Evolution and Host-Adaptation of the Mammalian Bordetellae

Evolution and host-adaptation of the mammalian bordetellae

D.A. Diavatopoulos Cover: “Tree partially hidden by fog banks”

Th e tree is a widely used symbol for the evolution of bacterial species, the fog represents the uncertainties scientists working on bacterial evolution need to cope with.

Adapted from http://www.sxc.hu

Printed by Ponsen & Looijen BV., Wageningen

Th e printing of this thesis was ﬁ nancially supported by: J.E. Jurriaanse Stichting, Dr. Ir. van de Laar Stichting, Eijkman Graduate School for Infection and Immunity. Evolution and host-adaptation of the mammalian bordetellae

Evolutie en gastheeradaptatie van de zoogdier bordetellae

(met een samenvatting in het Nederlands)

Proefschrift

ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de Rector Magniﬁ cus, Prof. Dr. W.H. Gispen ingevolge het besluit van het College voor Promoties in het openbaar te verdedigen op vrijdag 10 februari 2006 des middags te 12.45 uur

door

Dimitri Adriaan Diavatopoulos geboren op 4 december 1976, te Lage Zwaluwe Promotores: Prof. Dr. F.R. Mooi Prof. Dr. J. Verhoef

Co-promotor: Dr. L.M. Schouls

Paranimfen: M. Hijnen T.A. Diavatopoulos

Th e research presented in this thesis was performed at the Laboratory for Vaccine Preventable Diseases at the National Institute for Public Health and the Environment (RIVM), Bilthoven, the Netherlands and at the Eijkman-Winkler Institute, University Medical Center Utrecht, the Netherlands. Voor Cecile Science in a nutshell:

“For a moment, nothing happened.Th en, after a second or so, nothing continued to happen.” by Douglas Adams contents

Chapter 1 General Introduction 9

Chapter 2 Genetic Relationships of the Mammalian Bordetellae 31 Chapter 2&3, PLoS Pathogens, 2005

Chapter 3 Genetic Changes Associated With the Adaptation of a B. bronchiseptica Lineage to the Human Host 47 Chapter 2&3, PLoS Pathogens, 2005

Chapter 4 Identiﬁ cation and Characterization of B. bronchiseptica Complex IV-Speciﬁ c Sequences 69 Manuscript in preparation

Chapter 5 Transfer of an Iron-Uptake Island From Bordetella pertussis to Bordetella holmesii 81 Submitted

Chapter 6 Evolution of the Bordetella Autotransporter Pertactin: Identiﬁ cation of Regions Subject to Positive Selection 113 Submitted

Chapter 7 Summarizing Discussion 131

A Nederlandse samenvatting 141 B Dankwoord 145 C Curriculum Vitae 149 D Appendix 153

Chapter Chapter 1

General Introduction Chapter 

bordetella genus 1 Th e genus Bordetella, member of the β-proteobacterial family of the Alcaligenaceae, currently comprises nine species. Four Bordetella species are associated with respiratory disease in mammals: Bordetella bronchiseptica, Bordetella parapertussis, Bordetella pertussis and Bordetella holmesii (see Table 1). Of these species, B. pertussis and to a lesser extent B. parapertussis are the etiological agents of an acute respiratory illness in humans, designated whooping cough or pertussis. Until recent years, pertussis morbidity in adults was signiﬁ cantly underestimated, as it was assumed that vaccination induced long-lasting immunity. Pertussis is especially severe in young, unvaccinated children, but recent studies also showed that B. pertussis infection is a common cause of persistent cough in adults and adolescents, and this may have important consequences for its epidemiology 1,2. In many parts of Europe and America, mass pertussis vaccination was introduced in the 1940s and 1950s, resulting in a major decrease in pertussis incidence. Estimations of current pertussis disease suggest that, world-wide, approximately 48.5 million people contract pertussis resulting in 295,000 deaths 3. Morbidity and mortality due to pertussis is highest in developing countries, as mass vaccination is often not readily available in these countries. However, in the last two decades, the number of pertussis cases has also increased in countries with traditionally high vaccination coverage 4-6.

Table 1. Characteristics of Bordetella species that cause respiratory illness in mammalian species. Species Diseases Hosts B. pertussis Whooping cough Humans

B. parapertussishu Mild whooping cough Humans

B. parapertussisov Chronic nonprogressive pneumonia Sheep B. bronchiseptica Chronic tracheobronchitis; broncho- Mice, rats, guinea pigs, skunks, opossums, pneumonia; septicemia; kennel cough rabbits, raccoons, cats, dogs, ferrets, fox- (dogs); atrophic rhinitis (swine); catarrh es, pigs, hedgehogs, sheep, koalas, tur- (rabbits) keys, leopards, horses, lesser bushbabies, monkeys, humans B. holmesii Septicemia; endocarditis; respiratory ill- Immunocompromised humans ness

origin of pertussis Th e origin of the disease pertussis is still a mystery. Although it has very typical symptoms in children and was one of the major causes of child mortality previous to the introduction of vaccination, the ﬁ rst written reference to the disease in Europe is found in 1540 7. In contrast, historical descriptions of other major diseases with typical symptoms, such as diphtheria and tetanus, can be found in the ancient Greek literature. The first description of a pertussis epidemic, which occurred in Paris, was given by Baillon in

10 General Introduction

1578. In the 16th and 17th century, descriptions of whooping cough and epidemics in Europe are documented frequently in the literature 8. Th e absence of references to pertussis- 1 like symptoms in the ancient (European) literature has been taken as evidence that the association of B. pertussis with humans is of recent origin. properties of bordetella species In 1906, Bordet and Gengou were the ﬁ rst to describe B. pertussis, which was isolated from the sputum of a patient with whooping cough 9. B. parapertussis was ﬁ rst described in 1937 as a clinical entity in humans that resembled B. pertussis 10,11. B. pertussis and B. parapertussishu have been isolated from humans only and cause non-invasive acute and transient infections. No evidence exists that an animal reservoir exists for B. pertussis. B. parapertussis is comprised of two distinct lineages, one of which is found exclusively in humans, and one that is found exclusively in sheep, designated B. parapertussishu and B. parapertussisov, respectively. Previous research based on multilocus enzyme electrophoresis and the distribution of insertion sequence elements suggested that there is little or no 12 exchange between these reservoirs . Not only B. pertussis, but also B. parapertussishu has been described to cause epidemics in countries across the world (reviewed in 13). Usually,

B. parapertussishu infections result in less severe whooping cough-like symptoms in young infants and in adults, with paroxysmal coughing, but without lymphocytosis that is one of 14 the hallmarks of B. pertussis infections . It was shown that in 40% of the B. parapertussishu 15 infections, infected persons remained without symptoms . In sheep, B. parapertussisov has been described to cause chronic non-progressive pneumonia 16,17. B. bronchiseptica was fi rst described in literature in 1910 18 and has since been isolated from the respiratory tract of a diverse range of mammalian species, including humans. In contrast to B. pertussis and B. parapertussishu, which are non-invasive, B. bronchiseptica has also been described to cause septicemia in humans 19-22. Usually, however, B. bronchiseptica causes chronic and often asymptomatic infections of the upper respiratory tract. Diseases caused by this bacterium include kennel cough in dogs, atrophic rhinitis in swine and snuffl es in rabbits 23. Human infections by B. bronchiseptica, mainly in immune-compromised individuals 24,25, have also been described. It is generally assumed that humans are not a natural reservoir of B. bronchiseptica, and human infections are usually considered to be of zoonotic origin 26. B. bronchiseptica is increasingly being used as a model organism for B. pertussis, due to the fastidious nature of B. pertussis and the advantages of using B. bronchiseptica in a mouse or rat model, both of which are natural hosts for this bacterium. B. bronchiseptica has the ability, in contrast to B. pertussis and B. parapertussis, to survive outside the host for prolonged periods of time; it has been shown to survive and even replicate in lake water without added nutrients 27,28. B. holmesii was fi rst described in 1995 as a species that resembled B. pertussis, and was originally isolated from septicemic patients with immune disorders 29. During a pertussis

11 Chapter 

Figure 1. Respiratory tract infection BvgA-phosphorylation 1 model of humans by Bordetella bronchiseptica, Bordetella parapertussis and <26°C 35°C 37°C Bordetella pertussis. Bacteria enter the host through inhalation of aerosols. Bvg- Bvgi Bvg+ Adherence takes place in the nasal Flagella BipA BrkA/BrkB Early genes cavity, followed by colonization of the O-antigen TTSS FHA pharynx and the trachea and in rare Urease Prn Fim cases the lungs. They subsequently vrgs TcfA Late genes pass through diff erent Bvg phases Dnt Ptx during infection, depending on the vags CyA level of phosphorylation of BvgA, which is modulated by environmental signals such as temperature. The Bvg- phase (the avirulent phase) is thought to be important for survival Nasal cavity outside the host, and in this phase the virulence-repressed genes (vrgs) are Upper expressed. B. bronchiseptica expresses Pharynx respiratory fl agella for motility and urease in the Bordetella tract bvg- phase. The Bvgi phase is thought to be important for initial colonization Larynx of the nasopharynx, and is character- Lower respiratory ized by optimal expression of BipA. Trachea tract In the Bvg+ phase, the virulence-activated genes are expressed (vags), Primary bronchi including adhesins and toxins. This Lungs phase may be divided in an early and a late phase. In the early Bvg+ phase, adhesins are expressed but not toxins, and this phase is associated with colonization of the upper respiratory tract. The late Bvg+ phase is characterized by expression of toxins in addition to adhesins, and these toxins may modulate the immune system of the host for survival and may contribute to induction of disease symptoms that are important for transmission. The list of virulence factors is incomplete, and only represents the most studied virulence factors. Other virulence factors, either controlled by BvgAS or not, may be important for transmission as well. Abbreviations: Bvg, Bordetella virulence gene; vrgs, virulence-repressed genes; BipA, Bvg intermediate phase protein A; BrkA/B, Bordetella resistance to killing A/B; TTSS, type III secretion system; Prn, pertactin; TcfA, tracheal colonization factor A; Dnt, dermonecrotic toxin; FHA, fi lamentous haemagglutinin; Fim, fi mbriae; Ptx, pertussis toxin; CyA, adenylate cyclase. A color version of this fi gure is available in the appendix. epidemic in Massachusetts, it has also been isolated from the respiratory tract of previously healthy patients with pertussis symptoms 30. Although most B. holmesii infections occur in immune-compromised individuals, a serious infection of a healthy adolescent was also described. A recent study suggested that B. holmesii may be especially pathogenic to asplenic patients in which it causes bacteremia 31. Not much is known about the pathogenesis and transmission of this pathogen.

12 General Introduction virulence factors 1 Th e close evolutionary relationship between B. bronchiseptica, B. parapertussis and B. pertussis is refl ected in the fact that they express a similar set of virulence factors and contain nearly identical BvgAS global virulence control systems. On the other hand, they also diff er with respect to certain virulence factors, and this in turn may refl ect their diff erences in host range and disease symptoms. Th e bacteria enter the upper respiratory tract by inhalation of aerosols by the host, and subsequently binding of the bacteria to the ciliated epithelial cells is initiated through adhesins. Upon adherence, the mammalian bordetellae produce a range of immune-modulating substances, such as toxins, which interfere with clearance of the bacteria and enable them to colonize the host (Figure 1). Several secreted or surface- associated factors have been described to play an important role in the virulence of the mammalian bordetellae. the bvgas system Th e expression of many of almost all virulence factors is controlled by the BvgAS (Bordetella virulence genes) two-component system, encoded by the bvg locus. Th is two-component system is comprised of a sensor protein, BvgS, that senses and responds to external signals, and which can (de-)activate the transcriptional regulator protein BvgA that in turn controls the expression of many virulence-associated genes. Th e BvgAS system mediates the transition between what is now regarded as a spectrum of states. Th ese states can range between two extremes with on the one hand the virulent phase (bvg+) and on the other hand the avirulent phase (bvg-) (reviewed in 32). Th e virulent phase is important for infection of the host, while the avirulent phase is considered to be important for survival outside the host. Genes that are up-regulated in the bvg+ phase are called vags (virulence activated genes), while genes that are down-regulated in this phase are called vrgs (virulence repressed genes). In the bvg- phase, vags are repressed and vrgs are expressed. More recently, an intermediate state has been described, designated the bvgi-phase. Th is phase is characterized by the expression of adhesins, but not toxins, and also by the expression of BipA (Bordetella intermediate gene), which is expressed specifi cally in this phase 33-36. A third gene, bvgR, was also described in the bvg genetic locus, which encodes another transcriptional regulator, BvgR 37,38. Expression of bvgR is activated by BvgA, and BvgR in turn down-regulates the expression of virulence- repressed genes. A number of stimuli have been identifi ed that can modulate the transition between the bvg- phases, including temperature, nicotinic acid and sulfate. Temperature may be considered a natural stimulus, as the virulent phase is induced at temperatures corresponding to body temperature, while the avirulent phase is induced at temperatures of 25°C or lower 39. Other external signals may fi ne-regulate the expression of various virulence factors in vivo in response to changing micro-environments.

13 Chapter 

commonly and differently expressed virulence factors 1 B. bronchiseptica, B. parapertussis and B. pertussis express a similar set of virulence factors. Th ese include adhesins, such as fi lamentous hemagglutinin (FHA), the fi mbriae (Fim) and pertactin (Prn); but also toxins such as adenylate cyclase toxin (CyaA), tracheal cytotoxin (TCT) and dermonecrotic toxin (Dnt). Th ere are also numerous virulence factors that do not belong to either adhesins or toxins, and these include for example siderophores, which are involved in iron acquisition from the host, and the type III secretion system (TTSS), which can inject modulating substances into host-cells. Some virulence factors are expressed in only one species, e.g. expression of pertussis toxin (Ptx) and tracheal colonization factor A (TcfA) has only been observed in B. pertussis 40,41. Table 2 shows the commonly and diff erentially expressed virulence factors for the mammalian bordetellae. In the next paragraphs, the virulence factors lipopolysaccharide (LPS), Prn, Ptx, DNT and alcaligin are described in detail.

Table 2. Virulence factors of Bordetella species B. pertussis B. parapertussis B. bronchiseptica B. holmesii Gene(s) Production Gene(s) Production Gene(s) Production Gene(s) Production Toxins Ptx + + + - + - - - Cya + + + + + + - - Dnt + + + + + + - - Tct + + + -

Adhesins Prn + + + + + + ? - TcfA + + + - + - ? ? Fha + + + + + + - - Fimbriae + + + + + + - -

Other some BrkA + + + ? + ?? strains Alcaligin + + + + + + ? ? TTSS + - + - + + ? ?

Abbreviations: Ptx, pertussis toxin; Cya, adenylate cyclase; Dnt, dermonecrotic toxin; Tct, tracheal cytotoxin; Prn, pertactin; TcfA, tracheal colonization factor A; Fha, ﬁ lamentous haemagglutinin; BrkA, Bordetella resistance to killing protein A; TTSS, type III secretion system

lipopolysaccharide Lipopolysaccharide (LPS), also known as endotoxin, is a major component of the outer- membrane of gram-negative bacteria, and may constitute up to 75% of the surface. LPS is essential for survival of the bacteria, as it provides a protective barrier against the

14 General Introduction environment 42. LPS also plays an important role in the interaction with the host immune system. Van den Akker showed that production of LPS by Bordetella species was modulated 1 by both temperature and sulfate 43, and thus, LPS modulation is probably regulated by the BvgAS system. Th e LPS molecules of gram-negative bacteria usually consist of three, covalently linked, major domains: the lipid A, the branched chain oligosaccharide core and the hydrophilic O-antigen (reviewed in 42,44,45). Several genetic loci have been shown to play a role in the synthesis of these domains in Bordetella, although there probably are many other, unknown, genes that may play a role in the biosynthesis of LPS. Th ese genetic loci include the lpx locus (lipid A), the waa locus (inner core), the wlb locus (outer core) and the wbm locus (O-antigen) 46-50. Extensive polymorphism in the LPS molecular structure has been observed between Bordetella species, for instance with regard to the expression of the outer-membrane associated PagP. PagP is a palmitoyl transferase, that mediates acylation of the lipid A, in a BvgAS-dependent matter 51. Preliminary evidence suggests that B. parapertussis also expresses PagP. In contrast, although B. pertussis also contains the gene encoding PagP (pagP), it is not expressed due to the disruption of pagP by an insertion sequence element in the promoter region 51,52. PagP, which is regulated by the BvgAS system, was shown to be required for persistent colonization of the mouse by B. bronchiseptica 51,53. Polymorphism is also observed between the Bordetella species with regard to the O- antigen, which is only produced by B. bronchiseptica and B. parapertussis. Th is structure is absent from B. pertussis due to the deletion of the wbm locus 47. It was shown that for B. bronchiseptica, the O-antigen was essential for establishing chronic infections, but not for initial colonization 54. Further, B. bronchiseptica and B. parapertussis mutant strains lacking O-antigen were highly susceptible to complement-mediated killing. Possibly to compensate for this lack of O-antigen, B. pertussis expresses BrkA (Bordetella resistance to killing), which confers resistance to serum, and through another mechanism that has not been elucidated 55. Although the gene encoding BrkA (brkA) is also present in B. bronchiseptica, only some strains express BrkA 56. O-antigen may interfere with other virulence factors, possibly by steric hindrance by the long stretches of O-antigenic repeats. In Shigella, shortening of the O-antigen increased the virulence by enhancing the function of the type III secretion system, without loss of LPS protection against the immune system 57. Th e trisaccharide (the outer core) of the LPS has been shown, similar to the O-antigen, not to be required for initial colonization of the host by B. bronchiseptica, but is important for establishing persistent infections in the presence of adaptive immunity 58. It was also shown to confer resistance against surfactant A (SP-A), a bactericidal host factor that can bind to the lipid A 59-61. SP-A could inhibit adherence to ciliated epithelial cells of B. bronchiseptica mutants lacking the trisaccharide, but showed only limited eﬀ ects on ciliary adherence of wild type B. bronchiseptica strains 62.

15 Chapter 

pertactin 1 Th e genomes of B. bronchiseptica, B. parapertussis and B. pertussis contain 21 genes that encode autotransporters 63, of which B. bronchiseptica comprises the most complete set. Autotransporters are produced as precursors and have the ability to mediate transfer to the outer-membrane through their own C-terminal region. Autotransporters typically consist of a signal sequence, a passenger domain and a conserved autotransporter domain. Th e signal sequence at the N-terminus directs the transport of the passenger domain to the periplasmic space. In the periplasm, this sequence is cleaved off , and the C-terminal autotransporter domain subsequently forms a β-barrel-shaped pore-like structure in the outer-membrane through which the N-terminal passenger domain can translocate. Th e passenger domain may encode diverse functions such as proteases, adhesins, toxins and lipases. It may either be cleaved off after translocation to the outer-membrane or it may remain associated to the surface through non-covalent binding 64. One of the autotransporters that is expressed by all three species, in a BvgAS-dependent manner, is Prn. Th e three species produce mature Prn-molecules with diff erent apparent molecular sizes on gel, a 68-kDa protein (P.68) in B. bronchiseptica 65, a 69-kDa protein (P.69) in B. pertussis 66 and a 70-kDa protein (P.70) in B. parapertussis 67. Comparative analysis of the nucleotide sequences showed that the prn genes of B. bronchiseptica and B. parapertussis were more similar to each other than to B. pertussis 65. Prn contains a Arg- Gly-Asp (RGD) motif that has been proposed to play a role in adherence to host cells 68. Comparison of the prn genes of clinical isolates showed that Prn contains two amino acid repeat regions of which the number of repeat units is polymorphic 69,70. Th e fi rst amino acid repeat region, designated region 1, is located adjacent to the RGD motif, and the second repeat region, designated region 2, is located more proximate to the C-terminal autotransporter domain. P.69 was shown to elicit protective antibodies in a number of studies 71-74, and it is currently a component of many acellular pertussis vaccines. Th e crystal structure of the passenger domain of P.69 has been determined 75,76, and the location of linear epitopes has been studied in detail 73,77. Pertactin is a component of many acellular pertussis vaccines. Polymorphism has been observed in pertactin in B. pertussis and it was shown that the pertactin type represented by the vaccine has been gradually replaced by non-vaccine types in the years following vaccination 69,78. Further, the non-vaccine pertactin types have been associated with the reemergence of pertussis in the Netherlands 69. It has been hypothesized that vaccine pressure has selected for non-vaccine types that are better adapted to a situation in which high vaccination coverage exists.

pertussis toxin Pertussis toxin (Ptx) is an A-B toxin, composed of ﬁ ve subunits that are designated S1- S5. Th e subunits are encoded by the ptx operon, comprised of the genes ptxABCDE 79,80.

16 General Introduction

Translocation of Ptx through the outer membrane is mediated through the associated type IV secretion system Ptl (pertussis toxin liberation) that is comprised of nine proteins 81,82. 1 Th e ptl locus, which encodes Ptl, is located adjacent to the ptx locus and within the same transcriptional unit as Ptx. Ptx is thought to be expressed and produced exclusively by B. pertussis. Curiously, in both B. bronchiseptica and B. parapertussis the genes that are required for expression and secretion of Ptx appear to be largely intact 40. Interspecifi c diff erences in expression of Ptx are though to be the result of mutations in the promoter region of the ptx/ ptl locus. Indeed, replacement of the ptx/ptl promoter sequence in a B. parapertussishu and a B. bronchiseptica strain with the B. pertussis promoter actually resulted in the production and secretion of biologically active Ptx 83. Based upon interspecifi c comparative analysis of the ptx/ptl promoter sequences, Parkhill et al. postulated that Ptx expression has increased in B. pertussis in the course of its evolution, in contrast to the previous assumption that B. bronchiseptica and B. parapertussis have lost the ability to express Ptx 63. Th e expression of Ptx and Ptl is controlled by the BvgAS system, and Ptx is thought to play an important role in transmission of B. pertussis. Many studies, both in vitro and in mouse and rat models, have demonstrated major eff ects of Ptx on mammalian cell signaling pathways. Th ese may lead to a cascade of biological eff ects, such as insulinemia, histamine sensitization, leukocytosis 84,85 and both immunosuppression and stimulation 86-88. Important roles have been shown for Ptx in the early phases of respiratory tract infections of mice by B. pertussis. Ptx was further shown to suppress production of B. pertussis-specifi c antibodies 89,90. Several studies have also shown that Ptx may also function as an adhesin 91,92. Although Ptx was shown to aff ect the innate and the acquired immune system, a specifi c role for Ptx has so far not been demonstrated in transmission between humans. Ptx is a major component of all acellular pertussis vaccines, and polymorphism in Ptx was shown to be correlated to the emergence of nonvaccine-type B. pertussis strains in the Netherlands 4,69. dermonecrotic toxin Although the dermonecrotic toxin (Dnt) was one of the fi rst described B. pertussis virulence factors 93, it is one of the less studied toxins of B. pertussis. Th e toxin, encoded by the BvgAS-regulated gene dnt, is expressed by B. bronchiseptica, B. parapertussis and B. pertussis 94. When purifi ed Dnt is injected intradermally into animals, it produces localized necrotic lesions, hence the name dermonecrotic toxin 93,95,96. For mice, Dnt is lethal at low doses when injected intravenously 93,96,97. Dnt is an atypical toxin in the sense that it is not secreted but remains in the cytoplasm 98,99. Like the cytotoxic necrotizing toxin factor (Cnf) that is produced by Escherichia coli, Dnt activates the small Rho GTPase through deamidation or polyamination 100,101, which leads to stimulation of DNA replication, but blocks cell division. When administered to mammalian cells, it has been shown to induce drastic morphological changes in vitro 100. Although it has been shown to induce extensive mucosal damage in pigs, resulting in turbinate atrophy, and to enhance colonization of the upper

17 Chapter 

respiratory tract 102, its role in pertussis pathogenesis has not yet been elucidated. B. pertussis 1 transposon mutants lacking Dnt were no less virulent than wild type B. pertussis strains in a mouse model 103.

alcaligin Iron is essential to sustain growth of Bordetella species in the respiratory tract of their host. Mammalian hosts keep the concentration of free iron as low as 10-9 M through sequestration by heme, transferrin, lactoferrin and other iron-chelating factors. Bacterial pathogens have developed diff erent strategies to acquire the necessary iron from the host (reviewed in 104). One of these strategies is the production and secretion of low-molecular-weight compounds that can chelate iron with high affi nity, known as siderophores. B. bronchiseptica, B. parapertussis and B. pertussis can produce the siderophore alcaligin, and are able to utilize siderophores from other bacteria 105,106. Alcaligin is a hydroxamate siderophore, and is encoded by the alc operon, comprised of the genes alcABCDE 107-110. Transport of iron-bound siderophores across the outer membrane is mediated through a TonB-dependent mechanism. Th e genomes of the mammalian bordetellae comprise up to 16 TonB-dependent ferric complex receptors, of which B. bronchiseptica may contain the most complete set 63. Th e receptor for iron-bound alcaligin is FauA, which is encoded by the gene fauA, located adjacent to the alc operon 111. Th e genes alcABCDE encode enzymes for the biosynthesis of alcaligin. Under iron-rich conditions, expression of these genes is repressed by Fur. Expression of fauA and alcABCDE is also regulated by the AraC-like regulator AlcR, encoded by alcR 108,109,112. Recently, the gene alcS, previously designated bcr because of homology to the E. coli bicyclomycin resistance protein, was shown to encode the alcaligin sensor protein AlcS, which was shown to be important for alcaligin secretion 110. Although iron is essential for growth and virulence of Bordetella species, an AlcR defi cient B. pertussis, unable to produce and secrete alcaligin, was as virulent as the wild type in mouse respiratory tract infections 109. In contrast, alcaligin was required to confer maximal virulence for B. bronchiseptica in pigs 113.

18 General Introduction virulence, genome reduction, host adaptation and evolution 1 In recent years, developments in molecular biology have led to the ability to determine the complete genome sequences of a bacterium in a relatively short period of time. Recently, the complete sequence of a Mycoplasma genitalum strain was determined in just four hours 114. Th e comparative genomic analysis of closely related species, some of which may be pathogens, has led to an increased understanding of the mechanism by which bacteria evolve and adapt to changing environments (reviewed in 115). It was shown that bacterial genomes are not static and are often in a permanent state of genome fl ux. Also, the traditional belief that bacteria contain a single circular genome has been challenged; it was shown that Vibrio cholerae contains two circular chromosomes 116 and Borrelia burfdorferi comprises one linear chromosome in addition to at least 17 linear and circular plasmids 117. Several pathogens have evolved predominantly by acquisition of novel functions through DNA uptake. Many distinct mechanisms exist through which bacteria can acquire DNA, such as transformation, conjugation and bacteriophage 118. Examples of pathogenic bacteria in which DNA uptake has played a pivotal role in their evolution are H. pylori, which causes gastro-intestinal infections 119-122 and N. meningitidis, a commensal of the respiratory tract that can cause diseases such as meningitis and septicemia 123. Genomic islands are large genomic regions, often associated with tRNA-encoding genes and mobile genetic elements. Th ey usually diff er in GC-content in comparison to the rest of the genome. Genomic islands contribute to adaptation to the environment in diverse ways, illustrated by their functional association to virulence, symbiosis, antibiotic resistance, metabolism and many other functions (reviewed in 124). If these genomic islands confer an increase in virulence potential, they are called pathogenicity islands, and these have been detected in a number of pathogenic bacterial species (reviewed in 125-127). Yersinia pestis, the causative agent of the plague, and as such responsible for the deaths of countless millions in the Middle Ages, is a clone that has evolved from the enteropathogen Yersinia pseudotuberculosis approximately 6,500 years ago 128,129. Comparison of the genomes of these two closely related species showed that diff erences in virulence may be attributed to the acquisition of virulence factors, including a high-pathogenicity island (HPI), followed by the deletion and inactivation of genes no longer required for transmission 130,131. Th us, the mechanism of Y. pestis evolution is characterized by the uptake of DNA, followed by a “sifting” round of genes no longer necessary. Th is large-scale inactivation of genes, or genome decay, has also been suggested to be the primary driving force of evolution of some other bacterial pathogens. Genome decay may also be associated with increased adaptation to the host, as is observed in the obligate intracellular pathogens Rickettsia prowazekii 132 and Treponema pallidum 133. Th e genome of Mycobacterium leprae, which causes leprosy, is also riddled with pseudogenes and is also much smaller than the genomes of other, closely related Mycobacterial species, such as Mycobacterium tuberculosis 134,135. Large-scale inactivation of genes may be accomplished through diff erent means, such as the accumulation of mobile

19 Chapter 

insertion sequence elements that insert in and disrupt genes or the expansion of repetitive 1 DNA sequences. Th us, not only gain of function may enable a bacterium to evolve into a pathogen or restrict the host range, but genome reduction (e.g. through deletion of genes encoding proteins that are readily recognized by the host immune system) may also be an important strategy.

evolution of bordetella species causing respiratory infections in mammals Previous studies, based on DNA-DNA hybridization, multilocus enzyme electrophoresis (MLEE), the distribution of insertion sequence elements and the comparison of the 16S and 23S ribosomal RNA sequences suggested that B. bronchiseptica, B. parapertussis and B. pertussis are very closely related 12,136-139. Th ese species are commonly referred to as the mammalian bordetellae or the “classical bordetellae”. Musser et al. and Van der Zee et al.

showed that B. parapertussishu and B. pertussis independently evolved from a B. bronchiseptica- 12,136 like ancestor . Both B. pertussis and B. parapertussishu show very limited genetic diversity, whereas B. bronchiseptica is a much more diverse species, genetically 12,136,137,140. Recently, the genome sequences of single strains of B. bronchiseptica, B. parapertussis and B. pertussis have been described by Parkhill et al. 63. Th e genome sizes of these species are 5.3,

4.8 and 4.1 million basepairs, respectively. Th e genomes of B. pertussis and B. parapertussishu contain an unusually high number of pseudogenes (9.4% and 5%, respectively) and insertion sequence elements are widespread in both species. Comparative analysis of the genome

sequences showed that the host-restricted species B. pertussis and B. parapertussishu evolved from a B. bronchiseptica-like ancestor approximately 0.7-3.5 and 0.27-1.4 million years ago, respectively, and that their evolution was accompanied by extensive genome decay. Parkhill et al. suggested that adaptation to a single host by these species was primarily the consequence of loss of function, as there were no indications of signiﬁ cant DNA acquisition in both

B. pertussis and B. parapertussishu. Analysis of the genes that were most frequently deleted or inactivated showed that many of these were involved in membrane transport, small- molecule metabolism, regulation of gene expression and synthesis of surface structures 63. Th is was consistent with a comparative genomic hybridization study in which the genomes of various Bordetella strains were hybridized to microarrays representing the genomes of the sequenced Bordetella strains 141. B. holmesii is a recently identiﬁ ed Bordetella species, and not much is known about its evolution and epidemiology. Based upon the nearly identical 16S ribosomal RNA sequences of B. holmesii and B. pertussis, it was originally assumed that these two species were very closely related 29, an assumption that was consistent with the discovery of the B. pertussis insertion sequence element IS481 in B. holmesii 142. However, characterization of the BvgAS virulence control system and analysis of the cellular fatty acid composition suggested that B. holmesii may be more closely related to B. avium, an important pathogen of poultry 29,143. Several studies have shown that B. holmesii does not express many of the virulence factors

20 General Introduction that are commonly expressed by B. bronchiseptica, B. parapertussis and B. pertussis 144,145. Due to the lack of formal phylogenetic studies, the position of B. holmesii within the Bordetella 1 genus remains unclear.

21 Chapter 

outline of this thesis 1 Th e aim of this thesis was to elucidate the phylogenetic relationships between the Bordetella species causing respiratory disease in mammalian species, and to identify key molecular events that are important for adaptation to humans. In chapter 2 the population genetic relationships between the mammalian bordetellae (B. bronchiseptica, B. parapertussis and B. pertussis) are clarifi ed using a combination of multilocus sequence typing, sequencing of a virulence gene and by determining the distribution of several insertion sequence elements. Evidence is provided for the existence of a B. bronchiseptica lineage, associated with humans, that is closely related to B. pertussis. In chapter 3, the population structure of the mammalian bordetellae, as defi ned in chapter 2, is refi ned by comparative genomic hybridization to a Bordetella microarray. Host adaptation of B. pertussis and B. bronchiseptica complex IV strains is studied at a genome-wide level. Th e absence or polymorphism of several major virulence factors such as pertussis toxin, dermonecrotic toxin and the lipopolysaccharide is suggested to be associated with adaptation of complex IV strains to the human host, and may result in a decrease in cross-immunity to other human-specifi c Bordetella species in the human host. Gene content comparison suggested genome reduction in the B. bronchiseptica complex IV strains compared to its ancestral lineage, and this was further studied in chapter 4. Using subtractive hybridization, the genomes of B. bronchiseptica complex IV strains were compared to B. bronchiseptica complex I to identify complex IV-specifi c sequences. Further, by comparison of two B. bronchiseptica complex IV strains, the level of DNA uptake within complex IV is estimated. We provide evidence that DNA uptake has played little or no role in the evolution of B. bronchiseptica complex IV and B. pertussis. In chapter 5, the phylogenetic position of B. holmesii in the Bordetella genus is elucidated by sequencing of housekeeping genes and by comparative genomic hybridization. It is shown that B. holmesii is much closer related to B. avium and B. holmesii than to B. pertussis, despite the presence of B. pertussis-like 16S rRNA genes in the genome of B. holmesii. Comparative genomic hybridization identifi ed the presence of a genomic island in the B. holmesii genome conferring iron-uptake functions, and detailed analysis suggested that this island was most likely acquired from B. pertussis and may have played an important role in the evolution and emergence of B. holmesii. Th e evolution of the pertactin gene in the mammalian bordetellae is studied in chapter 6. Positive selection acting on pertactin is studied by comparing several nucleotide substitution models; and positively selected codons are compared to the location of epitopes. Evidence is provided for immune evasion of the pertactin gene. In the summarizing discussion in chapter 7, the evolution of the Bordetella species that cause respiratory disease in mammals is discussed.

22 General Introduction reference list cough due to Bordetella parapertussis: an unresolved problem. Expert. Rev Anti. Infect. Ther. 1 2, 447-454. 1. Cherry,J.D. (1999). Epidemiological, clinical, and laboratory aspects of pertussis in adults. Clin. 14. Heininger,U., Stehr,K., Schmitt-Grohe,S., Lorenz,C., Infect. Dis. 28 Suppl 2, S112-S117. Rost,R., Christenson,P.D., Uberall,M., and Cherry,J. D. (1994). Clinical characteristics of illness caused 2. Birkebaek,N.H., Kristiansen,M., Seefeldt,T., Degn,J., by Bordetella parapertussis compared with illness Moller,A., Heron,I., Andersen,P.L., Moller,J.K., and caused by Bordetella pertussis. Pediatr. Infect. Dis. Ostergard,L. (1999). Bordetella pertussis and J. 13, 306-309. chronic cough in adults. Clin. Infect. Dis. 29, 1239- 1242. 15. VYSOKA,B. (1958). The epidemiology of pertussis and parapertussis. J. Hyg. Epidemiol. Microbiol. 3. Crowcroft,N.S., Stein,C., Duclos,P., and Immunol. 2, 196-204. Birmingham,M. (2003). How best to estimate the global burden of pertussis? Lancet Infect. Dis. 3, 16. Weiss,A.A. (1992). The Genus Bordetella. In: The 413-418. prokaryotes: a handbook on the biology of bacteria, ed. A.Balows, H.G.Truper, M.Dworking, 4. Mooi,F.R., van Loo,I.H., and King,A.J. (2001). W.Harder, and K.H.SchleiferBerlin: Springer, 2530- Adaptation of Bordetella pertussis to Vaccination: 2543. A Cause for Its Reemergence? Emerg. Infect. Dis. 2001. ;7. (3 Suppl):526. -8. 7, 526-528. 17. Cullinane,L.C., Alley,M.R., Marshall,R.B., and Manktelow,B.W. (1987). Bordetella parapertussis 5. Crowcroft,N.S. and Britto,J. (2002). Whooping from lambs. N. Z. Vet. J. 35, 175. cough--a continuing problem. BMJ 324, 1537- 1538. 18. Ferry,N.S. (1910). A preliminary report of the bacterial fi ndings in canine distemper. Am. Vet. 6. Orenstein,W.A. (1999). Pertussis in adults: Rev. 37, 499-504. epidemiology, signs, symptoms, and implications for vaccination. Clin. Infect. Dis. 28 Suppl 2, S147- 19. Bauwens,J.E., Spach,D.H., Schacker,T.W., S150. Mustafa,M.M., and Bowden,R.A. (1992). Bordetella bronchiseptica pneumonia and bacteremia 7. Major R.H. (1945). Classic Descriptions of Disease, following bone marrow transplantation. J. Clin. Springfi eld, C. C. Thomas. Microbiol. 30, 2474-2475. 8. LAPIN,J.H. (1943). Whooping Cough, Springfi eld, 20. Espino,A.R., Gascon,L.F., Amor,T.J., Mongil,R., I, Illinois: Charles C. Thomas. Garcia,C.F., and de la Torre,C.C. (1992). [Bacteremia 9. Bordet,J. and Gengou,O. Le caused by Bordetella bronchiseptica in the course microbe de la coqueluche. Ann. of infl ammatory tinea capitis]. An. Esp. Pediatr. 36, Inst.Pasteur 20, 731-741. 1906. Paris. 323-325. Ref Type: Generic 21. Borras,S.M., Bonal,J., Bonet,J., Arnal,J., Roca,F., 10. Bradford,W.L. and Slavin,B. (1937). An organism and Caralps,A. (1991). Bordetella bronchiseptica resembling Hemophilus pertussis. Am. J. Pub. septicemia in a hemodialysis patient. Nephron 59, Health 27, 1277-1282. 676.

11. Eldering,G. and Kendrick,P. (1937). A group of 22. Katzenstein,D.A., Ciofalo,L., and Jordan,M.C. (1984). cultures resembling both Bacillus pertussis and Bordetella bronchiseptica bacteremia. West J. Med. Bacillus bronchisepticus but identical with neither. 140, 96-98. J. Bacteriol. 33. 23. Goodnow,R.A. (1980). Biology of Bordetella 12. van der Zee,A., Mooi,F., van Embden,J., and bronchiseptica. Microbiol. Rev. 44, 722-738. Musser,J. (1997). Molecular evolution and host 24. Gueirard,P., Weber,C., Le Coustumier,A., and adaptation of Bordetella spp.: phylogenetic Guiso,N. (1995). Human Bordetella bronchiseptica analysis using multilocus enzyme electrophoresis infection related to contact with infected animals: and typing with three insertion sequences. J. persistence of bacteria in host. J. Clin. Microbiol 33, Bacteriol. 179, 6609-6617. 2002-2006. 13. Watanabe,M. and Nagai,M. (2004). Whooping

23 Chapter 

25. Bauwens,J.E., Spach,D.H., Schacker,T.W., in reduced virulence and increased resistance 1 Mustafa,M.M., and Bowden,R.A. (1992). Bordetella to starvation, and identifi es a new class of Bvg- bronchiseptica pneumonia and bacteremia regulated antigens. Mol. Microbiol. 24, 671-685. following bone marrow transplantation. J. Clin. 36. Stockbauer,K.E., Fuchslocher,B., Miller,J. Microbiol 30, 2474-2475. F., and Cotter,P.A. (2001). Identifi cation and 26. Woolfrey,B.F. and Moody,J.A. (1991). Human characterization of BipA, a Bordetella Bvg- infections associated with Bordetella intermediate phase protein. Mol. Microbiol. 2001. bronchiseptica. Clin. Microbiol. Rev. 4, 243-255. Jan. ;39. (1):65. -78. 39, 65-78.

27. Porter,J.F., Parton,R., and Wardlaw,A.C. (1991). 37. Merkel,T.J. and Stibitz,S. (1995). Identifi cation of a Growth and survival of Bordetella bronchiseptica locus required for the regulation of bvg-repressed in natural waters and in buff ered saline without genes in Bordetella pertussis. J. Bacteriol. 177, added nutrients. Appl. Environ. Microbiol. 57, 2727-2736. 1202-1206. 38. Merkel,T.J., Barros,C., and Stibitz,S. (1998). 28. Porter,J.F. and Wardlaw,A.C. (1993). Long-term Characterization of the bvgR locus of Bordetella survival of Bordetella bronchiseptica in lakewater pertussis. J. Bacteriol. 180, 1682-1690. and in buff ered saline without added nutrients. 39. Martinez de Tejada,G., Miller,J.F., and Cotter,P. FEMS Microbiol. Lett. 110, 33-36. A. (1996). Comparative analysis of the virulence 29. Weyant,R.S., Hollis,D.G., Weaver,R.E., Amin,M. control systems of Bordetella pertussis and F., Steigerwalt,A.G., O’Connor,S.P., Whitney,A. Bordetella bronchiseptica. Molecular Microbiology M., Daneshvar,M.I., Moss,C.W., and Brenner,D.J. 22, 895-908. (1995). Bordetella holmesii sp. nov., a new gram- 40. Marchitto,K.S., Smith,S.G., Locht,C., and Keith,J. negative species associated with septicemia. J. M. (1987). Nucleotide sequence homology to Clin. Microbiol. 33, 1-7. pertussis toxin gene in Bordetella bronchiseptica 30. Yih,W.K., Silva,E.A., Ida,J., Harrington,N., Lett,S.M., and Bordetella parapertussis. Infect. Immun. 55, and George,H. (1999). Bordetella holmesii-like 497-501. organisms isolated from Massachusetts patients 41. Finn,T.M. and Stevens,L.A. (1995). Tracheal with pertussis-like symptoms. Emerg. Infect. Dis. 5, colonization factor: a Bordetella pertussis secreted 441-443. virulence determinant. Mol. Microbiol. 16, 625- 31. Shepard,C.W., Daneshvar,M.I., Kaiser,R.M., 634. Ashford,D.A., Lonsway,D., Patel,J.B., Morey,R. 42. Gronow,S. and Brade,H. (2001). Lipopolysaccharide E., Jordan,J.G., Weyant,R.S., and Fischer,M. biosynthesis: which steps do bacteria need to (2004). Bordetella holmesii bacteremia: a newly survive? J. Endotoxin. Res. 7, 3-23. recognized clinical entity among asplenic patients. Clin. Infect. Dis. 38, 799-804. 43. van den Akker,W.M. (1998). Lipopolysaccharide expression within the genus Bordetella: infl uence 32. Cotter,P.A. and Jones,A.M. (2003). Phosphorelay of temperature and phase variation. Microbiology control of virulence gene expression in Bordetella. 144 ( Pt 6), 1527-1535. Trends Microbiol 11, 367-373. 44. Caroff ,M., Karibian,D., Cavaillon,J.M., and Haeff ner- 33. Scarlato,V., Arico,B., Prugnola,A., and Rappuoli,R. Cavaillon,N. (2002). Structural and functional (1991). Sequential activation and environmental analyses of bacterial lipopolysaccharides. regulation of virulence genes in Bordetella Microbes. Infect. 4, 915-926. pertussis. EMBO J. 10, 3971-3975. 45. Erridge,C., Bennett-Guerrero,E., and 34. Prugnola,A., Arico,B., Manetti,R., Rappuoli,R., and Poxton,I.R. (2002). Structure and function of Scarlato,V. (1995). Response of the bvg regulon of lipopolysaccharides. Microbes. Infect. 4, 837-851. Bordetella pertussis to diff erent temperatures and short-term temperature shifts. Microbiology 141 ( 46. Isobe,T., White,K.A., Allen,A.G., Peacock,M., Raetz,C. Pt 10), 2529-2534. R., and Maskell,D.J. (1999). Bordetella pertussis waaA encodes a monofunctional 2-keto-3-deoxy- 35. Cotter,P.A. and Miller,J.F. (1997). A mutation in D-manno-octulosonic acid transferase that can the Bordetella bronchiseptica bvgS gene results complement an Escherichia coli waaA mutation. J.

24 General Introduction

Bacteriol. 181, 2648-2651. Science 307, 1313-1317.

47. Preston,A., Allen,A.G., Cadisch,J., Thomas,R., 58. Harvill,E.T., Preston,A., Cotter,P.A., Allen,A.G., 1 Stevens,K., Churcher,C.M., Badcock,K.L., Parkhill,J., Maskell,D.J., and Miller,J.F. (2000). Multiple roles for Barrell,B., and Maskell,D.J. (1999). Genetic basis Bordetella lipopolysaccharide molecules during for lipopolysaccharide O-antigen biosynthesis in respiratory tract infection. Infect. Immun. 68, 6720- bordetellae. Infect. Immun. 67, 3763-3767. 6728.

48. Preston,A. and Maskell,D. (2001). The molecular 59. Banemann,A., Deppisch,H., and Gross,R. (1998). The genetics and role in infection of lipopolysaccharide lipopolysaccharide of Bordetella bronchiseptica biosynthesis in the Bordetellae. J. Endotoxin. Res. acts as a protective shield against antimicrobial 7, 251-261. peptides. Infect. Immun. 66, 5607-5612.

49. Allen,A.G., Isobe,T., and Maskell,D.J. (1998). 60. Schaeff er,L.M., McCormack,F.X., Wu,H., and Weiss,A. Identifi cation and cloning of waaF (rfaF) from A. (2004). Interactions of pulmonary collectins with Bordetella pertussis and use to generate Bordetella bronchiseptica and Bordetella pertussis mutants of Bordetella spp. with deep rough lipopolysaccharide elucidate the structural basis lipopolysaccharide. J. Bacteriol. 180, 35-40. of their antimicrobial activities. Infect. Immun. 72, 7124-7130. 50. Allen,A. and Maskell,D. (1996). The identifi cation, cloning and mutagenesis of a genetic locus 61. Schaeff er,L.M., McCormack,F.X., Wu,H., and Weiss,A. required for lipopolysaccharide biosynthesis in A. (2004). Bordetella pertussis lipopolysaccharide Bordetella pertussis. Mol Microbiol 19, 37-52. resists the bactericidal eff ects of pulmonary surfactant protein A. J. Immunol. 173, 1959-1965. 51. Preston,A., Maxim,E., Toland,E., Pishko,E.J., Harvill,E. T., Caroff ,M., and Maskell,D.J. (2003). Bordetella 62. Edwards,J.A., Groathouse,N.A., and Boitano,S. bronchiseptica PagP is a Bvg-regulated lipid A (2005). Bordetella bronchiseptica Adherence to palmitoyl transferase that is required for persistent Cilia Is Mediated by Multiple Adhesin Factors and colonization of the mouse respiratory tract. Mol Blocked by Surfactant Protein A. Infect. Immun. 73, Microbiol 48, 725-736. 3618-3626.

52. Bishop,R.E., Gibbons,H.S., Guina,T., Trent,M. 63. Parkhill,J., Sebaihia,M., Preston,A., Murphy,L.D., S., Miller,S.I., and Raetz,C.R. (2000). Transfer of Thomson,N., Harris,D.E., Holden,M.T., Churcher,C. palmitate from phospholipids to lipid A in outer M., Bentley,S.D., Mungall,K.L., Cerdeno-Tarraga,A. membranes of gram-negative bacteria. EMBO J. M., Temple,L., James,K., Harris,B., Quail,M.A., 19, 5071-5080. Achtman,M., Atkin,R., Baker,S., Basham,D., Bason,N., Cherevach,I., Chillingworth,T., Collins,M., Cronin,A., 53. Bishop,R.E. (2005). The lipid A palmitoyltransferase Davis,P., Doggett,J., Feltwell,T., Goble,A., Hamlin,N., PagP: molecular mechanisms and role in bacterial Hauser,H., Holroyd,S., Jagels,K., Leather,S., pathogenesis. Molecular Microbiology 57, 900- Moule,S., Norberczak,H., O’Neil,S., Ormond,D., 912. Price,C., Rabbinowitsch,E., Rutter,S., Sanders,M., 54. Burns,V.C., Pishko,E.J., Preston,A., Maskell,D.J., and Saunders,D., Seeger,K., Sharp,S., Simmonds,M., Harvill,E.T. (2003). Role of Bordetella o antigen in Skelton,J., Squares,R., Squares,S., Stevens,K., respiratory tract infection. Infect. Immun. 71, 86- Unwin,L., Whitehead,S., Barrell,B.G., Maskell,D. 94. J. (2003). Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella 55. Pishko,E.J., Betting,D.J., Hutter,C.S., and Harvill,E. parapertussis and Bordetella bronchiseptica. Nat T. (2003). Bordetella pertussis Acquires Resistance Genet. 35, 32-40. to Complement-Mediated Killing In Vivo. Infect. Immun. 71, 4936-4942. 64. Henderson,I.R. and Nataro,J.P. (2001). Virulence functions of autotransporter proteins. Infect. 56. Rambow,A.A., Fernandez,R.C., and Weiss,A.A. Immun. 69, 1231-1243. (1998). Characterization of BrkA Expression in Bordetella bronchiseptica. Infect. Immun. 66, 65. Li,J., Fairweather,N.F., Novotny,P., Dougan,G., and 3978. Charles,I.G. (1992). Cloning, nucleotide sequence and heterologous expression of the protective 57. West,N.P. et al. (2005). Optimization of virulence outer-membrane protein P.68 pertactin from functions through glucosylation of Shigella LPS. Bordetella bronchiseptica. J. Gen. Microbiol. 138 (

25 Chapter 

Pt 8), 1697-1705. Isaacs,N.W. (1996). Structure of Bordetella pertussis virulence factor P.69 pertactin. Nature 381, 90-92. 1 66. Charles,I.G., Dougan,G., Pickard,D., Chatfi eld,S., Smith,M., Novotny,P., Morrissey,P., and 76. Emsley,P., McDermott,G., Charles,I.G., Fairweather,N. Fairweather,N.F. (1989). Molecular cloning and F., and Isaacs,N.W. (1994). Crystallographic characterization of protective outer membrane characterization of pertactin, a membrane- protein P.69 from Bordetella pertussis. Proc. Natl. associated protein from Bordetella pertussis. J. Acad. Sci. U. S. A 86, 3554-3558. Mol. Biol. 235, 772-773.

67. Li,L.J., Dougan,G., Novotny,P., and Charles,I.G. 77. Hijnen,M., Mooi,F.R., van Gageldonk,P.G., (1991). P.70 pertactin, an outer-membrane protein Hoogerhout,P., King,A.J., and Berbers,G.A. (2004). from Bordetella parapertussis: cloning, nucleotide Epitope structure of the Bordetella pertussis sequence and surface expression in Escherichia protein P.69 pertactin, a major vaccine component coli. Mol. Microbiol. 5, 409-417. and protective antigen. Infect. Immun. 72, 3716- 3723. 68. Leininger,E., Roberts,M., Kenimer,J.G., Charles,I. G., Fairweather,N., Novotny,P., and Brennan,M. 78. van Loo,I.H.M., Heuvelman,K.J., King,A.J., and J. (1991). Pertactin, an Arg-Gly-Asp-containing Mooi,F.R. (2002). Multilocus Sequence Typing of Bordetella pertussis surface protein that promotes Bordetella pertussis Based on Surface Protein adherence of mammalian cells. Proc. Natl. Acad. Genes. J. Clin. Microbiol. 40, 1994. Sci. U. S. A. 88, 345-349. 79. Locht,C. and Keith,J.M. (1986). Pertussis toxin gene: 69. Mooi,F.R., van Oirschot,H., Heuvelman,K., van der nucleotide sequence and genetic organization. Heide,H.G., Gaastra,W., and Willems,R.J. (1998). Science 232, 1258-1264. Polymorphism in the Bordetella pertussis virulence 80. Nicosia,A., Perugini,M., Franzini,C., Casagli,M.C., factors P.69/pertactin and pertussis toxin in The Borri,M.G., Antoni,G., Almoni,M., Neri,P., Ratti,G., Netherlands: temporal trends and evidence for and Rappuoli,R. (1986). Cloning and sequencing vaccine-driven evolution. Infect. Immun. 66, 670- of the pertussis toxin genes: operon structure and 675. gene duplication. Proc. Natl. Acad. Sci. U. S. A 83, 70. Boursaux-Eude,C. and Guiso,N. (2000). 4631-4635. Polymorphism of Repeated Regions of Pertactin 81. Weiss,A.A., Johnson,F.D., and Burns,D.L. (1993). in Bordetella pertussis, Bordetella parapertussis, Molecular characterization of an operon required and Bordetella bronchiseptica. Infect. Immun. 68, for pertussis toxin secretion. Proc. Natl. Acad. Sci. 4815-4817. U. S. A 90, 2970-2974. 71. Cherry,J.D., Gornbein,J., Heininger,U., and Stehr,K. 82. Farizo,K.M., Cafarella,T.G., and Burns,D.L. (1996). (1998). A search for serologic correlates of Evidence for a ninth gene, ptlI, in the locus immunity to Bordetella pertussis cough illnesses. encoding the pertussis toxin secretion system of Vaccine 16, 1901-1906. Bordetella pertussis and formation of a PtlI-PtlF 72. Kerr,J.R. and Matthews,R.C. (2000). Bordetella complex. J. Biol. Chem. 271, 31643-31649. pertussis infection: pathogenesis, diagnosis, 83. Hausman,S.Z., Cherry,J.D., Heininger,U., Wirsing management, and the role of protective immunity. Von Konig,C.H., and Burns,D.L. (1996). Analysis Eur. J. Clin. Microbiol Infect. Dis. 19, 77-88. of proteins encoded by the ptx and ptl genes 73. King,A.J., Berbers,G., van Oirschot,H.F., of Bordetella bronchiseptica and Bordetella Hoogerhout,P., Knipping,K., and Mooi,F.R. parapertussis. Infect. Immun. 64, 4020-4026. (2001). Role of the polymorphic region 1 of the 84. Reisine,T. (1990). Pertussis toxin in the analysis of Bordetella pertussis protein pertactin in immunity. receptor mechanisms. Biochem. Pharmacol. 39, Microbiology 147, 2885-2895. 1499-1504. 74. Storsaeter,J., Hallander,H.O., Gustafsson,L., and 85. Wong,W.S. and Rosoff ,P.M. (1996). Pharmacology Olin,P. (1998). Levels of anti-pertussis antibodies of pertussis toxin B-oligomer. Can. J. Physiol related to protection after household exposure to Pharmacol. 74, 559-564. Bordetella pertussis. Vaccine 16, 1907-1916. 86. Morse,S.I. and Morse,J.H. (1976). Isolation and 75. Emsley,P., Charles,I.G., Fairweather,N.F., and properties of the leukocytosis- and lymphocytosis-

26 General Introduction

promoting factor of Bordetella pertussis. J. Exp. Intracellular localization of the dermonecrotic Med. 143, 1483-1502. toxin of Bordetella pertussis. Infect. Immun. 25, 1 896-901. 87. Munoz,J.J., Arai,H., Bergman,R.K., and Sadowski,P. L. (1981). Biological activities of crystalline 99. Nakai,T., Sawata,A., and Kume,K. (1985). Intracellular pertussigen from Bordetella pertussis. Infect. locations of dermonecrotic toxins in Pasteurella Immun. 33, 820-826. multocida and in Bordetella bronchiseptica. Am. J. Vet. Res. 46, 870-874. 88. Pittman,M. (1979). Pertussis toxin: the cause of the harmful eff ects and prolonged immunity of 100. Kashimoto,T., Katahira,J., Cornejo,W.R., Masuda,M., whooping cough. A hypothesis. Rev. Infect. Dis. 1, Fukuoh,A., Matsuzawa,T., Ohnishi,T., and 401-412. Horiguchi,Y. (1999). Identifi cation of functional domains of Bordetella dermonecrotizing toxin. 89. Carbonetti,N.H., Artamonova,G.V., Andreasen,C., Infect. Immun. 67, 3727-3732. Dudley,E., Mays,R.M., and Worthington,Z.E. (2004). Suppression of Serum Antibody Responses by 101. Hoff mann,C. and Schmidt,G. (2004). CNF and DNT. Pertussis Toxin after Respiratory Tract Colonization Rev. Physiol Biochem. Pharmacol. by Bordetella pertussis and Identifi cation of an 102. Brockmeier,S.L., Register,K.B., Magyar,T., Lax,A.J., Immunodominant Lipoprotein. Infect. Immun. 72, Pullinger,G.D., and Kunkle,R.A. (2002). Role of the 3350-3358. Dermonecrotic Toxin of Bordetella bronchiseptica 90. Mielcarek,N., Riveau,G., Remoue,F., Antoine,R., in the Pathogenesis of Respiratory Disease in Capron,A., and Locht,C. (1998). Homologous and Swine. Infect. Immun. 70, 481-490. heterologous protection after single intranasal 103. Weiss,A.A. and Goodwin,M.S. (1989). Lethal administration of live attenuated recombinant infection by Bordetella pertussis mutants in the Bordetella pertussis. Nat. Biotechnol. 16, 454-457. infant mouse model. Infect. Immun. 57, 3757- 91. Relman,D., Tuomanen,E., Falkow,S., Golenbock,D. 3764. T., Saukkonen,K., and Wright,S.D. (1990). 104. Schaible,U.E. and Kaufmann,S.H. (2004). Iron and Recognition of a bacterial adhesion by an integrin: microbial infection. Nat Rev. Microbiol 2, 946-953. macrophage CR3 (alpha M beta 2, CD11b/CD18) binds fi lamentous hemagglutinin of Bordetella 105. Moore,C.H., Foster,L.A., Gerbig,D.G., Jr., Dyer,D.W., pertussis. Cell 61, 1375-1382. and Gibson,B.W. (1995). Identifi cation of alcaligin as the siderophore produced by Bordetella 92. Tuomanen,E. and Weiss,A. (1985). Characterization pertussis and B. bronchiseptica. J Bacteriol. 177, of two adhesins of Bordetella pertussis for human 1116-1118. ciliated respiratory-epithelial cells. J. Infect. Dis. 152, 118-125. 106. Vanderpool,C.K. and Armstrong,S.K. (2004). Integration of environmental signals controls 93. Bordet,J. and Gengou,O. (1909). L’endotoxine expression of Bordetella heme utilization genes. J coquelucheuse. Ann. Inst. Pasteur 23, 415-419. Bacteriol. 186, 938-948. 94. Walker,K.E. and Weiss,A.A. (1994). Characterization 107. Kang,H.Y., Brickman,T.J., Beaumont,F.C., and of the dermonecrotic toxin in members of the Armstrong,S.K. (1996). Identifi cation and genus Bordetella. Infect. Immun. 62, 3817-3828. characterization of iron-regulated Bordetella 95. Livey,I. and Wardlaw,A.C. (1984). Production and pertussis alcaligin siderophore biosynthesis genes. properties of Bordetella pertussis heat-labile toxin. J Bacteriol. 178, 4877-4884. J. Med. Microbiol. 17, 91-103. 108. Beaumont,F.C., Kang,H.Y., Brickman,T.J., and 96. Parton,R. (1985). Eff ect of prednisolone on the Armstrong,S.K. (1998). Identifi cation and toxicity of Bordetella pertussis for mice. J. Med. characterization of alcR, a gene encoding an AraC- Microbiol. 19, 391-400. like regulator of alcaligin siderophore biosynthesis and transport in Bordetella pertussis and Bordetella 97. Iida,T. and Okonogi,T. (1971). Lienotoxicity of bronchiseptica. J. Bacteriol. 180, 862-870. Bordetella pertussis in mice. J. Med. Microbiol. 4, 51-61. 109. Pradel,E., Guiso,N., and Locht,C. (1998). Identifi cation of AlcR, an AraC-type regulator 98. Cowell,J.L., Hewlett,E.L., and Manclark,C.R. (1979). of alcaligin siderophore synthesis in Bordetella

27 Chapter 

bronchiseptica and Bordetella pertussis. J. R. (1998). Simple sequence repeats in the 1 Bacteriol. 180, 871-880. Helicobacter pylori genome. Mol. Microbiol. 27, 1091-1098. 110. Brickman,T.J. and Armstrong,S.K. (2005). Bordetella AlcS Transporter Functions in Alcaligin Siderophore 123. Tettelin,H. et al. (2000). Complete genome Export and Is Central to Inducer Sensing in Positive sequence of Neisseria meningitidis serogroup B Regulation of Alcaligin System Gene Expression. J. strain MC58. Science 287, 1809-1815. Bacteriol. 187, 3650-3661. 124. Dobrindt,U., Hochhut,B., Hentschel,U., and 111. Brickman,T.J. and Armstrong,S.K. (1999). Essential Hacker,J. (2004). Genomic islands in pathogenic role of the iron-regulated outer membrane and environmental microorganisms. Nat. Rev. receptor FauA in alcaligin siderophore-mediated Microbiol. 2, 414-424. iron uptake in Bordetella species. J. Bacteriol. 181, 125. Hacker,J., Blum-Oehler,G., Muhldorfer,I., and 5958-5966. Tschape,H. (1997). Pathogenicity islands of 112. Brickman,T.J., Kang,H.Y., and Armstrong,S.K. (2001). virulent bacteria: structure, function and impact Transcriptional activation of Bordetella alcaligin on microbial evolution. Mol Microbiol 23, 1089- siderophore genes requires the AlcR regulator 1097. with alcaligin as inducer. J. Bacteriol. 183, 483-489. 126. Hacker,J. and Kaper,J.B. (2000). Pathogenicity 113. Register,K.B., Ducey,T.F., Brockmeier,S.L., and islands and the evolution of microbes. Annu. Rev. Dyer,D.W. (2001). Reduced virulence of a Bordetella Microbiol 54, 641-679. bronchiseptica siderophore mutant in neonatal 127. Schmidt,H. and Hensel,M. (2004). Pathogenicity swine. Infect. Immun. 69, 2137-2143. Islands in Bacterial Pathogenesis. Clin. Microbiol 114. Margulies,M. et al. (2005). Genome sequencing in Rev. 17, 14-56. microfabricated high-density picolitre reactors. 128. Achtman,M., Zurth,K., Morelli,G., Torrea,G., Nature. Guiyoule,A., and Carniel,E. (1999). Yersinia pestis, 115. Wren,B.W. (2000). Microbial genome analysis: the cause of plague, is a recently emerged clone insights into virulence, host adaptation and of Yersinia pseudotuberculosis. PNAS 96, 14043- evolution. Nat. Rev. Genet. 1, 30-39. 14048.

116. Heidelberg,J.F. et al. (2000). DNA sequence of both 129. Achtman,M. et al. (2004). Microevolution and chromosomes of the cholera pathogen Vibrio history of the plague bacillus, Yersinia pestis. PNAS cholerae. Nature 406, 477-483. 101, 17837-17842.

117. Fraser,C.M. et al. (1997). Genomic sequence of a 130. Chain,P.S.G. et al. (2004). Insights into the evolution Lyme disease spirochaete, Borrelia burgdorferi. of Yersinia pestis through whole-genome Nature 390, 580-586. comparison with Yersinia pseudotuberculosis. PNAS 0404012101. 118. Ochman,H., Lawrence,J.G., and Groisman,E.A. (2000). Lateral gene transfer and the nature of 131. Parkhill,J. et al. (2001). Genome sequence of bacterial innovation. Nature 405, 299-304. Yersinia pestis, the causative agent of plague. Nature 413, 523-527. 119. Tomb,J.F. et al. (1997). The complete genome sequence of the gastric pathogen Helicobacter 132. Andersson,S.G., Zomorodipour,A., Andersson,J. pylori. Nature 388, 539-547. O., Sicheritz-Ponten,T., Alsmark,U.C., Podowski,R. M., Naslund,A.K., Eriksson,A.S., Winkler,H.H., 120. Alm,R.A. et al. (1999). Genomic-sequence and Kurland,C.G. (1998). The genome sequence comparison of two unrelated isolates of the human of Rickettsia prowazekii and the origin of gastric pathogen Helicobacter pylori. Nature 397, mitochondria. Nature 396, 133-140. 176-180. 133. Fraser,C.M. et al. (1998). Complete genome 121. Saunders,N.J., Boonmee,P., Peden,J.F., and Jarvis,S. sequence of Treponema pallidum, the syphilis A. (2005). Inter-species horizontal transfer resulting spirochete. Science 281, 375-388. in core-genome and niche-adaptive variation within Helicobacter pylori. BMC. Genomics 6, 9. 134. Brosch,R., Gordon,S.V., Eiglmeier,K., Garnier,T., and Cole,S.T. (2000). Comparative genomics of the 122. Saunders,N.J., Peden,J.F., Hood,D.W., and Moxon,E.

28 General Introduction

leprosy and tubercle bacilli. Res. Microbiol. 151, and Gross,R. (2001). Evolutionary trends in the 135-142. genus Bordetella. Microbes and Infection 3, 61-72. 1 135. Brosch,R., Pym,A.S., Gordon,S.V., and Cole,S. T. (2001). The evolution of mycobacterial pathogenicity: clues from comparative genomics. Trends Microbiol. 2001. Sep. ;9. (9. ):452. -8. 9, 452- 458.

136. Musser,J.M., Hewlett,E.L., Peppler,M.S., and Selander,R.K. (1986). Genetic diversity and relationships in populations of Bordetella spp. J. Bacteriol. 166, 230-237.

137. Musser,J.M., Bemis,D.A., Ishikawa,H., and Selander,R. K. (1987). Clonal diversity and host distribution in Bordetella bronchiseptica. J. Bacteriol. 169, 2793- 2803.

138. Arico,B., Gross,R., Smida,J., and Rappuoli,R. (1987). Evolutionary relationships in the genus Bordetella. Mol. Microbiol. 1, 301-308.

139. Muller,M. and Hildebrandt,A. (1993). Nucleotide sequences of the 23S rRNA genes from Bordetella pertussis, B.parapertussis, B.bronchiseptica and B.avium, and their implications for phylogenetic analysis. Nucleic Acids Res. 21, 3320.

140. van Loo,I.H., van der Heide,H.G., Nagelkerke,N. J., Verhoef,J., and Mooi,F.R. (1999). Temporal trends in the population structure of Bordetella pertussis during 1949-1996 in a highly vaccinated population. J. Infect. Dis. 179, 915-923.

141. Cummings,C.A., Brinig,M.M., Lepp,P.W., Van De,P. S., and Relman,D.A. (2004). Bordetella species are distinguished by patterns of substantial gene loss and host adaptation. J Bacteriol. 186, 1484-1492.

142. Reischl,U., Lehn,N., Sanden,G.N., and Loeff elholz,M. J. (2001). Real-Time PCR Assay Targeting IS481 of Bordetella pertussis and Molecular Basis for Detecting Bordetella holmesii. J. Clin. Microbiol. 39, 1963-1966.

143. Gerlach,G., Janzen,S., Beier,D., and Gross,R. (2004). Functional characterization of the BvgAS two-component system of Bordetella holmesii. Microbiology 150, 3715-3729.

144. Njamkepo,E., Delisle,F., Hagege,I., Gerbaud,G., and Guiso,N. (2000). Bordetella holmesii isolated from a patient with sickle cell anemia: analysis and comparison with other Bordetella holmesii isolates. Clin. Microbiol. Infect. 2000. Mar. ;6. (3):131. -6. 6, 131-136.

145. Gerlach,G., von Wintzingerode,F., Middendorf,B.,

29 30

Chapter Chapter 2

Genetic Relationships of the Mammalian Bordetellae

Dimitri A. Diavatopoulos1,2, Craig A. Cummings3,5, Leo M. Schouls1, Mary M. Brinig3,5, David A. Relman3,4,5, Frits R. Mooi1,2

1Laboratory for Vaccine-Preventable Diseases, National Institute of Public Health and the Envi- ronment, Bilthoven, Th e Netherlands; 2Eijkman Winkler Institute, University Medical Center, Utrecht, Th e Netherlands; 3Departments of Microbiology and Immunology; and of 4Medicine, Stanford University School of Medicine, Stanford, California 94305, USA; 5VA Palo Alto Health Care System, Palo Alto, California 94304, USA

Chapter 2&3, PLoS Pathogens, 2005 Chapter 

abstract

Bordetella pertussis, Bordetella bronchiseptica, Bordetella parapertussishu and Bordetella parapertussis are closely related respiratory pathogens that infect mammalian species. 2 ov B. pertussis and B. parapertussishu are exclusively human pathogens and cause whooping cough or pertussis, a disease which kills 300,000 persons annually, and is reemerging despite vaccination. Although it most often infects animals, infrequently B. bronchiseptica is isolated from humans, and these infections are thought to be zoonotic. B. pertussis

and B. parapertussishu are assumed to have evolved from a B. bronchiseptica-like ancestor independently. To determine the phylogenetic relationships among these species, a total of seven housekeeping genes and one virulence gene (coding for pertactin) were sequenced. Further, the distribution of four insertion sequence elements was determined, using a collection of 132 strains. Th is approach distinguished four complexes, representing B.

pertussis, B. parapertussishu and two distinct B. bronchiseptica subpopulations, designated complexes I and IV. Of the two B. bronchiseptica complexes, complex IV was more closely related to B. pertussis. Of interest, while only 32% of the complex I strains were isolated from humans, 80% of the complex IV strains were human isolates. Th us, complex IV strains may comprise a human-associated lineage of B. bronchiseptica from which B. pertussis evolved. Th ese ﬁ ndings will facilitate the study of pathogen host-adaptation. Our results shed light on the origins of the disease pertussis and suggest that the association of B. pertussis with humans may be more ancient than previously assumed.

32 Multilocus Sequence Typing introduction Members of the genus Bordetella are predominantly pathogenic species, and three of these are presumed to be exclusively respiratory pathogens of mammalian hosts: Bordetella 2 bronchiseptica, Bordetella pertussis and Bordetella parapertussis (henceforth referred to as the mammalian bordetellae). B. bronchiseptica causes chronic and often asymptomatic respiratory tract infections in a wide variety of mammals. It is only sporadically isolated from humans 1,2. Th ese are usually immunocompromised individuals, and human infections have been considered to be zoonotic 3. B. parapertussis consists of two distinct lineages found in 4 humans and sheep (B. parapertussishu and B. parapertussisov, respectively) . B. pertussis and

B. parapertussishu have been isolated exclusively from humans and cause acute and transient infections, designated whooping cough or pertussis. Whooping cough is especially severe in young, unvaccinated children and has reemerged in recent years in vaccinated populations 5-7. Previous research, based on orthologous gene pairs between genomes, indicated that B. pertussis and B. parapertussishu independently evolved from a B. bronchiseptica-like ancestor about 0.7-3.5 and 0.27-1.4 million years ago, respectively 8,9. Despite their diff erent host tropisms, the mammalian bordetellae are very closely related 9,10, which makes the mammalian bordetellae attractive candidates to study host-adaptation. Such studies are facilitated by the availability of the genome sequences of B. bronchiseptica, B. pertussis and 8 B. parapertussishu . Th us far, only a single representative of each species has been sequenced, and it is important to determine their relationships to the Bordetella population on the whole. To that purpose, we used a combination of sequencing of housekeeping and virulence genes and the distribution of several insertion sequence elements (ISEs) to characterize 132 mammalian Bordetella strains with diverse host distributions. Th ese studies identifi ed four complexes, representing B. pertussis, B. parapertussishu and two distinct B. bronchiseptica lineages, designated complex I-IV. Th e two B. bronchiseptica complexes, designated complex I and IV, showed diff erent host-distributions and were isolated mainly from animals hosts and humans, respectively. Of these two complexes, B. bronchiseptica complex IV was more closely related to B. pertussis and may comprise strains that are human-adapted. Our data suggest that B. parapertussishu evolved from an animal-associated lineage of B. bronchiseptica, while B. pertussis evolved from a distinct B. bronchiseptica lineage. Members of this newly- identifi ed B. bronchiseptica lineage continue to circulate today and cause human disease. Th is may suggest that the association of B. pertussis to humans evolved before its recent clonal expansion.

33 Chapter 

experimental procedures

bacterial strains 2 A total of 132 Bordetella isolates were used in this study: 91 B. bronchiseptica; 9 B.

parapertussishu; 3 B. parapertussisov and 29 B. pertussis isolates (a table with strain information is available from the authors). Th e three strains from which the genome sequence has been determined, B. bronchiseptica RB50, B. pertussis Tohama and B. parapertussis 12822 8, were included. Th e collection included clinical isolates from humans and a broad range of animal species. Strains were grown on Bordet Gengou (BD, Franklin Lakes, NJ, USA) agar supplemented with 15% sheep blood at 37°C for two to ﬁ ve days. Chromosomal DNA was isolated using the Wizard Genomic DNA Puriﬁ cation Kit (Promega, Madison, WI), according to the manufacturers’ protocol for Gram-negative bacteria.

dna sequencing Th e nucleotide sequences were determined for internal regions of seven housekeeping genes for all strains (http://pubmlst.org/bordetella/ 11). Th e nucleotide sequence of the prn region encoding the extracellular domain of the surface-associated autotransporter pertactin, P.69, was determined for 116 strains, with the exclusion of the repeat regions 1 and 2 12. Th ese regions are comprised of amino acids repeats and are highly polymorphic due to insertion or deletion of the repeat unit. Primer characteristics are listed in Table 1, and can also be found at http://pubmlst.org/bordetella/.

detection of ises Th e distribution of IS481, IS1001, IS1002 and IS1663 was determined for all strains using PCR ampliﬁ cation. For PCR ampliﬁ cation of IS481, IS1001 and IS1002, primers were used that have been described previously 4,13. Primer characteristics are listed in Table 1.

sequence data analysis Analysis of nucleotide sequence data was performed using Bionumerics software package version 4.0 beta 4 (Applied Maths, Sint-Martens-Latem, Belgium). Th e Bordetella MLST database can be accessed at http://pubmlst.org/bordetella/ 11. Th e nucleotide sequences of pertactin have been deposited in GenBank under accession numbers DQ141700- DQ141711 and DQ141713-DQ141816. For each locus in the MLST analysis, the allele sequences for all strains were trimmed to a uniform length, and an allele number was assigned to each unique allele sequence. Th e combination of the allele numbers at the seven loci defi nes the sequence type (ST) or allelic profi le of each strain. Construction of trees based on allelic profi les may not accurately refl ect the true genetic distance because both single and multiple nucleotide polymorphisms

34 Multilocus Sequence Typing

Table 1. Primer characteristics for the genes used in multilocus sequence typing, pertactin sequencing and in the detection of the insertion sequence elements Gene1 Product1 Primer Name Sequence (5’-3’) adk adenylate kinase Adk-F AGCCGCCTTTCTCACCCAACACT 2 Adk-R TGGGCCCAGGACGAGTAGT fumC fumarate hydratase class II FumC-F CGTGAACCGGGGCCAGTCGTC FumC-R GGCCAGCCAGCGCACATCGTT icd isocitrate dehydrogenase Icd-F CTGGTCCACAAGGGCAACAT Icd-R ACACCTGGGTGGCGCCTTC glyA serine hydroxymethyltransferase GlyA-F CAACCAGGGCGTGTACATGGC GlyA-R CCGCGATGACGTGCATCAG tyrB aromatic amino-acid aminotransferase TyrB-F CGAGACCTACGCTTATTACGAT TyrB-R TGCCGGCCAGTTCATTTT pepA cytosol aminopeptidase PepA-F CGCCCCAGGTTGAAGAAAATCGTC PepA-R ATCAGGCCCACCACATCCAG pgm phoshoglucomutase Pgm-F CGCCCATGTCACCAGCACCGA Pgm-R CGCCGTCTATCGTAACCAG IS481 transposase IS481-F GGGGTCACCGCGCCGACTGT IS481-R GGGCCTGATGCTCGTAGCGC IS1001 transposase IS1001-F CGCCGCTTGATGACCTTGATA IS1001-R CACCGCCTACGAGTTGGAGAT IS1002 transposase IS1002-F TCCCAGCTCCACGCACACCG IS1002-R AACAACCATAAGCATGCGCG IS1663 transposase IS1663-F2 GGGTCTGTATCACGAGCAAGCGG IS1663-R CTTTGCGATTGAGCTCACGCAAC IS1663-F22 GCGAGACACTGGACGGTATCG IS1663-R2 GGGGACAGATACCGTCTTGGC IS1663-SPF1 CAGTTCAGCCCCTCGGCGC IS1663-SPR1 CTTTGCGATTGAGCTCACGCA prn pertactin Prn-SPF1 TCCCTGTTCCATCGCGGTG Prn-SPR3 GTTGGCGGCCAATCGATAGC Prn-SPF2 ATCGCGCTCTATGTGGCCG Prn-SPR1 CCTGAGCCTGGAGACTGGCAC Prn-SPF3 CACCGCACGGCAATGTCATC Prn-SPF4 GGCGACCTTTACCCTTGCCAA Prn-SPR2 CAGCGTCGCGTCCAGGTAGA Prn-SPR4 GCAAGGTGATCGACAGGGGC Prn-SPR5 TGGACCGTGACATTGGCGC

¹ Gene name and product as annotated by the Sanger Centre sequencing team ² IS1663-F and R were used only for detection of IS1663; IS1663-F2 and R2 were used for determination of the nucleotide sequence are given equal weight. Consequently, the degree of sequence diff erence between two alleles is not quantitatively refl ected in the MLST profi le. Conversely, tree construction based on concatenated allele sequences does not take into account the introduction of clustered multiple base substitutions due to a single recombinational event. As a result, trees based on MLST sequences often contain long braches, incorrectly suggesting a large genetic distance. Th erefore, we used a method designated as split-MLST, in which each locus is split into a user-defi ned number of equally sized sub-loci (D. A. Diavatopoulos, P. Vauterin, L. Vauterin, F.R. Mooi & L.M. Schouls, unpublished data). Using this method, the sensitivity of categorical clustering could be increased, without the perturbing eff ect of recombination.

35 Chapter 

Th e topology of the tree appeared to vary if the number of sub-loci per MLST locus was lower than fi ve. However, above the value four, increasing the number of sub-loci had no signifi cant eff ect on the topology of the tree, and we therefore selected the lowest possible 2 split-value, fi ve, resulting in a total of 35 sub-loci. For sequencing of the pertactin gene, the nucleotides encoding the extracellular domain of pertactin were determined (nucleotides 103-2132 of BP1054, 112-2150 of BB1366, and 103-2168 of BPP1150, the pertactin genes of B. pertussis Tohama, B. bronchiseptica RB50 and of B. parapertussis 12822, respectively). Based on these sequences, a UPGMA tree was constructed, with the exclusion of the polymorphic regions 1 and 2 (region 1 comprises nucleotides 825-871 of BP1054, 834-865 of BB1366 and 825-856 of BPP1150, respectively; region 2 comprises nucleotides 1754-1774 of BP1054, 1752-1790 of BB1366 and 1743-1808 of BPP1150, respectively).

genetic diversity of the complexes Th e genetic diversity for each complex was calculated using the Shannon-Weiner index of diversity (H) using the following formula:

H = −∑ Pi ⋅ln Pi

th 14 where Pi is the frequency of the i type .

calculation of divergence times For estimation of divergence times between complexes, we calculated the pair wise mean 15 distance (Ks) between alleles using DNASP 4.00 . Th e divergence time was calculated using the following formula: K Age = s r

where Ks is the number of synonymous substitutions per synonymous site and r is the molecular clock rate of Escherichia coli as determined by Whittam 16 or by Guttman and Dykhuizen 17. We used these two rates to calculate a range of divergence times. Th e divergence time was ﬁ rst calculated for each combination of STs between complexes, and from these the averaged age between complexes was calculated.

36 Multilocus Sequence Typing results population genetic relationships of the mammalian bordetellae 2 To determine the relationships between the mammalian bordetellae, we determined partial sequences of seven housekeeping genes from 132 strains (http://pubmlst.org/bordetella). We observed 32 sequence types (STs) among the 132 Bordetella isolates. Allele segments were divided into fi ve equally sized sub-loci and a minimum spanning tree algorithm was used to cluster the sub-loci 18. Complexes were defi ned as groups of strains diff ering at fewer than fi ve of 35 sub-loci with a minimum of 2 STs per complex. Using this criterion, strains could be assigned to one of four complexes, designated complexes I-IV (Figure 1).

B. bronchiseptica 31 complex I 3 5 27 2 2 3 1

IS1001 (100%) 4 1 B. pertussis 1 complex II 4 24 16 IS481 (100%) IS1002 (100%) 10 IS1663 (100%) 12 9 5 1 34 5 9 22 1 5 2 35 14 3 1 IS481 (40%) 3 17 3 25 10 4 3 2 6 29 15 2 1 18 1 5 3 1 1 32 28 33 3 8 26 21 1 7 B. bronchiseptica 7 23 complex IV IS1663 (80%) 19 1 IS1001 (83%) 30

B. parapertussis human origin hu animal origin complex III unknown IS1001 (100%) STs containing sequenced strains IS1002 (100%) (Tohama, RB50 or 12822)

Figure 1. Minimum spanning tree of B. bronchiseptica, B. pertussis and B. parapertussis. The tree was based on the sequence of seven housekeeping genes. Individual genes were split into fi ve sub-loci, and a categorical clustering was performed. In the minimum spanning tree, sequence types (STs) sharing the highest number of single locus variants were connected fi rst. Each circle represents an ST, the size of which is related to the number of isolates belonging to that particular ST. Colors within circles indicate host-distribution. The numbers between connected STs represent the number of diff erent sub-loci between those STs. The clonal complexes (I, II, III and IV) are indi-

cated by colored strips between connected STs. ST16 (B. bronchiseptica complex I) harbors the B. parapertussisov strains. STs containing strains of which the genome has been sequenced (B. pertussis Tohama, B. parapertussis 12822 or B. bronchiseptica RB50) are indicated by a thickset, dashed line. The distribution of the insertion sequence elements IS481, IS1001, IS1002 and IS1663 is shown in boxes; numbers between parentheses indicate the percentage of strains that contained the ISE as determined by PCR amplifi cation. A color version of this fi gure is available in the appendix.

37 Chapter 

Complexes II and III contained the B. pertussis and B. parapertussishu isolates, respectively. Both of these complexes showed very limited genetic diversity (H=0.65 and 0.35, respectively), as described previously 9,10. B. bronchiseptica was divided into two distinct 2 populations, designated complexes I and IV, respectively. Th e genetic diversity of these two complexes (H=2.16 and 2.45, respectively) was much higher than of complexes II and III. Complex I contained the majority of the B. bronchiseptica strains (76 of 91 strains),

including the sequenced RB50 strain. In addition, it contained the B. parapertussisov isolates in the study population (ST16). B. bronchiseptica complex IV was more closely related to B. pertussis than was B. bronchiseptica complex I. Furthermore, the host species associations of the two complexes were quite distinct. Of the B. bronchiseptica complex IV isolates, 80% were isolated from humans, while this was the case for only 32% of the complex I isolates. It should be noted that human B. bronchiseptica isolates were overrepresented in our strain collection, in comparison with their occurrence in naturally-occurring populations of B. bronchiseptica strains. However, the human complex IV isolates originated from diﬀ erent continents, comprising North America, South America, and Europe. Previously, phylogenetic analysis based on CGH suggested the existence of a distinct B. bronchiseptica lineage that was closely related to B. pertussis 19.

sequencing of the pertactin gene Th e relationship of the mammalian bordetellae inferred by housekeeping genes was conﬁ rmed by a tree based on the pertactin gene, which codes for a surface-associated virulence factor involved in adherence 20,21. A UPGMA tree was constructed from the aligned pertactin 95% 96 97 98 99 100% B. bronchiseptica Figure 2. UPGMA tree based on the analysis of complex IV 30% animal the pertactin gene of Bordetella isolates used in n=10 70% human the MLST analysis. The DNA segment coding for 100 the extracellular domain of pertactin (P.69) was B. pertussis used for analysis, with the exclusion of the repeat complex II regions 1 and 2. Bootstrap values are shown for n=26 100% human the nodes separating the complexes and are based on 500 bootstrap replicates. The scale indicates the genetic distance along the branches. Colors of the branches indicate the four complexes as defi ned by MLST. The number of strains of B. bronchiseptica each branch is shown in boxes, as well as the host complex I distribution. A color version of this fi gure is avail- 65% animal able in the appendix. n=71 31% human 4% unknown

100

B. parapertussishu complex III n=9 100% human

38 Multilocus Sequence Typing sequence data, and the topology of this tree was 99.5 very similar to the MLST tree (Figure 2). B. 98% 98.5 99 100% bronchiseptica strains grouped into two lineages, B0243 corresponding to complex I and IV in the B2490 2 MLST tree. Also, B. bronchiseptica complex IV B2114 B. bronchiseptica B0259 and B. pertussis strains clustered together in one complex IV branch, which was supported by bootstrapping. B2491 B1968 B. parapertussishu comprised a separate branch within a larger cluster that also contained the B2494 BP1717 B. bronchiseptica complex I strains. As was also BP1959 observed in the MLST tree, the B. parapertussis ov BP2721 strains were phylogenetically indistinguishable BP3216 from B. bronchiseptica complex I strains. BP1053

BP0118 distribution of insertion sequence BP3243 elements BP0812 Th e distribution of insertion sequence elements BP1388 (ISEs) has been used to reveal evolutionary BP1914 relationships between the Bordetella population BP3230 9. Towards this end, we screened our strain BP1044 collection for the presence of IS481, IS1001, BP2121 IS1002 and IS1663 using PCR. Th e distribution B. pertussis BP1035 of the ISEs was mapped onto the MST (Figure BP1365 1). IS481 was detected in all B. pertussis strains, BP3603 but in no other species, with the exception of Figure 3. UPGMA tree based on the IS1663 se- two B. bronchiseptica isolates, both from a horse quences of seven B. bronchiseptica complex IV (B1975, B0230, ST6), consistent with previous strains and the B. pertussis Tohama IS1663 or- 9,22 thologues. Bootstrap values are shown for the observations . IS1001 was detected in all nodes separating the complexes, and are based on 1000 bootstrap replicates. Branch lengths B. parapertussishu and B. parapertussisov strains. Additionally, IS1001 was detected in most (21 are shown, the scale indicates the genetic distance along the branches. A color version of this out of 25) B. bronchiseptica strains belonging fi gure is available in the appendix. to ST7 in complex I, but not in other STs, including STs in complex II and IV. IS1002 was detected only in B. pertussis and in B. 4,9 8 parapertussishu strains, conﬁ rming previous observations . IS1663 was detected in all B. pertussis isolates, but also in 10 out of 13 B. bronchiseptica complex IV strains. Th e three complex IV strains in which IS1663 was not detected belonged to STs 18 and 21. To determine if the IS1663 sequences of B. bronchiseptica complex IV and of B. pertussis Tohama were similar, the nucleotide sequence of IS1663 was determined from seven B. bronchiseptica complex IV strains. Figure 3 shows a UPGMA tree constructed from the

39 Chapter 

IS1663 sequences of the complex IV strains and of each IS1663 copy in the genome of B. pertussis Tohama. IS1663 sequences of the complex IV strains showed a sequence similarity of 97% to the IS1663 copies of B. pertussis Tohama. 2 Table 2. Calculated divergence times for combinations of complexes. Complex combination Divergence time (Mya) I & II 0.95 (+/-0.54)1 - 4.72 (+/-2.71) I & IV 1.11 (+/-0.48) - 5.57 (+/-2.4) IV & II 0.32 (+/-0.06) - 2.53 (+/-0.51) I & III 0.70 (+/-0.54) - 3.48 (+/-2.68)

¹ Between parentheses, the standard deviation calculated from the pair wise allele distances between STs of diﬀ erent complexes is shown

divergence times of complexes Under the assumption that the mutation rate in prokaryotes is relatively constant, the time since descent from the last common ancestor (LCA) can be estimated using pair 23,24 16 wise mean allele distances (KS) . Both the clock rates described by Whittam and by Guttman and Dykhuizen 17 were used to estimate a range of divergence times between complexes. Calculations indicated that B. pertussis and B. bronchiseptica complex IV separated approximately 0.3-2.5 million years ago (Mya), which suggests a more recent divergence time than B. pertussis and B. bronchiseptica complex I, estimated at 1.1-5.6 Mya.

B. parapertussishu and B. bronchiseptica complex I diverged between 0.7-3.5 Mya according to our calculations. Th e divergence times of combinations of complexes is shown in Table 2.

40 Multilocus Sequence Typing discussion Although it has long been speculated that B. pertussis evolved from a B. bronchiseptica strain 9,10, a specifi c lineage has not been identifi ed. Here we identify and characterize such a B. 2 bronchiseptica lineage. Analysis of MLST data from the mammalian bordetellae identifi ed four distinct complexes. Complex I and IV comprised B. bronchiseptica strains, while complex

II and III comprised the human pathogens B. pertussis and B. parapertussishu, respectively.

Our results suggest that B. pertussis and B. parapertussishu evolved from complexes I and IV, respectively, indicating that adaptation to humans occurred as two independent events, consistent with previous data 8,9. Th e population structure of the mammalian bordetellae inferred from MLST data largely corresponded with a maximum parsimony phylogeny derived from a previous CGH study 19 , with the exception of the relationship of B. parapertussisov and B. parapertussishu. In the current study, B. parapertussishu and B. parapertussisov are clearly derived from diﬀ erent STs in complex I. Further, in contrast to B. parapertussishu, B. parapertussisov is actually part of B. bronchiseptica complex I. In contrast, CGH analyses suggested a closer relationship between the sheep- and human-derived B. parapertussis lineages. However, this may be an artifact of the CGH analyses due to long-branch attraction in the maximum parsimony tree 25. 9,10,12 Consistent with previous studies , B. pertussis and B. parapertussishu showed a relatively low degree of genetic diversity, suggesting that they evolved recently or encountered a recent evolutionary bottleneck. Of the three B. pertussis STs observed, two were found exclusively before 1960, whereas all modern strains belonged to ST2. Th e temporal shift in B. pertussis STs is consistent with our previous studies on antigenic shifts which show major changes in the B. pertussis population after the introduction of mass vaccination against pertussis in the 1950s and 1960s 12,26. Most human disease, by far, is caused by B. pertussis and we therefore focused on the relationship of B. bronchiseptica complex IV with B. pertussis. B. bronchiseptica complex IV strains were found to be more closely related to B. pertussis than to the complex I B. bronchiseptica strains. A tree based on prn nucleotide sequences also suggested a closer relationship of complex IV strains to B. pertussis than to complex I strains. A number of other features of complex IV strains were consistent with their close relationship with B. pertussis. Most complex IV strains were isolated from humans (80%), while the majority of complex I was of animal origin (68%). Almost all B. bronchiseptica complex IV strains were isolated from patients with whooping cough symptoms. Further, complex IV strains and B. pertussis shared an IS element, IS1663, that was not found outside these two lineages. Th e sharing of an IS element may be explained by either vertical or horizontal transfer. Th e former suggests a common ancestry, while the latter would point to niche sharing of B. pertussis and B. bronchiseptica complex IV. It seems unlikely that the association of complex IV strains is due to a sampling artifact, as the strains analyzed were from widely separated geographic regions, including North America, South America and Europe. Th us,

41 Chapter 

these strains were not epidemiologically related. Th e high frequency of human isolates observed in complex IV may be due to the close interaction of humans with animal hosts in which these strains reside, or to the fact that 2 complex IV strains are better adapted to a human environment than B. bronchiseptica complex I strains. In either case, the B. bronchiseptica complex IV infections of humans would be zoonotic. Another intriguing possibility is that B. bronchiseptica complex IV strains are adapted to the human host and mainly transmitted between humans. Th ree STs (ST12, ST23 and ST27) found in complex I also contained a high percentage of humans strains (55%, 43% and 77%, respectively). All other, non-human isolates of these STs were collected from domesticated pet animals. Th us, both complex I and IV contain B. bronchiseptica strains that are well-adapted to the human host. However, the particular relevance of the human-associated lineage in complex IV appears to be its evolutionary relationship with B. pertussis. Th e origin of the disease whooping cough is still a mystery. Although the disease has very typical symptoms in children and was one of the major causes of child mortality previous to the introduction of vaccination, the ﬁ rst written reference to the disease in Europe is found in 1540 27. The first description of an epidemic, which occurred in Paris, was given by Baillon in 1578 28. Particularly interesting are the observations made by Nils Rosen von Rosenstein in 1766 who wrote 29 “Th e hooping cough never appeared in Europe originally, but was transported thither from other parts of the world by means of merchandise, seamen and animals. Its ﬁ rst appearance in Sweden cannot be determined with any certainty; but in France it began in the year 1414”. In contrast, 16th and 17th century descriptions of the disease and epidemics in Europe are documented frequently in the literature 28. Th e absence of references to pertussis-like symptoms in the ancient literature has been taken as evidence that the association of B. pertussis with humans is of recent origin. We propose that the association of B. pertussis with humans is, in fact, ancient, but that the introduction of B. pertussis into Europe may be more recent. Complex IV strains showed a degree of diversity that was comparable to complex I strains (H = 2.16 and 2.45, respectively), and thus, assuming that complex IV strains are primarily adapted to the human host, this association must be ancient. Parkhill et al. previously estimated the time to the LCA of a B. bronchiseptica complex I strain (RB50) and B. pertussis Tohama to be 0.7-3.5 Mya, based on the mean number of synonymous substitutions per synonymous site of orthologous gene pairs 8. Our data indicate that current B. pertussis strains expanded clonally from the B. pertussis-B. bronchiseptica complex IV LCA 0.32-2.53 Mya, further supporting an ancient association of B. pertussis with humans. However, we cannot rule out the possibility that more recent human-associated ancestors of B. pertussis are extinct or undiscovered. Such recent ancestors would indicate a more recent origin of B. pertussis. Although it is tempting to speculate that the LCA of B. pertussis and B. bronchiseptica

42 Multilocus Sequence Typing complex IV was associated with humans, the possibility remains that this association emerged after the split with B. pertussis. A possible evolutionary scenario (Figure 4) involves the adaptation of an ancestral B. bronchiseptica complex I strain to humans or their hominid ancestors. From this lineage the LCA of B. bronchiseptica complex IV and B. 2

Figure 4. Model of the evo- B. bronchiseptica lution of the mammalian complex I bordetellae. The bar on the left indicates increasing degrees of adaptation to the human host. Arrows indi-

cate descent. Abbreviations: broad host range LCA, last common ancestor. LCA complex IV A color version of this fi gure is available in the appendix.

B. bronchiseptica complex IV

B. pertussis B. parapertussishu

human-adapted complex II complex III pertussis evolved, subsequently giving rise to B. bronchiseptica complex IV and B. pertussis. Recent emergence of a pathogenic clone from a more ancient human-associated progenitor species has been proposed as the mechanism for the origin of Mycobacterium tuberculosis 31. Although previous genetic analysis had suggested that M. tuberculosis emerged as little as 20,000 years ago, phylogenetic analysis of M. tuberculosis and a closely related, but more diverse group of smooth tubercle bacilli indicated that this more broadly defi ned species has been associated with hominids for up to 3 million years. Yersinia pestis, the causative agent of plague, is a clone that evolved from Y. pseudotuberculosis 1,500–20,000 years ago, shortly before the fi rst known pandemics of human plague 30, and its recent origin is further suggested by the complete lack of polymorphism in housekeeping genes 31. Similarly, B. pertussis also shows limited diversity. However, in contrast to Y. pestis which reveals absolutely no polymorphisms in housekeeping genes, we observed three STs in B. pertussis. Th is may suggest an older origin of B. pertussis compared to Y. pestis although other factors, such as population size and bottlenecks could also explain these diff erences. Th e most plausible explanation from our data is that the association of B. pertussis with humans originated in the LCA of B. pertussis and B. bronchiseptica complex IV. Based on that assumption, the apparent emergence of pertussis in Europe within the last 500 years may be attributable to import via travel or

43 Chapter 

migration, or to the recent acquisition by B. pertussis of the ability to cause more severe, whooping cough-like symptoms. Although most of the B. bronchiseptica complex IV strains in our collection were isolated from patients suspected to have pertussis, we know 2 little of the severity of the symptoms caused by these strains. It is conceivable that B. bronchiseptica preceded B. pertussis in Europe and that its disease was not documented because of its relatively mild and nonspeciﬁ c course. Th e work presented here places the three sequenced mammalian Bordetella strains within a phylogenetic context, thereby facilitating rational selection of strains for further genomic sequencing. In particular, sequencing of one or more members of complex IV may shed more light on processes involved in host adaptation and immune-competition. Further, the identiﬁ cation of a B. bronchiseptica lineage which circulates in human populations may be important for public health. In recent years, whole cell vaccines have been replaced by acellular vaccines comprised of 1-5 antigens derived from B. pertussis 32. Th e acellular vaccines induce a less cross-reactive immune response compared to whole cell vaccines 33 and may therefore result in an increase in B. parapertussis and B. bronchiseptica infections in vaccinated human populations.

44 Multilocus Sequence Typing reference list Genet. 35, 32-40. 9. van der Zee,A., Mooi,F., van Embden,J., and 1. Gueirard,P., Weber,C., Le Coustumier,A., and Musser,J. (1997). Molecular evolution and host Guiso,N. (1995). Human Bordetella bronchiseptica adaptation of Bordetella spp.: phylogenetic 2 infection related to contact with infected animals: analysis using multilocus enzyme electrophoresis persistence of bacteria in host. J. Clin. Microbiol 33, and typing with three insertion sequences. J. 2002-2006. Bacteriol. 179, 6609-6617.

2. Bauwens,J.E., Spach,D.H., Schacker,T.W., 10. Musser,J.M., Hewlett,E.L., Peppler,M.S., and Mustafa,M.M., and Bowden,R.A. (1992). Bordetella Selander,R.K. (1986). Genetic diversity and bronchiseptica pneumonia and bacteremia relationships in populations of Bordetella spp. J. following bone marrow transplantation. J. Clin. Bacteriol. 166, 230-237. Microbiol 30, 2474-2475. 11. Jolley,K.A., Chan,M.S., and Maiden,M.C. (2004). 3. Woolfrey,B.F. and Moody,J.A. (1991). Human mlstdbNet - distributed multi-locus sequence infections associated with Bordetella typing (MLST) databases. BMC. Bioinformatics. 5, bronchiseptica. Clin. Microbiol. Rev. 4, 243-255. 86.

4. van der Zee,A., Groenendijk,H., Peeters,M., and 12. van Loo,I.H., van der Heide,H.G., Nagelkerke,N. Mooi,F.R. (1996). The diff erentiation of Bordetella J., Verhoef,J., and Mooi,F.R. (1999). Temporal parapertussis and Bordetella bronchiseptica trends in the population structure of Bordetella from humans and animals as determined by DNA pertussis during 1949-1996 in a highly vaccinated polymorphism mediated by two diff erent insertion population. J. Infect. Dis. 179, 915-923. sequence elements suggests their phylogenetic 13. van der Zee A., Agterberg,C., Peeters,M., relationship. Int. J. Syst. Bacteriol. 46, 640-647. Schellekens,J., and Mooi,F.R. (1993). Polymerase 5. Mooi,F.R., van Loo,I.H., and King,A.J. (2001). chain reaction assay for pertussis: simultaneous Adaptation of Bordetella pertussis to Vaccination: detection and discrimination of Bordetella A Cause for Its Reemergence? Emerg. Infect. Dis. pertussis and Bordetella parapertussis. J Clin. 2001. ;7. (3 Suppl):526. -8. 7, 526-528. Microbiol 31, 2134-2140.

6. Crowcroft,N.S. and Britto,J. (2002). Whooping 14. Margalef,R. (1958). Information theory in ecology. cough--a continuing problem. BMJ 324, 1537- General Systems 3, 36-71. 1538. 15. Rozas,J., Sanchez-DelBarrio,J.C., Messeguer,X., 7. Orenstein,W.A. (1999). Pertussis in adults: and Rozas,R. (2003). DnaSP, DNA polymorphism epidemiology, signs, symptoms, and implications analyses by the coalescent and other methods. for vaccination. Clin. Infect. Dis. 28 Suppl 2, S147- Bioinformatics. 19, 2496-2497. S150. 16. Whittam,T.S. (1996). Escherichia and Salmonella 8. Parkhill,J., Sebaihia,M., Preston,A., Murphy,L.D., Cellular and Molecular Biology, Washington, DC: Thomson,N., Harris,D.E., Holden,M.T., Churcher,C. Am. Soc. Microbiol. M., Bentley,S.D., Mungall,K.L., Cerdeno-Tarraga,A. 17. Guttman,D.S. and Dykhuizen,D.E. (1994). Clonal M., Temple,L., James,K., Harris,B., Quail,M.A., divergence in Escherichia coli as a result of Achtman,M., Atkin,R., Baker,S., Basham,D., Bason,N., recombination, not mutation. Science 266, 1380- Cherevach,I., Chillingworth,T., Collins,M., Cronin,A., 1383. Davis,P., Doggett,J., Feltwell,T., Goble,A., Hamlin,N., Hauser,H., Holroyd,S., Jagels,K., Leather,S., 18. Schouls,L.M., van der Heide,H.G.J., Vauterin,L., Moule,S., Norberczak,H., O’Neil,S., Ormond,D., Vauterin,P., and Mooi,F.R. (2004). Multiple-Locus Price,C., Rabbinowitsch,E., Rutter,S., Sanders,M., Variable-Number Tandem Repeat Analysis of Saunders,D., Seeger,K., Sharp,S., Simmonds,M., Dutch Bordetella pertussis Strains Reveals Rapid Skelton,J., Squares,R., Squares,S., Stevens,K., Genetic Changes with Clonal Expansion during Unwin,L., Whitehead,S., Barrell,B.G., Maskell,D. the Late 1990s. J. Bacteriol. 186, 5496-5505. J. (2003). Comparative analysis of the genome 19. Cummings,C.A., Brinig,M.M., Lepp,P.W., Van De,P. sequences of Bordetella pertussis, Bordetella S., and Relman,D.A. (2004). Bordetella species are parapertussis and Bordetella bronchiseptica. Nat distinguished by patterns of substantial gene loss

45 Chapter 

and host adaptation. J Bacteriol. 186, 1484-1492. 31. Gutierrez,M.C., Brisse,S., Brosch,R., Fabre,M., Omais,B., Marmiesse,M., Supply,P., and Vincent,V. 20. Leininger,E., Roberts,M., Kenimer,J.G., Charles,I. (2005). Ancient origin and gene mosaicism of the G., Fairweather,N., Novotny,P., and Brennan,M. progenitor of mycobacteriumtuberculosis. PLoS. J. (1991). Pertactin, an Arg-Gly-Asp-containing 2 Pathog. 1, e5. Bordetella pertussis surface protein that promotes adherence of mammalian cells. Proc. Natl. Acad. 32. Edwards,K.M. and Decker,M.D. (2004). Pertussis Sci. U. S. A. 88, 345-349. Vaccine. In: Vaccines, ed. S.A.Plotkin and W.A.Ore nsteinPhiladephia: WB Saunders Company, 2708- 21. Charles,I.G., Dougan,G., Pickard,D., Chatfi eld,S., 2720. Smith,M., Novotny,P., Morrissey,P., and Fairweather,N.F. (1989). Molecular cloning and 33. David,S., van,F.R., and Mooi,F.R. (2004). Effi cacies of characterization of protective outer membrane whole cell and acellular pertussis vaccines against protein P.69 from Bordetella pertussis. Proc. Natl. Bordetella parapertussis in a mouse model. Vaccine Acad. Sci. U. S. A 86, 3554-3558. 22, 1892-1898.

22. McLaff erty,M.A., Harcus,D.R., and Hewlett,E.L. (1988). Nucleotide sequence and characterization of a repetitive DNA element from the genome of Bordetella pertussis with characteristics of an insertion sequence. J. Gen. Microbiol. 134 ( Pt 8), 2297-2306.

23. Jukes,T.H. and Cantor,C.R. (1969). Mammalian Protein Metabolism, New York: Academic.

24. Achtman,M., Zurth,K., Morelli,G., Torrea,G., Guiyoule,A., and Carniel,E. (1999). Yersinia pestis, the cause of plague, is a recently emerged clone of Yersinia pseudotuberculosis. PNAS 96, 14043- 14048.

25. Felsenstein,J. (1978). Cases in which parsimony or compatibility methods will be positively misleading. Systematic Biology 27, 401-410.

26. Mooi,F.R., van Oirschot,H., Heuvelman,K., van der Heide,H.G., Gaastra,W., and Willems,R.J. (1998). Polymorphism in the Bordetella pertussis virulence factors P.69/pertactin and pertussis toxin in The Netherlands: temporal trends and evidence for vaccine-driven evolution. Infect. Immun. 66, 670- 675.

27. Major R.H. (1945). Classic Descriptions of Disease, Springfi eld, C. C. Thomas.

28. LAPIN,J.H. (1943). Whooping Cough, Springfi eld, Illinois: Charles C. Thomas.

29. Still,G.F. (1931). Th e History of Paediatrics., Oxford University Press, London.

30. Achtman,M., Zurth,K., Morelli,G., Torrea,G., Guiyoule,A., and Carniel,E. (1999). Yersinia pestis, the cause of plague, is a recently emerged clone of Yersinia pseudotuberculosis. Proc. Natl. Acad. Sci. U. S. A 96, 14043-14048.

Chapter Chapter 33

Genetic Changes Associated with the Adaptation of a B. bronchiseptica Lineage to the Human Host

Dimitri A. Diavatopoulos1,2, Craig A. Cummings3,5, Leo M. Schouls1, Mary M. Brinig3,5, David A. Relman3,4,5, Frits R. Mooi1,2

Chapter 2&3, PLoS Pathogens, 2005 Chapter 

abstract Th e population structure of the mammalian bordetellae, as deﬁ ned by multilocus sequence

typing, suggested that B. parapertussishu evolved from an animal-associated lineage of B. bronchiseptica, while B. pertussis and B. bronchiseptica complex IV evolved from a distinct B. bronchiseptica lineage. Extant members of B. bronchiseptica complex IV were mainly isolated 3 from humans suspected to have whooping cough. Comparative analysis of the genomes of

single isolates of B. bronchiseptica, B. parapertussishu and B. pertussis indicated a signiﬁ cant

genome reduction in the human-adapted species B. pertussis and B. parapertussishu, and it was suggested that DNA acquisition has played little or no role in the evolution of B.

pertussis and B. parapertussishu. A comparative genomic hybridization (CGH) study of a large collection of Bordetella strains provided further indication of genome decay in B.

pertussis and B. parapertussishu, although it is still unknown how this process commenced. Th e identifi cation of B. bronchiseptica complex IV provides an opportunity to study host adaptation by comparing the genomes of B. pertussis to that of B. bronchiseptica complex IV and B. bronchiseptica complex I. In this study, the genomes of 26 B. bronchiseptica complex I, 13 B. bronchiseptica complex IV strains and 12 B. pertussis isolates were compared by CGH to a whole-genome Bordetella microarray , in order to elucidate the host-adaptation of B. pertussis and B. bronchiseptica complex IV. Comparison of gene content identifi ed the absence of the pertussis toxin locus and dermonecrotic toxin gene, as well as a polymorphic LPS biosynthesis locus, as associated with adaptation of complex IV strains to the human host. LPS structural diversity among these strains was confi rmed by gel electrophoresis. CGH analysis suggested genome reduction in B. bronchiseptica complex IV strains, and it may be possible that this process may have started well before the evolution of B. pertussis. We hypothesize that the diff erences in key virulence genes between these two lineages may refl ect immune competition in the human host.

48 Comparative Genomic Hybridization introduction Bordetella pertussis and Bordetella parapertussis, both of which cause whooping cough in humans, have independently evolved from a Bordetella bronchiseptica-like ancestor (Chapter 2, 1-3). B. bronchiseptica has been isolated from many diff erent mammalian species, but only rarely from humans 4,5. Th e three species are very closely related, but display diff erent clinical 3 manifestations 6. Generally, B. bronchiseptica causes chronic infections, while B. pertussis and B. parapertussis cause acute disease. Analysis of the population structure of these three species by MLST revealed four distinct complexes (see Chapter 2). B. pertussis and B. parapertussishu comprised complexes II and III, respectively. Interestingly, B. bronchiseptica strains were divided into two complexes, I and IV. Complex I was isolated mainly from animals (67%) and included the B. bronchiseptica strain (RB50, ST12) of which the complete genome sequence was determined. Th e majority of complex IV strains was of human origin (80%), and these strains were more closely related to B. pertussis than B. bronchiseptica complex I strains. A tree inferred from the nucleotide sequences of the virulence factor pertactin (prn) showed a similar clustering as the MLST, B. pertussis and B. bronchiseptica complex IV strains comprised one branch while B. bronchiseptica complex I and B. parapertussis strains comprised another. Th e close genetic relationship of B. bronchiseptica complex IV to B. pertussis and the fact that a high percentage of these strains was isolated from humans suggested that these strains may comprise human-adapted pathogens. Th e fact that the mammalian bordetellae are so closely related but diff er signifi cantly in host tropism and clinical manifestations makes them very suitable to study host adaptation. Comparison of the complete genome sequences of single isolates of B. bronchiseptica, B. pertussis and B. parapertussishu showed that the evolution of B. pertussis and B. parapertussishu has been accompanied by genome decay 2. To identify genetic diff erences that may be associated with host adaptation or host restriction, B. bronchiseptica complex I and B. bronchiseptica complex IV were compared using comparative genomic hybridization (CGH) to a microarray representing the genomes of B. bronchiseptica, B. parapertussis and B. pertussis. Th is approach resulted in the identifi cation of several major virulence factors that were absent from B. bronchiseptica complex IV compared to B. bronchiseptica complex I. Diff erences between these complexes were also observed in the biosynthesis of LPS. Th ese diff erences may have been driven by immune-mediated competition between the human- associated lineages B. parapertussishu, B. pertussis and B. bronchiseptica complex IV.

49 Chapter 

experimental procedures

bacterial strains A total of 26 B. bronchiseptica complex I and 13 complex IV isolates (see Chapter 1) were 3 used in this study. Routine culturing of strains was performed on Bordet Gengou (BG) agar supplemented with 15% sheep blood (BD, Franklin Lakes, NJ, USA) at 37°C for two to ﬁ ve days. Chromosomal DNA was isolated using the Wizard Genomic DNA Puriﬁ cation Kit (Promega, Madison, WI), according to the manufacturers’ protocol for Gram-negative bacteria. Strain characteristics are listed in Table 1.

lps sds-page and western blotting BG-agar grown bacteria were harvested, boiled in 1x sample buﬀ er (7.5% glycerol, 0.125M Tris-HCl, pH 6.8, 1.5% SDS), and treated with proteinase K 7. Tricine-SDS-PAGE was then performed in 4% stacking and 16% separating gels, as previously described by Lesse et al. 8. Silver staining was performed as described by Tsai and Frasch 9. LPS was transferred to PVDF membranes (Amersham Biosciences, Buckinghamshire, UK) and blocked with 0.5% (w/v) Protifar nonfat dried milk, 0.5% bovine serum albumin (w/v) and 0.1% Tween-20 in PBS. Immunoblotting was performed with monoclonal antibodies (mAbs) 36G3 and BL-8, directed against band A and band B LPS, respectively 10,11.

dna microarrays Bordetella PCR product-based DNA microarrays were prepared largely as described by Cummings et al. 3. Th is study also employed a new array design that contained all of the probes from the fi rst array plus 1,417 additional probes that brought the theoretical ORF coverage of these arrays up to 97.4% for B. pertussis Tohama, 98.5% for B. bronchiseptica RB50 and 97.9% for B. parapertussis 12822. Like the previously used array probes, these additional probes were PCR products with a size of less than 300 bp and amplifi ed from the sequenced reference genomes with ORF-specifi c oligonucleotides (Illumina, San Diego, CA) designed with MicroarrayArchitect (C.A. Cummings and D.A. Relman, unpublished data).

comparative genomic hybridization Th e genomic DNA of B. bronchiseptica was labeled with Cy5 and hybridized to the array in conjunction with a Cy3 labeled genomic DNA reference comprising the three sequenced mammalian Bordetella genomes (B. pertussis Tohama, B. parapertussis 12822 and B. bronchiseptica RB50). CGH was carried out essentially as described by Cummings et al. 3, with minor modifi cations. Labeled probes were purifi ed using the Cyscribe GFX Purifi cation Kit (Amersham Biosciences, Freiburg, Germany) following the manufacturer’s

50 Comparative Genomic Hybridization

Table 1. Strain characteristics Strain Original Species Complex ST Host Country Year Number name B0084 BB I 31 dog unknown unknown PB383 B0189 BB I 27 dog South Africa 1984 6635 B0223 BB I 14 guinea pig Australia 1987 399 B0224 BB I 6 guinea pig Germany unknown 396 B0226 BB I 5 cat Netherlands unknown 373 3 B0227 BB I 23 cat Netherlands unknown 385 B0230 BB I 6 horse Denmark unknown 406 B0237 BB I 12 rabbit Denmark unknown 405 B0238 BB I 27 dog Netherlands unknown 207 B0242 BB I 7 pig USA unknown 688 B0251 BB I 23 dog Netherlands unknown 355 B0254 BB I 7 rabbit Netherlands 1987 372 B0258 BB I 10 rabbit USA unknown 705 B0260 BB I 23 cat Denmark unknown 723 B0505 BB I 12 human Netherlands 1997 9700219 B1965 BB I 10 dog unknown unknown 590 B1966 BB I 27 dog unknown unknown 599 B1967 BB I 4 dog unknown unknown 601 B1972 BB I 5 cat unknown unknown 782 B1975 BB I 6 horse unknown unknown 982 B1985 BB I 6 seal North Sea 1988 M861/99/1 B1986 BB I 6 seal North Sea 1997 M2936/97/3 B1987 BB I 33 seal Caspian Sea 2000 M85/00/1 B2104 BB I 23 human Netherlands 1995 95-638 B2112 BB I 7 human Netherlands 1998 98-061 B2116 BB I 4 human Netherlands 2000 00-257

B0006 BP II 1 human Netherlands 1950 509 B0213 BP II 1 human Japan 1952 Tohama I B0336 BP II 2 human Netherlands 1952 Hp291 B0352 BP II 2 human Netherlands 1994 9400526 B0441 BP II 1 human Netherlands 1952 459 B0540 BP II 2 human Netherlands 1994 9400450 B0558 BP II 2 human Netherlands 1949 Hp53 B0564 BP II 1 human Netherlands 1950 Hp199 B0604 BP II 2 human Netherlands 1995 9500451 B0782 BP II 2 human Netherlands 1996 9600427 B0939 BP II 2 human Netherlands 1996 9600498 B1121 BP II 24 human USA 1947 18-323

B0232 BB IV 18 human Germany unknown 397 B0243 BB IV 28 pig Ukraine unknown 654 B0259 BB IV 29 turkey USA unknown 707 B1968 BB IV 22 dog unknown unknown 591 B1969 BB IV 18 human unknown unknown 675 B2114 BB IV 25 human Netherlands 1999 99-003 B2490 BB IV 9 human USA unknown SBL-F6116 B2491 BB IV 8 human USA unknown SBL-F6368 B2492 BB IV 21 human USA 1996 GA96-01 B2494 BB IV 15 human USA unknown MO149 B2495 BB IV 17 human USA unknown MO211 B2496 BB IV 3 human USA unknown MO275 B2506 BB IV 34 human USA 2000 00-P-2730 B2586 BB IV 3 human Argentina >1998 203 B2588 BB IV 35 human Argentina >1998 5256

Abbreviations: BB= B. bronchiseptica; BPP-OV= B. parapertussisov; BPP-HU= B. parapertussishu; BP= B. pertussis

51 Chapter 

protocol for probes produced by the CyScribe First-Strand cDNA Labelling Kit. After purifi cation, the test and reference labeled DNA samples were concentrated to <8.5 μl using a Savant SpeedVac SVC-100H. Th e test and reference samples were combined and 150 μg of yeast tRNA (Invitrogen Life Technologies, San Diego, CA) was added to block nonspecifi c binding. Th e probe volume was adjusted to 24 μl with water, 3 and then 5.1 μl of 20x SSC (1x SSC is 0.15M NaCl plus 0.015M sodium citrate) and 0.9 μl of 10% sodium dodecyl sulfate (SDS) were added. Th irty μl of the probe were added to the array and covered with a 25x40 mm no.1 glass coverslip. Hybridization was performed in GeneMachines Hybchambers (Genomic Solutions, Ann Arbor, MI) with 2x30 μl of 3x SSC to maintain humidity and incubated at 65°C overnight. Arrays were washed in 0.5x SSC, 0.03% SDS for 30 s, 0.1x SSC, 0.01% SDS for 30 s, 0.05x SSC, 0.005% SDS for 1 min and 0.025x SSC for 1 min. Th e fi rst wash was performed at 65°C, the remaining washes at room temperature. Slides were dried using a Quick- Dry Filtered Air Gun (Matrix Technologies Corporation, Hudson, NH). Images were acquired on a PerkinElmer ScanArray 4000XL scanner using ScanaArray Express software (PerkinElmer Life and Analytical Sciences, Inc., Boston, MA). Images were analyzed with GenePix Pro software (Axon Instruments, Union City, CA).

microarray data analysis Processed two-color array image data were submitted to an in-house microarray database. Data were extracted using fi lters to eliminate automatically and manually fl agged spots and spots with very low background subtracted signal intensity (<150) in the reference channel. B. bronchiseptica complex I and complex IV enriched sequences were identifi ed using the Signifi cance Analysis for Microarrays software (SAM) 12. Th e probes were analyzed using 26 complex I and 13 complex IV strains that were hybridized to the arrays. SAM analysis was run using the two-class option with KNN missing value imputation. In addition to a statistically signifi cant diff erence, a 2-fold diff erence in mean signal intensity ratio for each probe was also required.

52 Comparative Genomic Hybridization results comparative genomic hybridization of b. bronchiseptica complexes i and iv Based on housekeeping gene and pertactin sequence data, a close relationship was observed between B. pertussis and B. bronchiseptica complex IV (Chapter 2). Complex IV strains 3 were also more frequently isolated from human hosts (80%) compared to complex I strains (33%), suggesting that these strains may be better adapted to humans. To identify genetic events that may have played a role in host adaptation or host restriction, we used comparative genomic hybridization (CGH) to Bordetella DNA microarrays. Genomic DNA from 26 B. bronchiseptica complex I and 13 complex IV strains was hybridized to microarrays, representing 97.4%, 98.5% and 97.9% of the open reading frames of B. pertussis Tohama, B. bronchiseptica RB50 and B. parapertussis 12822, respectively (CGH data fi les have been deposited in ArrayExpress, accession E-TABM-32). Signifi cance Analysis of Microarrays (SAM) was used to identify probes with statistically signifi cant log intensity ratio diff erences between the two complexes. Th is approach detected sequences that have been deleted more often in one of the two complexes, as well as DNA sequences diverging from the reference sequences in one of the two complexes (a table with microarray data is available from the authors). Th irty-one probes, representing 29 genes and IS1663, hybridized more strongly or frequently to the genomes of complex IV than to those of complex I. Two of these represented the B. pertussis alleles of the virulence genes prn and tcfA, encoding pertactin and tracheal colonization factor, respectively. Sequence analysis confi rmed that the B. bronchiseptica complex IV prn sequences were more similar to B. pertussis than to B. bronchiseptica complex I (see Chapter 2). Sixteen of these 31 probes had been identifi ed previously as B. pertussis- specifi c 3. In most cases, these probes hybridized to nine or more complex IV strains, but not to any complex I or III strain (see CGH data). Th e genes represented by these probes encode diverse functions such as metabolism, transport, regulation and transposition. With the possible exception of BP0703, which encodes a TonB-dependent iron receptor, no obvious virulence genes were observed among the 30 genes. In B. pertussis Tohama, BP0703 is interrupted by an IS481 element 2. Most of the 30 genes were located in small clusters along the B. pertussis Tohama chromosome, indicating that absence or presence were due to events (deletions or insertions) involving multiple genes. Th e presence of the 30 genes in B. bronchiseptica complex IV and B. pertussis but not in B. bronchiseptica complex I or B. parapertussishu suggested that they were acquired by the common ancestor of B. pertussis and B. bronchiseptica complex IV. We also identifi ed 248 probes, representing 237 genes, that exhibited signifi cantly stronger hybridization to complex I genomes compared to complex IV genomes, suggesting they were divergent or absent in B. bronchiseptica complex IV strains. Sixty-eight (27%) of these corresponded to genes associated with mobile elements such as prophages, while many of

53 Chapter 

the other probes represented metabolic, transport and regulatory genes. Surprisingly, several virulence-associated genes were found to be missing or divergent in the complex IV strains. Th ese included the B. bronchiseptica alleles of tcfA and prn, the Bvg-regulated intermediate phase gene A (bipA), the alcaligin biosynthesis locus (alcA/E), the pertussis toxin synthesis and transport locus (ptx/ptl), the dermonecrotic toxin gene (dnt) and the lipopolysaccharide 3 Complex I Complex IV RB50 Tohama B0227 ST23 B0242 ST7 B2112 ST7 B2490 ST9 B. pertussisB. bronchiseptica B. B0084 ST31 B0189 ST27 B0224 ST6 B0230 ST6 B0238 ST27 B0251 ST23 B0258 ST10 B1975 ST6 B1985 ST6 B1986 ST6 B0232 ST18 B0243 ST28 B0259 ST29 B1968 ST22 B1969 ST18 B2114 ST25 B2491 ST8 B2492 ST21 B2494 ST15 B2495 ST17 B2496 ST3 B2506 ST34 ptxA ptxB ptxB ptxD pertussis ptxE toxin ptxC ptxC ptlA ptlB ptlC ptlD pertussis ptlI toxin ptlE secretion ptlF ptlG ptlH dnt dermonecrotic toxin tcfA (BP) tracheal colonization factor A prn (BP) prn (BB/BPP) pertactin prn (BB/BPP) alcA alcB alcC alcD alcE alcaligin alcR alcS fauA -40 4

Figure 1. Gene content of the diff erentially hybridizing virulence loci between B. bronchiseptica complex I and IV, as determined by SAM analysis of CGH data. Each column represents one strain. Strain numbers and sequence types (ST) are indicated above the columns. Each row represents one ORF (in B. bronchiseptica RB50 gene order), ORF designations are shown to the right of the rows. In the case of tcfA and prn, the origins of the probes are indicated between parentheses. The BP probe of tcfA was 100% similar to B. pertussis Tohama and 85.1% similar to B. bronchiseptica RB50. The BP prn probe was 100% similar to B. pertussis Tohama and 86% similar to B. parapertussis 12822 and B. bronchiseptica RB50. The BB/BPP prn probes were both 100% similar to B. parapertussis 12822 (BPP) and B. bronchiseptica RB50 (BB) and 86% similar to B. pertussis Tohama. The yellow-black-blue color scale indicates the hybridization value relative to the reference; references are B. bronchiseptica RB50, B. parapertussis 12822 and B. pertussis Tohama. For B. bronchiseptica RB50 and B. pertussis Tohama, the data in the fi gure are based on the genomic sequences. Yellow, black and blue indicate decreased hybridization, hybridization values comparative to the references, and increased hybridization (e.g. due to gene duplication), respectively. Intermediate values indicate partial deletions or sequence divergence. Missing data are represented in grey. A color version of this fi gure is available in the appendix.

54 Comparative Genomic Hybridization

(LPS) biosynthetic locus (see Figure 1). Interestingly, 10 out of 13 complex IV strains harbored deletions in the adjacent ptx and ptl loci, which encode pertussis toxin (Ptx) and its secretion machinery, respectively 13,14 (Figure 1). While conditions under which Ptx is expressed by either B. parapertussis or B. bronchiseptica have not been identifi ed, the structural genes are generally conserved among these species 15, suggesting selective pressure to retain the ability to produce functional 3 Ptx under certain circumstances. Th e complex IV strains in which ptx/ptl was still present (ST3/17/29) were tested for expression of Ptx by immunoblotting, and these strains were found not to express Ptx under the growth conditions used (unpublished data). Another distinguishing characteristic of the complex IV strains was the deletion of the dermonecrotic toxin gene (dnt) in eight out of 13 strains. In contrast, this gene was detected in all B. pertussis, B. parapertussishu, B. parapertussisov and B. bronchiseptica complex I strains 3. CGH showed that the genomes of the complex IV strains hybridized stronger to the B. pertussis-derived probes of tcfA and prn than to the B. bronchiseptica/B. parapertussis-derived probes. Th e genes tcfA and prn encode two autotransporters of the same family, that are involved in adhesion of the bacterium to the epithelium 16. Of these two, prn is expressed by B. bronchiseptica, B. parapertussis and by B. pertussis 17,18, but tcfA has been suggested to be expressed exclusively by B. pertussis 19. Conversely, the genomes of the B. bronchiseptica complex I strains hybridized stronger to the B. bronchiseptica-derived probes of tcfA and prn, which are virtually identical to B. parapertussis. Sequencing of prn showed that the complex IV strains indeed had a prn gene that is more similar to B. pertussis than to B. bronchiseptica complex I (Chapter 2). Hybridization patterns suggested that the genomic sequence of the complex IV alcaligin biosynthetic locus, encoding a siderophore, might diff er from that of complex I and B. pertussis (Figure 1). Complex IV strains hybridized less well to alcABCDERS and fauA, with particularly weak hybridization to the alcE probe, of which the exact function in alcaligin synthesis has not yet been determined. Probably, complex IV B. bronchiseptica strains are competent to produce alcaligin, but these genes may be divergent compared to B. pertussis and B. bronchiseptica complex I strains. lps polymorphism Th e genetic structure of the LPS biosynthesis locus also diff ered substantially between complex I and IV strains. Th e LPS molecules of Gram-negative bacteria usually consist of three, covalently linked, major domains: the lipid A, the branched chain oligosaccharide core and the hydrophilic O-antigen. A number of genetic loci have been implicated in the synthesis of these domains in Bordetella, such as the lpx locus (lipid A), the waa locus (inner core), the wlb locus (trisaccharide) and the wbm locus (O-antigen) 20-24. B. pertussis usually produces a lipo-oligosaccharide (LOS) that consists of the lipid A and the inner core, to

55 Chapter 

A. Complex I Complex IV ST23 B0227 B0242 ST7 B2112 ST7 B2490 ST9 BP Tohama BB RB50 ST31 B0084 ST27 B0189 ST6 B0224 ST6 B0230 ST27 B0238 ST23 B0251 ST10 B0258 ST6 B1975 ST6 B1985 ST6 B1986 ST18 B0232 ST28 B0243 ST29 B0259 ST22 B1968 ST18 B1969 ST25 B2114 ST8 B2491 ST21 B2492 ST15 B2494 ST17 B2495 ST3 B2496 ST34 B2506 3 O-antigen

band A band B

O-antigen

band A

band B

B. 1 2 2 1 1 2 2 2 1 1 2 1 1 2 2 3 1 2 2 2 4 2 2 3 1 1 4 LPS genetic profile pagP pagL (de)acylation lipidA lpxA lpxB lpxC lipidA lpxD lpxH lpxK lpxK waaA inner core waaC wlbA wlbB wlbC trisaccharide wlbD wlbE wlbF wlbG wlbH wlbI wlbJK wlbL wbmA wbmB wbmC O-antigen wbmD wbmE wbmF wbmG wbmH wbmI wbmJ wbmK wbmL wbmM wbmN wbmO wbmR wbmS BB0124 BB0125 BB0126 BB0127 BB0127 wbmU wbmT alternative wbmS O-antigen wbmR wbmQ wbmP -40 4

56 Comparative Genomic Hybridization

Figure 2. Expression of LPS by B. bronchiseptica complex I and complex IV strains and gene content variation at the LPS biosynthesis locus. A. Top panel: Electrophoretic LPS profi les obtained by tricine-SDS-PAGE and silver staining. Middle panel: Western blot of the same samples with mAb 36G3, which detects band A. Bottom panel: Western blot of the same samples with mAb BL8, which detects band B. B. Gene content of the LPS biosynthesis locus as determined by CGH. See Figure 1 for details. For B. bronchiseptica RB50 and B. pertussis Tohama, the data in the fi gure are based on the genomic sequences. The genes wbmPQRSTU represent an alternative LPS O-antigen biosynthesis sublocus that is orthologous to the genes found in B. parapertussis 12822 [2] and B. bronchisep- 3 tica C7635E [21]. LPS genetic profi les as described in the text are indicated at the top of the columns. Color scale as in Figure 1. Missing data are represented in grey. A color version of this fi gure is available in the appendix. which the outer core (a trisaccharide) is attached; also referred to as band A. Th e genes wlbA-L encode the biosynthesis of this trisaccharide and the transport to the inner core 25. Certain B. pertussis strains produce a LOS molecule that is identical to band A except that it lacks the trisaccharide; this structure is called band B 26. Th e O-antigen, which is added to the trisaccharide, is only found in B. bronchiseptica and B. parapertussis. Th is structure is missing in B. pertussis due to the deletion of the genes wbmA-U 21. In Figure 2, the gene content of the LPS locus is shown for 13 complex I and 13 complex IV strains, and for B. pertussis Tohama and B. bronchiseptica RB50. Four LPS gene content profi les, designated LPS 1–4, could be distinguished among the B. bronchiseptica strains. Strains with the LPS 1 profi le had an LPS gene composition similar to RB50, characterized by the absence of wbmPQRSTU. Th e LPS 2 profi le was characterized by the absence of the genes wbmORS and BB0124-BB0127 and the presence of an alternative O-antigen locus, comprising wbmPQRSTU, orthologous to the B. parapertussis 12822 genes 2. Both of these genotypes appear competent for the production of a full length LPS. Strains with the LPS 3 profi le lacked wbmD-U and BB0124-BB0127, suggesting that they may not produce O- antigen. Th e LPS 4 profi le was similar to the LPS 3 profi le but additionally lacked wbmABC and wlbD-L, suggesting that strains of this genotype may be defi cient for the production of trisaccharide (wlb locus) as well as O-antigen (wbm locus). Th e deletion in the O-antigen genes of LPS 4 strains was similar to that observed in the O-antigen genes of B. pertussis Tohama. All complex I strains displayed either an LPS 1 or an LPS 2 profi le. Nine of 13 complex IV strains had either an LPS 1 or an LPS 2 profi le, while four complex IV strains showed more extensive deletions, resulting in LPS 3 and LPS 4 profi les. To study the eff ect of these deletions on LPS production, proteinase K-treated cell lysates were analyzed by Tricine-SDS-PAGE, followed by silver staining or immunoblotting with monoclonal antibodies directed against either band A or band B (mAbs 36G3 and BL-8, respectively 10,11, Figure 2). Silver-staining showed that all complex I strains produced a band co-migrating with band A of B. pertussis, except for B1985 and B2112, which produced a band co-migrating with B. pertussis band B and a band migrating at a position between bands A and B, respectively. Th is was confi rmed by immunoblotting with mAb 36G3, which further showed that these strains also produced O-antigen, again with the exception of B1985 and B2112. Th e

57 Chapter 

epitope recognized by mAb 36G3 is also present in the O-antigenic repeats, as was described previously 7. Further, most strains that produced band A and O-antigen also produced an additional band just above band A, which was also recognized by mAb 36G3 and is therefore likely derived from band A. Finally, immunoblotting with mAb BL8, directed against B. pertussis band B, showed that only fi ve of the complex I strains produced band 3 B (B0189, B0258, B1975, B1985 and B1986); and these strains also produced amounts of band B detectable by silver-staining. B1985 and B2112 showed no obvious deletions at their wlb locus, and the fact that they did not produce band A may be attributed to point mutations, e.g. in their wlb locus, or to regulatory eff ects. B1985 is thus the only complex I strain that produces exclusively band B, which has also been described for certain B. pertussis strains 26. All complex IV strains produced band B as detected by immunoblotting with mAb BL8. Th e nine complex IV strains with the LPS 1 or 2 profi le all produced band A and O- antigen as detected by silver-staining and immunoblotting with mAb 36G3. However, the two strains with the LPS 4 profi le, B2490 and B2506, produced a band smaller than band A, which failed to be recognized by mAb 36G3. Further, these strains produced no O- antigen detectable by immunoblotting. Of the two strains with a LPS 3 profi le, one strain (B0243) produced only band B. Unexpectedly, the other strain, B2494 produced both band A and O-antigen as detected by immunoblotting with mAb 36G3, indicating that this strain contains as yet unknown O-antigen biosynthesis genes. Silver-staining suggested that the two LPS 4 strains may also produce O-antigen, although the O-antigen failed to be recognized by mAb 36G3. Van den Akker et al. previously observed that the epitope in band A that is recognized by mAb 36G3 was also present in the O-antigen 7, and an explanation for the lack of O-antigen recognition by mAb 36G3 in these strains may thus be the observed change in band A.

58 Comparative Genomic Hybridization discussion Multilocus enzyme electrophoresis (MLEE) studies and the comparison of the genomes of the mammalian bordetellae previously indicated that B. parapertussishu and B. pertussis adapted to humans independently 1,2. Th e genesis of B. pertussis was further refi ned by MLST, which identifi ed a B. bronchiseptica lineage, designated complex IV, from which 3 this human pathogen evolved (Chapter 2). Until now it was assumed that B. bronchiseptica strains are mainly animal pathogens, and that human infections are zoonotic. Further, transmission of B. bronchiseptica between humans was assumed not to occur. Th erefore, it was intriguing that B. bronchiseptica complex IV strains were mainly isolated from humans, suggesting that they may be comprised of human-adapted strains. To identify genes or genetic events associated with host adaptation, we used comparative genomic hybridization (CGH) to microarrays. Both B. pertussis and B. parapertussis cause brief, transient infections in humans, with an infectious period of three weeks or more 6. Infection with B. pertussis initiates a response by the adaptive immune system, and antibodies are developed against a large number of B. pertussis antigens, such as Ptx, fi lamentous haemagglutinin, pertactin, fi mbriae and the LPS.

However, due to the short infectious period of B. pertussis and of B. parapertussishu, their transmission will likely be primarily aﬀ ected by the innate immune system in the absence of vaccine-induced protective immunity. Evasion of the host immune system occurs through adenylate cyclase, Ptx, FHA, LPS, the type III secretion system and likely also by other virulence factors. Th e paroxysmal coughing, characteristic of whooping cough and essential for transmission, is probably the result of damage to the epithelia, caused by toxins such as tracheal cytotoxin, dermonecrotic toxin and possibly adenylate cyclase.

As opposed to B. pertussis and B. parapertussishu, B. bronchiseptica usually causes chronic infections in many mammalian species, including humans, that often remain unnoticed. Th e long infectious period of B. bronchiseptica compared to the other mammalian bordetellae indicates that, in addition to the innate immune system, B. bronchiseptica must be aff ected by the adaptive immune system as well. Th us, the three species probably interact diff erently with the innate and adaptive arms of the immune system, and this will likely aff ect their transmission in diff erent ways. Th is may also be refl ected in the diff erential expression of major virulence factors such as Ptx and LPS. For example, B. pertussis expresses Ptx, but not B. parapertussishu and B. bronchiseptica. Further, B. bronchiseptica, B. parapertussishu and B. pertussis produce signifi cantly diff erent LPS molecules 27. Even the generally highly conserved lipid A is polymorphic in these species 28,29. Th e variability in these virulence factors may refl ect a diff erent role in their transmission. With respect to well-characterized virulence factors, diff erences between B. bronchiseptica complex I, complex IV and B. pertussis were observed in Ptx, Dnt and LPS. CGH analyses indicated that, although the ptx/ptl genes were conserved in B. bronchiseptica complex I strains, the complete ptx/ptl locus, with the exception of ptlD, was deleted in 10 of the 13

59 Chapter 

complex IV strains analyzed. No expression of Ptx was observed in the three complex IV strains that retained the ptx/ptl locus when cultured in vitro. Although conditions under which the Ptx genes are expressed in B. bronchiseptica complex I strains have not been identiﬁ ed, their conservation suggests that they may confer a selective advantage at some stage in the transmission cycle. Another characteristic that set the complex IV strains apart 3 from all other mammalian bordetellae studied is that in 8 out of 13 strains the gene for dermonecrotic toxin (dnt) was deleted. Further, the LPS genetic locus was generally more polymorphic in complex IV than in complex I, and deletions were observed in the O- antigen and trisaccharide biosynthesis genes in some complex IV strains. In complex IV strains with the LPS 4 proﬁ le, the extent of deletion in the O-antigen genes was very similar to B. pertussis Tohama. Like B. pertussis, four complex IV strains lacked the O-antigen genes

known to be present in the sequenced genomes of B. bronchiseptica and B. parapertussishu. However, despite carrying apparent large deletions in the LPS O-antigen biosynthesis locus, silver staining suggested that at least one, and possibly three of these strains did produce an O-antigen, suggesting that these strains may carry LPS genes distinct from those in RB50. Two of these strains, both isolated from humans, also lacked many genes required for the biosynthesis of the trisaccharide, and failed to produce trisaccharide detectable by immunoblotting. Th e trisaccharide is encoded by the wlb locus, and functional wlb genes were present in all three Bordetella species and are required to add a trisaccharide to the lipid A core 24. Strains with a deletion in this locus miss the trisaccharide and the O-antigen 27. Th e absence of the trisaccharide is intriguing in view of the fact that it was found to be otherwise conserved in all other mammalian Bordetella strains analyzed. Most of the above described virulence factors play complex and multiple roles in the transmission of these species, and often their exact role has not been determined. Ptx is an ADP-ribosylating toxin composed of fi ve subunits, encoded by the genes ptxABCDE 30,31. Secretion of the toxin is mediated by the associated Ptl system, consisting of nine proteins 14,32, which are encoded by the ptl genes that are located within the same transcriptional unit as the ptx genes. Th e properties and eff ects of Ptx have been studied extensively, both in vitro and in mouse and rat models. Ptx interferes with many mammalian cell signaling pathways, possibly leading to a cascade of biological eff ects, including insulinemia, histamine sensitization and both immunosuppression and stimulation 33-35. Ptx was found to play an important role especially in the early phases of respiratory tract infection of mice by B. pertussis, presumably by inhibiting neutrophil infl ux 36. Further, Ptx was also shown to suppress the antibody response to B. pertussis 37,38. Th us Ptx targets both the innate and adaptive immune system. Finally, Ptx has also been suggested to function as an adhesin 39,40. Ptx is thought to be expressed and produced exclusively by B. pertussis. Intriguingly, the coding sequences for these genes appear to be largely intact and conserved

in B. parapertussishu, B. parapertussisov and most B. bronchiseptica. Replacement of the ptx/ptl

promoter sequence in a B. parapertussishu and B. bronchiseptica strain with the B. pertussis

60 Comparative Genomic Hybridization promoter actually resulted in the production and secretion of biologically active Ptx 41. Dnt is an intracellular toxin that activates the small GTPase Rho through deamidation or polyamination 42. It has been shown that Dnt is important for turbinate atrophy and the colonization of the upper respiratory tract in pigs 43, but its role in pertussis pathogenesis has not been elucidated. For B. bronchiseptica, it was shown that the O-antigen was not required for initial 3 colonization of mice, however, mutants lacking O-antigen showed decreased colonization rates at the long-term 44. Th us, the O-antigen is important for B. bronchiseptica for establishing chronic infections. It was also shown that B. bronchiseptica strains defi cient in O-antigen elicited equal amounts of specifi c antibodies to the wild-type B. bronchiseptica strains 44. Although the O-antigen did not aff ect the amount of antibodies raised against B. bronchiseptica, it did provide protection against the innate immune system by inhibiting activation of the complement system 44. Long stretches of O-antigenic repeats may interfere with the function of other virulence factors. For Shigella, it was shown that shortening of the O-antigen signifi cantly enhanced the function of the type III secretion system, without losing protection by the LPS. Th us, partial loss of O-antigen may increase virulence of bacteria, possibly by enhancing accessibility to host tissues of other virulence factors such as adhesins, toxins or immunomodulatory factors, while allowing the bacteria to express a basal level of O-antigen level for protection against host defenses such as complement- mediated killing. Interestingly, Van den Akker showed that the LPS of the mammalian bordetella are modifi ed in a Bvg-dependent matter 7, suggesting that LPS modifi cation is important for virulence. It was previously shown that, for B. bronchiseptica, wlb-dependent LPS modifi cation was not required for initial infection of the lungs, but it was required for persistence. Further, the trisaccharide was apparently only required when adaptive immunity exists 27. Th e trisaccharide also protected Bordetella species against the bactericidal lung surfactants SP-A and SP-D of the innate immune system 45,46. Th us, the trisaccharide is important for resistance against both the innate and the adaptive immune system. Th is is possibly especially important for establishing persistent infections. Th us, both LPS and Ptx play a complex role in the ecology of the mammalian bordetellae, as they may at the same time protect against innate and adaptive immunity, and yet also activate these arms of the immune system. Various hypotheses can be put forward to explain the observed diff erences in major virulence factors between B. bronchiseptica complex I and complex IV. One possible explanation could be redundancy of virulence factors. For example, Dnt may not be required for B. bronchiseptica complex IV strains to infect their host, as the function of this toxin may be compensated for by other toxins in the genome, or by the type III secretion system. Alternatively, the genetic diff erences at the virulence loci may be explained by the adaptation to diff erent niches. Variation in LPS structures may for example refl ect adaptation to host receptors that are involved in innate immunity, such as the Toll-like receptor 4, which is critical to innate host defense against infections

61 Chapter 

with mammalian bordetellae 47-49. Another possibility is that these diﬀ erences have arisen in response to immune-competition between B. bronchiseptica complex IV strains and B.

pertussis. Bjørnstad & Harvill hypothesized that, since B. pertussis and B. parapertussishu both infect humans, they may have evolved to evade cross-immunity by the other pathogen 50,51. Th e authors propose that immune-competition provides an explanation for diﬀ erences 3 observed between B. pertussis and B. parapertussishu. For example, B. pertussis but not B.

parapertussishu expresses Ptx although both contain the required genes. Conversely, B.

Figure 3. Model of B. bronchiseptica the evolution of the complex I mammalian bordetellae. The bar on the left indicates increasing degrees of adap-

tation to the human broad host range LCA complex IV host. Arrows indi- Acquired cate descent; double BP0072 transposase IS1663 arrows between 16 lineage-specific genes complexes indicate Genome decay possible within- host immune-competition. In boxes, genetic events are B. bronchiseptica shown that may complex IV have played a role Deleted in speciation and Ptx Dnt niche adaptation. LPS polymorphism See text for details. Genome decay Abbreviations: LCA, last common ances- B. pertussis B. parapertussishu tor. A color version of complex II complex III this fi gure is available Ptx expression Immune No Ptx expression in the appendix. Type III secretion off competition Type III secretion off Deleted or inactivated Deleted or inactivated O-antigen locus Type II capsule Autotransporters Type II capsule Acquired IS1002 Acquired IS481, IS1002 Genome decay

human-adapted Genome decay

parapertussishu expresses O-antigen, while the corresponding genes have been deleted from B. pertussis. Similarly, the deletion of the genes for Ptx, Dnt and genes involved in trisaccharide syntheses by complex IV strains may have been driven by immune competition with B.

pertussis and possibly also B. parapertussishu. In Figure 3, the mammalian bordetellae evolutionary model from Chapter 2 is shown. Th e evolution and adaptation to humans of B. bronchiseptica complex IV and B. pertussis

is shown, as well as the evolution of B. parapertussishu. Additional to the model described

62 Comparative Genomic Hybridization in Chapter 2, genetic events that may have played an important role in the evolution of these lineages are indicated, such as the polymorphism in the LPS and the species-specifi c expression of Ptx. Potential immune-competition is indicated between species occupying the same host. Microarray-based CGH of B. pertussis, B. bronchiseptica complex I and complex IV identifi ed many genomic diff erences between these lineages. CGH revealed 29 genes and the insertion 3 sequence element IS1663 that were more frequently present in B. bronchiseptica complex IV compared to complex I strains. Of these genes, 16 were shared only by B. pertussis and B. bronchiseptica complex IV, suggesting they were acquired after the last common ancestor of complex IV and B. pertussis diverged from complex I. Conversely, 237 genes were absent or divergent in complex IV compared to complex I strains, suggesting that the B. bronchiseptica complex IV genome is decaying, as has been observed for B. pertussis and B. parapertussishu. Because putative complex IV-specifi c sequences were not represented on the microarray used here, we were unable to address the possibility that complex IV strains have acquired, through lateral transfer, genetic loci that may have promoted host restriction. However, gene acquisition appears to have been a rare event in the evolution of B. pertussis and B. parapertussis from B. bronchiseptica complex I 2 (see also Chapter 4). Th is study uses the phylogenetic framework phylogeny of the mammalian bordetellae provided in Chapter 2 as the basis for a more detailed analysis of the genomic changes that may be associated with the adaptation of B. bronchiseptica complex IV and B. pertussis to humans. Diff erences are observed in various major virulence factors that may be the result of immune-competition within the same hosts, to avoid cross-reactivity. Further, the genomes of B. bronchiseptica complex IV strains seem to have undergone genome reduction in comparison to B. bronchiseptica complex I strains, although the possibility remains that the complex IV strains have acquired DNA elements that are important for host-adaptation. Sequencing of the genomes of one or more members of complex IV may shed more light on processes involved in host adaptation and immune-competition.

63 Chapter 

acknowledgments We are grateful to Dr. Geoﬀ rey Foster (SAC Veterinary Science Division, Inverness) and to Dr. Gary Sanders (Centers for Disease Control and Prevention, Atlanta, Georgia, United States) for providing B. bronchiseptica strains. We thank Dr. Eric Harvill (Penn State University, Pennsylvania, United States) for sharing unpublished data and discussions and 3 Ing. Marjolein van Gent, Ing. Betsy Kuipers, and Ing. Hendrik-Jan Hamstra for assistance and introduction to LPS work. We also thank Dr. Martin Maiden and Dr. Keith Jolley for assistance with setting up the Bordetella MLST database. Th is work was supported by a travel grant from the Netherlands Organization for Scientiﬁ c Research (NWO).

64 Comparative Genomic Hybridization reference list 8. Lesse,A.J., Campagnari,A.A., Bittner,W.E., Apicella,M.A. (1990). Increased resolution of lipopolysaccharides and lipooligosaccharides 1. van der Zee,A., Mooi,F., van Embden,J., Musser,J. utilizing tricine-sodium dodecyl sulfate- (1997). Molecular evolution and host adaptation polyacrylamide gel electrophoresis. J.Immunol of Bordetella spp.: phylogenetic analysis using Methods 126, 109-117. multilocus enzyme electrophoresis and typing with three insertion sequences. The Journal of 9. Tsai,C.M., Frasch,C.E. (1982). A sensitive silver 3 Bacteriology 179, 6609-6617. stain for detecting lipopolysaccharides in polyacrylamide gels. Anal.Biochem. 119, 115-119. 2. Parkhill,J., Sebaihia,M., Preston,A., Murphy,L.D., Thomson,N., Harris,D.E., Holden,M.T., Churcher,C. 10. Martin,D., Peppler,M.S., Brodeur,B.R. (1992). M., Bentley,S.D., Mungall,K.L., Cerdeno-Tarraga,A. Immunological characterization of the M., Temple,L., James,K., Harris,B., Quail,M.A., lipooligosaccharide B band of Bordetella pertussis. Achtman,M., Atkin,R., Baker,S., Basham,D., Bason,N., Infection and Immunity 60, 2718-2725. Cherevach,I., Chillingworth,T., Collins,M., Cronin,A., 11. Poolman,J.T., Kuipers,B., Vogel,M.L., Hamstra,H.J., Davis,P., Doggett,J., Feltwell,T., Goble,A., Hamlin,N., Nagel,J. (1990). Description of a hybridoma bank Hauser,H., Holroyd,S., Jagels,K., Leather,S., towards Bordetella pertussis toxin and surface Moule,S., Norberczak,H., O’Neil,S., Ormond,D., antigens. Microb.Pathog. 8, 377-382. Price,C., Rabbinowitsch,E., Rutter,S., Sanders,M., Saunders,D., Seeger,K., Sharp,S., Simmonds,M., 12. Tusher,V.G., Tibshirani,R., Chu,G. (2001). Skelton,J., Squares,R., Squares,S., Stevens,K., Signifi cance analysis of microarrays applied to the Unwin,L., Whitehead,S., Barrell,B.G., Maskell,D. ionizing radiation response. Proc.Natl.Acad.Sci. J. (2003). Comparative analysis of the genome U.S.A 98, 5116-5121. sequences of Bordetella pertussis, Bordetella 13. Kotob,S.I., Hausman,S.Z., Burns,D.L. (1995). parapertussis and Bordetella bronchiseptica. Nat Localization of the promoter for the ptl genes Genet. 35, 32-40. of Bordetella pertussis, which encode proteins 3. Cummings,C.A., Brinig,M.M., Lepp,P.W., Van De,P. essential for secretion of pertussis toxin. Infection S., Relman,D.A. (2004). Bordetella species are and Immunity 63, 3227-3230. distinguished by patterns of substantial gene loss 14. Weiss,A.A., Johnson,F.D., Burns,D.L. (1993). and host adaptation. J Bacteriol. 186, 1484-1492. Molecular characterization of an operon required 4. Gueirard,P., Weber,C., Le Coustumier,A., Guiso,N. for pertussis toxin secretion. Proc.Natl.Acad.Sci. (1995). Human Bordetella bronchiseptica U.S.A 90, 2970-2974. infection related to contact with infected animals: 15. Arico,B., Rappuoli,R. (1987). Bordetella persistence of bacteria in host. J.Clin.Microbiol 33, parapertussis and Bordetella bronchiseptica 2002-2006. contain transcriptionally silent pertussis toxin 5. Bauwens,J.E., Spach,D.H., Schacker,T.W., Mustafa,M. genes. The Journal of Bacteriology 169, 2847- M., Bowden,R.A. (1992). Bordetella bronchiseptica 2853. pneumonia and bacteremia following bone 16. Leininger,E., Roberts,M., Kenimer,J.G., Charles,I.G., marrow transplantation. J.Clin.Microbiol 30, 2474- Fairweather,N., Novotny,P., Brennan,M.J. (1991). 2475. Pertactin, an Arg-Gly-Asp-containing Bordetella 6. Mattoo,S., Cherry,J.D. (2005). Molecular pertussis surface protein that promotes adherence Pathogenesis, Epidemiology, and Clinical of mammalian cells. Proc.Natl.Acad.Sci.U.S.A. 88, Manifestations of Respiratory Infections Due 345-349. to Bordetella pertussis and Other Bordetella 17. Li,L.J., Dougan,G., Novotny,P., Charles,I.G. (1991). Subspecies. Clinical Microbiology Reviews 18, 326- P.70 pertactin, an outer-membrane protein from 382. Bordetella parapertussis: cloning, nucleotide 7. van den Akker,W.M. (1998). Lipopolysaccharide sequence and surface expression in Escherichia expression within the genus Bordetella: infl uence coli. Mol.Microbiol. 5, 409-417. of temperature and phase variation. Microbiology 18. Li,J., Fairweather,N.F., Novotny,P., Dougan,G., 144 ( Pt 6), 1527-1535. Charles,I.G. (1992). Cloning, nucleotide sequence

65 Chapter 

and heterologous expression of the protective Richards,J.C., Therisod,H., Perry,M.B., Karibian,D. outer-membrane protein P.68 pertactin from (2001). Structural variability and originality of the Bordetella bronchiseptica. J.Gen.Microbiol. 138 ( Pt Bordetella endotoxins. J Endotoxin.Res. 7, 63-68. 8), 1697-1705. 29. Caroff ,M., Karibian,D., Cavaillon,J.M., Haeff ner- 19. Finn,T.M., Stevens,L.A. (1995). Tracheal colonization Cavaillon,N. (2002). Structural and functional factor: a Bordetella pertussis secreted virulence analyses of bacterial lipopolysaccharides. 3 determinant. Mol.Microbiol. 16, 625-634. Microbes.Infect. 4, 915-926. 20. Isobe,T., White,K.A., Allen,A.G., Peacock,M., Raetz,C. 30. Locht,C., Keith,J.M. (1986). Pertussis toxin gene: R., Maskell,D.J. (1999). Bordetella pertussis waaA nucleotide sequence and genetic organization. encodes a monofunctional 2-keto-3-deoxy-D- Science 232, 1258-1264. manno-octulosonic acid transferase that can 31. Nicosia,A., Perugini,M., Franzini,C., Casagli,M.C., complement an Escherichia coli waaA mutation. Borri,M.G., Antoni,G., Almoni,M., Neri,P., Ratti,G., The Journal of Bacteriology 181, 2648-2651. Rappuoli,R. (1986). Cloning and sequencing of the 21. Preston,A., Allen,A.G., Cadisch,J., Thomas,R., pertussis toxin genes: operon structure and gene Stevens,K., Churcher,C.M., Badcock,K.L., Parkhill,J., duplication. Proc.Natl.Acad.Sci.U.S.A 83, 4631- Barrell,B., Maskell,D.J. (1999). Genetic basis for 4635. lipopolysaccharide O-antigen biosynthesis in 32. Farizo,K.M., Cafarella,T.G., Burns,D.L. (1996). bordetellae. Infection and Immunity 67, 3763- Evidence for a ninth gene, ptlI, in the locus 3767. encoding the pertussis toxin secretion system of 22. Preston,A., Maskell,D. (2001). The molecular Bordetella pertussis and formation of a PtlI-PtlF genetics and role in infection of lipopolysaccharide complex. Journal of Biological Chemistry 271, biosynthesis in the Bordetellae. J.Endotoxin.Res. 7, 31643-31649. 251-261. 33. Morse,S.I., Morse,J.H. (1976). Isolation and 23. Allen,A.G., Isobe,T., Maskell,D.J. (1998). Identifi cation properties of the leukocytosis- and lymphocytosis- and cloning of waaF (rfaF) from Bordetella pertussis promoting factor of Bordetella pertussis. J.Exp. and use to generate mutants of Bordetella spp. Med. 143, 1483-1502. with deep rough lipopolysaccharide. The Journal 34. Munoz,J.J., Arai,H., Bergman,R.K., Sadowski,P. of Bacteriology 180, 35-40. L. (1981). Biological activities of crystalline 24. Allen,A., Maskell,D. (1996). The identifi cation, pertussigen from Bordetella pertussis. Infection cloning and mutagenesis of a genetic locus and Immunity 33, 820-826. required for lipopolysaccharide biosynthesis in 35. Pittman,M. (1979). Pertussis toxin: the cause of Bordetella pertussis. Mol Microbiol 19, 37-52. the harmful eff ects and prolonged immunity of 25. Allen,A.G., Thomas,R.M., Cadisch,J.T., Maskell,D.J. whooping cough. A hypothesis. Rev.Infect.Dis. 1, (1998). Molecular and functional analysis of the 401-412. lipopolysaccharide biosynthesis locus wlb from 36. Carbonetti,N.H., Artamonova,G.V., Mays,R.M., Bordetella pertussis, Bordetella parapertussis and Worthington,Z.E. (2003). Pertussis toxin plays Bordetella bronchiseptica. Mol Microbiol 29, 27- an early role in respiratory tract colonization by 38. Bordetella pertussis. Infection and Immunity 71, 26. Peppler,M.S. (1984). Two physically and 6358-6366. serologically distinct lipopolysaccharide profi les in 37. Carbonetti,N.H., Artamonova,G.V., Andreasen,C., strains of Bordetella pertussis and their phenotype Dudley,E., Mays,R.M., Worthington,Z.E. (2004). variants. Infection and Immunity 43, 224-232. Suppression of Serum Antibody Responses by 27. Harvill,E.T., Preston,A., Cotter,P.A., Allen,A.G., Pertussis Toxin after Respiratory Tract Colonization Maskell,D.J., Miller,J.F. (2000). Multiple roles for by Bordetella pertussis and Identifi cation of an Bordetella lipopolysaccharide molecules during Immunodominant Lipoprotein. Infection and respiratory tract infection. Infection and Immunity Immunity 72, 3350-3358. 68, 6720-6728. 38. Mielcarek,N., Riveau,G., Remoue,F., Antoine,R., 28. Caroff ,M., Aussel,L., Zarrouk,H., Martin,A., Capron,A., Locht,C. (1998). Homologous and

66 Comparative Genomic Hybridization

heterologous protection after single intranasal Murine Model of Bordetellosis. J Infect.Dis. 189, administration of live attenuated recombinant 833-836. Bordetella pertussis. Nat.Biotechnol. 16, 454-457. 49. Mann,P.B., Elder,K.D., Kennett,M.J., Harvill,E. 39. Relman,D., Tuomanen,E., Falkow,S., Golenbock,D.T., T. (2004). Toll-like receptor 4-dependent early Saukkonen,K., Wright,S.D. (1990). Recognition of a elicited tumor necrosis factor alpha expression is bacterial adhesion by an integrin: macrophage CR3 critical for innate host defense against Bordetella (alpha M beta 2, CD11b/CD18) binds fi lamentous bronchiseptica. Infection and Immunity 72, 6650- 3 hemagglutinin of Bordetella pertussis. Cell 61, 6658. 1375-1382. 50. Bjornstad,O.N., Harvill,E.T. (2005). Evolution and 40. Tuomanen,E., Weiss,A. (1985). Characterization of emergence of Bordetella in humans. Trends two adhesins of Bordetella pertussis for human Microbiol. ciliated respiratory-epithelial cells. J.Infect.Dis. 152, 51. Gupta,S., Maiden,M.C.J. (2001). Exploring the 118-125. evolution of diversity in pathogen populations. 41. Hausman,S.Z., Cherry,J.D., Heininger,U., Wirsing Von Trends in Microbiology 9, 181-185. Konig,C.H., Burns,D.L. (1996). Analysis of proteins encoded by the ptx and ptl genes of Bordetella bronchiseptica and Bordetella parapertussis. Infection and Immunity 64, 4020-4026.

42. Kashimoto,T., Katahira,J., Cornejo,W.R., Masuda,M., Fukuoh,A., Matsuzawa,T., Ohnishi,T., Horiguchi,Y. (1999). Identifi cation of functional domains of Bordetella dermonecrotizing toxin. Infection and Immunity 67, 3727-3732.

43. Brockmeier,S.L., Register,K.B., Magyar,T., Lax,A. J., Pullinger,G.D., Kunkle,R.A. (2002). Role of the Dermonecrotic Toxin of Bordetella bronchiseptica in the Pathogenesis of Respiratory Disease in Swine. Infection and Immunity 70, 481-490.

44. Burns,V.C., Pishko,E.J., Preston,A., Maskell,D.J., Harvill,E.T. (2003). Role of Bordetella o antigen in respiratory tract infection. Infection and Immunity 71, 86-94.

45. Schaeff er,L.M., McCormack,F.X., Wu,H., Weiss,A.A. (2004). Interactions of pulmonary collectins with Bordetella bronchiseptica and Bordetella pertussis lipopolysaccharide elucidate the structural basis of their antimicrobial activities. Infection and Immunity 72, 7124-7130.

46. Edwards,J.A., Groathouse,N.A., Boitano,S. (2005). Bordetella bronchiseptica Adherence to Cilia Is Mediated by Multiple Adhesin Factors and Blocked by Surfactant Protein A. Infection and Immunity 73, 3618-3626.

47. Miller,S.I., Ernst,R.K., Bader,M.W. (2005). LPS, TLR4 and infectious disease diversity. Nat Rev.Microbiol 3, 36-46.

48. Mann,P.B., Kennett,M.J., Harvill,E.T. (2004). Toll-Like Receptor 4 Is Critical to Innate Host Defense in a

67 68 Chapter

Chapter 44

Identiﬁ cation and Characterization of B. bronchiseptica Complex IV- Speciﬁ c Sequences

Dimitri A. Diavatopoulos1,2, Marjolein van Gent1, Frits R. Mooi1,2

Manuscript in preparation Chapter 

abstract Comparative genomics previously showed signiﬁ cant genome reduction in the human-

adapted species B. parapertussishu and B. pertussis compared to B. bronchiseptica. MLST suggested a close relationship of B. pertussis and B. bronchiseptica complex IV strains, and gene content comparison by hybridization to microarrays suggested that the genomes of B. bronchiseptica complex IV had also undergone reduction compared to B. bronchiseptica complex I. Although DNA acquisition seems to have played a minor role in the evolution 4 of both B. pertussis and B. parapertussis, based on microarray genome comparison, the possibility that DNA acquisition by horizontal transfer may also have played a role in host adaptation of B. bronchiseptica complex IV cannot be ruled out completely. To identify horizontally acquired genes specifi c for B. bronchiseptica complex IV, we used subtractive hybridization. To this purpose, the genome of a complex I strain (RB50) was “subtracted” from a mixture of the genomes of fi ve B. bronchiseptica complex IV strains, and the distribution of sequences not present in B. bronchiseptica RB50 was determined over a collection of complex I and IV strains. A limited number of complex IV-specifi c sequences were detected, suggesting that large scale gene acquisition by horizontal transfer has not played a major role in the evolution of complex IV strains.

70 Subtractive Hybridization introduction

Bordetella pertussis and Bordetella parapertussishu, the causative agents of whooping cough in humans, have independently evolved from a Bordetella bronchiseptica-like ancestor (1,2, chapter 2). Genome comparison indicated that the evolution and host-adaptation of B. 3 pertussis and B. parapertussishu was accompanied by extensive genome decay . Especially the genome of B. pertussis contains a very high percentage of pseudogenes (9.4%), and extensive reshuffl ing has occurred relative to B. bronchiseptica RB50 through IS481- mediated transposition events. Th e genome reduction is also refl ected in the genome sizes 4 of B. bronchiseptica RB50, B. parapertussishu 12822, and of B. pertussis Tohama, which are 5.3, 4.8 and 4.1 million bases, respectively. Th e genome of B. pertussis Tohama contains 1,161 less predicted genes than the genome of B. bronchiseptica RB50. Further, there are no indications of large-scale gene acquisition in B. pertussis compared to B. bronchiseptica RB50, and based on that observation, Parkhill et al. 3 suggested that host adaptation of B. pertussis and B. parapertussishu was associated with of genome reduction. It was shown by multilocus sequence typing (MLST) that B. bronchiseptica comprised two distinct lineages, designated complex I and complex IV (Chapter 2), isolated mainly from animals and humans, respectively. Based on MLST and pertactin sequence data, B. pertussis was more closely related to B. bronchiseptica complex IV than to B. bronchiseptica complex I. Comparative genomic hybridization (CGH) to microarrays identifi ed 30 genes that were overrepresented in the complex IV strains compared to complex I (Chapter 3). Sixteen of these 30 genes were previously believed to be unique to B. pertussis 3,4. Conversely, a total of 237 genes were identifi ed to be underrepresented in B. bronchiseptica complex IV strains compared to complex I, and these genes were in most cases absent from the genome of B. pertussis Tohama as well. Th ese observation further underline the closer relationship of B. pertussis to complex IV strains compared to complex I strains. Because gene acquisition has been suggested to have played a minor role in the evolution and host-adaptation of B. pertussis and B. parapertussis 3, the analysis of CGH data may suggest that the genomes of B. bronchiseptica complex IV have decayed compared to B. bronchiseptica complex I. However, CGH analysis is limited to genes that are represented on the array. To identify horizontally acquired genes specifi c for B. bronchiseptica complex IV, we used subtractive hybridization. To this purpose, the genome of a complex I strain (RB50) was “subtracted” from a mixture of the genomes of fi ve B. bronchiseptica complex IV strains. Th e B. bronchiseptica RB50 strain was chosen because its genome has been sequenced 3. Subsequently, the distribution of the identifi ed sequences, which were potentially unique for complex IV strains, was determined by PCR. Th is approach identifi ed a limited number of unique complex IV specifi c sequences, suggesting that large scale gene acquisition by horizontal transfer has not played a major role in the evolution of complex IV strains.

71 Chapter 

experimental procedures

bacterial strains To identify sequences specifi c to complex IV in comparison to complex I, fi ve B. bronchiseptica complex IV strains (B0232 (ST18), B2114 (ST25), B2494 (ST15), B2495 (ST17) and B2496 (ST3)) were compared to B. bronchiseptica RB50 (B1976, ST12) by 4 subtractive hybridization. For strain characteristics, see Table 1. Strains were cultured on Bordet Gengou agar supplemented with 15% sheep blood (BD, Franklin Lakes, NJ) at 37°C for one to two days. Chromosomal DNA was isolated using the Wizard Genomic DNA Purifi cation Kit (Promega, Madison, WI), according to the manufacturers’ protocol for Gram-negative bacteria.

Table 1. Characteristics of the strains used in the subtractive hybridization. Number Species1 Complex ST IS481 IS1001 IS1002 IS1663 Host Original name B1976 BB I 12 - - - - rabbit RB50 B0232 BB IV 18 - - - - human 397 B2114 BB IV 25 - - - + human 99-003 B2494 BB IV 15 - - - + human MO149 B2495 BB IV 17 - - - + human MO211 B2496 BB IV 3 - - - + human MO275

¹Abbreviations: BB= B. bronchiseptica; BPP-OV= B. parapertussisov; BPP-HU= B. parapertussishu; BP= B. pertussis

subtractive hybridization To identify complex IV-specifi c sequences, the genome of B. bronchiseptica RB50 (B1976; driver) was subtracted from the genomes of fi ve pooled B. bronchiseptica complex IV strains (consisting of equimolar amounts of B0232, B2114, B2494, B2495 and B2496; tester). B. bronchiseptica B1976 is a complex I strain, and its genome has been published 3. For each experiment, the PCR-Select Bacterial Genome Subtraction Kit (BD Biosciences Clontech) was used, with modifi cations to the standard protocol. First, 2 μg of genomic DNA of both tester and driver DNA was digested to completion in One-Phor-All Buff er (Amersham) with Alu I (Invitrogen), a four base pair cutter that generates 16552 restriction fragments in B. bronchiseptica RB50, with an average restriction fragment size of 323 nucleotides. Digestion with Alu I does not lead to restoration of the restriction site when the adaptors are ligated to the restriction fragments. After phenol extraction and ethanol precipitation, the tester DNA sample was divided into two samples. Adaptor 1 and 2R were ligated to 150 ng of each Tester DNA sample, respectively. An unsubtracted sample was generated as well, which served as a negative control for subtraction and consisted of 1.5 μl of Tester DNA Ligation mix with Adaptor 1 and 1.5 μl of Tester DNA Ligation mix with Adaptor 2R. Th e ligation mixture was incubated at 16°C for 16 hr. After incubation, 1 μl of 0.2 M EDTA (ethylenediaminetetraacetic acid) was added to stop the ligation reaction, followed by 5 min at 72°C to inactivate the ligase. A PCR control was performed to

72 Subtractive Hybridization determine the ligation effi ciency. In this PCR, a primer designed to hybridize to a randomly selected Alu I restriction fragment was combined to a primer designed to hybridize to the adaptors (supplied in the kit). Two hybridization rounds were performed with the ligated tester DNA. In the fi rst hybridization round, the two ligated tester DNA samples were each hybridized to a 27 times excess of driver DNA (400 ng of driver DNA to 15 ng of tester DNA) at 65°C for 1.5 hr. In this hybridization round, non-specifi c tester DNA sequences hybridize to the driver DNA, thereby enriching for tester-specifi c sequences. Th e second hybridization round was performed directly following the fi rst round; the two tester 4 DNA samples to which the diff erent adaptors were ligated were hybridized to each other in the presence of newly added, freshly denatured driver DNA. Th is step was performed to further enrich for tester-specifi c sequences. Hybridization was performed for 16 hr at 65°C. After the second hybridization, 150 μl of dilution buff er (Clontech) were added to the sample, and incubated for 7 min at 63°C to prevent non-specifi c hybridization. Selective amplifi cation of tester-specifi c sequences was performed by PCR. Th e mixture of the primary PCR consisted of 2.5 μl 10x BD Advantage 2 PCR buff er (Clontech), 0.5 μl 50x dNTP mix, 0.5 μl 50x BD Advantage 2 polymerase mix, 5 μl 5M betain (Sigma-

Aldrich), 1.3% dimethylsulfoxide (DMSO), 10 pmol PCR primer 1, 12.2 μl dH2O and 3 μl of the diluted subtracted sample. Th ermal cycling was as follows: 3 min 30 s at 72°C to extend the adaptors, 25 cycles of 1 min at 4°C, 30 s at 60°C, 2 min at 72°C. A ﬁ nal extension was performed for 7 min at 72°C. A secondary (nested) PCR was performed to increase the amount of tester-speciﬁ c sequences, consisting of 12.5 μl of HotStar Taq Master Mix (Qiagen), 1 μl of nested primer 1, 1 μl of nested primer 2R, 5 μl of 5M betain

(Sigma-Aldrich), 1.3% DMSO, 4.2 μl dH2O and 1 μl of the 10x diluted primary PCR sample. Cycling conditions were: 15 min at 95°C, 20 cycles of 1 min at 94°C, 30 s at 62°C, 2 min at 72°C. A fi nal extension step was performed for 30 min at 72°C, which generates the 3’A overhang that was used to clone the PCR fragments into the TOPO TA cloning vector (Invitrogen). cloning and sequencing of pcr fragments PCR fragments were purifi ed with the Qiaquick PCR purifi cation kit (Qiagen). Subsequently, the PCR fragments were cloned into the TOPO-TA cloning vector (Invitrogen) and transformed into One Shot Chemically Competent Cells (Invitrogen), according to the manufacturer’s protocol. Blue-white screening was performed to identify clones with inserts. Colony PCR was performed on white colonies using M13 primers (5’-GTT-GTA-AAA- CGA-CGG-CCA-GT-3’ and 5’-CAG-GAA-ACA-GCT-ATG-ACC-3’) and PCR products were analyzed by agarose gel electrophoresis. PCR products were purifi ed using ExoSAP- IT (USB, Cleveland, OH). All sequencing was performed with the M13 forward using standard dye-terminator chemistry.

73 Chapter 

data analysis Sequence trace data were analyzed using Kodon version 2.5 (Applied Maths, Sint-Martens- Latem, Belgium). Contigs were assembled using Kodon and SeqMan (DNASTAR, Inc., Madison, WI).

distribution of tester-specific sequences 4 For sequences that were potentially speciﬁ c to complex IV, the distribution was determined over six complex I (B0188, B0194, B0251, B1976, B2499 and B2501) and six complex IV strains (B0259, B2114, B2490, B2494, B2495 and B2506) by PCR. For these sequences, primers were designed and PCR ampliﬁ cation was performed with genomic DNA from various strains (a list with primer information is available from the authors).

74 Subtractive Hybridization results Gene content comparison of B. bronchiseptica complex IV and complex I strains by CGH suggested that the complex IV strains have undergone genome reduction compared to complex I (Chapter 3). However, one of the limitations of CGH is that it can detect only sequences that are represented on the array. Here we investigated whether complex IV strains contain additional DNA sequences compared to complex I. To enrich for sequences shared by multiple complex IV strains, the DNA of fi ve distinct complex IV (B0232, B2114, B2494, 4 B2495 and B2496) strains was combined and from this, the genome of the B. bronchiseptica complex I strain RB50 was subtracted by hybridization. By using B. bronchiseptica RB50 as the driver strain in the subtractive hybridization, identifi cation of non-RB50 sequences was facilitated, due to the fact that the genome of RB50 has been sequenced completely. Following subtractive hybridization, tester-specifi c sequences were further enriched by suppression PCR and PCR products were transformed into competent cells. Th e nucleotide sequences of 305 clones were determined. For the identifi cation of sequences present in the genome of RB50, a BLASTN search against the non-redundant GenBank database was performed. Nucleotide analysis showed that the 305 sequences contained 183 unique and 122 redundant sequences. Based on overlap of fl anking sequences, contigs were assembled. Out of the 183 unique sequences, 33 sequences could be assembled into 12 contigs; for the remaining 150 sequences, no overlap with other sequences could be detected. Fifty-six % of the 183 sequences (5 contigs representing 10 sequences and 92 unique sequences) showed >99% identity to genes present in the genome of B. bronchiseptica RB50. Th e average GC-content of these sequences was 60.9%, which is signifi cantly lower than the average GC-content of the B. bronchiseptica RB50 genome (68.1%) 3. We also identifi ed 4 contigs (14 sequences) and 2 sequences not assigned to a contig that showed homology to IS1663, which was previously shown to be present in complex IV strains (Chapter 2 and 3). Analysis of sequences adjacent to these IS1663 elements yielded seven diff erent RB50-encoded genes, as well as two sequences with no signifi cant homology to known genes. Previously, it was reported that B. pertussis Tohama contains 17 IS1663 copies 3, and comparison of the genes adjacent to IS1663 in B. bronchiseptica complex IV and B. pertussis Tohama indicated that the sequences fl anking IS1663 were diff erent in these two lineages. For 65 sequences, no homologous sequence could be identifi ed in the genome of B. bronchiseptica RB50, although for one contig (containing three clone sequences) and four single sequences, homologous genes were detected in B. pertussis Tohama (>99% sequence identity), including BP0207, which was previously shown to be specifi c to B. pertussis 4 and was shown to be overrepresented in complex IV strains by CGH analysis (Chapter 3). Five clone sequences showed homology to B. parapertussis genes (>99%), and three of these were related to O-antigen biosynthesis. In chapter 3, polymorphism in the O-antigen biosynthesis locus of complex IV strains was demonstrated by CGH analysis. It was shown

75 Chapter 

that some complex IV strains contained an alternative O-antigen biosynthesis locus. We identifi ed 47 sequences that showed no homology to any of the mammalian bordetellae, and the average GC-content was 55.7%, signifi cantly lower than that of any of the sequenced Bordetella species. As these 47 sequences were potentially specifi c to complex IV, their presence in six complex I (B0188, B0194, B0251, B1976, B2499 and B2501) and six complex IV strains (B0259, B2114, B2490, B2494, B2495 and B2506) was determined by PCR amplifi cation. Th e distribution was performed for two contigs (six sequences) as 4 well as for 34 single sequences. For one of two contigs (three sequences) and four out of 34 sequences, PCR screening indicated that these sequences were present in fi ve of six complex IV strains, and absent from all tested complex I strains. For the other sequences of which the distribution was determined, PCR indicated the sequence was present in a single B. bronchiseptica complex IV strain only. BLASTN indicated no signifi cant homology to known genes for these sequences, with the exception of one sequence that showed homology (E = 1e-111) to a hypothetical cytosolic protein encoding gene involved in arsenite oxidation in the species Alcaligenes faecalis, which is closely related to the genus Bordetella 5. For the remaining sequences of which the distribution was not determined, a putative function could not be identifi ed by homology search to known genes or proteins.

76 Subtractive Hybridization discussion Although comparative genomics of B. bronchiseptica complex IV strains and complex I strains suggested genome reduction of the complex IV strains, the possibility remained that acquisition of DNA may also have contributed to the evolution of B. bronchiseptica complex IV. In this study, the extent of DNA acquisition in complex IV strains in comparison to B. bronchiseptica complex I was investigated by comparing their genomes using subtractive hybridization. Th e RB50 strain was chosen as driver strain because of 4 the availability of its genome sequence, which facilitated the identifi cation of sequences specifi c to B. bronchiseptica complex IV. Screening of 305 clone sequences showed that 56% showed nearly 100% homology to genes in B. bronchiseptica RB50. Th e GC-content of these sequences was signifi cantly lower than the average GC-content of RB50, suggesting that sequences with divergent GC-content are possibly preferentially enriched using this method. Sequencing of 305 clones detected a high level of redundancy, and after removal of the redundant clones, 183 unique clones remained, which could be clustered into 12 contigs and 150 clones that, based on lack of overlap, could not be assigned to a contig. Of the 183 clones, 56% showed homology (>99%) to genes in B. bronchiseptica RB50. Th e relatively low GC-content (60.9%) of these clones suggested that sequences with a divergent GC- content may be more easily detected than sequences with a GC-content similar to the driver strain. Besides genes with homology to B. bronchiseptica RB50, we also isolated 10 clones with homology to genes in B. pertussis Tohama and B. parapertussis 12822. Most of the genes assumed to be specifi c to B. pertussis and B. parapertussis based on genome comparison 3, were later shown to be present in many other (non-RB50) B. bronchiseptica strains by CGH 4. Th e majority of the genes in B. parapertussis 12822 that were homologous to the clones obtained with subtractive hybridization were related to O-antigen biosynthesis. Th is can be explained by the diff erences at the O-antigen biosynthesis locus between B. bronchiseptica RB50 and B. bronchiseptica complex IV (see Chapter 3). Of the 183 unique clones, 47 showed no homology to genes in the genomes of the mammalian bordetellae. In only fi ve clones (one contig and four unique clones), PCR amplifi cation showed that the sequences were present in multiple complex IV strains. In most cases, a putative function for these clones could not be determined, due to lack of homology with nucleotide and protein databases. Th us, the majority of the clones that were found to be present in B. bronchiseptica complex IV and absent in RB50 were detected only in a single complex IV isolate. Th is suggested that (small) genomic diff erences exist between the complex IV strains, which may be expected due to the previously observed high genetic diversity in complex IV (discussed in Chapter 2). However, the failure to detect a high amount of sequences specifi c to complex IV strains suggests that large-scale DNA acquisition by horizontal transfer has not likely played a major role in the evolution of the B. bronchiseptica complex IV strains.

77 Chapter 

In Chapter 2 and 3, it was shown that the B. pertussis insertion sequence element IS1663 was also present in the majority of B. bronchiseptica complex IV isolates. Th e subtractive hybridization experiment resulted in 33 IS1663-containing clones. Eight of these clones were found to contain sequences fl anking IS1663. Seven of these fl anking sequences showed homology to genes in B. bronchiseptica RB50, and for one sequence, no homology was found. However, none of the fl anking sequences showed homology to sequences adjacent to the 17 B. pertussis Tohama IS1663 copies. Th is lack of homology may argue for lateral 4 transfer of IS1663 between B. pertussis and B. bronchiseptica complex IV instead of common descent. However, this issue requires further investigation, e.g. by analyzing all IS1663 fl anking sequences in both B. bronchiseptica complex IV and B. pertussis. In summary, we detected limited DNA acquisition in complex IV strains compared to B. bronchiseptica RB50 (25.6% of the clones isolated by subtractive hybridization), and in most cases these sequences were strain-specifi c and not shared by the whole complex. Th erefore, it seems unlikely that DNA acquisition has played an important role in the evolution of complex IV, although it may still have been important for niche-adaptation at the strain-level. Although our data suggest that the genomes of B. bronchiseptica complex IV strains have been reduced compared to B. bronchiseptica complex I strains, the ultimate proof lies in the sequencing of members of B. bronchiseptica complex IV and comparison of these sequences to the genome sequences of B. bronchiseptica complex I strains.

78 Subtractive Hybridization reference list

1. van der Zee,A., Mooi,F., van Embden,J., Musser,J. (1997). Molecular evolution and host adaptation of Bordetella spp.: phylogenetic analysis using multilocus enzyme electrophoresis and typing with three insertion sequences. The Journal of Bacteriology 179, 6609-6617. 2. Musser,J.M., Hewlett,E.L., Peppler,M.S., Selander,R. 4 K. (1986). Genetic diversity and relationships in populations of Bordetella spp. The Journal of Bacteriology 166, 230-237.

3. Parkhill,J., Sebaihia,M., Preston,A., Murphy,L.D., Thomson,N., Harris,D.E., Holden,M.T., Churcher,C. M., Bentley,S.D., Mungall,K.L., Cerdeno-Tarraga,A. M., Temple,L., James,K., Harris,B., Quail,M.A., Achtman,M., Atkin,R., Baker,S., Basham,D., Bason,N., Cherevach,I., Chillingworth,T., Collins,M., Cronin,A., Davis,P., Doggett,J., Feltwell,T., Goble,A., Hamlin,N., Hauser,H., Holroyd,S., Jagels,K., Leather,S., Moule,S., Norberczak,H., O’Neil,S., Ormond,D., Price,C., Rabbinowitsch,E., Rutter,S., Sanders,M., Saunders,D., Seeger,K., Sharp,S., Simmonds,M., Skelton,J., Squares,R., Squares,S., Stevens,K., Unwin,L., Whitehead,S., Barrell,B.G., Maskell,D. J. (2003). Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat Genet. 35, 32-40.

4. Cummings,C.A., Brinig,M.M., Lepp,P.W., Van De,P. S., Relman,D.A. (2004). Bordetella species are distinguished by patterns of substantial gene loss and host adaptation. J Bacteriol. 186, 1484-1492.

5. Gerlach,G., von Wintzingerode,F., Middendorf,B., Gross,R. (2001). Evolutionary trends in the genus Bordetella. Microbes and Infection 3, 61-72.

79 80

Chapter Chapter 55

Transfer of an Iron-Uptake Island From Bordetella pertussis to Bordetella holmesii

Dimitri A. Diavatopoulos1,2‡, Craig A. Cummings3,4‡, Han van der Heide1, Marjolein van Gent1, Sin-Yee Liew3,4, David A. Relman3,4,5, Frits R. Mooi1,2

‡ These authors contributed equally to this work

Submitted for publication Chapter 

abstract Th e recent discovery of Bordetella holmesii from the blood and airways of patients suggests that this species may represent a newly emerged human bacterial pathogen. However, the mechanism by which B. holmesii may have acquired the ability to infect humans is not understood. Furthermore, although the B. holmesii 16S rRNA sequence is most similar to that of B. pertussis, other evidence suggests a more distant relationship of these species. In order to address these questions, several B. holmesii genomes were analyzed by multilocus sequence analysis, and by comparative genome hybridization (CGH) using a microarray representing B. pertussis, B. parapertussis and B. bronchiseptica. Sequencing of housekeeping 5 genes indicated that B. holmesii is more closely related to B. hinzii and B. avium than to B. pertussis, and that B. holmesii strains are very closely related to each other. Overall, CGH indicated substantial sequence divergence between the genomes of B. pertussis and B. holmesii. However, we identiﬁ ed a putative pathogenicity island, encoding the biosynthesis, export, and uptake of the siderophore alcaligin, that showed near sequence identity to the orthologous B. pertussis genes, and that contained IS481 elements, suggesting that it may have been laterally acquired from B. pertussis. Expression of alcaligin biosynthesis and uptake genes in B. holmesii was induced in the absence of iron, and secreted alcaligin was detected, suggesting that this locus is functional. Th e B. holmesii locus also contains a fecIR- like locus adjacent to the alc operon that may have been acquired through a transposon- mediated recombination event, and that may contribute to transcriptional regulation of iron uptake. A possible role for this pathogenicity island in the evolution and emergence of B. holmesii as a human pathogen is proposed. Horizontal gene transfer between B. pertussis and B. holmesii may also explain the unusually high sequence identity of their 16S rRNA genes, and explain the failure of the 16S-based phylogeny to recapitulate the phylogeny based on housekeeping gene sequences.

82 Acquisition of Iron-Uptake Genes introduction Bordetella holmesii is a recently described human pathogen originally isolated from the blood of septicemic patients. Although most B. holmesii systemic infections occur in immunocompromised individuals 1, a serious systemic infection of a healthy adolescent has also been described 2. More recently, B. holmesii has also been isolated from the respiratory tracts of immunocompetent patients with coughing symptoms. For example, 0.6% of nasopharyngeal swab samples collected from patients with cough by the Massachusetts State Laboratory in 1998 tested positive for B. holmesii 3. In addition to B. holmesii, the genus Bordetella contains a number of other pathogenic species. Bordetella pertussis is the 5 causative agent of pertussis, which, in spite of widespread vaccination, is re-emerging in countries with traditionally high vaccination coverage. B. pertussis is very closely related to B. bronchiseptica and B. parapertussis, which also cause respiratory disease in mammalian hosts. Together, these three species are designated the mammalian, or classical, bordetellae. Th e non-classical Bordetella species include B. avium, B. hinzii, B. petrii and B. trematum, which form a genetically diverse group that is clearly distinct from the mammalian bordetellae 4. B. avium is the most thoroughly studied non-classical Bordetella species, due to the fact that it is an important respiratory pathogen of fowl 5. Comparative analysis of 16S rRNA sequences suggested that B. holmesii is very closely related to B. pertussis 6, a hypothesis that was supported by the discovery of the B. pertussis- specifi c insertion sequence element IS481 in B. holmesii 7. However, subsequent sequencing of housekeeping genes, analysis of cellular fatty acid composition, and characterization of the bvgAS locus suggested that B. holmesii may not be as closely related to B. pertussis as was fi rst assumed 6,8. In order to assess the sequence similarity of B. holmesii to the mammalian bordetellae at a high resolution, genome-wide level, B. holmesii strains were analyzed by comparative genome hybridization (CGH) to a “classical” Bordetella microarray. Th ese data indicated that the majority of B. holmesii genes did not hybridize to the Bordetella microarray, suggesting signifi cant sequence divergence between B. holmesii and the mammalian bordetellae. However, CGH detected a highly conserved genomic region of 66 kb that is proposed to have been transferred from B. pertussis to B. holmesii. Th is iron-uptake island (IUI) encoded the biosynthesis, export, and uptake of the siderophore alcaligin, as well as several IS481 elements. Further characterization of the island identifi ed a transposase-mediated insertion adjacent to the alcaligin operon that encoded a FecIR-like system that may be involved in regulation of iron-uptake. Transcription of the alcaligin operon and the production of secreted alcaligin were demonstrated under iron-depleted conditions. Sequencing of housekeeping genes indicated that B. holmesii is a uniform group that is more closely related to the non-classical B. hinzii and B. avium than to the mammalian bordetellae. Th e presence of 16S rRNA genes similar to those in B. pertussis in the genome of B. holmesii may be explained by horizontal gene transfer between B. pertussis and B.

83 Chapter 

holmesii. Horizontal gene transfer may also be a plausible explanation for the presence of similar 16S rRNA genes in the genomes of B. pertussis and B. holmesii.

84 Acquisition of Iron-Uptake Genes experimental procedures bacterial strains and culture conditions Characteristics of the strains used in this study are listed in Table 1. Strains were isolated either from the blood or from the respiratory tract of patients from Europe or the United States. Strains were routinely grown on BG agar, supplemented with 15% sheep blood (BD, Franklin Lakes, NJ, USA) at 37°C for 3-5 days. Genomic DNA was isolated using the Wizard Genomic DNA Puriﬁ cation Kit (Promega, Madison, WI) following the protocol for gram-negative bacteria. For iron depletion tests, a previously described chemically 5 deﬁ ned medium 9 was adapted by adding 330 μM L-cysteine, 114 μM ascorbic acid, 33 μM niacin, and 325 μM reduced glutathione. Chelex-100 resin was used for iron depletion, as described previously 9. Iron-replete media was made by adding 66 μM ferrous sulfate to iron-depleted media. Table 1. B. holmesii strain characteristics

Strain Original Name Source Isolation site Country City Isolation Year B0436 104394 Institut Pasteur unknown unknown unknown unknown B0437 104395 Institut Pasteur unknown unknown unknown unknown B1850 C690 Pam Cassiday Nasopharynx USA Massachusetts 1994 B1851 C691 Pam Cassiday Blood USA Massachusetts 1995 B1852 C692 Pam Cassiday Blood USA Ohio unknown B1853 C693 Pam Cassiday Blood USA Colorado unknown B1854 C694 Pam Cassiday Blood USA North Carolina unknown B1855 USA 8/8/00 Cathy Canthaboo unknown unknown unknown unknown Newport, Isle of B2738 BP244-01 Norman Fry Blood United Kingdom 2001 Wight B2739 BP246-01 Norman Fry Blood United Kingdom London 2001 B2767 RR 0000 0020 Norman Fry Blood United Kingdom unknown 2003 B2768 HO 4290 0199 Norman Fry Blood United Kingdom Oxford 2004 Bho29 G8350 Robbin Weyant Blood Switzerland unknown unknown comparative genomic hybridization Experiments utilized a comprehensive classical Bordetella genome-wide DNA microarray, based on the microarray described in 10, but with the inclusion of additional ORF probes and resulting, improved genome coverage. With 5,670 PCR products, 97.4% of the B. pertussis Tohama I ORFs and 98.5% of the B. bronchiseptica RB50 ORFs are represented. Probes were printed in duplicate on poly-L-lysine-coated glass slides as previously reported 10. B. holmesii genomic DNA was labeled with Cy5, and hybridized to arrays with Cy3-labeled reference DNA (equimolar amounts of B. pertussis Tohama I, B. bronchiseptica RB50 and B. bronchiseptica 12822 genomic DNA) as described in 10. Arrays were scanned using a GenePix 4000B scanner (Axon Instruments, Union City, CA) and analyzed with GenePix Pro version 6 (Axon Instruments). Data were ﬁ ltered to include only spots containing more than 30 pixels and with a mean background-subtracted Cy3 signal above 150 U.

85 Chapter 

Table 2. Characteristics of the ORFs in the B. holmesii IUI region.

Gene name B. pertussis ORF B. avium1 Gene product rpsO, secC BP0794 yes 30S ribosomal protein S15 BP0793 BP0793 yes putative lipoprotein pssA BP0792 yes putative CDP-diacylglycerol-serine O-phosphatidyltransferase protein ilvC BP0791 yes ketol-acid reductoisomerase ilvH, brnP BP0790 yes acetolactate synthase small subunit ilvI BP0789 yes acetolactate synthase large subunit BP0787/ BP0787/BP2450 yes N-and C-terminal regions of a putative membrane protein BP2450 BP2451 BP2451 no putative membrane transport protein 5 BP2452 BP2452 yes conserved hypothetical protein IS481 BP2453 no transposase for IS481 element BP2454 BP2454 yes putative oxidoreductase BP2455 BP2455 no putative membrane effl ux protein alcA BP2456 no alcaligin biosynthesis enzyme alcB BP2457 no alcaligin biosynthesis protein alcC BP2458 no alcaligin biosynthesis protein alcD BP2459 no hypothetical protein alcE BP2460 yes2 putative iron-sulfur protein alcR BP2461 no transcriptional regulator alcS BP2462 no alcaligin exporter fauA BP2463 no ferric alcaligin siderophore receptor mar BP2464 no possible membrane effl ux protein bhoA n/d no putative IS3-family transposase bhoB n/d no uncharacterized protein UPF0065 [Ralstonia metallidurans CH34] D-isomer specifi c 2-hydroxyacid dehydrogenase, NAD-binding [Ral- bhoC n/d no stonia metallidurans CH34] bhoD n/d no σ70 factor (ECF subfamily), FecI protein bhoE n/d no FecR protein BP2465 BP2465 yes hypothetical protein BP2466 BP2466 yes hypothetical protein BP2467 BP2467 no hypothetical protein vrg-6 BP2468 yes B. pertussis Bvg-repressed gene of unknown function BP2469 BP2469 yes hypothetical protein BP2470 BP2470 yes seryl-tRNA synthetase BP2471 BP2471 yes conserved hypothetical protein BP2472 BP2472 yes putative exported protein ftsK BP2473 yes putative cell division protein trxB BP2474 yes thioredoxin reductase BP2475 BP2475 yes hypothetical protein BP2476 BP2476 yes putative membrane transport protein IS4813 BP2477 no transposase for IS481 element BP2478 BP2478 yes putative membrane transport protein BP2479 BP2479 yes putative integral membrane protein kdpA BP2480 yes potassium-transporting ATPase A chain kdpB BP2481 yes potassium-transporting ATPase B chain kdpC BP2482 yes potassium-transporting ATPase C chain kdpD BP2483 yes two component sensor protein kdpE BP2484 yes two component system transcriptional regulatory protein BP2486 BP2486 no putative exported protein BP2487 BP2487 no hypothetical protein icd, icdA, icdE BP2488 yes isocitrate dehydrogenase [NADP] BP2489 BP2489 yes conserved hypothetical protein BP2490 BP2490 no putative cytochrome

86 Acquisition of Iron-Uptake Genes

Gene name B. pertussis ORF B. avium1 Gene product BP2490 BP2490 no putative cytochrome BP2491 BP2491 no putative cytochrome BP2493 BP2493 yes putative ATP-binding protein BP2494 BP2494 yes conserved hypothetical protein BP2495 BP2495 no conserved hypothetical protein BP2496 BP2496 yes hypothetical protein BP2497 BP2497 yes putative zinc protease dnaJ BP2498 yes molecular chaperone dnaK BP2499 yes molecular chaperone

¹Orthologue in the genome of B.avium identifi ed by BLASTN ²Homologue found outside the B. avium IUI syntenic region. 5 3Variably present among B. holmesii strains Background-subtracted Cy5/Cy3 ratios were averaged for replicate probes. Because most of the probes only had signal in the reference channel for the B. holmesii hybridizations, data could not be normalized by setting the mean Cy5/Cy3 ratio, or the mode of the ratios for “detected” genes, equal to 1. Instead, normalization was achieved by calculating the mean background-subtracted Cy5/Cy3 ratio for 16 probes within the IUI that were verifi ed by sequencing to be at least 99% identical to the B. holmesii genome, then dividing the background-subtracted Cy5/Cy3 ratio for each probe by this value. pcr and dna sequencing methodology PCR primer pairs were designed using Primer3 11 and Kodon 2.5 (Applied Maths, Sint-Martens-Latem, Belgium). PCR products were purifi ed prior to sequencing using the QIAquick PCR purifi cation Kit (Qiagen, Valencia, CA), or by ExoSAP-IT (USB Corporation, Cleveland, OH), according to the manufacturer’s instructions. All sequencing was performed using standard dye-terminator chemistry. Unless otherwise indicated, sequencing was performed directly from PCR products using one of the amplifi cation primers as the sequencing primer. DNA sequence data were analyzed using Kodon software package version 2.5 (Applied Maths, Sint-Martens-Latem, Belgium) and Sequencher version 4.2 (Gene Codes, Ann Arbor, MI). All amplifi cation and sequencing primers used in this study are listed in Table 3. organization of iui Primers were designed, based on the B. pertussis Tohama I sequence, to amplify overlapping 5-10 kb DNA fragments of IUI. PCR reactions were carried out on all B. holmesii isolates and Tohama I. PCR samples were analyzed by agarose gel electrophoresis to compare the fragment sizes. If the size of the PCR amplicons was similar to Tohama I, and the presence of these genes was confi rmed by CGH, the ORF organization of that particular amplicon was assumed to be similar between Tohama I and the tested B. holmesii strain. It should be noted that small insertions or deletions, or point mutations resulting in premature stop codons cannot be detected by CGH or comparison of PCR fragment sizes by agarose gel

87 Chapter 

electrophoresis. Th erefore, the possibility remains that some of the genes in IUI are actually pseudogenes. A total of 21 genes in IUI were partially sequenced from three B. holmesii isolates. Primer pairs used to amplify and sequence these genes were the same as those used to amplify the corresponding microarray probes. Mapping of IUI in B. holmesii by PCR identiﬁ ed a 4.8 kb insert in all B. holmesii strains examined. Th e nucleotide sequence of this DNA insert was determined by genome walking inwards from the adjacent genes. In order to assess the organization of the BB3888 homologue in B. pertussis and B. holmesii, PCR primers were designed in BB3888 that span 5 the IS481 insertion and rearrangement point from B. pertussis Tohama I.

determination of sequences flanking iui Nucleotide sequence data adjacent to the 5’ breakpoint of IUI was obtained by PCR using one primer inside IUI, in BB0794, and one primer outside IUI, in BB0795, which was not detected by CGH. Th e outside primer was designed in a region that is identical between all four sequenced Bordetella genomes as assessed by multiple sequence alignment with ClustalW. To obtain nucleotide sequences adjacent to the 3’ breakpoint of IUI, the TOPO Walker Kit (Invitrogen) was used. Genomic DNA from isolate B2768 was digested with the restriction enzyme Sph I, and concentrated by ethanol precipitation. Th e adaptor supplied in the kit was ligated to the restriction fragments according to the manufacturer’s instructions. Primer extension was performed using a gene-speciﬁ c primer designed to hybridize to the 5’ region of BP2499. Sequence data were obtained for 555 nucleotides downstream of BP2499. To determine the exact location of the breakpoint, PCR was performed using one primer in the 3’ region of BP2499, and one primer in the region downstream of BP2499, and the PCR product was sequenced by genome walking.

phylogenetic analysis PCR primers for the phylogenetic analysis were designed to hybridize to conserved regions of the atpD, rpoB, tuf and rnpB genes, as determined by multiple sequence alignment of B. bronchiseptica RB50, B. pertussis Tohama I, B. parapertussis 12822, and B. avium 197N sequences with ClustalW 12 (Table 3). Amplifi cation of the expected products was confi rmed in each of the four genomes from which the primers were designed. Most primer pairs also amplifi ed a product of similar size from all seven B. holmesii strains tested, as well as single isolates of the related species, B. hinzii, B. trematum, B. petrii, and Achromobacter xylosoxidans. In cases where PCR yielded a product from all strains examined, the nucleotide sequences of these PCR products were determined. Concatenated sequences of atpD, rpoB, tuf and rnpB genes were aligned using ClustalW 12 and used to construct trees with the NEIGHBOR program of the PHYLIP suite, which was run with the “only aligned

88 Acquisition of Iron-Uptake Genes

Table 3. Characteristics of primers and probes used in this study.

Purpose and target1 Primer name Sequence (5’-3’) Purpose Sequencing genes in the IUI BP0787 - BP0789 intergenic region BP0787 - BP0789 IR1 - F GTGACTATCCGAAGCCGGT A, S BP0787 - BP0789 IR1 - R GCTGGAGGTCGAAATGGAAG A, S BP2479 - BP2480 intergenic region BP2479 - BP2480 IR2 - F CGGAACATTTACCCCGTTTT A, S BP2479 - BP2480 IR2 - R ATTTCTCCGGATTGAACAGC A, S BP0787 BP0787 - F ATATTCTCGGTGTTCGGCAA A, S BP0787 - R CTTCACATCTAGGTCGGCGT A, S BP0792 BP0792 - F CCATCCACGACAGCATCAT A, S BP0792 - R CATGCCATCACCACCCAG A, S BP0794 BP0794 - F AAAAATCCGAAATCGTCGC A, S 5 BP0794 - R CCGAGTTTTTCGATCAGTGC A, S BP2452 BP2452 - F ACCTGGCCTCCGAAGAGTAT A, S BP2452 - R GGAACACCAGCTTGTCGG A, S BP2456 BP2456 - F TAATTACGCCATTGACCACG A, S BP2456 - R GTGGATGTCCATCAAGGGTC A, S BP2457 BP2457 - F TGGGCATGCATCTTCTCAT A, S BP2457 - R GCGTGCAGAAGGCAAGACT A, S BP2458 BP2458 - F TGGTTGAGACGGAACTGATG A, S BP2458 - R CGATAGGGTTGGGCAGATT A, S BP2462 BP2462 - F GCTTTATTCACCTGCGGATT A, S BP2462 - R CTCGATAAATGCTGCAACCA A, S BP2463 BP2463 - F CTCTACACCACGTACCGCCT A, S BP2463 - R CTGGCTGTAGAAGCCGATCT A, S BP2465 BP2465 - F ACCAGCGACGACCTGTGC A, S BP2465 - R ATCAGGTGCAGGCGATTCTT A, S BP2468 BP2468 - F AAAAAGTGGTTCGTTGCTGC A, S BP2468 - R ATGGCCCTTGTAGTAGCGG A, S BP2469 BP2469 - F CTGGTAAGAAACCTGGCCC A, S BP2469 - R CGCTTTTGGCAGGTAATCTC A, S BP2470 BP2470 - F CCAAGACCTACGACCTGGAG A, S BP2470 - R AGGTTCGAGAACGGTCAGG A, S BP2472 BP2472 - F CAGCTGAGCCAGGCGTTT A, S BP2472 - R GGTGAACTGGAATTCGGAGG A, S BP2476 BP2476 - F GAAATCATCGCCACCAGC A, S BP2476 - R GAACACGTTCATGACCACCA A, S BP2482 BP2482 - F CCAGTCGTTCACCGATCC A, S BP2482 - R GTTGAGGGCAAGCACATTG A, S BP2486 BP2486 - F GCGAACCCCACAATGATTT A, S BP2486 - R ACCAGCGTGGCACTGTAGAT A, S BP2491 BP2491 - F CGTGACAAGGTGTCGATGTG A, S BP2491 - R AGATGAGTAGTAGGCCGCCA A, S

Real-time RT-PCR BP2456 (alcA) AlcA-fw GGATCTGGGCTTCTGCTGC A AlcA-rev CGCGGTGGATGTCCATCAA A AlcA-FL CATCGAGACACGAACCGCCTTGC-F H AlcA-LC GGAGTTCTCTCCACCAGCAGACGG-L H BP2463 (fauA) FauA-fw CTGACGACCAGCGTGGAG A FauA-as CGGTACACATGTTTCATGAGATAGC A FauA-FL GCAGGATACCGAAGCCACCACCTA-F H FauA-LC TTCGTGGACCTCACGCACCGC-L H BP0015 (rpoB) RpoB-se CCCGAAGACTTCCTCTTTGGTC A RpoB-rev TCGTAGGACTCTTCGCTGTAGAA A

89 Chapter 

Purpose and target1 Primer name Sequence (5’-3’) Purpose RpoB-FL ATGCCAAGGTGCTGGATCTGCAG-F H CACTCTACACCAACGATCTGGAC- RpoB-LC H CGC-L

Southern blot hybridization bhoA bhoA-1 B-CGCAGGGACAAGAACCAGTGC H B-AACGGCACGGGCTAATATCCT- 16S rRNA (B. pertussis, B. holmesii) 16S-BP-BH-2 H GTG

Phylogenetic analysis BP3288 (atpD) pan-Bord_atpD_5’-for GCCGTGGTGGATATTCAGTT A, S pan-Bord_atpD_5’-rev CCATCATGTTGACGGTCTTG A, S 5 BP0015 (rpoB) pan-Bord_rpoB_1-for AAAGCGCATCCGCAAAAG A, S pan-Bord_rpoB_1-rev GAAGAACAGCACGTCCTTGG A, S pan-Bord_rpoB_2-for GAAGGCCATCGGCATGA A, S pan-Bord_rpoB_2-rev GAGATCGGCTTGGAGTTGAT A, S pan-Bord_rpoB_3-for GATCGAAACGCCGGAAG A, S pan-Bord_rpoB_3-rev GGCGTGAGCTGGGTTTC A, S BP0007 (tuf) pan-Bord_tuf_5’-for GAACGTGGGTACGATTGGTC A, S pan-Bord_tuf_5’-rev CTTCACGATCGGCGTGTC A, S pan-Bord_tuf_3’-for GACACGCCGATCGTGAAG A, S pan-Bord_tuf_3’-rev TAGAACTGCGGACGATAGCC A, S rnpB pan-Bord_rnpB-for AGGAACAGGGCCACAGAGAC A, S pan-Bord_rnpB-rev GCAGATCTATAGGCCGGATTC A, S

Screening for IS481 elements BP0787 - BP2450 BB3888-480f CTTCAGCCGCCTGTTCTG A,S BB3888-720r CATGACCCGGTCGTCTTC A,S BP2453-BP2452 IS481-3’-out GGCTTACGCTCACACCTACC A BP2452-screen-b TGAGGGTTTCCTTGGATTTG A BP2477-BP2478 IS481-3’-out see above A BP2478-screen CGATTATCGTGGTGTCGTTG A BP2485-BP2484 IS481-3’-out see above A BP-kdpE-screen GCTACCTGCTGACGGAAATC A BP2492-BP2493 IS481-3’-out see above A BP2493-screen ATGGTCTTGCGGCAATTATC A

Mapping genomic island BP0790-BP0794 BP0790 - F see above A BP0794 - R see above A BP0790-BP2450 BP0790 - BR GGTGACGTCGATGATGCGCCC A BP2450 - CR GCACCGCGAACAGGATGGTCG A BP2450-BP2452 BP2450 - BF GACGGAGAAGCACCAGGCC A BP2452 - F see above A BP2452-BP2457 BP2452 - R see above A BP2457 - R see above A BP2456-BP2458 BP2456 - F see above A BP2458 - R see above A BP2458-BP2463 BP2458 - F see above A BP2463 - R see above A BP2463-BP2465 BP2463 - F see above A BP2465 - R see above A BP2465-BP2472 BP2465 - F see above A BP2472 - F see above A BP2472-BP2476 BP2472 - R see above A BP2476 - R see above A BP2475-BP2479 BP2475 - F see above A BP2479 - F see above A

90 Acquisition of Iron-Uptake Genes

Purpose and target1 Primer name Sequence (5’-3’) Purpose BP2475-BP2479 BP2475 - F see above A BP2479 - F see above A BP2479-BP2482 BP2479 - R see above A BP2482 - R see above A BP2482-BP2486 BP2482 - BF TGGCGCTGGTCAGCAAAGG A BP2486 - BR TGGTCTATGGACTGGGCGGG A BP2486-BP2494 BP2486 - R see above A BP2494 - R see above A BP2493-BP2499 BP2493 - F see above A BP2499 - F see above A

Screening and sequencing bhoABCDE BP2463 - BP2465 BP2463 - F see above A,S 5 BP2465 - R see above A,S fauA-3’-out AGGTCGAGGGGATAGACCTG A Bho-specifi c-out CTGGCGAGTCGAGTACAACA A mar-BP2465-intergenic ACCTACCCGCCCCTGTGT A Bho-specifi c-out see above A BP2463 - SEQ1 GGCGCTCAACGCGATGTTC S BP2463 - SEQ2 CCGCCCCTGTGTGCAAAATAA S BP2463 - SEQ3 TCGCTTGTGATTTGCTTGCCA S BP2463 - SEQ4 GAGCAAATCATCGGCGTGCTC S BP2463 - SEQ5 CTACACTCGCGAATGCCTGGC S BP2463 - SEQ6 GCCGCATCTGGAGACCCAAG S BP2463 - SEQ7 AAGACGAAGGCCAGGCCCA S BP2465 - SEQ1 CTGCACCTGGCCATTGCTGTT S BP2465 - SEQ2 GCGGGACAGGAAATCGACCA S BP2465 - SEQ3 CCACGATGACCGCATGCTG S BP2465 - SEQ4 CTGATGAAGGAGGCGAACGCTT S BP2465 - SEQ5 TGCCAAAACCCGCAAGGAAC S BP2465 - SEQ6 AGCAGGCCTTCGACATTGCCT S BP2465 - SEQ7 TCTCTCCAGCGCCCGACTG S BP2465 - SEQ8 CCTTGCGGCAGGTCATTGC S BP2465 - SEQ9 CACGCCCCTTTGTGGTGCTTA S

Determining fl anking sequences BP0794 - BP0795 rpsO-screen1 GACGCATGTCCGTAGAGGTC A,S pnp-screen2 GCGGTACTCGCCCATCAG A,S BP2499 - BP2500 BP2499-GSP1 CATCTTGCGCAGCACTTCGG A,S BP2499-2-fl ank-F CGCTATGGCAATCCGTCGGT A,S BP2499-GSP2 GCCAGCTTCTTGCCGCGCA A BP2499-GSP3 TGTCGGCCTTGACGATCGAGT A BP2499-A1 TTCCACCTTGGCGTCGATCTC A,S BP2499-B1 CGGCGTTGGTCGTGGTCTC A,S BP2499-B2 CGTCAAGCGCCTCATCGGT S BP2498-A1 GAAGCCTCGAACTGGCGCA A,S BP2498-B1 GATGTACGCCGACATGCAGGC A,S

Broad-range ribosomal RNA PCR 16S rDNA 8F (broad-range) AGAGTTTGATCCTGGCTCAG A 806R (broad-range) GGACTACCAGGGTATCTAAT A 23S rDNA MS37 (broad-range) AGGATGTTGGCTTAGAAGCAGCCA A MS38 (broad-range) CCCGACAAGGAATTTCGCTACCTTA A Bb_16S-23S_1515-for CGGCTGGATCACCTCCTTTA A,S Bb_16S-23S_1751-rev CTCTCCCAGCTGAGCTACAC A,S Bb_16S-23S_1780-for GTTCGATCCCGTTCACCTC A,S

91 Chapter 

Purpose and target1 Primer name Sequence (5’-3’) Purpose Bb_16S-23S_1949-rev ACGTTGTGCTTCTTCCAAAT A,S Bb_16S-23S_1928-for CAATTTGGAAGAAGCACAACG A,S Bb_16S-23S_2166-rev GATCGCCAAGGCATCCAC A,S Abbreviations: A, amplifi cation; S, sequencing; H, hybridization; F, fl uorescein; L, LC Red 640; B, biotin ¹ORF numbers as annotated by the Sanger Centre sequencing team positions” option (Felsenstein, J. 2002. PHYLIP (Phylogeny Inference Package) version 3.6a3. Distributed by the author. Department of Genome Sciences, University of Washington, Seattle). Th e unrooted phylogeny was drawn using Phylodendron (http://iubio.bio.indiana. 5 edu/treeapp/treeprint-form.html). broad-range rdna pcr B. holmesii 16S rDNA was amplifi ed by broad-range PCR using the 8F and 806R primers 13, and 23S rDNA was amplifi ed using the MS37 and MS38 primers 14. PCR products were cloned using the Topo TA Cloning Kit (Invitrogen, Carlsbad, CA). Twelve clones were picked for each strain, and plasmids were purifi ed using the Wizard miniprep kit (Promega, Madison, WI) and sequenced using M13 forward and reverse primers.

southern blot hybridization Th e genomic DNA of B. holmesii, B. pertussis and B. avium isolates was digested with the restriction enzymes Cla I and Nco I. Southern blotting and hybridization with biotin- labeled oligonucleotide probes was performed essentially as described by Schouls et al. 15. Table 3 shows a list of the probes that were hybridized to the genomes of the Bordetella isolates.

transcription of alcaligin genes

B. holmesii strains were grown to mid-late logarithmic phase (OD600 0.7-1, approximately equivalent to 0.7-1·109 cells per ml) in iron-depleted or iron-replete medium (see above). Total RNA was isolated using the Qiagen RNeasy Kit (Qiagen Inc, Valencia, CA), according to the manufacturer’s instructions for the isolation of total RNA from bacteria. For each experiment several controls were included, a DNAse treatment, a RNAse treatment and a DNAse plus RNAse treatment. RNA was reverse transcribed with 40 pmol of antisense gene-speciﬁ c primers and heated at 94°C for 5 min. Real-time PCR experiments were performed on the LightCycler (Roche) using the LightCycler DNA Master Hybridization Probe kit (Roche). Oligonucleotide primers and hybridization probes were designed by TIB MolBiol (Berlin, Germany) for the genes alcA, fauA and rpoB (see Table 3). rpoB, which encodes the DNA-directed RNA polymerase beta chain, was presumed to be unregulated and was used as an internal reference to compare expression induction. PCR was performed in 20-μl reaction volumes, containing 2 μl of LightCycler DNA Master HybProbe, 0.32 μl TaqStart antibody (1.1 μg/μl, BD

92 Acquisition of Iron-Uptake Genes

Biosciences, Heidelberg, Germany), 1 M Betain (Sigma-Aldrich Chemie Zwijndrecht, Th e

Netherlands), 3 mM MgCl2, 10 pmol of each forward and reverse primer and 4 pmol of the FL530 and of the LC640 hybridization probes. LightCycler RNase free water was added to bring the volume up to 18 μl. For template, 2 μl of cDNA was added to the mixture. Each run consisted of an initial denaturation of 2 min at 95°C to activate the polymerase, followed by 45 cycles of 95°C for 10 s, 55 °C for 10 s and 72°C for 20 s. Melting curve analysis was used to analyze and optimize the PCR conditions. Real-time PCR experiments were analyzed using LightCycler software version 4.0. alcA and fauA transcript levels were compared to rpoB transcript levels in the iron-depleted and the non-depleted samples, and this was used as a measure to determine the induction of expression. 5 detection of alcaligin Liquid chromatography and mass spectrometry (LC/MS) was used to determine whether alcaligin was produced by the B. holmesii strains. LC/MS: Samples were analyzed by nanoscale reversed phase-liquid chromatography (HP1100 LC system, Hewlett Packard Gmbh, Waldbronn, Germany) coupled to electrospray mass spectrometry (LCQ Classic ion trap), essentially as described by Meiring et al. 16. Briefl y, 10 μl of culture supernatant was injected on a trapping column packed with AQUA C18 (5 μm, Phenomenex) at a fl ow rate of 5 μl/min and by using 100% solvent A (0.1M acetic acid in water) as eluent for 5 min. Analytes were subsequently separated by reversed phase chromatography using a 44-cm long x 50 μm inner diameter analytical column with Pepmap (5 μm; Dionex) at a fl ow rate of 125 nl/min. A linear gradient was started from 10% solvent B (0.1M acetic acid in acetonitrile) to 60% solvent B in 35 min. Next, the columns were equilibrated in 100% solvent A for 10 min. Analytes were measured in the MS1 mode (m/z 400-2000) to determine the molecular weight and retention time. A second LC/MS measurement was performed to obtain detailed structural information by collision-induced dissociation of the MH+-ion (406 Da). Th e collision energy was set to 35%.

93 Chapter 

results

comparative genome hybridization Th e genomes of 12 epidemiologically unrelated B. holmesii strains were hybridized to a microarray representing the nearly complete gene complement of B. pertussis Tohama I, B. parapertussis 12822, and B. bronchiseptica RB50 10. Although most microarray probes failed to hybridize signiﬁ cantly to the B. holmesii genomic DNA, 1.2% to 1.6% of the probes hybridized as strongly to the B. holmesii genomes as they did to the reference sample (Figure 5 1A; microarray data have been deposited at ArrayExpress, accession number E-TABM- 55). Most of these positive probes mapped to two contiguous regions in the Tohama I genome: BP0787 – BP0794 and BP2450 – BP2499 (Figure 1B). In the genomes of B. bronchiseptica RB50 and B. parapertussis 12822, the orthologous sequences are present in a single contiguous region, suggesting that the region has been rearranged in B. pertussis Tohama I. Th e rearrangement in the Tohama I genome appears to have been mediated by recombination within an IS481 element located in the gene for a putative membrane protein encoded by BB3888 and BPP3438 in B. bronchiseptica and B. parapertussis, respectively. In B. pertussis Tohama I, the orthologous gene is split into two ORFs, BP0787 and BP2450 with remnants of IS481 adjacent to both ORFs (Figure 2A). Hybridization of B. holmesii genomic DNA to a probe for IS481 was consistent with previous work indicating the presence of eight IS481 copies in the B. holmesii genome 7. Th e genomic region shared by the classical bordetellae and B. holmesii contained a virulence- associated locus comprising alcABCDERS, encoding the alcaligin biosynthetic pathway (AlcABCDE), the exporter for alcaligin (AlcS), a transcriptional activator of the locus (AlcR) and the alcaligin uptake receptor (FauA) 17-21. Alcaligin is a siderophore produced by B. pertussis, B. bronchiseptica, and Alcaligenes denitriﬁ cans 22 that scavenges free iron from the extracellular milieu. It is important for the acquisition of iron in the eukaryotic host environment in which free iron is sequestered by host factors. Accordingly, we designated this genomic fragment the B. holmesii iron-uptake island (IUI). No other genes in the Tohama I IUI region had sequence features suggesting a role in virulence, with the possible exception of the B. pertussis virulence-repressed gene, vrg-6, for which a function has not been determined 23 (Table 2).

molecular characterization of iui in b. holmesii Partial nucleotide sequences of 21 genes in the IUI that were indicated to be present by CGH were obtained from B. holmesii isolates B0436, B1850 and B1852 after PCR ampliﬁ cation using B. pertussis primer pairs (submitted to Genbank). All of these sequences were over 99.3% similar to the B. pertussis sequence, with the exception of BP0794 (94.1%), located at the putative right end of IUI. In B. pertussis Tohama I, the BB3888 orthologue is split into two partial ORFs, BP0787 and

94 Acquisition of Iron-Uptake Genes

A. B. log (B. holmesii intensity/reference intensity) 2 B holmesii -6 -5 -4 -3 -2 -1 0 1 BP0001 B1855 B1853 B1854 B1852 B0436 B2739 B2738 B1851 B2768 BP Tohama B0437 B1850 B2767 BP0783 BP0787

BP0794 BP0798 BP2447 BP2450 5

alcABCDERS fauA mar intergenic region BP2465 vrg-6

BP2475 BP2476

kdpABCDE

BP2499 BP2502 IS481 IS1001 insertion IS1002 IS1663 elements BB2492

-30 3 BP3871

Figure 1A. CGH of 12 B. holmesii isolates to a microarray comprising the genomes of B. pertussis Tohama, B. para-

pertussis 12822 and B. bronchiseptica RB50. The running average (window = 3) of the mean log2(Cy5/Cy3) of 12 B. holmesii genomes is plotted on the X-axis. Microarray probes are arranged on the Y-axis in B. pertussis Tohama genome order. B. Probes that hybridized to the B. holmesii genome with comparable strength to the reference, and adjacent non-hybridizing probes, are shown in detail for individual B. holmesii strains and for B. pertussis Tohama. A selection of probes, representing insertion sequence elements are also shown. Strain numbers are indicated above the columns. Each row represents one probe in B. pertussis Tohama gene order. ORF and gene

designations are shown for a selection of probes. The relative hybridization value (log2(Cy5/Cy3)) is indicated by the yellow-black-blue color scale. Missing data are represented in grey. A color version of this fi gure is available in the appendix. BP2450 (Figure 2A). PCR amplifi cation using one primer in BP0787 and one in BP2450 yielded a product from all B. holmesii strains, but it was approximately 1 kb longer than the product from B. bronchiseptica RB50. Nucleotide sequencing of this product confi rmed that both halves of the gene were adjacent in B. holmesii, but, unlike the intact RB50 gene, were interrupted by an IS481 element at the same position as the Tohama I breakpoint (Figure 2A). Screening of 45 B. pertussis strains by PCR with this primer pair identifi ed only

95 Chapter 

one strain, 18323, that carried the same IS481-inactivated allele as B. holmesii (submitted to Genbank). Th e genomic organization of IUI in the B. holmesii isolates was further characterized by PCR analysis and DNA sequencing. PCR primers were designed to amplify 14 overlapping fragments, covering the B. pertussis Tohama I sequence contained in the B. holmesii IUI. PCR screening was carried out on 12 B. holmesii isolates and B. pertussis Tohama I, and the products were compared by agarose gel electrophoresis (Table 4). In most cases PCR amplicons from B. holmesii and B. pertussis were identical in size, suggesting that the genomic organization was conserved between B. holmesii and B. pertussis Tohama I. However, size 5 diff erences were observed in four PCR fragments. In the case of two polymorphic fragments, which were approximately 1 kb smaller in B. holmesii, Table 4. Mapping of the IUI by overlapping PCR fragments. sequencing indicated PCR Forward primer Reverse primer PCR product size (nt) that the IS481 elements BH BP BP2485 and BP2492 1 BP0790 - F BP0794 - R 3412 3412 2 BP0790 - BR BP2450 - CR ~5000a no fragment were missing from the B. 3 BP2450 - BF BP2452 - F 3160 no fragment holmesii IUI. Similar to 4 BP2452 - F BP2457 - R 6620 6620 B. holmesii, B. pertussis 5 BP2456 - F BP2458 - R 2700 2700 6 BP2458 - F BP2463 - R 6955 6955 18323 also lacked these 7 BP2463 - F BP2465 - R 6050 2536 two IS481 elements. A 8 BP2465 - F BP2472 - F 5482 5482 9 BP2472 - R BP2476 - R 4941 4914 third IS481 element, 10 BP2475 - F BP2479 - F 4151 4151 BP2477, was variably 11 BP2479 - R BP2482 - R 5558b 5558 12 BP2482 - BF BP2486 - BR ~5500 6456 present among the B. 13 BP2486 - R BP2494 - R ~7000 8200 holmesii strains. 14 BP2493 - F BP2499 - F 7199 7199 PCR amplifi cation a Expected size: 3,972 kb of a genomic region, b PCR amplicon not detected in B. holmesii isolate B2767 downstream of the alc Abbreviations: BH, B. holmesii; BP, B. pertussis operon revealed a 4.8 kb insertion in the B. holmesii IUI (Figure 2B). Nucleotide sequencing of this region indicated that the left terminus of the insert was located 237 nucleotides downstream of the stop codon of BP2464, and its right terminus in BP2465, which has been partially deleted (submitted to Genbank). A BLASTX search of the 4.8 kb B.holmesii sequence against the GenBank non-redundant protein database identifi ed fi ve putative genes with highly signifi cant hits (E < 1x10-21) to known proteins, which were designated bhoABCDE (Table 5). Th e fi rst putative gene, bhoA, is predicted to encode an IS3-family transposase. A frameshift was detected in bhoA, resulting in two overlapping ORFs (orfA and orfB). Th e presence of two out-of-frame overlapping ORFs has been previously described for IS3 transposases 24. Th e 3’ ORF of this gene contains a second frameshift mutation that may render the putative transposase non-

96 Acquisition of Iron-Uptake Genes functional. Th e top GenBank BLASTN hit to bhoA was BB2492, which encodes a single- copy transposase pseudogene in B. bronchiseptica RB50 25. Southern blot hybridization with a bhoA-speciﬁ c probe showed that this sequence was present in 40-50 copies in the genome of B. holmesii (data not shown). Th e ORFs bhoB and bhoC are homologous to adjacent genes (RmetDRAFT_1101 and RmetDRAFT_1100, respectively) from the Ralstonia metallidurans CH34 genome (GenBank accession NZ_ AAAI03000009.1). Th ese genes encode a hypothetical protein with a Bug domain and a D-isomer speciﬁ c 2-hydroxyacid dehydrogenase, respectively. bhoD encodes a putative extracytoplasmic function (ECF) sigma factor, and bhoE encodes a homologue of FecR/PupR proteins (Table 1). Th e GC-content of the insert is 62%, similar 5 to the average GC-content of B. holmesii (61.5 – 62.3%), but lower than that of B. pertussis (67.7%) 6,25.

Table 5. Homologies of putative ORFs in the DNA insertion in B. holmesii as determined by TBLASTX. Gene Orientation Homologies Species E-value bhoA1 orfA - Putative IS3 family transposases, orfA many 6e-27 orfB - Putative IS3 family transposases, orfB many 4e-31 bhoB + Uncharacterized protein UPF0065 Ralstonia metallidurans 8e-72 Ralstonia metallidurans & bhoC + Putative dehydrogenases 3e-77 many other bhoD + ECF σ70 factors, FecI many 1e-21 bhoE + FecR many 2e-22

¹Putative pseudogene due to two frameshifts Th e IUI sequence was also compared to the unpublished complete genome sequence of B. avium 197N (produced by the B. avium Sequencing Group at the Sanger Institute and obtained from ftp://ftp.sanger.ac.uk/pub/pathogens/ba/). Pairwise comparison identifi ed a region in the B. avium genome syntenic to the IUI region of the other Bordetella genomes (Figure 2A). Th e region of synteny extends beyond the ends of IUI as defi ned by CGH, to include genes that were not detected by B. holmesii CGH. Sequences near the putative left and right ends of IUI have higher sequence identity to the B. avium genome (at least 88%) than sequences in the middle. Although gene order in the syntenic region is conserved, the B. avium genome shows no detectable homology to the alcABCDERS, fauA, bhoABCDE and IS481 loci (Figure 2A, Table 2). To determine the left breakpoint of IUI in B. holmesii, a PCR primer pair was designed with a forward primer in the B. pertussis BP0794 (rpsO) gene, which was detected in B. holmesii by CGH, and a reverse primer in a region, completely conserved in the classical Bordetella and B. avium genomes, of the adjacent undetected gene, BP0795 (pnp). PCR amplifi cation with this primer pair yielded a 1.4 kb product from B. pertussis, B. avium, and B. holmesii. Sequencing of the B. holmesii product (submitted to Genbank) and comparison to the B. pertussis sequence by BLASTN indicated that the BP0794 sequence is identical to B.

Chapter 

BP2499 BP2499 isolates 197N. binding site 5 kb 70

Fur binding site Fur σ

B. holmesii B. BP2465 bhoABCDE B. avium B. Representation of Representation A. .

Orthologs in all species not found Orf by replaced Conserved IUI-flanking ORFs in present Variably BP2492 FecIR * Tohama and Tohama

IUI. Arrows above ORFs indicate ORFs indicate above IUI. Arrows was derived from the published from was derived BP2485 B. bronchiseptica B.

5 pertussisB. hologous sequences. The ORF composi- The hologous sequences. bhoB bhoC bhoD bhoE insertion and B. holmesii B. elements RB50, orfA 481 Alcaligin operon Alcaligin holmesii B. IS BP0787/2450 orthologs Orthologs in all species found BhoA B. bronchiseptica bronchiseptica B. orfB

B. pertussisB.

and ,

marC BP2477 *

B. bronchiseptica bronchiseptica B. BP2477 B. holmesii B. B. pertussisB. , , B. avium B. B. avium B.

Fe-alcaligin receptor Fe-alcaligin regulation

+ alcaligin export+ alcaligin

alcR alcS fauA alcABCDE

BP2453 BP2453

and comparison to orthologous to in the genomes of and comparison regions

0 BP2450

alcA BP0787/245 in the appendix. is available gure of this fi version A color

alcaligin biosynthesis alcaligin

B. holmesii B.

BP0794 ORFs in the 4.8 kb locus and putative of the alcaligin DNA insertion Detailed organization detected in BP0794 alcA alcB alcC alcD alcE B.

IUI was deduced from PCR and CGH data, while the ORF organization of data, while the ORF organization PCR and CGH from IUI was deduced

BP2455 BP0787 island (IUI) in of the iron-uptake of the genomic organization analysis Comparative B. holmesii B. Figure 2. Figure genome sequences. genome sequences. function the putative of these ORFs. the ORF organization of the IUI in the ORF organization surfaces. of the ort Dashed lines connect grey Deletions or insertions the ORFs at borders by indicated species are between tion of B. A. B. avium B. holmesii B. pertussisB. bronchiseptica B.

98 Acquisition of Iron-Uptake Genes pertussis for the ﬁ rst 93 nt, but slightly divergent downstream, placing the IUI breakpoint at or near codon 31 of BP0794. Th e reading frames of BP0794 and BP0795 are intact within the region spanned by this sequence, with 98% and 91% amino acid identity to the B. pertussis protein sequences, respectively. To determine the right breakpoint of IUI, the TOPO Walker kit was used. Sequence data were obtained for the 5’ region of BP2499, extending 555 nucleotides downstream of BP2499 (dnaK) (submitted to Genbank). BLASTN indicated that the B. holmesii sequences downstream of BP2499 were orthologous to B. pertussis BP2500 and BP2501. Th e breakpoint of IUI was expected to be located between the array probes BP2499 (at the 3’ end of BP2499) and BP2500 (at the 3’ end of BP2500). Th erefore, a PCR primer pair 5 was designed with a forward primer in the 3’ array probe region of BP2499 and the reverse primer in the B. holmesii BP2500 sequence. Sequencing of this PCR product indicated 99% sequence identity to B. pertussis in the 3’ region of BP2499, but only 90% identity to B. pertussis in the the 5’ region of BP2499 (up until codon 276) (submitted to Genbank). Th is places the right breakpoint approximately at codon 276 of BP2499. Th e sequence in the immediate vicinity of the breakpoint was 91% identical to B. avium. Conservation of gene order around the breakpoints of IUI among the classical and non- classical Bordetella genomes suggests that the transferred island may have integrated into the orthologous location in the B. holmesii genome by a homologous recombination event. Furthermore, no evidence was found at either end of IUI for the presence of phage, plasmid, or IS sequences, indicating that integration of IUI into the genome was not mediated by a mobile DNA element. alcaligin expression and detection To determine whether the IUI-encoded alcaligin biosynthesis, export and uptake locus is expressed in B. holmesii, transcript levels of alcA and fauA were measured by quantitative RT- PCR. Because this locus is repressed in the presence of free iron by the Fur transcriptional repressor in B. pertussis and B. bronchiseptica 26, expression of these genes was determined under iron-depleted and iron-replete conditions. Th e transcript abundance of both alcA and fauA was 10-fold higher in the iron-depleted sample compared to the non-depleted sample (data not shown), indicating that the locus is expressed and iron-regulated. To determine if the expression of these genes resulted in the synthesis and secretion of alcaligin, we employed nanoscale capillary liquid chromatography-mass spectrometry (LC-MS) to analyze the supernatants of iron-starved B. holmesii strains for the presence of alcaligin 16,22. Th is approach detected a compound in the iron-depleted supernatant, but not in the iron-replete supernatant, with m/z 405 (M + H)+ (Figure 3), corresponding to the mass of desferri-alcaligin as described by Moore et al. 22. MS fragmentation analysis detected two fragment ions with m/z 283 and 387, corresponding to previously described alcaligin fragment ions 22.

99 Chapter 

100 -Fe

60 36.99

Relative Abundance 20

0 5 0 5 10 15 20 25 30 35 40 Time (min) 100 +Fe

Relative Abundance 20 38.87 0 0 5 10 15 20 25 30 35 40 Time (min)

a B. ab’ a’b N b 283 OH O 100 OH O NH

HN O HO 80 OOH b’ N 60 a’ Alcaligin 40 m/z=405 -H2O Relative Abundance 20 385

0 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 m/z

Figure 3. LC/MS spectra of B. holmesii supernatants cultured in iron-depleted and iron-replete medium. A. LC/MS spectra of iron-depleted and iron-replete B. holmesii supernatants, depicting ions with m/z 405 (corresponding to alcaligin), with retention times between 0 and 45 min. B. Fragment ions detected after collisional activation (35% energy) of the peak from (A) with retention time t=36.99 min. After measurement of the reference standard, the calibration deviated approximately 1.5 mass units, explaining the diff erence in m/z values compared to the fragment ions as detected by 22. Nomenclature of fragment ions is according to 22.

phylogeny of b. holmesii In contrast to 16S rRNA gene sequencing data that suggested a close phylogenetic relationship between B. pertussis and B. holmesii, our CGH data suggested substantial sequence divergence between the two species. To determine the phylogenetic position of B. holmesii within the

100 Acquisition of Iron-Uptake Genes

Burkholderia Figure 4. Neighbor-joining phylogeny of Bordetella and pseudomallei related β-proteobacteria. The unrooted tree is based on 3,559 fully informative nucleotides from an alignment of concatenated sequences of atpD, rpoB, tuf and rnpB from B. holmesii, B. pertussis, B. parapertussis, B. bronchiseptica, B. avium, B. hinzii, B. trematum, B. petrii, Achro- mobacter xylosoxidans, and Burkholderia pseudomallei. Bootstrapping values greater than 50% (based on 1000 resamplings) are indicated at branches.

Bordetella genus, conserved regions of the Achromobacter xylosoxidans atpD, rpoB, tuf and rnpB genes, which have 5 been used for multilocus sequence typing of other organisms, were sequenced from seven 100 B. holmesii strains and from representative B. trematum B. petrii isolates of B. hinzii, B. trematum, B. petrii, and

89 95 the closely related β-proteobacterial species, 67 B. avium Achromobacter xylosoxidans (submitted to 100 GenBank). Orthologous sequences from 100 B. bronchiseptica the classical Bordetella species, B. avium, B. parapertussis B. hinzii B. pertussis and another β-proteobacterial species, Burkholderia pseudomallei, were obtained B. holmesii 0.1 from available genome sequence data. A neighbor-joining tree, based on 3,559 fully informative characters in the alignment of concatenated nucleotide sequences of these genes, indicated that B. holmesii is more closely related to B. avium and B. hinzii than to the mammalian bordetellae (Figure 4). Th e exclusion of B. holmesii from the mammalian Bordetella clade was supported by all individual gene trees. Comparison of atpD, rpoB, tuf and rnpB sequences from seven independent B. holmesii strains revealed only two single nucleotide polymorphisms (one synonymous and one non-synonymous) among 3,666 aligned bases, both of which were variant only in Bho29. molecular characterization of the s rrna loci in b. holmesii In light of the phylogeny determined above, the near identity of the B. pertussis and B. holmesii 16S rRNA genes (99.7%) is anomalous. For comparison, the B. pertussis 16S rRNA sequence is 98.5% and 99.2% identical to the 16S rRNA sequences of B. avium and B. hinzii, respectively. One possible explanation for this discrepancy is that the 16S rRNA gene, like the IUI, was laterally transferred from B. pertussis to B. holmesii. If such a transfer has occurred, it is possible that B. holmesii harbors one or more copies of a divergent “native” 16S rRNA gene in addition to the proposed B. pertussis-derived locus, although the presence of heterogeneous rRNA loci within a genome is rather limited 27. No alternate B.

101 Chapter 

B. pertussis/B. holmesii Broad-range 16S probe 16S probe (Tohama) (Tohama) (Tohama) (Tohama) (B0436) (B0436) (B0436) (B0436) B0021) (B0021) (B0021) (B0021) MW Marker pertussisB. holmesii B. avium ( B. pertussisB. holmesii B. avium B. pertussisB. holmesii B. avium B. pertussisB. holmesii B. avium B.

Cla I Nco I Cla I Nco I 5 9,416 Figure 5. Southern blot hybridizations of B. pertussis 6,557 Tohama, B. holmesii B0436 and B. avium B0021 with 16S rDNA probes. Genomic DNA 4.361 was hybridized to a B. pertussis-specifi c probe (left panel) and a broad-range 16S rDNA probe (B-16S8F; right panel) (2). For each hybridization experiment, the probe was stripped off the membrane, 2.322 and the membrane was 2,027 re-used. Biotinylated DNA markers are in the fi rst lane.

holmesii 16S rRNA sequences have been reported to date. To determine the number of B. pertussis-like 16S rRNA gene copies and search for divergent 16S rRNA genes, Southern blot hybridization was performed using 16S rDNA probes. A B. pertussis-specifi c 16S rDNA probe that did not hybridize to the B. avium genome detected three copies in the B. holmesii genome (Figure 5). A broad-range 16S rDNA probe 15 that hybridized equally well to B. pertussis and B. avium did not detect any additional bands in B. holmesii, suggesting that the only 16S rRNA loci are the three B. pertussis-like copies (Figure 5). 16S rRNA broad-range PCR was also employed to identify possible variant 16S rRNA sequences. Twelve cloned PCR products from each of three B. holmesii strains were sequenced. All were at least 99.5% identical to the B. pertussis 16S rRNA, further suggesting that all copies of the 16S rRNA in B. holmesii are essentially identical to the B. pertussis gene (submitted to GenBank). No copies of the 16S rRNA gene were identifi ed within the boundaries of the IUI as defi ned above, suggesting that this gene was not transferred in the same recombination event. To further test whether a 16S rRNA gene was located near either end of IUI in B. holmesii, PCR amplifi cation was attempted using primers in the 16S rRNA gene and in the putative 5’ and 3’ termini of IUI. PCR amplifi cation was unsuccessful, suggesting that no

102 Acquisition of Iron-Uptake Genes

16S rRNA genes are in close proximity to IUI. Most of the region between the 16S and 23S rRNA genes, including two tRNA genes, was also amplifi ed and sequenced (submitted to GenBank). Th is region is more variable among the sequenced bordetellae than the structural rRNA genes; for example, B. pertussis is only 96.4% identical to B. bronchiseptica and 72.5% identical to B. avium over the 550 nucleotides examined in this study. However, sequences from seven B. holmesii strains, all of which were identical to each other, were 99.6% identical to B. pertussis, but only 72.2% identical to B. avium. Partial sequence of the 23S rRNA gene (B. pertussis bases 1043-1919) was also determined following broad-range PCR from three B. holmesii strains. Nine sequences (three from 5 each strain) were identical except single-nucleotide diff erences, possibly due to PCR errors, that were each restricted to a single clone (submitted to GenBank). Over 882 bases the B. holmesii consensus sequence was identical to B. pertussis at 857 positions and to B. avium at 869 positions. For comparison, B. pertussis and B. bronchiseptica vary at only a single nucleotide in this region. Th ese data suggest that this fragment of the 23S rRNA gene has not been replaced by a B. pertussis-like sequence, placing the right breakpoint of the proposed gene transfer and recombination event within the fi rst 1000 bases of the 23S rRNA gene.

103 Chapter 

discussion

evolutionary relationships of b. holmesii to other bordetella species Th e evolutionary relationship of B. holmesii to other members of the Bordetella genus has been controversial. Although 16S rRNA sequencing and clinical criteria placed B. holmesii close to B. pertussis, other evidence, including sequencing of protein coding genes and analysis of cellular fatty acid composition, suggested a more distant evolutionary relationship between these species 4,6,8. In this work, we provide a possible explanation for these discrepancies. 5 Our microarray-based CGH analysis indicated that, with the exception of part of the rRNA operon, and a highly conserved genomic island, the B. holmesii genomic sequences diverge considerably from those of the mammalian bordetellae, B. pertussis, B. parapertussis and B. bronchiseptica. Likewise, our multilocus sequence analysis suggested that B. holmesii is phylogenetically more closely related to B. hinzii and B. avium than to the mammalian bordetellae. We propose that lateral transfer of the 16S rRNA gene from B. pertussis is the most likely explanation for the presence of a highly similar gene in B. holmesii. Th is exchange may have occurred in conjunction with transfer of IUI, but failure to identify a 16S rRNA gene inside or in close proximity to IUI suggests that it may have been independently transferred from B. pertussis, or that the 16S rRNA gene has been transposed after horizontal exchange. Southern blot hybridization and sequencing of individual 16S rRNA genes showed that the genome of B. holmesii contained three B. pertussis-like 16S rRNA genes. Th e presence of only B. pertussis-like 16S rRNA genes in B. holmesii suggests that putative endogenous copies have either been lost or converted by recombination after acquisition of the B. pertussis-like 16S rRNA gene. Relatively low sequence identity between the B. pertussis and B. holmesii 23S rRNA genes suggests that the proposed exchange may have involved only the 16S rRNA gene and the sequence upstream of the 23S rRNA gene, including two tRNA genes. Only two out of 3,666 nucleotide positions in the housekeeping genes were polymorphic among seven B. holmesii strains. Th is high degree of sequence conservation among geographically distinct isolates argues for the recent emergence of a clonal B. holmesii population. For comparison, B. pertussis, which diverged and expanded clonally from the last common ancestor of B. bronchiseptica and B. pertussis two million to ﬁ ve million years ago (25, Diavatopoulos et al.)), diﬀ ers from B. bronchiseptica at only three bases in this sequence sample.

evidence for transfer of a genomic island from b. pertussis to b. holmesii Comparative genomic hybridization identiﬁ ed a 66 kb DNA region, IUI, which, in contrast to the rest of the B. holmesii genome, was highly conserved in B. pertussis, B. bronchiseptica, and B. parapertussis. Th is observation could be explained by the transfer to B. holmesii of a

104 Acquisition of Iron-Uptake Genes genomic fragment from any of these three species. But, the presence in IUI of IS481 elements, which have only been found in B. pertussis and B. holmesii, argues that a B. pertussis strain was the donor in the putative DNA transfer event. Because the IUI region is rearranged in B. pertussis Tohama I relative to B. holmesii and the other mammalian bordetellae, transfer from this strain to B. holmesii would have required two independent events, which seems unlikely. However, chromosome order is known to be highly variable in B. pertussis 28. One strain, 18323, had a genomic architecture and an IS481 element distribution very similar to those observed in the B. holmesii IUI, suggesting that the B. holmesii IUI may have been derived from an 18323-like B. pertussis strain. Several genotyping methods have shown that 18323 is genetically distinct from other B. pertussis strains 10,29, but related strains have 5 been isolated from pertussis patients as recently as 1993, indicating that they still circulate in the human population 30. B. pertussis and B. holmesii have both been isolated from the respiratory tract of humans, making the human airway the most likely environment in which transfer of IUI could have occurred. At the level of resolution of CGH and PCR analysis, the genomic composition of IUI was identical, with the exception of a variably present IS481 element, in all B. holmesii isolates examined. Th is result suggests that IUI was acquired recently, or alternatively, that IUI has been under selective pressure to maintain its genomic organization. A region syntenic to the B. holmesii IUI was detected in the genome of the avian pathogen B. avium (Figure 2A). Th e region of synteny between B. avium, B. holmesii, and the classical bordetellae extended beyond the boundaries of IUI, suggesting that the backbone of this chromosomal region is conserved across distantly related Bordetella species. Th is ﬁ nding, together with the failure to identify signatures of mobile elements (e.g., phage or conjugative transposon genes) at the ends of IUI, suggests that the most likely mechanism for the proposed integration of IUI into the B. holmesii genome is homologous recombination between the laterally transferred IUI and the ancestral B. holmesii genome. Th e degree of nucleotide sequence similarity between the B. holmesii IUI and the orthologous sequences in B. avium is highest in the vicinity of the left and right insertion breakpoints, indicating that these genes may be more conserved across the Bordetella genus than the average gene. BP0794, located at the left end of IUI, encodes the 30S ribosomal protein S15, which is highly conserved in the Bordetellae. At the right end of IUI, dnaK is located, which encodes a chaperone that is also highly conserved across bacterial species. Th ese conserved sequences are proposed to have served as the substrates for the putative double homologous recombination event that replaced the B. holmesii sequence with the transferred B. pertussis fragment. iron-acquisition function conferred by iui Sequestration of free iron by the eukaryotic host creates a hostile environment for bacterial pathogens. Consequently, most, if not all bacterial pathogens have evolved various strategies

105 Chapter 

to obtain iron from the host. B. pertussis and B. bronchiseptica possess a number of diff erent iron acquisition systems with specifi cities for diff erent environmental iron sources, but their key mechanism for scavenging free iron is production of the siderophore, alcaligin. Production of alcaligin by B. bronchiseptica has been shown to be required for maximal virulence in a piglet model of infection 31. In bacteria, expression of iron acquisition genes is negatively regulated by the Fur transcriptional repressor (reviewed in 32. In B. pertussis and B. bronchiseptica, Fur regulates expression of the alc operon, as well as several other iron acquisition loci 26,33. Interestingly, the complete locus encoding alcaligin biosynthesis (alcABCDE), export 5 (alcS), uptake (fauA), and regulation (alcR), was present in the B. holmesii IUI 17-21. Reverse transcriptase real-time PCR experiments demonstrated that alcA and fauA transcription in B. holmesii was induced 10-fold in the absence of iron. Furthermore, detection of alcaligin in the supernatant of iron-limited B. holmesii cultures indicated that the alcaligin biosynthesis and export locus was functional in B. holmesii. Alcaligin production has not been detected in B. avium, and its genome does not appear to encode an alcaligin biosynthesis locus (data not shown), indicating that not all Bordetella species possess the ability to produce this siderophore. Th erefore, prior to IUI acquisition, the hypothetical progenitor of B. holmesii may not have been competent to produce alcaligin. Wholesale acquisition of this function by lateral transfer from B. pertussis could have provided B. holmesii with a new, highly effi cient iron uptake system, leading to an immediate enhancement of its ability to colonize a eukaryotic host. In many cases, maximal expression of iron-uptake loci requires the activity of a transcriptional activator in addition to de-repression by Fur 34. An example of such a system is the FecIRA regulatory system in Escherichia coli K12, which controls the expression of the ferric citrate transport (fec) locus 35-38. FecA is an outer membrane protein that responds to environmental signals and transmits these signals to the cytoplasmic membrane protein FecR. FecR in turn transmits this signal to the σ70 factor, FecI, which then regulates the transcription of fecABCDE 35,36. Homologous FecIR regulatory systems (with or without a FecA homologue) have been identifi ed in a large number of bacterial genomes including B. bronchiseptica, B. pertussis and B. avium, in which they control expression of the heme uptake locus 39-42. Th e products of the two putative ORFs in the 3’ region of the insert, bhoD and bhoE, are homologous to ECF sigma factors (σ70) and FecR family proteins, respectively. We hypothesize that the ORFs bhoD and bhoE encode a FecIR-like regulatory system. Although we did not identify an obvious FecA-homologue in the insertion, FauA may fulfi ll this role, as FecIR systems are able to associate with one or more FecA-like partners, sometimes from distal genomic loci 37. Consistent with this model, FauA is highly similar to FpvA, the Pseudomonas aeruginosa ferripyoverdine receptor, which is an established member of a FecIRA system, FpvIRA. Expression of alcaligin in B. bronchiseptica and B. pertussis is assumed to be regulated solely

106 Acquisition of Iron-Uptake Genes by Fur and the AraC-like regulator AlcR. Interestingly, we identifi ed a putative Fur-binding site centered 43 nucleotides upstream of the start codon of bhoD, suggesting that bhoDE is Fur-regulated. Th is FecIR system could, in turn, regulate transcription of the alc operon. Although this has not been described to be the case for B. pertussis or B. bronchiseptica, the possibility that transcriptional regulators besides AlcR may participate in activation of the alc operon cannot be ruled out. In fact, detailed analysis of the B. pertussis region upstream of the alcaligin operon indicated the presence of a putative σ70 recognition site 75 nucleotides upstream of the transcriptional start site of alcA, although it should be mentioned that the typical ECF sigma factor box is located at 35 nucleotides upstream of the start codon, so the possibility remains that the σ70 recognition site is too far upstream 5 to aff ect transcription of alcABCDE. By enhancing transcription of the alcaligin operons under iron-limiting conditions, BhoDE could further increase the ability of B. holmesii to acquire iron in the host environment. Th e eff ects of IUI acquisition on the evolution of virulence and transmission of B. holmesii in humans are unknown, but it is reasonable to propose that the ability to produce alcaligin would improve the colonization rates of B. holmesii. Experimental testing of this hypothesis using B. holmesii mutants awaits the development of an appropriate animal model. B. holmesii was identifi ed as a pathogen in humans relatively recently 6. Th e fi rst cases concerned immunocompromised patients with septicemia, although it was shown more recently that B. holmesii may also cause pertussis-like disease 3. Th e results discussed here strongly suggest that the acquisition of B. pertussis DNA, by conferring an increased capacity to scavenge free iron in the host environment, has played a key role in the emergence of B. holmesii and its adaptation to humans.

107 Chapter 

acknowledgements We would like to thank Ir. J. ten Hove and Dr. A. de Jong (NVI, Unit Research and Development) for assistance with LC/MS experiments and analysis. Th is work was supported by a travel grant from the Netherlands Organization for Scientiﬁ c Research (NWO).

108 Acquisition of Iron-Uptake Genes reference list distinguished by patterns of substantial gene loss and host adaptation. J Bacteriol. 186, 1484-1492.

1. Shepard,C.W., Daneshvar,M.I., Kaiser,R.M., 11. Rozen,S., Skaletsky,H. (2000). Primer3 on the WWW Ashford,D.A., Lonsway,D., Patel,J.B., Morey,R. for general users and for biologist programmers. E., Jordan,J.G., Weyant,R.S., Fischer,M. (2004). Methods Mol.Biol. 132, 365-386. Bordetella holmesii bacteremia: a newly recognized 12. Chenna,R., Sugawara,H., Koike,T., Lopez,R., clinical entity among asplenic patients. Clin.Infect. Gibson,T.J., Higgins,D.G., Thompson,J.D. (2003). Dis. 38, 799-804. Multiple sequence alignment with the Clustal 2. Russell,F.M., Davis,J.M., Whipp,M.J., Janssen,P.H., series of programs. Nucleic Acids Res. 31, 3497- Ward,P.B., Vyas,J.R., Starr,M., Sawyer,S.M., Curtis,N. 3500. (2001). Severe Bordetella holmesii Infection in a 13. Relman,D.A., Schmidt,T.M., MacDermott,R.P., Previously Healthy Adolescent Confi rmed by Gene Falkow,S. (1992). Identifi cation of the uncultured 5 Sequence Analysis. Clin Infect Dis 33, 129-130. bacillus of Whipple’s disease. N.Engl.J.Med. 327, 3. Yih,W.K., Silva,E.A., Ida,J., Harrington,N., Lett,S. 293-301. M., George,H. (1999). Bordetella holmesii-like 14. Kotilainen,P., Jalava,J., Meurman,O., Lehtonen,O.P., organisms isolated from Massachusetts patients Rintala,E., Seppala,O.P., Eerola,E., Nikkari,S. (1998). with pertussis-like symptoms. Emerg.Infect.Dis. 5, Diagnosis of meningococcal meningitis by broad- 441-443. range bacterial PCR with cerebrospinal fl uid. J.Clin. 4. Gerlach,G., von Wintzingerode,F., Middendorf,B., Microbiol 36, 2205-2209. Gross,R. (2001). Evolutionary trends in the genus 15. Schouls,L.M., Schot,C.S., Jacobs,J.A. (2003). Bordetella. Microbes and Infection 3, 61-72. Horizontal transfer of segments of the 16S rRNA 5. Skeeles,J.K., Arp,L.H. (1997). Bordetellosis (Turkey genes between species of the Streptococcus Coryza). In: Diseases of poultry., ed. B.W.Chalnek, anginosus group. The Journal of Bacteriology 185, H.J.Barnes, C.W.Beard, L.R.McDougal, Y.M.SaifAmes: 7241-7246. Iowa State University Press, 275-288. 16. Meiring,H.D., van der Heeft,E., ten Hove,G.J., de 6. Weyant,R.S., Hollis,D.G., Weaver,R.E., Amin,M. Jong,A.P.J.M. (2002). Nanoscale LC-MS(n): technical F., Steigerwalt,A.G., O’Connor,S.P., Whitney,A.M., design and applications to peptide and protein Daneshvar,M.I., Moss,C.W., Brenner,D.J. (1995). analysis. J.Sep.Sci. 25, 557-568. Bordetella holmesii sp. nov., a new gram-negative 17. Kang,H.Y., Brickman,T.J., Beaumont,F.C., species associated with septicemia. Journal of Armstrong,S.K. (1996). Identifi cation and Clinical Microbiology 33, 1-7. characterization of iron-regulated Bordetella 7. Reischl,U., Lehn,N., Sanden,G.N., Loeff elholz,M. pertussis alcaligin siderophore biosynthesis genes. J. (2001). Real-Time PCR Assay Targeting IS481 J Bacteriol. 178, 4877-4884. of Bordetella pertussis and Molecular Basis for 18. Beaumont,F.C., Kang,H.Y., Brickman,T.J., Detecting Bordetella holmesii. Journal of Clinical Armstrong,S.K. (1998). Identifi cation and Microbiology 39, 1963-1966. characterization of alcR, a gene encoding an AraC- 8. Gerlach,G., Janzen,S., Beier,D., Gross,R. (2004). like regulator of alcaligin siderophore biosynthesis Functional characterization of the BvgAS two- and transport in Bordetella pertussis and Bordetella component system of Bordetella holmesii. bronchiseptica. The Journal of Bacteriology 180, Microbiology 150, 3715-3729. 862-870.

9. Connell,T.D., Dickenson,A., Martone,A.J., Militello,K. 19. Pradel,E., Guiso,N., Locht,C. (1998). Identifi cation T., Filiatraut,M.J., Hayman,M.L., Pitula,J. (1998). of AlcR, an AraC-type regulator of alcaligin Iron starvation of Bordetella avium stimulates siderophore synthesis in Bordetella bronchiseptica expression of fi ve outer membrane proteins and and Bordetella pertussis. The Journal of regulates a gene involved in acquiring iron from Bacteriology 180, 871-880. serum. Infection and Immunity 66, 3597-3605. 20. Brickman,T.J., Armstrong,S.K. (1999). Essential 10. Cummings,C.A., Brinig,M.M., Lepp,P.W., Van De,P. role of the iron-regulated outer membrane S., Relman,D.A. (2004). Bordetella species are receptor FauA in alcaligin siderophore-mediated

109 Chapter 

iron uptake in Bordetella species. The Journal of (1997). Molecular evolution and host adaptation Bacteriology 181, 5958-5966. of Bordetella spp.: phylogenetic analysis using multilocus enzyme electrophoresis and typing 21. Brickman,T.J., Armstrong,S.K. (2005). Bordetella with three insertion sequences. The Journal of AlcS Transporter Functions in Alcaligin Siderophore Bacteriology 179, 6609-6617. Export and Is Central to Inducer Sensing in Positive Regulation of Alcaligin System Gene Expression. 30. Boursaux-Eude,C., Thiberge,S., Carletti,G., Guiso,N. The Journal of Bacteriology 187, 3650-3661. (1999). Intranasal murine model of Bordetella pertussis infection: II. Sequence variation and 22. Moore,C.H., Foster,L.A., Gerbig,D.G., Jr., Dyer,D.W., protection induced by a tricomponent acellular Gibson,B.W. (1995). Identifi cation of alcaligin as vaccine. Vaccine 17, 2651-2660. the siderophore produced by Bordetella pertussis and B. bronchiseptica. J Bacteriol. 177, 1116-1118. 31. Register,K.B., Ducey,T.F., Brockmeier,S.L., Dyer,D. W. (2001). Reduced virulence of a Bordetella 5 23. Knapp,S., Mekalanos,J.J. (1988). Two trans-acting bronchiseptica siderophore mutant in neonatal regulatory genes (vir and mod) control antigenic swine. Infection and Immunity 69, 2137-2143. modulation in Bordetella pertussis. The Journal of Bacteriology 170, 5059-5066. 32. Escolar,L., Perez-Martin,J., de Lorenzo,V. (1999). Opening the Iron Box: Transcriptional 24. Sekine,Y., Eisaki,N., Ohtsubo,E. (1994). Translational Metalloregulation by the Fur Protein. The Journal control in production of transposase and in of Bacteriology 181, 6223-6229. transposition of insertion sequence IS3. J.Mol.Biol. 235, 1406-1420. 33. Beall,B.W., Sanden,G.N. (1995). Cloning and initial characterization of the Bordetella pertussis fur 25. Parkhill,J., Sebaihia,M., Preston,A., Murphy,L.D., gene. Curr.Microbiol. 30, 223-226. Thomson,N., Harris,D.E., Holden,M.T., Churcher,C. M., Bentley,S.D., Mungall,K.L., Cerdeno-Tarraga,A. 34. Venturi,V., Weisbeek,P., Koster,M. (1995). Gene M., Temple,L., James,K., Harris,B., Quail,M.A., regulation of siderophore-mediated iron Achtman,M., Atkin,R., Baker,S., Basham,D., Bason,N., acquisition in Pseudomonas: not only the Fur Cherevach,I., Chillingworth,T., Collins,M., Cronin,A., repressor. Mol.Microbiol. 17, 603-610. Davis,P., Doggett,J., Feltwell,T., Goble,A., Hamlin,N., 35. Van Hove,B., Staudenmaier,H., Braun,V. (1990). Hauser,H., Holroyd,S., Jagels,K., Leather,S., Novel two-component transmembrane Moule,S., Norberczak,H., O’Neil,S., Ormond,D., transcription control: regulation of iron dicitrate Price,C., Rabbinowitsch,E., Rutter,S., Sanders,M., transport in Escherichia coli K-12. The Journal of Saunders,D., Seeger,K., Sharp,S., Simmonds,M., Bacteriology 172, 6749-6758. Skelton,J., Squares,R., Squares,S., Stevens,K., Unwin,L., Whitehead,S., Barrell,B.G., Maskell,D. 36. Enz,S., Mahren,S., Stroeher,U.H., Braun,V. (2000). J. (2003). Comparative analysis of the genome Surface signaling in ferric citrate transport gene sequences of Bordetella pertussis, Bordetella induction: interaction of the FecA, FecR, and FecI parapertussis and Bordetella bronchiseptica. Nat regulatory proteins. The Journal of Bacteriology Genet. 35, 32-40. 182, 637-646.

26. Brickman,T.J., Armstrong,S.K. (1995). Bordetella 37. Visca,P., Leoni,L., Wilson,M.J., Lamont,I.L. (2002). pertussis fur gene restores iron repressibility of Iron transport and regulation, cell signalling siderophore and protein expression to deregulated and genomics: lessons from Escherichia coli and Bordetella bronchiseptica mutants. J Bacteriol. 177, Pseudomonas. Mol.Microbiol. 45, 1177-1190. 268-270. 38. Braun,V., Mahren,S., Ogierman,M. (2003). 27. Coenye,T., Vandamme,P. (2003). Intragenomic Regulation of the FecI-type ECF sigma factor by heterogeneity between multiple 16S ribosomal transmembrane signalling. Curr.Opin.Microbiol. 6, RNA operons in sequenced bacterial genomes. 173-180. FEMS Microbiol.Lett. 228, 45-49. 39. Pradel,E., Locht,C. (2001). Expression of the putative 28. Stibitz,S., Yang,M.S. (1999). Genomic plasticity in siderophore receptor gene bfrZ is controlled natural populations of Bordetella pertussis. The by the extracytoplasmic-function sigma factor Journal of Bacteriology 181, 5512-5515. BupI in Bordetella bronchiseptica. The Journal of Bacteriology 183, 2910-2917. 29. van der Zee,A., Mooi,F., van Embden,J., Musser,J.

110 Acquisition of Iron-Uptake Genes

40. Vanderpool,C.K., Armstrong,S.K. (2001). The Bordetella bhu locus is required for heme iron utilization. The Journal of Bacteriology 183, 4278- 4287.

41. Vanderpool,C.K., Armstrong,S.K. (2004). Integration of environmental signals controls expression of Bordetella heme utilization genes. J Bacteriol. 186, 938-948.

42. Kirby,A.E., Metzger,D.J., Murphy,E.R., Connell,T. D. (2001). Heme utilization in Bordetella avium is regulated by RhuI, a heme-responsive extracytoplasmic function sigma factor. Infection 5 and Immunity 69, 6951-6961.

111 112

Chapter Chapter 66

Evolution of the Bordetella Autotrans- porter Pertactin: Identiﬁ cation of Re- gions Subject to Positive Selection

Dimitri A. Diavatopoulos1,2‡, Marcel Hijnen1,2‡, Frits R. Mooi1,2

‡ These authors contributed equally to this work

Submitted for publication Chapter 

abstract Th e virulence factor Pertactin is expressed by all of the closely related bacterial pathogens Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. B. pertussis and B. parapertussis both cause whooping cough in humans, which kills 300,000 people annually. B. bronchiseptica is usually an animal pathogen, but recently it was shown that a human- associated lineage also exists. Pertactin is an autotransporter that is involved in adherence of Bordetellae to the lung epithelium. It is an important component of most current acellular pertussis vaccines, which are based on B. pertussis. Th ese three species produce immunologically distinct pertactin molecules, and it has been shown that the acellular pertussis vaccines do not provide protection against B. parapertussis and probably also not against B. bronchiseptica. Extensive variation has been observed in the Pertactin repeat regions 6 1 and 2, as well as in other regions of the protein, and this variation has been associated with the recent resurgence of pertussis. Th is variation is not only inter-speciﬁ c, but also occurs between isolates from the same species. Knowledge about codons that are under positive selection could possibly facilitate the development of more broadly protective vaccines.

In this study, a large number of Pertactin genes from B. bronchiseptica, B. parapertussishu,

B. parapertussisov and B. pertussis were compared using diﬀ erent nucleotide substitutions models, and positively selected codons were identiﬁ ed using an empirical Bayesian approach. Th is approach yielded 15 codons subject to diversifying selection pressure. Th e results were interpreted in an immunological context and may help in improving future pertussis vaccines.

114 Evolution of Pertactin introduction Th e very closely related pathogens Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica (referred to as the mammalian bordetellae) express a similar array of virulence factors, including Pertactin. B. pertussis is a strictly human pathogen that causes pertussis or whooping cough. B. parapertussis comprises two distinct lineages, found in humans and sheep, designated B. parapertussishu and B. parapertussisov, respectively. B. bronchiseptica has been isolated from a large number of mammalian host species, and recently it was shown that a human-associated lineage also exists (Diavatopoulos et al., PLoS pathogens, 2005). It was further shown that B. pertussis and B. parapertussishu evolved from distinct branches of a B. bronchiseptica-like ancestor 1,2 (Diavatopoulos et al., PLoS Pathogens, 2005). Pertactin (Prn) belongs to the type V autotransporter protein family 3-5, and these proteins 6 are characterized by the ability to catalyze their own transport through the outer membrane. After secretion, autoproteolytic activities reduce the 69-kDa protein to its fi nal 60.37 or 58.34 kDa forms 6, which remain non-covalently bound to the bacterial cell surface 7. X-ray crystallography indicated that Prn consists of a 16-stranded parallel β-helix with a V-shaped cross-section 8. From this helix, several loops protrude, one of which contains the Arg-Gly- Asp (RGD) motif that is associated with adherence to host tissues 8-11. Th e protein further contains two hyper-variable regions, designated region 1 (R1) and region 2 (R2), which are comprised of amino acid (AA) repeats (Gly-Gly-X-X-Pro and Pro-Gln-Pro, respectively). R1 is located proximal to the N-terminus and directly adjacent to the RGD motif, and R2 is located at the C-terminus. One of the known biological functions of Prn is that it serves as an adhesin to the epithelium 12. Th e exact host receptor to which Prn binds is unknown. Pertactin elicits high antibody titers, and anti-Prn antibodies (Abs) have been shown to confer protective immunity 13 (Hijnen et al., submitted). Furthermore, anti-Prn Abs, but not anti-Ptx, anti-fi mbriae, or anti-FHA antibodies, were found to be crucial for B. pertussis phagocytosis 14, indicating an important role of Prn in immunity to pertussis. Especially polymorphism in R1 and R2 has been suggested to be important for evasion of antibody responses 13,15. Mice vaccinated with B. pertussis Prn1 were less well protected against B. pertussis Prn2 strains then to B. pertussis Prn1 strains 13, and the only diff erence between these Prn types is the presence of an additional repeat unit in R1 in Prn2. Further, vaccination of mice with B. pertussis Prn1 did not protect against infection with B. parapertussis 16,17, suggesting a lack of Prn cross-reactivity between these species. Th ese observations indicate that anti-Prn Abs signifi cantly aff ect transmission of the Bordetellae, at least for B. pertussis. A new light has been cast on the observed variation in Prn with the recent identifi cation of a Bordetella phage (BPP-1, Bvg Plus tropic Phage–1) 18. Th e mammalian bordetellae can switch between Bvg-phases (Bordetella virulence gene), depending on environmental stimuli. Phase switching results in a diff erent expression of surface-associated molecules,

115 Chapter 

including Prn, which is specifi cally expressed in the Bvg+ phase 19. BPP-1 showed a marked tropism for the Bvg+ phase of B. pertussis, B. parapertussis, and B. bronchiseptica 18, and its primary receptor for BPP-1 was shown to be Prn. Phase switching of the bacteria would expectedly result in the loss of the receptor for BPP-1. However, BPP-1 has been shown to specifi cally generate polymorphism in its ligand-binding domain, resulting in phages with increased binding capacity to alternative surface receptors for host cell entry 18,20,21. Extensive variation has been observed in the Prn repeat regions 1 and 2, as well as in other regions of the protein, both between species but also between strains belonging to the same species. Prn is an important component of many current acellular pertussis vaccines. Th erefore, knowledge about codons that are under positive selection could possibly facilitate the development of more broadly protective vaccines. Our aim was to identify regions 6 in Prn that are subject to positive selection within the Prn gene and to put this into an immunological framework. Pertactin genes of the mammalian bordetellae were compared using diff erent models of nucleotide substitutions, and positively selected codons were identifi ed using an empirical Bayesian approach. Th e location of positively selected sites was visualized in the crystal structure of B. pertussis Prn and compared to the location of epitopes.

116 Evolution of Pertactin experimental procedures sequence data and alignment In this study, the nucleotide sequence encoding the extracellular domain of Prn was used, represented by AAs 1-Asp to 677-Gly in the B. pertussis Tohama sequence. Regions 1 (232- Gly to 256-Pro) and 2 (545-Pro to 566-Pro), which contain repeats, were excluded from the positive selection analysis, which resulted in 1908 nucleotides in total (or 636 codons). Nucleotide sequences were obtained from a previous study (Diavatopoulos et al., PLOS pathogens, 2005), and additionally the Genbank database was searched for Prn nucleotide sequences that included the region encoding the extracellular domain. Th is search yielded 147 prn sequences, of which 25 were unique (Figure 1). Nucleotide sequences were aligned 6 using Kodon 2.5 (Applied Maths, Sint-Martens-Latem, Belgium) and alignment gaps were omitted. detection of selection Detection of selective pressures acting on individual codons within genes is generally estimated from varying ratios (ω) of non-synonymous (dN) to synonymous mutations (dS). In the case of positive selection, ω is expected to be >1. However, positive selection often occurs only at a limited number of codons and therefore, the ω for the complete gene may be <1, although the ω for individual codons can still be >1. A maximum-likelihood approach can be used to estimate varying ω-values across sequences using diff erent models of codon evolution 22. Th ese models assume a certain statistical distribution of ω and estimate the likelihood for the model, thereby also accounting for the phylogenetic relationships of the sequences. Th e following models are commonly compared to estimate the distribution of ω: model M1A to M2A and model M7 to M8. Th e fi rst nested model pair consists of the “nearly neutral model” M1A and the positive selection model M2A 23. In M1A, codons are assigned to two classes of ω of which the value is between 0 and 1, thus always assuming essentially neutral evolution. Th is model is compared to the “positive selection” model M2A, which has exactly the same codon classes as the M1A model, but in this model an additional class of codons is allowed with ω free to assume a value >1. Th e second nested pair consists of models M7 and M8. M7 assumes eight codon classes that are distributed in a β-shaped manner with 0<ω<1. Model M8 is similar to model M7, but diff ers in the existence of an additional codon class with ω>1. A likelihood ratio test respectively compares each nested model pair, and this gives an estimation of the extent of positive selection acting on the sequences under investigation. If M1A is rejected in favor of M2A, positive selection may be concluded. Similarly, positive selection can also be concluded if M7 is rejected in favor of M8. In the case of positive selection, individual positively selected codons can be identifi ed using a Bayes empirical Bayes approach 24. Likelihood ratio tests were performed using the CODEML program from the PAML software package

117 Chapter 

25. Statistical signifi cance for the fi t of the models to the actual data was obtained using 2 a χ -test. Twice the diff erence in log-likelihood (2Δl=2(l1-l0) with l the log-likelihood of a model was calculated, and this was compared to a χ2 distribution with two degrees of freedom and a 95% confi dence interval. Maximum likelihood trees were reconstructed from the aligned 25 prn sequences using Treefi nder (Jobb, G. 2005 Treefi nder version of June 2005, Munich, Germany, http://www.treefi nder.de) under the general-time-reversible (GTR) model of nucleotide substitution, with 1000 bootstrap replicates.

epitopes in pertactin In previous studies, a Pepscan was used to map the location of linear epitopes recognized by mAbs 26, and site-directed mutagenesis (SDM) was used to identify the location of 6 conformational epitopes recognized by monoclonal antibodies (mAbs) (Hijnen et al, submitted). In addition to these experimentally identiﬁ ed epitopes, the location of additional, putative discontinuous epitopes was determined using the Conformational Epitope Prediction (CEP) server (http://bioinfo.ernet.in/cep.htm) 27. CEP predicts the location of conformational epitopes, based on the exposure of stretches of AAs that are located in a 6Å proximity of each other. For this analysis, the crystal structure of B. pertussis Prn (1DAB.pdb) was used 8,28.

solvent exposure and secondary structure Th e solvent accessibility of AAs is a measure for their surface exposure; AAs with very high solvent accessibility are thus more likely targets for e.g. the immune system while AAs with low solvent accessibility will not likely come into contact with antibodies. Th e solvent accessibility was determined for each AA in Prn 1DAB.pdb with the program Getarea1.1 29. Since Getarea1.1 was only able to determine the exposure of the AAs present in the crystal structure, we also used the PredictProtein server to determine the solvent accessibility for the remainder of Prn 30. Th e secondary structure of the entire Pertactin protein was determined with the PredictProtein server as well.

visualization of sites Th e coordinates of Prn were downloaded from the Protein Data Bank, under code 1DAB. pdb 8,28. Th e location of epitopes identiﬁ ed by the above described methods and the location of positively selected sites were visualized in Chimera 31.

118 Evolution of Pertactin results Based on housekeeping gene sequence data, comparative genomic hybridization and Pertactin sequence data, it was previously shown that B. pertussis forms a separate branch with B. bronchiseptica complex IV, and B. parapertussis clusters together with B. bronchiseptica complex I (Diavatopoulos et al., PLoS Pathogens, 2005). Interestingly, B. bronchiseptica complex I strains and B. parapertussishu cannot be discriminated based on their prn genes,

B. bronchiseptica (DQ141702) 100 B. bronchiseptica (DQ141714) 21 G ~ Q 1.7 ~ 24 P G B. bronchiseptica (DQ141701) ~ 213 Q D 330 Q~ R 100 270 S ~ T B. bronchiseptica (DQ141768) 334 H ~ R 31.5 358 D~ Q 6 B. parapertussis (DQ141780) 488 A ~ G ov 91.6 376 I ~ S B. bronchiseptica (DQ141722) 0 B. bronchiseptica I B. bronchiseptica (DQ141751) and 0 B. parapertussishu III B. parapertussishu (DQ141779) 41.5 187 S ~ F B. bronchiseptica (AJ245927) 83.2 B. bronchiseptica (DQ141771) 84.7 B. bronchiseptica (AY376325) 88.9 438 V ~ C B. bronchiseptica (X54815)

B. pertussis (AF456357; prn6) 98.6 B. pertussis (AJ133245; prn8) 87.3 B. pertussis (AF218785; prn9) 79.5 498 R ~ L B. pertussis (AJ007362)

B. bronchiseptica (DQ141764) 86.1 B. bronchiseptica (DQ141719) 398 S ~ A 87.8 B. bronchiseptica IV B. bronchiseptica (DQ141725) and 96.9 B. pertussis II B. bronchiseptica (DQ141765) 83.2 B. bronchiseptica (DQ141732) 98.5 350 K ~ Q B. bronchiseptica (DQ141762) 85.3 B. bronchiseptica (DQ141741) 75.9 B. bronchiseptica (DQ141766) 77.7 B. bronchiseptica (DQ141774)

0.0 0.004 0.008 0.012 0.016 0.02 0.024 0.028

Figure 1. Maximum likelihood tree of 25 unique prn sequences, encoding the exposed domain of Prn, with the exclusion of regions 1 and 2 and alignment gaps. Accession numbers are indicated between parentheses. Posi- tively selected amino acids and their substitutions are indicated in boxes. Numbers near the branches indicate the bootstrap values, based on 1,000 bootstrap replicates. The scale indicates the evolutionary distance in substitutions per site.

119 Chapter 

while the housekeeping gene tree does allow distinction between the two complexes. A Genbank search yielded 25 unique prn sequences, out of a total of 147 sequences that included the nucleotide region encoding the surface-exposed domain. Th ese sequences included ten B. bronchiseptica complex I, nine B. bronchiseptica complex IV, four B. pertussis,

one B. parapertussisov and one B. parapertussishu sequences. A maximum-likelihood tree of these sequences is shown in Figure 1.

sequences coding for prn are subject to positive selection Positive selection can be estimated from the ratio (ω) of non-synonymous substitutions

(dN) to synonymous substitutions (dS). For genes and codons evolving under positive selection pressure, ω is expected to be larger than one, indicating that mutations in those 6 codons resulting in amino acid (AA) changes are selected for. We used a likelihood ratio test (LRT) to determine if Prn was evolving under positive selection pressure. In the LRT, the likelihood of a model assuming positive selection is compared to a model that assumes a diff erent (non-positive) selection. Th e LRT indicated that for Prn, both models M2A and M8 had a signifi cantly better likelihood (p>0.95) than models M1A and M7, respectively. Th is suggests that positive selection could be detected for some codons in Prn. Although the average ω for Prn was 0.29, the ω-values for the additional codon class in M2A and M8 were well above 1. An empirical Bayesian approach identifi ed 11 codons in model M2A and 15 codons in model M8 to be under positive selection. Th e 15 codons identifi ed in model M8 also contained the 11 codons that were identifi ed under the M2A model (Table 1). Of the 15 positively selected codons in Prn, all but one (codon 22) resulted in an AA change. Interestingly, although the fi rst two nucleotides of codon 22 had been substituted, this did not result in an AA change.

characterization of positively selected codons in prn In Figure 2, the positively selected codons are indicated on the primary structure of Prn. Of the 15 sites predicted to be positively selected for, the majority was located in the N- terminus or in the center (87%). Only two (13%) positively selected sites could be identiﬁ ed in the C-terminus of Prn, and these were only detected using the M8 model, which has been described to be less conserved than the M2A model 23,24. Amino acids may be part of a putative conformational or discontinuous epitope if they are within a 6Å proximity, and their solvent accessibility is more than 25% 27. Positively selected codons that correspond to these criteria may be recognized by a single antibody species and were therefore designated regions. In total, we identiﬁ ed four regions (A-D), representing 11 codons. Four codons could not be assigned to a region under these criteria (Fig. 2, 3 and Table 1).

120 Evolution of Pertactin 1 -sheet; 498R β , 677-Gly β 1 C' 488A 1 y selected codons. Yellow Yellow y selected codons. cated by black triangles. AA- black triangles. by cated the connecting lines. Numbering the connecting lines. complex I, C, loop or coil; reen triangles indicate loops that after triangles indicate reen Not characterized by by Not characterized crystallography X-ray ively 6 B. bronchiseptica B. bronchiseptica 488 498 ; BB-1; ov anks pepscan epitope A color version of this fi gure is available in the appendix. is available gure of this fi version A color 398 438 B. parapertussis 376 358 ; BPP-ov ; BPP-ov 350 hu 334 330 Region C Region D B. parapertussis complexes Bordetella erent 270 213Q 270S 330Q 334H 350K 358D 376I 398S 438V RGD R1 R2 1 ect in binding with mAbs. or increase 232-Gly to 256-Pro232-Gly to 566-Pro to 545-Pro complex IV; BPP-hu complex IV; 187 213 B. bronchiseptica B. bronchiseptica model M8 ed only by ------L 21G 22S 24P 187S ; BB-4, B. pertussis rst amino acid of the mature protein. Green boxes indicate the location of conformational epitopes co-localized with positivel epitopes the location of conformational indicate boxes Green protein. rst amino acid of the mature Region A Region B 21, 22, 24 22, 21, 123 4 56 910 13 14 1517 18 1920 2 Location of positively selected codons and regions on the primary structure of Tohama Prn. Positively selected codons are indi selected are codons Positively Prn. Tohama on the primary selected and regions codons of positively structure of Location Characteristics of positively selected sites in the diff Characteristics of positively N' -sheet adjacent to coil; CEP, conformational epitope prediction; SDM, site-directed mutagenesis; S+P, ; f-P, fl ; f-P, mutagenesis; S+P, SDM, site-directed conformational epitope prediction; -sheet adjacent to coil; CEP, β boxes indicate the co-localization of both conformational and linear epitopes with positively selected codons. Red, white and g white Red, selected codons. with positively and linear epitopes the co-localization indicate of both conformational boxes no eff a decrease, respectively SDM showed mutation by Figure 2. Figure residues with a maximal distance of 6Å and a minimum solvent accessibility of 25% have been designated as regions, indicated by indicated as regions, been designated of 25% have accessibility of 6Å and a minimum solvent with a maximal distance residues starts with the fi BP (4; 60) BB-4 (9; 9) (1; 10)BPP-hu (1; 3)BPP-ov BB-1 (10) (%) exposure Solvent inLocated Q 65.8CEP -SDM/PEPSCAN Q 87.2 - - Q 88.2 - C S+P G 0.1 - - G S+P 59.9 C Yes F S+P G - 26.2 - Yes C D No 36.2 - Yes - D 41.1 No β T No D 44.4 - T No C No - 39.3 T P - - 25.2 No β R 35.5 R P Yes - R β-C 29.2 - Yes R β-C S 64.5 Q - No Q β-C 84.7 - No - Q Yes β-C S f-P Q Yes S S C No - No S A - β No No - - β-C No - - Yes G C C S Yes - G - C G - - - -C, Numbers between brackets indicate the number of unique Prn sequences and the total number of represented Prn sequences, respect Prn sequences and the total number of represented brackets indicate the number of unique Prn between Numbers Positively selected codons identifi Positively 1-Asp Table 1. Table a b Abbreviations: BP, BP, Abbreviations: β

121 Chapter  Gly ⊂ +488-Ala Arg ⊂ +498-Leu +498-Leu Cys codons not present in a region are colored colored are in a region not present codons ⊂ 6 +438-Val +438-Val Ala ⊂ Prn1 (1DAB.pdb). Numbers indicate the positively selected the positively Numbers indicate (1DAB.pdb). Prn1 Gln ⊂ Ser +398-Ser ⊂ Gly ⊂ B. pertussisB. Region D +358-Asp +358-Asp +376-Ile Gln ⊂ +334-His Region C +350-Lys +350-Lys Arg ⊂ crystal structure of

+330-Gln Thr ⊂ +270-Ser Region B Phe Asp ⊂ ⊂ +187-Ser +213-Gln Gly Ser Silent Projection of the positively selected codons and regions on the selected and regions codons of the positively Projection ⊂ ⊂ ⊂ A color version of this fi gure is available in the appendix. is available gure of this fi version A color Region A +21-Gly +22-Ser +24-Pro +24-Pro Figure 3. Figure codons. Codons that are part of a region are colored in red, and regions are indicated by black rectangles. Positively selected Positively black rectangles. by indicated are and regions in red, colored part are that are Codons of a region codons. in blue.

122 Evolution of Pertactin positively selected regions in prn Region A, located in the N-terminus of Prn, consisted of three positively selected codons that were in very close proximity (AAs 21, 22 and 24) and located in an exposed loop of Prn, designated loop 2 (Fig. 3). In a previous report we described that the N-terminus contained a number of conformational epitopes (Hijnen et al., submitted), and one of these N-terminal conformational epitopes was found to co-localize with Region A. Th e highly variable loop 2 contains 6 codons (20-Gln to 25-Gly), and 12 of the respective 18 nucleotides were found to be polymorphic between B. pertussis and B. bronchiseptica complex IV on the one hand and B. parapertussis and B. bronchiseptica complex I on the other hand. Th e solvent accessibility of the AAs in loop two was also very high, suggesting they are all well exposed (Table 1). 6 Region B comprised two codons (213-Gln and 270-Ser), located partially in the N-terminus and partially in the center of Prn. Although separated by 57 AAs in the primary sequence, they are within a 4Å radius of each other in the crystal structure. Of these two AAs, 213- Gln is well exposed (59.9%), but 270-Ser is only 26.2% exposed to solvent. In the -sheet where 270-Ser is located (AAs 261-274), only three AAs (including 270-Ser) are exposed to solvent. Both codons were not part of previously identifi ed epitopes 26 (Hijnen et al., submitted), or epitopes predicted by CEP. Th ree positively selected codons comprise region C (330-Gln, 334-His and 358-Asp). Th e solvent accessibility of these three AAs indicates they are all well exposed to the environment. Further, these residues were predicted to co-localize with a putative conformational epitope, as predicted by CEP. In the center of Prn, three closely located positively selected codons were identifi ed within a 6Å radius (Region D; 350-Lys, 376-Ile and 398-Ser). Although 350-Lys and 398-Ser are well exposed, 376-Ile is only 25.2% exposed. In contrast, the α-helix adjacent to 376-Ile (373-Gly to 375-Ser) is well exposed. It is likely that the mutation of the hydrophobic 376- Ile to a hydrophilic serine may have an eff ect on the tertiary structure, or on the location of the adjacent α-helix. positively selected codons in prn not assigned to a region Located in the N-terminus of Prn, 187-Ser was predicted to be under positive selection. Although 187-Ser was in a 6Å radius of Region B (see above), the β-sheet in which 187- Ser is located is inaccessible to solvent, suggesting that this β-sheet does not constitute an epitope. Th e mutation of the hydrophilic 187-Ser to a bulky aromatic phenylalanine, as observed in a number of Prn variants, will likely aff ect the local tertiary structure of the protein, possibly indirectly aff ecting the exposure of adjacent epitopes or leading to a change in receptor binding. Th is may suggest an indirect role of this loop in antigenic variation. Positively selected residue 438-Val was mutated into a cysteine in two B. bronchiseptica

123 Chapter 

Prn sequences. Th is mutation is very unusual, as cysteine residues are not normally present in Prn. Residue 488-Ala, located in the beginning of the C-terminus, was also predicted to be under positive selection. Th e loop in which this AA resides was predicted to be part of fi ve distinct putative conformational epitopes, suggesting that it is well exposed and possibly very immunogenic. Two of these predicted epitopes also contained the loop that is comprised of AAs 428-436. Th is loop is fl anked by residue 438-Val (see above) which was also found to be under positive selection. It is likely that these two mutations aff ect the structure and location of several of these conformational epitopes. Th e last C-terminally located positively selected site was the well exposed residue 498-Leu, which was predicted by CEP to be part of two conformational epitopes. In a previous study, we also identifi ed this residue as part of a conformational epitope recognized by both 6 human and mouse Abs (Hijnen et al., submitted).

analysis of repeat regions  and  R1 and R2 are located in the N-terminus and in the C-terminus, respectively. R1 is comprised of repeats of fi ve AA in length, which may also vary in composition (GXXXP), and it is located adjacent to the RGD-site. Th e RGD site has been implicated in adherence of the bacterium to host cells 9, but it is likely that other, uncharacterized domains may also be involved. R1 has been shown to induce Prn-specifi c Abs and variation in this region aff ected the effi cacy of the Dutch whole cell vaccine 13,32. Th is suggests that variation in R1 is important for evasion of host immunity. Diversity in R1 and R2 was also observed in prn sequences which were otherwise conserved (Table 2). Th e length of R1 was found to be statistically signifi cantly associated with the length of R2. Longer R1 sequences were associated with shorter R2 sequences, and vice versa (Pearson correlation P<10e-16). Further, the ratio of R1 to R2 length was found to be phylogenetically associated. Long R1 sequences and short R2 sequences were found almost exclusively in the Table 2. Region 1 and 2 characteristics human-associated B. pertussis and B. Length in amino acids bronchiseptica complex IV strains. In Complex Region 1 Region 2 contrast, short R1 sequences combined I 16.7 (± 3.1)1 25.5 (±2.9) II 26.5 (± 3.3) 14.6 (± 1) with long R2 sequences were observed III 20 (± 0) 31.4 (±1.3) predominantly in the B. parapertussis IV 25 (± 6.6) 22.2 (± 2.4) and B. bronchiseptica complex I strains 1Numbers between parentheses indicate the standard deviation (Table 2).

124 Evolution of Pertactin discussion In this study, we provide evidence for the presence of positively selected codons in the autotransporter protein Pertactin. Th e majority of these codons were well exposed to solvent and located in linear or conformational epitopes, suggesting that adaptive changes in these codons may lead to immune escape, decreased phage-recognition or a better fi t with the host receptor. Further, the length of repeat region 1 was found to be signifi cantly associated to that of region 2, and possible explanations are put forward for this association. characterization of positively selected codons An analysis of Prn from which the hyper-variable R1 and R2 were excluded resulted in 25 unique Prn sequences. Th ese sequences were analyzed for positive selection using a 6 likelihood ratio test and empirical Bayes estimates. Th is approach identifi ed 15 codons that were subject to positive selection. Out of these 15 codons, 14 were exposed for more than 25% to solvent, indicating they are surface exposed and therefore likely to be aff ected by the immune system or phage binding. Th e majority of the positively selected codons was located in (n=6), or directly adjacent to a loop (n=5). In contrast, only three positively codons were located in a β-sheet, including the only non-exposed codon (Table 1). Th ese data indicate that in Prn, amino acids located in or near loops are more amenable to diversifying selection than those in β-sheets. Th e backbone of Prn is comprised of mainly β-sheets and variation in the composition of these sheets may result in structural changes and thus possibly loss of biological function. Variation in the exposed loops however, is not likely to aff ect the overall structure and function of the protein, and thus these loops may be important for immune evasion. Th e majority of positively selected codons (n=10) co-localized with linear and conformational epitopes that were predicted by CEP or experimentally identifi ed previously (Hijnen et al., submitted) 26,27. A total of six codons were located in linear epitopes recognized by human Abs to B. pertussis Prn (Table 1) 26. Further, we recently modifi ed exposed loops of B. pertussis Prn by site-directed mutagenesis (SDM), after which the binding of well-characterized monoclonal antibodies (mAbs) to these Prn variants was investigated (Hijnen et al., submitted). Several of these Prn variants showed a decreased affi nity for a number of mAbs, indicating that mutations in these loops may be important for immune evasion. Out of the 15 positively selected codons, fi ve were located in loops of which modifi cation by SDM resulted in decreased affi nity to at least three mAbs. In the same study, several loops in Prn were identifi ed that upon SDM showed an increase in binding with mAbs (Hijnen et al., submitted). We hypothesized that these mutations aff ected the conformation of the loop, thereby unmasking epitopes. Since these loops could be important for masking of epitopes, they are possibly under purifying selection pressure. Consistent with this hypothesis, none of the codons that we identifi ed to be under positive selection in this work were located in these loops. Th e majority of the positively selected codons were found to be diff erent between B. bronchiseptica complex I

125 Chapter 

and B. parapertussis versus B. bronchiseptica complex IV and B. pertussis. Previously, it was shown that, although vaccination with B. pertussis Prn1 protected at least partially against B. pertussis strains (including those with diﬀ erent Prn sequences) 13,33, it did not protect against B. parapertussis 16,17. Th e latter observation is consistent with an important role of the variable codons in immune evasion. We previously provided evidence that B. pertussis and B. bronchiseptica complex IV strains were subject to immune competition resulting in antigenic divergence between these two species (Diavatopoulos et al., PLoS Pathogens, 2005). Th is analysis was based on the presence or absence of genes coding for dermonecrotic toxin, pertussis toxin and LPS. Here we looked for more subtle changes due to amino acid substitutions in Prn. Two substitutions, in the codons 350 and 398, were found that may have been caused by immune 6 competition between B. pertussis and B. bronchiseptica complex IV strains. In B. pertussis en B. bronchiseptica complex I strains, these codons code for Lys and Ser, respectively. In contrast, in B .bronchiseptica complex IV strains, the residues Gln and Ala are found at these positions, respectively. Similarly, the polymorphism in codon 187 may be due to immune

competition between B. pertussis and B. parapertussishu. All Bordetella species code for Ser at

this position, except for B. parapertussishu in which Phe is found at this position. Although data about the location of epitopes were available, functional data concerning receptor specifi city and residues possibly involved in this interaction were unavailable. We therefore compared the location of the positively selected AA residues present in human adapted strains with the animal adapted strains, in order to locate residues possibly involved in host receptor specifi city. Th is approach yielded two residues, 213 and 270 that could possibly play a role in host receptor specifi city. Both residues are identical in B. pertussis and

B. bronchiseptica complex IV strains, but diff erent in B. parapertussishu and B. bronchiseptica complex I strains. Furthermore, both residues were not previously identifi ed as an epitope, or predicted to be part of an epitope. Both residues are located closely together in between two large loops (Fig. 3). Th is creates a groove that could be a potential receptor binding site. Th e subtle variations observed for these residues (Q>D and S>T), which are located on the bottom of the groove, could potentially enhance the affi nity or the fi t to the human receptor.

Th e fact that B. parapertussishu and B. pertussis Prn are distinct at these two positions may

reﬂ ect the more recent adaptation of B. parapertussishu to humans.

polymorphism in repeat regions  and  Comparison of R1 and R2 sequences between the 147 isolates revealed a striking correlation between the length of R1 and R2. Long R1 sequences were found to be accompanied with short R2 sequences, and vice versa (Pearson correlation P<10e-16). Th is association was correlated to the phylogenetic tree, high R1 to R2 ratios were found almost exclusively in the human-associated B. pertussis and B. bronchiseptica complex IV isolates; low R1:R2 ratios were observed mainly in the B. bronchiseptica complex I and B. parapertussis isolates. We

126 Evolution of Pertactin previously provided evidence that one of the roles of R1 was masking of epitopes (Hijnen et al., submitted). Further, we observed that R1 and R2 may be part of a single discontinuous epitope, implicating close proximity of these regions. In the light of these observations, it is plausible that variation in the length of R1 is compensated by variation in the length of R2 to maintain the close proximity of the variable epitope, or to maintain masking of underlying epitopes.

In this study we have identifi ed codons of Prn that are under diversifying selection. Th e results we obtained are largely consistent with immunological and structural data. Our analyses may facilitate the development of more eff ective vaccines against pertussis by identifying regions which induce an eff ective immune response and are not subject to diversifying selection. It should be noted that variation in Prn may not only be driven by the interaction 6 with the host. Recently, phage BBP-1 was described that infects Bvg+ bordetellae via Prn as its main receptor. It is likely that this phage has had a diversifying eff ect on Prn.

127 Chapter 

reference list Comparative roles of the Arg-Gly-Asp sequence present in the Bordetella pertussis adhesins pertactin and fi lamentous hemagglutinin. Infect. 1. Musser,J.M., Hewlett,E.L., Peppler,M.S., and Immun. 60, 2380-2385. Selander,R.K. (1986). Genetic diversity and relationships in populations of Bordetella spp. J. 12. Everest,P., Li,J., Douce,G., Charles,I., De Azavedo,J., Bacteriol. 166, 230-237. Chatfi eld,S., Dougan,G., and Roberts,M. (1996). Role of the Bordetella pertussis P.69/pertactin 2. van der Zee,A., Mooi,F., van Embden,J., and protein and the P.69/pertactin RGD motif in the Musser,J. (1997). Molecular evolution and host adherence to and invasion of mammalian cells. adaptation of Bordetella spp.: phylogenetic Microbiology 142 ( Pt 11), 3261-3268. analysis using multilocus enzyme electrophoresis and typing with three insertion sequences. J. 13. King,A.J., Berbers,G., van Oirschot,H.F., Bacteriol. 179, 6609-6617. Hoogerhout,P., Knipping,K., and Mooi,F.R. (2001). Role of the polymorphic region 1 of the 3. Henderson,I.R., Navarro-Garcia,F., and Nataro,J.P. Bordetella pertussis protein pertactin in immunity. 6 (1998). The great escape: structure and function of Microbiology 147, 2885-2895. the autotransporter proteins. Trends Microbiol. 6, 370-378. 14. Hellwig,S.M., Rodriguez,M.E., Berbers,G.A., Van De Winkel,J.G., and Mooi,F.R. (2003). Crucial Role 4. Henderson,I.R., Cappello,R., and Nataro,J.P. (2000). of Antibodies to Pertactin in Bordetella pertussis Autotransporter proteins, evolution and redefi ning Immunity. J. Infect. Dis. 188, 738-742. protein secretion. Trends Microbiol. 8, 529-532. 15. Mooi,F.R., van Loo,I.H., and King,A.J. (2001). 5. Henderson,I.R. and Nataro,J.P. (2001). Virulence Adaptation of Bordetella pertussis to Vaccination: functions of autotransporter proteins. Infect. A Cause for Its Reemergence? Emerg. Infect. Dis. Immun. 69, 1231-1243. 2001. ;7. (3 Suppl):526. -8. 7, 526-528.

6. Gotto,J.W., Eckhardt,T., Reilly,P.A., Scott,J.V., 16. Khelef,N., Danve,B., Quentin-Millet,M.J., and Cowell,J.L., Metcalf,T.N., III, Mountzouros,K., Guiso,N. (1993). Bordetella pertussis and Bordetella Gibbons,J.J., Jr., and Siegel,M. (1993). Biochemical parapertussis: two immunologically distinct and immunological properties of two forms of species. Infect. Immun. 61, 486-490. pertactin, the 69,000-molecular-weight outer membrane protein of Bordetella pertussis. Infect. 17. David,S., van,F.R., and Mooi,F.R. (2004). Effi cacies of Immun. 61, 2211-2215. whole cell and acellular pertussis vaccines against Bordetella parapertussis in a mouse model. Vaccine 7. Miller,E. (1999). Overview of recent clinical trials of 22, 1892-1898. acellular pertussis vaccines. Biologicals 27, 79-86. 18. Liu,M. et al. (2002). Reverse transcriptase-mediated 8. Emsley,P., Charles,I.G., Fairweather,N.F., and tropism switching in Bordetella bacteriophage. Isaacs,N.W. (1996). Structure of Bordetella pertussis Science 295, 2091-2094. virulence factor P.69 pertactin. Nature 381, 90-92. 19. Kinnear,S.M., Boucher,P.E., Stibitz,S., and 9. Leininger,E., Roberts,M., Kenimer,J.G., Charles,I. Carbonetti,N.H. (1999). Analysis of BvgA activation G., Fairweather,N., Novotny,P., and Brennan,M. of the pertactin gene promoter in Bordetella J. (1991). Pertactin, an Arg-Gly-Asp-containing pertussis. J. Bacteriol. 181, 5234-5241. Bordetella pertussis surface protein that promotes adherence of mammalian cells. Proc. Natl. Acad. 20. Doulatov,S., Hodes,A., Dai,L., Mandhana,N., Sci. U. S. A. 88, 345-349. Liu,M., Deora,R., Simons,R.W., Zimmerly,S., and Miller,J.F. (2004). Tropism switching in Bordetella 10. Roberts,M., Fairweather,N.F., Leininger,E., bacteriophage defi nes a family of diversity- Pickard,D., Hewlett,E.L., Robinson,A., Hayward,C., generating retroelements. Nature 431, 476-481. Dougan,G., and Charles,I.G. (1991). Construction and characterization of Bordetella pertussis 21. Liu,M. et al. (2004). Genomic and genetic analysis mutants lacking the vir-regulated P.69 outer of Bordetella bacteriophages encoding reverse membrane protein. Mol. Microbiol. 5, 1393-1404. transcriptase-mediated tropism-switching cassettes. J Bacteriol. 186, 1503-1517. 11. Leininger,E., Ewanowich,C.A., Bhargava,A., Peppler,M.S., Kenimer,J.G., and Brennan,M.J. (1992). 22. Yang,Z. (2000). Maximum likelihood estimation

128 Evolution of Pertactin

on large phylogenies and analysis of adaptive Sequence variation in pertussis S1 subunit toxin evolution in human infl uenza virus A. J. Mol. Evol. and pertussis genes in Bordetella pertussis 51, 423-432. strains used for the whole-cell pertussis vaccine produced in Poland since 1960: effi ciency of the 23. Wong,W.S., Yang,Z., Goldman,N., and Nielsen,R. DTwP vaccine-induced immunity against currently (2004). Accuracy and power of statistical methods circulating B. pertussis isolates. Vaccine 22, 2122- for detecting adaptive evolution in protein coding 2128. sequences and for identifying positively selected sites. Genetics 168, 1041-1051.

24. Yang,Z., Wong,W.S., and Nielsen,R. (2005). Bayes empirical bayes inference of amino acid sites under positive selection. Mol. Biol. Evol. 22, 1107- 1118.

25. Yang,Z. (1997). PAML: a program package for phylogenetic analysis by maximum likelihood. 6 Comput Appl. Biosci. 13, 555-556.

26. Hijnen,M., Mooi,F.R., van Gageldonk,P.G., Hoogerhout,P., King,A.J., and Berbers,G.A. (2004). Epitope structure of the Bordetella pertussis protein P.69 pertactin, a major vaccine component and protective antigen. Infect. Immun. 72, 3716- 3723.

27. Kulkarni-Kale,U., Bhosle,S., and Kolaskar,A.S. (2005). CEP: a conformational epitope prediction server. Nucleic Acids Res. 33, W168-W171.

28. Berman,H.M., Westbrook,J., Feng,Z., Gilliland,G., Bhat,T.N., Weissig,H., Shindyalov,I.N., and Bourne,P. E. (2000). The Protein Data Bank. Nucleic Acids Res. 28, 235-242.

29. Fraczkiewicz,R. and Braun,W. (1998). Exact and effi cient analytical calculation of the accessible surface areas and their gradients for macromolecules. J Comput Chem 19, 319-333.

30. Rost,B., Yachdav,G., and Liu,J. (2004). The PredictProtein server. Nucleic Acids Res. 32, W321- W326.

31. Pettersen,E.F., Goddard,T.D., Huang,C.C., Couch,G. S., Greenblatt,D.M., Meng,E.C., and Ferrin,T.E. (2004). UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem 25, 1605-1612.

32. He,Q., Makinen,J., Berbers,G., Mooi,F.R., Viljanen,M. K., Arvilommi,H., and Mertsola,J. (2003). Bordetella pertussis protein pertactin induces type-specifi c antibodies: one possible explanation for the emergence of antigenic variants? J. Infect. Dis. 187, 1200-1205.

33. Gzyl,A., Augustynowicz,E., Gniadek,G., Rabczenko,D., Dulny,G., and Slusarczyk,J. (2004).

129 130

Chapter Chapter 77

General Discussion Chapter 

summarizing discussion Th e Bordetella genus comprises a diverse group of bacterial species, most of which are pathogens. Th e mammalian (or “classical”) bordetellae are comprised of the closely related species B. bronchiseptica, B. parapertussis and B. pertussis. Historically, these are the most thoroughly studied Bordetella species, due to the fact that they are important pathogens to humans, and farm and domestic animals. Although these species are very closely related, they are also remarkably diﬀ erent with respect to their host tropism and clinical course. B. bronchiseptica comprises a genetically diverse lineage that has been isolated from many diﬀ erent mammalian hosts and causes chronic respiratory tract diseases such as kennel cough in dogs and atrophic rhinitis in pigs 1. Th e genetically limited human pathogen B. pertussis is the etiological agent of the acute respiratory disease known as whooping cough, against which is widely vaccinated. For B. parapertussis, two distinct clonal lineages have

been described, a human-speciﬁ c lineage (B. parapertussishu) and a lineage found exclusively 7 2 in sheep (B. parapertussisov), with no evidence of exchange between the two reservoirs . Both

B. pertussis and B. parapertussishu cause whooping cough in humans, a highly contagious respiratory tract infection with an infectious period of three weeks or more 3. In general, B.

parapertussishu infections are less severe than B. pertussis infections. In contrast to B. pertussis

and B. parapertussishu, B. bronchiseptica usually causes chronic infections in many mammalian species, including humans, which often remain unnoticed. Further, B. bronchiseptica but not B. pertussis and B. parapertussis has been shown to be able to survive in the environment for a prolonged period of time 4,5. Recently, a third species has presumably causes pertussis-like symptoms, Bordetella holmesii. It was ﬁ rst described in 1995 as a human pathogen that was isolated from the blood of septicemic, immunocompromised patients 6,7. More recently however, B. holmesii has also been isolated from the respiratory tracts of immunocompetent patients with pertussis-like symptoms 8. B. holmesii has been suggested to be very closely related to B. pertussis, based on the presence of B. pertussis-like 16S rRNA genes and the B. pertussis-speciﬁ c insertion sequence element IS481 6,9, although other evidence suggests that it may not be closely related to B. pertussis 10. Not much is known about the biology and epidemiology of B. holmesii. Th e origin of the disease pertussis remains an enigma. Pertussis, which has very typical symptoms, was one of the major causes of childhood mortality prior to the introduction of vaccines in the 1940s and 1950s. However, no historical descriptions have been found in the European literature prior to the Middle Ages 11. In comparison, numerous descriptions have been found in the ancient Greek literature for diseases such as diphtheria and tetanus, which also had a major impact on child mortality and cause diseases with characteristic symptoms. Due to the lack of references to pertussis-like symptoms in the ancient (European) literature and the limited genetic diversity of B. pertussis, it has been assumed that B. pertussis only recently adapted to humans.

132 General Discussion

Previous studies indicated that the human-specifi c pathogens B. pertussis and B. parapertussishu are derived from distinct B. bronchiseptica-like lineages 2,12, although a specifi c ancestral lineage for both species has not yet been described. Recently, the complete genome sequences of the mammalian bordetellae have been published, and it was shown that the human- restricted pathogens B. pertussis and B. parapertussishu have undergone signifi cant genome reduction compared to B. bronchiseptica 13. It was suggested that this genome reduction was associated with host restriction of B. pertussis and B. parapertussishu, although the key molecular events are unknown. Th e phylogenetic position of B. holmesii in the Bordetella genus is still controversial, and it is unknown if and how this species emerged as a human pathogen.

Th e goal of this thesis was to elucidate the phylogenetic relationships between the Bordetella species causing respiratory infections in mammals, and to identify genetic factors that are 7 important for adaptation to the human host in order to further the understanding of host- adaptation in general and that of B. pertussis in particular.

In chapter 2, the genetic diversity and evolutionary relationships between B. bronchiseptica, B. parapertussis and B. pertussis are studied by sequencing of housekeeping genes from a large collection of strains. Th e results supported a phylogeny with four distinct complexes, representing B. pertussis (complex II), B. parapertussishu (complex III), and two distinct B. bronchiseptica populations (complexes I and IV). Sequencing of the pertactin virulence gene provided further support for this newly-described population structure. Th e distribution of four insertion sequence elements was determined, and it was shown that B. bronchiseptica complex IV and B. pertussis shared IS1663, implying either common descent or horizontal exchange. Divergence times were calculated for combinations of complexes from the housekeeping gene sequence data. Th ese calculations indicated that B. pertussis and B. bronchiseptica complex IV separated approximately 0.3-2.5 million years ago (Mya), which suggested a more recent divergence time than B. pertussis and B. bronchiseptica complex I, estimated at 1.1-5.6 Mya. Th e results also suggested that B. parapertussishu evolved from an animal-associated lineage of B. bronchiseptica (complex I), while B. pertussis evolved from a distinct B. bronchiseptica lineage (complex IV) that may already had a preference for hominids up to 2.5 million years ago. Extant members of this newly-identiﬁ ed B. bronchiseptica lineage were found to circulate in human populations and cause pertussis- like disease. Th e most plausible explanation from these data is that the association of B. pertussis with humans originated in the last common ancestor (LCA) of B. pertussis and B. bronchiseptica complex IV, although it should be noted that the adaptation of B. bronchiseptica complex IV strains to humans is not absolute. Th erefore, the possibility remains that the association of B. bronchiseptica complex IV to humans emerged

133 Chapter 

independently after the divergence of B. bronchiseptica complex IV and B. pertussis. Based on the assumption that the LCA of B. bronchiseptica and B. pertussis had an increased preference to humans, the apparent emergence of pertussis in Europe within the last 500 years may be attributable to import via travel or migration; or to the recent acquisition by B. pertussis of the ability to cause more severe, whooping cough-like symptoms. Th e newly-proposed evolutionary model provides a framework to investigate adaptation to the human host by comparison of B. pertussis to B. bronchiseptica complex I and complex IV.

In chapter 3, the gene content of B. bronchiseptica complex I, complex IV and B. pertussis strains were compared at a high resolution genome-wide level using comparative genomic hybridization (CGH) to a Bordetella microarray 14, in order to identify genetic events that are associated with adaptation of B. bronchiseptica complex IV and B. pertussis to the human host. Th e results of the CGH are interpreted in the context of the mammalian bordetellae 7 evolutionary scenario as proposed in chapter 2. CGH analysis identifi ed the absence of the complete pertussis toxin locus and dermonecrotic toxin gene, as well as a polymorphic LPS biosynthesis locus in B. bronchiseptica complex IV strains compared to B. bronchiseptica complex I and B. pertussis. Th e observed diff erences in gene content of the LPS biosynthesis locus were further investigated by LPS gel electrophoresis. In a similar vein to Gupta et al. 15 and Harvill et al. [15], it was proposed that diff erences between the human-associated Bordetella complexes with regard to these major virulence factors are the result of competition to avoid cross-immunity. Although immune- competition provides an attractive explanation for the observed diff erences, the possibility that these genetic diff erences may refl ect diff erences in niche-occupation or transmission cannot be ruled out. In addition to the observed gene content diff erences in various major virulence loci, analysis of CGH data also suggested a genome reduction of the genomes of B. bronchiseptica complex IV strains relative to B. bronchiseptica complex I strains. It should be noted that the possibility remains that the complex IV strains have acquired DNA elements, not covered by the microarray, that are important for host-adaptation. Gene content comparison identifi ed a total of 30 genes (including IS1663) that were statistically signifi cantly overrepresented in B. bronchiseptica complex IV strains compared to B. bronchiseptica complex I strains, and 16 of these 30 genes were previously shown to be specifi c to B. pertussis 14. In contrast, 237 genes were found to be absent or divergent in complex IV strain relative to complex I strains. Th ese results suggest that the genome of B. bronchiseptica complex IV is decreasing in size,

as was also observed for the human-adapted pathogens B. pertussis and B. parapertussishu. Genome reduction has been implied in host-restriction or a change of niche in a number 16-18 13 of pathogens , including B. pertussis and B. parapertussishu , and our results suggest that this process may have already started before the divergence of B. bronchiseptica complex IV and B. pertussis.

134 General Discussion

Based on CGH analysis, it was suggested that the genomes of B. bronchiseptica complex IV strains may have undergone genome reduction (chapter 3). Although DNA uptake seems to have played an insigniﬁ cant role in the evolution of the human-adapted species

B. pertussis and B. parapertussishu, the possibility remains that horizontally acquired DNA elements may have been important for host-adaptation and evolution of B. bronchiseptica complex IV strains. In order to identify putative horizontally acquired DNA sequences in B. bronchiseptica complex IV, the genome of the complex I strain B. bronchiseptica RB50, of which the complete genome sequence has been described 13, was subtracted from a mixture of ﬁ ve B. bronchiseptica complex IV genomes by subtractive hybridization (chapter 4), and the distribution of sequences speciﬁ c to complex IV was determined over complexes I and IV. Our results suggested that limited DNA acquisition has occurred in complex IV strains compared to complex I, and it seems unlikely that uptake of DNA elements has been pivotal to the evolution of B. bronchiseptica complex IV strains. Sequencing of the complete 7 genomes of several members of complex IV may provide the ultimate proof to determine the extent of genome reduction.

In chapter 5, the phylogenetic relationship between B. holmesii and the mammalian bordetellae was investigated using a combination of CGH to Bordetella microarrays and sequencing of multiple housekeeping genes. Analysis of CGH data indicated that the majority of the B. holmesii genome diverged signifi cantly from the mammalian Bordetella genomes, suggesting a more distant relationship than was previously assumed from comparison of 16S rRNA genes between these species. Th is was confi rmed by analysis of housekeeping gene sequence data, which demonstrated that B. holmesii was in fact much closer related to B. avium, an important pathogen of fowl, and B. hinzii than to the mammalian bordetellae. In spite of the relatively distant phylogenetic relationship between B. holmesii and the mammalian bordetellae, CGH detected a genomic region of 66 kb in the genome of B. holmesii, designated iron-uptake island (IUI) that was highly conserved in the mammalian bordetellae. Partial sequencing of genes in the IUI-region from several B. holmesii strains indicated near-identity to the orthologous sequences in the genomes of the mammalian bordetellae. Due to the presence of several IS481 elements in IUI, it was proposed that B. holmesii acquired this genomic island from B. pertussis. Screening of B. pertussis strains for B. holmesii IUI-specifi c features identifi ed the lineage to which the atypical B. pertussis strain 18323 belonged as a potential donor. By sequencing of 16S rRNA genes and Southern blot hybridization with 16S rRNA probes, it was shown that B. holmesii contained three 16S rRNA gene copies, which were all nearly identical to B. pertussis. Th e near identity of their 16S rRNA genes was proposed to be the result of horizontal gene transfer between B. pertussis and B. holmesii, possibly in conjunction with IUI.

135 Chapter 

Comparative genomic analysis identifi ed a region syntenic to the B. holmesii IUI in the genome of the avian pathogen B. avium. Using genome walking, it was shown that the region of synteny between B. avium, B. holmesii, and the classical bordetellae extended beyond the boundaries of IUI, suggesting that the backbone of this chromosomal region is conserved across distantly related Bordetella species. Th e presence of highly conserved sequences at the 5’ and 3’ borders of IUI were proposed to have served as the substrates for homologous recombination. Th e IUI contained all genes necessary for the biosynthesis, export, and uptake of the siderophore alcaligin, which is involved in iron scavenging in a eukaryotic environment. Th e alcaligin locus was shown to be transcriptionally activated under iron-depleted conditions, and under these conditions, alcaligin production was demonstrated as well. Characterization of IUI in 12 B. holmesii strains identifi ed a transposase-mediated insertion adjacent to the alcaligin operon that encoded a Fur-regulated FecIR-like system that is presumably involved 7 in regulation of iron-uptake. Detailed analysis of the region upstream of the alcaligin biosynthetic locus identifi ed a putative σ70 (FecI) binding site in the promoter region of alcABCDE, suggesting that this locus may also be regulated by the FecIR system, thereby shedding a new light on regulation of alcaligin production, which has been assumed to be solely regulated by Fur and the AraC-like regulator AlcR 19-21. Alcaligin production has not been detected in B. avium, and its genome does not appear to encode an alcaligin biosynthesis locus, indicating that not all Bordetella species possess the ability to produce this siderophore. Th erefore, prior to IUI acquisition, the progenitor of B. holmesii may not have been competent to produce alcaligin. Acquisition of this function by lateral transfer from B. pertussis may have provided B. holmesii with a new, highly effi cient iron uptake system, leading to an immediate enhancement of its ability to colonize a eukaryotic host. Th ese results suggest that the acquisition of B. pertussis DNA, by conferring an increased capacity to scavenge free iron in the host environment, has played a key role in the emergence of B. holmesii and its adaptation to humans.

In chapter 6, evidence is provided for the presence of 15 positively selected codons in the autotransporter protein pertactin (Prn). Prn is a virulence factor, produced by all mammalian bordetellae, which is used in many current acellular vaccines. Th e location of these positively selected codons was compared to the location of previously identiﬁ ed epitopes, as well as to putative conformational epitopes as predicted by in silico analysis. Th e majority of these codons were shown to be located in linear or conformational epitopes (60%), and well exposed to solvent (93%). It is proposed that adaptive changes in these codons may result in immune escape. Possibly, variation in these codons may also reﬂ ect adaptation to host receptors or evasion of recognition by BPP-1. Th ese results could also explain the previously described lack of cross-protection of Abs between B. pertussis and B. parapertussis Prn 22,23.

136 General Discussion

Prn contains two amino acid repeat regions (region 1 and 2, or R1 and R2), which have been implicated in evasion of host immunity 24. A statistically signiﬁ cant inversed correlation was found between the length of R1 and R2. Prn sequences with a long R1 sequence combined with a short R2 sequence were found almost exclusively in the human-associated B. pertussis and B. bronchiseptica complex IV strains, while Prn sequences with short R1 and long R2 sequence lengths were found predominantly in B. bronchiseptica complex I and B. parapertussis. Previously, it was observed that R1 and R2 may be part of a single discontinuous epitope (Hijnen et al., unpublished), implicating close proximity of these regions. It is proposed that sequence length variation in R1 may be compensated for by variation in the length of R2, and vice versa, to maintain the close proximity of the variable epitope, or to maintain masking of underlying epitopes.

In summary, our data provide a model for the evolution of the Bordetella species that cause respiratory infections in mammalians, and we show that the pertussis-causing human- 7 specifi c lineages B. pertussis, B. parapertussishu and B. holmesii have evolved through diff erent means and from diff erent lineages. Th e results presented in this thesis have improved our understanding of how Bordetella species have adapted to the human host, and provide a framework for further studies on host adaptation in general and of the Bordetella species in particular.

137 Chapter 

reference list Functional characterization of the BvgAS two- component system of Bordetella holmesii. Microbiology 150, 3715-3729. 1. Goodnow,R.A. (1980). Biology of Bordetella 11. Major R.H. (1945). Classic Descriptions of Disease. bronchiseptica. Microbiol.Rev. 44, 722-738. Springfi eld, C. C. Thomas. 2. van der Zee,A., Mooi,F., van Embden,J., Musser,J. 12. Musser,J.M., Hewlett,E.L., Peppler,M.S., Selander,R. (1997). Molecular evolution and host adaptation K. (1986). Genetic diversity and relationships in of Bordetella spp.: phylogenetic analysis using populations of Bordetella spp. The Journal of multilocus enzyme electrophoresis and typing Bacteriology 166, 230-237. with three insertion sequences. The Journal of Bacteriology 179, 6609-6617. 13. Parkhill,J., Sebaihia,M., Preston,A., Murphy,L.D., Thomson,N., Harris,D.E., Holden,M.T., Churcher,C. 3. Mattoo,S., Cherry,J.D. (2005). Molecular M., Bentley,S.D., Mungall,K.L., Cerdeno-Tarraga,A. Pathogenesis, Epidemiology, and Clinical M., Temple,L., James,K., Harris,B., Quail,M.A., Manifestations of Respiratory Infections Due Achtman,M., Atkin,R., Baker,S., Basham,D., Bason,N., to Bordetella pertussis and Other Bordetella Cherevach,I., Chillingworth,T., Collins,M., Cronin,A., Subspecies. Clinical Microbiology Reviews 18, 326- Davis,P., Doggett,J., Feltwell,T., Goble,A., Hamlin,N., 382. 7 Hauser,H., Holroyd,S., Jagels,K., Leather,S., 4. Porter,J.F., Parton,R., Wardlaw,A.C. (1991). Growth Moule,S., Norberczak,H., O’Neil,S., Ormond,D., and survival of Bordetella bronchiseptica in Price,C., Rabbinowitsch,E., Rutter,S., Sanders,M., natural waters and in buff ered saline without Saunders,D., Seeger,K., Sharp,S., Simmonds,M., added nutrients. Applied and Environmental Skelton,J., Squares,R., Squares,S., Stevens,K., Microbiology 57, 1202-1206. Unwin,L., Whitehead,S., Barrell,B.G., Maskell,D. J. (2003). Comparative analysis of the genome 5. Porter,J.F., Wardlaw,A.C. (1993). Long-term survival sequences of Bordetella pertussis, Bordetella of Bordetella bronchiseptica in lakewater and in parapertussis and Bordetella bronchiseptica. Nat buff ered saline without added nutrients. FEMS Genet. 35, 32-40. Microbiol.Lett. 110, 33-36. 14. Cummings,C.A., Brinig,M.M., Lepp,P.W., Van De,P. 6. Weyant,R.S., Hollis,D.G., Weaver,R.E., Amin,M. S., Relman,D.A. (2004). Bordetella species are F., Steigerwalt,A.G., O’Connor,S.P., Whitney,A.M., distinguished by patterns of substantial gene loss Daneshvar,M.I., Moss,C.W., Brenner,D.J. (1995). and host adaptation. J Bacteriol. 186, 1484-1492. Bordetella holmesii sp. nov., a new gram-negative species associated with septicemia. Journal of 15. Gupta,S., Maiden,M.C., Feavers,I.M., Nee,S., May,R. Clinical Microbiology 33, 1-7. M., Anderson,R.M. (1996). The maintenance of strain structure in populations of recombining 7. Shepard,C.W., Daneshvar,M.I., Kaiser,R.M., infectious agents. Nat.Med. 2, 437-442. Ashford,D.A., Lonsway,D., Patel,J.B., Morey,R. E., Jordan,J.G., Weyant,R.S., Fischer,M. (2004). 16. Parkhill,J., Wren,B.W., Thomson,N.R., Titball,R.W., Bordetella holmesii bacteremia: a newly recognized Holden,M.T., Prentice,M.B., Sebaihia,M., James,K. clinical entity among asplenic patients. Clin.Infect. D., Churcher,C., Mungall,K.L., Baker,S., Basham,D., Dis. 38, 799-804. Bentley,S.D., Brooks,K., Cerdeno-Tarraga,A.M., Chillingworth,T., Cronin,A., Davies,R.M., Davis,P., 8. Mazengia,E., Silva,E.A., Peppe,J.A., Timperi,R., Dougan,G., Feltwell,T., Hamlin,N., Holroyd,S., George,H. (2000). Recovery of Bordetella holmesii Jagels,K., Karlyshev,A.V., Leather,S., Moule,S., from Patients with Pertussis-Like Symptoms: Use Oyston,P.C., Quail,M., Rutherford,K., Simmonds,M., of Pulsed-Field Gel Electrophoresis To Characterize Skelton,J., Stevens,K., Whitehead,S., Barrell,B.G. Circulating Strains. Journal of Clinical Microbiology (2001). Genome sequence of Yersinia pestis, the 38, 2330-2333. causative agent of plague. Nature 413, 523-527.

9. Loeff elholz,M.J., Thompson,C.J., Long,K.S., 17. Chain,P.S.G., Carniel,E., Larimer,F.W., Lamerdin,J., Gilchrist,M.J.R. (2000). Detection of Bordetella Stoutland,P.O., Regala,W.M., Georgescu,A.M., holmesii Using Bordetella pertussis IS481 PCR Vergez,L.M., Land,M.L., Motin,V.L., Brubaker,R.R., Assay. Journal of Clinical Microbiology 38, 467. Fowler,J., Hinnebusch,J., Marceau,M., Medigue,C., 10. Gerlach,G., Janzen,S., Beier,D., Gross,R. (2004). Simonet,M., Chenal-Francisque,V., Souza,B.,

138 General Discussion

Dacheux,D., Elliott,J.M., Derbise,A., Hauser,L.J., Garcia,E. (2004). Insights into the evolution of Yersinia pestis through whole-genome comparison with Yersinia pseudotuberculosis. Proceedings of the National Academy of Sciences 0404012101.

18. Nierman,W.C., DeShazer,D., Kim,H.S., Tettelin,H., Nelson,K.E., Feldblyum,T., Ulrich,R.L., Ronning,C. M., Brinkac,L.M., Daugherty,S.C., Davidsen,T.D., Deboy,R.T., Dimitrov,G., Dodson,R.J., Durkin,A. S., Gwinn,M.L., Haft,D.H., Khouri,H., Kolonay,J. F., Madupu,R., Mohammoud,Y., Nelson,W.C., Radune,D., Romero,C.M., Sarria,S., Selengut,J., Shamblin,C., Sullivan,S.A., White,O., Yu,Y., Zafar,N., Zhou,L., Fraser,C.M. (2004). Structural fl exibility in the Burkholderia mallei genome. Proceedings of the National Academy of Sciences 0403306101.

19. Beaumont,F.C., Kang,H.Y., Brickman,T.J., Armstrong,S.K. (1998). Identifi cation and 7 characterization of alcR, a gene encoding an AraC- like regulator of alcaligin siderophore biosynthesis and transport in Bordetella pertussis and Bordetella bronchiseptica. The Journal of Bacteriology 180, 862-870.

20. Brickman,T.J., Kang,H.Y., Armstrong,S.K. (2001). Transcriptional activation of Bordetella alcaligin siderophore genes requires the AlcR regulator with alcaligin as inducer. The Journal of Bacteriology 183, 483-489.

21. Pradel,E., Guiso,N., Locht,C. (1998). Identifi cation of AlcR, an AraC-type regulator of alcaligin siderophore synthesis in Bordetella bronchiseptica and Bordetella pertussis. The Journal of Bacteriology 180, 871-880.

22. Khelef,N., Danve,B., Quentin-Millet,M.J., Guiso,N. (1993). Bordetella pertussis and Bordetella parapertussis: two immunologically distinct species. Infect.Immun. 61, 486-490.

23. David,S., van,F.R., Mooi,F.R. (2004). Effi cacies of whole cell and acellular pertussis vaccines against Bordetella parapertussis in a mouse model. Vaccine 22, 1892-1898.

24. King,A.J., Berbers,G., van Oirschot,H.F., Hoogerhout,P., Knipping,K., Mooi,F.R. (2001). Role of the polymorphic region 1 of the Bordetella pertussis protein pertactin in immunity. Microbiology 147, 2885-2895.

139 140

Nederlandse Samenvatting Nederlandse Samenvatting Dutch Summary

Het genus Bordetella bestaat uit een negental aanverwante bacteriesoorten, waarvan de meeste infectieziektes kunnen veroorzaken. Drie soorten veroorzaken frequent luchtweginfecties in verschillende zoogdieren, te weten Bordetella bronchiseptica, Bordetella parapertussis en Bordetella pertussis. Deze drie bacteriesoorten worden samen ook wel de “klassieke” of “zoogdier” bordetellae genoemd. Ze zijn zeer nauw aan elkaar verwant, maar desondanks vertonen ze toch ook opmerkelijke verschillen wat betreft hun gastheerbereik en ziekteverloop. B. bronchiseptica heeft een wijd uiteenlopend gastheerbereik en veroorzaakt voornamelijk chronische infecties die vaak onopgemerkt kunnen blijven. Zo veroorzaakt hij bijvoorbeeld kennelhoest bij honden, maar daarnaast kan B. bronchiseptica ook mensen infecteren. Algemeen wordt echter aangenomen dat deze mensen geïnfecteerd raken als gevolg van een verminderde weerstand. Als enige van de zoogdier bordetellae kan B. bronchiseptica ook buiten de gastheer overleven en zich zelfs daar vermenigvuldigen, bijvoorbeeld in het drinkwater van besmet vee. B. pertussis is de belangrijkste veroorzaker van kinkhoest in de mens, een zeer ernstige A en besmettelijke ziekte van de bovenste luchtwegen. In Nederland wordt vrijwel ieder kind gevaccineerd tegen B. pertussis. Naast B. pertussis kan ook B. parapertussis kinkhoest veroorzaken in mensen, al zijn B. parapertussis infecties in het algemeen minder ernstig. B. parapertussis bestaat eigenlijk uit twee verschillende groepen, waarvan de één exclusief in

mensen voorkomt (B. parapertussishu) en de ander uitsluitend in schapen (B. parapertussisov). Kinkhoest is vooral zeer ernstig in jonge, niet (volledig) gevaccineerde kinderen, al kunnen ook volwassenen ernstige ziektesymptomen vertonen. Geïnfecteerde personen zijn gedurende drie of meer weken besmettelijk. Besmetting vindt plaats via aërosolen die vrijkomen tijdens hevige, repeterende hoestaanvallen. Vrij recent is een nieuwe Bordetella soort beschreven waarvan wordt aangenomen dat deze ook mogelijk kinkhoest kan veroorzaken, Bordetella holmesii. Deze soort werd voor het eerst in 1995 ontdekt, in eerste instantie in het bloed van patiënten met een verzwakt immuunsysteem, maar later ook in de luchtwegen van voorheen gezonde patiënten met kinkhoest-symptomen. Er is tot op heden niet veel bekend over de transmissie en epidemiologie van B. holmesii. Er is veel onduidelijkheid over de origine van kinkhoest. De symptomen van kinkhoest zijn zeer karakteristiek, en voordat er op grote schaal werd gevaccineerd, kreeg vrijwel ieder kind kinkhoest. Daarom is het verrassend dat er vóór de Middeleeuwen geen enkele historische beschrijving is terug te vinden van kinkhoest in de Europese literatuur. Ter illustratie, van andere ziektes met zeer karakteristieke symptomen en een grote mortaliteit onder kinderen, zoals difterie en tetanus, zijn referenties terug gevonden tot in de Oudgriekse literatuur. Door het ontbreken van pré-Middeleeuwse beschrijvingen van kinkhoest en het homogene karakter van B. pertussis wordt algemeen aangenomen dat deze soort zich pas zeer recent heeft aangepast aan de mens.

Verschillende studies hebben aangetoond dat zowel B. pertussis als B. parapertussishu zijn

142 Nederlandse Samenvatting geëvolueerd uit B. bronchiseptica, al is onduidelijk hoe dit precies is gebeurd. Uit een vergelijking van hun genomen blijkt dat zowel de genomen van B. pertussis als van B. parapertussishu aan verval onderhevig zijn. In beide genomen vindt grootschalige inactivatie van genen plaats door het accumuleren van zogenaamde insertie-elementen, dit zijn mobiele DNA-elementen die door het genoom kunnen “springen” en zodoende genen kunnen inactiveren. Dit verval wordt toegeschreven aan de verregaande aanpassing aan de mens, waardoor veel genen niet meer nodig zijn, waaronder bijvoorbeeld genen die belangrijk zijn voor overleving in het milieu. Het is echter onbekend wat de belangrijkste genetische veranderingen zijn geweest die invloed hebben gehad op hun aanpassing aan de mens. De relatie van B. holmesii ten opzichte van de andere zoogdier bordetellae is nog steeds aan discussie onderhevig, en het is onbekend of, en hoe deze soort als menselijke pathogeen is ontstaan.

Het doel van dit onderzoek was om inzicht te verkrijgen in de onderlinge relaties van de Bordetella soorten die luchtweginfecties in zoogdieren veroorzaken, teneinde genetische factoren te identiﬁ ceren die belangrijk zijn geweest voor hun aanpassing aan de mens. Dit A kan belangrijke inzichten geven over hoe bacteriën zich aanpassen aan de mens.

In hoofdstuk 2 worden de evolutionaire relaties tussen de drie zoogdier bordetellae bestudeerd. Voor dit doel is van een groot aantal bacteriestammen de DNA volgordes van verschillende genen bepaald; daarnaast is ook de distributie van verschillende insertie- elementen over deze stammen bepaald. Dit resulteerde in een opdeling van de populatie in vier verschillende groepen (complexen), B. pertussis (complex II), B. parapertussishu (complex III) en twee afzonderlijke B. bronchiseptica complexen, complex I en IV. B. bronchiseptica stammen uit complex IV bleken vooral uit mensen te zijn geïsoleerd en waren nauw verwant aan B. pertussis. Mogelijk vertegenwoordigt B. bronchiseptica complex IV de evolutionaire voorouder van B. pertussis.

Aan de hand van de populatiestructuur zoals bepaald in hoofdstuk 2, is in hoofdstuk 3 een gedetailleerde genoom-analyse gedaan van B. bronchiseptica complex I en IV en B. pertussis. Hierbij werden verschillende belangrijke genetische factoren geïdentiﬁ ceerd die waarschijnlijk belangrijk kunnen zijn geweest bij de aanpassing van B. bronchiseptica complex IV aan de mens. De uitkomst van dit onderzoek suggereerde dat de genomen van de complex IV stammen aan verval onderhevig zijn.

In hoofdstuk 4 wordt verder bestudeerd of de genomen van complex IV inderdaad aan verval onderhevig zijn. De genomen van de twee B. bronchiseptica complexen, complex I en IV, zijn met elkaar vergeleken om genen te identiﬁ ceren die uniek zijn voor complex IV stammen en niet voorkomen in complex I. Doordat deze genen vrijwel niet ontdekt werden,

143 Summary in Dutch

werd bevestigd dat complex IV stammen inderdaad aan genoomverval onderhevig zijn. Om de evolutionaire positie van B. holmesii op te helderen is in hoofdstuk 5 gekeken naar de relatie van B. holmesii ten opzichte van B. bronchiseptica, B. pertussis en B. parapertussis. Hieruit bleek dat B. holmesii, in tegenstelling tot wat eerst werd aangenomen, niet nauw verwant is met B. pertussis. Wel werd er een groot DNA element (genomisch eiland) geïdentiﬁ ceerd dat waarschijnlijk van B. pertussis naar B. holmesii is overgedragen. Dit eiland bevat genen die betrokken zijn bij actieve opname van ijzer onder omstandigheden met lage concentraties vrij ijzer. In mensen (en andere zoogdieren) wordt de hoeveelheid vrij ijzer kunstmatig laag gehouden door ﬁ xatie met bijvoorbeeld hemoglobine, zodat bacteriën zich niet vrij kunnen vermenigvuldigen in de mens. Dit eiland kan bij hebben gedragen aan de aanpassing van B. holmesii aan de mens.

In hoofdstuk 6 is onderzocht welke regio’s van het eiwit pertactine aan selectieve druk bloot staan om te veranderen. Pertactine is een eiwit dat betrokken is bij aanhechting van de bacterie aan epitheelcellen in de luchtwegen, en is een belangrijke component van veel A huidige kinkhoest vaccins. Na vergelijking van een groot aantal pertactine sequenties is met behulp van modellering een aantal regio’s en aminozuren geïdentiﬁ ceerd die belangrijk kunnen zijn voor ontwijking van het immuunsysteem. Daarnaast werd een correlatie gevonden tussen de lengtes van twee variabele regio’s in pertactine. Dit kan invloed hebben op het afschermen van bepaalde delen van het eiwit en zodoende herkenning door het immuunsysteem te voorkomen.

Het onderzoek beschreven in dit proefschrift geeft inzichten in hoe bacteriën zich aanpassen aan de menselijke gastheer in het algemeen, en in het bijzonder hoe de Bordetella soorten die luchtweginfecties in mensen veroorzaken zich genetisch hebben aangepast aan de mens. Tenslotte kunnen de in dit proefschrift afgeleide fylogenetische verwantschappen tussen stammen die een verschillende gastheerspeciﬁ citeit vertonen, gebruikt worden om de processen betrokken bij gastheeradaptatie te onderzoeken.

144

Dankwoord Dankwoord Word of Gratitude

So long and thanks for all the ﬁ sh!!!!

Zoals Arthur Dent in “Th e Hitchhiker’s Guide to the Galaxy” een rondreis door het heelal begon, is mijn AIO-tijd voor mij, wellicht op een iets kleinere schaal, ook een reis geweest, met veel ontdekkingen en verrassingen. Tijdens deze reis heb ik veel dingen bijgeleerd, over mijzelf en ook over het omgaan met moeilijke situaties in je werk en daarbuiten. Er zijn een aantal mensen geweest die tijdens deze reis voor mij een speciale betekenis hebben gehad, en ik wil graag deze gelegenheid benutten om hun persoonlijk bedanken. Ik zal ongetwijfeld ook mensen vergeten, maar ik ben hun ook zeker dankbaar.

Frits, het moet met momenten moeilijk zijn geweest om niet één, maar twee eigenwijze jonge AIO’s onder je hoede te hebben gehad. Ik wil je graag bedanken voor al je hulp en begeleiding, en vooral voor het feit dat je altijd tijd over had voor mij. Hopelijk breekt er nu een rustigere periode aan!

Leo, toen ik na twee jaar weer terugkwam op het RIVM, heb ik erg veel gehad aan onze discussies, en ik heb het erg op prijs gesteld dat ik vaak “zomaar” even binnen kon lopen B om met je te overleggen over het één of het ander. Ik kijk er naar uit om met je samen te gaan werken!

Marjolein en Han, ik heb het ontzettend ﬁ jn gevonden dat jullie de afgelopen twee jaar zoveel “klusjes” hebben opgeknapt voor mij! Ik heb in die tijd veel van jullie geleerd, zowel op persoonlijk vlak als op de labvloer. De rest van de “kinkhoestgroep”, Kees en Audrey, dank jullie voor alles, de WOL’s zullen straks een stuk korter duren denk ik.

De rest van S2: Sanne, Ingrid, Sandra, Corrie, bedankt voor alle mooie momenten op het RIVM. Het zal straks een stuk rustiger worden tijdens de lunchpauzes denk ik! De rest van het LTR, bedankt voor een mooie tijd, ik heb me altijd erg thuis gevoeld op het LTR. Het ga jullie goed!

Betsy, bedankt voor het inwerken in het LPS werk, ik heb onze discussies laat op de avond altijd ontzettend leuk gevonden. Als al onze plannen met het RIVM nu eens werkelijkheid zouden worden..... Ik ga er van uit dat je je volleybalploeg gewoon weer afbelt op 10 februari!

Mijn studenten, Miranda en Zena, wil ik bij deze ook graag bedanken voor hun inzet. Zena, ik wens je veel succes de komende tijd, en ga vooral niet aan jezelf twijfelen, dat is altijd het laatste wat je moet gaan doen!

146 Dankwoord

Mijn kamergenootjes van G04.612. Toen ik nog in mijn eentje de complete kinkhoestgroep vormde, heb ik altijd erg veel steun gehad aan jullie! Suzan en Noortje, die lunchafspraak is er dan toch nog van gekomen! Niki, je Griekse bijles heeft helaas niet veel geholpen, maar hopelijk leer ik straks vanzelf wat meer Grieks bij.

Bent en Pieter-Jan, we hebben samen een geweldige tijd gehad! BBQ-en bij -10°C, tijdens een windhoos, of gewoon bij een hittegolf, BBQ-en kan altijd! Horror Night moeten we zeker een keer overdoen, dan kunnen we gelijk die Soccaroos laten zien hoe het echt moet! Jullie zijn in ieder geval altijd welkom!

Los Passionatos was voor mij altijd heel bijzonder, met zijn 10-en voetballen op 20 vierkante meter, maar wel altijd met passie! Ik heb er van genoten! De rest van het EWI, ik heb mijn tijd bij jullie erg ﬁ jn gevonden!

I would also like to thank all the people in the Relman lab, I’ve really enjoyed my time B in your lab, and you have always made me feel at home. Elies, thanks for taking care of me. I always enjoyed our weekendtrips. I really should email more often! Craig, thanks for introducing me into the world of microarrays, and for always being critical about our manuscripts. Mary, thanks for all the help when this tall Dutch guy came over from Europe to work on Bordetella microarrays. David, thanks for giving me the opportunity to work in your lab, I have really enjoyed our collaboration, and who knows, we may meet again in the future!

Dan kom ik nu bij de mensen die voor mij een extra speciale betekenis hebben (gekregen).

Peter en Tineke (en Ben en Loes), jullie ook bedankt voor alle steun. Nu ga ik dan toch echt “afstuderen”!

Marcel, ik ken je eigenlijk pas vijf jaar, maar het feit dat je mijn getuige was tijdens mijn bruiloft spreekt denk ik voor zich. Jij en Candida zijn in de afgelopen vijf jaar erg goede vrienden van ons geworden. Bedankt voor je steun in alle momenten dat ik het even niet meer zag zitten. Ach, 15 km is niet zo ver, toch?!

Tineke, zoals je pas geleden nog zo mooi zei, ik ken je al mijn hele leven. Onze tijd in Wageningen heeft ons denk ik pas echt dicht bij elkaar gebracht als broer en zus, en een paar

147 Word of Gratitude

jaar aan de andere kant van de wereld gaat dat echt niet veranderen! Je zult altijd mijn grote zus(je) blijven! Bedankt voor alle steun, en we komen elkaar zeker opzoeken!

Pap, mam, zonder jullie steun, belangstelling en liefde zou ik dit alles niet hebben kunnen doen. Nu weten jullie net wat ik nu doe, ga ik weer aan iets anders met een moeilijke naam werken! Ik hoop dat jullie trots op me kunnen blijven in de toekomst. Bedankt voor alles!

Tenslotte...... lieve Cecile, hoe zou ik ooit in woorden kunnen uitdrukken wat jij voor mij betekent en wat voor steun ik aan je heb gehad? Ik weet heel goed wat dit allemaal voor jou heeft betekent, elke keer als ik ‘s avonds laat en in het weekend weer aan mijn proefschrift zat te werken.

Ik wil je ontzettend bedanken voor alles wat je voor me hebt gedaan. Jij bent vanaf het begin mijn “guide” geweest! Zonder jou zou ik verdwaald zijn in die eindeloze ruimte. B

148

Curriculum Vitae Curriculum Vitae Curriculum Vitae

Dimitri Diavatopoulos was born on December 4th, 1976 in Lage Zwaluwe, the Netherlands, where he grew up. He graduated from high school (Nassau College) in Breda in 1995. In the same year he went to study Bioprocess Engineering at Wageningen University. He did a seven month internship at the Department of Virology (Wageningen University), during which he worked on the White Spot Syndrome Virus, under the supervision of Prof. Dr. J.M. Vlak and Dr. M.C. van Hulten. Th is was followed by a second internship of seven months at the Department of Cell Biology and Immunology, during which he studied aspects of crustacaean immunology to the White Spot Syndrome Virus, supervised by Dr. R.J. Stet. After this he did a ﬁ nal internship of six months at the Laboratory for Vaccine-Preventable Diseases (National Institute for Public Health and the Environment, Bilthoven, Th e Netherlands), during which he ﬁ rst studied whooping cough (supervised by Prof. Dr. F.R. Mooi and Dr. I.H. van Loo). After receiving his Master of Science degree in 2000, he commenced his graduate training at the Eijkman-Winkler Institute (University Medical Center Utrecht) and the Laboratory for Vaccine-Preventable Diseases, under the supervision of Prof. Dr. F.R. Mooi and Dr. L.M. Schouls. During that time, he also spent three months at the Relman Laboratory at Stanford University. After his graduation, he hopes to continue his career in Australia to work on pneumococci.

150 Curriculum Vitae

Dimitri Diavatopoulos werd op 4 december 1976 geboren in Lage Zwaluwe. In 1995 behaalde hij het VWO diploma aan het Nassau College te Breda. In hetzelfde jaar ging hij Bioprocestechnologie studeren aan Wageningen Universiteit, waar in 1996 het propedeuse werd behaald. Na een zeven maanden durende afstudeerstage bij de vakgroep Virologie in Wageningen aan het White Spot Syndrome Virus (onder begeleiding van Prof. Dr. J.M. Vlak en Dr. M.C. van Hulten) werd een afstudeerstage van zeven maanden gevolgd bij de vakgroep Celbiologie en Immunologie, waarbij de immunologische aspecten van het White Spot Syndrome Virus werden bestudeerd (begeleid door Dr. R.J. Stet). Hierna werd een stage gevolgd bij Prof. Dr. F.R. Mooi en Dr. I.H. van Loo, waarbij hij kennismaakte met kinkhoest. Na het behalen van het doctoraal in 2000 werd direct begonnen met zijn promotieonderzoek, wat plaats heeft gevonden onder begeleiding van Prof. Dr. F.R. Mooi en Dr. L.M. Schouls. Dit onderzoek vond plaats aan twee instanties, het Laboratorium ter Toetsing van het Rijksvaccinatieprogramma (Rijksinstituut voor Volksgezondheid en het Milieu, Bilthoven) en aan het Eijkman-Winkler Instituut (Universitair Medisch Centrum Utrecht). In het kader van zijn promotieonderzoek werd in 2003 een bezoek van drie maanden gebracht aan het Relman Laboratorium (Stanford University, CA, USA). Hij hoopt zijn loopbaan te vervolgen in Australië om daar te gaan werken aan pneumokokken.

151 152

Appendix Appendix Color Figures Chapter 

Figure 1. Respiratory tract infection BvgA-phosphorylation model of humans by Bordetella bronchiseptica, Bordetella parapertussis and <26°C 35°C 37°C Bordetella pertussis. Bacteria enter the Bvg- Bvgi Bvg+ host through inhalation of aerosols. Adherence takes place in the nasal Flagella BipA BrkA/BrkB Early genes cavity, followed by colonization of the O-antigen TTSS FHA Urease Prn pharynx and the trachea and in rare Fim vrgs TcfA cases the lungs. They subsequently Dnt Late genes pass through diff erent Bvg phases vags Ptx during infection, depending on the CyA level of phosphorylation of BvgA, which is modulated by environmental signals such as temperature. The Bvg- phase (the avirulent phase) is thought to be important for survival Nasal cavity outside the host, and in this phase the virulence-repressed genes (vrgs) are Upper Pharynx respiratory expressed. B. bronchiseptica expresses tract fl agella for motility and urease in the Bordetella bvg- phase. The Bvgi phase is thought Larynx to be important for initial colonization Lower of the nasopharynx, and is character- Trachea respiratory ized by optimal expression of BipA. tract In the Bvg+ phase, the virulence-activated genes are expressed (vags), Primary bronchi including adhesins and toxins. This Lungs phase may be divided in an early and D a late phase. In the early Bvg+ phase, adhesins are expressed but not toxins, and this phase is associated with colonization of the upper respiratory tract. The late Bvg+ phase is characterized by expression of toxins in addition to adhesins, and these toxins may modulate the immune system of the host for survival and may contribute to induction of disease symptoms that are important for transmission. The list of virulence factors is incomplete, and only represents the most studied virulence factors. Other virulence factors, either controlled by BvgAS or not, may be important for transmission as well. Abbreviations: Bvg, Bordetella virulence gene; vrgs, virulence-repressed genes; BipA, Bvg intermediate phase protein A; BrkA/B, Bordetella resistance to killing A/B; TTSS, type III secretion system; Prn, pertactin; TcfA, tracheal colonization factor A; Dnt, dermonecrotic toxin; FHA, fi lamentous haemagglutinin; Fim, fi mbriae; Ptx, pertussis toxin; CyA, adenylate cyclase.

154 Color Figures Chapter  8 18 3 2 22 (100%) (100%) (100%) (80%) 3 1663 1002 481 3 complex II complex B. pertussisB. 1663 IS IS IS IS 1 complex IV complex 25 2 B. bronchiseptica B. 34 35 1 3 9 1 2 STs containing sequenced strains sequenced containing STs RB50 or 12822) (Tohama, human origin animal origin unknown 1 17 31 5 28 1 3 2 24 3 21 15 27 9 5 4 3 hu 30 5 1 29 2 4 (100%) (100%) 19 1001 1002 IS IS complex III complex 4 B. parapertussis B. 7 10 26 10

3 D (40%) 32 16 481 1 5 23 1 IS (100%) complex I complex 33 5 1 1001 B. bronchiseptica B. IS 1 6 (83%) 12 1 1 1001 IS 7 14 , B. B. B. rst. 481 B. bron- B. and is shown in is shown Tohama, Tohama, B. bronchisep- B. 1663 B. pertussisB. , strains. STs containing containing STs strains. and IS B. pertussisB. ov . The tree was based on tree The . 12822 or cation. complex I) harbors the complex 1002 Minimum of spanning tree , IS erent sub-locierent those between RB50) are indicated by a thickset, by indicated RB50) are 1001 sub-loci, cluster- ve and a categorical Figure 1. Figure B. bronchiseptica B. strains of which the genome has been ( sequenced chiseptica STs. The clonal complexes (I, II, III and clonal complexes The STs. IV) strips be- colored by indicated are ST16 ( connectedtween STs. Each circle represents an ST, the size the size an ST, represents Each circle the number of to of which is related that particular to belonging isolates host- indicate within circles Colors ST. numbers between The distribution. the number represent connected STs of diff parapertussis ing was performed. In the minimum types sequence (STs) spanning tree, sharing the highest number of single connected fi locus variants were the sequence of seven housekeeping of seven the sequence split into Individual genes were genes. fi parapertussis tica distribution of the The dashed line. insertion elements IS sequence IS parentheses numbers between boxes; of strains that the percentage indicate by the ISE as determined contained PCR amplifi parapertussis

155 Color Figures Chapter  95% 96 97 98 99 100% B. bronchiseptica Figure 2. UPGMA tree based on the analy- complex IV sis of the pertactin gene of Bordetella iso- 30% animal n=10 70% human lates used in the MLST analysis. The DNA 100 segment coding for the extracellular domain of pertactin (P.69) was used for analy- B. pertussis sis, with the exclusion of the repeat regions complex II 1 and 2. Bootstrap values are shown for n=26 100% human the nodes separating the complexes and are based on 500 bootstrap replicates. The scale indicates the genetic distance along the branches. Colors of the branches indicate the four complexes as defi ned by MLST. The number of strains of each B. bronchiseptica branch is shown in boxes, as well as the complex I host distribution. 65% animal n=71 31% human 4% unknown

100

B. parapertussishu complex III n=9 100% human D

Figure 4. Model B. bronchiseptica of the evolution complex I of the mammalian bordetellae. The bar on the left indicates increasing degrees of adap- broad host range tation to the human host. Arrows LCA complex IV indicate descent. Abbreviations: LCA, last common ancestor.

B. bronchiseptica complex IV

B. pertussis B. parapertussishu

human-adapted complex II complex III

156 Color Figures Chapter 

Complex I Complex IV RB50 Tohama B0227 ST23 B0242 ST7 B2112 ST7 B2490 ST9 B. pertussisB. bronchiseptica B. B0084 ST31 B0189 ST27 B0224 ST6 B0230 ST6 B0238 ST27 B0251 ST23 B0258 ST10 B1975 ST6 B1985 ST6 B1986 ST6 B0232 ST18 B0243 ST28 B0259 ST29 B1968 ST22 B1969 ST18 B2114 ST25 B2491 ST8 B2492 ST21 B2494 ST15 B2495 ST17 B2496 ST3 B2506 ST34 ptxA ptxB ptxB ptxD pertussis ptxE toxin ptxC ptxC ptlA ptlB ptlC ptlD pertussis ptlI toxin ptlE secretion ptlF ptlG ptlH dnt dermonecrotic toxin tcfA (BP) tracheal colonization factor A prn (BP) prn (BB/BPP) pertactin prn (BB/BPP) alcA alcB alcC alcD alcE alcaligin D alcR alcS fauA -40 4

157 Color Figures Chapter 

O-antigen

band A band B

O-antigen

band A

band B

B. 1 2 2 1 1 2 2 2 1 1 2 1 1 2 2 3 1 2 2 2 4 2 2 3 1 1 4 LPS genetic profile pagP pagL (de)acylation lipidA lpxA lpxB lpxC lipidA lpxD lpxH lpxK lpxK waaA inner core waaC wlbA wlbB wlbC trisaccharide wlbD D wlbE wlbF wlbG wlbH wlbI wlbJK wlbL wbmA wbmB wbmC O-antigen wbmD wbmE wbmF wbmG wbmH wbmI wbmJ wbmK wbmL wbmM wbmN wbmO wbmR wbmS BB0124 BB0125 BB0126 BB0127 BB0127 wbmU wbmT alternative wbmS O-antigen wbmR wbmQ wbmP -40 4

158 Color Figures Chapter 

Figure 3. Model of B. bronchiseptica the evolution of the complex I mammalian bordetellae. The bar on the left indicates increasing degrees of adaptation to the human broad host range LCA complex IV host. Arrows indi- Acquired cate descent; double BP0072 transposase D arrows between IS1663 16 lineage-specific genes complexes indicate Genome decay possible within- host immune-competition. In boxes, genetic events are B. bronchiseptica shown that may complex IV have played a role in Deleted speciation and niche Ptx adaptation. See text Dnt for details. Abbrevia- LPS polymorphism tions: LCA, last com- Genome decay mon ancestor. B. pertussis B. parapertussishu complex II complex III

Ptx expression Immune No Ptx expression Type III secretion off competition Type III secretion off Deleted or inactivated Deleted or inactivated O-antigen locus Type II capsule Autotransporters Type II capsule Acquired IS1002 Acquired IS481, IS1002 Genome decay

human-adapted Genome decay

159 Color Figures Chapter 

A. B. log (B. holmesii intensity/reference intensity) 2 B holmesii -6 -5 -4 -3 -2 -1 0 1 BP0001 B0436 B1855 B2738 B1851 B1853 B1854 B1852 B2768 BP Tohama B0437 B2739 B1850 B2767 BP0783 BP0787

BP0794 BP0798 BP2447 BP2450

alcABCDERS fauA mar intergenic region BP2465 vrg-6

BP2475 BP2476

kdpABCDE

BP2499 D BP2502 IS481 IS1001 insertion IS1002 IS1663 elements BB2492

-30 3

Figure 1A. CGH of 12 B. holmesii isolates to a microarray comprising the genomes of B. pertussis Tohama, B. para-

designations are shown for a selection of probes. The relative hybridization value (log2(Cy5/Cy3)) is indicated by the yellow-black-blue color scale. Missing data are represented in grey.

160

Color Figures Chapter 

BP2499 BP2499 isolates 197N. binding site 5 kb 70

Fur binding site Fur σ

B. holmesii B. BP2465 bhoABCDE B. avium B. Representation of Representation A. .

Orthologs in all species not found Orf by replaced Conserved IUI-flanking ORFs in present Variably BP2492 FecIR * Tohama and Tohama

IUI. Arrows above ORFs indicate ORFs indicate above IUI. Arrows was derived from the published from was derived BP2485 B. bronchiseptica B. B. pertussisB. hologous sequences. The ORF composi- The hologous sequences. bhoB bhoC bhoD bhoE insertion and B. holmesii B. elements RB50, orfA 481 Alcaligin operon Alcaligin holmesii B. IS BP0787/2450 orthologs Orthologs in all species found BhoA B. bronchiseptica bronchiseptica B. orfB

B. pertussisB.

and ,

marC BP2477 *

B. bronchiseptica bronchiseptica B. BP2477 B. holmesii B. B. pertussisB. , , B. avium B. B. avium B. Fe-alcaligin receptor Fe-alcaligin

regulation D

+ alcaligin export+ alcaligin

alcR alcS fauA alcABCDE

BP2453 BP2453

and comparison to orthologous to in the genomes of and comparison regions BP2450

alcA BP0787/2450

alcaligin biosynthesis alcaligin

B. holmesii B.

BP0794 ORFs in the 4.8 kb locus and putative of the alcaligin DNA insertion Detailed organization detected in BP0794 alcA alcB alcC alcD alcE B.

IUI was deduced from PCR and CGH data, while the ORF organization of data, while the ORF organization PCR and CGH from IUI was deduced

BP2455 BP0787 island (IUI) in of the iron-uptake of the genomic organization analysis Comparative B. holmesii B. the putative function the putative of these ORFs. Figure 2. Figure genome sequences. genome sequences. the ORF organization of the IUI in the ORF organization surfaces. of the ort Dashed lines connect grey Deletions or insertions the ORFs at borders by indicated species are between tion of B. A. B. avium B. holmesii B. pertussisB. bronchiseptica B.

161 Color Figures Chapter  677-Gly C' y selected codons. Yellow Yellow y selected codons. cated by black triangles. AA- black triangles. by cated the connecting lines. Numbering the connecting lines. reen triangles indicate loops that after triangles indicate reen Not characterized by by Not characterized crystallography X-ray 498 488 438 398 376 358 350 334 330 D Region C Region D 270 RGD R1 R2 ect in binding with mAbs. or increase 232-Gly to 256-Pro232-Gly to 566-Pro to 545-Pro 213 187 rst amino acid of the mature protein. Green boxes indicate the location of conformational epitopes co-localized with positivel epitopes the location of conformational indicate boxes Green protein. rst amino acid of the mature Region A Region B 21, 22, 24 22, 21, 123 4 56 910 13 14 1517 18 1920 Location of positively selected codons and regions on the primary structure of Tohama Prn. Positively selected codons are indi selected are codons Positively Prn. Tohama on the primary selected and regions codons of positively structure of Location N' boxes indicate the co-localization of both conformational and linear epitopes with positively selected codons. Red, white and g white Red, selected codons. with positively and linear epitopes the co-localization indicate of both conformational boxes no eff a decrease, respectively SDM showed mutation by Figure 2. Figure residues with a maximal distance of 6Å and a minimum solvent accessibility of 25% have been designated as regions, indicated by indicated as regions, been designated of 25% have accessibility of 6Å and a minimum solvent with a maximal distance residues starts with the fi 1-Asp

162 Color Figures Chapter  Gly ⊂ +488-Ala Arg ⊂ +498-Leu +498-Leu Cys codons not present in a region are colored colored are in a region not present codons ⊂ +438-Val +438-Val Ala ⊂ Prn1 (1DAB.pdb). Numbers indicate the positively selected the positively Numbers indicate (1DAB.pdb). Prn1 Gln ⊂ Ser +398-Ser ⊂ Gly ⊂ B. pertussisB. Region D +358-Asp +358-Asp +376-Ile Gln ⊂ +334-His Region C +350-Lys +350-Lys Arg ⊂ D crystal structure of

+330-Gln Thr ⊂ +270-Ser Region B Phe Asp ⊂ ⊂ +187-Ser +213-Gln Gly Ser Silent Projection of the positively selected codons and regions on the selected and regions codons of the positively Projection ⊂ ⊂ ⊂ Region A +21-Gly +22-Ser +24-Pro Figure 3. Figure codons. Codons that are part of a region are colored in red, and regions are indicated by black rectangles. Positively selected Positively black rectangles. by indicated are and regions in red, colored part are that are Codons of a region codons. in blue.

163 Notes

164 Notes

165 Notes

166 Notes

167 Notes

168