is regulated by the two-component regulatory system BvrR/BvrS, a master regulator of Brucella virulence that modulates outer membrane homeostasis and controls unknown aspects of LPS structure (Manterola et al., 2005; Viadas et al., 2010). It is an isolated ORF located in chromosome I and surrounded by genes implicated in cell cycle (Figure 2.2). The LPS of mutants in BAB1_1620 presented smooth and rough fractions similar to that of the parental strain (Gil-Ramírez, 2011). Mutants in BAB1_1620 behaved similarly to Ba-parental in polycationic peptide resistance and virulence in mice (data not shown) (Gil-Ramírez, 2011).

Expression of BAB1_0326, BAB2_0133 and BAB2_0135 has been shown to be controlled by MucR, a general virulence regulator in Brucella (Caswell et al., 2013). However, although it has been reported that B. abortus and B. melitensis mucR mutants have a defective LPS core since they fail to react with a monoclonal antibody that recognizes this part of the LPS, the glycosyltransferases involved have not been identified (Caswell et al., 2013; Mirabella et al., 2013). This difference with the parental strain could account for the attenuated phenotype of the mucR mutant (Caswell et al., 2013; Mirabella et al., 2013). It is known that six of the mucR regulated genes (BAB1_0326, BAB1_0560, BAB1_1465, BAB2_0133, BAB2_0134 and BAB2_0135) are predicted to be related to polysaccharide biosynthesis or modification (Caswell et al., 2013). However, no work has been conducted to elucidate if any of them is involved in the synthesis of the core LPS. To clarify this point, we analysed in detail the role of all these ORFs. BAB1_0560 was not considered as a candidate for core synthesis since it codes for

ManBOAg, a previously identified phosphomannomutase required for the synthesis of

126

Chapter 2 Results perosamine, the subunit of the O-PS (Godfroid et al., 2000). Thus, we constructed insertion mutants in BAB1_0326 and BAB1_1465 (since the downstream ORFs are oriented in the opposite direction) and non-polar deletion mutants in BAB2_0133, BAB2_0134 and BAB2_0135 (part of an operon, see below). We then extracted their LPS with SDS-proteinase-K and analysed the profile by SDS-PAGE and Western-Blot. The LPS of insertion mutants in ORFs BAB1_0326 and BAB1_1465 were similar to that of the parental strain (Figure 2.3), suggesting that they do not play a role in LPS synthesis.

Figure 2.3. BAB1_1465 and BAB1_0326 are not related to LPS synthesis. Insertion mutants Ba::pJQK- BAB1_1465 and Ba::pSKoriT-BAB1_0326 showed the same reactivity to polyclonal serum against S-Brucella as the parental strain (Ba-parental) in Western-blot analysis.

Interestingly, a genomic search of the region surrounding the other three mucR- regulated ORFs (coordinates 129088 to 132626 of B. abortus chromosome II), showed that they are part of an operon encoding a hypothetical protein (BAB2_0132), a putative glycosyltransferase (BAB2_0133) that would belong to the family 2 of glycosyltransferases, currently associated with LPS synthesis, a dehydrogenase (BAB2_0134) and a dolichyl-phosphate-mannose-protein mannosyltransferase (BAB2_0135) (Figure 2.4).

127

Chapter 2 Results

Figure 2.4. Schematic representation of the chromosomal region between coordinates 129088 and 132626, containing ORFs BAB2_0133, BAB2_0134 and BAB2_0135 of the mucR-regulated operon.

We constructed non-polar deletion mutants in ORFs BAB2_0133, BAB2_0134 and BAB2_0135 and analysed their characteristics in different tests. Any of them differed from Ba-parental in growth rate or in the crystal violet exclusion test, classically used to differentiate smooth and rough Brucella (Alton et al., 1988). The migration profile of the LPS extracted from the mutants was also similar to that of the parental strain and all reacted similarly to Ba-parental LPS with a polyclonal serum against S-Brucella (Figure 2.5A and B). Moreover, all mutants showed the same reactivity as the parental strain with an anti R-Brucella serum (Figure 2.5C) and with MoAb A68/24G12/A08 directed against R-LPS (Figure 2.5D). These results suggest that mutants in these ORFs are not affected in the structure of the LPS.

128

Chapter 2 Results

Figure 2.5. The hypothetical glycosyltransferases controlled by MucR are not required for the synthesis of a complete core LPS. (A) 18% SDS-PAGE electrophoresis and silver staining of LPS samples extracted with the SDS-proteinase K protocol. (B) Western-blot analysis of LPS extracts with polyclonal serum against S-Brucella. (C) Western-blot with a polyclonal serum against R-Brucella. (D) Western-blot analysis with monoclonal anti-core antibody A68/24G12/A08.

129

Chapter 2 Results

It is known that mutants with a defect in the core lateral branch are affected in their interaction with elements of the innate immune system and their virulence in the murine model (Conde-Álvarez et al., 2012; Gil-Ramírez et al., 2014). Therefore, even if we did not appreciate differences in LPS pattern, we wondered whether mutants in the ORFs of this operon were affected in their ability to multiply in mice and explained, at least in part, the mucR attenuated phenotype. We thus inoculated BALB/c mice with mutants in BAB2_0133, BAB2_0134 or BAB2_0135 and Ba-parental and compared the number of CFU/spleen 2 and 8 weeks post-infection. As can be seen in Figure 2.6, mutants produced CFU/spleen that did not differ significantly from those of Ba- parental at weeks 2 (p=0.78; 0.39 and 0.39 respectively) and 8 (p=0.63; 0.62 and 0.92) after infection.

Figure 2.6. Mutants in ORFs BAB2_0133, BAB2_0134 and BAB2_0135 did not show attenuation in mice. Spleen CFU in infected BALB/c mice were counted at 2 and 8 weeks after inoculation of 5 x 104 CFU.

To gain further insight into the properties of the mutants, we selected those in the first glycosyltransferase of the operon (BAB2_0133) and the dehydrogenase (BAB2_0134), and compared their sensitivity to components of the innate immune system with that of the parental strain and BaΔwadC mutant, defective in the synthesis of the core lateral branch (Conde-Álvarez et al., 2012). First, we measured their MIC to polymyxin B and poly-L-ornithine and we did not see any difference with the parental strain (Table 2.4). Secondly, we measured the mutant sensitivity to new- born bovine serum. Moreover, Figure 2.7 shows that after 90 minutes of exposure, the percentage of survival of both mutants was parallel to that of the parental strain (p>0.05) while BaΔwadC showed a statistically significant (p < 0.01) difference.

130

Chapter 2 Results

Table 2.4. Sensitivity to polycationic peptides of mutants in BAB2_0133 and BAB2_0134.

MIC1 (µg/ml) to: Strain Polymyxin B Poly-L-ornithine

Ba-parental 4.27 108

BaΔwadC 3.20 32

BaΔBAB2_0133 4.27 108

BaΔBAB2_0134 4.27 108

1MIC: Minimal inhibitory concentration

Figure 2.7. Mutants in ORFs BAB2_0133 and BAB2_0134 were as sensitive to the lytic action of complement present in bovine serum as the parental strain.

These results are consistent with the idea that the putative glycosyltransferases regulated by MucR or BvrR/BvrS are not involved in the synthesis of LPS or of other components implicated in virulence, at least under the conditions used in this study.

Is MucR implicated in the synthesis of the Brucella core LPS? An interesting observation is that, in a previous study, colonies from a B. abortus mucR mutant, analysed with the crystal violet exclusion test, showed a R-LPS phenotype, and complementation with the wild type gene reverts to a S-LPS phenotype.

131

Chapter 2 Results

However, in the same study, SDS-PAGE gel electrophoresis indicated that, although the core section was affected, the O-PS was intact and this was confirmed with Western-blot analysis with O-PS specific antibodies (Caswell et al., 2013). These observations are intriguing since is the first case in which the results of the crystal violet assay do not correspond to the LPS profiles. It is also in contradiction with a B. melitensis mucR mutant, that showed a smooth phenotype in both assays (Mirabella et al., 2013). In view of this apparent controversy, we decided to construct a B. abortus mucR mutant using the same suicide vector as Caswell et al. As already observed by these authors, the growth of our mucR mutant was delayed in the first hours (generation time 7.4 hours versus 4.7 hours for Ba-parental), but reached stationary phase as the wild type strain (Figure 2.8). More interestingly, our mutant was unstable in solid media and repeated isolation and re-isolation always gave a mixture of colonies that were different both in size and, independently, in their ability to take up or exclude the crystal violet. We then extracted the LPS of Ba∆mucR, but instead of following the hot phenol protocol used by Caswell et al., we used the SDS-proteinase-K method, that has been successful in the extraction and detection of core LPS defective mutants in wadC and wadB (Conde-Álvarez et al., 2012; Gil-Ramírez et al., 2014).

Figure 2.8. Ba∆mucR showed a higher generation time (7.4 hours) than the parental strain (4.7 hours), but reached stationary phase as the wild type strain. Growth curves of B. abortus 2308 (Ba-parental) and mucR mutant (Ba∆mucR) on Mueller-Hinton medium. Each point represents the mean of triplicate samples.

132

Chapter 2 Results

Our results confirmed that Ba∆mucR kept the O-PS although, surprisingly, it reacted as the parental LPS with the same monoclonal antibody against core epitopes used by Caswell et al. (A68/24G12/A08) (Figure 2.9A), and with a polyclonal serum anti S-Brucella (Figure 2.9B). Thus, in our hands, mucR is not involved in the synthesis of LPS core section. Moreover, our results are in agreement with the fact that mutants in the ORFs coding putative glycosyltransferases regulated by mucR seem to keep an intact core LPS (see above). The final role of mucR in the synthesis of the core LPS needs the analysis of the chemical structure of mutants LPS.

Figure 2.9. Ba∆mucR kept the O-PS and its LPS reacted as that of Ba-parental in the core LPS section. Western-blot analysis of LPS extracts with monoclonal anti-core antibody A68/24G12/A08 (A) and with polyclonal serum (against S-Brucella) of an infected rabbit with B. melitensis 16M (45 days) (B).

BAB1_0953 encodes WadD, a previously unidentified glycosyltransferase involved in the synthesis of the LPS core lateral branch. To address whether the other seven putative glycosyltransferases identified in the bioinformatic search (BAB1_0953, BAB2_0105, BAB2_0693, BAB1_0607, BAB1_0114, BAB1_0932 and BAB1_0417) were required for LPS synthesis in B. abortus, we constructed insertion mutants in each one. All of them behaved as smooth Brucella in the crystal violet assay and the analysis of the extracted LPS confirmed that all mutants presented smooth fractions with a migration profile parallel to that of Ba-parental LPS and reacted similarly with a polyclonal serum against S-Brucella (Figure 2.10, Table

133

Chapter 2 Results

S2.3 and data not shown). Interestingly, although keeping the S fraction, mutant in BAB1_0953 lost reactivity in the R fraction, suggesting a defect in the core and/or lipid A epitope(s) recognized by polyclonal sera of infected animals (Rojas et al., 2001). Since this was not observed for the other mutants, we investigated further BAB1_0953 and the phenotype associated with its mutation.

Figure 2.10. The LPS of insertion mutants in the signaled ORFs reacted with a polyclonal serum as the parental strain, except for BAB1_0953 mutant, that failed to react in the rough fraction. Western-blot analysis of LPS extracts with a polyclonal serum from a rabbit infected with B. melitensis 16M (smooth Brucella). 1 For BaΔwboX see Annex Chapter 2.

BAB1_0953 is an isolated gene and the adjacent ORFs are encoded in the complementary strand. Thus, it was very unlikely that a polar effect caused the LPS phenotype of the insertion mutant. However, to rule out such a possibility, we constructed a non-polar deletion mutant, hereafter named BaΔwadD following the nomenclature previously established for Brucella LPS core genes (Gil-Ramírez et al., 2014; Reeves et al., 1996). BaΔwadD LPS showed a migration profile similar to that of Ba-parental in the high molecular weight S-LPS fraction and an increased mobility in the R-LPS one, and Western-blot analysis showed that the latter failed to react with the polyclonal serum. Assignment of the defect to the core fraction was probed with MoAbs A68/24G12/A08 and A68/24D08/G09. Since they recognize outer core (i.e. GlcN related) epitopes (Conde-Álvarez et al., 2012; Gil-Ramírez et al., 2014), the observed failure to react with these MoAbs proved that the insertion and deletion

134

Chapter 2 Results mutants in BAB1_0953 had a core defect involving the GlcN tetrasaccharide. Moreover, complementation by stable insertion of a complete wadD gene in the bacterial chromosome of the deletion mutant (BaΔwadD::Tn7-PwadD) restored the migration pattern of the R-fraction to the level of the Ba-parental LPS and normal reactivity to polyclonal serum and monoclonal anti-core antibodies (Figure 2.11). An ELISA with several anti-core MoAbs and whole bacteria confirmed the core defect (Figure 2.12A). We only observed little but constant differences in reactivity of Ba- parental and Ba∆wadD with anti-Outer Membrane Proteins (OMP) antibodies (Figure 2.12B), consistent with the presence of the O-PS and the subsequent steric hindrance to the access of antibodies to the OMPs (Bowden et al., 1995).

Figure 2.11. Mutant BaΔwadD presented a defective LPS core oligosaccaride. A. SDS-PAGE electrophoresis and silver staining of LPS samples extracted SDS-proteinase K. B. Western-blot analysis of LPS extracts with polyclonal serum of an infected rabbit with smooth B. melitensis 16M (anti-S- Brucella). C. Western-blot analysis with monoclonal anti-core antibody A68/24G12/A08.

135

Chapter 2 Results

Figure 2.12. Outer membrane epitopes in BaΔwadD and Ba-parental. Anti-O-PS and anti-R-LPS (upper panel) and anti-OMP antibodies (lower panel), measured by ELISA test.

All these results strongly suggest that wadD encodes a previously unidentified glycosyltransferase involved in the synthesis of the core lateral branch.

WadD orthologs are present in all Brucella spp. but in a recently characterized isolate from amphibians. As signaled in General Introduction, the genus Brucella is classified in several species that form a core group including B. abortus, B. melitensis, B. suis, B. canis, B. ovis and B. neotomae (often referred to as “classical” Brucella species) as well as more recent isolates from a variety of mammals (B. ceti, B. pinnipedialis, B. microti, B. papionis and B. vulpis). They are separated from several early-diverging brucellae that

136

Chapter 2 Results produce and atypical LPS and include B. inopinata BO1 and BO2, a novel atypical strain isolated from rodents in Australia and several strains isolated from amphibians (Al Dahouk et al., 2017; Scholz et al., 2010; Soler-Lloréns et al., 2016; Tiller et al., 2010). In silico analysis (Figure 2.13) showed that wadD was highly conserved in the core brucellae including the “classical” spp. B. melitensis, B. suis (smooth LPS) and B. ovis and B. canis, (rough LPS) and also in other “non-classical” Brucella spp. (B. pinnipedialis, B. microti, B. ceti and B. vulpis) with minor aminoacid modifications. This changes observed in the aminoacid sequence (Table S2.4) could account for the differences in the LPS of those spp. when compared to B. abortus (Zygmunt et al., 2012; Cloeckaert personal communication). We also anlayzed the presence of wadD in the clade of early-diverging brucellae that depart from the core spp. Interestingly, wadD was present in all of them but absent in Brucella spp. B13-0095, one of the four Brucella strains isolated from frogs that have been completely sequenced, although this strain conserved wadB and wadC. Finally, WadD was 72% and 71% homologous to Ochrobactrum anthropi and O. intermedium orthologs respectively, two species that also belong to the α-2 Proteobacteria subclass and are the closest genetic neighbor of Brucella.

As pointed out above, WadD belongs to the family 2 of glycosyltransferases and some of the members of this family are multifunctional, and conserve two DXD domains (Coutinho et al., 2003). As shown in Figure 2.13, this aminoacid sequences are present in all the Brucella WadD orthologs.

137

Chapter 2 Results

B. abortus MPIFKIIIATTTRNRPKMLINLYKSLGDLEIPSNIDVEFLIVENNRTSTSESWLHEIRS B. melitensis MPIFKIIIATTTRNRPKMLINLYKSLGDLEIPSNIDVEFLIVENNRTSTSESWLHEIRS B. suis bv. 1 MPIFKIIIATTTRNRPKMLINLYKSLGDLEIPSNIDVEFLIVENNRTSTSESWLHEIRS B. suis bv. 2 MPIFKIIIATTTRNRPKMLINLYKSLGDLEIPSNIDVEFLIVENNRTSTSESWLHEIRS B. suis bv. 5 MPIFKIIIATTTRNRPKMLINLYKSLGDLEIPSNIDVEFLIVENNRTSTSESWLHEIRS B. ovis MPIFKIIIATTTRNRPKMLINLYKSLGDLEIPSNIDVEFLIVENNRTSTSESWLHEIRS B. canis MPIFKIIIATTTRNRPKMLINLYKSLGDLEIPSNIDVEFLIVENNRTSTSESWLHEIRS B. microti MPIFKIIIATTTRNRPKMLINLYKSLGDLEIPSNIDVEFLIVENNRTSTSESWLHEIRS B. pinnipedialis MPIFKIIIATTTRNRPKMLINLYKSLGDLEIPSNIDVEFLIVENNRTSTSESWLHEIRS B. vulpis MPIFKIIIATTTRNRPKMLINLYKSLSDLEIPSNIDVEFLIVENNRTSTSENWLHEIRC B. ceti MPIFKIIIATTTRNRPKMLINLYKSLGDLEIPSNIDVEFLIVENNRTSTSESWLHEIRS B. inopinata MPIFKIIIATTTRNRPKMLINLYKSLSDLEIPSNIDVEFLIVENNRTSTSENWLHEIRS B. inopinata BO2 MPIFKIIIATTTRNRPKMLINLYKSLSDLEIPSNIDVEFLIVENNRTSTSENWLHEIRS NF2653 (Austr.) MPIFKIIIATTTRNRPKMLINLYKSLSDLEIPSNIDVEFLIVENNRTSTSENWLHEIRS 09RB8471(frog) MPIFKIIIATTTRNRPKMLINLYKSLSDLEIPSNIDVEFLIVENNKTSTSENWLHEIRS * * * *

B. abortus SISPSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. melitensis SISPSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. suis bv. 1 SISPSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. suis bv. 2 SISPSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. suis bv. 5 SISPSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. ovis SISPSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. canis SISPSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. microti SISPSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. pinnipedialis SISPSAVVYILETSIDISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. vulpis SISPSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. ceti SISPSAVVYILETSIDISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. inopinata SISSSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR B. inopinata BO2 SISPSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR NF2653 (Austr.) SISPSAVVYILETSIGISCARNRALDYAQEAGADFLAFVDDDEFVEPDWLKQLFAEQQRR 09RB8471(frog) SIPSSAVVYILETSIGISCARNRALDYAQEASADFLAFVDDDEFVEPDWLKQLFAEQQRR ** * *

B. abortus DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. melitensis DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. suis bv. 1 DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. suis bv. 2 DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. suis bv. 5 DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. ovis DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. canis DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. microti DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. pinnipedialis DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. vulpis DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. ceti DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. inopinata DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL B. inopinata BO2 DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL NF2653 (Austr.) DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL 09RB8471(frog) DLDLVGSPVRPVPQNSKLSLWQRFVWSGVERNGTRAEDRARRKWQENKADTIKIATGSWL

B. abortus GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH B. melitensis GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH B. suis bv. 1 GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH B. suis bv. 2 GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH B. suis bv. 5 GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH B. ovis GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH B. canis GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH B. microti GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH B. pinnipedialis GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH B. vulpis GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH B. ceti GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH

138

Chapter 2 Results

B. inopinata GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKHGAKTGWAPDAIVYETVPHCRISFSYH B. inopinata BO2 GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPHCRISFSYH NF2653 (Austr.) GRIDFFRRTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH 09RB8471(frog) GRIDFFRKTGLRFDSKLGLTGGEDWNLWLEAKKLGAKTGWAPDAIVYETVPYCRISFSYH * * *

B. abortus FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGQALISLAMAL B. melitensis FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGQALISLAMAL B. suis bv. 1 FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGQALISLAMAL B. suis bv. 2 FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGQALISLAMAL B. suis bv. 5 FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLITAILTLPFKGGQALISLAMAL B. ovis FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGQALISLAMAL B. canis FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGQALISLAMAL B. microti FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGQALISLAMAL B. pinnipedialis FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGQALISLAMAL B. vulpis FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGQALISLAMAL B. ceti FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGQALISLAMAL B. inopinata FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFRGGQALISLAMAL B. inopinata BO2 FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFRGGQALISLAMAL NF2653 (Austr.) FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGQALISLAMAL 09RB8471(frog) FRRNRDHNATEFTLLYSKSPRRAWMRLPSRILSRVWKLLTAILTLPFKGGRLSSRWRWLW * * **********

B. abortus GGIVGLVQACCGKQQLHYKETTGS B. melitensis GGIVGLVQACCGKQQLHYKETTGS B. suis bv. 1 GGIVGLVQACCGKQQLHYKETTGS B. suis bv. 2 GGIVGLVQACCGKQQLHYKETTGS B. suis bv. 5 GGIVGLVQACCGKQQLHYKETTGS B. ovis GGIVGLVQACCGKQQLHYKETTGS B. canis GGIVGLVQACCGKQQLHYKETTGS B. microti GGIVGLVQACCGKQQLHYKETTGS B. pinnipedialis GGIVGLVQACCGKQQLHYKETTGS B. vulpis GGIVGLVQACCGKQQLHYKETTGS B. ceti GGIVGLVQACCGKQQLHYKETTGS B. inopinata GGIVGLVQACCGKQQLHYKETTGS B. inopinata BO2 GGIVGLVQACCGKQQLHYKETTGS NF2653 (Austr.) GGIVGLVQACCGKQQLHYKETTGS 09RB8471(frog) EELSGWSRPAAENSSFIIRKQPAL **** *******************

Figure 2.13. Alignment of the protein encoded by wadD in B. abortus (biovar 2308); B. melitensis (biovar 1, 16M), B. suis bv. 1 (strain 1330, biovar 1); B. suis bv. 2 (strain ATCC 23445, biovar 2, Thomsen); B. suis bv. 5 (strain 513, biovar 5); B. ovis (strain ATCC 25840); B. canis (strain ATCC 23365); B. microti (strain CCM 4915); B. pinnipedialis (strain B2/94); B. vulpis; B. ceti; B. inopinata (strain BO1); B. inopinata BO2 (strain BO2); NF2653 (Austr.) (strain NF 2653, Australian isolate) and 09RB8471 (frog) (strain 09RB8471 isolated from amphibians). The changes in nucleotides are marked in grey.

Dysfunction of wadD generates increased sensitivity to cationic peptides and normal serum. Brucella is resistant to polycationic peptides (Freer et al., 1996), a property attributed to the O-PS density and to the low negative charge of the core and lipid A sections (Martínez de Tejada et al., 1995; Velasco et al., 2000). It is known that B. abortus wadC and wadB mutants, defective in the core lateral branch, are more

139

Chapter 2 Results sensitive to polymyxin B (Conde-Álvarez et al., 2012; Gil-Ramírez et al., 2014). To test if the core defect displayed by Ba∆wadD affected the resistance to polycationic peptides, we used poly-L-ornithine, a mildly bactericidal cationic peptide and the two known core mutants Ba∆wadB and Ba∆wadC plus the parental strain as references. The results (Figure 2.14A) showed that wadD dysfunction brought about a sensitivity similar to that of Ba∆wadB but less than that of Ba∆wadC. These differences in sensitivity were not due to growth defects because Ba∆wadD had a growth rate similar to that of Ba-parental, and experiments with the highly bactericidal lipopeptide polymyxin B confirmed the role of wadD (Ba∆wadD MIC = 0.094 µg/ml versus MIC = 2 µg/ml for both the mutant complemented with wild-type wadD and Ba-parental) (Figure S.17). S brucellae are resistant to the bactericidal action of normal serum, a property associated with both the O-PS hindrance to inner OM targets such as OMPs and the PAMP modifications of the core that reduce binding of the complement activators of the antibody-independent classical pathway (Conde-Álvarez et al., 2012; Fontana et al., 2016; Gil-Ramírez et al., 2014). We compared the sensitivity to newborn bovine and ovine serum of Ba-parental and wadB, wadC or wadD mutants and observed that the three core mutants were more sensitive than Ba-parental. The effect was more remarkable for mutant wadC than for wadD or wadB mutants (Figure 2.14B).

140

Chapter 2 Results

Figure 2.14. A Mutant in wadD was more sensitive than Ba-parental to components of innate immune system. A. MIC determined by the serial dilution method to poly-L-ornithine. The results shown are representative of three experiments in which Ba-parental, BaΔwadC, BaΔwadB and BaΔwadD were assayed simultaneously. B. Survival of after incubation in non-immune serum for 90 minutes. Data are the media ± standard error of triplicate assays.

Dysfunction of wadD generates attenuation detectable in the chronic phase in the mouse model. To analyse the role of wadD in virulence, we infected BALB/c mice (n=5) with Ba∆wadD or Ba-parental and compared the CFU/spleen at weeks 2, 8 and 12 (Figure 2.15). At weeks 8 and 12 post-infection, the CFU numbers of Ba∆wadD were significantly lower than those of Ba-parental (p = 0.0003 and p = 0.0073 respectively), showing that wadD is required for full Brucella virulence in mice. This result is in line with previous observations with wadB and wadC mutants (Conde-Álvarez et al., 2012; Gil-Ramírez et al., 2014) and further confirms that an intact LPS core is necessary for virulence.

141

Chapter 2 Results

Figure 2.15. Mutants in wadD showed slight attenuation at late stage of infection. Spleen CFU in infected BALB/c mice were counted at 2, 8 and 12 weeks after inoculation of 5 x 104 CFU. Means were compared by t-Student independent-samples test.

142

Chapter 2 Discussion and Future Directions

DISCUSSION AND FUTURE DIRECTIONS

In this work we have analysed the role of ORFs BAB1_0114, BAB1_0417, BAB2_0693, BAB1_0932, BAB1_0607, BAB2_0105, BAB1_1620, BAB2_0133, BAB2_0135, BAB1_0326 annotated as hypothetical glycosyltransferases in B. abortus genome. Our results indicate that mutants in these ORFs react similarly to the parental strain in the smooth and rough LPS fractions and suggest that, in the studied conditions and with the available techniques, they seem not to be required for the synthesis of a complete LPS. Interestingly, the last three ORFs and BAB1_1465 have been shown to be controlled by mucR, a regulator of Brucella virulence. Although, as signalled above, it has been reported that B. abortus and B. melitensis mucR mutants have a defect in the core LPS (Caswell et al., 2013; Mirabella et al., 2013), the glycosyltransferases responsible for this defect have not been identified. In this work we have shown that mutation of the mucR-regulated genes BAB2_0133, BAB2_0134, BAB2_0135, BAB1_0326 and BAB1_1465 (Caswell et al., 2013) does not affect to the synthesis of the core, at least in the conditions tested. Nevertheless, since the expression of these genes seems to be repressed by mucR (Figure S2.18) (Caswell et al., 2013), a single mutation in the ORF could not be sufficient for the complete clarification of their role in LPS synthesis and further work would be required.

We have also analysed in more detail the role of the hypothetical glycosyltransferase BAB1_1620, as it is regulated by the master two-component regulator BvrR/BvrS that controls Brucella virulence and the expression of surface components. According to our results, this ORF is not required for the synthesis of a complete LPS and is not implicated in surface-dependent characteristics that confer resistance to polycationic peptides or in virulence in the mouse model.

More interestingly, we report the identification of wadD, a gene encoding a new glycosyltransferase involved in the synthesis of the core section not linked to the O-PS and thus, corroborate and extend previous work indicating that the Brucella LPS core is a branched structure that constitutes a steric impairment preventing the elements of the innate immune system to fight against Brucella (Conde-Álvarez et al., 2012;

143

Chapter 2 Discussion and Future Directions

Fontana et al., 2016; Gil-Ramírez et al., 2014; Kubler-Kielb and Vinogradov, 2013), and thus contribute to Brucella virulence.

The discovery of genes wadC and wadB, involved in the synthesis of the lateral branch not linked to the O-PS, was critical for the understanding of the structure and the role of the core section in virulence. It has been clearly demostrated that in a wadC mutant, the complete core lateral branch is absent because this mutant cannot incorporate the mannose residue that is the depart of the lateral branch and it links to the lipid A-core section (Conde-Álvarez et al., 2012; Fontana et al., 2016). In accordance, deletion of wadC results in higher sensitivity to polycationic peptides and complement, better recognition by the CD14-MD2-TLR4 receptor complex, maturation of dendritic cells, secretion of pro-inflammatory cytokines (including Th1-type cytokines IL-12 and IFN-γ) and attenuation in mice (Conde-Álvarez et al., 2012; Fontana et al., 2016). A wadB mutant is also more sensitive to elements of the innate immune system and shows attenuation in mice, although not to the levels of the wadC mutant (Gil-Ramírez et al., 2014). As we show here, disruption of wadD in B. abortus leads to a smooth strain with a core defect less severe than that of the wadC mutant, more sensitive to components of the innate immune system than the parental strain and less virulent in the murine model. Interestingly, its sensitivity to polycationic peptides is similar to that of the wadB mutant (Gil-Ramírez et al., 2014) and not as strong as that of the wadC mutant (Conde-Álvarez et al., 2012), that has lost the complete branch. This is in accordance with the fact that removal of glucosamine residues would cause an increase in overall negative charge of the remaining LPS inner section that will facilitate the binding of polycationic peptides. In the mice model, WadD attenuation was significant when compared with the parental strain at later stages of infection. However, in comparison with wadB, the attenuation caused by wadD disruption was not significant. These results would be compatible with the loss of one external glucosamine residue of the tetrasaccharide branch in the wadD mutant, while WadB would act in a more internal part of the branch and thus the wadB mutation would result in the loss of several

144

Chapter 2 Discussion and Future Directions glucosamine residues. These results would suggest that a gradual defect in the core LPS correlates with virulence.

According to chemical studies performed in B. melitensis, the core lateral branch contains a mannose and four glucosamines residues assambled as follows: β-D-GlcpN-

(16)-β-D-GlcpN-(14)-[β-D-GlcpN-(16)]-β-D-GlcpN-(13)-α-D-Manp-(15) (Figure 2.1). Considering that wadC, wadB and wadD are perfectly conserved in B. melitensis and B. abortus, and since WadC adds the mannose (Conde-Álvarez et al., 2012; Fontana et al., 2016), in all likelihood, the four glucosamines should be added by WadB (Gil-Ramírez et al., 2014) and WadD. These glucosamines are bound to each other by

β-(16), or β-(14) links, and the one bound to mannose by β-(13) is also linked to two glucosamine residues, both in β-(16) and β-(14). If our study is complete and WadB and WadD are the only glycosyltransferases involved in the assembly of the glucosamine tetrasacchride and its binding to the manose residue, one of them (or both) could be bi-functional and thus, able to add sugars in different linkage. Most glycosyltransferase enzymes involved in LOS (lipooligosaccharide) or LPS core biosynthesis are responsible for one type of sugar addition onto the growing chain (Raetz and Whitfield, 2002). However, some bacterial glycosyltransferase enzymes of the GT-2 family, to which WadD belongs, can be multifunctional and are characterized by the presence of tandems of two active sites (DXD) on one polypeptide (Coutinho et al., 2003) as is the case of Lgt3, responsible for the addition of three glucoses with different linkages [β-(1-3), β-(1-4) and β-(1-6)] onto the inner core of Moraxella catarralis LOS (Coutinho et al., 2003; Luke-Marshall et al., 2013). Interestingly, WadD from B. melitensis, B. abortus and all the orthologs in the other Brucella spp. conserves two DXD domains, opening the door to the possibility of a bifunctional role for this glycosyltransferase (Figure 2.13). Nevertheless, the understanding of the particular role of each glycosyltransferase in the linkage of the different glucosamines to form the pentasacharide would require the elucidation of the core chemical structure of wadB and wadD mutants.

For this purpose, we have constructed double mutants in the genes coding the core glycosyltransferases and in per, the gene coding for the enzyme that synthesizes perosamine that forms the O-PS polysaccharide in Brucella. The LPS of mutants

145

Chapter 2 Discussion and Future Directions

Ba∆per∆wadB, Ba∆per∆wadC and Ba∆per∆wadD will be purified and analysed by NMR spectroscopy in collaboration with Dr. Widmalm from the Stockholm University (work in progress).

Contrary to most of the genes encoding glycosyltransferases implicated in the synthesis of the LPS, that are clustered in the same or related regions of the Brucella genome (González et al., 2008; Monreal et al., 2003; Rajashekara et al., 2004, 2008; Vemulapalli et al., 2000), wadC, wadB and wadD (BAB1_1522, BAB1_0351 and BAB1_0953 respectively), although all situated in chromosome I, are isolated and surrounded by other ORFs apparently not related to LPS synthesis. This, and the fact that some other genes involved in the synthesis of the core (manBcore and manCcore) are situated in chromosome II (González et al., 2008; Monreal et al., 2003), makes even more intriguing the identification of genes needed for the synthesis of Brucella core and its lateral branch.

In a chemical characterization of the core LPS previously performed in a B. melitensis strain different from the one used in our studies, a glucose residue was found linked to the mannose that is the depart of the lateral branch, and, if this were the case, a new glycosyltransferase could be needed (Kubler-Kielb and Vinogradov, 2013). However, it should be taken into account that the LPS extraction method and the B. melitensis biovar used for the determination of the core structure in this experiement were different from those used in our genetic and biochemical studies (Figure 2.1). It is important to notice that the chemical structure we discuss in Figure 2.1 has been elucidated in the same B. melitensis strain where the wadC gene was mutated (Fontana et al., 2016) and, in this case, no glucose residues were detected. The fact that, as discussed above, wadC, wadB and wadD are perfectly conserved in B. melitensis, reinforces the interpretation of our results and the idea that the glycosyltransferases encoded by the last two genes would be involved in the assembly of glucosamine residues.

Nevertheless, we should consider that, although the phenotype of wadC mutant in B. melitensis and B. abortus is similar (Conde-Álvarez et al., 2012; Fontana et al., 2016), previous results suggest that there could exist differences in the structure of the core

146

Chapter 2 Discussion and Future Directions in these two Brucella spp., since they react differently with monoclonal antibodies against core epitopes (González et al., 2008). Moreover, we have already seen that some of the studied ORFs (and discarded in our first screening since they reacted as the parental strain in the rough and smooth LPS fractions) present differences between B. abortus and B. melitensis (Table 2.3). Thus, we could not discard them as the responsible for these differences. To understand the final role of wadB, wadC and wadD, it would be necessary to analyse and compare the chemical structure of the core section in mutants in these genes in both spp.

As we will discuss in more detail in Chapter 3, the fact that mutants defective in the core lateral branch modulate the immune response has been proposed as a strategy for the design of Brucella vaccines (Conde-Álvarez et al., 2012; Zhao et al., 2017). The demonstration that this modulation presents different intensity according to the gene that has been disrupted, also opens new perspectives for the genetic engineering of vaccines. Thus, a wadC mutant can be interesting since induces the stronger immune response. However, if the persistence of the wadC-derived vaccine is too long, we could think about the possibility of combining a “less aggressive” deletion of the core lateral branch (mutants wadB or wadD) with a mutation in a metabolic gene. This would allow engineering a vaccine candidate with a reduced capability of multiplying in vivo and, at the same time, able to induce a stronger immune response than that of the wild type strain or classical vaccines. In this line, several mutants in Brucella metabolic genes have been shown to be attenuated (Barbier et al., 2011, 2018; Ronneau et al., 2014; Zúñiga-Ripa et al., 2014).

Finally, it has been shown that B. ovis core mutants (WadB and WadC) protect against brucellosis caused by the rough spp. B. ovis (Soler-Lloréns et al., 2014). Since we have demonstrated that wadD is also present in B. ovis, it would be interesting to compare the phenotype of wadB, wadC and wadD mutants in B. ovis and analyse whether the later also protects against B. ovis infection in this background.

147

Chapter 2 References

REFERENCES

Al Dahouk, S., Köhler, S., Occhialini, A., Jiménez de Bagüés, M. P., Hammerl, J. A., Eisenberg, T., et al. (2017). Brucella spp. of amphibians comprise genomically diverse motile strains competent for replication in macrophages and survival in mammalian hosts. Sci. Rep. 7, 44420. doi:10.1038/srep44420. Allen, C. A., Adams, L. G., and Ficht, T. A. (1998). Transposon-derived Brucella abortus rough mutants are attenuated and exhibit reduced intracellular survival. Infect. Immun. 66, 1008–16. Alton, G. G., Jones, L. M., Angus, R. D., and Verger, J. M. (1988). Techniques for the brucellosis laboratory. Paris (France): INRA (Institut National de la Recherche Agronomique). Barbier, T., Nicolas, C., and Letesson, J. J. (2011). Brucella adaptation and survival at the crossroad of metabolism and virulence. FEBS Lett. 585, 2929–2934. doi:10.1016/J.FEBSLET.2011.08.011. Barbier, T., Zúñiga-Ripa, A., Moussa, S., Plovier, H., Sternon, J. F., Lázaro-Antón, L., et al. (2018). Brucella central carbon metabolism: an update. Crit. Rev. Microbiol. 44, 182–211. doi:10.1080/1040841X.2017.1332002. Bowden, R., Cloeckaert, A., Zygmunt, M., Bernard, S., and Dubray, G. (1995). Surface exposure of outer membrane protein and lipopolysaccharide epitopes in Brucella species studied by Enzyme-Linked Immunosorbent Assay and flow cytometry. Infect. Immun. 63, 3945–3952. Briones, G., Iñón de Iannino, N., Roset, M., Vigliocco, A., Paulo, P. S., and Ugalde, R. A. (2001). Brucella abortus cyclic beta-1,2-glucan mutants have reduced virulence in mice and are defective in intracellular replication in HeLa cells. Infect. Immun. 69, 4528–35. doi:10.1128/IAI.69.7.4528-4535.2001. Caswell, C. C., Elhassanny, a. E. M., Planchin, E. E., Roux, C. M., Weeks-Gorospe, J. N., Ficht, T. a., et al. (2013). Diverse genetic regulon of the virulence-associated transcriptional regulator MucR in Brucella abortus 2308. Infect. Immun. 81, 1040– 1051. doi:10.1128/IAI.01097-12. Choi, K.-H., Gaynor, J. B., White, K. G., Lopez, C., Bosio, C. M., Karkhoff-Schweizer, R. R., et al. (2005). A Tn7-based broad-range bacterial cloning and expression system. Nat. Methods 2, 443–448. doi:10.1038/nmeth765. Choi, K. H., and Schweizer, H. P. (2006). mini-Tn7 insertion in bacteria with single attTn7 sites: example Pseudomonas aeruginosa. Nat. Protoc. 1, 153–161. doi:10.1038/nprot.2006.24.

149

Chapter 2 References

Cloeckaert, A., De Wergifosse, P., Dubray, G., and Limet, J. N. (1990). Identification of seven surface-exposed Brucella outer membrane proteins by use of monoclonal antibodies: Immunogold labeling for electron microscopy and Enzyme-Linked Immunosorbent Assay. Infect. Immun. 58. Cloeckaert, A., Zygmunt, M. S., Dubray, G., and Limet3, J. N. (1993). Characterization of O-polysaccharide specific monoclonal antibodies derived from mice infected with the rough Brucella melitensis strain B115. J. Gen. Microbiol. 139, 1551–1556. doi: 10.1099/00221287-139-7-1551 Conde-Álvarez, R., Arce-Gorvel, V., Gil-Ramírez, Y., Iriarte, M., Grilló, M. J., Gorvel, J. P., et al. (2013). Lipopolysaccharide as a target for brucellosis vaccine design. Microb. Pathog. 58, 29–34. doi:10.1016/j.micpath.2012.11.011. Conde-Álvarez, R., Arce-Gorvel, V., Iriarte, M., Manček-Keber, M., Barquero-Calvo, E., Palacios-Chaves, L., et al. (2012). The lipopolysaccharide core of Brucella abortus acts as a shield against innate immunity recognition. PLoS Pathog. 8, e1002675. doi:10.1371/journal.ppat.1002675. Coutinho, P. M., Deleury, E., Davies, G. J., and Henrissat, B. (2003). An evolving hierarchical family classification for glycosyltransferases. J. Mol. Biol. 328, 307– 317. doi:10.1016/S0022-2836(03)00307-3. Dehio, C., and Meyer, M. (1997). Maintenance of broad-host-range incompatibility group P and group Q plasmids and transposition of Tn5 in Bartonella henselae following conjugal plasmid transfer from Escherichia coli. J. Bacteriol. 179, 538– 40. Dubray, G., and Limet, J. (1987). Evidence of heterogeneity of lipopolysaccharides among Brucella biovars in relation to A and M specificities. Ann. Inst. Pasteur. Microbiol. 138, 27–37. Fontana, C., Conde-Álvarez, R., Ståhle, J., Holst, O., Iriarte, M., Zhao, Y., et al. (2016). Structural studies of lipopolysaccharide defective mutants from Brucella melitensis identify a core oligosaccharide critical in virulence. J. Biol. Chem. 291, 7727–7741. doi:10.1074/jbc.M115.701540. Freer, E., Moreno, E., Moriyón, I., Pizarro-Cerdá, J., Weintraub, A., and Gorvel, J. P. (1996). Brucella-Salmonella lipopolysaccharide chimeras are less permeable to hydrophobic probes and more sensitive to cationic peptides and EDTA than are their native Brucella sp. counterparts. J. Bacteriol. 178, 5867. Garin-Bastuji, B., Bowden, R. A., Dubray, G., and Limet, J. N. (1990). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblotting analysis of smooth-lipopolysaccharide heterogeneity among Brucella biovars related to A and M specificities. J. Clin. Microbiol. 28, 2169–74. Gil-Ramírez, Y. (2011). Papel de una etanolamina-fosfato transferasa y una glicosil transferasa en la síntesis del lipopolisacárido de Brucella y utilización de una acetil transferasa para la modificación epitópica de la cadena O. (PhD thesis). Universidad de Navarra.

150

Chapter 2 References

Gil-Ramírez, Y., Conde-Álvarez, R., Palacios-Chaves, L., Zúñiga-Ripa, A., Grilló, M.-J., Arce-Gorvel, V., et al. (2014). The identification of wadB, a new glycosyltransferase gene, confirms the branched structure and the role in virulence of the lipopolysaccharide core of Brucella abortus. Microb. Pathog. 73, 53–9. doi:10.1016/j.micpath.2014.06.002. Godfroid, F., Cloeckaert, A., Taminiau, B., Danese, I., Tibor, A., De Bolle, X., et al. (2000). Genetic organisation of the lipopolysaccharide O-antigen biosynthesis region of Brucella melitensis 16M (wbk). Res. Microbiol. 151, 655–668. doi:10.1016/S0923-2508(00)90130-X. Goldman, R. C., and Kohlbrenner, W. E. (1985). Molecular cloning of the structural gene coding for CTP:CMP-3-deoxy-manno-octulosonate cytidylyltransferase from Escherichia coli K-12. J. Bacteriol. 163, 256–61. González, D., Grilló, M.-J. M., De Miguel, M. M.-J., Ali, T., Arce-Gorvel, V., Delrue, R. R.- M., et al. (2008). Brucellosis vaccines: assessment of Brucella melitensis lipopolysaccharide rough mutants defective in core and O-polysaccharide synthesis and export. PLoS One 3, e2760. doi:10.1371/journal.pone.0002760. Iriarte, M., González, D., Delrue, R.-M., Monreal, D., Conde, R., López-Goñi, I., et al. (2004). “Brucella Lipopolysaccharide: Structure, biosynthesis and genetics,” in Brucella. Molecular and Cellular Biology (Wymondham, England: Horizon Bioscience), 159–192. Kubler-Kielb, J., and Vinogradov, E. (2013). The study of the core part and non- repeating elements of the O-antigen of Brucella lipopolysaccharide. Carbohydr. Res. 366, 33–7. doi:10.1016/j.carres.2012.11.004. Lapaque, N., Moriyón, I., Moreno, E., and Gorvel, J.-P. (2005). Brucella lipopolysaccharide acts as a virulence factor. Curr. Opin. Microbiol. 8, 60–66. doi:10.1016/j.mib.2004.12.003. Llobet, E., March, C., Giménez, P., and Bengoechea, J. A. (2009). Klebsiella pneumoniae OmpA confers resistance to antimicrobial peptides. Antimicrob. Agents Chemother. 53, 298–302. doi:10.1128/AAC.00657-08. Luke-Marshall, N. R., Edwards, K. J., Sauberan, S., St Michael, F., Vinogradov, E. V., Cox, A. D., et al. (2013). Characterization of a trifunctional glucosyltransferase essential for Moraxella catarrhalis lipooligosaccharide assembly. Glycobiology 23, 1013–21. doi:10.1093/glycob/cwt042. Mancilla, M., Grilló, M.-J., de Miguel, M.-J., López-Goñi, I., San-Román, B., Zabalza- Baranguá, A., et al. (2013). Deletion of the GI-2 integrase and the wbkA flanking transposase improves the stability of Brucella melitensis Rev 1 vaccine. Vet. Res. 44, 105. doi:10.1186/1297-9716-44-105. Mancilla, M., López-Goñi, I., Moriyón, I., Zárraga, A. M., López-Goñi, I., Moriyón, I., et al. (2010). Genomic island 2 is an unstable genetic element contributing to Brucella lipopolysaccharide spontaneous smooth-to-rough dissociation. J. Bacteriol. 192, 6346–6351. doi:10.1128/JB.00838-10.

151

Chapter 2 References

Manterola, L., Moriyón, I., Moreno, E., Sola-Landa, A., Weiss, D. S., Koch, M. H. J., et al. (2005). The lipopolysaccharide of Brucella abortus BvrS/BvrR mutants contains lipid A modifications and has higher affinity for bactericidal cationic peptides. J. Bacteriol. 187, 5631–9. doi:10.1128/JB.187.16.5631-5639.2005. Martínez-Gómez, E., Ståhle, J., Gil-Ramírez, Y., Zúñiga-Ripa, A., Zaccheus, M., Moriyón, I., et al. (2018). Genomic insertion of a heterologous acetyltransferase generates a new lipopolysaccharide antigenic structure in Brucella abortus and Brucella melitensis. Front. Microbiol. 9, 1092. doi:10.3389/FMICB.2018.01092. Martínez de Tejada, G., Pizarro-Cerdá, J., Moreno, E., and Moriyón, I. (1995). The outer membranes of Brucella spp. are resistant to bactericidal cationic peptides. Infect. Immun. 63, 3054. McQuiston, J. R., Vemulapalli, R., Inzana, T. J., Schurig, G. G., Sriranganathan, N., Fritzinger, D., et al. (1999). Genetic characterization of a Tn5-disrupted glycosyltransferase gene homolog in Brucella abortus and its effect on lipopolysaccharide composition and virulence. Infect. Immun. 67, 3830–5. Mirabella, A., Terwagne, M., Zygmunt, M. S., Cloeckaert, A., De Bolle, X., and Letesson, J. J. (2013). Brucella melitensis MucR, an orthologue of Sinorhizobium meliloti MucR, is involved in resistance to oxidative, detergent, and saline stresses and cell envelope modifications. J. Bacteriol. 195, 453–65. doi:10.1128/JB.01336-12. Monreal, D., Grilló, M. J., González, D., Marín, C. M., De Miguel, M. J., López-Goñi, I., et al. (2003b). Characterization of Brucella abortus O-polysaccharide and core lipopolysaccharide mutants and demonstration that a complete core is required for rough vaccines to be efficient against Brucella abortus and Brucella ovis in the mouse model. Infect. Immun. 71, 3261–71. doi:10.1128/iai.71.6.3261-3271.2003. Raetz, C. R. H., and Whitfield, C. (2002). Lipopolysaccharide endotoxins. Annu. Rev. Biochem. 71, 635–700. doi:10.1146/annurev.biochem.71.110601.135414. Rajashekara, G., Covert, J., Petersen, E., Eskra, L., and Splitter, G. (2008). Genomic island 2 of Brucella melitensis is a major virulence determinant: Functional analyses of genomic islands. J. Bacteriol. 190, 6243–6252. doi:10.1128/JB.00520- 08. Rajashekara, G., Glasner, J. D., Glover, D. A., and Splitter, G. A. (2004). Comparative whole-genome hybridization reveals genomic islands in Brucella species. J. Bacteriol. 186, 5040–51. doi:10.1128/JB.186.15.5040-5051.2004. Reeves, P. R., Hobbs, M., Valvano, M. A., Skurnik, M., Whitfield, C., Coplin, D., et al. (1996). Bacterial polysaccharide synthesis and gene nomenclature. Trends Microbiol. 4, 495–503. doi:10.1016/S0966-842X(97)82912-5. Rojas, N., Freer, E., Weintraub, A., Ramirez, M., Lind, S., and Moreno, E. (1994). Immunochemical identification of Brucella abortus lipopolysaccharide epitopes. 1, 206–213. Rojas, N., Zamora, O., Cascante, J., Garita, D., and Moreno, E. (2001). Comparison of the antibody response in adult cattle against different epitopes of Brucella

152

Chapter 2 References

abortus lipopolysaccharide. J. Vet. Med. Ser. B 48, 623–629. doi:10.1111/j.1439- 0450.2001.00476.x. Ronneau, S., Moussa, S., Barbier, T., Conde-Álvarez, R., Zúñiga-Ripa, A., Moriyón, I., et al. (2014). Brucella , nitrogen and virulence. Crit. Rev. Microbiol. 42, 1–19. doi:10.3109/1040841X.2014.962480. Sangari, F., and Agüero, J. (1991). Mutagenesis of Brucella abortus: comparative efficiency of three transposon delivery systems. Microb. Pathog. 11, 443–6. Scholz, H. C., Nockler, K., Gollner, C., Bahn, P., Vergnaud, G., Tomaso, H., et al. (2010). Brucella inopinata sp. nov., isolated from a breast implant infection. Int. J. Syst. Evol. Microbiol. 60, 801–808. doi:10.1099/ijs.0.011148-0. Scupham, A. J., and Triplett, E. W. (1997). Isolation and characterization of the UDP- glucose 4′-epimerase-encoding gene, galE, from Brucella abortus 2308. Gene 202, 53–59. doi:10.1016/S0378-1119(97)00453-8. Simon, R., Priefer, U., and Pühler, A. (1983). A broad host range mobilization system for in vivo genetic engineering: Transposon mutagenesis in gram negative bacteria. Bio/Technology 1, 784–791. doi:10.1038/nbt1183-784. Soler-Lloréns, P. F., Quance, C. R., Lawhon, S. D., Stuber, T. P., Edwards, J. F., Ficht, T. A., et al. (2016). A Brucella spp. isolate from a Pac-Man frog (Ceratophrys ornata) reveals characteristics departing from classical Brucellae. Front. Cell. Infect. Microbiol. 6, 116. doi:10.3389/fcimb.2016.00116. Soler-Lloréns, P., Gil-Ramírez, Y., Zabalza-Baranguá, A., Iriarte, M., Conde-Álvarez, R., Zúñiga-Ripa, A., et al. (2014). Mutants in the lipopolysaccharide of Brucella ovis are attenuated and protect against B. ovis infection in mice. Vet. Res. 45, 72. doi:10.1186/s13567-014-0072-0. Tibor, A., Wansard, V., Bielartz, V., Delrue, R.-M., Danese, I., Michel, P., et al. (2002). Effect of omp10 or omp19 deletion on Brucella abortus outer membrane properties and virulence in mice. Infect. Immun. 70, 5540–6. doi:10.1128/IAI.70.10.5540-5546.2002. Tiller, R. V, Gee, J. E., Frace, M. A., Taylor, T. K., Setubal, J. C., Hoffmaster, A. R., et al. (2010). Characterization of novel Brucella strains originating from wild native rodent species in North Queensland, Australia. Appl. Environ. Microbiol. 76, 5837– 45. doi:10.1128/AEM.00620-10. Tsai, C.-M., and Frasch, C. E. (1982). A sensitive silver stain for detecting lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 119, 115–119. doi:10.1016/0003-2697(82)90673-X. Ugalde, J. E., Czibener, C., Feldman, M. F., and Ugalde, R. A. (2000). Identification and characterization of the Brucella abortus phosphoglucomutase gene: role of lipopolysaccharide in virulence and intracellular multiplication. Infect. Immun. 68, 5716–23. Velasco, J., Bengoechea, J. A., Brandenburg, K., Lindner, B., Seydel, U., González, D., et al. (2000). Brucella abortus and its closest phylogenetic relative, Ochrobactrum

153

Chapter 2 References

spp., differ in outer membrane permeability and cationic peptide resistance. Infect. Immun. 68, 3210–3218. doi:10.1128/IAI.68.6.3210-3218.2000. Vemulapalli, R., He, Y., Buccolo, L. S., Boyle, S. M., Sriranganathan, N., and Schurig, G. G. (2000). Complementation of Brucella abortus RB51 with a functional wboA gene results in O-antigen synthesis and enhanced vaccine efficacy but no change in rough phenotype and attenuation. Infect. Immun. 68, 3927–32. Viadas, C., Rodríguez, M. C., Sangari, F. J., Gorvel, J.-P., García-Lobo, J. M., and López- Goñi, I. (2010). Transcriptome analysis of the Brucella abortus BvrR/BvrS two- component regulatory system. PLoS One 5, e10216. doi:10.1371/journal.pone.0010216. Zhao, Y., Hanniffy, S., Arce-Gorvel, V., Conde-Álvarez, R., Oh, S., Moriyón, I., et al. (2017). Immunomodulatory properties of Brucella melitensis lipopolysaccharide determinants on mouse dendritic cells in vitro and in vivo. Virulence, 0. doi:10.1080/21505594.2017.1386831. Zúñiga-Ripa, A., Barbier, T., Conde-Álvarez, R., Martínez-Gómez, E., Palacios-Chaves, L., Gil-Ramírez, Y., et al. (2014). Brucella abortus depends on pyruvate phosphate dikinase and malic enzyme but not on Fbp and GlpX fructose-1,6-bisphosphatases for full virulence in laboratory models. J. Bacteriol. 196, 3045–57. doi:10.1128/JB.01663-14. Zygmunt, M. S., Jacques, I., Bernardet, N., and Cloeckaert, A. (2012). Lipopolysaccharide heterogeneity in the atypical group of novel emerging Brucella species. Clin. Vaccine Immunol. 19, 1370–3. doi:10.1128/CVI.00300-12.

154

Chapter 2 Supplemental material

Supplemental material

157

Chapter 2 Supplemental material

5´ BAB2_0133-F1 3´ GCGTTGGACAAGTTGAGGTTGAAAATCCAGATATGACA Start BAB2_0132 CCAGGTCGTAATTCCAGCTATTTTAGCGATGGCGCCCCGCGCGATGTTTAAACGCTTCCC

CATTTTTGTCGGCGGCTCCGCAATAGGCGCTGTCATTGATTATGTCGTGACGCTGTCTGC

GAGCAATTACATAAATTTATATCCGCCGGTTGCGCTGGCGCTGGCGATGCTTATCAGCGG

TTCTGTGGTATTCTTTTTCCATATCCGCATTACATTTCAGTATTCAACGCACAAAATATT

TCGGCAATACGCTCTTTTTATGGGCTGGACATGTGCGATTTATTTTCTTCGTGCTGCGAT

GTTACAAATATGCTTGTATGCGGGCTTGCCACTTGCGGTTGCGCTAATTCTGGCGATCGG

GCTCGCCTCAGTCATTAACTTCGTTATCTCGTCCGTCATCATTTTTGCAAAGAGTCCAGC Stop BAB2_0132 Start BAB2_0133 1 ATGAACGCAAAGACCCTCTGCATTATCGTTCCTGTCTATAATGAGGCGGAAGGCTTGTCT M N A K T L C I I V P V Y N E A E G L S 20

3´ BAB2_0133-R2 5´ CATAGCGGTCGGTTAAATGC 61 GATCTGCTCGACCGCTTGCACTCTGCCGCAGACAGTATCGCCAGCCAATTTACGCTCAAT D L L D R L H S A A D S I A S Q F T L N 40

121 ATCGAGTTTATCTTTATTGACGATGGCAGCAGTGACGGCAGTTTCGCGCTGTTGAAGGCG I E F I F I D D G S S D G S F A L L K A 60

181 CATGATTTCGGGTCCCGTCCGGTTCGCCTGCTGCGTTTTTCACGCAATTTCGGCAAGGAA H D F G S R P V R L L R F S R N F G K E 80

241 GCAGCGCTATCTGCCGGGATAGACGCCGCTGAAGGGGCAGATGCGGCCATCCTGATGGAT A A L S A G I D A A E G A D A A I L M D 100

301 GCCGATCTTCAGCACCCGCCTGAAATGATCGCGGATTTCGTTCGTATCTGGCAAGAGGAA A D L Q H P P E M I A D F V R I W Q E E 120

361 GATGCTGATAGCGTTTATGCCTACAAGGCCAGTCGTCTCGCCTCGGAAGGGCCGGTAAAG D A D S V Y A Y K A S R L A S E G P V K 140

3´ BAB2_0133-R5 5´ AAGACCCAGTAGTTAGCACT 421 GCTGCATTGTCACGGGCGTTCTTCTGGGTCATCAATCGTGATCAGCGATATAAAATTCCG A A L S R A F F W V I N R D Q R Y K I P 160

481 CCGGGGGCGGGAGGTTTTCGCCTGGTCAATCGTCGCTTCATGGCCGCCCTGCGCAGCCTG P G A G G F R L V N R R F M A A L R S L 180

541 CCGGAGAGCGACCGTTTCATGAAGGGGCTCTATGGCTGGGTCGGCTTTAATCAGGTCGGC P E S D R F M K G L Y G W V G F N Q V G 200

601 CTTCCTCTCCAGCCGCCTCCCCGTATGCGGGGTACCAGTTATTACAATCCGCTCCAGCTA L P L Q P P P R M R G T S Y Y N P L Q L 220

661 TTATTGATGTCGCTTGGCGCAATGACCAGTTTCTCGACAACGCCCCTGCGCCTCATGGCA L L M S L G A M T S F S T T P L R L M A 240

721 TTGGCAGGTATTGTCGTGGCCGGCCTCAGCGCTGCATATGGCCTTTATGTCCTGTCGGAA L A G I V V A G L S A A Y G L Y V L S E 260

159

Chapter 2 Supplemental material

781 TATTTCTTCTTCCCCGGCGTTCCGGTTGGCCTTACCTCGATTCTGGCATTGACTGCATTC Y F F F P G V P V G L T S I L A L T A F 280

5´ BAB2_0133-R2 3´ GTATCGCCAGCCAATTTACG 5´ 841 TTCGGAGGGATCCAGCTGATGTTCCTCGGGCTGCTTGGCGAGTATATCGGTAAGTCCGTA F G G I Q L M F L G L L G E Y I G K S V 300

BAB2_0133-F3 3´ 901 TTGGAAGCCAAGAAACGCCCGATCTACATTCTGGCTGAAGATATCCGCCGCAAGGAAAGC L E A K K R P I Y I L A E D I R R K E S 320 Start BAB2_0134 Stop BAB2_0133 961 GATGATGCGGATCGCGGATGATTTTGGGCTAGGGCGCGGACATGACCGGGTCATCCTGTC D D A D R G * 326

GCTGCTCGAAACCGGGCGGCTTGATGGTACGTCAGTCATGATGAACGATGCAATGAACCC

TGAAGACATCGCCCGGCTGCGGAAATTGCGTGCCGGTGGTGCGAAAGTCGGGTTGCATCT

GAACCTGACACAGGCACTTCCGGGTGGCGGCCCCATCTGGCCGCTGGGCGAATTGATACG

GCCGTTATTGGGTGCACCCTTTCTGGGCGCGATCACAGCATCTCTGGTGCGGCAGGTAGA

TGCGTTCGTGACACAATTCGGCAGCCTGCCCGATTATTATGACGGCCATCAGCATTGTCA CAGT TTGTTTTCCGGCGATA AACAAAAGGCCGCTAT 3´ BAB2_0133-R4 5´

Figure S2.1. Sequence of BAB2_0133 and its upstream and downstream regions. Start and stop codons are in green and red characters respectively. Blue characters denote intergenic nucleotides. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids deleted in BAB2_0133 mutant.

160

Chapter 2 Supplemental material

5´ BAB2_0134-F1 3´ TGCGGGGTACCAGTTATTACAATCCGCTCCAGCTATTAT

TGATGTCGCTTGGCGCAATGACCAGTTTCTCGACAACGCCCCTGCGCCTCATGGCATTGG

CAGGTATTGTCGTGGCCGGCCTCAGCGCTGCATATGGCCTTTATGTCCTGTCGGAATATT

TCTTCTTCCCCGGCGTTCCGGTTGGCCTTACCTCGATTCTGGCATTGACTGCATTCTTCG

GAGGGATCCAGCTGATGTTCCTCGGGCTGCTTGGCGAGTATATCGGTAAGTCCGTATTGG

AAGCCAAGAAACGCCCGATCTACATTCTGGCTGAAGATATCCGCCGCAAGGAAAGCGATG 3´ BAB2_0134-R2 Start BAB2_0134 Stop BAB2_0133 TAGGACAGCGAC 1 ATGCGGATCGCGGATGATTTTGGGCTAGGGCGCGGACATGACCGGGTCATCCTGTCGCTG M R I A D D F G L G R G H D R V I L S L 20 5´ GAGCTT 61 CTCGAAACCGGGCGGCTTGATGGTACGTCAGTCATGATGAACGATGCAATGAACCCTGAA L E T G R L D G T S V M M N D A M N P E 40

121 GACATCGCCCGGCTGCGGAAATTGCGTGCCGGTGGTGCGAAAGTCGGGTTGCATCTGAAC D I A R L R K L R A G G A K V G L H L N 60

181 CTGACACAGGCACTTCCGGGTGGCGGCCCCATCTGGCCGCTGGGCGAATTGATACGGCCG L T Q A L P G G G P I W P L G E L I R P 80

241 TTATTGGGTGCACCCTTTCTGGGCGCGATCACAGCATCTCTGGTGCGGCAGGTAGATGCG L L G A P F L G A I T A S L V R Q V D A 100

301 TTCGTGACACAATTCGGCAGCCTGCCCGATTATTATGACGGCCATCAGCATTGTCATTGT F V T Q F G S L P D Y Y D G H Q H C H C 120

361 TTTCCGGCGATAGCGCCTCTCGTGGCCCGCTTGTCCTATGGCCCGGCTACATGGGTGCGT F P A I A P L V A R L S Y G P A T W V R 140

421 GTGCCGTTGCCTGCCACATGGGAGGGGCGCTGGCTCAATATTCGCGCTGGCGGTGCAAAG V P L P A T W E G R W L N I R A G G A K 160

3´ BAB2_0134-R5 5´ TAAAAACAGTCCCGTCCTGA 481 GTTCTGCTTATCCTGGCGCTGGCGGCAAGAGCGCGCGCCATTTTTGTCAGGGCAGGACTG V L L I L A L A A R A R A I F V R A G L 180

541 AAAACAAACCATGACTTTTCAGGGTTTCTGCGCCTTGGCGATCCAGCCAGTGTTCGGCGC K T N H D F S G F L R L G D P A S V R R 200

601 TGGTTGCCTGAGCTTCTGGCCCGGGCAACACCGGACTGCCTGATAATGCTCCATCCCGGT W L P E L L A R A T P D C L I M L H P G 220

5´ BAB2_0134-R2 3´ AAGCTCGTCGCTGTCCTA 5` BAB2_0134-F3 661 GATGGCGCAGATCCTGTGCAATGCGCCGGGCACGCGGCTGGCAGCCGCGCCATCGAAGCC D G A D P V Q C A G H A A G S R A I E A 240

161

Chapter 2 Supplemental material

Start BAB2_0135 3` Stop BAB2_0134 721 CGAATCCTGTTTGAAAGTCCGATAAAATGATGAAGAACCGTGCTATCGTCTGGGGCCTGT R I L F E S P I K * 249

TTCTGCTGCTTTTTGTGCGGCTCCTCGCGATGATCTGGGTTCCCCTGACAGACCCGACCG

AGGCACGCTATGCCGAAATTGCCCGCAAGATGTTTGAAACGGGCAATTGGATCACGCCCC

AATTCGATTATGGGGTGCCGTTCTGGGCCAAGCCGCCACTGCATACATGGCTGTCGGCGG

CGGGCATCGCCATTTTCGGCACCACAGCCTTTACCGCGCGACTGGGAATATTGCTGGCAA

GCCTGGCAACATTGACCATCCTGTGGCAATGGGCCTGCACACTGACCGACCGCCGCACTG

CGACAATTGCCGTATTGGTTGCCGCCAGTTCAGGCCTGTTCTATGTCTCTGCCGTATTCG ACAGAGACGGCATAAGC 3´ BAB2_0134-R4 5´ TCC AGG

Figure S2.2. Sequence of BAB2_0134 and its upstream and downstream regions. Start and stop codons are in green and red characters respectively. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids deleted in BAB2_0134 mutant.

162

Chapter 2 Supplemental material

5´ BAB2_0135-F1 3´ AACACCGGACTGCCTGATAATGCTCCATCCCGGTGATGGCGCAGATCCTGTGCAATGCGC

CGGGCACGCGGCTGGCAGCCGCGCCATCGAAGCCCGAATCCTGTTTGAAAGTCCGATAAA Start BAB2_0135 1 ATGATGAAGAACCGTGCTATCGTCTGGGGCCTGTTTCTGCTGCTTTTTGTGCGGCTCCTC M M K N R A I V W G L F L L L F V R L L 20

3´ BAB2_0135-R2 5´ GATACGGCTTTAACGGGC 61 GCGATGATCTGGGTTCCCCTGACAGACCCGACCGAGGCACGCTATGCCGAAATTGCCCGC A M I W V P L T D P T E A R Y A E I A R 40

121 AAGATGTTTGAAACGGGCAATTGGATCACGCCCCAATTCGATTATGGGGTGCCGTTCTGG K M F E T G N W I T P Q F D Y G V P F W 60

181 GCCAAGCCGCCACTGCATACATGGCTGTCGGCGGCGGGCATCGCCATTTTCGGCACCACA A K P P L H T W L S A A G I A I F G T T 80

241 GCCTTTACCGCGCGACTGGGAATATTGCTGGCAAGCCTGGCAACATTGACCATCCTGTGG A F T A R L G I L L A S L A T L T I L W 100

301 CAATGGGCCTGCACACTGACCGACCGCCGCACTGCGACAATTGCCGTATTGGTTGCCGCC Q W A C T L T D R R T A T I A V L V A A 120

361 AGTTCAGGCCTGTTCTATGTCTCTGCCGTATTCGTCCAGACCGATATGGTGCTGACGCTT S S G L F Y V S A V F V Q T D M V L T L 140

421 GGGGTCACCGCCAGCATGGCGGGTTTCTACAACGGGTTGGCAGGAAGCCGCCGCTGGGGC G V T A S M A G F Y N G L A G S R R W G 160

3´ BAB2_0135-R5 5´ GAAAGACCCTGACCGTTAGC 481 TGGCTCTTCTTTCTGGGACTGGCAATCGGCTTGCTGGCCAAGGGGCCTGTGGCCGTGGTT W L F F L G L A I G L L A K G P V A V V 180

541 CTTTCCGCGACACCGATTGCGGTGTGGATGCTTTGGCGCGGCAATTGGCGGGATCTGAAA L S A T P I A V W M L W R G N W R D L K 200

601 CATTTGCCCTGGGCGGGCGGCCTGACACTATGCGCAGTGCTTGTCATACCGTGGTATGTG H L P W A G G L T L C A V L V I P W Y V 220

661 GCTGCGGAAATCGCCCCTCCCGGTTTCCTCAAATATTTCCTGATAGGCGAACATATCCAG A A E I A P P G F L K Y F L I G E H I Q 240

721 CGCTTCTTGCAGCCCGGCTGGTCGGGCGATCTTTATGGCGCAGGCCGCGAACATGCCAGA R F L Q P G W S G D L Y G A G R E H A R 260

781 GGATCGATCTGGCTGTTCTGGATCGTGGCCACATTGCCGTGGAGCCCGCTGCTGCCCGTT G S I W L F W I V A T L P W S P L L P V 280

841 CTCATCTGGCGGTTAAGAAAGAACGGTATGCCCGATGGCAAAGGGCTTCATCTCTACCTG L I W R L R K N G M P D G K G L H L Y L 300

901 CTCCTGTTCGCGCTGACACCGTTGGTGTTCTTTACCCCGGCGGCCAATATCCTCCTTGCC L L F A L T P L V F F T P A A N I L L A 320

961 TATGTGCTGCCGGGCGTGCCGGCGGCAGCACTGCTTGCCGTTATATTGTGGACCCGAACC Y V L P G V P A A A L L A V I L W T R T 340

163

Chapter 2 Supplemental material

1021 GGCCCCGCTGGCGCGGGATGGCTCAAGCTGGGGGTGGTCGAGGTTCTGCTGATCTTTCTT G P A G A G W L K L G V V E V L L I F L 360

1081 GTCATCACCATCGGCTCATTCTTTGGGCTGGGCCGGTCGTTTCTGCCCACGCAAGCGCCC V I T I G S F F G L G R S F L P T Q A P 380

1141 CTGATTGCAAGCTATTCCGGCCCGGGGCGTCTGGCGATTCTGGGCGGGCGCAGCTTTTCG L I A S Y S G P G R L A I L G G R S F S 400

1201 GCCGAGTTCTATACGCAAGACAAGATCGTCCGGTTTCAAACGCTGAATGAACTGGCCGCC A E F Y T Q D K I V R F Q T L N E L A A 420

1261 TGGCTCAAGCCGGGTGACGGGGTGCTGGTGCCCAAAAGTGCATATGATGCCTTTGCCGCC W L K P G D G V L V P K S A Y D A F A A 440

5´ BAB2_0135-R2 3´ CTATGCCGAAATTGCCCG 5` BAB2_0135-F3 3` 1321 CGGTTTGGAAATGCGGTCAAACCGATTTCCGAAGATCGGAAATATCTGCTGTTCATTCCC R F G N A V K P I S E D R K Y L L F I P 460 Stop BAB2_0135 1381 ATACCCGCTGCATCTGTCGCAAAAGGCACGGGTGCCGCCGCCGGTGCGGTGATATGAAGG I P A A S V A K G T G A A A G A V I * 478

CGATGCAAACGGAGATGAGCGCGCGCTTTTTGCTGGAGGCGACAACGACGCATTTGCCGA

AACCGAAGAACATCATTACGCCACGCTGGACGAAACGGTGGCGGCTACATAA TTTGCCACCGCCGATGTATT 3´ BAB2_0135-R4 5`

Figure S2.3. Sequence of BAB2_0135 and its upstream and downstream regions. Start and stop codons are in green and red characters respectively. Blue characters denote intergenic nucleotides. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids deleted in BAB2_0135 mutant.

164

Chapter 2 Supplemental material

5´ BAB2_0105-F1 3´ GCGTGTTCTACAGCCATGAAATGAATGTTTCTTTAAATG

TTTATCAATAGTGGCCCGATAGATAGCATTGTAAGAATATTGAAAATAGCCAGGGACGAA Start BAB2_0105 1 ATGTCGGCTCTACCTGATCATATTGAAGTAATTTCTAAAACATCTGATCTCAGGTTTGAA M S A L P D H I E V I S K T S D L R F E 20

3´ BAB2_0105-R2 5´ GTGAAGGATGTAAAGCCGCC 61 GACATCGATATCGTCGTTACACTTCCTACATTTCGGCGGCCCGATCATCTCATCCGTACC D I D I V V T L P T F R R P D H L I R T 40

121 CTTGATACGCTGAAAGCGCAGAAGACCGAAAAGCGCTTTGCGGTGATCGTGATCGAAAAT L D T L K A Q K T E K R F A V I V I E N 60

181 GAGGCTGAAAAGAGGGAAGGGGCCGAGGTGGCCGCTCCGCTGTTTGCGGATGGCACGTAC E A E K R E G A E V A A P L F A D G T Y 80

241 CGGGGAATGCTGATTGTCGAGTCGCATCGCGGCAATTGCAATGCCTATAATGCGGGCTGG R G M L I V E S H R G N C N A Y N A G W 100

301 CTTACGGCCATGACCTGTTTTCCGAACTTCAAATATATCCTCGTCATCGATGATGATGAA L T A M T C F P N F K Y I L V I D D D E 120

361 CTGGCCGATCCGATGTGGATCGAGAACATGGTTTCCGCGGCCGGGCGGTTCGATGCAAGC L A D P M W I E N M V S A A G R F D A S 140

3´ BAB2_0105-R5 5´ GGGCGTTATAGGCCAGAAAC 421 CTCGTCGGAGGCCCGCAATATCCGGTCTTTGAAAAGCCGAATTCGCAACAATGGGCCAAA L V G G P Q Y P V F E K P N S Q Q W A K 160

481 CACCCGGTTTTTATGCCGCCTTATACAAAGACAGGCGCTGTTCCGATCATCTATTCGTCG H P V F M P P Y T K T G A V P I I Y S S 180

541 GGAAACCTGCTGATCGCGCGTCCTGTGCTGGAGGCCATCGGCTATCCGTTCCTGGACCTG G N L L I A R P V L E A I G Y P F L D L 200

601 ATGTTCAATTTTACCGGCGGCGGCGATTCCGATTTCGTGGGCCGCTCCAAGGCGAAAGGT M F N F T G G G D S D F V G R S K A K G 220

661 TTCCGGATTGCCTGGTGCGCGGAAGGTATCGTGCGTGAAACCATTCCGGCGCGCCGTCTG F R I A W C A E G I V R E T I P A R R L 240

721 GAAGGCGACTGGATCAGGGCGCGGGGCTTGCGCAACGGCGTTCTTTCCACTCTTACTGAA E G D W I R A R G L R N G V L S T L T E 260

781 CAGCGTCGGCGCAAGGATGAGCCTTGCGGGCAGATACGTGTGTTCCTGAAGAGCCTTGCG Q R R R K D E P C G Q I R V F L K S L A 280

841 CTATTGGCTTATTCGCCGATCAAGGCCTTGCGGAGCGGTCTTGCGGCAGGCTTTGCGCCG L L A Y S P I K A L R S G L A A G F A P 300

5´ BAB2_0105-R2 3´ CACTTCCTACATTTCGGCGG 5` BAB2_0105-F3 3` 901 GCTGGAATGTATTTTATCTATGTTGGATTGGGACGGGTTTTGGCACATTTTGGATATCTG A G M Y F I Y V G L G R V L A H F G Y L 320

165

Chapter 2 Supplemental material

Start BAB2_0106 Stop BAB2_0105 961 AATGAGCAATATCGTAACCCGGACAAGAACTGACGCCCGTATGCCGGATGCGACAGGGCC N E Q Y R N P D K N * 331

ACAGGCACGGATACGCAGGATTGCTCTTGTTATAGCAAGCGTCATATTCGGC ATCGTTCGCAGTATAAGCCG 3´ BAB2_0105-R4 5`

Figure S2.4. Sequence of BAB2_0105 and its upstream and downstream regions. Start and stop codons are in green and red characters respectively. Blue characters denote intergenic nucleotides. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids deleted in BAB2_0105 mutant.

166

Chapter 2 Supplemental material

5´ wadD-F1 3´ TCTATAATGAGAGGCGGCTTTTTAATCCCGCTCT

CGATACCGTTTTCCGATATTGTCGTAATACGCAGCTTATAAATTATCACTCACTTTTACA

TTTACCGCTTATATTTGCATATGCTAAATTTCAAAAAATCGAATCTATAAAGTTTATAAG Start wadD 1 ATGCCTATATTTAAAATAATTATCGCGACCACCACGCGCAACAGACCTAAGATGCTGATT M P I F K I I I A T T T R N R P K M L I 20

61 AATCTCTATAAATCGCTGGGCGATTTAGAGATACCCTCTAATATTGATGTCGAATTCCTG N L Y K S L G D L E I P S N I D V E F L 40 3´ wadD-R2 5´ TGTTGTCCTGGTCGTGAAGA 121 ATCGTAGAAAACAACAGGACCAGCACTTCTGAAAGCTGGCTTCATGAAATCCGTTCCAGC I V E N N R T S T S E S W L H E I R S S 60

181 ATTTCCCCGTCTGCGGTGGTTTATATTTTAGAAACAAGTATCGGTATTTCCTGCGCTCGC I S P S A V V Y I L E T S I G I S C A R 80

241 AATCGTGCGCTTGATTATGCCCAGGAAGCTGGCGCCGATTTTCTGGCCTTTGTGGACGAT N R A L D Y A Q E A G A D F L A F V D D 100

301 GATGAATTTGTCGAACCCGATTGGCTGAAGCAACTTTTCGCCGAGCAACAAAGGCGGGAT D E F V E P D W L K Q L F A E Q Q R R D 120

361 CTCGATCTGGTCGGCTCTCCGGTGCGCCCTGTTCCCCAGAACAGCAAACTGAGCTTATGG L D L V G S P V R P V P Q N S K L S L W 140

421 CAAAGATTTGTCTGGTCCGGCGTGGAACGGAACGGCACGAGGGCCGAGGACCGGGCACGC Q R F V W S G V E R N G T R A E D R A R 160

481 AGGAAATGGCAGGAAAATAAAGCTGATACGATCAAGATAGCGACCGGAAGCTGGCTTGGA R K W Q E N K A D T I K I A T G S W L G 180

541 AGGATCGATTTCTTCCGCAGAACCGGCCTCAGATTCGATTCAAAACTTGGCCTGACCGGT R I D F F R R T G L R F D S K L G L T G 200

601 GGGGAAGACTGGAACCTTTGGCTTGAAGCCAAGAAGCTCGGCGCAAAAACGGGCTGGGCG G E D W N L W L E A K K L G A K T G W A 220

3´ wadD-R5 5´ AGCAGATACTTTGGCACGGA 661 CCAGATGCAATCGTCTATGAAACCGTGCCTTATTGCAGGATCAGCTTTTCCTATCATTTC P D A I V Y E T V P Y C R I S F S Y H F 240

721 CGCCGGAACCGGGATCATAACGCAACAGAGTTTACGCTTCTGTATAGCAAAAGCCCCCGA R R N R D H N A T E F T L L Y S K S P R 260

781 CGCGCGTGGATGCGGCTTCCCTCCCGCATTCTCAGTCGTGTATGGAAACTGCTAACCGCA R A W M R L P S R I L S R V W K L L T A 280

5´ wadD-R2 3´ ACAACAGGACCAGCACTTCT

5` wadD-F3 3` 841 ATCCTCACCCTGCCATTCAAGGGGGGACAGGCTCTCATTTCGCTGGCGATGGCTTTGGGC I L T L P F K G G Q A L I S L A M A L G 300

901 GGAATTGTCGGGCTGGTCCAGGCCTGCTGCGGAAAACAGCAGCTTCATTATAAGGAAACA G I V G L V Q A C C G K Q Q L H Y K E T 320 167

Chapter 2 Supplemental material

Stop wadD 961 ACCGGTTCTTAGGGCAGCTCCGACGATTCTGTTAAAACAGGGCGTCTAGTACCAG T G S * 323 TTGTCCCGCAGATCATGGTC 3´ wadD-R4 5`

Figure S2.5. Sequence of BAB1_0953 (wadD) and its upstream and downstream regions. Start and stop codons are in green and red characters respectively. Blue characters denote intergenic nucleotides. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids deleted in BaΔwadD mutant.

168

Chapter 2 Supplemental material

5´ mucR-F1 3´ ACAATGTTATCGCCCACCATCACCAATAACCGGCTTGCGCCTTAATCGGAAGTTGCGCTG

TCTGTAAGTCTCAATTTTCTTGCGGTGCCCTGTTTAATATCATTTTATTTGTCGATCTAA

GAAGAGTTGCCTATTATTAATGTAATATGGTTTGACAATTCTATTGCAAATGGCATCGTC

AATTGATATTTCCATAAGGGATCGAGTTGGGCCGGATTATGAAATACGCAGCGGCGGCAA

GGGGTGGGTTGCCATTGTCAGCCGCTGCGCGGACAACAAAAAATTAAAAAAGGAAAACTT 3´ mucR-R2 5´ Start mucR TTGCTTGCTGCTTTCGTGG 1 ATGGAAAATCTGGAAACGAACGACGAAAGCACCGAGCTGCTTTTGAGTTTGACCGCCGAT M E N L E T N D E S T E L L L S L T A D 20

61 GTCGTTGCAGCCTATGTCGGCAATAATTCCATTCGTGCAGGCGAACTCCCGGTTCTGATT V V A A Y V G N N S I R A G E L P V L I 40

121 GCTGAAGTTCATGCTGCTTTCAAGCGCCACGTCGAACGCGAAGAGGCCCCGGTCGTTGTT A E V H A A F K R H V E R E E A P V V V 60

181 GAAAAGCCGAAGCCAGCGGTGAACCCGAAGAAGTCTGTTCATGACGACTATATTGTCTGC E K P K P A V N P K K S V H D D Y I V C 80

241 CTTGAAGACGGCAAGAAGTTCAAGTCGCTGAAGCGTCACCTGGTGACCCATTACAACATG L E D G K K F K S L K R H L V T H Y N M 100

301 ACGCCTGAGCAGTACCGCGAAAAGTGGGATCTCGATCCGAACTATCCTATGGTTGCTCCG T P E Q Y R E K W D L D P N Y P M V A P 120

361 AACTATGCTGCAGCGCGCTCGCGCCTTGCCAAGAAGATGGGCCTCGGCCGCAAGCCGAAG N Y A A A R S R L A K K M G L G R K P K 140

5´ mucR-R2 3´ AACGAACGACGAAAGCACC

5´ mucR-F3 3´ 421 GACGCCTGATTCTTCAGCGAGTGAATCACGAAGCGCCTATGCGGGGAAATCAGTCCACTG D A * 142 GGTGAC

AACTGATTTCCGGT TTGACTAAAGGCCA 3´ mucR-R4 5´

Figure S2.6. Sequence of BAB1_0594 (mucR) and its upstream and downstream regions. Start and stop codons are in green and red characters respectively. Blue characters denote intergenic nucleotides. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids deleted in BaΔmucR mutant.

169

Chapter 2 Supplemental material

5´ wboX-F1 3´ AAACGATGCCCTGACCACCCCAACAAACTAAGAAACCCACACCCTGCCTTGGCGGGGCGT

TTTGCTATGGAGCATTCGCATGACCGTTCGCACGATTGACGAGATTTTCCGCGATTTCGT

GATTGATGTGTTCCGGCATCAGGGCCGTTTCATCCTTACAAGCCTGACCTTCGCGATACG 3´ wboX-R2 5´ Start wboX AGGGACTATTGGCGCAGTAA 1 TTGAAAGCGTTGCTTGAGGGCCTTTCGGCGTTCCCTGATAACCGCGTCATTCGTCTGAAC L K A L L E G L S A F P D N R V I R L N 20

61 AACGCTGACGAAGGTACAGCCAATAATATCGTCGTGACATCTTCTGTCGCGATCCCGACG N A D E G T A N N I V V T S S V A I P T 40

121 GCTGCGTATCAGGTGCTTTATATCCTGAATGTCACTCAGGAGAACACGGGGCCGGTGACG A A Y Q V L Y I L N V T Q E N T G P V T 60

181 GTATCTGGCACAATTAATCGCGATCTGGTTACGAATATCAATCAGCCAGTCCCGGCAGGA V S G T I N R D L V T N I N Q P V P A G 80

241 TACCTGATGCCTGGCATGGCGCTGCTATGCATTGATACCGGCACAGAATTGCGTTTGCTG Y L M P G M A L L C I D T G T E L R L L 100

301 TCCTATGGCGATGCAGAAGCCATCCTTGCCGCCGCTGAAGACGCAGCCGCACGGGCGGAG S Y G D A E A I L A A A E D A A A R A E 120

361 GCGGCGGCATCCGGTGTTAATCTACCCAAAATCAGCGCCGCCGATGCAGGTAAGTCTCTT A A A S G V N L P K I S A A D A G K S L 140

421 ATCGTTAATCAAGATGGGACCGGTTATGATGTTGGCCTTGCTGTAGAGTTTGCCACCGAA I V N Q D G T G Y D V G L A V E F A T E 160

481 ACTCAAGCAAAAGCGCCGGGTACGCCAGAGAACGATGGGTTAGTCGTAAACCTTCTCCGA T Q A K A P G T P E N D G L V V N L L R 180

541 ACTCTGCAACAAGGTCAGGCTCGCGGGTTTTTCTCGCTGAAAGACTTTGCGACTACAGTG T L Q Q G Q A R G F F S L K D F A T T V 200

601 ATTGACAATGGAAGCAACAGTGCAATTTCTGCGTTGCAGTCTGCCGTTAATTCAGGTGAA I D N G S N S A I S A L Q S A V N S G E 220

661 ACAGTCGTAATGCCAAAGACAGACGACAGTTATCTGTTTGATGGAGACGTTATAATCAAT T V V M P K T D D S Y L F D G D V I I N 240

3´ wboX-R5 CAGCCACGATAA 721 AACCCTGTCAAGTTGCGTGGCGCATCTGGTGGCTCAAAAATAAAGCGCGTCGGTGCTATT N P V K L R G A S G G S K I K R V G A I 260 5` CCAGCAAA 781 GGTCGTTTTTTACTTAGAGCAAATCATATAAATATTGGAGATTTCATAGAAGATGGGTCT G R F L L R A N H I N I G D F I E D G S 280

841 GAAATGACGGGGCAAGCAGCGTTGTTCCATTTCGACACTATCGAAAATGATGCATCCAGA E M T G Q A A L F H F D T I E N D A S R 300

901 TACATTCAACATACGCTTATTGAGAATGTTCGGTCTTTTAATTCCGGTGGGTTGATTGAC Y I Q H T L I E N V R S F N S G G L I D 320

170

Chapter 2 Supplemental material

961 GATAACAATGTAAAGGGGGCCGCAGCCACACCTTATGTTTGGGACATTTTAATTAGGAAC D N N V K G A A A T P Y V W D I L I R N 340

1021 GTTTATGCCTTCGGATGCCGTGGGCGTGGCATCCGCATGAGGGATGGCTTTGCGTTTCTG V Y A F G C R G R G I R M R D G F A F L 360

1081 CGCTTTGAAGATGTTGAAATTGGGTTAGGGCCGCAATATGGTGGTGTAAGCCTACCAGCA R F E D V E I G L G P Q Y G G V S L P A 380

1141 TATGAGTTTCGAAATTTTGAGGGGTTATTTCTTCGCCGCACCGAAGCCACTGGATTGGAT Y E F R N F E G L F L R R T E A T G L D 400

1201 GGATATACAGACCTTGGATACGTCCCAGATGCTGAACAGCGGGGGTTCATCTTTGCAGAC G Y T D L G Y V P D A E Q R G F I F A D 420

1261 GGAGCAGCCCTCCACATCAATAGCCTGTTTGCAGATAATAACAAGGGTGATGGCGTGTTT G A A L H I N S L F A D N N K G D G V F 440

5´ wboX-R2 3´ TCCCTGATAACCGCGTCATT

5´ wboX-F3 3´ 1321 TGCCAAAACGTCCAATACGTAGATGGAAACGATCTCAATTCATCCATCGACGGCGGAACT C Q N V Q Y V D G N D L N S S I D G G T 460

1381 GGGTTCAATTTTATCAACGTAGATCGCATAAACATCAATACGATCCGCAGTGGTGGCCGC G F N F I N V D R I N I N T I R S G G R 480

1441 CGGAATATGGCACCAGGAAATCTTAACACTGTTTCCCAAGGTATCTCTTTGAATGCAAAT R N M A P G N L N T V S Q G I S L N A N 500 Stop wboX 1501 TGTCAGACTGTAATTATAGGCAACGCAGTTACCCACAACTGGTGAAGTCACGGTTTTTAT C Q T V I I G N A V T H N W * 514

AGCCAAGCTCAGGACATTTTGGTTAATGGTCTGATATCACGTGATAATGGCGGAAGGGGG

TACGTTGCAGAGGGTTCAGCAGGGTCATCTCTCCTAAATGGGGCCGTTTTCAGAGATAAT

GTAGCAGGGAATTATTTTACAGGAGGGACAAGCGT AAATGTCCTCCCTGTTCGCA 3´ wboX-R4 5´

Figure S2.7. Sequence of BAB1_0998 (wboX) and its upstream and downstream regions. Start and stop codons are in green and red characters respectively. Blue characters denote intergenic nucleotides. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids deleted in BaΔwboX mutant.

171

Chapter 2 Supplemental material

Start BAB1_0114 1 ATGGCATCGCGAGACGACAGGAACAGACGCATCGAACCTTCCTTCGGTGGGGAAGAGAAG M A S R D D R N R R I E P S F G G E E K 20

61 CGGGAAGATGAGGTCTTTCAGGTCGATGAGGAGGATCGCACGGCTCGCTCGCAACAGCGC R E D E V F Q V D E E D R T A R S Q Q R 40

121 AGGCAAAAGCAATCCGCAGATCGTGCTCCCCGCTCTTCGCGGGGCGGTAAAAGAAAGCAA R Q K Q S A D R A P R S S R G G K R K Q 60

181 CGCGGCTTTTTCGGTTTTCTGCGCCGTTCGATCTACTGGTTTCTCGTGCTGGGCCTTTGG R G F F G F L R R S I Y W F L V L G L W 80

241 GGCGGCATCGCCTTTGCTGGCCTTCTGGTCTATTTCGCGGCCAAGATGCCGCCGACGACC G G I A F A G L L V Y F A A K M P P T T 100

301 GCCTGGGCCATTCCTGACAGGCCACCCAATGTGCGCATTGTCGATGTGAACAGCAATCTC A W A I P D R P P N V R I V D V N S N L 120 5´ 361 ATCGCCAACCGTGGCACGACGGGCGGCGAGGCCGTGAGCCTGCATGAAATGTCGCCCTAT I A N R G T T G G E A V S L H E M S P Y 140

BAB1_0114-F3 3´ 421 ATTCCCCAGGCCGTTATCGCCATTGAGGACCGCCGCTTCAATTCGCATTTCGGTATCGAC I P Q A V I A I E D R R F N S H F G I D 160

481 CCGATCGGGCTTGCGCGCGCTATCGTAACCAATGTGGTTTCGGGCCGCGCGGTGCAGGGC P I G L A R A I V T N V V S G R A V Q G 180

541 GGCTCCACGCTGACACAGCAGCTTGCCAAGAATCTCTTTCTTTCGCCTGACCGCACGCTG G S T L T Q Q L A K N L F L S P D R T L 200

601 GAGCGCAAGGTGCAGGAAGTGATGCTCGCCCTTTGGCTGGAGCATAAATATACCAAGGAC E R K V Q E V M L A L W L E H K Y T K D 220

661 CAGATATTGGAGATGTATCTGAACCGGGTTTATCTGGGTTCGGGCGCTTTCGGCGTCGAT Q I L E M Y L N R V Y L G S G A F G V D 240

5´ BAB1_0114-F1 3´ 721 GCGGCATCGCGGCGCTATTTCAACAAATCGGCCAAGGACGTGAACCTGATGGAAGCGGCA A A S R R Y F N K S A K D V N L M E A A 260

781 ACGCTTGCAGGGCTCCTGAAAGCGCCTTCGCGCCTCTCCCCTGCCCGCGACCCGGAAGCC T L A G L L K A P S R L S P A R D P E A 280

841 GCCGCAGCGCGCGCCAGGCTGGTGCTGGGCGCCATGCGCGAGGAAGGCATGATCGACGAC A A A R A R L V L G A M R E E G M I D D 300

901 AGGCAGGTGGCGATTGCCGAAAGCGAACCACCGACGCGCGCGCCTTCCTACTGGCAGGGT R Q V A I A E S E P P T R A P S Y W Q G 320

961 TCGGAAAACTATGTGGCCGACAAGGTTGTCGCCGAATTGCCCGAACTGATCGGCGAAGCA S E N Y V A D K V V A E L P E L I G E A 340

1021 AAGGATGACATCATCGTCCAGACGACCATCGATCTTAACCTCCAGCGCGCTGGCGAAGAG K D D I I V Q T T I D L N L Q R A G E E 360

1081 GCGATCAAGGATCAGATCGAGCAGAACGGCAAAAAGATGAATGCGAGCCAGGGCGCGCTT A I K D Q I E Q N G K K M N A S Q G A L 380

172

Chapter 2 Supplemental material

1141 GTTTCCATCGATAATACGGGCGCGGTGCGCGCCATGGTGGGCGGGGTGGACTATGCGGCC V S I D N T G A V R A M V G G V D Y A A 400

5´ BAB1_0114-R2 3´ GGTCAAACTGGCGCACTG 1201 AGCCAGTTTGACCGCGTGACCGATGCGCATCGCCAACCGGCTTCCTCGTTCAAGCCCTTT S Q F D R V T D A H R Q P A S S F K P F 420

1261 GTCTATCTTTCTGCGCTGGAACAGGGCTGGACACCGGATTCCGTGCGCAACGACGCGCCT V Y L S A L E Q G W T P D S V R N D A P 440

1321 GTGCGCATCGGCAAATGGACACCGAGTAATGATAACGGCAAGTATATGGGGCCAGTGACG V R I G K W T P S N D N G K Y M G P V T 460

1381 CTCGCCACCGCACTTTCCCATTCGCTGAACTCGGTCGCAGCCCAGCTTGTGATGGAAGTC L A T A L S H S L N S V A A Q L V M E V 480

1441 GGCCCCCAAACCGTGATCGACACGGCCCACCGCCTCGGTGTGCAGTCGAAGCTTGAACCC G P Q T V I D T A H R L G V Q S K L E P 500

1501 AATGCTTCGCTGGCGCTAGGCACGTCGGAAGTGACACTGCTTGAATTAACCGACGCCTAT N A S L A L G T S E V T L L E L T D A Y 520

1561 GTTCCCTTCGCCAATGGCGGCTATCGCGCGCCGGTCTATTTCATTACCAAGGTGACGGAT V P F A N G G Y R A P V Y F I T K V T D 540

1621 TCCGATGGAAACGTGCTTTACCAGAAGGAAGATGGTGTCGGCCCGCGCGTGATCGACGAG S D G N V L Y Q K E D G V G P R V I D E 560

1681 CGCAATGTCGGCATGATGAACGCCATGCTGCGCCGCACCGTGGAAGATGGCACCGCCAAA R N V G M M N A M L R R T V E D G T A K 580

1741 CGCGCCGCCTTCGGCTGGCCGGCCGCAGGCAAAACCGGCACTAGCCAGAATTTCCGCGAT R A A F G W P A A G K T G T S Q N F R D 600

1801 GCCTGGTTCATCGGCTACACAGCCAATCTTACGACCGGCGTTTGGTTCGGCAATGACGAC A W F I G Y T A N L T T G V W F G N D D 620

1861 GGCAAGGGCATGAAGCGCGTTTTTGGCGCAACACTTCCCGTGGCGGCCTGGAAAAGCTTC G K G M K R V F G A T L P V A A W K S F 640

1921 ATGAAGGAAGCACATAAGGGTGTTCCCATCGCCGAGCTTCCCGGCCATTACGATATCCAG M K E A H K G V P I A E L P G H Y D I Q 660

1981 AACGTGCTGCCCGGCGGCAGCGACGGCATAGACCCCAATGCGCCTTATGATCCCGGCATG N V L P G G S D G I D P N A P Y D P G M 680

2041 ATGCAGCAGGATAATTATGGCAACCAGCCGATGGCAGGCACCGGCGACGATGGCTACTGG M Q Q D N Y G N Q P M A G T G D D G Y W 700

2101 CCACCGGCGCCGGATTTGCCCGGCCGAAGCGGTATCCAGCAGGGCGAAACCGGACAGCCC P P A P D L P G R S G I Q Q G E T G Q P 720

2161 CTTGCCTCTTCCGCCAGGAATAACGGCGATTACCGGCCCATGCCTGCGGGCAATGTGGGC L A S S A R N N G D Y R P M P A G N V G 740

2221 GAACAGCCGCAAGCCCAGCCTGCGAGAAAATCCACCACCCTGCTCGATATCATCATGGGT E Q P Q A Q P A R K S T T L L D I I M G 760

173

Chapter 2 Supplemental material

Stop BAB1_0114 2281 AATTCACAGTAA N S Q * 763

Figure S2.8. Sequence of BAB1_0114. Start and stop codons are in green and red characters respectively. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids inserted in the BAB1_0114 mutant.

174

Chapter 2 Supplemental material

5´ BAB1_0417-F3 3` GCCGGCCTGTTTCCCGACCAGCTTGCTGATGAATCTATTCCTATCGCAATTGTCAGAACA

AATCAAGAACAAAAATAGGGGCACCCTTTCCGGCCGGGTAAACCTCAGCTATGAACGGGC

Start BAB1_0417 1 ATGTTGTCCGTGATTATCGAAACGCTCAATTCAGAACGTCCACTCGCCTGCACGCTGTCA M L S V I I E T L N S E R P L A C T L S 20

5´ BAB1_0417-F1 61 GGGCTGGTGCCGGCTGTGGTGGAAGGGCTGCTGCGGCGCGTGACGGTGATCGACCATGGC G L V P A V V E G L L R R V T V I D H G 40

3´ 121 TCGGGTGACGGCACGGCGCTTGTCGCCGAAGGTGCCGGTTGCGCTTTCCATGGGCTTGCG S G D G T A L V A E G A G C A F H G L A 60

181 GAAAGGGATGTCGCGCTCAATGGAATCCGCACGGATTGGGTGCTTTTCCTGAAGCCCGGT E R D V A L N G I R T D W V L F L K P G 80

241 GCGATGCCGCAGGAGGGCTGGGGGGAGGCCGTGCGCCGCCATATGGAAGGCGGGGGAGGC A M P Q E G W G E A V R R H M E G G G G 100

301 CCGGCCCGTTTCTCGCTGCCCGAAGAAGGCAGGCCTGGAAATATCAGGAAAATATTTGGC P A R F S L P E E G R P G N I R K I F G 120

5´ BAB1_0417-R2 3´ CCGAAGACCAGTCCGAACT 361 GGCGGGCCGTCGC]TTGATGCCGGGCTTCTGGTCAGGCTTGATCTGGTGCAGCCGCTTTTT G G P S L D A G L L V R L D L V Q P L F 140

421 CTGGAAGGCATAGGATCTGATGATCTGCCGCGGCGGCTGAAACCGTTGCGGCTGAAGCAT L E G I G S D D L P R R L K P L R L K H 160

Stop BAB1_0417 481 CGTATTCTGCCGCCTGAACGGCACTGA R I L P P E R H * 168

Figure S2.9. Sequence of BAB1_0417 and its upstream region. Start and stop codons are in green and red characters respectively. Blue characters denote intergenic nucleotides. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids inserted in the BAB1_0417 mutant.

175

Chapter 2 Supplemental material

Start BAB1_1465 5´ BAB1_1465-F3 3` 1 ATGAACCGTCTGGCCAGCAGCACGATCCTTGCACTCGGCGCCTTTCTTCTCTGCGCCTGC M N R L A S S T I L A L G A F L L C A C 20

61 ACGACAGTGGATTACGACTTTTCCGATGTGAAGAAATCCCGCAGTTCCGTAGCGCGCATC T T V D Y D F S D V K K S R S S V A R I 40

121 ACCCCGCCTTCTGGCCCTGTCAAGGCGCCCCGCTTCGGCGACCGCGATCCGCATGAATGG T P P S G P V K A P R F G D R D P H E W 60

5´ BAB1_1465-F1 3` 181 ACAGGCAAAACGCCGTGGCATTATCCCATCCATGGCACGGACGTCTCAAAATACCAGAGC T G K T P W H Y P I H G T D V S K Y Q S 80

241 GATGTGGACTGGAGCGCGGTACGCGCCAGCGGCATTTCGTTTGCCTTCATAAAGGCCACC D V D W S A V R A S G I S F A F I K A T 100

301 GAAGGTGGCGACCGGGTGGACGAACGCTTCAACGAACACTGGAGCGGCACGCGCGAAGCG E G G D R V D E R F N E H W S G T R E A 120

361 CGCCTGCCACGCGGGGCCTATCATTTCTATTATTTCTGCCGCCCGGCCATTGAGCAGGCA R L P R G A Y H F Y Y F C R P A I E Q A 140

421 CGCTGGTACATCCAGAATGTCCCGCGCGAACAATCCGCCCTGCCGCCCGTGCTGGACATG R W Y I Q N V P R E Q S A L P P V L D M 160

481 GAGTGGAACCCCCATTCACCAACCTGCAAGCTGCGCCCCAATGCGGCGGTGGTGCGCAGG E W N P H S P T C K L R P N A A V V R R 180

3´ BAB1_1465-R2 5´ GAAAGGACGTTCGTCAGCTC 541 GAAATGCGCACTTTCCTGCAAGCAGTCGAGAAACATTACGGCAAACGCCCTGTCATCTAT E M R T F L Q A V E K H Y G K R P V I Y 200

601 ACGACAGTCGATTTCTTCGATGACAACGACCTGCGCCAGCTTTCGGAATATCCGTTCTGG T T V D F F D D N D L R Q L S E Y P F W 220

661 CTGCGCTCGGTCGCAGGGCACCCCGACGAGAAATACGGGCCCCATCCATGGACCTTCTGG L R S V A G H P D E K Y G P H P W T F W 240

721 CAATATACCGGAACCGGCTCCATTCCCGGCATCCGCGGCGATGCGGACATCAACACCTTT Q Y T G T G S I P G I R G D A D I N T F 260

Stop BAB1_1465 781 GCCGGTGACAGTGCATCCTGGAAAAAGTGGCTGGAAAGCAACAAAGTGCGCTAA A G D S A S W K K W L E S N K V R * 277

Figure S2.10. Sequence of BAB1_1465. Start and stop codons are in green and red characters respectively. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids inserted in the BAB1_1465 mutant.

176

Chapter 2 Supplemental material

Start BAB2_0693 1 ATGAAAATGCGTTCCACCGATAGCAGGGCCGATCAGGTTTCGGCATGGGCTCCGGTATTT M K M R S T D S R A D Q V S A W A P V F 20

61 GGCGGTTGGCGGCGTGTCGAATATCTGCTGGGCGCTGGCGTCTGGGTTGCTGCTATCATC G G W R R V E Y L L G A G V W V A A I I 40

121 TATTTCTGGCACTGGTGGCTGCAACCCGCAAACCACACATATCTCGTTGGCTCCATCCTT Y F W H W W L Q P A N H T Y L V G S I L 60

181 GTGACGGCAATTCTTGCCTGGGTCACGCTTTTGCCAGCCTATTTTATCTTCCTGTTCTTT V T A I L A W V T L L P A Y F I F L F F 80

241 CGCGCCCGGCGCCCGGTCGGACCTTTGAACCTTCCTGCTGGTAGCCGTATCGCCATGGTG R A R R P V G P L N L P A G S R I A M V 100

301 GTGACGAAGGCTCCGTCAGAACCGTTCGGGGTGATGGCCGAAACGCTCCGCGCAATGCTG V T K A P S E P F G V M A E T L R A M L 120

361 GCTCAGAATGTTCCGCATGATACCTGGCTGGCCGATGAAGATCCTTCACCGGAAACACTC A Q N V P H D T W L A D E D P S P E T L 140

421 GCCTGGTGTCGCGAACACGGCGTCTTCGTTTCCACCCGCAGAGGACGGGCCGACTATCAT A W C R E H G V F V S T R R G R A D Y H 160

481 CTCCACACCTGGCCGCGCCGGACGCGCTGCAAGGAAGGCAATCTCGCTTTTTTTTATGAC L H T W P R R T R C K E G N L A F F Y D 180

5´ BAB2_0693-F3 3` 541 CATTACGGGTACGAGCGCTATGATTTCGTCGCGCAGCTCGATGCCGATCATGTTCCGACC H Y G Y E R Y D F V A Q L D A D H V P T 200

601 GAAGGTTATCTTTTCGAGATGCTGCGCCCGTTTGCCGATCCGGCGATAGGCTATGTATCC E G Y L F E M L R P F A D P A I G Y V S 220

661 GCGCCCAGCATATGCGACCGTAACGCGCATGAAAGCTGGTCGGCGCGCGGGCGCCTCTAC A P S I C D R N A H E S W S A R G R L Y 240

5´ BAB2_0693-F1 3` 721 GCAGAAGCCAGTATGCATGGCTCACTGCAAGCCGGTTACAATAACGGCCTCGCGCCCCTC A E A S M H G S L Q A G Y N N G L A P L 260

781 TGTATCGGGTCCCACTATGCTGTCCGTACAGTTGCGTTGCGGGAAATTGGTGGTCTTGGG C I G S H Y A V R T V A L R E I G G L G 280

841 CCCGAACTTGCGGAGGATCATTCCACCACATTGATGATGAATGCCGGTGGCTGGCGTGGC P E L A E D H S T T L M M N A G G W R G 300

901 GCGCATGCACTCGATGCAATCGCCCATGGCGACGGACCGCGCACCTTTGCCGACCTGGTG A H A L D A I A H G D G P R T F A D L V 320

961 ACACAGGAGTTCCAGTGGTCGCGCAGTCTGGTGATGATCTTGTTGCAATATTCGCCCAGA T Q E F Q W S R S L V M I L L Q Y S P R 340

1021 CTTGTGCGTACCCTTCCGTTTCGGATGAAATTCCAGTTCCTGTTCTCGCAGTTCTGGTAT L V R T L P F R M K F Q F L F S Q F W Y 360

177

Chapter 2 Supplemental material

GG 1081 CCGCTGTTCTCCGCTTTCATGGCGTTGATGTTCATGCTGCCGATTATAGCGCTCGTCGCC P L F S A F M A L M F M L P I I A L V A 380

3´ BAB2_0693-R2 5´ CCTGTCTTAAAGCAACGT 1141 GGACAGAATTTCGTTGCAGTTACCTATCCGGACTTTCTGGCGCATTTTGTGCCGCAGGCC G Q N F V A V T Y P D F L A H F V P Q A 400

1201 GTCATACTTATCCTGCTGGCCTATCGCTGGCGTGCATCCGGCACCTTTCGTCCTTTTGAC V I L I L L A Y R W R A S G T F R P F D 420

1261 GCGAAAATACTGAGTTTTGAAGCGACACTTTTCCTGTTCGCGCGTTGGCCCTGGGCTCTA A K I L S F E A T L F L F A R W P W A L 440

1321 GCGGGTACGCTTGCGGCACTGAGGGACTGGGCCACGGGTTCATTCGTGGATTTCCGGGTC A G T L A A L R D W A T G S F V D F R V 460

1381 ACGCCGAAAGGTGCTTCGGAGGTGGATCCCCTTCCACTTCGTGTCCTGGCGCCATACGGC T P K G A S E V D P L P L R V L A P Y G 480

1441 GTGATGGCAATTCTCTCTATTATGCCGGTCCTGCTAATCCGGGATGCGAGCGAAAGCAGC V M A I L S I M P V L L I R D A S E S S 500

1501 CGGGGGTTCTACATCTTCGCCATTATGAATGCGCTGGCTTATGTCGTTCTGCTGCTCGTC R G F Y I F A I M N A L A Y V V L L L V 520

1561 ATTATCGTCCAGCATGCCCGCGAGAACACCGTGCGCTATCGCATGCGGTTCTATCAGCCG I I V Q H A R E N T V R Y R M R F Y Q P 540

1621 GCTCTGGCCGCCAGCCTTCTCGCGCTCATGGCCCTGCCGGCTGTCGCGACGGTCGAAAGG A L A A S L L A L M A L P A V A T V E R 560

1681 GGGCGCGACGGGATCGTGGCACTATCGTGGGGCACCCGCAGCTTCAGCCTTTTCGACGAC G R D G I V A L S W G T R S F S L F D D 580

1741 CGCTTCTCCGTCGCAGGCGCCGGTATCGGCGGGCGCGACATCCACAAAGTCATCTTCAAT R F S V A G A G I G G R D I H K V I F N 600 Stop BAB2_0693 1801 CCGCACTGGCGCGTCGGCGCGACCGGCAATCCTGATCAGAAATAG P H W R V G A T G N P D Q K * 614

Figure S2.11. Sequence of BAB2_0693. Start and stop codons are in green and red characters respectively. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids inserted in the BAB2_0693 mutant.

178

Chapter 2 Supplemental material

Start BAB1_0953 1 ATGCCTATATTTAAAATAATTATCGCGACCACCACGCGCAACAGACCTAAGATGCTGATT M P I F K I I I A T T T R N R P K M L I 20

61 AATCTCTATAAATCGCTGGGCGATTTAGAGATACCCTCTAATATTGATGTCGAATTCCTG N L Y K S L G D L E I P S N I D V E F L 40

5´ BAB1_0953-F3 3` 121 ATCGTAGAAAACAACAGGACCAGCACTTCTGAAAGCTGGCTTCATGAAATCCGTTCCAGC I V E N N R T S T S E S W L H E I R S S 60

181 ATTTCCCCGTCTGCGGTGGTTTATATTTTAGAAACAAGTATCGGTATTTCCTGCGCTCGC I S P S A V V Y I L E T S I G I S C A R 80

241 AATCGTGCGCTTGATTATGCCCAGGAAGCTGGCGCCGATTTTCTGGCCTTTGTGGACGAT N R A L D Y A Q E A G A D F L A F V D D 100

5´ BAB1_0953-F1 3` 301 GATGAATTTGTCGAACCCGATTGGCTGAAGCAACTTTTCGCCGAGCAACAAAGGCGGGAT D E F V E P D W L K Q L F A E Q Q R R D 120

361 CTCGATCTGGTCGGCTCTCCGGTGCGCCCTGTTCCCCAGAACAGCAAACTGAGCTTATGG L D L V G S P V R P V P Q N S K L S L W 140

421 CAAAGATTTGTCTGGTCCGGCGTGGAACGGAACGGCACGAGGGCCGAGGACCGGGCACGC Q R F V W S G V E R N G T R A E D R A R 160

481 AGGAAATGGCAGGAAAATAAAGCTGATACGATCAAGATAGCGACCGGAAGCTGGCTTGGA R K W Q E N K A D T I K I A T G S W L G 180

541 AGGATCGATTTCTTCCGCAGAACCGGCCTCAGATTCGATTCAAAACTTGGCCTGACCGGT R I D F F R R T G L R F D S K L G L T G 200

601 GGGGAAGACTGGAACCTTTGGCTTGAAGCCAAGAAGCTCGGCGCAAAAACGGGCTGGGCG G E D W N L W L E A K K L G A K T G W A 220

3´ BAB1_0953-R2 5´ AGCAGATACTTTGGCACGGA 661 CCAGATGCAATCGTCTATGAAACCGTGCCTTATTGCAGGATCAGCTTTTCCTATCATTTC P D A I V Y E T V P Y C R I S F S Y H F 240

721 CGCCGGAACCGGGATCATAACGCAACAGAGTTTACGCTTCTGTATAGCAAAAGCCCCCGA R R N R D H N A T E F T L L Y S K S P R 260

781 CGCGCGTGGATGCGGCTTCCCTCCCGCATTCTCAGTCGTGTATGGAAACTGCTAACCGCA R A W M R L P S R I L S R V W K L L T A 280

841 ATCCTCACCCTGCCATTCAAGGGGGGACAGGCTCTCATTTCGCTGGCGATGGCTTTGGGC I L T L P F K G G Q A L I S L A M A L G 300

901 GGAATTGTCGGGCTGGTCCAGGCCTGCTGCGGAAAACAGCAGCTTCATTATAAGGAAACA G I V G L V Q A C C G K Q Q L H Y K E T 320

Stop BAB1_0953 961 ACCGGTTCTTAG T G S * 323

Figure S2.12. Sequence of BAB1_0953. Start and stop codons are in green and red characters respectively. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids inserted in the BAB1_0953 mutant. 179

Chapter 2 Supplemental material

Start BAB1_0932 1 ATGGTAAGGCTTATCGGATATTTCTTGGGGATAGGCACCGTTTTGATGCTGTTGGTGGCA M V R L I G Y F L G I G T V L M L L V A 20

61 GGCGGTGCGGCGCTTTATATTGGAAGCCTCGCGAAGGATTTGCCGGATTATGAGGTGCTC G G A A L Y I G S L A K D L P D Y E V L 40

121 GCCAAGTACGAGCCACCGGTGATGACGCGAGTTCATGCTTCCGATGGCAGCCTGATGGCC A K Y E P P V M T R V H A S D G S L M A 60

181 GAGTTTGCCCGTGAGCGACGCCTTTATCTGCCGATTCAGGCCGTGCCGAATCGCGTGAAG E F A R E R R L Y L P I Q A V P N R V K 80

241 GCGGCTTTCGTTTCCGCCGAAGACAAGACCTTTTATGAGCATCATGGCCTCGATTTCGGT A A F V S A E D K T F Y E H H G L D F G 100

301 GGCCTTGCGCGTGCGGTGGTCACGAATCTGAAGAATATAGGTTCTGGCCGCCGTCCGGTC G L A R A V V T N L K N I G S G R R P V 120

361 GGCGCCTCCACCATCACGCAACAGGTGGCGAAGAACTTCCTGCTGAGCTCCAAGCAGACC G A S T I T Q Q V A K N F L L S S K Q T 140

421 TATGATCGCAAGATCAAGGAAGCCATCCTTGCGATGCGCATTGAGCAGGCATATTCCAAG Y D R K I K E A I L A M R I E Q A Y S K 160

481 GATCGCATTCTTGAGCTTTACCTGAACGAGATTTTCTTTGGTCTCGGTTCCTATGGCATT D R I L E L Y L N E I F F G L G S Y G I 180

541 GCCAGTGCGGCACTGACCTATTTCGACAAGTCGGTGGGCGAACTGACCATTGCGGAAAGT A S A A L T Y F D K S V G E L T I A E S 200

601 GCCTATCTTGCGGCCCTGCCCAAGGGGCCGAACAATTACCATCCCTTCCGCCAGCCCGAG A Y L A A L P K G P N N Y H P F R Q P E 220

661 CGCGCCATCGAGCGCCGCAACTGGGTGATCGACCGCATGGCCGAGAATGGCTATATCACG R A I E R R N W V I D R M A E N G Y I T 240

721 ACTGCCGAAGCGGAAGACGCAAAGAAGCAGCCCCTTGGCGTCACCCCCCGCACGGCTACC T A E A E D A K K Q P L G V T P R T A T 260

781 CATTATCTCTTTGCTTCCGAATATTTCACCGAAGAAGTGCGCCGCCAGATCATTCAGAAA H Y L F A S E Y F T E E V R R Q I I Q K 280

841 TACGGTGTCGATGCGCTTTACGAAGGCGGCTTGTCGGTTCGCACGACCCTCAATCCCAAA Y G V D A L Y E G G L S V R T T L N P K 300

901 TTGCAGCTTGAAGCGCGCAAATCCTTGCAGGCGGCATTGATACGCTATGACGAGGCGCGC L Q L E A R K S L Q A A L I R Y D E A R 320

961 GGCTGGCGCGGCCCGTTGAAAAATGTCGAGCTTGGCGGCGATTGGGGTACGGCTTTCGGC G W R G P L K N V E L G G D W G T A F G 340

5´ BAB1_0932-F1 3` 1021 AATATGGAGCCCTATTCGGATGTGCCGGAATGGCAGCTTGCCGTCGTCCTGAATGTTACG N M E P Y S D V P E W Q L A V V L N V T 360

1081 CCTGCCGGTGCTGATATCGGCCTCCAGCCGCAGTTTGACGCGTCCGGTTCACGCAGCAAG P A G A D I G L Q P Q F D A S G S R S K 380

180

Chapter 2 Supplemental material

1141 GAGCGCAAGCGCGCTTTCATCGCTGCCGACGATATGAAATGGGCAATGCGCATCATCAAT E R K R A F I A A D D M K W A M R I I N 400

1201 ATCGGCGGAAAGCGCACCAGTGCCAAGTCGCCGGAAGGCGTGCTGAAGCCGGGCGATATT I G G K R T S A K S P E G V L K P G D I 420

1261 ATCTATGTGTCAAAGACGGGCGATTCCCGGTATAGTCTCCAGCAGCCGCCGAAGCTTGAA I Y V S K T G D S R Y S L Q Q P P K L E 440

1321 GGAGCACTGGTTGCCATGGACCCCCATACGGGCCGTGTGCTGGCCATGGTTGGCGGTTTT G A L V A M D P H T G R V L A M V G G F 460

1381 TCCTTTGGCGAATCGGAGTTTAATCGCGCCACGCAGGCCTATCGCCAGCCGGGTTCGTCC S F G E S E F N R A T Q A Y R Q P G S S 480

3´ BAB1_0932-R2 5` CCGACGTGACCTATTACCG 1441 TTCAAGCCGTTTGTCTATGCGGCTGCACTGGATAATGGCTATACGCCTGCTTCCGTGGTG F K P F V Y A A A L D N G Y T P A S V V 500

1501 CTGGACGGGCCGCTGGAAATTGACCAGGGCGGGTCGCTCGGTATCTGGGCTCCGAAAAAC L D G P L E I D Q G G S L G I W A P K N 520

1561 TTTTCGGGCAAGTTCGCCGGCCCCTCTACGCTGCGCTACGGCATCGAGCAGTCCCGTAAC F S G K F A G P S T L R Y G I E Q S R N 540

1621 GTCATGACCGTGCGTCTGGCGCAGGATATGGGCATGAAGCTCGTGGTCGAATATGCCGAG V M T V R L A Q D M G M K L V V E Y A E 560

1681 CGTTTCGGTATCTACGACAAGATGTTGCCTGTTCTCTCCATGGTGCTTGGCGCGGGCGAA R F G I Y D K M L P V L S M V L G A G E 580

1741 ACCACGGTGCTTCGTATGGTGACGGCTTATTCGATTATAGCCAATGGCGGGCAGAGCATC T T V L R M V T A Y S I I A N G G Q S I 600

1801 ACGCCATCGATGATCGACCGCATTCAGGACCGCTACGGCAAGACCGTGTTCAAGCATGAC T P S M I D R I Q D R Y G K T V F K H D 620

1861 AACCGCCAGTGCGAAGGTTGCAATGCGCAGGAATGGGCCAATCAGGACGAGCCGACACTG N R Q C E G C N A Q E W A N Q D E P T L 640

1921 ATCGACAATCGCGATCAGGTTCTCGATCCCATGACGGCCTATCAGATCACGTCGATGATG I D N R D Q V L D P M T A Y Q I T S M M 660

1981 GAAGGTGTCGTGCAGCGCGGAACGGCACAGATATTGAAGAGCCTGGACCGCCCGCTGGCG E G V V Q R G T A Q I L K S L D R P L A 680

2041 GGCAAGACCGGAACGACCAACGAGGAAAAGGATGCCTGGTTCGTGGGCTTCACACCTGAT G K T G T T N E E K D A W F V G F T P D 700

2101 CTCGTGGTTGGCGTTTTCATGGGATATGATACGCCTACTCCGCTTGGCCGCGGCAATACG L V V G V F M G Y D T P T P L G R G N T 720

2161 GGCGGTGGCCTTGCCGCGCCGGTGTTCAAGTCCTTCATGGAGCAGGCTTTGGCTCGAACG G G G L A A P V F K S F M E Q A L A R T 740

2221 CCGAAGGTTGATTTCCGTGTGCCGGAAGGCATGACCGTCATCGCGATCGACCGCAAGACG P K V D F R V P E G M T V I A I D R K T 760

2281 GGCATGCGCGCCAATCCGGGCGATCCGAATGTCATTATGGAAGCTTTCAAGCCGGGCACC G M R A N P G D P N V I M E A F K P G T 780

181

Chapter 2 Supplemental material

2341 GGCCCGTCCGACAGCTATTCGGTCATCGGCATGGACACGTTCCGCGAAGGCTCCCCGGTT G P S D S Y S V I G M D T F R E G S P V 800

Stop BAB1_0932 2401 GCGCCGCAGTCCCCGCAGGCAACCCGCGCGATCAATTCCGGTTCCGGCGGACTTTACTGA A P Q S P Q A T R A I N S G S G G L Y * 819

Figure S2.13. Sequence of BAB1_0932. Start and stop codons are in green and red characters respectively. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids inserted in the BAB1_0932 mutant.

182

Chapter 2 Supplemental material

Start BAB1_0326 1 ATGAAAGTCGGATCGAATCTTGTCGATCTGGAGATATGCAGCGCCATATCTGCCAGATTG M K V G S N L V D L E I C S A I S A R L 20

61 AAGCGATGCGGTATGCCGCAGGACAGAATTGAAAATGCATTCAGTGCTGCACATCTCCAC K R C G M P Q D R I E N A F S A A H L H 40

121 GGCACCGGCTTTTTTGAAGAGGCGGTTCATCAGCCCGGCGTGACGGAGCGCATGGTTTAT G T G F F E E A V H Q P G V T E R M V Y 60

181 GAGGCGATCGCGGATTATCTCCAGCTTCCCTATACGGAAGAGATCCTGCCTTCGCGCATT E A I A D Y L Q L P Y T E E I L P S R I 80

241 TTCGTTTCCGGTGATGAAAATATCTGTTTTGCCCATATCCGGCAGGTGATGGTCATCGGC F V S G D E N I C F A H I R Q V M V I G 100

301 CGCGACGGCATCTCCTATCTTTATATCGCGCCGGGCGAACGGCGCATCGCGCGGCTGCGC R D G I S Y L Y I A P G E R R I A R L R 120

361 GAACATCTGAAAGCCGCGCCGCAATTGCGCGAGCGCATTCGCATCTGCACACCATCGGTT E H L K A A P Q L R E R I R I C T P S V 140

421 TTGAAAAGCATATTGAGACAGCGTCATCAGCGCGACCTCGTGATGCGGGCCCACGCCATG L K S I L R Q R H Q R D L V M R A H A M 160

481 ACGCCGCAATTGTTCTCCGCCGCACGTGTGCTTGATACGCCGCAGGCTTTTGTCATCGCC T P Q L F S A A R V L D T P Q A F V I A 180

541 ATGGCGGTCTACGCATTTCTGGGGTGTATTGTGAACTGGCCGATGAAAACCATGCTGGCG M A V Y A F L G C I V N W P M K T M L A 200

5´ BAB1_0326-F3 3` 601 CTGCATGTTGCCATGTCGCTGTTTTTCTTCGGCTGTGTTCTTATCCGCCTTTTTGCCGCC L H V A M S L F F F G C V L I R L F A A 220

661 GCTTCTGGCAAGCGGCTGCAATTCACGGAAATTGCGCCTTTCAAGCCGCGTGATCTGCCG A S G K R L Q F T E I A P F K P R D L P 240

721 GTCTATTCTATATTGGTGCCGCTCTATCGCGAAAAGGACGTGGTTGCGCAGCTTATAGCA V Y S I L V P L Y R E K D V V A Q L I A 260

5´ BAB1_0326-F1 3` 781 GCACTCAACCGGCTCAATTGGCCACGCAGCAAGCTCGACATCAAGCTTGTCTGCGAGAAG A L N R L N W P R S K L D I K L V C E K 280

841 GACGATTACGAGACCATTGCCGCGATCAGGTGCAACACTATGCCGTCCAATTTCGAGCTG D D Y E T I A A I R C N T M P S N F E L 300

901 GTCCTTGTGCCGCCCGGCGGTCCGCGCACCAAGCCCAAGGCGCTGAATTATGCGCTGCAA V L V P P G G P R T K P K A L N Y A L Q 320

961 TTTGCCCGTGGCGAGATTGTCGCCGTTTTCGATGCCGAGGATCGGCCGCATCCCGACCAA F A R G E I V A V F D A E D R P H P D Q 340

1021 TTGCTGGAGGCTTGGCAGGCCTTCCGGCGGGGCGGCAGCAAGCTTGCCTGTGTGCAGGCG L L E A W Q A F R R G G S K L A C V Q A 360

183

Chapter 2 Supplemental material

3´ BAB1_0326-R2 CGGAAACTTATA 1081 CCGCTCATTATCGGTAACTTCCGGCGCAACCTGCTCACCCGCATGTTTGCCTTTGAATAT P L I I G N F R R N L L T R M F A F E Y 380

5´ CGCCACGA 1141 GCGGTGCTGTTTCGCGGCCTTCTGCCATGGCTTGCCAGGCGCGGGCTGGTTATTCCGCTT A V L F R G L L P W L A R R G L V I P L 400

1201 GGCGGAACATCCAATCACTTTCGCCGCTCCTGTCTGGAACAGGTGGGCGGATGGGATGCC G G T S N H F R R S C L E Q V G G W D A 420

1261 TATAATGTCACCGAGGATGCCGATCTCGGCATGCGGTTGGCGCGGTTTGGCTATCGCATC Y N V T E D A D L G M R L A R F G Y R I 440

1321 GATGTCATCAGCCGCGGCACGATAGAGGATGCGCCGGAAGAACACGGGGTCTGGCACAGG D V I S R G T I E D A P E E H G V W H R 460

1381 CAGCGCACGCGCTGGATCAAGGGGTGGATGCAGACATGGCTGGTGCATGGCCGCCAGCCG Q R T R W I K G W M Q T W L V H G R Q P 480

1441 ATGAACACCTGGCGCGAGCTTGGCTGGTGGCGTTTCGTCGTGAGCCAGATTTATACGCTC M N T W R E L G W W R F V V S Q I Y T L 500

1501 GGCATCATCGGTTCGGCGCTGCTGCACCCGCTGATGCTGCTCATGCTGGCAGGGCTTTGC G I I G S A L L H P L M L L M L A G L C 520

1561 CTGCGCATGGCGTTCGGGCCGCTGACACCGCAGGGCCTGTGGCTTCTGGCGCTCGATGTC L R M A F G P L T P Q G L W L L A L D V 540

1621 ATCAATATCCTGATGGCTTATATGAGCTTCCATATGCTCGGCGCCAAGACCATGGAGCCG I N I L M A Y M S F H M L G A K T M E P 560

1681 ACGGAACTTGGCGGCTATGCCTATTTCCTTGCCATTCCCATCTACTGGGTGCTGATCTCG T E L G G Y A Y F L A I P I Y W V L I S 580

1741 CTTTCGGCATGGCGGGCTGTGTGGCAGCTGGTGCGCCAGCCTCATCTCTGGGAAAAGACC L S A W R A V W Q L V R Q P H L W E K T 600

1801 CCGCACCAGCCAAACCTGTTCTACATTCCGCTGGAGGATGCCTCTTCTGAATTCGACGCG P H Q P N L F Y I P L E D A S S E F D A 620

Stop BAB1_0326 1861 AGGAATGCCGCCCCGCGCCCGATAATGGATTGA R N A A P R P I M D * 630

Figure S2.14. Sequence of BAB1_0326. Start and stop codons are in green and red characters respectively. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids inserted in the BAB1_0326 mutant.

184

Chapter 2 Supplemental material

Start BAB1_0607 1 ATGCGTAATACGGCAGAAAAAGATCGCGGAAACGACAAGCGTAAGCCCTTCCGGGTTTCG M R N T A E K D R G N D K R K P F R V S 20

61 CGCCTGATCGGACTCGACGCCTGGATTGATTCCGCGCTCTACAATTTACATTTCCGCTTG R L I G L D A W I D S A L Y N L H F R L 40

121 GGCGAATGGTGGGAAAACATCACGATTTTCTCCCGGCGTTTTCGCGTGCGCGGCTTTCGC G E W W E N I T I F S R R F R V R G F R 60

181 CGTTTCACGGTGGAAGTGCTGGACGAAGGCTTTACGCTCGGCGTCGCCGGTTCCGTGCTG R F T V E V L D E G F T L G V A G S V L 80

241 ATGTTGATGCTGGCGCTGCCAGCCTTTGAGGAAACCAAGAAGGACTGGCGCGCGCAGGAC M L M L A L P A F E E T K K D W R A Q D 100

301 GATTATGCCGTCACCTTTCTTGACCGTTATGGCAACGAGATTGGCCGGCGCGGTATTCTG D Y A V T F L D R Y G N E I G R R G I L 120

361 CACCGGCAGGCGGTGCCTATCGATGAATTGCCGGATCATGTCATCAAGGCTGTGCTCGGC H R Q A V P I D E L P D H V I K A V L G 140

421 ACCGAGGACCGTCGCTTCTTCGACCATTACGGCATTGATCTCATGGGGCTTAGCCGAGCG T E D R R F F D H Y G I D L M G L S R A 160

481 TTAAGCCAGAACATGCGCGCCAACGGCGTGGTTCAGGGCGGCTCGACCATCACGCAGCAG L S Q N M R A N G V V Q G G S T I T Q Q 180

5´ BAB1_0607-F3 3` 541 CTTGCCAAAAACCTGTTTCTTTCCAATGAGCGCACCATCGAACGCAAGGTCAAGGAAGCC L A K N L F L S N E R T I E R K V K E A 200

601 TTTCTCGCGCTCTGGCTTGAAAGCAATCTCAGCAAGAAGGAAATCCTCCAGCTTTACCTC F L A L W L E S N L S K K E I L Q L Y L 220

661 GACCGTGCCTATATGGGCGGCGGCACTTTCGGCATTGCAGCCGCGTCCGAGTTTTATTTC D R A Y M G G G T F G I A A A S E F Y F 240

721 GGCAAGAATGTAAAGGACATCTCGCTGGCCGAGGCAGCCATGCTGGCCGGCCTGTTCAAG G K N V K D I S L A E A A M L A G L F K 260

5´ BAB1_0607-F1 781 GCTCCGGCGAAATTCGCGCCGCACGTCAACCTGCCTGCAGCTCGCGCCCGCGCCAATGTC A P A K F A P H V N L P A A R A R A N V 280 3` 841 GTTCTCTCCAACATGGTGGAAAGCGGCTTTCTCAGCGAAGGTCAGGTTGCCGTCGCTCGC V L S N M V E S G F L S E G Q V A V A R 300

901 CGCCACCCGGCCAGCGTCATAGATCGCGCCAAGGATGAAAGCCCCGACTATTTCCTGGAC R H P A S V I D R A K D E S P D Y F L D 320

961 TGGGCCTTCGACGAGGTGAAAAAGGTTGCAGACAGGTTCAACCAGCATACGCTGATCGTG W A F D E V K K V A D R F N Q H T L I V 340

1021 CGCACCACGCTTGACCGCAATATCCAGAAAGCCGCCGAAGAATCGCTTGAATTTCACCTT R T T L D R N I Q K A A E E S L E F H L 360

1081 CGCCAATATGGCAAGGAATATAACGTATCGGAGGCGGCCACCGTCGTGCTTGCCAATGAC R Q Y G K E Y N V S E A A T V V L A N D 380

185

Chapter 2 Supplemental material

1141 GGTTCCGTGCGCGCGCTCGTTGGCGGGCGCGACTATGGCGAAAGCCAGTTCAACCGCGCA G S V R A L V G G R D Y G E S Q F N R A 400

1201 ACCAGGGCGCTGCGGCAGGCCGGTTCATCCTTCAAGCCCTATGTCTATGCGGCTGCCATG T R A L R Q A G S S F K P Y V Y A A A M 420

3´ BAB1_0607-R2 5` CTTTTCCCCGACTGTGGTTC 1261 GAAAAGGGGCTGACACCAAGCACCATCGTTTCCGATGCACCGATCAGCTGGGGCAACTGG E K G L T P S T I V S D A P I S W G N W 440

1321 TCGCCGCGCAATTACGGGCGCAGTTTTGCTGGCCGGGTTGATCTCACCACCGCGCTCGTC S P R N Y G R S F A G R V D L T T A L V 460

1381 CGCTCGCTCAACAGCGTGCCGGTGCGCCTTGCCCGCGATTATCTGACCACGGCGCCCGTT R S L N S V P V R L A R D Y L T T A P V 480

1441 GTCGCGCTCACAAAGGCAATGGGTGTGGAATCGCCCATCTCCTCCCATAAGACGATGGTG V A L T K A M G V E S P I S S H K T M V 500

1501 CTCGGCACATCCGAAATGACCGTCATGGATCAGGCAACAGGCTTTAACGTCTTCGCCAAT L G T S E M T V M D Q A T G F N V F A N 520

1561 GCGGGCATGGCCGGAAACCGTCACGCTTTCACGCAAATCCTGGCATCCGACGGAAAGGTT A G M A G N R H A F T Q I L A S D G K V 540

1621 CTTTGGGATTTTGGCCGTGACGCACCAAAGCCGCACCGCGCCCTTTCGGAAAAGGCCGCA L W D F G R D A P K P H R A L S E K A A 560

1681 CTTGAGATGAACTCCATGCTGGTGCAGGTGCCCGAGCGCGGCACGGGCCGGCGCGCCGCA L E M N S M L V Q V P E R G T G R R A A 580

1741 CTTACGATGACGCGCGTGGCAGGCAAAACGGGTACGACCCAGAACTATCGCGATGCCTGG L T M T R V A G K T G T T Q N Y R D A W 600

1801 TTCGTGGGCTTTACCGGAAATTTCACCGCCGCCGTCTGGTTCGGCAACGATAATTTCACG F V G F T G N F T A A V W F G N D N F T 620

1861 CCCATGAAGGAACTGACCGGCGGTGTTCTGCCAGCCATGGCCTGGCAGCGCATGATGGCA P M K E L T G G V L P A M A W Q R M M A 640

1921 TACGCCCATCAGAATATCGAATTGAAGCCGCTGCCGGGCGTCAACCCGCCCTTCCCCGCC Y A H Q N I E L K P L P G V N P P F P A 660

1981 CAGCCGAAGAACCCGCCCGCACAGGTGGCCGATACGAGGCCGCACGAAACAATGGCTGCC Q P K N P P A Q V A D T R P H E T M A A 680

2041 CCACCACGGGTTTTATCGCCGCTGGCGACAAAAATTCTCAAGGAACTGCATGACCGCTTC P P R V L S P L A T K I L K E L H D R F 700

Stop BAB1_0607 2101 CTTGCCGCGCCGCCTCTGCCTACGATAGCCGAGCGCACAAAAGTTTCGGTACTCTGA L A A P P L P T I A E R T K V S V L * 718

Figure S2.15. Sequence of BAB1_0607. Start and stop codons are in green and red characters respectively. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids inserted in the BAB1_0607 mutant.

186

Chapter 2 Supplemental material

CGGGCTGCAGGAATT

5´ Tn7-wadD-F1 3´ GCGATTCCTTTGTGCCAGATGAAACAAAATGAAGCGCAAGCGATGTGTCTTTCATGCTCTC

AGGTCTCGTTTCTGGCAAATCTGCCGCTATTTTTTCCCGCCGCCTATTGGCTATCACATG

CAAAACATGGTTGCCATAATCTCAACATGGTGATTTTTCATTTTACAGGCGCGGTATTTA

TGCTAATCGGGCGCCTGTGCCCTTATAGCTCAGTTGGTAGAGCACCTGATTTGTAATCAG

GGGGTCGGGAGTTCGAGTCTCTCTGGGGGCACCACTTTTCGACAGATAGCGTATATAATC

5´ wadD-F1 3´ GGTCTTTTTACATTGGTATCGTCACCTCTATAATGAGAGGCGGCTTTTTAATCCCGCTCT

CGATACCGTTTTCCGATATTGTCGTAATACGCAGCTTATAAATTATCACTCACTTTTACA

TTTACCGCTTATATTTGCATATGCTAAATTTCAAAAAATCGAATCTATAAAGTTTATAAG Start wadD 1 ATGCCTATATTTAAAATAATTATCGCGACCACCACGCGCAACAGACCTAAGATGCTGATT M P I F K I I I A T T T R N R P K M L I 20

61 AATCTCTATAAATCGCTGGGCGATTTAGAGATACCCTCTAATATTGATGTCGAATTCCTG N L Y K S L G D L E I P S N I D V E F L 40 3´ wadD-R2 5´ TGTTGTCCTGGTCGTGAAGA 121 ATCGTAGAAAACAACAGGACCAGCACTTCTGAAAGCTGGCTTCATGAAATCCGTTCCAGC I V E N N R T S T S E S W L H E I R S S 60

181 ATTTCCCCGTCTGCGGTGGTTTATATTTTAGAAACAAGTATCGGTATTTCCTGCGCTCGC I S P S A V V Y I L E T S I G I S C A R 80

241 AATCGTGCGCTTGATTATGCCCAGGAAGCTGGCGCCGATTTTCTGGCCTTTGTGGACGAT N R A L D Y A Q E A G A D F L A F V D D 100

301 GATGAATTTGTCGAACCCGATTGGCTGAAGCAACTTTTCGCCGAGCAACAAAGGCGGGAT D E F V E P D W L K Q L F A E Q Q R R D 120

361 CTCGATCTGGTCGGCTCTCCGGTGCGCCCTGTTCCCCAGAACAGCAAACTGAGCTTATGG L D L V G S P V R P V P Q N S K L S L W 140

421 CAAAGATTTGTCTGGTCCGGCGTGGAACGGAACGGCACGAGGGCCGAGGACCGGGCACGC Q R F V W S G V E R N G T R A E D R A R 160

481 AGGAAATGGCAGGAAAATAAAGCTGATACGATCAAGATAGCGACCGGAAGCTGGCTTGGA R K W Q E N K A D T I K I A T G S W L G 180

541 AGGATCGATTTCTTCCGCAGAACCGGCCTCAGATTCGATTCAAAACTTGGCCTGACCGGT R I D F F R R T G L R F D S K L G L T G 200

601 GGGGAAGACTGGAACCTTTGGCTTGAAGCCAAGAAGCTCGGCGCAAAAACGGGCTGGGCG G E D W N L W L E A K K L G A K T G W A 220

661 CCAGATGCAATCGTCTATGAAACCGTGCCTTATTGCAGGATCAGCTTTTCCTATCATTTC P D A I V Y E T V P Y C R I S F S Y H F 240

721 CGCCGGAACCGGGATCATAACGCAACAGAGTTTACGCTTCTGTATAGCAAAAGCCCCCGA R R N R D H N A T E F T L L Y S K S P R 260

187

Chapter 2 Supplemental material

781 CGCGCGTGGATGCGGCTTCCCTCCCGCATTCTCAGTCGTGTATGGAAACTGCTAACCGCA R A W M R L P S R I L S R V W K L L T A 280

5` wadD-F3 3` 841 ATCCTCACCCTGCCATTCAAGGGGGGACAGGCTCTCATTTCGCTGGCGATGGCTTTGGGC I L T L P F K G G Q A L I S L A M A L G 300

901 GGAATTGTCGGGCTGGTCCAGGCCTGCTGCGGAAAACAGCAGCTTCATTATAAGGAAACA G I V G L V Q A C C G K Q Q L H Y K E T 320 Stop wadD 961 ACCGGTTCTTAGGGCAGCTCCGACGATTCTGTTAAAACAGGGCGTCTAGTACCAGCCGAC T G S * 323 TTGTCCCGCAGATCATGGTC 3´ wadD-R4 5`

GGCAACCTGCGCTTCTTCGGACATACGGTCCGGTGTCCATGGCGGATCGAAGGTCATGGT

GGCCTCCACGAACGAAACGCCCTCCACGGCGCTGACCGCATTTTCCACCCAGCCCGGCAT

TTCGCCGGCGACCGGGCAGCCGGGAGCCGTCAGCGTCATTTCGATCTTGACCGTGCGGTC

ATCCTCAATATCGATCTTGTAGATGAGGCCGAGTTCGTAGATATCCGCCGGAATTTCCGG

ATCATAGACGGTCTTCAATGCGGCGATGAT CAGAAGTTACGCCGCTACTA TAAGGAGCTCTTCG 3´ Tn7-wadD-R2 5`

Figure S2.16. Sequence of wadD and its upstream and downstream regions. Start and stop codons are in green and red characters respectively. Blue characters denote intergenic nucleotides. Primers used for complementation are in bold characters. Underlined nucleotides at the 5’ end correspond to 15 bases that are homologous to the pUC18 R6KTmini Tn7T KmR linearized vector with EcoRI, to which it will be joined.

188

Chapter 2 Supplemental material

A B C

Figure S2.17. Polymyxin B sensitivity of Ba-parental (A), BaΔwadD (B) and the complemented strain BaΔwadD::Tn7-PwadD (C).

189

Chapter 2 Supplemental material

Figure S2.18. Hypothetical role of mucR regulation in the synthesis of Brucella LPS.

190

Chapter 2 Supplemental material

Table S2.1. Oligonucleotides.

Oligonucleotide Sequence 5`  3`

BAB2_0133-F1 GCGTTGGACAAGTTGAGGTT BAB2_0133-R2 CATAGCGGTCGGTTAAATGC BAB2_0133-F3 GTATCGCCAGCCAATTTACGTCCGTATTGGAAGCCAAGAA BAB2_0133-R4 CAGTAACAAAAGGCCGCTAT BAB2_0133-R5 AAGACCCAGTAGTTAGCACT BAB2_0134-F1 TGCGGGGTACCAGTTATTACA BAB2_0134-R2 TAGGACAGCGACGAGCTT BAB2_0134-F3 AAGCTCGTCGCTGTCCTAATCGAAGCCCGAATCCTG BAB2_0134-R4 GGACGAATACGGCAGAGACA BAB2_0134-R5 AGTGCTGCCCTGACAAAAAT BAB2_0135-F1 TGGCGGCCGCTCTAGAACACCGGACTGCCTGATAA * BAB2_0135-R2 CGGGCAATTTCGGCATAG BAB2_0135-F3 CTATGCCGAAATTGCCCGCCGGTTTGGAAATGCGGTCAA BAB2_0135-R4 ATCCACTAGTTCTAGTTATGTAGCCGCCACCGTTT * BAB2_0135-R5 CGATTGCCAGTCCCAGAAAG BAB2_0105-F1 GCGTGTTCTACAGCCATGAA BAB2_0105-R2 CCGCCGAAATGTAGGAAGTG BAB2_0105-F3 CACTTCCTACATTTCGGCGGTATGTTGGATTGGGACGGGT BAB2_0105-R4 GCCGAATATGACGCTTGCTA BAB2_0105-R5 CAAAGACCGGATATTGCGGG BAB1_1465-F1 GGCACGGACGTCTCAAAATA BAB1_1465-R2 CTCGACTGCTTGCAGGAAAG BAB1_1465-F3 CCAGCAGCACGATCCTTG wadD-F1 TGGCGGCCGCTCTAGTCTATAATGAGAGGCGGCTTTT * wadD-R2 AGAAGTGCTGGTCCTGTTGT wadD-F3 ACAACAGGACCAGCACTTCTATCCTCACCCTGCCATTCAA wadD-R4 ATCCACTAGTTCTAGCTGGTACTAGACGCCCTGTT * wadD-R5 AGGCACGGTTTCATAGACGA mucR-F1 ACAATGTTATCGCCCACCAT mucR-R2 GGTGCTTTCGTCGTTCGTT

191

Chapter 2 Supplemental material

Oligonucleotide Sequence 5`  3` mucR-F3 AACGAACGACGAAAGCACCTGATTCTTCAGCGAGTGAATCACG mucR-R4 ACCGGAAATCAGTTCAGTGG mucR-F5 AACGAACGACGAAAGCACC mucR-R6 GCTCAGGCGTCATGTTGTAA wboX-F1 CTAAGAAACCCACACCCTGC wboX-R2 AATGACGCGGTTATCAGGGA wboX-F3 TCCCTGATAACCGCGTCATTCTCAATTCATCCATCGACGG wboX-R4 ACGCTTGTCCCTCCTGTAAA wboX-R5 AAACGACCAATAGCACCGAC BAB1_0114-F1 TCAACAAATCGGCCAAGGAC BAB1_0114-R2 GTCACGCGGTCAAACTGG

BAB1_0114-F3 CCTATATTCCCCAGGCCGTT BAB1_0417-F1 TGATCGACCATGGCTCGG BAB1_0417-R2 TCAAGCCTGACCAGAAGCC

BAB1_0417-F3 CTGTTTCCCGACCAGCTTG BAB2_0693-F1 CACTGCAAGCCGGTTACAAT BAB2_0693-R2 TGCAACGAAATTCTGTCCGG BAB2_0693-F3 ACGAGCGCTATGATTTCGTC BAB1_0953-F1 ACTTTTCGCCGAGCAACAAA BAB1_0953-R2 AGGCACGGTTTCATAGACGA BAB1_0953-F3 GCTGGCTTCATGAAATCCGT BAB1_0932-F1 GCCGTCGTCCTGAATGTTAC BAB1_0932-R2 GCCATTATCCAGTGCAGCC BAB1_0932-F3 GGCCGAGAATGGCTATATCA BAB1_0326-F1 GCACTCAACCGGCTCAATTG

BAB1_0326-R2 AGCACCGCATATTCAAAGGC BAB1_0607-F1 GCCAATGTCGTTCTCTCCAA BAB1_0607-R2 CTTGGTGTCAGCCCCTTTTC Tn7-wadD-F1 CGGGCTGCAGGAATTGCGATTCCTTTGTGCCAGAT * Tn7-wadD-R2 GCTTCTCGAGGAATTATCATCGCCGCATTGAAGAC *

*Underlined nucleotides at the 5’ end correspond to 15 bases that are homologous to the corresponding linearized vector to which it will be joined. The 3’ end of the primers contains the sequence that is specific to the target gene.

192

Chapter 2 Supplemental material

Table S2.2. Previously identified ORFs acting on the LPS of Brucella abortus.

ORF Name Function / target Reference BAB1_0639 wadA Core glycosyltransferase Monreal et al., 2003

BAB1_0351 wadB Core glycosyltransferase Gil-Ramírez et al., 2014

BAB1_1522 wadC Core glycosyltransferase Conde-Álvarez et al., 2012, 2013

BAB2_0209 waaA Kdo formation Iriarte et al., 2004

BAB1_1171 lpxB Lipid A Iriarte et al., 2004

BAB1_0553 wbkA O-PS Godfroid et al., 2000

BAB1_0563 wbkE O-PS González et al., 2008

BAB1_1000 wboA O-PS McQuiston et al., 1999

BAB1_1000 wboB O-PS McQuiston et al., 1999

193

Chapter 2 Supplemental material

Table S2.3. ORFs coding for B. abortus hypothetical glycosyltransferases, family to which they belong, predicted function and the corresponding mutant LPS phenotype by Western-Blot analysis.

Mutant LPS reactivity ORF Family Predicted function (KEGG) O-PSa R-LPSb

BAB1_0326 2 Glycosyltransferase + +

BAB2_0133 2 Glycosyltransferase + +

BAB2_0105 2 Glycosyltransferase + +

BAB2_0693 2 Glycosyltransferase + +

BAB1_0953 2 Glycosyltransferase + -

BAB1_1620 25 Glycosyltransferase + +

Penicillin-binding protein 1A transpeptidase BAB1_0607 51 + + domain - Glycosyltransferase

Penicillin-binding protein transpeptidase BAB1_0114 51 domain: ATP/GTP-binding site motif A (P- + + loop) - Glycosyltransferase

Penicillin-binding protein 1A transpeptidase BAB1_0932 51 + + domain - Glycosyltransferase

Possible dolichyl-phosphate-mannose- BAB2_0135 83 + + protein mannosyltransferase family protein

BAB1_0417 ncc Conserved hypothetical protein + + a Reactivity to polyclonal serum from a rabbit infected with B. melitensis 16M. b Reactivity to monoclonal antibodies anti R-LPS: A68/24G12/A08 and A68/24D08/G09. c Glycosyltransferase family non-classified.

194

Chapter 2 Supplemental material

Table S2.4. Aminoacid alterations in the protein encoded by wadD in different Brucella species.

Species Altered aminoacids

B. abortus G27 S52 S59 P63 G75 L213 Y231 L278 K287 B. melitensis B. suis bv.1 B. suis bv.2

B. suis bv.5 I278 B. ovis B. canis B. microti B. D75 pinnipedialis

B. vulpis S27 N52 C59

B. ceti D75

B. inopinata S27 N52 S63 H213 H231 R287 B. inopinata S27 N52 H231 R287 BO2 NF2653 S27 N52 (Austr.)

195

Chapter 2 Annex

Annex A new gene involved in the synthesis of Brucella O-polysaccharide?

197

Annex

Chapter 2 Annex

A new gene involved in the synthesis of Brucella O-polysaccharide?

Previous investigations characterized genomic island GI-2 in B. abortus (Mancilla et al., 2010) and B. melitensis (Rajashekara et al., 2008). This island includes several ORFs that are represented in Figure S2.20.

In B. melitensis background, BMEI0998 (wboA) and BMEI0997 (wboB) have been assigned to O-PS synthesis (McQuiston et al., 1999). Deletion of either of these two ORFs or the complete GI-2 resulted in a rough phenotype. Interestingly, complementation analyses indicated that, in addition to wboA and wboB, the immediately upstream ORF (BMEI0999) was also required to revert the rough LPS phenotype and the in vitro and in vivo virulence properties of the mutants either in these two ORFs or in the one that lacks the complete GI-2 island (Rajashekara et al., 2008). Similar observations were made in Rev1, where plasmid pBGI-997-99c (containing BMEI0999, wboA and wboB), but not plasmid pMM76 (carrying only wboA and wboB) restored the smooth phenotype of mutants lacking the complete GI-2 (Mancilla et al., 2013).

Figure S2.20. Genetic organization of the ORFs in GI-2 island of B. melitensis 16M and B. abortus 2308 genomes. The arrows indicate the direction of transcription (the picture is not drawn to scale).

199

Chapter 2 Annex

All the above results, together with the observation that BMEI0999 has homology to an outer membrane autotransporter protein of Rhodopseudomonas palustris (bacteria closely related to Brucella), suggest that BMEI0999 could participate in O-PS synthesis or transport in B. melitensis (Rajashekara et al., 2008). However, its exact role in LPS synthesis has never been analysed.

Studies in B. abortus confirmed the results shown in B. melitensis and showed that spontaneous mutants lacking the complete GI-2 resulted in the same phenotype as the individual mutants in wboA and wboB (rough-LPS) (Mancilla et al., 2010) and the phenotype could not be complemented with a plasmid carrying only these two genes. This suggests again that, as is the case in B. melitensis, the upstream ORF of wboA-B in B. abortus could be important for the synthesis of a complete LPS.

To clarify this point, we first investigated the role of ORF BAB1_0998 (provisionally renamed wboX) immediately upstream wboA-wboB in B. abortus that, although annotated as a pseudogene, could encode the orthologue of BMEI0999. According to genomic databases and sequence analysis, the start of BAB1_0998 (515 aa) is different from that of BMEI0999 (483 aa) and this last ORF would lack 32 aminoacids in the amino-terminal end.

We constructed a non-polar mutant in B. abortus (BAB1_0998) by deleting the region that codes from aminoacids 18 to 451 to obtain Ba∆wboX (Figure S2.7 in Supplemental material, Chapter 2). The mutant had the same growth profile as Ba- parental (data not shown) and showed a smooth phenotype in the crystal violet assay. LPS extracted by SDS-proteinase K and analysed by SDS-PAGE and silver staining showed a similar profile to Ba-parental LPS, and reacted similarly in Western-blot with a polyclonal serum against S-Brucella (Figure 2.10) and a monoclonal antibody that specifically recognizes the O-PS (Figure S2.21). These results suggest that BAB1_0998 is not required for the synthesis of the O-PS in B. abortus.

200

Chapter 2 Annex

Figure S2.21. BAB1_0998 is not required for the synthesis of the O-PS in B. abortus. BaΔper mutant is included as a control of a LPS lacking the complete O-PS. 12% SDS-PAGE electrophoresis and silver staining of LPS samples extracted with the SDS-proteinase K protocol (A). Western-blot analysis of BaΔwboX LPS extract with a monoclonal antibody that recognizes specifically the O-PS (B).

To explain the situation in B. melitensis, the same suicide vector was used to construct a mutant in BMEI0999 but, if the genomic annotation is correct and this ORF is 32 aminoacids shorter, this mutant would only conserve the genomic region that hypothetically codifies the last 64 amino acids. When analysed with the crystal violet assay, the mutant showed a smooth phenotype similar to Bme-parental.

As a conclusion, our results in B. melitensis and B. abortus indicate that the ORF situated directly upstream wboA-wboB are not required for the synthesis of a S-LPS in these two Brucella spp.

201

Chapter 3

Chapter 3 Towards a new vaccine against B. ovis infection 1, 2

1 Studies in mice were done in cooperation with the Centro de Investigación y Tecnología Agroalimentaria de Aragón (CITA) and Instituto de Agrobiotecnología de Navarra (IdAB). 2 Double agar gel immunodiffusion tests were performed in cooperation with Centro de Investigación y Tecnología Agroalimentaria de Aragón (CITA).

205

Chapter 3 Abbreviations

Abbreviations used in Chapter 3

AGID double agar gel immunodiffusion Amp ampicillin ba2308-CAI Brucella abortus Carbonic anhydrase I ba2308-CAII Brucella abortus Carbonic anhydrase II BAB Blood Agar Base bp base pairs BLAST Basic Local Alignment Search Tool CA Carbonic anhydrase CFT complement fixation test CFU Colony Forming Units CIMA Centro de Investigación Médica Aplicada CITA Centro de Investigación y Tecnología Agroalimentaria de Aragón Cm chloramphenicol

CO2 Carbon dioxide DIVA Differentiating Infected from Vaccinated Animals DMSO Dimethyl sulfoxide DNA Deoxyribonucleic acid EMBL European Molecular Biology Laboratory Gm gentamicin HS hot-saline IdAB Instituto de Agrobiotecnología i-ELISA indirect enzyme-linked immunosorbent assay IP Intraperitoneal kb kilobase Kdo 3-deoxy-D-manno-2-octulosonic acid KEGG Kyoto Encyclopedia of Genes and Genomes Km Kanamycin LPS Lipopolysaccharide MoAbs monoclonal antibodies Nal Nalidixic acid NCBI National Centre of Biotechnology Information OD Optical Density OIE World Organization for Animal Health OMP Outer Membrane Protein O-PS O-polysaccharide ORF Open Reading Frame PAMP Pathogen-Associated Molecular Pattern PVDF Polyvinylidene fluoride PAGE Polyacrylamide Gel Electrophoresis

207

Chapter 3 Abbreviations

PBS 10 mM Phosphate buffer saline PCR Polymerase Chain Reaction PI Post-infection Pmx Polymixin B R-LPS rough lipopolysaccharide rpm revolutions per minute Sac Sucrose SC Subcutaneous SD Standard Deviation SDS Sodium Dodecyl Sulphate S-LPS smooth lipopolysaccharide TSA Tripticase Soy Agar TSA-YE-S TSA supplemented with 0.5% Yeast Extract and 5% serum TSB Tripticase Soy Broth TSB-YE-S TSB supplemented with 0.5% Yeast Extract and 5% serum WHO World Health Organization YE Yeast extract Δ deletion mutant :: insertion mutant

208

Chapter 3 Abstract

ABSTRACT

Brucella ovis causes a disease that affects sheep and produces great economic losses. The only existing vaccine is Rev1, a live attenuated B. melitensis strain which can cause abortions and infertility. When B. melitensis becomes eradicated, the use of Rev1 is banned, but in some of those areas there is still a high prevalence of B. ovis. Therefore, the development of new vaccines against B. ovis infection is essential. The core section of Brucella lipopolysaccharide (LPS) has been shown to be a crucial virulence factor in smooth Brucella strains, and also in the rough species B. ovis. Genes wadA, wadB, wadC and wadD encode core glycosyltransferases involved in its synthesis and mutants in wadB and wadC in a B. ovis PA are attenuated and protect against B. ovis infection. However, this strain requires a 5–10% CO2 atmosphere for its growth. A recent work has shown that this requirement is due to the lack of two carbonic anhydrases, present in the Brucella CO2-independent species and introduction of functional carbonic anhydrases in B. ovis PA allowed its growth in atmospheric conditions. In the present work, we analysed the presence of wadD in B. ovis and constructed mutants in wadB, wadC and wadD in a CO2-independent B. ovis PA strain. We showed that mutants in the three core glycosyltransferases presented a defect in the core LPS section and were attenuated in the murine model. In particular, mutants in wadD showed a growth defect when compared with the parental strain. Interestingly, mutants in wadB protected against B. ovis infection in mice and are currently being tested in the natural host.

209

Chapter 3 Introduction

INTRODUCTION

Brucella ovis produces a disease unique to sheep. It causes epididymitis and orchitis in rams, and is responsible for placentitis, fertility decrease and abortion in ewes and increase of neonatal mortality in lambs (OIE, 2015). The disease is transmitted by direct and venereal contact through ewes so that control policies based on male culling have been unsuccessful and lead in fact to the expansion of the disease within flocks. Transmission between rams occurs passively through venereal contact with female or, less frequently, by direct ram-to-ram transmission (Blasco, 1990). Transmission to lambs occurs as infected sheep can excrete B. ovis in vaginal discharges.

1. Diagnosis of B. ovis infection. The most efficient and widely used serological test for the diagnosis of B. ovis infection are the complement fixation test (CFT), the double agar gel immunodiffusion (AGID) test and the indirect enzyme-linked immunosorbent assay (i-ELISA). i-ELISA is more sensitive than either the CFT or AGID tests although international standardisation is lacking. Moreover, it has been demonstrated that the AGID test shows similar sensitivity to the CFT, and it is a simpler test to perform (OIE, 2015). The most widely used antigen is a hot saline (HS) extract of total cells that is enriched in rough lipopolysaccharide (R-LPS), Outer Membrane Proteins (OMP) and other outer membrane components (Gamazo et al., 1989; Riezu-Boj et al., 1986). However, this extract contains LPS components specific of B. ovis, but also additional antigenic epitopes shared with smooth Brucella (Santos et al., 1984), which account for the cross-reactivity that can be observed when using this antigen with sera from sheep infected with B. melitensis or vaccinated with B. melitensis Rev1 (Riezu-Boj et al., 1986). According to the OIE, HS antigens should be prepared from B. ovis REO198, a

CO2- and serum-independent strain (OIE, 2015).

211

Chapter 3 Introduction

2. The problem of vaccination against B. ovis infection. Animal vaccination is the only means to prevent and control brucellosis in sheep. As both rams and ewes can play a role in the transmission of infection (Grilló et al., 1999), vaccination of both is probably the most economical and practical means to control B. ovis in areas with a high prevalence of infection (OIE, 2015). Brucellosis in sheep can be also caused by B. melitensis and, indeed, B. melitensis Rev1 is considered the best available vaccine for the prophylaxis of B. melitensis and B. ovis infection in small ruminants (OIE, 2016). However, Rev1 often causes abortion and excretion in milk in pregnant animals, does not provide complete protection and interferes with the serological diagnosis of brucellosis since vaccination may lead to long-term persistence of antibodies. To avoid these problems, vaccination with Rev1 should be performed when the animals are not pregnant or during the lambing season (Blasco, 1997). Moreover, Rev1 is virulent for humans and resistant to streptomycin, an antibiotic of choice to treat brucellosis (Elberg and Faunce, 1956). A particular problem occurs in countries affected by B. ovis but free of B. melitensis, where vaccination with Rev1 is banned. In this scenario, B. ovis infection becomes an emerging problem causing serious economic losses (OIE 2015). There have been a large number of attempts to develop better vaccines with the idea of preventing the problems derived from Rev1 vaccination.

2.1. B. melitensis rough mutants against B. ovis infection. Since B. ovis is naturally rough, R vaccines could be a satisfactory alternative, provided they are complemented with a test that allows the differentiation between infected and vaccinated animals (DIVA test). Indeed, B. melitensis R mutants by transposon insertion in genes per (perosamine synthesis), wbkD (bactoprenol priming for O-PS polymerization), wbkF (quinovosamine-third O-PS sugar synthesis) (Kubler- Kielb and Vinogradov, 2013), wadA (sugar linking the core with the O-PS) and wzm (O- PS transport) provide total protection against B. ovis in the mouse model (unpublished results). However, despite lacking the O-PS, R vaccines cause significant interference in diagnostic assays especially in immunosorbent, lateral flow immunochromatography and fluorescence polarization.

212

Chapter 3 Introduction

Thus, the development of a B. ovis specific vaccine in a B. ovis background is urgently needed. Moreover, this vaccine would not pose the problems of biohazards and drawbacks of Rev1 since B. ovis is not zoonotic and the infection is limited to sheep, and could help to avoid the problems in diagnosis.

2.2. Specific B. ovis vaccines. Several subcellular vaccines based on the hot-saline extract of B. ovis or in recombinant OMP proteins or DNA have been investigated. All of them are costly and use bacterial components of limited availability, and most require adjuvants and repeated immunization protocols that would severely restrain their implementation (Blasco et al., 1993; Cassataro et al., 2007; Da Costa Martins et al., 2010, 2012, Estein et al., 2003, 2009; Muñoz et al., 2006; Murillo et al., 2001; Silva et al., 2015a, 2015b).

Live attenuated vaccines represent the best approximation to generate new vaccines against brucellosis because reproduce more closely cell invasion, intracellular trafficking and antigen presentation (Pandey et al., 2016). Live attenuated B. ovis vaccine candidates have also been developed and include single or combined mutants deleted in one or up to three OMP (OMP10, OMP22, OMP25c, OMP25d and OMP31) (Sancho et al., 2014; Vizcaíno et al., personal communication). Also, the virulence- related genes cgs (cyclic glucan synthesis), bacA (lipid A acylation), vjbR (virulence regulator), and virB (type IV secretion system) have been investigated either as single mutants or in combination with OMP deletions (Martín-Martín et al., 2012). The results indicate that B. ovis OMP25d, OMP22 and OMP10-ugpB-OMP31 mutants are not only attenuated but also provide better protection than Rev1 in mice (Sancho et al., 2014; Vizcaíno et al., personal communication). Nevertheless, these vaccines lack a good associated DIVA test.

As a conclusion, although some candidates have been proposed, no B. ovis specific vaccine has proven to be sufficiently efficient against B. ovis and any of the tested candidates is officially accepted by the OIE (Blasco et al., 2016; OIE, 2015).

213

Chapter 3 Introduction

2.3. B. ovis vaccines based in modifications of the core lateral branch. As we have already discussed, B. abortus and B. melitensis glycosiltransferases WadB, WadC and WadD are involved in the synthesis of the core lateral branch (Conde-Álvarez et al., 2012; Gil-Ramírez et al., 2014; Fontana et al., 2016 and Chapter 2 of this work). Since the removal of this core section increases significantly the immune response induced by Brucella (Conde-Álvarez et al., 2012; Fontana et al., 2016; Zhao et al., 2017), it has been suggested that partial or total modifications of this core section could be a good approach for the development of new vaccines against brucellosis (Conde-Álvarez et al., 2013). Genes wadB and wadC (BOV_0337 and BOV_1453 respectively) are highly conserved in strains B. ovis ATCC 25840, B. ovis PA (BoPA) and B. ovis REO. B. ovis PA mutants in both genes (BoPAΔwadB and BoPAΔwadC) are deficient in the synthesis of the core LPS, attenuated in mice and induce a good protection against B. ovis infection in the mouse model, being mutant wadB the most effective (Soler-Lloréns et al., 2014). These results in the laboratory model open the door for the development of B. ovis vaccines based in modifications of the core lateral branch. However, the presence of wadD in B.ovis and the effect of a mutation in this gene has never been investigated. Morover, working in a B. ovis PA background for developing a specific vaccine creates a practical problem since its growth requires an atmosphere containing 5–10%

CO2 (Blasco et al., 1993). Strain B. ovis REO198 is CO2-independent, but it has a defect in the LPS (Suarez et al., 1990) and, although is used to prepare antigens for diagnosis, is not used for vaccination.

A recent work (Pérez-Etayo et al., 2018) has shown that requirement of CO2 for growth is due to the fact that B. ovis PA lacks carbonic anhydrases I and II (CAI and

CAII), that are present in Brucella CO2-independent spp. and needed for the conversion of CO2 to bicarbonate (Joseph et al., 2010, 2011). The region coding for these enzymes in B. ovis PA carries a deletion in CAI and an insertion in CAII, that could explain the absence of a functional carbonic anhydrase and the CO2-dependence of B. ovis PA. Introduction of B. abortus genes coding for these enzymes in B. ovis PA allowed growth in atmospheric conditions. The resultant CO2-independent construction was called B. ovis PA::Tn7-ba2308-CAI-CAIIΔKm (from now Bov::CA-parental) (Pérez-Etayo et al., 2018). Since the introduction of genes coding CAI and CAII does not affect bacterial

214

Chapter 3 Introduction growth, Bov::CA-parental could be considered a good background for the development of B. ovis specific vaccines (Pérez-Etayo et al., 2018). In this chapter, we investigate whether the new gene wadD, described in this thesis and required for the synthesis of a complete core lateral branch in B. abortus (Chapter 2), is present in the rough spp. B. ovis and whether it is possible to combine the genetic modifications that affect core synthesis with those required for B. ovis CO2- independent growth for the design of new B. ovis specific vaccines. We also investigate whether the HS extracted from a modified CO2 independent strain defective in the synthesis of the core lateral branch can be used to improve diagnosis of B. ovis infection.

215

Chapter 3 Experimental procedures

EXPERIMENTAL PROCEDURES

Bacterial strains and growth conditions. The bacterial strains, plasmids and oligonucleotides used are listed in Tables 3.1, 3.2 and Table S3.1 respectively. All bacteria were grown either on tryptic soy broth (TSB, Scharlau) or agar (TSA, Pronadisa) supplemented with 0.5% Yeast Extract (YE, Merck) and 5% fetal bovine serum (S, Gibco). Alternatively, Blood Agar Base (BAB nº 2, Oxoid) supplemented with 5% of fetal bovine serum (S) plates and broth were used. Where indicated, growth media were supplemented with kanamycin (Km) at 50 mg/ml, nalidixic acid (Nal) at 12.5 mg/ml, chloramphenicol (Cm) at 20 mg/ml, ampicillin (Amp) at 100 mg/ml and/or sucrose at 5%. All strains were grown at 37ºC and stored in TSB-YE-S (0.5%) and dimethyl sulfoxide (DMSO) at –80ºC. When needed, atmosphere was enriched with 10% CO2. Work with Brucella was performed at the Biosafety Level 3 laboratory facilities of the University of Navarra, Instituto de Agrobiotecnología (IdAB) from the Public University of Navarra and the Centro de Investigación y Tecnología Agroalimentaria de Aragón (CITA).

DNA manipulations and analyses. Sequence data of Brucella ovis ATCC 25840 were obtained from Kyoto Encyclopedia of Genes and Genomes (KEGG, http://www.genome.jp/kegg/). Searches for DNA and protein homologies between Brucella species were carried out using KEGG, Basic Local Alignment Sequence Tool (BLAST, http://blast.ncbi.nlm.nih.gov/Blast.cgi) and Clustal Omega (http://www.ebi.ac.uk/Tools/msa/clustalo) from the European Molecular Biology Laboratory-European Bioinformatics Institute (EMBL-EBI, http://www.ebi.ac.uk/). To translate nucleotides into amino acid sequences, Prettyseq from EMBOSS (European Molecular Biology Open Software Suite) was used (http://www.bioinformatics.nl/cgi- bin/emboss/prettyseq). Oligonucleotides were designed using Primer 3 input (http://bioinfo.ut.ee/primer3-0.4.0/) and synthesized by Sigma-Aldrich. Plasmid DNA was extracted with Qiaprep spin Miniprep (Qiagen GmbH). When needed, DNA was sequenced by the Servicio de Secuenciación de CIMA (Centro de Investigación Médica Aplicada, Universidad de Navarra, Spain) or Secugen (Madrid).

217

Chapter 3 Experimental procedures

Table 3.1. Bacterial strains. Strain Characteristics Reference Brucella ovis BoPA-parental B. ovis PA virulent strain, natural NalR, CITA1 CO2-dependent collection

BoPA-GmR Challenge strain, B. ovis PA GmR UNAV2 collection

BoPAΔwadB BoPA in frame deletion mutant in Soler-Lloréns wadB (BOV_0337Δ49-195). et al., 2014

Bov::CA-parental B. ovis PA::Tn7-ba2308-CAI-CAII Km Pérez-Etayo et al., 2018

Bov::CAwadB Bov-parental deletion mutant in wadB This work (BOV_0337∆49-195)

Bov::CAwadC Bov-parental deletion mutant in wadC This work (BOV_1453∆16-308)

Bov::CAwadD Bov-parental deletion mutant in wadD This work (BOV_0930∆50-281)

Bov::CAwadB-pMR10 wadB Bov∆wadB harbouring pMR10 KmR-wadB. This work

Bov::CAwadD::Tn7-PwadD Bov∆wadD complemented strain with miniTn7T This work KmR harbouring BAB1_0953 complete ORF with its own promoter.

Brucella melitensis Elberg and Rev1 Reference vaccine Faunce, 1956. CITA1 collection Escherichia coli Stellar F–, endA1, supE44, thi-1, recA1, relA1, gyrA96 Clontech phoA, Φ80d lacZΔ M15, Δ(lacZYA-argF) U169 Δ(mrr-hsdRMS-mcrBC), ΔmcrA, λ–

S17-λpir Mating strain with plasmid RP4 inserted Simon et al., into the chromosome 1983

PIR1 F-∆lac169 rpoS(Am) robA1 creC510 hsdR514 Invitrogen endA recA1 uidA(∆MluI)::pir-116

HB101- pRK2013 pRK2013 is a helper plasmid for conjugation. Choi et al., KmR 35μg/ml 2005

HST08 F–, endA1, supE44, thi-1, recA1, relA1, gyrA96, Takara phoA, Φ80d lacZΔ M15, Δ(lacZYA-argF) U169, Δ(mrr-hsdRMS-mcrBC), ΔmcrA, λ–

1 Centro de Investigación y Tecnología Agroalimentaria de Aragón. 2 Universidad de Navarra.

218

Chapter 3 Experimental procedures

Table 3.2. Plasmids.

Plasmid Characteristics Reference/Source pCR2.1 Cloning vector, KmR Invitrogen pJQK Derivate of pJQ200KS+. Suicide vector, KmR SacS Scupham and Triplett, 1997 pRH001 Derivative of pMR10, KmR CmR Hallez et al., 2007 pUC18 R6KT Broad host-range mini-Tn7 vector Llobet et al., 2009 mini Tn7T KmR pMSB-34 XbaI fragment of 479 bp from B. abortus This work chromosomal DNA containing the wadD deletion allele cloned into the corresponding sites of pJQK by InFusion HD Cloning System. pRCI-26 BamHI-XbaI fragment from pRCI-23 (containing 1019-bp of Conde-Álvarez et al., Ba-parental chromosomal DNA with the wadC deletion allele) 2012 cloned into the corresponding sites of pJQK. pYRI-2 BamHI-XbaI fragment from pYRI-1 (containing 570-bp of Gil-Ramírez et al., Ba-parental chromosomal DNA with the wadB deletion allele) 2014 cloned into the corresponding sites of pJQK. pwadB attL1-attL2 fragment of pYRI-3 (containing the complete Gil-Ramírez et al., wadB gene with attB sites) cloned into the attR1-attR2 sites of 2014 pRH001.

Construction of non-polar mutants. For the construction of mutants in wadB, wadC and wadD in B. ovis, we used the same suicide vectors used for B. abortus mutagenesis since their nucleotide sequences are identical.

Bov::CAΔwadB was constructed using the suicide vector pYRI-2, a pJQK containing the wadB deletion allele (Gil-Ramírez et al., 2014). Oligonucleotides wadB-F1 (5'- GCATGATTACCCCGCTGAT -3') and wadB-R2 (5´-CGCAATCTCGTCTTTGTTGAG- 3´) amplified a 296 base pairs (bp) fragment including codons 1 to 48 of BAB1_0351, as well as 152 bp upstream of the BAB1_0351 start codon. Oligonucleotides wadB-F3 (5´- CTCAACAAAGACGAGATTGCGGGTGGCGTGAAGGAAATCT -3´) and wadB-R4 (5´- TGATAGCCGAGCCTCTTCAG -3´) were used to amplify a 274 bp fragment including codons 196 to 239 of BAB1_0351 and 139 bp downstream of the stop codon (Figure

219

Chapter 3 Experimental procedures

S3.1 in Supplemental material, Chapter 3). The suicide plasmid pYRI-2 was introduced into B. ovis PA::Tn7-ba2308-CAI-CAIIΔKm (Bov::CA-parental) by conjugation between E. coli S17 λpir (pYRI-2) and Bov::CA-parental. The first recombination event (integration of the suicide vector in the chromosome) was selected by Nal and Km resistance, and the second recombination (excision of the mutator plasmid leading to construction of the mutant by allelic exchange), was selected by Nal and sucrose resistance and Km sensitivity. The resulting colonies were screened by PCR with primers wadB-F1 and wadB-R4 which amplified a fragment of 570 bp in the mutant and 1011 bp in the sibling strain that keeps the wild type gene. Primers wadB-F1 and wadB-R5 (5´- ATGCACCCATGAAGTTTTCC- 3´) amplified a fragment of 471 bp only in the wild type strain. The mutation generated results in the loss of the 61.5% of the ORF. The mutant was called Bov::CAΔwadB. For Bov::CAΔwadC construction, we followed the same strategy using the suicide vector pRCI-26 (Conde-Álvarez et al., 2012). Oligonucleotides wadC-F1 (5´- CTGGCGTCAGCAATCAGAG-3´) and wadC-R2 (5´- GTGCAACGACCTCAACTTCC-3´) were used to amplify a 476 bp fragment including codons 1 to 16 of the wadC ORF, as well as 424 bp upstream of its start codon, and oligonucleotides wadC-F3 (5´-GGAAGTTGAGGTCGTTGCACACGCCATCGAACCTTATCTG- 3’) and wadC-R4 (5´-CGGCTATCGTGCGATTCT-3´) amplified a 453 bp fragment including codons 308 to 354 of the wadC ORF and 320 bp downstream of the stop codon. Primers wadC-F1 and wadC-R4 amplified a fragment of 929 bp in the mutant and 1805 bp in the sibling strain that keeps the wild type gene and wadC-F1 and wadC-R5 (5`- GCAATGGAATGAGCTGAACA- 3`) amplified a fragment of 533 bp only in the wild type strain. The mutation generated results in the loss of the 82% of the ORF. The mutant was called Bov::CAΔwadC (Figure S3.2). For the construction of Bov::CAΔwadD, plasmid pMSB-34, containing the deletion allele of BAB1_0953 cloned into pJQK, was used. Oligonucleotides wadD-F1 (5'- TCTATAATGAGAGGCGGCTTTT -3') and wadD-R2 (5´- AGAAGTGCTGGTCCTGTTGT - 3´) were used to amplify a 304 bp fragment including codons 1 to 50 of wadD, as well as 154 bp upstream of the start codon. Oligonucleotides wadD-F3 (5´- ACAACAGGACCAGCACTTCTATCCTCACCCTGCCATTCAA -3´) and wadD-R4 (5´- CTGGTACTAGACGCCCTGTT -3´) were used to amplify a 175 bp fragment including

220

Chapter 3 Experimental procedures codons 281 to 324 of this ORF and 43 bp downstream of the wadD stop codon (see Figure S2.5 in Supplemental material, Chapter 2). Conjugation with Bov-parental was performed following the same protocol used for Bov::CAΔwadB and Bov::CAΔwadC mutants, and the resulting colonies were screened by PCR with primers wadD-F1 and wadD-R4 which amplified a fragment of 479 bp in the mutant and 1169 bp in the sibling strain which keeps the wild type gene. Primers wadD-F1 and wadD-R5 (5´- AGGCACGGTTTCATAGACGA -3´) amplified a fragment of 844 bp only in the wild type strain. The mutation generated results in the loss of the 71% of the ORF and the mutant was called Bov::CAΔwadD.

Complementation of mutants. For the complementation of Bov::CAΔwadB, plasmid pwadB was used (Gil-Ramírez et al., 2014). This plasmid harbors ORF BAB1_0351 of B. abortus 2308, inserted into the Gateway-compatible vector pDONR221 by site-specific recombination and then subcloned in plasmid pRH001. Plasmid pwadB was introduced into Bov::CAΔwadB by mating with E. coli S17-λpir and the conjugates harboring pwadB were selected by plating onto TSA-YE-S-Nal-Cm plates, which were incubated at 37ºC for 6 days. Oligonucleotides wadB-F6 (5`-CCGCAGGTGGAAAAACTC-3`) and wadB-R7 (5`- TCTTTTTGTGCTGGATCGTG-3`) were used to screen the colonies that had incorporated plasmid pwadB. wadB-F6 and wadB-R7 amplified both the deleted wadB copy and the complete one inserted with the pwadB plasmid, with 151 bp and 592 bp respectively. The complemented strain was named Bov::CAΔwadB-pMR10 wadB.

For Bov::CAΔwadD complementation, we used plasmid pMSB-44, already used in Chapter 2, carrying miniTn7 transposon with the complete ORF wadD and its promoter. For the construction of the vector, primers Tn7-wadD-F1 (5`- CGGGCTGCAGGAATTGCGATTCCTTTGTGCCAGAT-3`) and Tn7-wadD-R2 (5`- GCTTCTCGAGGAATTATCATCGCCGCATTGAAGAC-3`) were used to amplify a 1771 bp fragment including codons 1 to 323 of BAB1_0953, 481 bp upstream of the wadD start codon to ensure the amplification of its promoter and 318 bp downstream of the stop codon (see Figure S2.16, in supplemental material, Chapter 2). pMSB-44 was introduced into Bov::CAΔwadD mutant by tetra-parental conjugation as described in

221

Chapter 3 Experimental procedures

Chapter 2. The conjugants harbouring pMSB-44 were selected by plating the mating mixture onto TSA-YE-S-Nal-Km plates which were incubated at 37ºC for 6 days. The resulting colonies were screened using oligonucleotides wadD-F1 and wadD-R4 (used for deletion mutagenesis in Chapter 1), which amplified two fragments in the complemented strain: 479 bp for the deleted copy of the gen and 1169 bp for the complete copy. The complemented strain was named Bov::CAΔwadD::Tn7-PwadD.

LPS extraction. For LPS extraction, bacteria were grown in 10 ml of TSB-YE-S at 37ºC with orbital agitation for 48 hours. Cells were hot-killed (100ºC) for 30 minutes.

SDS-PAGE and Western blot. Samples were mixed 1:1 with Sample buffer 2X (Bio-Rad), heated at 100ºC for 10 minutes and analysed in Tris-HCl-glycine-12 polyacrylamide gels (37.5:1 acrylamide/methylene-bisacrylamide ratio). 15 μl of each sample were run at 30 mA constant current for 140 minutes. For Western blots, gels were electro-transferred onto PVDF sheets (Whatman, Schleicher & Schuell, WESTRAN S; 0.2 μm pore size) in a transfer buffer (pH 8.3) containing 0.025 M Tris, 0.192 M glycine, and 20 % (vol/vol) methanol. Transfer was performed at a constant voltage of 8V and 200 mA for 30 minutes in a Trans-Blot Semi- Dry Transfer Cell (Bio-Rad). After transfer, membranes were blocked overnight with 3% skim milk in PBS containing 0.05% Tween 20, and next washed with the same buffer. They were then incubated overnight at room temperature with immune sera diluted 1:500 in the same solution. After washing, the corresponding peroxidase-conjugated secondary antibody was added, and incubation continued for at least 4 hours at room temperature. Bound immunoglobulins were detected using 4-chloro-1-naphthol-H2O2. Antibodies used in this study were A68/24G12/A08 [monoclonal that recognizes core epitopes (Bowden et al., 1995)], a polyclonal serum of 45 days of an infected rabbit with 109 colony forming unit (CFU) of B. melitensis 16M and a serum of an infected ram with B.ovis.

222

Chapter 3 Experimental procedures

Growth curves. To study the growth rates of each strain and to analyse if any of the mutants had growth deficiencies, bacteria were grown either in 10 ml of TSB-YE-S or BAB-S in a 50- ml flask and incubated at 37ºC with orbital agitation for 18 hours. The number of bacteria was determined by optical density (OD) at 600 nm in a spectrophotometer. Bacteria were harvested by centrifugation at 13,000 rpm for 10 minutes and resuspended in 1 ml of TSB-YE-S or BAB-S medium at an OD600nm of 0.1. Starting with this OD, cells were cultivated in a Bioscreen C (Lab Systems) with continuous shaking in a working volume of 200 μl/well. Temperature was controlled at 37ºC and OD at 420- 580 nm. Absorbance values of cell suspensions were automatically recorded at regular 0.5-hours intervals, over a 120 to 300 hours period. All experiments were performed in triplicate. Wells containing only the culture medium (TSB-YE-S or BAB-S) were used as sterility controls.

Virulence and protection experiments. For virulence, seven-week-old female BALB/c mice (ENVIGO, Harlan) (n=5) were inoculated intraperitoneally (IP) with 5–7 x 106 CFU/mouse of mutants Bov::CAΔwadB, Bov::CAΔwadC and Bov::CAΔwadD, and viable spleen counts were determined at 3 and 8 weeks post-inoculation. As controls, additional groups of mice (n=5) were inoculated similarly with the virulent B. ovis PA reference strain (BoPA) and the CO2-independent strain Bov::CA-parental. The identity of the spleen isolates was confirmed throughout the experiment by both PCR and B. ovis phenotypic features. Spleen infections were expressed as mean ± SD (n=5) of individual log10 CFU/spleen at the times indicated.

Efficacy of the Bov::CAΔwadB mutant as vaccine was evaluated in BALB/c mice (n=5) vaccinated with 7 x 106 CFU/mouse of the mutant both IP or subcutaneously (SC). As controls, we used animals (n=5) vaccinated with 7 x 106 CFU/mouse of the mutant BoPAΔwadB, both IP and SC, or with B. melitensis Rev1 (as standard vaccinated control), via SC (n=10), and animals (n=5) inoculated with 0.1 mL of PBS (pH 6.8) as the unvaccinated control. Four weeks after vaccination, all mice were challenged with 5 x 104 CFU/mouse of the virulent B. ovis PA-GmR strain, and the number of challenge bacteria in spleens was determined. Differentiation between challenge and residual vaccine bacteria was performed by duplicate plating on BAB-S and BAB-S 223

Chapter 3 Experimental procedures supplemented with Gm. Results are expressed as the mean and SD of the individual

R log10 CFU/spleen of B. ovis PA-Gm challenge strain.

Hot-Saline extraction. For HS extraction, Bov::CAΔwadB or Bov::CA-parental were grown in TSA-YE-S plates and incubated for 72 hours at 37ºC. After resuspending in 3-4 ml of saline, cells were incubated in six 800 ml-TSB-YE-S flasks for 36 hours at 36ºC with orbital agitation (200 rpm). Bacteria were then inactivated with 0.5% phenol and concentrated by centrifugation (7.000 x g, 20 min) in 1g/10ml of saline. Then, the samples were fluent vapor-extracted (100ºC) for 15 min. When cooled, they were centrifuged (12.000 x g, 15 min, 4ºC) and the supernatant was collected and dialyzed with 3 volumes of distilled water. Finally, samples were ultracentrifuged for 20 hours (100.000 x g, 4ºC) and lyophilized.

224

Chapter 3 Results and Discussion

RESULTS AND DISCUSSION

WadD, together with WadB and WadC, are required for the synthesis of a complete

LPS core in B. ovis PA CO2-independent strain. To anlayse whether wadD was present in the rough spp. B. ovis, we first looked for the ortholog in B. ovis ATCC25840 and identified BOV_0930 as 100% identical to B. abortus wadD. We also wanted to analyse the presence of wadD in the virulent species B. ovis PA and B. ovis REO. For this purpose, we sequenced the region sorrounding wadD in both species. As can be seen in Table 3.3, the gene was intact in B. ovis PA but, in B. ovis REO, the serine in position 73 had been substituted by an asparagine.

Table 3.3. Comparison between B. abortus and B. ovis species of the core glycosyltransferases.

Percentage of homology with respect to B. abortus 2308 Strain wadB1 wadC1 wadD

B. ovis ATCC 25840 99.6 100 100 B. ovis PA 99.6 100 100 B. ovis REO 99.6 100 99.7

1 Results from (Soler-Lloréns et al., 2014).

We then constructed non-polar mutants in the genes encoding the core glycosytransferases WadB, WadC and WadD in the CO2-independent B. ovis PA background (Bov::CA-parental) to obtain Bov::CAΔwadB, Bov::CAΔwadC and Bov::CAΔwadD. The LPS from Bov::CA-parental and the three mutants was extracted and analysed by SDS-PAGE and Western-blot. As shown in Figure 3.1, all mutants failed to react with monoclonal antibody A68/24G12/A08 that recognize the core section of the LPS.

Complementation was performed only for wadB and wadD mutants, since BoPAΔwadB was the most effective protecting against B. ovis infection (Soler-Lloréns et al., 2014) and wadD had not been tested previously. As can be seen in Figure 3.1, 225

Chapter 3 Results and Discussion introduction of an intact wadB in Bov::CAΔwadB (Bov::CAΔwadB-pMR10wadB) or wadD in Bov::CAΔwadD (Bov::CAΔwadD::Tn7-PwadD) restored the Bov::CA-parental phenotype. These results prove that wadB, wadC and wadD were also required for the synthesis of a complete core LPS in CO2-independent B. ovis PA, and that the phenotypes of the mutants were due to the respective gene deletions and not to a polar effect.

Figure 3.1. The LPS of mutants in wadB, wadC and wadD in the B. ovis CO2-independent background failed to react in the core LPS section. Western-blot analysis of LPS extracts of Bov::CA-parental, the mutants Bov::CAΔwadB, Bov::CAΔwadC and Bov::CAΔwadD and the complemented strains Bov::CAΔwadB-pMR10 wadB and Bov::CAΔwadD::Tn7-PwadD with monoclonal anti-core antibody A68/24G12/A08.

226

Chapter 3 Results and Discussion

Mutants in glycosyltransferases required for the synthesis of the core lateral branch in B. ovis CO2 -independent are attenuated. The colonization and persistance of Bov::CAΔwadB, Bov::CAΔwadC and Bov::CAΔwadD was tested in BALB/c mice. Groups of 5 mice were inoculated with 5-7 x 106 CFU/mouse of the different mutants, Bov::CA-parental or BoPA-parental strains by the IP route, and sacrified 3 and 8 weeks after infection to determine the number of CFU/spleen. As can be seen in Figure 3.2, the three mutants were attenuated from week 3, the time-point at which highest peak splenic concentrations are reached by virulent strains. The results concerning wadB and wadC mutants expand and confirm those previously observed in the B. ovis PA background.

Figure 3.2. Mutants in the core glycosyltransferases in B. ovis CO2-independent background were attenuated in mice. Multiplication of B. ovis PA (BoPA), Bov::CA-parental, Bov::CAΔwadB, Bov::CAΔwadC and Bov::CAΔwadD in the spleens of BALB/c mice (each point represents the logarithm of CFU in the spleen of a mouse).

To study if the mutations in the genes involved in synthesis of the core LPS of B. ovis (Bov::CA-parental) generated growth deficiencies, what could explain the attenuation observed, mutants were grown in different media. Bov::CAΔwadB grew as the parental strain and Bov::CAΔwadC showed a slight growth defect in BAB-S (Figure 3.3). Interestingly, Bov::CAΔwadD had a marked growth defect when compared to Bov::CA-parental both in TBS-YE-S and BAB-S. As shown in Figure 3.3C, the complemented strain Bov::CAΔwadD::Tn7-PwadD also showed a growth defect, not restoring the levels of Bov::CA-parental. This could be due to an aleatory mutation generated during the construction of the wadD mutant that affected bacterial fitness. 227

Chapter 3 Results and Discussion

However, introduction of an empty miniTn7-KmR (Figure 3.3C) or pMR10-KmR (Figure 3.3D) also resulted in similar growth delay suggesting that the deficiency to multiply of Bov::CAΔwadD is not related to the lack of wadD and could be influenced by the presence of the gene that confer resistance to kanamycin (Figure 3.3C and D). Taking into account this result, we could not conclude that the low multiplication of Bov::CAΔwadD in the spleen of mice is related to the wadD disfunction.

228

Chapter 3 Results and Discussion

Bov::CA-parental Bov::CAΔwadC

Bov::CAΔwadB Bov::CAΔwadD

A B

Bov parental

C D

Bov::CA-parental Bov::CA-parental

Bov::CAΔwadD Bov::CAΔwadB Bov::CAΔwadD::Tn7-PwadD Bov::CAΔwadB-pMR10 Bov::CA-parental::Tn7 BoPAwadB -pMR10 CAII

Figure 3.3. wadD mutation generated a marked growth defect when compared to the parental strain. Growth curves of Bov::CA-parental and the mutants Bov::CAΔwadB, Bov::CAΔwadC and Bov::CAΔwadD in TSB-YE-S (A) and BAB-S (B) and the complemented strains Bov::CAΔwadD::Tn7-PwadD (C) and Bov::CAΔwadB-pMR10 wadB (D) in TSB-YE-S, grown either with or without 50µg/ml Km. As controls, Bov::CA-parental::Tn7 (C) and BoPA-pMR10 CAII (D) were used. Each point represents the mean of optical density (O.D.) values for triplicate samples. The experiment was repeated three times with similar results.

229

Chapter 3 Results and Discussion

A wadB mutant in a CO2-independent B. ovis background protects against B. ovis PA infection in mice. Taking into account that (i), the B. ovis wadD mutants studied above showed a growth defect that remains unexplained and (ii), the B. ovis PA wadB mutant is the most effective in protection against B. ovis PA infection (Soler-Lloréns et al., 2014), we continued the protection studies with Bov::CAΔwadB (a wadB mutant in a CO2- independent B. ovis PA). The protection efficacy of Bov::CAΔwadB was compared with that of BoPAΔwadB (Soler-Lloréns et al., 2014) and Rev1, using as controls unvaccinated mice. Mice were vaccinated with 7 x 106 CFU/mice via IP or SC routes and challenged 4 weeks later with 5 x 104 CFU/mice of B. ovis PA-GmR. As can be seen in Table 3.4, wadB mutants in both

CO2-dependent and independent backgrounds, protected against infection with the challenge strain when the vaccine was administered by the IP or the SC route, since the number of CFU of the challenge strain/spleen was significantly lower than that of unvaccinated mice. However, the protection was better when the vaccine was administered by the IP route, and it was similar to the protection conferred by Rev1 administered by the SC route.

Table 3.4. Protection induced against B. ovis PA infection.

Group Vaccination route Log challenge/spleen IP 2.47 ± 2.45 BoPAΔwadB SC 4.69 ± 1.43 IP 2.98 ± 2.40 Bov::CAΔwadB SC 4.31 ± 2.36 Rev1 SC 2.75 ± 1.26 PBS 6.21 ± 0.45

Thus, a wadB mutant in a CO2-independent B. ovis background could be a good candidate for protection against brucellosis and currently, it is being tested as a vaccine in the natural host.

230

Chapter 3 Results and Discussion

Could antigen from Bov::CAΔwadB be used for differentiation between vaccinated and infected animals with B. ovis? Since wadB mutation in Bov::CA presented a defect in the core LPS, we questioned whether the HS extracted from this mutant could allow differentiation between sheep infected wih field strain (B. ovis PA) and animals vaccinated with Bov::CAΔwadB. To this end, we extracted the HS from Bov::CAΔwadB and Bov::CA-parental, following the method described to obtain REO198 HS, and compared them by SDS- PAGE and Western Blot with a serum from an infected ram. As can be seen in Figure 3.4, this serum reacted with HS from REO or Bov::CA-parental, but not with the antigen extracted from Bov::CAΔwadB. As expected, this antigen failed to react with a R-LPS specific MoAb. This opens the door to design and test whether a serological test using the HS antigen from Bov::CAΔwadB could help to differentiate between infected and vaccinated animals (DIVA test). It would be also interesting to test whether the HS from Bov::CAΔwadB could avoid the cross-reactivity that is sometimes observed with the HS-REO method and sera of sheep infected with B. melitensis or vaccinated with B. melitensis Rev1 (Riezu-Boj et al., 1986).

231

Chapter 3 Results and Discussion

SDS-PAGE Anti-R Brucella Anti-core

Western-Blot

Figure 3.4. The HS extracted from mutant Bov::CAΔwadB presented a defective LPS core oligosaccaride. SDS-PAGE electrophoresis and silver staining of HS extract samples (left panel). Western- blot analysis with a serum of an infected ram with B. ovis (anti-R-Brucella) or monoclonal anti-core antibody A68/24G12/A08 (central and right panels).

232

Chapter 3 References

REFERENCES

Blasco, J. M. (1990). “Brucella ovis,” in Animal brucellosis. Nielsen, Klaus. Duncan, J. Robert. (Bocaratón,Florida: CRC Press), 351–378. Blasco, J. M. (1997). A review of the use of B. melitensis Rev 1 vaccine in adult sheep and goats. Prev. Vet. Med. 31, 275–283. doi:10.1016/S0167-5877(96)01110-5. Blasco, J. M., Gamazo, C., Winter, A. J., Jiménez de Bagüés, M. P., Marín, C., Barberán, M., et al. (1993). Evaluation of whole cell and subcellular vaccines against Brucella ovis in rams. Vet. Immunol. Immunopathol. 37, 257–70. doi.org/10.1016/0165- 2427(93)90198-D. Blasco, J. M., Moreno, E., and Moriyón, I. (2016). “Brucellosis vaccines and vaccine candidates.,” in Veterinary Vaccines for Developing Countries., ed. A. Metwally, S and Viljoen, G J and El Idrissi (Rome: FAO), 1–66. Bowden, R., Cloeckaert, A., Zygmunt, M., Bernard, S., and Dubray, G. (1995). Surface exposure of outer membrane protein and lipopolysaccharide epitopes in Brucella species studied by Enzyme-Linked Immunosorbent Assay and flow cytometry. Infect. Immun. 63, 3945–3952. Cassataro, J., Pasquevich, K. A., Estein, S. M., Laplagne, D. A., Zwerdling, A., de la Barrera, S., et al. (2007). A DNA vaccine coding for the chimera BLSOmp31 induced a better degree of protection against B. ovis and a similar degree of protection against B. melitensis than Rev.1 vaccination. Vaccine 25, 5958–5967. doi:10.1016/J.VACCINE.2007.05.049. Choi, K.-H., Gaynor, J. B., White, K. G., Lopez, C., Bosio, C. M., Karkhoff-Schweizer, R. R., et al. (2005). A Tn7-based broad-range bacterial cloning and expression system. Nat. Methods 2, 443–448. doi:10.1038/nmeth765. Conde-Álvarez, R., Arce-Gorvel, V., Gil-Ramírez, Y., Iriarte, M., Grilló, M. J., Gorvel, J. P., et al. (2013). Lipopolysaccharide as a target for brucellosis vaccine design. Microb. Pathog. 58, 29–34. doi:10.1016/j.micpath.2012.11.011. Conde-Álvarez, R., Arce-Gorvel, V., Iriarte, M., Manček-Keber, M., Barquero-Calvo, E., Palacios-Chaves, L., et al. (2012). The lipopolysaccharide core of Brucella abortus acts as a shield against innate immunity recognition. PLoS Pathog. 8, e1002675. doi:10.1371/journal.ppat.1002675. Da Costa Martins, R., Irache, J. M., Blasco, J. M., Muñoz, M. P., Marín, C. M., Jesús Grilló, M., et al. (2010). Evaluation of particulate acellular vaccines against Brucella ovis infection in rams. Vaccine 28, 3038–3046. doi:10.1016/J.VACCINE.2009.10.073. Da Costa Martins, R., Irache, J. M., and Gamazo, C. (2012). Acellular vaccines for ovine brucellosis: A safer alternative against a worldwide disease. Expert Rev. Vaccines 11, 87–95. doi:10.1586/erv.11.172.

233

Chapter 3 References

Elberg, Sanford S, and Kenneth Faunce. (1956). Immunization against Brucella Infection Vi. Immunity conferred on goats by a nondependent mutant from a streptomycin- dependent mutant strain of Brucella melitensis. Journal of Bacteriology 73. Estein, S. M., Cassataro, J., Vizcaíno, N., Zygmunt, M. S., Cloeckaert, A., and Bowden, R. A. (2003). The recombinant Omp31 from Brucella melitensis alone or associated with rough lipopolysaccharide induces protection against Brucella ovis infection in BALB/c mice. Microbes Infect. 5, 85–93. doi:10.1016/S1286-4579(02)00075-8. Estein, S. M., Fiorentino, M. A., Paolicchi, F. A., Clausse, M., Manazza, J., Cassataro, J., et al. (2009). The polymeric antigen BLSOmp31 confers protection against Brucella ovis infection in rams. Vaccine 27, 6704–6711. doi:10.1016/J.VACCINE.2009.08.097. Fontana, C., Conde-Álvarez, R., Ståhle, J., Holst, O., Iriarte, M., Zhao, Y., et al. (2016). Structural studies of lipopolysaccharide defective mutants from Brucella melitensis identify a core oligosaccharide critical in virulence. J. Biol. Chem. 291, 7727–7741. doi:10.1074/jbc.M115.701540. Gamazo, C., Winter, A. J., Moriyón, I., Riezu-Boj, J. I., Blasco, J. M., and Diaz, A. R. (1989). Comparative analyses of proteins extracted by hot saline or released spontaneously into outer membrane blebs from field strains of Brucella ovis and Brucella melitensis. Infect. Immun. 57, 1419–1426. Gil-Ramírez, Y., Conde-Álvarez, R., Palacios-Chaves, L., Zúñiga-Ripa, A., Grilló, M.-J., Arce-Gorvel, V., et al. (2014). The identification of wadB, a new glycosyltransferase gene, confirms the branched structure and the role in virulence of the lipopolysaccharide core of Brucella abortus. Microb. Pathog. 73, 53–9. doi:10.1016/j.micpath.2014.06.002. Grilló, M. J., Marín, C. M., Barberán, M., and Blasco, J. M. (1999). Experimental Brucella ovis infection in pregnant ewes. Vet. Rec. 144, 555–8. Hallez, R., Letesson, J.-J., Vandenhaute, J., and De Bolle, X. (2007). Gateway-based destination vectors for functional analyses of bacterial ORFeomes: Application to the min system in Brucella abortus. Appl. Environ. Microbiol. 73, 1375–1379. doi:10.1128/AEM.01873-06. Joseph, P., Ouahrani-Bettache, S., Montero, J.-L., Nishimori, I., Minakuchi, T., Vullo, D., et al. (2011). A new β-carbonic anhydrase from Brucella suis, its cloning, characterization, and inhibition with sulfonamides and sulfamates, leading to impaired pathogen growth. Bioorg. Med. Chem. 19, 1172–1178. doi:10.1016/j.bmc.2010.12.048. Joseph, P., Turtaut, F., Ouahrani-Bettache, S., Montero, J.-L., Nishimori, I., Minakuchi, T., et al. (2010). Cloning, characterization, and inhibition studies of a β-carbonic anhydrase from Brucella suis. J. Med. Chem. 53, 2277–2285. doi:10.1021/jm901855h. Kubler-Kielb, J., and Vinogradov, E. (2013). The study of the core part and non- repeating elements of the O-antigen of Brucella lipopolysaccharide. Carbohydr. Res. 366, 33–7. doi:10.1016/j.carres.2012.11.004.

234

Chapter 3 References

Llobet, E., March, C., Giménez, P., and Bengoechea, J. A. (2009). Klebsiella pneumoniae OmpA confers resistance to antimicrobial peptides. Antimicrob. Agents Chemother. 53, 298–302. doi:10.1128/AAC.00657-08. Martín-Martín, A. I., Sancho, P., de Miguel, M. J., Fernández-Lago, L., and Vizcaíno, N. (2012). Quorum-sensing and BvrR/BvrS regulation, the type IV secretion system, cyclic glucans, and BacA in the virulence of Brucella ovis: similarities to and differences from smooth brucellae. Infect. Immun. 80, 1783–93. doi:10.1128/IAI.06257-11. Monreal, D., Grilló, M. J., González, D., Marín, C. M., De Miguel, M. J., López-Goñi, I., et al. (2003). Characterization of Brucella abortus O-polysaccharide and core lipopolysaccharide mutants and demonstration that a complete core is required for rough vaccines to be efficient against Brucella abortus and Brucella ovis in the mouse model. Infect. Immun. 71, 3261–71. doi:10.1128/iai.71.6.3261-3271.2003. Muñoz, P. M., Estevan, M., Marín, C. M., Jesús De Miguel, M., Jesús Grilló, M., Barberán, M., et al. (2006). Brucella outer membrane complex-loaded microparticles as a vaccine against Brucella ovis in rams. Vaccine 24, 1897–1905. doi:10.1016/J.VACCINE.2005.10.042. Murillo, M., Grilló, M. J., Reñé, J., Marın,́ C. M., Barberán, M., Goñi, M. M., et al. (2001). A Brucella ovis antigenic complex bearing poly-ε-caprolactone microparticles confer protection against experimental brucellosis in mice. Vaccine 19, 4099– 4106. doi:10.1016/S0264-410X(01)00177-3. OIE (2015). “OIE: Chapter 2.7.8. Ovine epididymitis (Brucella ovis),” in Manual of Diagnostic Tests and Vaccines for Terrestrial Animals. OIE (2016). “OIE: Chapter 2.1.4. Brucellosis (Brucella abortus, B. melitensis and B. suis) (infection with B. abortus, B. melitensis and B. suis),” in Manual of Diagnostic Tests and Vaccines for Terrestrial Animals. OIE - World Organisation for Animal Health (2016). Brucellosis (Brucella abortus, B. melitensis and B. suis) (Infection with Brucella abortus, B. melitensis and B. suis). Pandey, A., Cabello, A., Akoolo, L., Rice-Ficht, A., Arenas-Gamboa, A., McMurray, D., et al. (2016). The case for live attenuated vaccines against the neglected zoonotic diseases Brucellosis and Bovine Tuberculosis. PLoS Negl. Trop. Dis. 10, e0004572. doi:10.1371/journal.pntd.0004572. Pérez-Etayo, L., de Miguel, M. J., Conde-Álvarez, R., Muñoz, P. M., Khames, M., Iriarte, M., et al. (2018). The CO2-dependence of Brucella ovis and Brucella abortus biovars is caused by defective carbonic anhydrases. Vet. Res. Riezu-Boj, J. I., Moriyón, I, Blasco, J. M., Marin, C. M., and Diaz, A. R. (1986). Comparison of lipopolysaccharide and outer membrane protein- lipopolysaccharide extracts in an Enzyme-Linked Immunosorbent Assay for the diagnosis of Brucella ovis infection. J. Clin. Microbiology, 938–942. Sancho, P., Tejedor, C., Sidhu-Muñoz, R. S., Fernández-Lago, L., and Vizcaíno, N. (2014). Evaluation in mice of Brucella ovis attenuated mutants for use as live vaccines

235

Chapter 3 References

against B. ovis infection. Vet. Res. 45, 61. doi:10.1186/1297-9716-45-61. Santos, J. M., Verstreate, D. R., Perera, V. Y., and Winter, A. A. J. (1984). Outer membrane proteins from rough strains of four Brucella species. Infect. Immun. 46, 188–194. Scupham, A. J., and Triplett, E. W. (1997). Isolation and characterization of the UDP- glucose 4′-epimerase-encoding gene, galE, from Brucella abortus 2308. Gene 202, 53–59. doi:10.1016/S0378-1119(97)00453-8. Silva, A. P. C., Macêdo, A. A., Costa, L. F., Rocha, C. E., Garcia, L. N. N., Farias, J. R. D., et al. (2015a). Encapsulated Brucella ovis lacking a putative ATP-binding cassette transporter (ΔabcBA) protects against wild type Brucella ovis in rams. PLoS One 10, e0136865. doi:10.1371/journal.pone.0136865. Silva, A. P. C., Macêdo, A. A., Silva, T. M. A., Ximenes, L. C. A., Brandão, H. M., Paixão, T. A., et al. (2015b). Protection provided by an encapsulated live attenuated ΔabcBA strain of Brucella ovis against experimental challenge in a murine model. Clin. Vaccine Immunol. 22, 789–97. doi:10.1128/CVI.00191-15. Simon, R., Priefer, U., and Pühler, A. (1983). A broad host range mobilization system for in vivo genetic engineering: Transposon mutagenesis in gram negative bacteria. Bio/Technology 1, 784–791. doi:10.1038/nbt1183-784. Soler-Lloréns, P., Gil-Ramírez, Y., Zabalza-Baranguá, A., Iriarte, M., Conde-Álvarez, R., Zúñiga-Ripa, A., et al. (2014). Mutants in the lipopolysaccharide of Brucella ovis are attenuated and protect against B. ovis infection in mice. Vet. Res. 45, 72. doi:10.1186/s13567-014-0072-0. Suarez, C. E., Pacheco, G. A., and Vigliocco, A. M. (1990). Immunochemical studies of oligosaccharides obtained from the lipopolysaccharide of Brucella ovis. Vet. Microbiol. 22, 329–34. Zhao, Y., Hanniffy, S., Arce-Gorvel, V., Conde-Álvarez, R., Oh, S., Moriyón, I., et al. (2017). Immunomodulatory properties of Brucella melitensis lipopolysaccharide determinants on mouse dendritic cells in vitro and in vivo. Virulence, 0. doi:10.1080/21505594.2017.1386831.

236

Chapter 3 Supplemental material

Supplemental material

239

Chapter 3 Supplemental material

Table S3.1. Oligonucleotides.

Oligonucleotide Sequence 5`  3` wadB-F1 GCATGATTACCCCGCTGAT wadB-R2 CGCAATCTCGTCTTTGTTGAG wadB-F3 CTCAACAAAGACGAGATTGCGGGTGGCGTGAAGGAAATCT wadB-R4 TGATAGCCGAGCCTCTTCAG wadB-R5 ATGCACCCATGAAGTTTTCC wadB-F6 CCGCAGGTGGAAAAACTC wadB-R7 TCTTTTTGTGCTGGATCGTG wadC-F1 CTGGCGTCAGCAATCAGAG wadC-R2 GTGCAACGACCTCAACTTCC wadC-F3 GGAAGTTGAGGTCGTTGCACACGCCATCGAACCTTATCTG wadC-R4 CGGCTATCGTGCGATTCT wadC-R5 GCAATGGAATGAGCTGAACA wadD-F1 TGGCGGCCGCTCTAGTCTATAATGAGAGGCGGCTTTT * wadD-R2 AGAAGTGCTGGTCCTGTTGT wadD-F3 ACAACAGGACCAGCACTTCTATCCTCACCCTGCCATTCAA wadD-R4 ATCCACTAGTTCTAGCTGGTACTAGACGCCCTGTT * wadD-R5 AGGCACGGTTTCATAGACGA

Tn7-wadD-F1 CGGGCTGCAGGAATTGCGATTCCTTTGTGCCAGAT *

Tn7-wadD-R2 GCTTCTCGAGGAATTATCATCGCCGCATTGAAGAC *

*Underlined nucleotides at the 5’ end correspond to 15 bases that are homologous to the corresponding linearized vector to which it will be joined. The 3’ end of the primers contains the sequence that is specific to the target gene.

241

Chapter 3 Supplemental material

5` wadB-F1 3` Stop BOV_0338 CAGTGGATCGAGGAAAAAATCAAATCGGGCATGATTACCCCGCTGATCTGAGCCGGTAAT

TATACCGGCCCTTATGCCTCTCTCCGGTGGATTTCGGTCGTTTTGCATGCCATTTGTCGA

AATGCCAAATTTCGTGCATATAACGCGCTTTCCTGAATCACAGCCATCGGTCCCGCGTCC Start wadB 1 ATGAACGTCAAGGCATTCATCATTCACCTCAAACGCGCGACCGATCGTGCGCCGCAGGTG M N V K A F I I H L K R A T D R A P Q V 20

61 GAAAAACTCATCAAGGAATTGCCGGTCAAGGCCGAAGTCATCGAAGCCGTCGATAGCCGT E K L I K E L P V K A E V I E A V D S R 40

3` wadB-R2 5` GAGTTGTTTCTGCTCTAACGC 121 GCGCTCAACAAAGACGAGATTGCGCGCATTTATAAACGCCGCCTGCACACTCCGCGCTAT A L N K D E I A R I Y K R R L H T P R Y 60

181 CCCTTTGCCCTCAGCCGCAACGAGATTGCCTGTTTCCTGTCGCACCGCAAGGCATGGCAG P F A L S R N E I A C F L S H R K A W Q 80

C 241 GCGATTATCGACCGGAAGCTTGATGCGGGCTTCATCGTGGAAGACGATATTGCACTGACG A I I D R K L D A G F I V E D D I A L T 100

3` wadB-R5 5` CTTTTGAAGTACCCACGTA 301 GAAAACTTCATGGGTGCATACCGGGCCGCGGTCGATCATCTGGAACCGGGTGGCTTCATC E N F M G A Y R A A V D H L E P G G F I 120

361 CGCTTTACATTCCGGGACGACCGCGAGCATGGCCGTGAGGTTTTCCGGGACGAAGCGGTG R F T F R D D R E H G R E V F R D E A V 140

421 CGGATCATCATCCCGAACCCGATCGGCCTTGGCATGGTTGCGCAGTTTGTTTCCTATGAT R I I I P N P I G L G M V A Q F V S Y D 160

481 GCGGCCCAAAAACTGCTCGACATCACCCAGACGTTCGACCGGCCTGTCGATACGACGGTG A A Q K L L D I T Q T F D R P V D T T V 180

5` wadB-R2 3` CTCAACAAAGACGAGATTGCG 5` wadB-F3 541 CAGATGCGCTGGGTAACGGGCCTGCAACCCCTTTCGGTCATTCCGGGTGGCGTGAAGGAA Q M R W V T G L Q P L S V I P G G V K E 200

3` 601 ATCTCGTCACAACTGGGCGGAACCACGATCCAGCACAAAAAGAACTTTTCAGACAAGCTT I S S Q L G G T T I Q H K K N F S D K L 220 Stop wadB 661 GCGCGCGAAATTTTACGCCCCATCTACCGGATGCGTGTGCGCGCCTATTCATCGAAATGA A R E I L R P I Y R M R V R A Y S S K * 239

Start BAB1_0350 GTTTTATGATGCCTGTACCTATTCTTTTGTATCACCAGATCGCGCCATTACCGGCCAAAA

ACATTCCATTCCGCGGCTTGCTGGTTCATCCTGACCGCTTCCGCAGCCAGATGCGCTGGC G TGAAGAGGCTCGGCTATCA ACTTCTCCGAGCCGATAGT 3` wadB-R4 5`

242

Chapter 3 Supplemental material

Figure S3.1. Sequence of wadB (BOV_0337) and its upstream and downstream regions. Start and stop codons are in green and red characters respectively. Blue characters denote intergenic nucleotides. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids deleted in Bov::CAΔwadB mutant.

243

Chapter 3 Supplemental material

5` wadC-F1 3` CTGGCGTCAGCAATCAGAGCGCCCTCGGATCATTGAGGCAATCGCCGAAGCGCGCGCCCATGGT

GATCTGTCGGAAAACGCCGAATATCACGCCGCCAAGGAAGCCCAGAGCCTCAATGAAGGG

CGCATCAACGAATTGGAAGACCTGGTTGCACGCGCCGAGGTGATCGATGTTTCCAAGTTG

ACGGGCGACAGGATCAAATTCGGTGCGACCGTTACGATGATCGATGAAGACACCGAAGAA

GAAAAAATCTACCAGATCGTCGGCGATCAGGAAGCCGATGTGAAGGAAGGCCGTATTTCG

ATTTCCTCTCCGATCGCGCGCGCCCTTATCGGCAAGGGCGAAGGCGACACAATTGAGGTC Stop BOV_1454 AACGCACCGGGCGGCTCGCGTTCCTACGAAATCATCGCTTTGAAATTCGTCTGATTTTCC

3` wadC-R2 5` Start wadC CCTTCAACTCCAGCAACGTG 1 GTGACTTTATCAGGGCAAGTTCCGGTTCGGGAAGTTGAGGTCGTTGCACCCAACTTCAAG V T L S G Q V P V R E V E V V A P N F K 20

61 CGCCGCCTTTCCGGCGTGACCTCGACGATTGTTCAGCTCATTCCATTGCAGCGCGCAATG R R L S G V T S T I V Q L I P L Q R A M 40

121 GGGCTGAAAATTGCCACTATGGGGCCGGGCCTGCCTGACACTCTCCCGCATCTTGGATGG G L K I A T M G P G L P D T L P H L G W 60

181 AGCGCGTTGCCCTCCTTCTGGTCGCGACCTAAAACCAGACGGTTTCGCATCTGGCATGCG S A L P S F W S R P K T R R F R I W H A 80

241 CGCCGCAACATTGAAATGCTTGCCGGAATTTTCATGCGGGATGTCCTGCGCATGAAACTG R R N I E M L A G I F M R D V L R M K L 100

301 CGGCTCGTTTTTACCTCGGCCGCCCAGCGCCATCACAAACCTTTCACCAAATGGCTCATC R L V F T S A A Q R H H K P F T K W L I 120

361 CGCCGCATGAATGCGGTGATCGCGACAAGTGTGCGTTCGGGAAGTTTTCTCGAAGTCCCG R R M N A V I A T S V R S G S F L E V P 140

421 CATCAGGTCATCATGCATGGTGTCGATCTGGAACGGTTTCACCCGCCTCTTGCGGAGGAC H Q V I M H G V D L E R F H P P L A E D 160

481 GATGATTTTTCCGCTTCCGGCCTGCCGGGTAAATATGCCGTGGGGTGTTTTGGACGGGTT D D F S A S G L P G K Y A V G C F G R V 180

541 CGGCCTTCAAAAGGAACGGACCTTTTCGTTGATGCGATGATCGCGCTCCTGCCGAAATAT R P S K G T D L F V D A M I A L L P K Y 200

601 CCTGACTGGACGGCCATTGTCACGGGGCGCACAACCGCCGAATATCAGGCCTTTGAAGCC P D W T A I V T G R T T A E Y Q A F E A 220

661 GAGCTGCGCACCAGGATTGCGGCAGCCGGTTTGCAGGACCGTATTCTTATTCTGGGCGAA E L R T R I A A A G L Q D R I L I L G E 240

721 GTGCCGGACGTGCGCGTCTGGTATCGCCGGCTCACGCTTTACGTCGCGCCTTCCCGCAAT V P D V R V W Y R R L T L Y V A P S R N 260

781 GAAGGCTTTGGCCTGACGCCGCTTGAAGCCATGGCCTCGAAAACAGCCGTCGTTGCCAGT E G F G L T P L E A M A S K T A V V A S 280

841 GATGCCGGGGCTTATGCGGAAATGGTCGTTGAGGATACGGGCCGCTTTGTGCCAGCCGGT

244

Chapter 3 Supplemental material

D A G A Y A E M V V E D T G R F V P A G 300

5` wadC-R2 3` GGAAGTTGAGGTCGTTGCAC 5` wadC-F3 3` 901 GACGGGCGCGCATTGACGAACGCCATCGAACCTTATCTGGCCGATCCGGCCATGACAAAA D G R A L T N A I E P Y L A D P A M T K 320

961 CGCTGCGGCGAAAATGCTCTGGCGCATGTTCGCGAAGCCTTTCCGCTCCAAAAGGAAGCG R C G E N A L A H V R E A F P L Q K E A 340 Stop wadC 1021 GCTGCTATCAGCAGCGTTTATGAGCAGGTTTTCGCAGGAAGATAATCCGGGTCAGAGGAT A A I S S V Y E Q V F A G R * 354

CATCAACGGTTCGTCCTTGACGGGACGAATGGTGAGGGCTGTGCGTACCGAATCCACGTT

TGGAGCGGCGGTCAGTTCCTCGATCACGAAAGTCTGGAACGTGTTGAGATCGCGCGCAAC

GCAATGGAGCAGGAAGTCGGATTCACCCGAAATCATCCATGCGCGGCGAACGAGCGGCCA

TTGCTTGGTCTTTTCGGCGAAAGCCTTGAGATCCGCATCGGCCTGACGGTGCAGCCCAAC

CGAGCAGAATGCCACAAGGTCCTGCCCCAGCGTATTGCCGTTGAGAATCGCACGATAGCCG TCTTAGCGTGCTATCGGC 3` wadC-R4 5`

Figure S3.2. Sequence of wadC (BOV_1453) and its upstream and downstream regions. Start and stop codons are in green and red characters respectively. Blue characters denote intergenic nucleotides. Primers used for mutagenesis are in bold characters. Grey characters mark amino acids deleted in Bov::CAΔwadC mutant.

245

General discussion

General discussion

249

General discussion

The present thesis is a continuation of our team’s research focused on elucidating the structure of Brucella LPS, one of the main virulence factors in this bacterium.

It has been established for a long time that the presence of a smooth LPS is critical for Brucella virulence and most of the genes involved in the synthesis and assembly of the O-polysaccharide (O-PS) have been described (González et al., 2008; Lacerda et al., 2010; Mancilla et al., 2010, 2013, 2016; Martínez-Gómez et al., 2018; Monreal et al., 2003; Rajashekara et al., 2008; Vemulapalli et al., 1999). More recently, coordinated chemical and genetic analysis clearly demonstrated that the core is a branched structure and that the lateral branch not linked to the O-PS plays a critical role in the interaction of Brucella with elements of the innate immune system and in virulence, both in the animal model and in the natural host (Conde-Álvarez et al., 2012, 2013; Fontana et al., 2016; Gil-Ramírez et al., 2014; González et al., 2008; Monreal et al., 2003; Soler-Lloréns et al., 2014). In addition, the role of some of the genes that modify the lipid A backbone to rend the bacteria undetectable by bactericidal peptides have been investigated (see Appendix). Even though the progress in the understanding of LPS genetics and structure has been considerable in the last few years, there are still some points that need to be clarified and some interesting questions to answer. And this is even more crucial since the complete understanding of the LPS genetics can help to design new vaccine candidates against brucellosis (Conde-Álvarez et al., 2012, 2013; Zhao et al., 2017), the only way to control this worldwide disease.

With these ideas in mind, we first addressed an open question about genes responsible for O-PS synthesis in Brucella. Previous works had characterized genomic island-2 in Brucella. This island allocates genes wboA and wboB, responsible for O- polysaccharide synthesis. Deletion of either genes or the complete island resulted in a rough phenotype, but reversion to the smooth LPS phenotype was only possible when using the immediately upstream ORF BAB1_0998 / BMEI0999 (provisionally named wboX) (Mancilla et al., 2010; Rajashekara et al., 2008). A valid explanation for the role of these ORFs has remained evasive. Our work with mutants constructed in BAB1_0998 and BMEI0999 showed that they are not required for the synthesis of a smooth LPS in B. abortus and B. melitensis respectively. Thus, we think that the 251

General discussion explanation of the necessity of this ORF for complementation of wboA-wboB mutants could be related with genetic annotations of the genes or with the need of a promoter region still undescribed. More work is under progress to clarify this point.

We then focused on the elucidation of the genes involved in the synthesis of the diaminoglucose disaccharide that forms a lipid A in Brucella that clearly departs from the canonical glucosamine lipid A disaccharide, characteristic of gram-negative bacteria. The results suggested that BAB1_1617 and BAB1_1616 could encode GnnA and GnnB respectively, the enzymes responsible for diaminoglucose synthesis in other bacteria (Boon Hinckley et al., 2005; Sweet et al., 2004; van Mourik et al., 2010). However, confirmation that these genes account for the synthesis of the lipid A backbone was not possible. In this work, we were unable to construct mutants in those ORFs using the well-known and established protocols for Brucella mutagenesis that have been proven to be successful for mutagenesis of other genes. Interestingly, other scientific group, expert in Brucella molecular genetics that uses a completely different approach, encountered the same problem. Both approaches and results were in keeping with the idea that these two genes could be needed to construct a LPS backbone essential for the stability of Brucella outer membrane. It is coherent with the idea that Brucella has developed other strategies to modify the lipid A in order to be undetected by the immune system (see Appendix) and the resulting lipid A backbone could be unalterable to maintain bacterial viability. This seems to be different from the situation in other bacteria such as Campylobacter, that also carries diaminoglucose in their lipid A, since, in this bacterium, both gnnA and gnnB have been successfully disrupted (Mourik et al., 2010).

Our assays to investigate the enzymatic activity of BAB1_1617 and BAB1_1616 were not conclusive and should be repeated in the future using appropriate controls such as the enzymes GnnA and GnnB responsible for the synthesis of Acidithiobacillus diaminoglucose-backbone-lipid A, whose enzymatic activity has been clearly established (Sweet et al., 2004). We also envisage to carry out genetic cross- complementation between Brucella BAB1_1617 and/or BAB1_1616 and the corresponding orthologs in other bacteria that carry a diaminoglucose-derived lipid A. 252

General discussion

As an example, and since we have already successfully cloned both genes in expression vectors, we could investigate whether Brucella gnnA and/or gnnB could complement a mutant in the corresponding Campylobacter orthologues (van Mourik et al., 2010). Moreover, and as has been done in Acidithiobacillus (Sweet et al., 2004) using directed mutagenesis, we could construct a gnnB mutant in the expression plasmid pET- BAB1_1616 by replacing the lysine residue of the catalytic domain with alanine in order to inactivate the enzyme. If Brucella BAB1_1616 (gnnB) is functional, this plasmid carrying the mutated gene would serve as a negative control in the cross- complementation experiments.

The third aspect was to gain further insight into the Brucella LPS core section. Previous work had clearly demonstrated that this section is critical for interaction with elements of the immune system and virulence and the structural studies suggest that, in addition to the previously described WadB and WadC, other glycosyltransferases could be necessary for the construction of the core lateral branch not linked to the O- polysaccharide. After a careful genomic search and mutant construction and analysis, we could discard a role of ORFs BAB1_0114, BAB1_0417, BAB2_0693, BAB1_0932, BAB1_0607 or BAB2_0105 in the synthesis of a complete LPS. More interestingly, ORFs BAB1_1620, BAB2_0133, BAB2_0135, BAB1_0326 and BAB1_1465, whose expression depends on master regulators BvrR/S and MucR that control virulence and outer membrane components, were neither required for the synthesis of a smooth LPS.

One of the most interesting results of our work was the identification of a new core glycosyltransferase, WadD, required for the synthesis of the core section of the LPS not linked to the O-polysaccharide. This, together with the observation that the wadD mutant was affected in its interaction with polycationic peptides and components of serum and attenuated in later stages of infection, confirmed that Brucella LPS core oligosaccharide is a branched structure critical for cell envelope and virulence. The correlative analysis of genetics and chemical structure suggest that WadD, together with the previously identified WadB, are implicated in the assembly of the glucosamine tetrasaccharide. To complete this work, we have constructed several mutants (BmeΔperΔwadB and BmeΔperΔwadD) that will help to clarify the exact role 253

General discussion of these glycosyltransferases in the different chemical links between the glucosamine residues and whether more glycosyltransferases are needed. In our mind, and taking into account our careful genomic search, this last possibility seems improbable and, in keeping with this, analysis of the amino acid sequence of both genes already predicted a bifunctional role for WadD that could act adding glucosamine residues in two different links. Directed mutagenesis of the WadD domains that could be responsible for its bifunctional role, together with the chemical analysis of the double mutants BmeΔperΔwadB and BmeΔperΔwadD, will help to solve this interesting question.

Finally, we applied the knowledge acquired in this and previous works to address the development of new vaccines against brucellosis and tools that could help in the diagnosis of Brucella infection. As has been commented in Chapter 3, B. melitensis Rev1 is the only available vaccine to fight against B. ovis infection. However, vaccination with Rev1 is banned in countries or scenarios where B. melitensis has been eradicated (OIE, 2016). Thus, the development of B. ovis specific vaccines, in a B. ovis background is an urgent need. Previous works had demonstrated that a B. ovis wadB mutant protected against B. ovis infection (Soler-Lloréns et al., 2014). In this work, we extended these results by constructing a wadB mutant in a B. ovis CO2-independent background (Pérez-Etayo et al., 2018) and confirmed that it protected against B. ovis in the mouse model. This mutant is currently being tested in the natural host. In addition, new vaccine candidates could be investigated combining mutations in genes that affect Brucella metabolism and fitness with other(s) that alter the core lateral branch and allow a better detection by the immune system.

Concerning the diagnosis problem of B. ovis infection, as already mentioned in Chapter 3, the most widely used antigen is HS extract (Gamazo et al., 1989; Riezu-Boj et al., 1986). However, this extract contains in addition to LPS components specific of B. ovis, other antigenic epitopes shared with smooth Brucella (Santos et al., 1984), which account for the cross-reactivity that can be observed when using this antigen with sera from sheep infected with B. melitensis or vaccinated with B. melitensis Rev1 (Riezu-Boj et al., 1986). Following the OIE protocols, HS antigen is usually prepared

254

General discussion

from B. ovis REO198, a CO2- and serum-independent strain (OIE, 2015), easy to manipulate in laboratory conditions. Since Bov::CAΔwadB protected againts B. ovis infection in the mice model, we wondered whether the HS extracted from this strain could be used in a serological test for diagnosis of B. ovis. Western-blot analysis showed that the sera from an infected ram reacted with the HS from B. ovis REO198 but not with that from Bov::CAΔwadB. These results should be extended in the future using more sera from infected and vaccinated sheep. It would also be very interesting to test if the HS from Bov::CAΔwadB does not give cross reaction with sera of sheep vaccinated with Rev1. If this were the case and protection by Bov::CAΔwadB is confirmed in the natural host, this strain could be used both as a vaccine and DIVA test.

255

General discussion References

REFERENCES

Boon Hinckley, M., Reynolds, C. M., Ribeiro, A. A., McGrath, S. C., Cotter, R. J., Lauw, F. N., et al. (2005). A Leptospira interrogans enzyme with similarity to yeast Ste14p that methylates the 1-phosphate group of lipid A. J. Biol. Chem. 280, 30214–24. doi:10.1074/jbc.M506103200. Conde-Álvarez, R., Arce-Gorvel, V., Gil-Ramírez, Y., Iriarte, M., Grilló, M. J., Gorvel, J. P., et al. (2013). Lipopolysaccharide as a target for brucellosis vaccine design. Microb. Pathog. 58, 29–34. doi:10.1016/j.micpath.2012.11.011. Conde-Álvarez, R., Arce-Gorvel, V., Iriarte, M., Manček-Keber, M., Barquero-Calvo, E., Palacios-Chaves, L., et al. (2012). The lipopolysaccharide core of Brucella abortus acts as a shield against innate immunity recognition. PLoS Pathog. 8, e1002675. doi:10.1371/journal.ppat.1002675. Fontana, C., Conde-Álvarez, R., Ståhle, J., Holst, O., Iriarte, M., Zhao, Y., et al. (2016). Structural studies of lipopolysaccharide defective mutants from Brucella melitensis identify a core oligosaccharide critical in virulence. J. Biol. Chem. 291, 7727–7741. doi:10.1074/jbc.M115.701540. Gil-Ramírez, Y., Conde-Álvarez, R., Palacios-Chaves, L., Zúñiga-Ripa, A., Grilló, M.-J., Arce-Gorvel, V., et al. (2014). The identification of wadB, a new glycosyltransferase gene, confirms the branched structure and the role in virulence of the lipopolysaccharide core of Brucella abortus. Microb. Pathog. 73, 53–9. doi:10.1016/j.micpath.2014.06.002. González, D., Grilló, M.-J. M., De Miguel, M. M.-J., Ali, T., Arce-Gorvel, V., Delrue, R. R.- M., et al. (2008). Brucellosis vaccines: assessment of Brucella melitensis lipopolysaccharide rough mutants defective in core and O-polysaccharide synthesis and export. PLoS One 3, e2760. doi:10.1371/journal.pone.0002760. Lacerda, T. L. S., Cardoso, P. G., Augusto de Almeida, L., Camargo, I. L. B. da C., Afonso, D. A. F., Trant, C. C., et al. (2010). Inactivation of formyltransferase (wbkC) gene generates a Brucella abortus rough strain that is attenuated in macrophages and in mice. Vaccine 28, 5627–5634. doi:10.1016/j.vaccine.2010.06.023. Mancilla, M. (2016). Smooth to rough dissociation in Brucella: The missing link to virulence. Front. Cell. Infect. Microbiol. 5, 98. doi:10.3389/fcimb.2015.00098. Mancilla, M., Grilló, M.-J., de Miguel, M.-J., López-Goñi, I., San-Román, B., Zabalza- Baranguá, A., et al. (2013). Deletion of the GI-2 integrase and the wbkA flanking transposase improves the stability of Brucella melitensis Rev 1 vaccine. Vet. Res. 44, 105. doi:10.1186/1297-9716-44-105. Mancilla, M., López-Goñi, I., Moriyón, I., Zárraga, A. M., López-Goñi, I., Moriyón, I., et al. (2010). Genomic island 2 is an unstable genetic element contributing to Brucella lipopolysaccharide spontaneous smooth-to-rough dissociation. J. Bacteriol. 192, 6346–6351. doi:10.1128/JB.00838-10. Martínez-Gómez, E., Ståhle, J., Gil-Ramírez, Y., Zúñiga-Ripa, A., Zaccheus, M., Moriyón, I., et al. (2018). Genomic insertion of a heterologous acetyltransferase generates a 257

General discussion References

new lipopolysaccharide antigenic structure in Brucella abortus and Brucella melitensis. Front. Microbiol. 9, 1092. doi:10.3389/FMICB.2018.01092. Monreal, D., Grilló, M. J., González, D., Marín, C. M., De Miguel, M. J., López-Goñi, I., et al. (2003). Characterization of Brucella abortus O-polysaccharide and core lipopolysaccharide mutants and demonstration that a complete core is required for rough vaccines to be efficient against Brucella abortus and Brucella ovis in the mouse model. Infect. Immun. 71, 3261–71. doi:10.1128/iai.71.6.3261-3271.2003. OIE (2016). “OIE: Chapter 2.1.4. Brucellosis (Brucella abortus, B. melitensis and B. suis) (infection with B. abortus, B. melitensis and B. suis),” in Manual of Diagnostic Tests and Vaccines for Terrestrial Animals. Rajashekara, G., Covert, J., Petersen, E., Eskra, L., and Splitter, G. (2008). Genomic island 2 of Brucella melitensis is a major virulence determinant: Functional analyses of genomic islands. J. Bacteriol. 190, 6243–6252. doi:10.1128/JB.00520- 08. Soler-Lloréns, P., Gil-Ramírez, Y., Zabalza-Baranguá, A., Iriarte, M., Conde-Álvarez, R., Zúñiga-Ripa, A., et al. (2014). Mutants in the lipopolysaccharide of Brucella ovis are attenuated and protect against B. ovis infection in mice. Vet. Res. 45, 72. doi:10.1186/s13567-014-0072-0. Sweet, C. R., Ribeiro, A. A., and Raetz, C. R. H. (2004). Oxidation and transamination of the 3″-position of UDP- N -acetylglucosamine by enzymes from Acidithiobacillus ferrooxidans. J. Biol. Chem. 279, 25400–25410. doi:10.1074/jbc.M400596200. van Mourik, A., Steeghs, L., van Laar, J., Meiring, H. D., Hamstra, H.-J., van Putten, J. P. M., et al. (2010). Altered linkage of hydroxyacyl chains in lipid A of Campylobacter jejuni reduces TLR4 activation and antimicrobial resistance. J. Biol. Chem. 285, 15828–36. doi:10.1074/jbc.M110.102061. Vemulapalli, R., McQuiston, J. R., Schurig, G. G., Sriranganathan, N., Halling, S. M., and Boyle, S. M. (1999). Identification of an IS711 element interrupting the wboA gene of Brucella abortus vaccine strain RB51 and a PCR assay to distinguish strain RB51 from other Brucella species and strains. Clin. Diagn. Lab. Immunol. 6, 760–4. Zhao, Y., Hanniffy, S., Arce-Gorvel, V., Conde-Álvarez, R., Oh, S., Moriyón, I., et al. (2017). Immunomodulatory properties of Brucella melitensis lipopolysaccharide determinants on mouse dendritic cells in vitro and in vivo. Virulence, 0. doi:10.1080/21505594.2017.1386831.

258

Conclusions

Conclusions

261

Conclusions

1. We have identified the Brucella orthologs to A. ferrooxidans gnnA and gnnB, that could be responsible for the synthesis of the N-acetyl-diaminoglucose disaccharide of the LPS lipid A section. The fact that it was not possible to construct mutants in these ORFs suggests that these genes are essential for outer membrane stability and Brucella viability and that diaminoglucose can not be substituted by glucosamine in Brucella lipid A.

2. BvrR/BvrS and MucR regulated hypothetical glycosyltransferases (ORFs BAB1_1620 and BAB1_0326, BAB1_1465, BAB2_0133, BAB2_0134 and BAB2_0135 respectively) are not apparently related to the synthesis of a complete LPS in vitro. Mutants in BAB2_0133, BAB2_0134 or BAB2_0135, that are clustered in the same operon, or in BAB1_1620 are neither affected in resistance to innate immunity components nor attenuated in mice.

3. Despite their hypothetical function as glycosyltransferases, ORFs BAB2_0105, BAB2_0693, BAB1_0607, BAB1_0114, BAB1_0932 and BAB1_0417 are apparently not involved in synthesis of B. abortus LPS in vitro.

4. We have identified a new glycosyltransferase, WadD, involved in the synthesis of the core LPS section in B. abortus. Mutants in wadD generate a LPS that maintains a complete O-polysaccharide but are deficient in the synthesis of the core section. This indicates that wadD, together with the previously identified wadC and wadB, is implicated in the synthesis of the core lateral branch not linked to the O- polysaccharide that is essential for Brucella virulence.

5. In B. abortus, the core LPS section that depends on WadD hampers binding of polycationic peptides and serum components and disruption of the gene generates attenuation in the mouse model.

6. wadD is present, with minor aminoacid modifications, in the core brucellae including those that have a smooth or rough LPS, and also in the clade of early-diverging spp. that have an atypical LPS. Unexpectedly, it is absent in Brucella spp. B13-0095, one of the strains isolated from amphibians although this strain

263

Conclusions conserves wadB and wadC. Finally, WadD has an ortholog in Ochrobactrum that also belongs to the α-2 Proteobacteria and is the closest genetic neighbor of Brucella.

7. It is possible to design B. ovis specific vaccines modifying the core LPS lateral branch of a genetically engineered B. ovis strain able to grow in atmospheric conditions. A wadB mutant in this background protects against B. ovis infection in mice.

8. ORF BMEI0999 // BAB1_0998, situated immediately upstream the O-polysaccharide genes wboA and wboB is not required for the synthesis of a smooth LPS in B. melitensis or B. abortus.

264

Appendix

Appendix

267

fmicb-08-02657 January 9, 2018 Time: 16:56 # 1

ORIGINAL RESEARCH published: 10 January 2018 doi: 10.3389/fmicb.2017.02657

Identification of lptA, lpxE, and lpxO, Three Genes Involved in the Remodeling of Brucella Cell Envelope

Raquel Conde-Álvarez1, Leyre Palacios-Chaves2, Yolanda Gil-Ramírez1, Miriam Salvador-Bescós1, Marina Bárcena-Varela1, Beatriz Aragón-Aranda1, Estrella Martínez-Gómez1, Amaia Zúñiga-Ripa1, María J. de Miguel3, Toby Leigh Bartholomew4, Sean Hanniffy5, María-Jesús Grilló2, Miguel Ángel Vences-Guzmán6, José A. Bengoechea4, Vilma Arce-Gorvel5, Jean-Pierre Gorvel5, Ignacio Moriyón1* and Maite Iriarte1* Edited by: 1 Universidad de Navarra, Facultad de Medicina, Departamento de Microbiología y Parasitología, Instituto de Salud Tropical Axel Cloeckaert, (ISTUN) e Instituto de Investigación Sanitaria de Navarra (IdISNA), Pamplona, Spain, 2 Instituto de Agrobiotecnología, Institut National de la Recherche Consejo Superior de Investigaciones Científicas – Universidad Pública de Navarra – Gobierno de Navarra, Pamplona, Spain, Agronomique (INRA), France 3 Unidad de Producción y Sanidad Animal, Instituto Agroalimentario de Aragón, Centro de Investigación y Tecnología Reviewed by: Agroalimentaria de Aragón – Universidad de Zaragoza, Zaragoza, Spain, 4 Wellcome-Wolfson Institute for Experimental Diego J. Comerci, Medicine, Queen’s University Belfast, Belfast, United Kingdom, 5 Institut National de la Santé et de la Recherche Médicale, Instituto de Investigaciones U1104, Centre National de la Recherche Scientifique UMR7280, Centre d’Immunologie de Marseille-Luminy, Aix-Marseille Biotecnológicas (IIB-INTECH), University UM2, Marseille, France, 6 Centro de Ciencias Genómicas, Universidad Nacional Autónoma de México, Argentina Cuernavaca, Mexico Roy Martin Roop II, East Carolina University, United States The brucellae are facultative intracellular bacteria that cause a worldwide extended *Correspondence: Ignacio Moriyón zoonosis. One of the pathogenicity mechanisms of these bacteria is their ability to [email protected] avoid rapid recognition by innate immunity because of a reduction of the pathogen- Maite Iriarte [email protected] associated molecular pattern (PAMP) of the lipopolysaccharide (LPS), free-lipids, and other envelope molecules. We investigated the Brucella homologs of lptA, lpxE, and Specialty section: lpxO, three genes that in some pathogens encode enzymes that mask the LPS This article was submitted to Infectious Diseases, PAMP by upsetting the core-lipid A charge/hydrophobic balance. Brucella lptA, which a section of the journal encodes a putative ethanolamine transferase, carries a frame-shift in B. abortus but Frontiers in Microbiology not in other Brucella spp. and phylogenetic neighbors like the opportunistic pathogen Received: 26 October 2017 Ochrobactrum anthropi. Consistent with the genomic evidence, a B. melitensis lptA Accepted: 20 December 2017 Published: 10 January 2018 mutant lacked lipid A-linked ethanolamine and displayed increased sensitivity to Citation: polymyxin B (a surrogate of innate immunity bactericidal peptides), while B. abortus Conde-Álvarez R, Palacios-Chaves L, carrying B. melitensis lptA displayed increased resistance. Brucella lpxE encodes a Gil-Ramírez Y, Salvador-Bescos M, Bárcena-Varela M, Aragón-Aranda B, putative phosphatase acting on lipid A or on a free-lipid that is highly conserved in all Martínez-Gómez E, Zúñiga-Ripa A, brucellae and O. anthropi. Although we found no evidence of lipid A dephosphorylation, de Miguel MJ, Bartholomew TL, a B. abortus lpxE mutant showed increased polymyxin B sensitivity, suggesting the Hanniffy S, Grilló M-J, Vences-Guzmán MÁ, existence of a hitherto unidentified free-lipid involved in bactericidal peptide resistance. Bengoechea JA, Arce-Gorvel V, Gene lpxO putatively encoding an acyl hydroxylase carries a frame-shift in all brucellae Gorvel J-P, Moriyón I and Iriarte M (2018) Identification of lptA, lpxE, except B. microti and is intact in O. anthropi. Free-lipid analysis revealed that lpxO and lpxO, Three Genes Involved corresponded to olsC, the gene coding for the ornithine lipid (OL) acyl hydroxylase in the Remodeling of Brucella Cell active in O. anthropi and B. microti, while B. abortus carrying the olsC of O. anthropi Envelope. Front. Microbiol. 8:2657. doi: 10.3389/fmicb.2017.02657 and B. microti synthesized hydroxylated OLs. Interestingly, mutants in lptA, lpxE, or olsC

Frontiers in Microbiology| www.frontiersin.org 1 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 2

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

were not attenuated in dendritic cells or mice. This lack of an obvious effect on virulence together with the presence of the intact homolog genes in O. anthropi and B. microti but not in other brucellae suggests that LptA, LpxE, or OL β-hydroxylase do not significantly alter the PAMP properties of Brucella LPS and free-lipids and are therefore not positively selected during the adaptation to intracellular life.

Keywords: lipopolysaccharide, Brucella, lipids, cell envelope, PAMP

INTRODUCTION is a eukaryotic-type phospholipid required for Brucella full virulence (Comerci et al., 2006; Conde-Alvarez et al., 2006). Brucellosis is the collective name of a group of zoonotic Among the amino lipids, only the ornithine lipids (OL) have diseases afflicting a wide range of domestic and wild mammals been investigated which unlike their counterparts in Bordetella, (Whatmore, 2009; Zheludkov and Tsirelson, 2010). In domestic do not trigger the release of IL-6 or TNF-α by macrophages, livestock brucellosis is manifested mostly as abortions and possibly on account of their longer acyl chains that reduce the OL infertility, and contact with infected animals and consumption PAMP (Palacios-Chaves et al., 2011). Concerning the LPS, most of unpasteurized dairy products are the sources of human bacteria carry C1 and C40 glucosamine disaccharides with C12 brucellosis, an incapacitating condition that requires prolonged and C14 acyl and acyl-oxyacyl chains. This highly amphipathic antibiotic treatment (Zinsstag et al., 2011). Eradicated in a structure, named lipid A, is adjacent to additional negatively handful of countries, brucellosis is endemic or even increasing in charged groups of the core oligosaccharide, namely the heptose many areas of the world (Jones et al., 2013; Ducrotoy et al., 2017; phosphates and 2-keto-3-deoxyoctulosonate carboxyl groups Lai et al., 2017). (Kastowsky et al., 1992; Moriyón, 2003). This lipid A-core PAMP This disease is caused by facultative intracellular parasites is so efficiently detected by the innate immunity system that some of the genus Brucella. Taxonomically placed in the α-2 pathogens partially conceal it by removing phosphate groups or Proteobacteria (Moreno et al., 1990), the brucellae are close substituting them with arabinosamine and/or ethanolamine, or to plant pathogens and endosymbionts such as Agrobacterium, by hydroxylating the acyl chains (Takahashi et al., 2008; Lewis Sinorhizobium, and Rhizobium and to soil bacteria such et al., 2013; Moreira et al., 2013; Needham and Trent, 2013; Llobet as Ochrobactrum, the latter including some opportunistic et al., 2015; Trombley et al., 2015). In contrast, Brucella lipid pathogens, and comparative analyses suggest that soil bacteria A is a diaminoglucose disaccharide amide-linked to long (C16, of this group are endowed with properties that represent a first C18) and very long (C28–C30) acyl chains (Velasco et al., 2000; scaffold on which an intracellular life style develops (Velasco Iriarte et al., 2004; Fontana et al., 2016). Furthermore, negative et al., 2000; Moreno and Moriyón, 2007; Barquero-Calvo et al., charges in lipid A phosphates and 2-keto-3-deoxyoctulosonate 2009). The brucellae owe their pathogenicity mainly to their are counterbalanced by four glucosamine units present in the ability to multiply within dendritic cells, macrophages, and a core (Kubler-Kielb and Vinogradov, 2013; Fontana et al., 2016). variety of other cells. Due to their ability to control intracellular As illustrated by the unusually reduced endotoxicity of the trafficking and be barely detected by innate immunity, these Brucella LPS this structure is defectively detected by the innate bacteria are able to reach a safe intracellular niche before immune response (Lapaque et al., 2005; Martirosyan et al., 2011; an effective immune response is mounted, and to multiply Conde-Álvarez et al., 2012). It remains unknown, however, extensively (Gorvel and Moreno, 2002; Barquero-Calvo et al., whether Brucella LPS undergoes post-synthetic modifications 2007). A mechanism used by Brucella to scape from the that have been described for other bacteria that could alter host immune response is the interference with the toll-like its PAMP potential and contribution to virulence. In this receptor (TLR) signaling pathway by the injection of active work, we investigated in Brucella the role of gene homologs to effectors such as BtpA and BtpB through the Type IV secretion phosphatases, phospho-ethanolamine (pEtN) transferases, and system T4SS. Both effector proteins contain a TIR domain acyl hydroxylases (Figure 1) that have been shown in other that interferes with TLR signaling by directly interacting with Gram-negative pathogens to act on LPS and to contribute to MyD88 (Cirl et al., 2008; Salcedo et al., 2008, 2013; Chaudhary overcoming innate immunity defenses. et al., 2012) and contribute to the control of dendritic cell (DC) activation during infection. Moreover, Brucella has modified outer membrane (OM) components in order to reduce the MATERIALS AND METHODS pathogen-associated molecular patterns (PAMP) of the cell envelope. In Gram-negative bacteria, these PAMP are created by Bacterial Strains and Growth Conditions the conserved composition of the OM lipopolysaccharide (LPS) The bacterial strains and plasmids used in this study are listed and the free lipids on which the topology of the OM also depends. in Supplementary Table S1. Bacteria were routinely grown in However, in addition to free-lipid species present in most Gram- standard tryptic soy broth or agar either plain or supplemented negative bacteria (i.e., cardiolipin, phosphatidylglycerol, with kanamycin at 50 µg/ml, or/and nalidixic at 5 or 25 µg/ml and phosphatidylethanolamine), Brucella also possesses or/and 5% sucrose. All strains were stored in skim milk phosphatidylcholine and amino lipids. Phosphatidylcholine at −80◦C.

Frontiers in Microbiology| www.frontiersin.org 2 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 3

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

FIGURE 1 | Brucella lipid A and hypothetical sites of action of putative LpxE, LptA, and LpxO. The structure proposed is based on acyl-chain and mass spectrophotometry analyses and genomic predictions. The predicted sites of action of LpxE (phosphatase), LptA (pEtN transferase), and LpxO (acyl chain hydroxylase) are indicated, and the corresponding ORF of B. microti (green), B. melitensis (blue), and B. abortus (red) presented (NA, not annotated). The B. abortus lptA homolog and the B. melitensis and B. abortus lpxO homologs carry a frame-shift mutation.

DNA Manipulations Mutagenesis Genomic sequences were obtained from the Kyoto Encyclopedia To obtain Bme1lptA, Ba1lpxE, and Bmi1olsC in-frame deletion of Genes and Genomes (KEGG) database1. Searches for DNA mutants, directed mutagenesis by overlapping PCR were and protein homologies were carried out using the National performed using genomic DNA as template and pJQK (Scupham Center for Biotechnology Information (NCBI2) and the and Triplett, 1997) as the suicide vector. The corresponding gene European Molecular Biology Laboratory (EMBL) – European was deleted using allelic exchange by double recombination as Bioinformatics Institute server3. Primers were synthesized previously described (Conde-Alvarez et al., 2006). by Sigma-Genosys (Haverhill, United Kingdom). DNA For the construction of the Bme1lptA mutant, we first sequencing was performed by the “Servicio de Secuenciación generated two PCR fragments: oligonucleotides lptA-F1 (50-GAA del Centro de Investigación Médica Aplicada” (Pamplona, CGCGAGACTATGGAAAC-30) and lptA-R2 (50-TGGT Spain). Restriction–modification enzymes were used under the GAACGCCAGAAGATAGA-30) were used to amplify a conditions recommended by the manufacturer. Plasmid and 400-bp fragment including codons 1–26 of BmelptA ORF, chromosomal DNA were extracted with Qiaprep Spin Miniprep as well as 324 bp upstream of the BmelptA start codon, and (Qiagen) and Ultraclean Microbial DNA Isolation Kits (Mo Bio oligonucleotides lptA-F3 (50-TCTATCTTCTGGCGTTCACC Laboratories), respectively. When needed, DNA was purified GCACGACAATCTCTTC-30) and lptA-R4 (50-AATATTCCAT from agarose gels using the Qiack Gel Extraction Kit (Qiagen). GGCGCATTTC-30) were used to amplify a 472-bp fragment including codons 506–544 of the lptA ORF and 353-bp 1http://www.genome.jp/kegg/ downstream of the lptA stop codon. Both fragments were ligated 2http://www.ncbi.nlm.nih.gov/ by overlapping PCR using oligonucleotides lptA-F1 and lptA-R4 3http://www.ebi.ac.uk/ for amplification, and the complementary regions between

Frontiers in Microbiology| www.frontiersin.org 3 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 4

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

lptA-R2 and lptA-F3 for overlapping. The resulting fragment, the corresponding ORF was subcloned in plasmid pRH001 containing the lptA deleted allele, was cloned into pCR2.1 (Hallez et al., 2007) to produce pBMElpxE and pBMElptA. (Invitrogen, Barcelona, Spain), sequenced to ensure maintenance For pBMIolsC, olsC was amplified using genomic DNA of of the reading frame, and subcloned into the BamHI and the Bmi-parental as DNA template. The primers used were olsC- XbaI sites of the suicide plasmid pJQK. The resulting mutator F6 (50-GCTTTCCGAACAAGCACTGA-30) and olsC-R7 (50- plasmid (pRCI-32) was introduced in B. melitensis 16M by GCCTCCCTTCACCGGTTATT-30). The resulting PCR product, conjugation using the Escherchia coli S.17 strain (Simon et al., containing the ORF from 342 bp upstream to 84 bp downstream, 1983). was then cloned into pCR2.1 TOPO (Invitrogen) plasmid by For the construction of the Ba1lpxE mutant, we first “TA cloning” (Life Technologies). The resulting plasmid was generated two PCR fragments: oligonucleotides lpxE-F1 (50- sequenced to ensure that the gene was correctly cloned. Then, CGCGTGTGCCATAGGTATATT-30) and lpxE-R2 (50-TATAGG the gene was subcloned into the BamHI and the XbaI sites CAGGGCGCAGAA-30) were used to amplify a 482-bp fragment of the replicative plasmid pBBR1 MCS (Kovach et al., 1994) including codons 1–29 of lpxE ORF, as well as 394 bp upstream of pBMElpxE, pBMElptA, and pBMIolsC were introduced into the lpxE-1 start codon, and oligonucleotides lpxE-F3 (50-TTCTG Brucella by conjugation using E. coli S.17-1 strain and the CGCCCTGCCTATAGATTCGTTTCCGCATGGT-30) and lpxE- conjugants harboring corresponding plasmid were selected by R4 (50-CCAATACAC CCGTCATGAGA-30) were used to amplify plating onto TSA-Nal-Cm plates. a 577-bp fragment including codons 226–255 of the lpxE ORF and 488-bp downstream of the lpxE stop codon. Both fragments Sensitivity to Cationic Peptides were ligated by overlapping PCR using oligonucleotides lpxE-F1 Exponentially growing bacteria were adjusted to an optical and lpxE-R4 for amplification, and the complementary regions density equivalent to one of the McFarland scale and the between lpxE-R2 and lpxE-F3 for overlapping. The resulting minimal inhibitory concentrations (MICs) of polymyxin B were fragment, containing the lpxE deleted allele, was cloned into determined by the e-test method on Müller–Hinton agar (Izasa) pCR2.1 (Invitrogen, Barcelona, Spain), sequenced to ensure or by the serial dilution method in a similar broth. maintenance of the reading frame, and subcloned into the BamHI and the XbaI sites of the suicide plasmid pJQK (Scupham and LPS Preparation Triplett, 1997). The resulting mutator plasmid (pRCI-36) was Lipopolysaccharide was obtained by methanol precipitation of introduced in B. abortus 2308 by conjugation using the E. coli the phenol phase of a phenol–water extract (Leong et al., 1970). S.17 strain (Simon et al., 1983). This fraction [10 mg/ml in 175 mM NaCl, 0.05% NaN3, 0.1 M For the construction of the Bmi1olsC mutant, we first Tris–HCl (pH 7.0)] was then purified by digestion with nucleases generated two PCR fragments: oligonucleotides olsC-F1 (50- [50 µg/ml each of DNase-II type V and RNase-A (Sigma, St. TGCTGGATCGTATTCGTCTG-30) and olsC-R2 (50-GCCATAA Louis, MO, United States), 30 min at 37◦C] and three times GCCGATGGAACTA-30) were used to amplify a 334-bp fragment with proteinase K (50 µg/ml, 3 h at 55◦C), and ultracentrifuged including codons 1–15 of olsC ORF, as well as 289 bp upstream (6 h, 100,000 × g)(Aragón et al., 1996). Free lipids (OLs and of the olsC start codon, and oligonucleotides olsC-F3 (50-TA phospholipids) were then removed by a fourfold extraction with GTTCCATCGGCTTATGGCAGGAGGGGCTAGACAACCAC- chloroform–methanol [2:1 (vol/vol)] (Velasco et al., 2000). 30) and olsC-R4 (50-AACCAGCGACAGGGTAAGC-30) were used to amplify a 320-bp fragment including codons 286–313 Infections in Mice of the olsC ORF and 237-bp downstream of the olsC stop Seven-week-old female BALB/c mice (Charles River, Elbeuf, codon. Both fragments were ligated by overlapping PCR using France) were kept in cages with water and food ad libitum and oligonucleotides olsC-F1 and olsC-R4 for amplification, and accommodated under biosafety containment conditions 2 weeks the complementary regions between olsC-R2 and olsC-F3 for before the start of the experiments. To prepare inocula, tryptic overlapping. The resulting fragment, containing the lptA deleted soy agar (TSA) grown bacteria were harvested and suspended in allele, was cloned into pCR2.1 (Invitrogen, Barcelona, Spain), 10 mM phosphate buffered saline (pH 6.85), and 0.1 ml/mouse sequenced to ensure maintenance of the reading frame, and containing approximately 5 × 104 colony forming units (CFU) subcloned into the BamHI and the XbaI sites of the suicide for B. melitensis or B. abortus and 1 × 104 CFU for B. microti plasmid pJQK (Scupham and Triplett, 1997). The resulting was administered intraperitoneally. The exact doses assessed mutator plasmid (pRCI-65) was introduced in B. microti CM445 retrospectively by plating dilutions of the inocula. Number of by conjugation using the E. coli S.17 strain (Simon et al., 1983). CFU in spleens was determined at diferent time after inoculation. Deletion of each gene was checked with oligonucleotides For this, the spleens were aseptically removed and individually gene-F1 and gene-R4 and internal primers hybridizing in the weighed and homogenized in 9 volumes of PBS. Serial 10-fold non-deleted regions. dilutions of each homogenate were performed and each dilution was plated by triplicate. Plates were incubated at 37◦C for 5 days. Complementation of Deleted Genes At several points during the infection process, the identity of For pBMElpxE and pBMElptA construction we took advantage the spleen isolates was confirmed by PCR. The individual data of the Brucella ORFeome constructed with the Gateway cloning were normalized by logarithmic transformation, and the mean Technology (Invitrogen) (Dricot et al., 2004). The clones carrying log CFU/spleen values and the standard deviations (n = 5) were BmelpxE or BmelptA were extracted and the DNA containing calculated.

Frontiers in Microbiology| www.frontiersin.org 4 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 5

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

Intracellular Multiplication Assays Extraction and Analysis of Envelope Bone marrow cells were isolated from femurs of 7–8- Lipids week-old C57Bl/6 female and differentiated into dendritic The free-lipid fraction was extracted as described by Bligh cells [bone-marrow derived dendritic cells (BMDCs)] as and Dyer(1959), and analyzed on a silica gel 60 high- described by Inaba et al.(1992). Infections were performed performance thin layer chromatography (HPLC) plates (Merck, by centrifuging the bacteria onto the differentiated cells Darmstadt, Germany). Chromatography was performed either ◦ (400 x g for 10 min at 4 C; bacteria:cells ratio of 30:1 monodimensionally with chloroform–methanol–water [14:6:1 ◦ followed by incubation at 37 C for 30 min under a 5% (volume)] or bidimensionally with chloroform–methanol–water CO2 atmosphere). BMDCs were gently washed with medium [14:6:1 (volume)] first and chloroform–methanol–acetic acid to remove extracellular bacteria before incubating in medium [13:5:2 (volume)] in the second dimension (Weissenmayer et al., supplemented with 50 µg/ml gentamicin for 1 h to kill 2002). Plates were developed with 0.2% ninhydrin in acetone at extracellular bacteria. Thereafter, the antibiotic concentration 180◦C or 15% sulfuric acid in ethanol at 180◦C. was decreased to 10 µg/ml. To monitor Brucella intracellular survival at different time-points post-infection, BMDC were lysed with 0.1% (vol/vol) Triton X-100 in H2O and serial RESULTS dilutions of lysates were plated onto TSA plates to enumerate the CFU. The Brucella lptA Orthologs Encode a Lipid A Phosphate-Ethanolamine Flow Cytometry Transferase To assess activation and maturation, BMDC were analyzed A genomic search in the KEGG database revealed that all Brucella for surface expression of classical maturation markers at 24 h spp. carry an ORF (BMEI0118 in B. melitensis) homologous post-treatment with the different Brucella strains and derived to Neisseria meningitidis lptA, a pEtN transferase that modifies mutants. Cells were labeled with fluorochrome-conjugated lipid A (Cox et al., 2003). Strikingly, in B. abortus but not antibodies specific for mouse CD11c:APC-Cy7 (clone N418), in other Brucella spp., all genomic sequences available at IA-IE:PE (MHC class II clone M5/114.15.2) (PE), CD86:FITC KEGG show a deletion of a thymine in position 774 that (Clone GL-1), CD40:APC (clone 3/23), and CD80:PE-Cy5 (clone should result in a truncated protein lacking the amino acids 16-10A1), all from BioLegend. Labeled cells were then subjected related to the enzymatic activity (Naessan et al., 2008; Figure 1 to multi-color cytometry using a LSR II UV (Becton Dickinson) and Supplementary Figure S1 and Supplementary Table S2). and the data analyzed using FlowJo Software by first gating on In addition to LptA, two other pEtN transferases have been the CD11c+ population (100,000 events) prior to quantifying identified in N. meningitidis: Lpt-3 and Lpt-6, which, respectively, expression of receptors. Cells were stimulated with E. coli LPS modify the LPS core at the third and sixth position of heptose II (055:B5) as a positive control. (Mackinnon et al., 2002; Wright et al., 2004). By multiple sequence alignment, the B. melitensis putative pEtN transferase Lipid A Extraction showed highest homology with Neisseria LptA and also displayed Five milligrams of LPS was hydrolyzed in 5 ml 1% acetic acid by ◦ the LptA membrane-associated domains not present in Lpt-3 and sonication, heating to 100 C for 30 min, and cooling to room Lpt-6 (ORFs NMB1638, NMB2010, and NMA0408, respectively). temperature. Concentrated HCl was added to the mixture until Accordingly, it can be predicted that ORF BMEI0118 (henceforth the pH was 1–2. The solution was converted to a two-phase BMElptA) encodes a pEtN transferase that acts on lipid A, a acidic Bligh–Dyer mixture by adding 5.6 ml of chloroform and hypothesis fully consistent with the absence of heptose in the 5.6 ml of methanol. Phases were mixed by inverting the tubes Brucella LPS core (Iriarte et al., 2004; Fontana et al., 2016). and separated by centrifugation at 4000 × g for 20 min. The To test this hypothesis, we constructed a B. melitensis lower phases containing lipid A were collected, washed two times non-polar mutant (Bme1lptA) lacking the LptA enzymatic with water, and dried under a stream of nitrogen. Extraction was domain (amino acids 26–506), which as expected maintained a repeated, and the lower phases (11.2 ml) were combined and smooth (S) phenotype (negative crystal violet test and positive neutralized with a drop of pyridine. Samples were evaporated to coagglutination with anti-S-LPS antibodies). As a consequence dryness under a stream of nitrogen. of the increased positive charge of the amino group, pEtN has been shown to decrease binding of the polycationic lipopeptide Mass Spectrometry polymyxin B to LPS, and to increase resistance to this antibiotic Mass spectrometra were acquired on a Bruker Autoflex R Speed in a variety of bacteria (Needham and Trent, 2013; Trombley TOF/TOF Mass Spectrometer (Bruker Daltonics Inc.) in negative et al., 2015; Herrera et al., 2017). In keeping with this possibility, reflective mode with delayed extraction. The ion-accelerating the Bme1lptA mutant was more sensitive to polymyxin B voltage was set at 20 kV. Each spectrum was an average of 300 than the parental strain B. melitensis 16M (Bme-parental) shots. A peptide calibration standard (Bruker Daltonics Inc.) was (Figure 2A). In contrast, and consistent with the frame-shift used to calibrate the Matrix Assisted Laser Desorption/Ionization in its lptA homolog, B. abortus 2308 (Ba-parental) displayed Time-of-Flight (MALDI-TOF), and lipid A extracted from E. coli polymyxin B sensitivity similar to that of Bme1lptA. Moreover, strain MG1655 grown in LB medium at 37◦C. complementation of Bme1lptA with the multi-copy plasmid

Frontiers in Microbiology| www.frontiersin.org 5 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 6

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

both constructs kept the S type features (negative crystal violet test and positive coagglutination with anti-S-LPS antibodies) of the parental strains. N. gonorrhoeae shows increased resistance to the action of complement in non-immune serum that is dependent on lipid A-linked pEtN (Lewis et al., 2013). Testing for a similar contribution here, we found that Bme1lptA was more sensitive than either the parental strain or the complemented mutant (25% vs. no decrease in viability after 3 h of incubation in normal sheep serum) relevant given that B. melitensis is characteristically resistant to killing by normal serum. By MALDI-TOF analysis, the lipid A of Bme-parental was found to contain four main clusters of ions (A, B, C, and D in Figure 2B). Bme1lptA lipid A was qualitatively identical to Bme-parental with respect to groups A, B, and C but clearly differed in group D (Figure 2B and Supplementary Table S3). In group D, the 2191 m/z00 species of Ba-parental was consistent with the isotopic mass of a molecule (C120H232N4O25P2) formed by a hexaacylated and bisphosphorylated diaminoglucose disaccharide carrying the hydroxylated long and very long chain acyl groups characteristic of Brucella (Velasco et al., 2000; Ferguson et al., 2004). According to this interpretation, the signal(s) at 2112 m/z (mass of - H2PO3 o - HPO3 -, 80.9 - 79.9) could correspond to a monophosphorylated (C120H232N4O25P) 2191 m/z00 equivalent. Substitution of this monophosphorylated + form with pEtN ( H3NCH2-CH2- HPO3 mass 125) should account for signal m/z 2237, in keeping with the fact that m/z 2237 did not appear in the spectrum of the lipids A from either Bme1lptA or Ba-parental (Figure 2B). Although a clear cut demonstration requires direct analyses of the enzymatic analyses of LptA, these results and the homologies with LptA of other bacteria are consistent with the hypothesis that LptA acts as a pEtN transferase in B. melitensis and lacks functionality in B. abortus. It is remarkable that pEtN activity was detected for only a fraction (D) of lipid A species. This could be explained by a preferential activity of the enzyme for higher MW lipid A molecules.

The Brucella lpxE Orthologs Encode a Phosphatase Involved in the Remodeling of the OM As described above, MALDI-TOF analyses showed the presence of molecular species with a mass compatible with monophosphorylated lipid A. Since lipid A synthesis 0 FIGURE 2 | The Brucella lptA orthologs are involved in polymyxin B resistance produces C1 and C4 bisphosphorylated disaccharide backbones and code for a phosphate-ethanolamine transferase acting on lipid A. (Qureshi et al., 1994), a possible explanation could be its (A) Polymyxin B sensitivity of B. melitensis wild-type (Bme-parental), dephosphorylation by a phosphatase such as LpxE, an inner B. melitensis non-polar lptA mutant (BmelptA), the cognate complemented membrane enzyme that in the phylogenetic neighbor Rhizobium mutant (Bme1lptApBMElptA), B. abortus wild-type (Ba-parental), and B. abortus wild-type carrying a plasmid with the B. melitensis lptA gene leguminosarum removes the lipid A phosphate at C1 (Raetz (Ba-parentalpBMElptA) (the results are representative of three independent et al., 2009). A search in KEGG showed that all Brucella experiments). (B) MALDI-TOF analysis of the lipid A of Bme-parental, spp. carry an ORF homologous to R. leguminosarum lpxE Bme1lptA, and Ba-parental. (Supplementary Table S2). However, the start codon in the B. melitensis 16M homolog (BMEI1212) is annotated to a position different from that determined for other brucellae pBMElptA or its introduction into B. abortus 2308 leads to (Supplementary Table S2), including other B. melitensis strains. restoration of polymyxin B resistance in Bme1lptA or an increase Thus, whereas the B. abortus homolog (BAB1_0671) is predicted up to B. melitensis level in B. abortus (Figure 2A). As expected to encode a protein of 255 amino acids, the B. melitensis

Frontiers in Microbiology| www.frontiersin.org 6 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 7

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

one could encode a protein of either 235 or 255 amino acids this ORF presents a frame-shift leading to a truncated protein that (Figure 1). Both proteins conserve the consensus sequence of lacks the consensus of the aspartyl/asparaginyl β-hydroxylases the lipid phosphatase superfamily [KX6RP-(X12–54)-PSGH- family to which LpxO belongs (Figure 1 and Supplementary (X31–54)-SRX5HX3D] (Stukey and Carman, 2008) which is also Table S2). These characteristics are consistent with chemical present in LpxE from R. leguminosarum, Sinorhizobium meliloti, studies that previously failed to observe S2 hydroxylated fatty and Agrobacterium tumefaciens (Karbarz et al., 2003). Although acids in B. abortus lipid A (Velasco et al., 2000). Moreover, a BAB1_0671 and BMEI1212 code for proteins that contain the lpxO homolog is present in Ochrobactrum anthropi where S2 three motifs conserved in the LpxF phosphatase from Francisella, hydroxylated fatty acids were also not observed in the lipid A they lack two amino acids of the central motif, NCSFX2G, which (Velasco et al., 2000), indicating that a role similar to that of seems LpxF specific (Wang et al., 2006, 2007). Thus, the Brucella Salmonella LpxO is unlikely. Thus, the lpxO homologs present proteins were named BALpxE and BMELpxE. in these B. microti and O. anthropi could be acting on a free To study whether BALpxE actually acts as a lipid A lipid and, in fact, it has been reported that the corresponding phosphatase, we constructed a non-polar mutant (Ba1lpxE) R. tropici homolog is a β-hydroxylase acting on OLs (Vences- and tested it against polymyxin B, since the permanence of a Guzmán et al., 2011). If this were the case in O. anthropi phosphate group in an OM molecule should increase sensitivity and the brucellae, the end product [a hydroxylated OL (OH– to this antibiotic. Mutant Ba1lpxE was eight times more sensitive OL)] of the pathway described previously in members of the than the parental strain (MIC 0.2 and 1.6 µg/ml, respectively). Rhizobiaceae (Figure 3A) should be observed in O. anthropi Moreover, when we introduced a plasmid containing the and B. microti (and B. vulpis) but not in other Brucella BMElpxE ortholog into Ba1lpxE, the resistance to polymyxin B spp. was restored (MIC 1.6 µg/ml). Although final confirmation of To investigate these hypotheses, we compared the free lipids this interpretation would require to assay the enzymatic activity of B. abortus, B. melitensis, B. suis, B. ovis, B. microti, and of the protein, these results are consistent with the predicted role O. anthropi. As can be seen in Figure 3B, B. microti but of lpxE as a phosphatase and its functionality in both B. abortus not B. abortus produced an amino lipid with the migration and B. melitensis 16M, a strain where the annotation of the start pattern predicted for OH-OL (Vences-Guzmán et al., 2011), codon was a source of ambiguity. and results similar to those of B. microti were obtained for By MALDI-TOF analysis, the Ba-parental lipid A spectrum O. anthropi but not for the other Brucella spp. tested (not showed three of the four predominant clusters of ions (A, B, shown). These observations support the interpretation that and C) found in B. melitensis (Figure 2B and Supplementary O. anthropi and B. microti LpxO are OL hydroxylases and are Table S3). Cluster A (m/z 2173) was consistent with an fully consistent with the aforementioned genomic and chemical hexaacylated bisphoshorylated diaminoglucose disaccharide evidence. Accordingly, Brucella lpxO should be named olsC. To (C120H232N4O24P2) and the signal at 2093 m/z, which differed confirm this, we examined the amino lipids of a non-polar olsC in the mass of one phosphate group (i.e., 80), was consistent with mutant in B. microti (Bmi1olsC). As predicted, this mutant did the cognate monophosphorylated lipid A (C120H232N4O25P) not synthesize OH–OL and complementation with a plasmid (A-Pi, Figure 2). Other signals differing in a mass of 14 or 28 containing B. microti olsC restored the wild-type phenotype units should result from the heterogeneity in acyl chain length (Figure 3C). Furthermore, introducing this plasmid or a plasmid that is typical of lipid A. The B and C clusters also contained carrying O. anthropi olsC into B. abortus resulted in the signals differing in 80 mass units that could correspond to synthesis of OH–OL (Figure 3C and Supplementary Figure S2). bis- and mono-phosphorylated species. The mass spectrum of No difference in polymyxin sensitivity was observed in these Ba1lpxE lipid A (not shown) did not differ significantly from constructs or the mutant Bmi1olsC when compared to the that of Ba-parental, and again showed acyl chain heterogeneity corresponding parental strains. in the A, B, C clusters, as well as the −80 m/z signals indicative of mono- and bisphoshorylated lipid A species. As mutation of LptA, LpxE, and OlsC Are Not Required lpxE is concomitant with an increase in polymyxin B sensitivity, it is tempting to speculate that LpxE directly or indirectly for Brucella Virulence in Laboratory modulates Brucella cell envelope by removing an accessible Models phosphate group from a substrate different from lipid A. Brucella abortus, B. melitensis, and B. suis have been shown Further studies need to be performed to clarify the role of to multiply in murine and human monocyte-derived dendritic LpxE. cells while interfering with their activation and maturation and reducing both antigen presentation and an effective adaptive The Brucella lpxO Orthologs Encode an response (Billard et al., 2007; Martirosyan et al., 2011; Conde- Álvarez et al., 2012; Gorvel et al., 2014; Papadopoulos et al., Acyl Hydroxylase Acting on Ornithine 2016). To assess whether LptA, LpxE, and OL β-hydroxylase Lipids (OlsC) were involved, we compared parental and mutant strains The genomes of all Brucella species available at KEGG contain of B. melitensis, B. abortus, and B. microti in mouse BMDCs. an ORF homologous to Salmonella lpxO (Gibbons et al., 2000), As shown in Figure 4, the kinetics of multiplication of the which encodes an enzyme hydroxylating the 30-secondary acyl mutants and wild-type strains were similar. We also performed chain of lipid A. In all Brucella spp. except B. microti and B. vulpis a phenotypic characterization of MHC II and co-stimulatory

Frontiers in Microbiology| www.frontiersin.org 7 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 8

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

FIGURE 3 | The Brucella lpxO orthologs encode an acyl hydroxylase acting on ornithine lipids. (A) Pathway of synthesis of ornithine lipids in α-2 Proteobacteria (adapted from Geiger et al., 2010); the ORFs of B. abortus and B. microti are indicated, whereas B. microti, B. vulpis, and O. anthropi contain an intact olsC acyl hydroxylase gene, B. abortus and other Brucella spp. carry a frame-shift in the olsC homolog. (B) Lipid profile of B. abortus wild-type (Ba-parental) and B. microti wild-type (Bmi-parental) showing the absence or presence, respectively, of OH–OL. (C) Amino lipid profile of B. abortus wild-type (Ba-parental), B. microti wild-type (Bmi-parental), B. microti deleted in olsC (Bmi1olsC), the cognate reconstituted mutant (Bmi1olsCpOlsC), and B. abortus wild-type carrying a plasmid with the B. microti olsC gene (Ba-parentalpOlsC).

FIGURE 4 | LptA, LpxE, and OlsC deletions do not alter the Brucella interaction with dendritic cells. Intracellular replication in BMDCs (each point receptors CD86 and CD80 (Figure 5). In agreement with represents the mean ± standard error of the logarithm of CFU in dendritic previous studies, these analyses showed that activation and cells). maturation was only partially induced in BMDC infected with B. melitensis and B. abortus (Martirosyan et al., 2011). In addition, a similar partial-activation profile was evident both for alter the CFU/spleen profile produced by this species which is B. microti, for which no previous studies exist in infected BMDC, characterized by a lower lethal dose in mice as well as a faster and all of the tested mutants obtained for each of the three clearance from mouse spleens (Jiménez de Bagüés et al., 2010; Brucella spp. Figure 6, lower left panel). Moreover, when we tested whether The mouse model has been widely used for testing Brucella the expression of B. microti olsC in B. abortus could affect virulence (Grilló et al., 2012). In this model, the LptA and virulence, we found no differences between the B. microti olsC- LpxE mutants and the parental strains behaved identically carrying and the wild-type B. abortus strains (Figure 6, lower (Figure 6 upper panels). Deletion of olsC in B. microti did not right panel).

Frontiers in Microbiology| www.frontiersin.org 8 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 9

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

FIGURE 5 | LptA, LpxE, and OlsC deletions do not significantly impact the intrinsic immunogenicity of Brucella. Each point represents the mean ± standard error of the median intensity of surface receptor expression in dendritic cells treated with Brucella strains or derived mutants. E. coli LPS was used as a positive control for dendritic cell activation.

DISCUSSION pEtN promotes binding to N. gonorrhoeae lipid A of factors that downregulate the complement cascade and thwart building of In this work we investigated three Brucella ORFs that according the membrane-attack complex and opsonophagocytosis (Lewis to homologies with genes of known function in other pathogens et al., 2013). N. meningitidis pEtN also promotes adhesion of could modify the lipid A and contribute to further altering the non-encapsulated bacteria to endothelial cells (Takahashi et al., LPS PAMP of representative Brucella species. The results show 2008). Indeed, properties that parallel some of those observed that, whereas Brucella LptA modifies the lipid A, this is not the for the above-listed pathogens can also be attributed to the pEtN case for lpxE and lpxO (redesignated olsC), the former encoding a transferase counterpart in Brucella. An intact lptA was related to putative phosphatase acting on an unidentified OM molecule and polymyxin B resistance in B. melitensis and the introduction of the latter for an enzyme with OlsC activity. B. melitensis lptA into B. abortus increases polymyxin B resistance Our data strongly suggest that B. melitensis LptA is involved to the level of B. melitensis, suggesting that LptA function is in the addition of pEtN to lipid A, homologous proteins severally impaired in B. abortus. This is in agreement with carrying out this function are not uncommon in Gram-negative the presence of a frame-shift in B. abortus lptA encompassing pathogens and modulate the properties of lipid A. In Salmonella the consensus sequence, which makes likely that it codes for a Typhimurium, Shigella flexneri, E. coli, Vibrio cholerae, protein with no or residual enzymatic activity. Previous analyses Helicobacter pylori, Haemophilus ducreyi, N. gonorrhoeae, are contradictory with regard to the presence (Casabuono et al., and N. meningitidis pEtN reduces the binding of cationic 2017) or absence (Moreno et al., 1990) of ethanolamine in bactericidal peptides by balancing the negative charge of lipid A B. abortus lipid A but the materials analyzed differ in methods (Needham and Trent, 2013; Trombley et al., 2015). Conversely, of extraction and presence of B. abortus lipid A markers, such as

Frontiers in Microbiology| www.frontiersin.org 9 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 10

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

FIGURE 6 | The OM properties that depend on LptA, LpxE, and OlsC are not required for Brucella virulence in the mouse model. BALB/c mice were inoculated intraperitoneally with 5 × 104 (Bme-parental, Bme1lptA, Ba-parental, and Ba1lpxE) or 1 × 104 (Bmi-parental and Bmi1olsC) CFU/mouse and CFU/spleen determined at the indicated times. Each point represents the mean ± standard error of the logarithm of CFU in the spleens of five animals.

very long chain fatty acids (VCLFA). Although further chemical displayed by the LPS of B. abortus and B. melitensis (Martínez and enzymatic analyses are necessary for a definite conclusion, de Tejada et al., 1995; Lapaque et al., 2006; Conde-Álvarez our results strongly suggest that, if present, pEtN is in much et al., 2012), our results do not support a role for BALpxE as less amounts in B. abortus than in B. melitensis lipid A. It is a lipid A phosphatase. This is consistent with genomic analysis also worth noting that such genetic and phenotypic differences showing that, whereas in bacteria where LpxE acts on lipid in the lipid A of B. abortus and B. melitensis could relate to A the gene is located together with lptA in an operon (Tran differences in biological properties. The LPS of B. abortus and et al., 2006; Renzi et al., 2015), Brucella lpxE is instead located B. melitensis is a poor activator of the complement cascade, and upstream of three sequences annotated as pseudogenes and this property has been traced to the core and lipid A structure downstream, but in the opposite direction, of a cystathionine (Moreno et al., 1981; Conde-Álvarez et al., 2012; Fontana et al., beta-lyase. On the basis of the data shown here, the origin of 2016). Since B. abortus is less resistant than B. melitensis to monophosphoryl lipid A in Brucella remains to be explained. normal serum (González et al., 2008), it is tempting to suggest Further, we believe it unlikely to be an artifact resulting from that, like in N. gonorrhoeae, B. melitensis pEtN could sequester the hydrolytic steps used to obtain lipid A and instead favor regulatory elements enhancing complement resistance in this the hypothesis of the existence of an as yet unidentified lipid A species. phosphatase. Concerning LpxE, phosphatases acting on lipid A have at least LpxE belongs to the type 2 family of phosphatases that can been shown in Francisella tularensis, H. pylori, Porphyromonas act on lipid A but also on phosphatidylglycerol phosphate, gingivalis, and Capnocytophaga canimorsus, bacteria where lipid phosphatidic acid, sphingosine phosphate, and lysophosphatidic A dephosphorylation is involved both in resistance to bactericidal acid (Brindley and Waggoner, 1998; Sciorra and Morris, peptides and the reduction of TLR-4-dependent recognition 2002). Significantly, LpxE from Agrobacterium, although (Needham and Trent, 2013). Although these properties are predicted to be a lipid A phosphatase, dephosphorylates

Frontiers in Microbiology| www.frontiersin.org 10 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 11

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

phosphatidyl glycerophosphate (Karbarz et al., 2009) to generate lpxE, or olsC do not play a role in the ability of Brucella to replicate phosphatidylglycerol, a cell envelope phospholipid. Indeed, in BMDC and do not modulate the activation and maturation a hypothetical phosphatidyl glycerophosphate phosphatase profile in these cells. Similarly, the mouse model did not reveal activity of Brucella LpxE could account for both the polymyxin any effect on its ability to colonize and multiply in the spleen. B sensitivity of the mutated bacteria and the unaltered mass However, further experimental work in the natural hosts and spectra of the lipid A of the mutant. Such a modification alternative routes of infection might provide evidence on the of a phospholipid could be meaningful by itself on account role in virulence of these genes. The fact that lptA and olsC are of the LpxE-dependent bactericidal peptide resistance but not functional in all Brucella spp. must therefore be considered there are other possibilities. In some bacteria (i.e., Rhizobium) in the context of the models used. While the absence of a phosphatidylglycerol is a precursor for the synthesis of amino functional lptA in B. abortus suggests that the gene is not essential lipids such as lysyl-phosphatidylglycerol. This synthesis is for the virulence of this species we cannot conclude it to be induced by acid pH and brings about resistance to daptomycin totally irrelevant. Differences between B. melitensis and B. abortus and polymyxin B (Sohlenkamp et al., 2007; Ernst and Peschel, related to lptA could explain the higher invasiveness of the former 2011; Arendt et al., 2012). Interestingly, whereas the Ba1lpxE species noted by early researchers in studies carried out in guinea mutant is impaired for growth at pH 6, the parental B. abortus pigs, animals that are highly susceptible to brucellosis (Braude, becomes more resistant to cationic peptides (L. Palacios-Chaves 1951). This possibility together with the presence of intact lptA and R. Conde-Álvarez, Unpublished observations). These and olsC in Ochrobactrum and B. microti is also compatible with observations suggest the existence in Brucella of pH-dependent the hypothesis that they represent ancestral characters that are envelope modifications that require a functional LpxE. Research liable to be lost in the absence of a selective pressure during is in progress to elucidate the mechanisms behind the increased the intracellular life cycle or, in the case of lptA, that is no resistance at acid pH and the implication regarding a role for longer present in the ruminant host species (i.e., cattle) to which LpxE. B. abortus is characteristically associated. In S. Typhimurium, Pseudomonas aeruginosa, Bordetella bronchispetica, Legionella pneumophila, and Klebsiella pneumonia, LpxO is a Fe2+/α-ketoglutarate-dependent ETHICS STATEMENT dioxygenase that catalyzes the hydroxylation of the 30-secondary acyl chain of lipid A. LpxO has been implicated indirectly in Female BALB/c mice (Charles River, France) were kept in stress responses at the envelope level (Needham and Trent, cages with water and food ad libitum under P3 biosafety 2013) and, in K. pneumoniae, it has been shown to be relevant conditions in the facilities of “Centro de Investigación Médica in vivo by increasing bactericidal peptide resistance and reducing Aplicada” (registration code ES31 2010000132) 2 weeks before the inflammatory responses (Llobet et al., 2015). However, as and during the experiments. The procedures were in accordance discussed above, previous chemical analysis (Velasco et al., with the current European (directive 86/609/EEC) and Spanish 2000) of lipid A and the evidence presented here indicate that (RD 53/2013) legislations, supervised by the Animal Welfare the Brucella lpxO homolog is not a lipid A hydroxylase but Committee of the University of Navarra, and authorized by the rather an OlsC whose mutation, in contrast with LpxO, does “Gobierno de Navarra” [CEEA045/12 and E36-14 (045-12E1)]. not result in increased sensitivity to polymyxin B. This absence of an effect on polycation resistance is in keeping with both the lack of activity on lipid A and the fact that OL do no play AUTHOR CONTRIBUTIONS a major role in resistance to polycationic bactericidal peptides in B. abortus (Palacios-Chaves et al., 2011). At the same time, it IM, MI, J-PG, JB, and RC-Á conceived the study. RC-Á, LP-C, would also appear to rule out, the involvement of this protein in YG-R, MB-V, MS-B, BA-A, EM-G, AZ-R, MdM, TLB, SH, M-JG, the metabolism of succinate in B. microti as has been previously MV-G, and VA-G carried out the experimental work. IM, MI, suggested (Audic et al., 2009). and RC-Á wrote the paper. All authors participated in the Previous data showing lptA, lpxE, and lpxO to be involved presentation and discussion of results. in modulating the properties of the OM in a way that in some cases confers in vitro resistance to innate immunity bactericidal peptides, complement, and cytokine responses (Needham and FUNDING Trent, 2013) have been drawn upon as evidence for a role in virulence. However, to the best of our knowledge, a This research was supported by the Institute for Tropical Health role in vivo has thus far been shown only for lpxO from funders (Obra Social la CAIXA, Fundaciones Caja Navarra K. pneumoniae (Llobet et al., 2015). Moreover, contrasting results and Roviralta, PROFAND, Ubesol, ACUNSA, and Artai) and have been obtained with mutants both showing bactericidal grants MINECO (AGL2014-58795-C4-1-R, Bru-Epidia 291815- peptide sensitivity in vitro and no phenotype in vivo have been FP7/ERANET/ANIHWA), Aragón Government (Consolidated reported for at least H. ducreyi (Trombley et al., 2015) and may Group A14), and Marie Curie Career Integration Grant U-KARE reflect the complexities of the infection processes and/or the (PCIG13-GA-2013-618162). TLB is the recipient of a Ph.D. inadequacies of the currently available in vivo models. Despite Fellowship funded by the Department for Employment and their effect on the envelope, our results show that Brucella lptA, Learning (Northern Ireland, United Kingdom).

Frontiers in Microbiology| www.frontiersin.org 11 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 12

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

ACKNOWLEDGMENT SUPPLEMENTARY MATERIAL

The authors thank A. Delgado-López for excellent The Supplementary Material for this article can be found technical assistance in the extraction and purification online at: https://www.frontiersin.org/articles/10.3389/fmicb. of LPS. 2017.02657/full#supplementary-material

REFERENCES Cox, A. D., Wright, J. C., Li, J., Hood, D. W., Moxon, E. R., and Richards, J. C. (2003). Phosphorylation of the lipid A region of meningococcal Aragón, V., Díaz, R., Moreno, E., and Moriyón, I. (1996). Characterization of lipopolysaccharide: identification of a family of transferases that add Brucella abortus and Brucella melitensis native haptens as outer membrane phosphoethanolamine to lipopolysaccharide. J. Bacteriol. 185, 3270–3277. O-type polysaccharides independent from the smooth lipopolysaccharide. doi: 10.1128/jb.185.11.3270-3277.2003 J. Bacteriol. 178, 1070–1079. doi: 10.1128/jb.178.4.1070-1079.1996 Dricot, A., Rual, J. F., Lamesch, P., Bertin, N., Dupuy, D., Hao, T., et al. (2004). Arendt, W., Hebecker, S., Jäger, S., Nimtz, M., and Moser, J. (2012). Generation of the Brucella melitensis ORFeome version 1.1. Genome Res. 14, Resistance phenotypes mediated by aminoacyl-phosphatidylglycerol synthases. 2201–2206. doi: 10.1101/gr.2456204 J. Bacteriol. 194, 1401–1416. doi: 10.1128/JB.06576-11 Ducrotoy, M., Bertu, W. J., Matope, G., Cadmus, S., Conde-Álvarez, R., Gusi, Audic, S., Lescot, M., Claverie, J.-M., and Scholz, H. C. (2009). Brucella microti: A. M., et al. (2017). Brucellosis in sub-saharan Africa: current challenges for the genome sequence of an emerging pathogen. BMC Genomics 10:352. management, diagnosis and control. Acta Trop. 165, 179–193. doi: 10.1016/j. doi: 10.1186/1471-2164-10-352 actatropica.2015.10.023 Barquero-Calvo, E., Chaves-Olarte, E., Weiss, D. S., Guzmán-Verri, C., Chacón- Ernst, C. M., and Peschel, A. (2011). Broad-spectrum antimicrobial peptide Díaz, C., Rucavado, A., et al. (2007). Brucella abortus uses a stealthy strategy resistance by MprF-mediated aminoacylation and flipping of phospholipids. to avoid activation of the innate immune system during the onset of infection. Mol. Microbiol. 80, 290–299. doi: 10.1111/j.1365-2958.2011.07576.x PLOS ONE 2:e631. doi: 10.1371/journal.pone.0000631 Ferguson, G. P., Datta, A., Baumgartner, J., Roop, R. M., Carlson, R. W., Walker, Barquero-Calvo, E., Conde-Alvarez, R., Chacón-Díaz, C., Quesada-Lobo, L., G. C., et al. (2004). Similarity to peroxisomal-membrane protein family reveals Martirosyan, A., Guzmán-Verri, C., et al. (2009). The differential interaction that Sinorhizobium and Brucella BacA affect lipid-A fatty acids. Proc. Natl. Acad. of Brucella and Ochrobactrum with innate immunity reveals traits related to the Sci. U.S.A. 101, 5012–5017. doi: 10.1073/pnas.0307137101 evolution of stealthy pathogens. PLOS ONE 4:e5893. doi: 10.1371/journal.pone. Fontana, C., Conde-Álvarez, R., Ståhle, J., Holst, O., Iriarte, M., Zhao, Y., 0005893 et al. (2016). Structural studies of lipopolysaccharide-defective mutants from Billard, E., Dornand, J., and Gross, A. (2007). Brucella suis prevents human Brucella melitensis identify a core oligosaccharide critical in virulence. J. Biol. dendritic cell maturation and antigen presentation through regulation of tumor Chem. 291, 7727–7741. doi: 10.1074/jbc.M115.701540 necrosis factor alpha secretion. Infect. Immun. 75, 4980–4989. doi: 10.1128/IAI. Geiger, O., González-Silva, N., López-Lara, I. M., and Sohlenkamp, C. (2010). 00637-07 Amino acid-containing membrane lipids in bacteria. Prog. Lipid Res. 49, 46–60. Bligh, E. M., and Dyer, W. J. (1959). A rapid method of total lipid extraction and doi: 10.1016/j.plipres.2009.08.002 purification. Can. J. Biochem. Physiol. 37, 911–917. doi: 10.1139/y59-099 Gibbons, H. S., Lin, S., Cotter, R. J., and Raetz, C. R. H. (2000). Oxygen Braude, A. I. (1951). Studies in the pathology and pathogenesis of experimental requirement for the biosynthesis of the S -2-hydroxymyristate moiety in brucellosis: I. A comparison of the pathogenicity of Brucella abortus, Brucella Salmonella typhimurium lipid A. J. Biol. Chem. 275, 32940–32949. doi: 10.1074/ melitensis, and Brucella suis for Guinea pigs. J. Infect. Dis. 89, 76–86. jbc.M005779200 doi: 10.1093/infdis/89.1.76 González, D., Grilló, M.-J., De Miguel, M.-J., Ali, T., Arce-Gorvel, V., Delrue, Brindley, D. N., and Waggoner, D. W. (1998). Mammalian lipid phosphate R.-M., et al. (2008). Brucellosis vaccines: assessment of Brucella melitensis phosphohydrolases. J. Biol. Chem. 273, 24281–24284. doi: 10.1074/JBC.273.38. lipopolysaccharide rough mutants defective in core and O-polysaccharide 24281 synthesis and export. PLOS ONE 3:e2760. doi: 10.1371/journal.pone.000 Casabuono, A. C., Czibener, C., Del Giudice, M. G., Valguarnera, E., Ugalde, J. E., 2760 and Couto, A. S. (2017). New features in the lipid A structure of Brucella Gorvel, J.-P., and Moreno, E. (2002). Brucella intracellular life: from invasion suis and Brucella abortus lipopolysaccharide. J. Am. Soc. Mass Spectrom. 28, to intracellular replication. Vet. Microbiol. 90, 281–297. doi: 10.1016/S0378- 2716–2723. doi: 10.1007/s13361-017-1805-x 1135(02)00214-6 Chaudhary, A., Ganguly, K., Cabantous, S., Waldo, G. S., Micheva-Viteva, S. N., Gorvel, L., Textoris, J., Banchereau, R., Ben Amara, A., Tantibhedhyangkul, W., Nag, K., et al. (2012). The Brucella TIR-like protein TcpB interacts with the von Bargen, K., et al. (2014). Intracellular bacteria interfere with dendritic death domain of MyD88. Biochem. Biophys. Res. Commun. 417, 299–304. cell functions: role of the type I interferon pathway. PLOS ONE 9:e99420. doi: 10.1016/j.bbrc.2011.11.104 doi: 10.1371/journal.pone.0099420 Cirl, C., Wieser, A., Yadav, M., Duerr, S., Schubert, S., Fischer, H., et al. (2008). Grilló, M.-J., Blasco, J. M., Gorvel, J. P., Moriyón, I., and Moreno, E. (2012). Subversion of Toll-like receptor signaling by a unique family of bacterial What have we learned from brucellosis in the mouse model? Vet. Res. 43:29. Toll/interleukin-1 receptor domain–containing proteins. Nat. Med. 14, doi: 10.1186/1297-9716-43-29 399–406. doi: 10.1038/nm1734 Hallez, R., Letesson, J.-J., Vandenhaute, J., and De Bolle, X. (2007). Gateway-based Comerci, D. J., Altabe, S., de Mendoza, D., and Ugalde, R. A. (2006). destination vectors for functional analyses of bacterial ORFeomes: application Brucella abortus synthesizes phosphatidylcholine from choline provided to the min system in Brucella abortus. Appl. Environ. Microbiol. 73, 1375–1379. by the host. J. Bacteriol. 188, 1929–1934. doi: 10.1128/JB.188.5.1929-1934. doi: 10.1128/AEM.01873-06 2006 Herrera, C. M., Henderson, J. C., Crofts, A. A., and Trent, M. S. (2017). Novel Conde-Álvarez, R., Arce-Gorvel, V., Iriarte, M., Mancek-Keber,˘ M., Barquero- coordination of lipopolysaccharide modifications in Vibrio cholerae promotes Calvo, E., Palacios-Chaves, L., et al. (2012). The lipopolysaccharide core of CAMP resistance. Mol. Microbiol. 106, 582–596. doi: 10.1111/mmi.13835 Brucella abortus acts as a shield against innate immunity recognition. PLOS Inaba, K., Inaba, M., Romani, N., Aya, H., Deguchi, M., Ikehara, S., Pathog. 8:e1002675. doi: 10.1371/journal.ppat.1002675 et al. (1992). Generation of large numbers of dendritic cells from mouse Conde-Alvarez, R., Grilló, M. J., Salcedo, S. P., de Miguel, M. J., Fugier, E., bone marrow cultures supplemented with granulocyte/macrophage colony- Gorvel, J. P., et al. (2006). Synthesis of phosphatidylcholine, a typical eukaryotic stimulating factor. J. Exp. Med. 176, 1693–1702. doi: 10.1084/jem.176. phospholipid, is necessary for full virulence of the intracellular bacterial parasite 6.1693 Brucella abortus. Cell. Microbiol. 8, 1322–1335. doi: 10.1111/j.1462-5822.2006. Iriarte, M., González, D., Delrue, R. M., Monreal, D., Conde-Álvarez, R., López- 00712.x Goñi, I., et al. (2004). “Brucella LPS: structure, biosynthesis and genetics,” in

Frontiers in Microbiology| www.frontiersin.org 12 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 13

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

Brucella: Molecular and Cellular Biology, eds I. López-Goñi and I. Moriyón Moreno, E., and Moriyón, I. (2007). “The genus Brucella,” in The Prokaryotes: (Gurgaon: Horizon Bioscience), 159–192. Proteobacteria, Vol. 5, eds S. Falkow, E. Rosenberg, K. H. Schleifer, E. Jiménez de Bagüés, M. P., Ouahrani-Bettache, S., Quintana, J. F., Mitjana, O., Stackebrandt, and M. Dworkin (Berlin: Springer-Verlag), 315–455. Hanna, N., Bessoles, S., et al. (2010). The new species Brucella microti replicates Moreno, E., Stackebrandt, E., Dorsch, M., Wolters, J., Busch, M., and in macrophages and causes death in murine models of infection. J. Infect. Dis. Mayer, H. (1990). Brucella abortus 16S rRNA and lipid A reveal a 202, 3–10. doi: 10.1086/653084 phylogenetic relationship with members of the alpha-2 subdivision of the class Jones, B. A., Grace, D., Kock, R., Alonso, S., Rushton, J., Said, M. Y., et al. (2013). Proteobacteria. J. Bacteriol. 172, 3569–3576. doi: 10.1128/jb.172.7.3569-3576. Zoonosis emergence linked to agricultural intensification and environmental 1990 change. Proc. Natl. Acad. Sci. U.S.A. 110, 8399–8404. doi: 10.1073/pnas. Moriyón, I. (2003). “Against gram-negative bacteria: the lipopolysaccharide case,” 1208059110 in Intracellular Pathogens in Membrane Interactions and Vacuole Biogenesis, ed. Karbarz, M. J., Kalb, S. R., Cotter, R. J., and Raetz, C. R. H. (2003). Expression J. P. Gorvel (Dordrecht: Kluwer Academic Publishers), 204–230. cloning and biochemical characterization of a Rhizobium leguminosarum Naessan, C., Egge-Jacobsen, W., Heiniger, R., Wolfgang, M., Aas, F., Røhr, A., et al. lipid A 1-phosphatase. J. Biol. Chem. 278, 39269–39279. doi: 10.1074/jbc. (2008). Genetic and functional analyses of PptA, a phospho-form transferase M305830200 targeting type IV pili in Neisseria gonorrhoeae. J. Bacteriol. 190, 387–400. Karbarz, M. J., Six, D. A., and Raetz, C. R. H. (2009). Purification and doi: 10.1128/JB.00765-07 characterization of the lipid A 1-phosphatase LpxE of Rhizobium Needham, B. D., and Trent, M. S. (2013). Fortifying the barrier: the impact of leguminosarum. J. Biol. Chem. 284, 414–425. doi: 10.1074/jbc.M8083 lipid A remodelling on bacterial pathogenesis. Nat. Rev. Microbiol. 11, 467–481. 90200 doi: 10.1038/nrmicro3047 Kastowsky, M., Gutberlet, T., and Bradaczek, H. (1992). Molecular modelling Palacios-Chaves, L., Conde-Álvarez, R., Gil-Ramírez, Y., Zuñiga-Ripa, A., of the three-dimensional structure and conformational flexibility of bacterial Barquero-Calvo, E., Chacón-Díaz, C., et al. (2011). Brucella abortus ornithine lipopolysaccharide. J. Bacteriol. 174, 4798–4806. doi: 10.1128/jb.174.14.4798- lipids are dispensable outer membrane components devoid of a marked 4806.1992 pathogen-associated molecular pattern. PLOS ONE 6:e16030. doi: 10.1371/ Kovach, M. E., Phillips, R. W., Elzer, P. H., Roop, R. M., and Peterson, K. M. journal.pone.0016030 (1994). pBBR1MCS: a broad-host-range cloning vector. Biotechniques 16, Papadopoulos, A., Gagnaire, A., Degos, C., de Chastellier, C., and Gorvel, J.- 800–802. P. (2016). Brucella discriminates between mouse dendritic cell subsets upon Kubler-Kielb, J., and Vinogradov, E. (2013). The study of the core part and in vitro infection. Virulence 7, 33–44. doi: 10.1080/21505594.2015.1108516 non-repeating elements of the O-antigen of Brucella lipopolysaccharide. Qureshi, N., Takayama, K., Seydel, U., Wang, R., Cotter, R. J., Agrawal, Carbohydr. Res. 366, 33–37. doi: 10.1016/j.carres.2012.11.004 P. K., et al. (1994). Structural analysis of the lipid A derived from Lai, S., Zhou, H., Xiong, W., Gilbert, M., Huang, Z., Yu, J., et al. (2017). Changing the lipopolysaccharide of Brucella abortus. J. Endotoxin Res. 1, 137–148. epidemiology of human brucellosis, China, 1955-2014. Emerg. Infect. Dis. 23, doi: 10.1177/096805199400100303 184–194. doi: 10.3201/eid2302.151710 Raetz, C. R. H., Guan, Z., Ingram, B. O., Six, D. A., Song, F., Wang, X., et al. Lapaque, N., Moriyon, I., Moreno, E., and Gorvel, J.-P. (2005). Brucella (2009). Discovery of new biosynthetic pathways: the lipid A story. J. Lipid Res. lipopolysaccharide acts as a virulence factor. Curr. Opin. Microbiol. 8, 60–66. 50(Suppl.), S103–S108. doi: 10.1194/jlr.R800060-JLR200 doi: 10.1016/j.mib.2004.12.003 Renzi, F., Zähringer, U., Chandler, C. E., Ernst, R. K., Cornelis, G. R., and Ittig, Lapaque, N., Takeuchi, O., Corrales, F., Akira, S., Moriyon, I., Howard, J. C., et al. S. J. (2015). Modification of the 1-phosphate group during biosynthesis of (2006). Differential inductions of TNF-alpha and IGTP, IIGP by structurally Capnocytophaga canimorsus lipid A. Infect. Immun. 84, 550–561. doi: 10.1128/ diverse classic and non-classic lipopolysaccharides. Cell. Microbiol. 8, 401–413. IAI.01006-15 doi: 10.1111/j.1462-5822.2005.00629.x Salcedo, S. P., Marchesini, M. I., Degos, C., Terwagne, M., Von Bargen, K., Leong, D., Díaz, R., Milner, K., Rudbach, J., and Wilson, J. B. (1970). Some Lepidi, H., et al. (2013). BtpB, a novel Brucella TIR-containing effector structural and biological properties of Brucella endotoxin. Infect. Immun 1, protein with immune modulatory functions. Front. Cell. Infect. Microbiol. 3:28. 174–182. doi: 10.3389/fcimb.2013.00028 Lewis, L. A., Shafer, W. M., Dutta Ray, T., Ram, S., and Rice, P. A. (2013). Salcedo, S. P., Marchesini, M. I., Lelouard, H., Fugier, E., Jolly, G., Balor, S., Phosphoethanolamine residues on the lipid A moiety of Neisseria gonorrhoeae et al. (2008). Brucella control of dendritic cell maturation is dependent on the lipooligosaccharide modulate binding of complement inhibitors and resistance TIR-containing protein Btp1. PLOS Pathog. 4:e21. doi: 10.1371/journal.ppat. to complement killing. Infect. Immun. 81, 33–42. doi: 10.1128/IAI.00751-12 0040021 Llobet, E., Martínez-Moliner, V., Moranta, D., Dahlström, K. M., Regueiro, V., Sciorra, V. A., and Morris, A. J. (2002). Roles for lipid phosphate phosphatases Tomás, A., et al. (2015). Deciphering tissue-induced Klebsiella pneumoniae lipid in regulation of cellular signaling. Biochim. Biophys. Acta 1582, 45–51. A structure. Proc. Natl. Acad. Sci. U.S.A. 112, E6369–E6378. doi: 10.1073/pnas. doi: 10.1016/S1388-1981(02)00136-1 1508820112 Scupham, A. J., and Triplett, E. W. (1997). Isolation and characterization of the Mackinnon, F. G., Cox, A. D., Plested, J. S., Tang, C. M., Makepeace, K., Coull, UDP-glucose 4‘-epimerase-encoding gene, galE, from Brucella abortus 2308. P. A., et al. (2002). Identification of a gene (lpt-3) required for the addition Gene 202, 53–59. doi: 10.1016/S0378-1119(97)00453-8 of phosphoethanolamine to the lipopolysaccharide inner core of Neisseria Simon, L. D., Randolph, B., Irwin, N., and Binkowski, G. (1983). Stabilization of meningitidis and its role in mediating susceptibility to bactericidal killing and proteins by a bacteriophage T4 gene cloned in Escherichia coli. Proc. Natl. Acad. opsonophagocytosis. Mol. Microbiol. 43, 931–943. doi: 10.1046/j.1365-2958. Sci. U.S.A. 80, 2059–2062. doi: 10.1073/pnas.80.7.2059 2002.02754.x Sohlenkamp, C., Galindo-Lagunas, K. A., Guan, Z., Vinuesa, P., Robinson, S., Martínez de Tejada, G., Pizarro-Cerdá, J., Moreno, E., and Moriyón, I. (1995). The Thomas-Oates, J., et al. (2007). The lipid Lysyl-phosphatidylglycerol is present outer membranes of Brucella spp. are resistant to bactericidal cationic peptides. in membranes of Rhizobium tropici CIAT899 and confers increased resistance Infect. Immun. 63, 3054–3061. to polymyxin B under acidic growth conditions. Mol. Plant Microbe Interact. Martirosyan, A., Moreno, E., and Gorvel, J.-P. (2011). An evolutionary strategy 20, 1421–1430. doi: 10.1094/MPMI-20-11-1421 for a stealthy intracellular Brucella pathogen. Immunol. Rev. 240, 211–234. Stukey, J., and Carman, G. M. (2008). Identification of a novel phosphatase doi: 10.1111/j.1600-065X.2010.00982.x sequence motif. Protein Sci. 6, 469–472. doi: 10.1002/pro.55600 Moreira, C. G., Herrera, C. M., Needham, B. D., Parker, C. T., Libby, S. J., Fang, 60226 F. C., et al. (2013). Virulence and stress-related periplasmic protein (VisP) Takahashi, H., Carlson, R. W., Muszynski, A., Choudhury, B., Kim, K. S., in bacterial/host associations. Proc. Natl. Acad. Sci. U.S.A. 110, 1470–1475. Stephens, D. S., et al. (2008). Modification of lipooligosaccharide with doi: 10.1073/pnas.1215416110 phosphoethanolamine by LptA in Neisseria meningitidis enhances Moreno, E., Berman, D. T., and Boettcher, L. A. (1981). Biological activities of meningococcal adhesion to human endothelial and epithelial cells. Infect. Brucella abortus lipopolysaccharides. Infect. Immun. 31, 362–370. Immun. 76, 5777–5789. doi: 10.1128/IAI.00676-08

Frontiers in Microbiology| www.frontiersin.org 13 January 2018| Volume 8| Article 2657 fmicb-08-02657 January 9, 2018 Time: 16:56 # 14

Conde-Álvarez et al. Genes Remodeling Brucella Cell Envelope

Tran, A. X., Whittimore, J. D., Wyrick, P. B., Mcgrath, S. C., Cotter, R. J., and Trent, Whatmore, A. M. (2009). Current understanding of the genetic diversity of M. S. (2006). The lipid A 1-phosphatase of Helicobacter pylori is required for Brucella, an expanding genus of zoonotic pathogens. Infect. Genet. Evol. 9, resistance to the antimicrobial peptide polymyxin. J. Bacteriol. 188, 4531–4541. 1168–1184. doi: 10.1016/j.meegid.2009.07.001 doi: 10.1128/JB.00146-06 Wright, J. C., Hood, D. W., Randle, G. A., Makepeace, K., Cox, A. D., Li, J., et al. Trombley, M. P., Post, D. M. B., Rinker, S. D., Reinders, L. M., Fortney, (2004). lpt6, a gene required for addition of phosphoethanolamine to inner- K. R., Zwickl, B. W., et al. (2015). Phosphoethanolamine transferase LptA core lipopolysaccharide of Neisseria meningitidis and Haemophilus influenzae. in Haemophilus ducreyi modifies lipid A and contributes to human defensin J. Bacteriol. 186, 6970–6982. doi: 10.1128/JB.186.20.6970-6982.2004 resistance in vitro. PLOS ONE 10:e0124373. doi: 10.1371/journal.pone.0124373 Zheludkov, M. M., and Tsirelson, L. E. (2010). Reservoirs of Brucella infection in Velasco, J., Bengoechea, J. A., Brandenburg, K., Lindner, B., Seydel, U., nature. Biol. Bull. 37, 709–715. doi: 10.1134/S106235901007006X González, D., et al. (2000). Brucella abortus and its closest phylogenetic relative, Zinsstag, J., Schelling, E., Solera, J., Blasco, J. M., and Moriyón, I. (2011). Ochrobactrum spp., differ in outer membrane permeability and cationic peptide “Brucellosis,” in Handbook of Zoonoses, eds S. R. Palmer, L. Soulsby, P. R. resistance. Infect. Immun. 68, 3210–3218. doi: 10.1128/IAI.68.6.3210-3218. Torgeson, and D. G. Brown (Oxford: Oxford University Press), 54–62. 2000 Vences-Guzmán, M. Á., Guan, Z., Ormeño-Orrillo, E., González-Silva, N., López- Conflict of Interest Statement: The authors declare that the research was Lara, I. M., Martínez-Romero, E., et al. (2011). Hydroxylated ornithine lipids conducted in the absence of any commercial or financial relationships that could increase stress tolerance in Rhizobium tropici CIAT899. Mol. Microbiol. 79, be construed as a potential conflict of interest. 1496–1514. doi: 10.1111/j.1365-2958.2011.07535.x Wang, X., McGrath, S. C., Cotter, R. J., and Raetz, C. R. H. (2006). Expression Copyright © 2018 Conde-Álvarez, Palacios-Chaves, Gil-Ramírez, Salvador-Bescós, cloning and periplasmic orientation of the Francisella novicida lipid A 4’- Bárcena-Varela, Aragón-Aranda, Martínez-Gómez, Zúñiga-Ripa, de Miguel, phosphatase LpxF. J. Biol. Chem. 281, 9321–9330. doi: 10.1074/jbc.M600435200 Bartholomew, Hanniffy, Grilló, Vences-Guzmán, Bengoechea, Arce-Gorvel, Gorvel, Wang, X., Ribeiro, A. A., Guan, Z., Abraham, S. N., and Raetz, C. R. H. (2007). Moriyón and Iriarte. This is an open-access article distributed under the terms Attenuated virulence of a Francisella mutant lacking the lipid A 4’-phosphatase. of the Creative Commons Attribution License (CC BY). The use, distribution or Proc. Natl. Acad. Sci. U.S.A. 104, 4136–4141. doi: 10.1073/pnas.0611606104 reproduction in other forums is permitted, provided the original author(s) or licensor Weissenmayer, B., Gao, J. L., Lopez-Lara, I. M., and Geiger, O. (2002). are credited and that the original publication in this journal is cited, in accordance Identification of a gene required for the biosynthesis of ornithine-derived lipids. with accepted academic practice. No use, distribution or reproduction is permitted Mol. Microbiol. 45, 721–733. doi: 10.1046/j.1365-2958.2002.03043.x which does not comply with these terms.

Frontiers in Microbiology| www.frontiersin.org 14 January 2018| Volume 8| Article 2657

Miriam Salvador Bescós

Frontiers: Your manuscript submission - 409514

Frontiers Microbiology Editorial Office 21 de junio de 2018, 19:57 Responder a: Frontiers Microbiology Editorial Office Para: [email protected]

Dear Dr Salvador,

We are pleased to inform you that we have received the manuscript ""WadD, a new Brucella lipopolysaccharide core glycosyltransferase identified by genomic search and phenotypic characterization"" to be considered for publication in Frontiers in Microbiology, section Infectious Diseases.

You can access the review forum and track the progress of your manuscript using the following link: http://www.frontiersin.org/Journal/MySubmission.aspx?stage=100

You will receive a notification as soon as the interactive review forum is activated and you receive access the review reports. You will then be able to interact directly with the reviewers in the interactive review forum and also re-submit a revised manuscript.

Best regards,

Your Frontiers in Microbiology team

Frontiers | Editorial Office - Collaborative Peer Review Team www.frontiersin.org Avenue du Tribunal Fédéral 34, 1005 Lausanne, Switzerland Office T 41 21 510 17 25

For technical issues, please contact our IT Helpdesk ([email protected]) or visit our Frontiers Help Center (zendesk.frontiersin.org/hc/en-us)

------MANUSCRIPT DETAILS------Manuscript title: WadD, a new Brucella lipopolysaccharide core glycosyltransferase identified by genomic search and phenotypic characterization Manuscript ID: 409514 Submitted By: Maite Iriarte Authors: Maite Iriarte, Raquel Conde-Alvarez, Miriam Salvador, Yolanda Gil-Ramirez, Amaia Zúñiga-Ripa, Estrella Martínez-Gómez, María Jesús De Miguel, Pilar Maria Muñoz, Axel Cloeckaert, Michel Stanislas Zygmunt, Ignacio Moriyón Journal: Frontiers in Microbiology, section Infectious Diseases

Research Topic: Pathogenomics of the genus Brucella and beyond Article type: Original Research Submitted on: 21 Jun 2018

------ADDITIONAL INFORMATION------In order to enable a smooth and efficient review process, please familiarize yourself with the Frontiers review guidelines: http://www.frontiersin.org/Journal/ReviewGuidelines.aspx?s=1404&name=infectious_diseases

To take part in the Resource Identification Initiative please cite antibodies, genetically modified organisms, software tools, data, databases and services using the corresponding catalog number and RRID in the text of your article. Please see here for more information: http://www.frontiersin.org/files/pdf/letter_to_author.pdf 1 WadD, a new Brucella lipopolysaccharide core 2 glycosyltransferase identified by genomic search and 3 phenotypic characterization.

4 1 1∫ 1 5 Miriam Salvador-Bescós , Yolanda Gil-Ramírez , Amaia Zúñiga-Ripa , Estrella 6 Martínez-Gómez1#, María J. de Miguel2, Pilar M. Muñoz2, Axel Cloeckaert3, Michel S. 7 Zygmunt3, Ignacio Moriyón1, Maite Iriarte1*, Raquel Conde-Álvarez1*. 8 9 1Instituto de Salud Tropical, Instituto de Investigación Sanitaria de Navarra, and 10 Departamento de Microbiología y Parasitología, Universidad de Navarra, c/Irunlarrea 1, 11 31008 Pamplona, Spain. 12 2 Instituto Agroalimentario de Aragón (IA2), Unidad de Producción y Sanidad 13 Animal del Centro de Investigación y Tecnología Agroalimentaria de Aragón (CITA)- 14 Universidad de Zaragoza, Zaragoza, Spain 15 3 Institut National de la Recherche Agronomique, Université François Rabelais de 16 Tours, UMR 1282, Nouzilly, France. 17 ∫ Current address: Centro Nacional de Tecnología y Seguridad Alimentaria (CNTA). 18 Crta-Na134-km 53. San Adrian. 31570. Navarra 19 # Current address: Unidad de Gestión Clínica de Aparato Digestivo, Servicio de 20 Farmacología Clínica, Instituto de Investigación Biomédica de Málaga-IBIMA, 21 Hospital Universitario Virgen de la Victoria, Universidad de Málaga, CIBERehd, 22 Málaga,In Spain. review 23 24 25 *To whom correspondence should be addressed: Raquel Conde-Álvarez, and Maite 26 Iriarte. Departamento de Microbiología y Parasitología, Universidad de Navarra, 27 c/Irunlarrea 1, 31008 Pamplona, Spain. Telephone: +34 948425600; E-mail address: 28 [email protected] ; [email protected] 29 30 31 32 33 Running title: Brucella WadD and lipopolysaccharide synthesis 34 35 Key words Lipopolysaccharide (LPS), bacterial pathogenesis, vaccine 36 development, virulence factor, glycosyltransferase, brucellosis, Brucella.

37

1 38 Abstract 39 Brucellosis, an infectious disease caused by Brucella, is one of the most extended 40 bacterial zoonosis in the world and an important cause of economic losses and human 41 suffering. The lipopolysaccharide (LPS) of Brucella plays a major role in virulence as it 42 impairs normal recognition by the innate immune system and delays the immune 43 response. The LPS core is a branched structure involved in resistance to complement 44 and polycationic peptides, and mutants in glycosyltransferases required for the synthesis 45 of the lateral branch not linked to the O-polysaccharide (O-PS) are attenuated and have 46 been proposed as vaccine candidates. For this reason, the complete understanding of the 47 genes involved in the synthesis of this LPS section is of particular interest. The 48 chemical structure of the Brucella LPS core suggests that, in addition to the already 49 identified WadB and WadC glycosyltransferases, others could be implicated in the 50 synthesis of this lateral branch. To clarify this point, we identified and constructed 51 mutants in 11 ORFs encoding putative glycosyltransferases in B. abortus.

52 Four of these ORFs, regulated by the virulence regulator MucR (involved in LPS 53 synthesis) or the BvrR/BvrS system (implicated in the synthesis of surface 54 components), were not required for the synthesis of a complete LPS neither for 55 virulence or interaction with polycationic peptides and/or complement. Among the other 56 seven ORFIns, six seemed notreview to be required for the synthesis of the core LPS since the 57 corresponding mutants kept the O-PS and reacted as the wild type with polyclonal sera. 58 Interestingly, mutant in ORF BAB1_0953 (renamed wadD) lost reactivity against 59 antibodies that recognize the core section while kept the O-PS. This suggests that WadD 60 is a new glycosyltransferase adding one or more sugars to the core lateral branch. WadD 61 mutants were more sensitive than the parental strain to components of the innate 62 immune system and played a role in chronic stages of infection. These results 63 corroborate and extend previous work indicating that the Brucella LPS core is a 64 branched structure that constitutes a steric impairment preventing the elements of the 65 innate immune system to fight against Brucella.

66

2 67 INTRODUCTION 68 69 Members of the genus Brucella are the etiologic agents of brucellosis, a worldwide 70 spread zoonosis that affects ruminants, camelids, swine, dogs and several forms of 71 marine and terrestrial wildlife and causes abortions, infertility and the subsequent 72 economic losses in livestock. Humans become infected via direct contact with affected 73 animals and through consumption of unpasteurized dairy products, and develop a 74 chronic and debilitating condition that requires prolonged antibiotic treatment, being 75 lethal in 1-5% of untreated cases (Ariza, 1999). Because of its impact on animal 76 production and Public Health, it is estimated that brucellosis imposes a heavy burden in 77 the developing world (McDermott et al., 2013).

78 The genus includes several nominal species that show host preferences 79 (http://www.bacterio.net/-allnamesac.html). Those that have been known for a long time 80 (often referred to as “classical” Brucella species) include B. abortus and B. melitensis 81 (the brucellae that infect domestic ruminants), B. suis (infecting swine, reindeer, hares 82 and several species of wild rodents), B. canis (infecting dogs), B. ovis (not zoonotic and 83 restricted to sheep) and B. neotomae (infecting the desert woodrat). Because of their 84 early identification and their economic and public health importance, B. abortus, B. 85 melitensis and B. suis are the best-characterized members of the genus, and all of them 86 produceIn smooth (S) glossy reviewcolonies, a morphology that reflects the existence of a 87 lipopolysaccharide (LPS) carrying an O-polysaccharide (O-PS) linked to the core-lipid 88 A section that anchors the molecule to the outer membrane (OM). These Brucella spp. 89 behave as facultative intracellular parasites of professional and non-professional 90 phagocytes, an ability that depends on a number of virulence factors, chiefly a type IV 91 secretion system and a peculiar OM structure. Critical OM components such as the S- 92 LPS, lipoproteins and ornithine lipids differ in relevant molecular details from the 93 homologous molecules that in other bacteria bear the Pathogen-Associated Molecular 94 Patterns (PAMP) readily detected by innate immunity pattern recognition receptors 95 (PRR). Consequently, these brucellae induce comparatively low and delayed 96 proinflammatory responses, which creates a time window allowing the pathogen to 97 traffic intracellularly in dendritic cells and macrophages to reach a safe niche before 98 effective phagocyte activation takes place (Lapaque et al., 2005; Barquero-Calvo et al., 99 2007; Palacios-Chaves et al., 2011). In this regard, the Brucella S-LPS carries the most 100 significant PAMP modifications and is thus a major virulence factor (Lapaque et al.,

3 101 2005). Whereas the structure of the O-PS (a N-formylperosamine homopolymer) and its 102 role in virulence in animal models and in the natural host have been known for a long 103 time, the importance of the core and its structure have only recently been established.

104 The core of B. melitensis LPS (Figure 1) is a branched oligosaccharide built of lipid 105 A-linked 3-deoxy-D-manno-2-octulosonic acid (Kdo), glucose, 2-amino-2,6-dideoxy- 106 D-glucose (quinovosamine), mannose, and 2-amino-2-deoxy-D-glucose (glucosamine 107 [GlcN]) (Conde-Álvarez et al., 2012; Fontana et al., 2016; Gil-Ramírez et al., 2014; 108 Iriarte et al., 2004; Kubler-Kielb and Vinogradov, 2013). This structure accounts for the 109 Brucella LPS core overlapping epitopes (Rojas et al., 1994) an inner one comprising the 110 Kdo residues plus the glucose bridging KdoII with the O-PS and an outer epitope 111 encompassing the mannose and GlcN residues (Fontana et al., 2016; González et al., 112 2008; Iriarte et al., 2004). This last epitope plays a critical role in the binding of 113 monoclonal antibodies such as A68/24G12/A08 and A08/24D08/G09 (Conde-Álvarez 114 et al., 2012; Fontana et al., 2016).

115 Accordingly, the reactivity with R-LPS-specific monoclonal antibodies (MoAbs) 116 strongly suggests that the structure elucidated for B. melitensis is conserved in the 117 classical species (Bowden et al., 1995; Zygmunt et al., 2012). Moreover, availability of 118 the corresponding structure of several mutants has also allowed assigning genes that 119 upon mutationIn generate LPSsreview that lack (i.e. rough [R] LPS) or carry O-PS (Figure 1). 120 Gene wadA corresponds to the enzyme linking KdoII and glucose, wadC to the 121 mannosyltransferase acting on KdoI, and wadB to a glucosaminyltransferase involved in 122 the assembly of the GlcN branch (Conde-Álvarez et al., 2012; Gil-Ramírez et al., 2014; 123 González et al., 2008). These genes are highly conserved in the classical Brucella 124 species (Conde-Álvarez et al., 2012; Gil-Ramírez et al., 2014; González et al., 2008; 125 Iriarte et al., 2004; Monreal et al., 2003; Soler-Llorens et al., 2016) and as expected, all 126 Brucella genomes also carry a waaA homologue, the essential gene coding for the Kdo 127 transferase of Gram-negative bacteria (Iriarte et al., 2004; Raetz and Whitfield, 2002). 128 However, since most but not all glycosyltransferases involved in LPS synthesis are 129 monofunctional (Raetz and Whitfield, 2002), it remains to be determined whether 130 glucosaminyltransferases other than WadB are required for the synthesis of the GlcN 131 branch.

132 Based on the complete structure of the core and the phenotype of mutants in wadB 133 and wadC, it is postulated that the lack of acidic groups other than the two Kdo and

4 134 lipid A phosphates and the mannose-GlcN branch account for the role of Brucella core 135 in virulence both in cellular and animal models. By virtue of the density of amino 136 groups and close position to the inner core and lipid A, the GlcN tetrasaccharide both 137 neutralizes and sterically protects those inner anionic groups, thereby hampering 138 binding of bactericidal peptides and PRRs such as the activators of the antibody- 139 independent classical complement pathway and MD2, the TLR4 co-receptor. 140 Accordingly, core defects bolster proinflammatory responses causing an activation of 141 innate immunity earlier than that of the wild-type, thereby generating attenuation 142 (Conde-Álvarez et al., 2012; Fontana et al., 2016; Gil-Ramírez et al., 2014; Soler- 143 Lloréns et al., 2014). Also, although both wadB and wadC mutants maintain an intact 144 O-PS, attenuation in mice is more severe for the latter (Conde-Álvarez et al., 2012; 145 Fontana et al., 2016; Gil-Ramírez et al., 2014) strongly suggesting a correlation between 146 the extent of core damage and the intensity of the immunoactivation that brings about 147 attenuation. A complete elucidation of the genetics of Brucella LPS core could confirm 148 such a correlation and, since LPS core mutants represent a tool for developing a new 149 generation of brucellosis vaccines (Conde-Álvarez et al., 2013; Zhao et al., 2017), also 150 provide a graded array of possibilities. With these possibilities in mind, we investigated 151 B. abortus 2308 genes annotated as glycosyltransferases for their possible involvement 152 in LPS Incore synthesis and relevantreview biological effects. 153 MATERIALS AND METHODS

154 Bacterial strains and growth conditions

155 The bacterial strains and plasmids used in this study are listed in Table S1. All 156 bacteria were grown either on tryptic soy agar (TSA, Pronadisa) plates or in tryptic soy 157 (TSB, Scharlau) or Mueller-Hinton (Becton Dickinson, Difco) broths at 37ºC. Where 158 indicated, growth media were supplemented with kanamycin (Km) at 50 mg/ml, 159 nalidixic acid (Nal) at 25 mg/ml, ampicillin (Amp) at 100 mg/ml and/or 5% sucrose. 160 Bacterial growth rates were determined at 37ºC in Mueller-Hinton broth (Becton 161 Dickinson, Difco), using a Bioscreen C apparatus (Lab Systems). All strains were stored 162 in skim milk at –80ºC. Work with Brucella was performed at the Biosafety Level 3 163 (BSL-3) laboratory facilities of the “Centro de Investigación Médica Aplicada de la 164 Universidad de Navarra” (CIMA), and “Centro de Investigación y Tecnología 165 Agroalimentaria de Aragón” (CITA), Spain.

5 166 DNA manipulations and analyses.

167 Sequence data were obtained from Kyoto Encyclopedia of Genes and Genomes 168 (KEGG1). Searches for DNA and protein homologies between Brucella species and 169 other α-proteobacteria such as Ochrobactrum, Rhizobium or Agrobacterium were 170 carried out using KEGG, Basic Local Alignment Sequence Tool (BLAST2) and Clustal 171 Omega3 from the European Molecular Biology Laboratory-European Bioinformatics 172 Institute (EMBL-EBI4). New glycosyltransferase identification, using B. abortus 2308 173 was supported by Carbohydrate-Active enZymes database (CAZy5). Primers were 174 designed using Primer 3 input6 and synthesized by Sigma-Aldrich. Plasmid DNA was 175 extracted with Qiaprep spin Miniprep (Qiagen GmbH). When needed, DNA was 176 purified from agarose gels using Qiack Gel extraction kit (Qiagen) and sequenced by the 177 Servicio de Secuenciación of CIMA.

178 1 http://www.genome.jp/kegg/

179 2 http://blast.ncbi.nlm.nih.gov/Blast.cgi

180 3 http://www.ebi.ac.uk/Tools/msa/clustalo

181 4 http://www.ebi.ac.uk/

182 5 http://www.cazy.org 6 In review 183 http://bioinfo.ut.ee/primer3-0.4.0/

184

185 Construction of mutants.

186 Open Reading Frames (ORFs) BAB2_0133, BAB2_0135, BAB2_0105 and 187 BAB1_1620 were mutagenized by in frame non-polar deletion in B. abortus 2308W 188 (Table S1).

189 For the construction of BaΔBAB2_0133 mutant, we first generated two PCR 190 fragments: oligonucleotides BAB2_0133-F1 (5'-GCGTTGGACAAGTTGAGGTT-3') 191 and BAB2_0133-R2 (5´-CATAGCGGTCGGTTAAATGC- 3´) were used to amplify a 192 572 base pairs (bp) fragment including codons 1 to 38 of BAB2_0133, as well as 458 bp 193 upstream of the BAB2_0133 start codon. Oligonucleotides BAB2_0133-F3 (5´- 194 GTATCGCCAGCCAATTTACGTCCGTATTGGAAGCCAAGAA-3´) and 195 BAB2_0133-R4 (5´-CAGTAACAAAAGGCCGCTAT-3´) were used to amplify a 442

6 196 bp fragment including codons 299 to 326 of BAB2_0133 and 355 bp downstream of the 197 BAB2_0133 stop codon. Both fragments were ligated by overlapping PCR using 198 oligonucleotides F1 and R4 for amplification, and the complementary regions between 199 R2 and F3 for overlapping. The resulting fragment, containing the BAB2_0133 deletion 200 allele, was cloned into pCR2.1 (Invitrogen), to generate plasmid pMSB-01, sequenced 201 to ensure the maintenance of the reading frame, subsequently subcloned into the BamHI 202 and the XbaI sites of the suicide plasmid pJQK (Scupham and Triplett, 1997) and 203 transformed into competent E. coli S17 λpir (Simon et al., 1983). The resulting suicide 204 pJQK-derived plasmid was introduced into B. abortus 2308 by conjugation. The first 205 recombination event (integration of the suicide vector in the chromosome) was selected 206 by Nal and Km resistance, and the second recombination (excision of the mutator 207 plasmid leading to construction of the mutant by allelic exchange), was selected by Nal 208 and sucrose resistance and Km sensitivity. The resulting colonies were screened by PCR 209 with primers F1 and R4 which amplified a fragment of 1014 bp in the mutant and 1794 210 bp in the sibling strain that keeps the wild type gene. Primers BAB2_0133-F1 and 211 BAB2_0133-R5 (5´-AAGACCCAGTAGTTAGCACT- 3´) amplified a fragment of 919 212 bp only in the wild type strain. The mutation generated results in the loss of the 80% of 213 the ORF. 214 BaΔInBAB2_0135 mutant review was constructed following the same procedure and using 215 oligonucleotides BAB2_0135-F1 (5´- 216 TGGCGGCCGCTCTAGAACACCGGACTGCCTGATAA -3´) and BAB2_0135-R2 217 (5´- CGGGCAATTTCGGCATAG -3´) that amplified a 240 bp fragment including 218 codons 1 to 40 of BAB2_0135, as well as 120 bp upstream of the BAB2_0135 start 219 codon, and oligonucleotides BAB2_0135-F3 (5´- 220 CTATGCCGAAATTGCCCGCCGGTTTGGAAATGCGGTCAA -3´) and 221 BAB2_0135-R4 (5´- ATCCACTAGTTCTAGTTATGTAGCCGCCACCGTTT -3´) that 222 amplified a 232 bp fragment including codons 441 to 478 of BAB2_0135 and 115 bp 223 downstream of the BAB2_0135 stop codon. The resulting colonies were screened by 224 PCR with primers F1 and R4 that amplified a fragment of 472 bp in the mutant and 225 1672 bp in the sibling strain that keeps the wild type gene. Primers BAB2_0135-F1 and 226 BAB2_0135-R5 (5´- CGATTGCCAGTCCCAGAAAG -3´) amplified a fragment of 227 628 bp only in the wild type strain. The mutation generated results in the loss of the 228 84% of the ORF.

7 229 For the construction of BaΔBAB2_0105 mutant, oligonucleotides BAB2_0105-F1 230 (5´- GCGTGTTCTACAGCCATGAA -3´) and BAB2_0105-R2 (5´- 231 CCGCCGAAATGTAGGAAGTG -3´) amplified a 198 bp fragment including codons 1 232 to 33 of BAB2_0105, as well as 99 bp upstream of the BAB2_0105 start codon. 233 Oligonucleotides BAB2_0105-F3 (5´- 234 CACTTCCTACATTTCGGCGGTATGTTGGATTGGGACGGGT -3´) and

235 BAB2_0105-R4 (5´- GCCGAATATGACGCTTGCTA -3´) amplified a 154 bp 236 fragment including codons 307 to 330 of BAB2_0105 and 79 bp downstream of the 237 BAB2_0105 stop codon. The resulting colonies were screened by PCR with primers F1 238 and R4 which amplified a fragment of 352 bp in the mutant and 1171 bp in the sibling 239 strain which keeps the wild type gene. Primers BAB2_0105-F1 and BAB2_0105-R5

240 (5´- CAAAGACCGGATATTGCGGG -3´) amplified a fragment of 550 bp only in the 241 wild type strain. The mutation results in the loss of the 83% of the ORF.

242 BaΔBAB1_1620 mutant was constructed using oligonucleotides 1620-F1 (5´- 243 GTACGCGGTCGTAGCTCAGT-3´) and 1620-R2 (5´- 244 CTCAAACTGAGACGCCATGA-3´), that amplified a 475 bp fragment including 245 codon 1 to 23 of BAB1_1620 as well as 406 bp upstream of the ORF start codon. 246 Oligonucleotides 1620-F3 (5´- 247 TCATGGCGTCTCAGTTTGAGATAGCCAACGTCACCAAAACAIn review-3´) and 1620-R4 248 (5´-CTCTGCAATTCTTGCGATCA-3´) were used to amplify a 410 bp fragment 249 including codons 241 to 261 of the BAB1_1620 ORF and 347 bp downstream of the 250 BAB1_1620 stop codon. Both fragments were ligated, cloned into pCR2.1 to generate 251 plasmid pYRI-16, and subcloned into the suicide pJQK (pYRI-17). After conjugation 252 with B. abortus, the resulting colonies were screened by PCR with primers 1620-F1 and 253 1620-R4 which amplified a 885 bp fragment in the mutant and 1536 bp in the parental 254 strain. The mutation generated results in the loss of the 83% of the BAB1_1620 ORF.

255 The rest of the ORFs were mutagenized by recombination and gene disruption using 256 as suicide vectors pJQK or pSKoriT (Tibor et al., 2002) carrying an internal fragment of 257 the ORF.

258 For the construction of Ba::pJQK-BAB1_0114 mutants, we generated a PCR 259 fragment using oligonucleotides BAB1_0114-F1 (5'- 260 TCAACAAATCGGCCAAGGAC -3') and BAB1_0114-R2 (5´- 261 GTCACGCGGTCAAACTGG - 3´) which amplified a 481 bp fragment containing the

8 262 region that codes for aminoacids 248 to 407. The fragment was cloned into pCR2.1, to 263 generate plasmid pMSB-17, sequenced and subcloned into the BamHI and the XbaI 264 sites of the suicide plasmid pJQK to obtain pMSB-28, and then transformed into 265 competent E. coli S17 and transferred into B. abortus 2308 by conjugation. The 266 integration of the suicide vector and disruption of the target gene was selected by Nal

267 and Km resistance and by PCR combining BAB1_0114-F3 (5´- 268 CCTATATTCCCCAGGCCGTT - 3´) with M13 Forward (5'- 269 CTGGCCGTCGTTTTAC -3') or with M13 Reverse (5'- CAGGAAACAGCTATGAC- 270 3'). These last two primers hybridize in the suicide vector inserted in the chromosome. 271 BAB1_0114-F3 and M13 Forward amplified a fragment of 881 bp only in the mutant 272 strain. Following the same strategy, we constructed the rest of insertion mutants:

273 Mutant Ba::pSKoriT-BAB1_0417 was obtained using oligonucleotides 274 BAB1_0417-F1 (5'- TGATCGACCATGGCTCGG-3') and BAB1_0417-R2 (5´- 275 TCAAGCCTGACCAGAAGCC - 3´) which amplified a 295 bp fragment of 276 BAB1_0417 (codon 37 to 134). The fragment was first cloned in pCR2.1 (pMSB-05), 277 subcloned into the suicide plasmid pSKoriT (pMSB-06), and transferred into B. abortus

278 2308 by conjugation. Primers M13 Reverse and BAB1_0417-F3 (5´- 279 CTGTTTCCCGACCAGCTTG - 3´) amplified a fragment of 649 bp only in the mutant. 280 OligonucleotidesIn BAB2_0693 review-F1 (5'- CACTGCAAGCCGGTTACAAT -3') and 281 BAB2_0693-R2 (5´- TGCAACGAAATTCTGTCCGG - 3´) were used for the 282 construction of Ba::pJQK-BAB2_0693 mutant. F1 and R2 amplified a fragment of 416 283 bp (codons 249 to 386). We generated plasmid pMSB-16, subsequently subcloned into 284 the suicide plasmid pJQK (pMSB-24), and conjugated into B. abortus 2308. Primers

285 M13 Forward and BAB2_0693-F3 (5´- ACGAGCGCTATGATTTCGTC - 3´) 286 amplified a fragment of 684 bp only in the mutant.

287 For the construction of Ba::pJQK-BAB1_0932 mutants we used oligonucleotides 288 BAB1_0932-F1 (5'- GCCGTCGTCCTGAATGTTAC -3') and BAB1_0932-R2 (5´- 289 GCCATTATCCAGTGCAGCC - 3´) which amplified a 420 bp fragment of 290 BAB1_0932 (codons 354 to 493). We generated plasmid pMSB-28, subsequently 291 subcloned into the suicide plasmid pJQK (pMSB-29), and conjugated into B. abortus 292 2308. The resulting Nal-Km resistant colonies were screened by PCR. Primers M13 293 Reverse and BAB1_0932-F3 (5´- GGCCGAGAATGGCTATATCA - 3´) amplified a 294 fragment of 915 bp only in the mutant.

9 295 Mutant Ba::pSKoriT-BAB1_0326 was obtained using oligonucleotides 296 BAB1_0326-F1 (5'- GCACTCAACCGGCTCAATTG -3') and BAB1_0326-R2 (5´- 297 AGCACCGCATATTCAAAGGC - 3´) which amplified a 368 bp fragment of 298 BAB1_0326 (codons 261 to 383) that was cloned into pCR2.1 to obtain pMSB-07. 299 The fragment was then subcloned into the suicide pSKoriT (pMSB-10), and conjugated 300 into B. abortus 2308. The resulting Nal-Km resistant colonies were screened by PCR. 301 Primers M13 Reverse and BAB1_0326-F3 (5´- ATGTTGCCATGTCGCTGTTT - 3´) 302 amplified a fragment of 678 bp only in the mutant strain.

303 Construction of Ba::pJQK-BAB1_0607 mutants was constructed using 304 oligonucleotides BAB1_0607-F1 (5'- GCCAATGTCGTTCTCTCCAA -3') and 305 BAB1_0607-R2 (5´- CTTGGTGTCAGCCCCTTTTC - 3´) which amplified a 449 bp 306 fragment of BAB1_0607 (codons 278 to 427). We generated the pCR2.1-derived 307 plasmid pMSB-19, and then subcloned into the suicide plasmid pJQK (pMSB-21), and 308 conjugated into B. abortus 2308. Primers M13 Forward and BAB1_0607-F3 (5´- 309 TTCTTTCCAATGAGCGCACC - 3´) amplified a fragment of 800 bp only in the 310 mutant.

311 We constructed two different mutants in ORF BAB1_0953 (wadD). The first, 312 Ba::pSKoriT-BAB1_0953, carried the suicide vector inserted in the gene and was 313 obtainedIn with oligonucleotides review BAB1_0953-F1 (5'- ACTTTTCGCCGAGCAACAAA - 314 3') and BAB1_0953-R2 (5´- AGGCACGGTTTCATAGACGA - 3´) which amplified a 315 358 bp fragment of BAB1_0953 (codon 112 to 230). We generated plasmid pMSB-11, 316 subsequently subcloned into the suicide plasmid pSKoriT (pMSB-12), and conjugated

317 into B. abortus 2308. Primers M13 Forward and BAB1_0953-F3 (5´- 318 GCTGGCTTCATGAAATCCGT - 3´) amplified a fragment of 612 bp in mutant.

319 We also constructed a non-polar wadD mutant (BaΔwadD) by in frame deletion. 320 Oligonucleotides wadD-F1 (5'- TCTATAATGAGAGGCGGCTTTT -3') and wadD-R2 321 (5´- AGAAGTGCTGGTCCTGTTGT - 3´) were used to amplify a 304 (bp) fragment 322 including codons 1 to 50 of BAB1_0953, as well as 154 bp upstream of the 323 BAB1_0953 start codon. Oligonucleotides wadD-F3 (5´- 324 ACAACAGGACCAGCACTTCTATCCTCACCCTGCCATTCAA -3´) and wadD-R4 325 (5´- CTGGTACTAGACGCCCTGTT -3´) were used to amplify a 175 bp fragment 326 including codons 281 to 324 of BAB1_0953 and 43 bp downstream of the BAB1_0953 327 stop codon. Both fragments were ligated by overlapping PCR using oligonucleotides F1

10 328 and R4 for amplification, and the complementary regions between R2 and F3 for 329 overlapping. The resulting fragment, containing the BAB1_0953 deletion allele, was 330 cloned directly into pJQK by the InFusion technique to generate pMSB-34. This suicide 331 vector was sequenced to ensure the maintenance of the reading frame and transferred 332 into B. abortus 2308 by conjugation. The resulting colonies were screened by PCR with 333 primers F1 and R4 that amplified a fragment of 479 bp in the mutant and 1169 bp in the 334 sibling strain which keeps the wild type gene. Primers wadD-F1 and wadD-R5 (5´- 335 AGGCACGGTTTCATAGACGA -3´) amplified a fragment of 844 bp only in the wild 336 type strain. The mutation generated results in the loss of the 71% of the ORF and the 337 mutant was called BaΔwadD.

338 Complementation of wadD mutants.

339 For complementation experiments, we performed a stable insertion of the miniTn7 340 transposon into the chromosome of BaΔwadD (Choi and Schweizer, 2006). Por this 341 purpose, we first generated a PCR product using oligonucleotides Tn7-wadD-F1 (5`- 342 CGGGCTGCAGGAATTGCGATTCCTTTGTGCCAGAT-3`) and Tn7-wadD-R2 (5`- 343 GCTTCTCGAGGAATTATCATCGCCGCATTGAAGAC-3`), which amplified a 1771 344 bp fragment including codons 1 to 323 of BAB1_0953 together with 481 bp upstream of 345 the ORF start codon including the putative wadD promoter and 318 bp downstream the 346 ORF stopIn codon. This PCRreview product was cloned into the corresponding sites of the 347 linearized pUC18 R6KT miniTn7T KmR vector (Llobet et al., 2009) to generate plasmid 348 pMSB-44. The plasmid was sequenced to ensure the maintenance of the reading frame 349 transformed into E. coli S17 and transferred to BaΔwadD mutant by tetra-parental 350 conjugation between E. coli S17 (pMSB44), E. coli SM10 λpir (pTNS2) and E. coli 351 HB101 (pRK2013). The conjugants harboring pMSB-44 were selected by plating the 352 mating mixture onto TSA-Nal-Km plates that were incubated at 37ºC for 4 days. To 353 confirm that the transposon was inserted between genes glmS and recG (Choi and 354 Schweizer, 2006), we performed PCR using different oligonucleotides: Tn7F (5`- 355 TGGCTAAAGCAAACTCTTCATTT – 3`) and Tn7R (5`- 356 GCGGATTTGTCCTACTCAGG – 3`) allowed to confirm that the Tn7 was inserted, 357 oligonucleotides Glms_B (5`- GTCCTTATGGGAACGGACGT – 3`) and PTn7-R (5`- 358 CACAGCATAACTGGACTGATT -3`) confirmed that the transposon was inserted 359 immediately after the gene glmS, and RecG-R (5`- TATATTCTGGCGAGCGATCC – 360 3`) and PTn7-L (5`- ATTAGCTTACGACGCTACACCC – 3`) confirmed that the

11 361 transposon was inserted before the gene recG (Choi and Schweizer, 2006). The 362 resulting strain was named BaΔwadD::Tn7-PwadD.

363 Crystal violet exclusion test

364 To study if the mutants had smooth (complete LPS) or rough (O-PS-lacking LPS) 365 phenotype, 5 ml of a crystal violet solution at 0.1 mg/ml in distilled water were used to 366 cover isolated colonies on TSA plates for 20 seconds. Smooth colonies excluded crystal 367 violet and looked white, whereas rough colonies captured crystal violet and looked 368 violet.

369 LPS extraction and characterization

370 LPS was extracted by the proteinase-K sodium dodecyl sulphate (SDS) protocol 371 (Dubray and Limet, 1987; Garin-Bastuji et al., 1990) with some modifications. Bacteria 372 grown overnight in 10 ml of TSB were killed with 0.5% phenol during 3 days in 373 agitation at 37ºC. After that, samples were weighed and pipetted into small 374 polycarbonate cap tubes and then suspended by ultrasounds in 2% SDS-60 mM Tris- 375 HCl buffer (pH 6.8) at a concentration of 0.5g (wet weight) of bacteria per 10 ml of 376 buffer. Samples were then heated at 100ºC for 10 minutes, and lysates were cooled to 377 55ºC. This treatment was followed by digestion with 60µl of proteinase-K at 2.5 mg/ml 378 in HCl-InTris per ml of sample review (Merck KGaA) for 3 hours at 55ºC, and overnight 379 incubation at 20ºC. Afterwards, they were centrifuged at 20,000 x g for 30 minutes at 380 room temperature, and the LPS was precipitated from the supernatant by addition of 3 381 volumes of methanol containing 1% sodium acetate-saturated methanol at -20ºC. After 382 60 minutes, the precipitate was harvested by centrifugation at 5,000 x g for 15 minutes 383 at 4ºC and resuspended by sonication in 10 ml of distilled water. After a second 384 methanol precipitation and centrifugation, the pellets were resuspended by sonication in 385 2-3 ml of 60 mM HCl-Tris (pH 6.8) and left at 37ºC. Then samples were treated with 20 386 µl/ml of RNase and DNase stock solutions at 0.5 mg/ml in HCl-Tris (MP Biomedicals 387 and Sigma-Aldrich, respectively) at 37ºC for 30 minutes. Subsequently, the LPS was 388 treated again with 5 µl/ml of proteinase K at 2.5 mg/ml in HCl-Tris, at 55ºC for 3 hours 389 and then, at room temperature overnight. After a third methanol precipitation in the 390 same conditions described above, the pellet containing LPS was recovered in 1 ml of 391 distilled water and frozen at -20ºC.

392 SDS-PAGE and Western blots.

12 393 Samples were mixed 1:1 with Sample buffer 2X (Bio-Rad), heated at 100ºC for 10 394 minutes and analyzed in Tris-HCl-glycine-12, 15 or 18% polyacrylamide gels (37.5:1 395 acrylamide/methylene-bisacrylamide ratio). Fifteen µl of each sample were run at 30 396 mA constant current for 140 minutes. Finally, LPS molecules were revealed by the 397 periodate-alkaline silver method (Tsai and Frasch, 1982).

398 For Western blot, gels were electro-transferred onto PVDF sheets (Whatman, 399 Schleicher & Schuell, WESTRAN S.; 0.2 µm pore size) in a transfer buffer (pH 8.3) 400 containing 0.025M Tris, 0.192M glycine, and 20% (vol/vol) methanol. Transfer was 401 performed at a constant voltage of 8V and 200mA for 30 minutes in a Trans-Blot Semi- 402 Dry Transfer Cell (Bio-Rad). Antibodies used were MoAbs A68/24D08/G09 and 403 A68/24G12/A08, which recognize core epitopes (Bowden et al., 1995), and a polyclonal 404 serum from a rabbit infected with B. melitensis 16M and bled at day 45.

405 Enzyme Linked Immunosorbent Assay (ELISA).

406 ELISA using whole bacteria (sonicated cells) as the antigen were performed as 407 described previously (Cloeckaert et al., 1993a). MoAbs used were directed against O- 408 PS, R-LPS and the outer membrane lipoproteins Omp10, Omp16, and Omp19 409 (Cloeckaert et al., 1990, 1993a, 1993b, 1998). The anti- R-LPS MoAbs used were 410 A68/03F03/D05 (IgG2b), A68/10A06/B11 (IgM), A68/24D08/G09 (IgG1) and 411 A68/24G12/A08In (IgG3). Thereview MoAbs specific for the O-PS epitopes were 2E11 (IgG3; 412 M epitope), 12G12 (IgG1; C [A=M] epitope), 07F09 (IgG1; C [A=M] epitope), 12B12 413 (IgG3; C [M>A] epitope), 18H08 (IgA; C/Y [A=M] epitope), 04F9 (IgG2a; C/Y [A>M] 414 epitope) and 05D4 (IgG1; C/Y [A>M] epitope). The MoAbs specific for outer 415 membrane lipoproteins were A68/08E07/B11 (Omp10; IgG2a), A68/04G01/C06 416 (Omp16; IgG2a), A76/08C03/G03 (Omp16; IgG2a), and A76/10D03/H02 (Omp19; 417 IgG2b). All MoAbs were used as hybridoma supernatants in ELISA.

418 Sensitivity to polycationic bactericidal peptides.

419 The minimal inhibitory concentration (MIC) of polymyxin B and poly-L-ornithine 420 (both from Sigma-Aldrich), were determined in Mueller-Hinton medium. Exponentially 421 growing bacteria were adjusted to an O.D. equivalent to 1 of the McFarland scale, and 422 exposed to serial dilutions of the bactericidal peptides. MICs were determined by 423 technical duplicates after 2 days of incubation at 37ºC. Experiments were performed in 424 triplicate.

13 425 Sensitivity to the bactericidal action of nonimmune serum.

426 Exponentially growing bacteria were adjusted to 104 Colony Forming Units 427 (CFU)/ml in saline and dispensed in duplicate in microtiter plates (30µl/well) containing 428 60µl of new-born bovine serum. After 90 minutes of incubation at 37ºC with gentle 429 agitation, complement action was blocked by adding brain heart infusion (BHI) broth 430 (150µl/well). After mixing the BHI broth with the bacterial suspension, 75µl were 431 plated by triplicate on TSA plates. 5 days after incubation at 37ºC, results were 432 expressed as the percentage of CFU recovered with respect to control samples where 433 new-born bovine serum was substituted by PBS. The experiment was repeated three 434 different times.

435 Virulence in mice.

436 Seven-week-old female BALB/c mice (ENVIGO, Harlan) were lodged in cages in 437 BSL-3 facilities with water and food ad libitum for 2, 8 or 12 weeks. Six groups of 5 438 mice each were inoculated with BaΔwadD or Ba-parental. Inocula were prepared in 439 sterile PBS and each mouse was administered intraperitoneally approximately with 5 x 440 104 CFU in 0,1 ml. To assess the exact dose retrospectively, dilutions of each inoculum 441 were plated by triplicate on TSA plates. Spleen CFUs in infected mice were counted at 442 2, 8 and 12 weeks after inoculation. The CFU counts were normalized by logarithmic 443 transformationIn and the meanreview log CFU/spleen values and the standard deviations were 444 calculated. The spleens were weighed and homogenized in 9 volumes of PBS and serial 445 ten-fold dilutions were accomplished and plated by triplicate on TSA plates. After 5 446 days of incubation at 37ºC the colonies were checked by crystal violet exclusion test 447 and PCR.

448 Statistical analysis.

449 Statistical significance for sensitivity to normal serum was evaluated with one-way 450 ANOVA followed by Dunnett´s multiple comparisons test (∗∗∗∗p < 0.0001). For 451 virulence analysis, statistical significance between the parental strain and the wadD 452 mutant was evaluated using t-Student independent-samples test (∗p < 0.05).

453 RESULTS

454 Screening for putative LPS core glycosyltransferases

14 455 A bioinformatic search in the Carbohydrate-Active Enzymes database CAZy 456 (www.cazy.org) revealed 23 ORFs in the genome of B. abortus 2308 (Table S1) that 457 could code for glycosyltransferases. We excluded from further analysis BAB1_0108- 458 cgs, which is involved in cyclic glucan synthesis (Briones et al., 2001), BAB1_1786- 459 mtgA and BAB1_1450-murG, both related to peptidoglycan synthesis, BAB1_1171- 460 lpxB, probably implicated in lipid A formation (Iriarte et al., 2004) and BAB1_0553- 461 wbkA, BAB1_0563-wbkE; BAB1_1000-wboA and BAB1_1000-wboB, 4 genes that 462 belong to the O-PS synthesis route (Godfroid et al., 2000; González et al., 2008; 463 McQuiston et al., 1999). Similarly, 4 ORFs correspond to those glycosyltransferases 464 already known to be involved in the synthesis of the LPS core: BAB1_0639-wadA 465 (Monreal et al., 2003), BAB1_0351-wadB (Gil-Ramírez et al., 2014), BAB1_1522- 466 wadC (Conde-Álvarez et al., 2012) and BAB2_0209-waaA (Iriarte et al., 2004) (see 467 Introduction). The remaining 11 ORFs are listed in Table 1, and data on their presence 468 in other Brucella spp. and genetic location are in the Supplemental Material (Table S2 469 and Figure S1). Of these, 7 (BAB1_0953, BAB2_0105, BAB2_0133, BAB2_0135, 470 BAB1_1620, BAB1_0607 and BAB1_0932) were highly conserved in all Brucella spp., 471 but 4 (BAB2_0693, BAB1_0417, BAB1_0114 and BAB1_0326) presented significant 472 differences when compared to B. abortus sequences, mainly due to frameshifts 473 generatingIn shorter proteins review(Table S2). Perusal of the literature revealed some 474 information on 4 of those 11 putative glycosyltransferases. Expression of BAB1_0326, 475 BAB2_0133 and BAB2_0135 has been shown to be controlled by MucR, a general 476 virulence regulator in Brucella (Caswell et al., 2013). However, although it has been 477 reported that B. abortus and B. melitensis mucR mutants have a defective LPS core, the 478 glycosyltransferases involved have not been identified (Caswell et al., 2013; Mirabella 479 et al., 2013). Also, BAB1_1620 expression has been reported to be controlled by 480 BvrR/BvrS, a master regulator of Brucella virulence that modulates OM homeostasis 481 and undetermined aspects of LPS structure (Manterola et al., 2005; Viadas et al., 2010).

482 We first analyzed the MucR and BvrR/BvrS controlled ORFs for involvement in 483 LPS synthesis. To this end, we constructed an insertion mutant in BAB1_0326 (since 484 the downstream ORF is oriented in the opposite direction) as well as non-polar deletion 485 mutants in BAB2_0133 and BAB2_0135 (both part of an operon), and BAB1_1620 486 (which, although isolated, is surrounded by genes implicated in the cell cycle). These 487 four mutants maintained the S phenotype in the crystal violet assay, suggesting that they

15 488 kept an intact O-PS. Then, SDS-proteinase-K LPSs were analyzed by SDS-PAGE and 489 Western-blot with both a polyclonal serum against S brucellae and anti-core MoAbs 490 A68/24G12/A08 and A68/24D08/G09, using as controls LPS from B. abortus 2308W 491 (Ba-parental), a B. abortus wadC mutant (Ba∆wadC) and a R per mutant (Ba∆per) 492 (Martínez-Gómez et al., 2018). The LPS of the four mutants presented S and R fractions 493 with migration profiles identical to those of Ba-parental LPS and reacted similarly with 494 the serum and MoAbs (Figure S2) strongly suggesting that the corresponding ORFs are 495 not required for normal LPS synthesis. When we complemented these observations by 496 inoculating BALB/c mice with BAB2_0133, BAB2_0135 and BAB1_1620 mutants, 497 they produced CFU/spleen that did not differ from those of Ba-parental at weeks 2 498 (p=0.99; 0.75 and 0.45 respectively) and 8 (p=0.95; 0.99 and 0.99) after infection. 499 Moreover, mutants in BAB2_0133 and BAB1_1620 behaved similarly to Ba-parental in 500 polycationic peptide resistance, and the former also performed as Ba-parental in 501 sensitivity to normal serum (Figure S3). These results are consistent with the idea that 502 the putative glycosyltransferases regulated by MucR or BvrR/BvrS are not involved in 503 the synthesis of LPS or of other components implicated in virulence, at least under the 504 conditions used in this study.

505 To investigate whether the remaining 7 putative glycosyltransferases (BAB1_0953, 506 BAB2_0105,In BAB2_0693, review BAB1_0607, BAB1_0114, BAB1_0932 and BAB1_0417) 507 were required for LPS synthesis, we constructed B. abortus 2308W insertion mutants in 508 each of them. All mutants were S by the crystal violet assay and the analysis of the 509 extracted LPS showed S fractions with a migration profile like that of Ba-parental and 510 reacted similarly with the anti S-Brucella polyclonal serum (Figure 2 and Figure S4). 511 Interestingly, although keeping the S fraction, mutant in BAB1_0953 lost reactivity in 512 the R fraction, suggesting a defect in the core and/or lipid A epitope(s) recognized by 513 polyclonal sera of infected animals (Rojas et al., 2001). Since this was not observed for 514 the other mutants, we investigated further BAB1_0953 and the phenotype associated 515 with its mutation.

516 BAB1_0953 encodes WadD, a previously unidentified glycosyltransferase 517 involved in the synthesis of the LPS core lateral branch.

518 BAB1_0953 is an isolated gene and the adjacent ORFs are encoded in the 519 complementary strand. Thus, it was very unlikely that a polar effect caused the LPS 520 phenotype of the insertion mutant. However, to rule out such a possibility, we

16 521 constructed a non-polar deletion mutant, hereafter named BaΔwadD following the 522 nomenclature previously established for Brucella LPS core genes (Gil-Ramírez et al., 523 2014; Reeves et al., 1996). BaΔwadD LPS showed a migration profile similar to that of 524 Ba-parental in the high molecular weight S-LPS fraction and an increased mobility in 525 the R-LPS one. Western-blot analysis with a polyclonal serum showed that, while the 526 former fraction kept the reactivity with this serum, the latter failed to react indicating a 527 significant alteration of the core-lipid A epitopes. To assign the defect to the core 528 oligosaccharide, we probed the LPS with MoAbs A68/24G12/A08 and 529 A68/24D08/G09, the binding of which to the R-LPS requires an intact mannose-GlcN 530 tetrasaccharide (Conde-Álvarez et al., 2012; Fontana et al., 2016). Both antibodies 531 failed to react with the R-LPS fraction and this failure was reverted upon insertion of a 532 complete wadD gene in the bacterial chromosome of the deletion mutant 533 (BaΔwadD::Tn7-PwadD). Moreover, this complementation restored both the migration 534 pattern of the R-LPS fraction to the level of the Ba-parental LPS and the reactivity with 535 the polyclonal serum. An ELISA with several anti-core MoAbs and whole bacteria 536 confirmed the core defect (Figure 4, upper panel).

537 We only observed small but constant differences in reactivity of Ba-parental and 538 Ba∆wadD with anti-Outer Membrane Proteins (OMP) antibodies (Figure 4, lower 539 pannel).In Although considering review the limitations of the method used, this suggests that the 540 presence of the O-PS and the defect in the core LPS could generate a steric hindrance 541 that would allow the access of antibodies to the OMPs (Bowden et al., 1995) or with the 542 possibility that wadD mutation could affect the amount of LPS in the outer membrane 543 or how it is inserted in the bacterial surface.

544 Although as signaled above, the final confirmation would require a complete 545 chemical analysis, all these results strongly suggest that wadD encodes a previously 546 unidentified glycosyltransferase involved in the synthesis of the core lateral branch.

547 WadD orthologs are present in all Brucella spp. but in a recently 548 characterized isolate from amphibians.

549 In silico analysis (Figure S5) showed that wadD was highly conserved in the core 550 brucellae including the “classical” spp. B. melitensis, B. suis (smooth LPS), and B. ovis 551 and B. canis (rough LPS) and also in other “non-classical” smooth Brucella spp.: B. 552 pinnipedialis, B. microti, B. ceti and B. vulpis. We also analyzed the presence of wadD 553 in the group of early-diverging brucellae that depart from the classical spp. and includes

17 554 B. inopinata strain BO1, B. inopinata-like strain BO2, an isolated strain from native 555 rodents in Australia (NF2653), and the Brucella spp. recently isolated from amphibians. 556 These early-diverging Brucella produce an atypical LPS (Al Dahouk et al., 2017; 557 Scholz et al., 2010; Soler-Lloréns et al., 2016; Tiller et al., 2010). Unexpectedly, wadD 558 was present in all of them but absent in Brucella spp. B13-0095, one of the four 559 Brucella strains isolated from frogs that have been completely sequenced (Soler-Lloréns 560 et al., 2016). In contrast, this strain conserves wadB and wadC. Finally, WadD was 72% 561 and 71% homologous to Ochrobactrum anthropi and O. intermedium orthologs, 562 respectively, two species that also belong to the α-2 Proteobacteria subclass and are the 563 closest genetic neighbors of Brucella.

564 Dysfunction of wadD generates increased sensitivity to cationic peptides and 565 normal serum

566 To test if the core defect displayed by Ba∆wadD affected resistance to polycationic 567 peptides, we used poly-L-ornithine, a mildly bactericidal cationic peptide and the two 568 known core mutants Ba∆wadB and Ba∆wadC plus the parental strain as controls. The 569 results (Figure 5A) showed that wadD dysfunction brought about a sensitivity similar to 570 that of Ba∆wadB but inferior to that of Ba∆wadC. These differences in sensitivity were 571 not due to growth defects because Ba∆wadD had a growth rate similar to that of Ba- 572 parental,In and experiments withreview the highly bactericidal lipopeptide polymyxin B 573 confirmed the role of wadD (Ba∆wadD MIC = 0.094 µg/ml versus MIC = 2 µg/ml for 574 both the mutant complemented with wild-type wadD and Ba-parental).

575 S brucellae are resistant to the bactericidal action of normal serum, a property 576 associated with both the O-PS hindrance to inner OM targets such as OMPs and the 577 PAMP modifications of the core that reduce binding of the complement activators of the 578 antibody-independent classical pathway (Conde-Álvarez et al., 2012; Fontana et al., 579 2016; Gil-Ramírez et al., 2014). We compared the sensitivity to newborn bovine and 580 ovine serum of Ba-parental and wadB, wadC or wadD mutants and observed that the 581 three core mutants were more sensitive than Ba-parental. The effect was more 582 remarkable for mutant wadC than for wadD or wadB mutants (Figure 5B).

583 Dysfunction of wadD generates attenuation detectable in the chronic phase in 584 the mouse model.

18 585 To analyze the role of wadD in virulence, we infected BALB/c mice (n=5) with 586 Ba∆wadD or Ba-parental and compared the CFU/spleen at weeks 2, 8 and 12 (Figure 587 6). At weeks 8 and 12 post-infection, the CFU numbers of Ba∆wadD were significantly 588 lower than those of Ba-parental (p=0.0003 and p=0.0073 respectively), showing that 589 wadD is required for full Brucella virulence in mice. This result is in line with previous 590 observations with wadB and wadC mutants (Conde-Álvarez et al., 2012; Gil-Ramírez et 591 al., 2014) and further confirms that an intact LPS core is necessary for virulence.

592 DISCUSSION

593 In this work we have analyzed the role of ORFs BAB1_0114, BAB1_0417, 594 BAB2_0693, BAB1_0932, BAB1_0607, BAB2_0105, BAB1_1620, BAB2_0133, 595 BAB2_0135, BAB1_0326 annotated as glycosyltransferases in the B. abortus genome. 596 Our results indicate that mutants in these ORFs react similarly to the parental strain in 597 the S and R LPS fractions and suggest that, in the studied conditions and with the 598 available techniques, they seem not to be required for the synthesis of a complete LPS. 599 Interestingly, the last three ORFs have been shown to be controlled by mucR, a 600 regulator of Brucella virulence. Although it has been reported that B. abortus and B. 601 melitensis mucR mutants have a defect in the core LPS (Caswell et al., 2013; Mirabella 602 et al., 2013), the glycosyltransferases responsible for this defect have not been 603 identified.In In this work we reviewhave shown that mutation of the mucR-regulated putative 604 glycosyltransferases BAB2_0133, BAB2_0135 and BAB1_0326 (Caswell et al., 2013) 605 does not affect the synthesis of the core, at least in the growth conditions tested. 606 Nevertheless, since the expression of these genes seems to be repressed by mucR 607 (Caswell et al., 2013), a single mutation in the ORF could not be sufficient for the 608 complete clarification of their role in LPS synthesis and further work would be required.

609 We have also analyzed in detail the role of the hypothetical glycosyltransferase 610 BAB1_1620, as it is regulated by the master two-component regulator BvrR/BvrS that 611 controls Brucella virulence and the expression of surface components. According to our 612 results, this ORF is not required for the synthesis of a complete LPS and is not 613 implicated in surface-dependent characteristics that confer resistance to polycationic 614 peptides or in virulence in the mouse model.

615 More interestingly, we report the identification of wadD, a gene encoding a 616 previously unidentified glycosyltransferase involved in the synthesis of the core section

19 617 not linked to the O-PS and thus, corroborate and extend previous work indicating that 618 the Brucella LPS core is a branched structure that constitutes a steric impairment 619 preventing the elements of the innate immune system to fight against Brucella (Conde- 620 Álvarez et al., 2012; Fontana et al., 2016; Gil-Ramírez et al., 2014; Kubler-Kielb and 621 Vinogradov, 2013), and thus contribute to Brucella virulence.

622 The discovery of genes wadC and wadB, involved in the synthesis of the lateral 623 branch not linked to the O-PS, was critical for the understanding of the structure and the 624 role of the core section in virulence. It has been clearly demostrated that in a wadC 625 mutant, the complete core lateral branch is absent because this mutant cannot 626 incorporate the mannose residue that is the depart of the lateral branch and it links to the 627 lipid A-core section (Conde-Álvarez et al., 2012; Fontana et al., 2016). In accordance, 628 deletion of wadC results in higher sensitivity to polycationic peptides and complement, 629 better recognition by the CD14-MD2-TLR4 receptor complex, maturation of dendritic 630 cells, secretion of pro-inflammatory cytokines (including Th1-type cytokines IL-12 and 631 IFN-γ), and attenuation in mice (Conde-Álvarez et al., 2012; Fontana et al., 2016). A 632 wadB mutant is also more sensitive to elements of the innate immune system and shows 633 attenuation in mice, although not to the levels of the wadC mutant (Gil-Ramírez et al., 634 2014). As we show here, disruption of wadD in B. abortus leads to a S strain with a core 635 defect lessIn severe than that reviewof the wadC mutant, more sensitive to polycationic peptides 636 and normal serum than the parental strain and attenuated in the murine model. 637 Interestingly, its sensitivity to polycationic peptides is similar to that of the wadB 638 mutant (Gil-Ramírez et al., 2014) and not as strong as that of the wadC mutant (Conde- 639 Álvarez et al., 2012), that has lost the complete branch, and its role in virulence became 640 apparent already at 8 weeks post-infection.

641 The resistance to polycationic peptides and the bactericidal action of normal serum 642 of mutant wadD strongly suggest a role in thwarting some effectors of innate immunity 643 and that this could be manifested in the early stages of infection. However, to observe 644 these effects would depend on the virulence model used. Indeed, the results obtained in 645 an in vitro test, where the bacteria are put directly in contact with the polycationic 646 peptides or with the serum can not be completely extrapolated to all situations in the in 647 vivo model where other factors apart from polycationic peptides and complement are 648 clearly taking part during the infection process. Still, the mouse model is the only well- 649 characterized laboratory model for Brucella virulence studies and requires that mice are

20 650 inoculated by the IP route. By this route, bacteria are not in contact with polycationic 651 peptides or serum proteins at the beginning of the infection process as they are taken up 652 and transported to the spleen rapidly. Nevertheless, the attenuation observed for 653 Brucella LPS core mutants is caused by an early activation of innate immunity, as we 654 have proved before (Conde-Álvarez et al., 2012; Fontana et al., 2016).

655 Our results would be compatible with the loss of one or few glucosamine residues 656 in the lateral branch of the wadD mutant, and with the fact that removal of these 657 residues would cause an increase in overall negative charge of the remaining LPS inner 658 section that will facilitate the binding of polycationic peptides.

659 According to chemical studies performed in B. melitensis, the core lateral branch 660 contains a mannose and four glucosamines residues assambled as follows: β-D-GlcpN-

661 (1à6)-β-D-GlcpN-(1à4)-[β-D-GlcpN-(1à6)]-β-D-GlcpN-(1à3)-α-D-Manp-(1à5) 662 (Figure 1). Taking into account that wadC, wadB and wadD are perfectly conserved in 663 B. melitensis and B. abortus, and since WadC adds the mannose (Conde-Álvarez et al., 664 2012; Fontana et al., 2016), in all likelihood the four glucosamines should be added by 665 WadB (Gil-Ramírez et al., 2014) and WadD. These glucosamines are bound to each

666 other by β-(1à6), or β-(1à4) links, and the one bound to mannose by β-(1à3) is also

667 linked to two glucosamine residues, both in β-(1à6) and β-(1à4). If WadB and WadD 668 are theIn only glycosyltransferases review involved in the assembly of the glucosamine 669 tetrasacchride and its binding to the manose residue, one of them (or both) could be 670 multi-fiunctional and thus able to add sugars in different linkage. Most 671 glycosyltransferase enzymes involved in lipooligosaccharide (LOS) or LPS core 672 biosynthesis are responsible for one type of sugar addition onto the growing chain 673 (Raetz and Whitfield, 2002). However, some bacterial glycosyltransferase enzymes of 674 the GT-2 family, to which WadD belongs, can be multi-functional and are 675 characterized by the presence of tandems of two active domains (DXD) on one 676 polypeptide, as is the case of Lgt3, responsible for the addition of three glucoses with 677 different linkages [β-(1-3), β-(1-4) and β-(1-6)] onto the inner core of Moraxella 678 catarralis LOS (Coutinho et al., 2003; Luke-Marshall et al., 2013). Interestingly, WadD 679 from B. melitensis, B. abortus and all the orthologs in the other Brucella spp. conserves 680 two DXD domains, opening the door to the possibility of a bi-functional role for this 681 glycosyltransferase (Figure S2). This DXD domain is not present in WadB. 682 Nevertheless, the understanding of the particular role of each glycosyltransferase in the

21 683 linkage of the different glucosamines to form the pentasacharide (glucosamine 684 tetrascharide bound to manose) would require the elucidation of the core chemical 685 structure of wadB and wadD mutants.

686 Contrary to most of the genes encoding glycosyltransferases implicated in the 687 synthesis of the LPS, that are clustered in the same or related regions of the Brucella 688 genome (González et al., 2008; Monreal et al., 2003; Rajashekara et al., 2004, 2008; 689 Vemulapalli et al., 2000), wadC, wadB and wadD (BAB1_1522, BAB1_0351 and 690 BAB1_0953 respectively), although all situated in chromosome I, are isolated and 691 surrounded by other ORFs apparently not related to LPS synthesis. This, and the fact 692 that some other genes involved in the synthesis of the core (manBcore and manCcore) 693 are situated in chromosome II (González et al., 2008; Monreal et al., 2003), makes even 694 more intriguing the identification of genes needed for the synthesis of Brucella core 695 LPS and its lateral branch. Thus, although we think all glycosyltransferases have been 696 identified, we can not ruled out that other glycosyltransferases could be required for the 697 assembly of the pentasaccharide that forms the core lateral branch.

698 In a chemical characterization of the core LPS previously performed in a B. 699 melitensis strain different from the one used in our studies, a glucose residue was found 700 linked to the mannose that is the depart of the lateral branch, and, if this were the case, a 701 new glycosyltransferaseIn couldreview be needed (Kubler-Kielb and Vinogradov, 2013). 702 However, it should be taken into account that the LPS extraction method and the B. 703 melitensis biovar used for the determination of the core structure in this experiement 704 were different from those used in our genetic and biochemical studies (Figure 1). It is 705 important to notice that the chemical structure we discuss in Figure 1 has been 706 elucidated in the same B. melitensis strain where the wadC gene was mutated (Fontana 707 et al., 2016) and, in this case, no glucose residues were detected. The fact that, as 708 discussed above, wadC, wadB and wadD are perfectly conserved in B. melitensis, 709 reinforces the interpretation of our results and the idea that the glycosyltransferases 710 encoded by the last two genes would be involved in the assembly of glucosamine 711 residues.

712 Nevertheless, we should consider that, although the phenotype of wadC mutant in 713 B. melitensis and B. abortus is similar (Conde-Álvarez et al., 2012; Fontana et al., 714 2016), previous results suggest that there could exist differences in the structure of the

22 715 core in these two Brucella spp., since they react differently with MoAbs against core 716 epitopes (González et al., 2008). Moreover, we have already seen that some of the 717 studied ORFs (and discarded in our first screening since they reacted as the parental 718 strain in the rough and smooth LPS fractions) present differences between B. abortus 719 and B. melitensis (Table S2). Thus, we could not discard them as the responsible for 720 these differences. To understand the final role of wadB, wadC and wadD, it would be 721 necessary to analyze and compare the chemical structure of the core section in mutants 722 in these genes in both spp.

723

724 ETHICS STATEMENT

725 Female BALB/c mice (ENVIGO, Harlan) were kept in cages with water and food 726 ad libitum under P3 biosafety conditions in the facilities of CIMA (registration code 727 ES31 2010000132) or CITA (registration code ES502970012005) 2 weeks before and 728 during the experiments. The procedures were in accordance with the current European 729 (directive 86/609/EEC) and Spanish (RD 53/2013) legislations, supervised by the 730 Animal Welfare Committee of the University of Navarra or CITA, and authorized by 731 “Gobierno de Navarra” (protocol number R102/2007) or “Gobierno de Aragón” 732 (protocolIn number R108/2009) review. 733 FUNDING

734 This research was supported by the Institute for Tropical Health funders (Obra 735 Social la CAIXA, “Fundaciones Caja Navarra” and “Roviralta, PROFAND, Ubesol, 736 ACUNSA”, and “Artai”) and grants MINECO (AGL2014-58795-C4-1-R). MSB is the 737 recipient of a Ph.D. Fellowship funded by the “Asociación de Amigos de la Universidad 738 de Navarra”. Work at CITA-Spain was also sustained by Grants from MINECO 739 (AGL2008-04514-C03-03/GAN and AGL2014-58795-C4-1-R).

740 ACKNOWLEDGMENT

741 The authors thank A. Delgado-López for excellent technical assistance. 742 743 Conflict of Interest Statement: 744 The authors declare that the research was conducted in the absence of any 745 commercial or financial relationships that could be construed as a potential conflict of 746 interest.

23 747 REFERENCES 748 Al Dahouk, S., Köhler, S., Occhialini, A., Jiménez de Bagüés, M. P., Hammerl, J. A., 749 Eisenberg, T., et al. (2017). Brucella spp. of amphibians comprise genomically 750 diverse motile strains competent for replication in macrophages and survival in 751 mammalian hosts. Sci. Rep. 7, 44420. doi:10.1038/srep44420. 752 Ariza, J. (1999). Brucellosis: an update. The perspective from the Mediterranean basin. 753 Rev.Med.Microbiol. 10, 125–135. 754 Barbier, T., Nicolas, C., and Letesson, J. J. (2011). Brucella adaptation and survival at 755 the crossroad of metabolism and virulence. FEBS Lett. 585, 2929–2934. 756 doi:10.1016/J.FEBSLET.2011.08.011. 757 Barbier, T., Zúñiga-Ripa, A., Moussa, S., Plovier, H., Sternon, J. F., Lázaro-Antón, L., 758 et al. (2018). Brucella central carbon metabolism: an update. Crit. Rev. Microbiol. 759 44, 182–211. doi:10.1080/1040841X.2017.1332002. 760 Barquero-Calvo, E., Chaves-Olarte, E., Weiss, D. S., Guzmán-Verri, C., Chacón-Díaz, 761 C., Rucavado, A., et al. (2007). Brucella abortus uses a stealthy strategy to avoid 762 activation of the innate immune system during the onset of infection. PLoS One 2, 763 e631. doi:10.1371/journal.pone.0000631. 764 Bowden, R., Cloeckaert, A., Zygmunt, M., Bernard, S., and Dubray, G. (1995). Surface 765 exposure of outer membrane protein and lipopolysaccharide epitopes in Brucella 766 species studied by enzyme-linked immunosorbent assay and flow cytometry. 767 Infect. Immun. 63, 3945–3952. 768 Briones, G., Iñón de Iannino, N., Roset, M., Vigliocco, A., Paulo, P. S., and Ugalde, R. 769 A. (2001). Brucella abortus cyclic beta-1,2-glucan mutants have reduced virulence 770 in mice and are defective in intracellular replication in HeLa cells. Infect. Immun. 771 69, 4528–35. doi:10.1128/IAI.69.7.4528-4535.2001. Caswell,In C. C., Elhassanny, review a. E. M., Planchin, E. E., Roux, C. M., Weeks-Gorospe, J. 772 773 N., Ficht, T. a., et al. (2013). Diverse genetic regulon of the virulence-associated 774 transcriptional regulator MucR in Brucella abortus 2308. Infect. Immun. 81, 1040– 775 1051. doi:10.1128/IAI.01097-12. 776 Choi, K. H., and Schweizer, H. P. (2006). mini-Tn7 insertion in bacteria with single 777 attTn7 sites: example Pseudomonas aeruginosa. Nat. Protoc. 1, 153–161. 778 doi:10.1038/nprot.2006.24. 779 Cloeckaert, A., De Wergifosse, P., Dubray, G., and Limet, J. N. (1990). Identification of 780 seven surface-exposed Brucella outer membrane proteins by use of monoclonal 781 antibodies: immunogold labeling for electron microscopy and enzyme-linked 782 immunosorbent assay. Infect. Immun. 58. 783 Cloeckaert, A., Grayon, M., Verger, J.-M., Letesson, J.-J., and Godfroid, F. (2000). 784 Conservation of seven genes involved in the biosynthesis of the lipopolysaccharide 785 O-side chain in Brucella spp. Res. Microbiol. 151, 209–216. doi:10.1016/S0923- 786 2508(00)00141-8. 787 Cloeckaert, A., Jacques, I., Bowden, R. ., Dubray, G., and Limet, J. N. (1993a). 788 Monoclonal antibodies to Brucella rough lipopolysaccharide: characterization and 789 evaluation of their protective effect against B. abortus. Res. Microbiol. Paris 144, 790 475–484. doi:10.1016/0923-2508. 791 Cloeckaert, A., Jacques, I., Grilló, M. J., Marín, C. M., Grayon, M., Blasco, J. M., et al.

24 792 (2004). Development and evaluation as vaccines in mice of Brucella melitensis 793 Rev.1 single and double deletion mutants of the bp26 and omp31 genes coding for 794 antigens of diagnostic significance in ovine brucellosis. Vaccine 22, 2827–2835. 795 doi:10.1016/j.vaccine.2004.01.001. 796 Cloeckaert, A., and Vizcaíno, N. (2004). “DNA Polymorphism and Taxonomy of 797 Brucella Species,” in Brucella. Molecular and Cellular Biology (Wymondham, 798 England: Horizon Bioscience), 1–24. 799 Cloeckaert, A., Weynants, V., Godfroid, J., Verger, J.-M., Grayon, M., and Zygmunt, 800 M. S. (1998). O-Polysaccharide Epitopic Heterogeneity at the Surface of Brucella 801 spp. Studied by Enzyme-Linked Immunosorbent Assay and Flow Cytometry. Clin. 802 Diagn. Lab. Immunol. 5, 862–870. 803 Cloeckaert, A., Zygmunt, M. S., Dubray, G., and Limet, J. N. (1993b). Characterization 804 of 0-polysaccharide specific monoclonal antibodies derived from mice infected 805 with the rough Brucella melitensis strain B115. J. Gen. Microbiol. 139, 1551– 806 1556. doi: 10.1099/00221287-139-7-1551 807 Conde-Álvarez, R., Arce-Gorvel, V., Gil-Ramírez, Y., Iriarte, M., Grilló, M. J., Gorvel, 808 J. P., et al. (2013). Lipopolysaccharide as a target for brucellosis vaccine design. 809 Microb. Pathog. 58, 29–34. doi:10.1016/j.micpath.2012.11.011. 810 Conde-Álvarez, R., Arce-Gorvel, V., Iriarte, M., Manček-Keber, M., Barquero-Calvo, 811 E., Palacios-Chaves, L., et al. (2012). The lipopolysaccharide core of Brucella 812 abortus acts as a shield against innate immunity recognition. PLoS Pathog. 8, 813 e1002675. doi:10.1371/journal.ppat.1002675. 814 Coutinho, P. M., Deleury, E., Davies, G. J., and Henrissat, B. (2003). An Evolving 815 Hierarchical Family Classification for Glycosyltransferases. J. Mol. Biol. 328, 816 307–317. doi:10.1016/S0022-2836(03)00307-3. 817 Dubray,In G., and Limet, J. (1987).review Evidence of heterogeneity of lipopolysaccharides 818 among Brucella biovars in relation to A and M specificities. Ann. Inst. Pasteur. 819 Microbiol. 138, 27–37. 820 Fontana, C., Conde-Álvarez, R., Ståhle, J., Holst, O., Iriarte, M., Zhao, Y., et al. (2016). 821 Structural studies of lipopolysaccharide defective mutants from Brucella melitensis 822 identify a core oligosaccharide critical in virulence. J. Biol. Chem. 291, 7727– 823 7741. doi:10.1074/jbc.M115.701540. 824 Garin-Bastuji, B., Bowden, R. A., Dubray, G., and Limet, J. N. (1990). Sodium dodecyl 825 sulfate-polyacrylamide gel electrophoresis and immunoblotting analysis of 826 smooth-lipopolysaccharide heterogeneity among Brucella biovars related to A and 827 M specificities. J. Clin. Microbiol. 28, 2169–74. 828 Gil-Ramírez, Y., Conde-Álvarez, R., Palacios-Chaves, L., Zúñiga-Ripa, A., Grilló, M.- 829 J., Arce-Gorvel, V., et al. (2014). The identification of wadB, a new 830 glycosyltransferase gene, confirms the branched structure and the role in virulence 831 of the lipopolysaccharide core of Brucella abortus. Microb. Pathog. 73, 53–9. 832 doi:10.1016/j.micpath.2014.06.002. 833 Godfroid, F., Cloeckaert, A., Taminiau, B., Danese, I., Tibor, A., De Bolle, X., et al. 834 (2000). Genetic organisation of the lipopolysaccharide O-antigen biosynthesis 835 region of Brucella melitensis 16M (wbk). Res. Microbiol. 151, 655–668. 836 doi:10.1016/S0923-2508(00)90130-X.

25 837 González, D., Grilló, M.-J. M., De Miguel, M. M.-J., Ali, T., Arce-Gorvel, V., Delrue, 838 R. R.-M., et al. (2008). Brucellosis vaccines: assessment of Brucella melitensis 839 lipopolysaccharide rough mutants defective in core and O-polysaccharide synthesis 840 and export. PLoS One 3, e2760. doi:10.1371/journal.pone.0002760. 841 Iriarte, M., González, D., Delrue, R.-M., Monreal, D., Conde, R., López-Goñi, I., et al. 842 (2004). “Brucella Lipopolysaccharide: Structure, Biosynthesis and Genetics,” in 843 Brucella. Molecular and Cellular Biology (Wymondham, England: Horizon 844 Bioscience), 159–192. 845 Kubler-Kielb, J., and Vinogradov, E. (2013). The study of the core part and non- 846 repeating elements of the O-antigen of Brucella lipopolysaccharide. Carbohydr. 847 Res. 366, 33–7. doi:10.1016/j.carres.2012.11.004. 848 Lapaque, N., Moriyon, I., Moreno, E., and Gorvel, J.-P. (2005). Brucella 849 lipopolysaccharide acts as a virulence factor. Curr. Opin. Microbiol. 8, 60–66. 850 doi:10.1016/j.mib.2004.12.003. 851 Llobet, E., March, C., Giménez, P., and Bengoechea, J. A. (2009). Klebsiella 852 pneumoniae OmpA Confers Resistance to Antimicrobial Peptides. Antimicrob. 853 Agents Chemother. 53, 298–302. doi:10.1128/AAC.00657-08. 854 Luke-Marshall, N. R., Edwards, K. J., Sauberan, S., St Michael, F., Vinogradov, E. V., 855 Cox, A. D., et al. (2013). Characterization of a trifunctional glucosyltransferase 856 essential for Moraxella catarrhalis lipooligosaccharide assembly. Glycobiology 23, 857 1013–21. doi:10.1093/glycob/cwt042. 858 Manterola, L., Moriyón, I., Moreno, E., Sola-Landa, A., Weiss, D. S., Koch, M. H. J., et 859 al. (2005). The lipopolysaccharide of Brucella abortus BvrS/BvrR mutants 860 contains lipid A modifications and has higher affinity for bactericidal cationic 861 peptides. J. Bacteriol. 187, 5631–9. doi:10.1128/JB.187.16.5631-5639.2005. 862 MartínezIn-Gómez, E., Ståhle, review J., Gil-Ramírez, Y., Zúñiga-Ripa, A., Zaccheus, M., 863 Moriyón, I., et al. (2018). Genomic insertion of a heterologous acetyltransferase 864 generates a new lipopolysaccharide antigenic structure in Brucella abortus and 865 Brucella melitensis. Front. Microbiol. 9, 1092. doi:10.3389/FMICB.2018.01092. 866 McDermott, J., Grace, D., and Zinsstag, J. (2013). Economics of brucellosis impact and 867 control in low-income countries. Rev. Sci. Tech. 32, 249–61. 868 McQuiston, J. R., Vemulapalli, R., Inzana, T. J., Schurig, G. G., Sriranganathan, N., 869 Fritzinger, D., et al. (1999). Genetic characterization of a Tn5-disrupted 870 glycosyltransferase gene homolog in Brucella abortus and its effect on 871 lipopolysaccharide composition and virulence. Infect. Immun. 67, 3830–5. 872 Mirabella, A., Terwagne, M., Zygmunt, M. S., Cloeckaert, A., De Bolle, X., and 873 Letesson, J. J. (2013). Brucella melitensis MucR, an orthologue of Sinorhizobium 874 meliloti MucR, is involved in resistance to oxidative, detergent, and saline stresses 875 and cell envelope modifications. J. Bacteriol. 195, 453–65. doi:10.1128/JB.01336- 876 12. 877 Monreal, D., Grilló, M. J., González, D., Marín, C. M., De Miguel, M. J., López-Goñi, 878 I., et al. (2003). Characterization of Brucella abortus O-polysaccharide and core 879 lipopolysaccharide mutants and demonstration that a complete core is required for 880 rough vaccines to be efficient against Brucella abortus and Brucella ovis in the 881 mouse mo. Infect. Immun. 71, 3261–71. doi:10.1128/iai.71.6.3261-3271.2003.

26 882 Palacios-Chaves, L., Conde-Álvarez, R., Gil-Ramírez, Y., Zúñiga-Ripa, A., Barquero- 883 Calvo, E., Chacón-Díaz, C., et al. (2011). Brucella abortus ornithine lipids are 884 dispensable outer membrane components devoid of a marked pathogen-associated 885 molecular pattern. PLoS One 6, e16030. doi:10.1371/journal.pone.0016030. 886 Raetz, C. R. H., and Whitfield, C. (2002). Lipopolysaccharide endotoxins. Annu. Rev. 887 Biochem. 71, 635–700. doi:10.1146/annurev.biochem.71.110601.135414. 888 Rajashekara, G., Covert, J., Petersen, E., Eskra, L., and Splitter, G. (2008). Genomic 889 island 2 of Brucella melitensis is a major virulence determinant: functional 890 analyses of genomic islands. J. Bacteriol. 190, 6243–6252. doi:10.1128/JB.00520- 891 08. 892 Rajashekara, G., Glasner, J. D., Glover, D. A., and Splitter, G. A. (2004). Comparative 893 whole-genome hybridization reveals genomic islands in Brucella species. J. 894 Bacteriol. 186, 5040–51. doi:10.1128/JB.186.15.5040-5051.2004. 895 Reeves, P. R., Hobbs, M., Valvano, M. A., Skurnik, M., Whitfield, C., Coplin, D., et al. 896 (1996). Bacterial polysaccharide synthesis and gene nomenclature. Trends 897 Microbiol. 4, 495–503. doi:10.1016/S0966-842X(97)82912-5. 898 Rojas, N., Freer, E., Weintraub, A., Ramirez, M., Lind, S., and Moreno45, E. (1994). 899 Immunochemical identification of Brucella abortus lipopolysaccharide epitopes. 1, 900 206–213. doi:10.1111/j.1439-0450.2001.00476.x 901 Rojas, N., Zamora, O., Cascante, J., Garita, D., and Moreno, E. (2001). Comparison of 902 the antibody response in adult cattle against different epitopes of Brucella abortus 903 lipopolysaccharide. J. Vet. Med. Ser. B 48, 623–629. doi:10.1111/j.1439- 904 0450.2001.00476.x. 905 Ronneau, S., Moussa, S., Barbier, T., Conde-Álvarez, R., Zúñiga-Ripa, A., Moriyon, I., 906 et al. (2014). Brucella , nitrogen and virulence. Crit. Rev. Microbiol. 42, 1–19. 907 doi:10.3109/1040841X.2014.962480.In review 908 Scholz, H. C., Nockler, K., Gollner, C., Bahn, P., Vergnaud, G., Tomaso, H., et al. 909 (2010). Brucella inopinata sp. nov., isolated from a breast implant infection. Int. J. 910 Syst. Evol. Microbiol. 60, 801–808. doi:10.1099/ijs.0.011148-0. 911 Scupham, A. J., and Triplett, E. W. (1997). Isolation and characterization of the UDP- 912 glucose 4′-epimerase-encoding gene, galE, from Brucella abortus 2308. Gene 202, 913 53–59. doi:10.1016/S0378-1119(97)00453-8. 914 Simon, R., Priefer, U., and Pühler, A. (1983). A broad host range mobilization system 915 for in vivo genetic engineering: transposon mutagenesis in Gram negative bacteria. 916 Bio/Technology 1, 784–791. doi:10.1038/nbt1183-784. 917 Soler-Lloréns, P. F., Quance, C. R., Lawhon, S. D., Stuber, T. P., Edwards, J. F., Ficht, 918 T. A., et al. (2016). A Brucella spp. isolate from a Pac-Man Frog (Ceratophrys 919 ornata) reveals characteristics departing from classical Brucellae. Front. Cell. 920 Infect. Microbiol. 6, 116. doi:10.3389/fcimb.2016.00116. 921 Soler-Lloréns, P., Gil-Ramírez, Y., Zabalza-Baranguá, A., Iriarte, M., Conde-Álvarez, 922 R., Zúñiga-Ripa, A., et al. (2014). Mutants in the lipopolysaccharide of Brucella 923 ovis are attenuated and protect against B. ovis infection in mice. Vet. Res. 45, 72. 924 doi:10.1186/s13567-014-0072-0. 925 Suárez-Esquivel, M., Ruiz-Villalobos, N., Castillo-Zeledón, A., Jiménez-Rojas, C., 926 Roop II, R. M., Comerci, D. J., et al. (2016). Brucella abortus Strain 2308

27 927 Wisconsin genome: importance of the definition of reference strains. Front. 928 Microbiol. 7, 1–6. doi:10.3389/fmicb.2016.01557 929 Tibor, A., Wansard, V., Bielartz, V., Delrue, R.-M., Danese, I., Michel, P., et al. (2002). 930 Effect of omp10 or omp19 deletion on Brucella abortus outer membrane properties 931 and virulence in mice. Infect. Immun. 70, 5540–6. doi:10.1128/IAI.70.10.5540- 932 5546.2002. 933 Tiller, R. V, Gee, J. E., Frace, M. A., Taylor, T. K., Setubal, J. C., Hoffmaster, A. R., et 934 al. (2010). Characterization of novel Brucella strains originating from wild native 935 rodent species in North Queensland, Australia. Appl. Environ. Microbiol. 76, 936 5837–45. doi:10.1128/AEM.00620-10. 937 Tsai, C.-M., and Frasch, C. E. (1982). A sensitive silver stain for detecting 938 lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 119, 115–119. 939 doi:10.1016/0003-2697(82)90673-X. 940 Vemulapalli, R., He, Y., Buccolo, L. S., Boyle, S. M., Sriranganathan, N., and Schurig, 941 G. G. (2000). Complementation of Brucella abortus RB51 with a functional wboA 942 gene results in O-antigen synthesis and enhanced vaccine efficacy but no change in 943 rough phenotype and attenuation. Infect. Immun. 68, 3927–32. 944 doi:10.1128/IAI.68.7.3927-3932.2000 945 Viadas, C., Rodríguez, M. C., Sangari, F. J., Gorvel, J.-P., García-Lobo, J. M., and 946 López-Goñi, I. (2010). Transcriptome analysis of the Brucella abortus BvrR/BvrS 947 two-component regulatory system. PLoS One 5, e10216. 948 doi:10.1371/journal.pone.0010216. 949 Zhao, Y., Hanniffy, S., Arce-Gorvel, V., Conde-Alvarez, R., Oh, S., Moriyón, I., et al. 950 (2017). Immunomodulatory properties of Brucella melitensis lipopolysaccharide 951 determinants on mouse dendritic cells in vitro and in vivo. Virulence, 0. 952 doi:10.1080/21505594.2017.1386831.In review 953 Zúñiga-Ripa, A., Barbier, T., Conde-Álvarez, R., Martínez-Gómez, E., Palacios- 954 Chaves, L., Gil-Ramírez, Y., et al. (2014). Brucella abortus depends on pyruvate 955 phosphate dikinase and malic enzyme but not on Fbp and GlpX fructose-1,6- 956 bisphosphatases for full virulence in laboratory models. J. Bacteriol. 196, 3045–57. 957 doi:10.1128/JB.01663-14. 958 Zygmunt, M. S., Jacques, I., Bernardet, N., and Cloeckaert, A. (2012). 959 Lipopolysaccharide heterogeneity in the atypical group of novel emerging Brucella 960 species. Clin. Vaccine Immunol. 19, 1370–3. doi:10.1128/CVI.00300-12. 961 962 FIGURE CAPTIONS 963 Figure 1. Schematic representation of Brucella melitensis LPS core. The site of 964 action of the glycosyltransferases identified thus far is indicated with arrows. Kdo (3- 965 deoxy-d-manno-octulosonic acid), Glc (glucose), Man (mannose), GlcN (glucosamine) 966 (adapted from Fontana et al. 2016 and Gil-Ramírez et al. 2014).

967

968 Figure 2. Mutation of B. abortus BAB1_0953 but not of other candidate for LPS 969 glycosyltransferase genes abrogates reactivity of the R-LPS but not of the S-LPS 970 fraction with antibodies from B. melitensis infected rabbits. Ba∆per is a N-formyl-

28 971 perosamine synthase mutant lacking the LPS O-polysaccharide, and Ba∆wadC is an 972 LPS mutant bearing O-polysaccharides but defective in the mannose-tetraglucosamine 973 lateral branch of the core oligosaccharide (see Figure 1). 974

975 Figure 3. Mutation of wadD generates an LPS core defect. Left panel, SDS-PAGE 976 electrophoresis and silver staining of SDS-proteinase K extracts; central panel, Western- 977 blot analysis of SDS-proteinase K extracts with a polyclonal serum of a B. melitensis 978 infected rabbit; right panel, Western-blot analysis of SDS-proteinase K extracts with 979 monoclonal anti-core antibody A68/24G12/A08. 980 981 Figure 4. Outer membrane epitopes in BaΔwadD and Ba-parental. Upper panel, 982 anti-O-PS and anti-R-LPS; lower panel, anti-OMP antibodies measured by ELISA test. 983 984 Figure 5. Dysfunction of wadD increases the sensitivity to poly-L-ornithine and 985 non-immune sera. (A) Poly-L-ornithine MIC determined by the serial dilution method 986 (results representative of three independent experiments in which Ba-parental, 987 BaΔwadC, BaΔwadB and BaΔwadD where assayed simultaneously). (B) Survival after 988 incubation for 90 minutes in bovine non-immune serum (media ± standard error of 989 technical triplicates). Means were compared by one-way ANOVA followed by 990 Dunnett´s multiple comparisons test (∗∗∗∗p < 0.0001). Differences were significative 991 between wadC and wadB mutants (p=0.0023), between wadC and wadD mutants 992 (p<0.0001), but not between wadB and wadD mutants (p=0.1958). 993 994 Figure 6. Mutants in wadD showed slight attenuation at late stage of infection. 995 Spleen CFUIns in infected BALB/creview mice were counted at 2, 8 and 12 weeks after 996 inoculation of 5 x 104 CFU. Means were compared by t-Student independent-samples 997 test (∗ p<0.05). For mice infected with the parental strain the counts were 6.25+/-0.38; 998 6.98+/-0.24 and 5.85+/-0.32 respectively, and for mice infected with the wadD mutant 999 6.56+/-0.22; 6.28+/-0.54 and 5.17+/-0.27. 1000 1001 1002 1003 1004

29 1005 Table 1. ORF coding for B. abortus hypothetical glycosyltransferases, family to which 1006 they belong, predicted function and the corresponding mutant LPS phenotype by 1007 Western-Blot analysis. 1008

Mutant LPS reactivity ORF Family Predicted function (KEGG) O-PSa R-LPSb

BAB1_0326 2 Glycosyltransferase + +

BAB2_0133 2 Glycosyltransferase + +

BAB2_0105 2 Glycosyltransferase + +

BAB2_0693 2 Glycosyltransferase + +

BAB1_0953 2 Glycosyltransferase + -

BAB1_1620 25 Glycosyltransferase + +

Penicillin-binding protein 1A transpeptidase BAB1_0607 51 + + domain - Glycosyltransferase

Penicillin-binding protein transpeptidase BAB1_0114 51 domain: ATP/GTP-binding site motif A (P- + + loop) - Glycosyltransferase

Penicillin-binding protein 1A transpeptidase BAB1_0932 51 + + In reviewdomain - Glycosyltransferase

Possible dolichyl-phosphate-mannose- BAB2_0135 83 + + protein mannosyltransferase family protein

BAB1_0417 ncc Conserved hypothetical protein + +

1009 a Reactivity to polyclonal serum from a rabbit infected with B. melitensis 16M. 1010 b Reactivity to monoclonal antibodies anti R-LPS: A68/24G12/A08 and A68/24D08/G09. 1011 c Glycosyltransferase family non-classified.

30 Figure 1.JPEG

In review Figure 2.JPEG

In review Figure 3.JPEG

In review Figure 4.JPEG

In review Figure 5.JPEG

In review Figure 6.JPEG

In review