This work is dedicated to my dearest and devoted friend, Sandra Rodin, for being my constant inspiration and support

And to my dear Mom, whose encouragement and belief in me made this possible

In loving memory of my morfar, Sven Henriksson, who I can only imagine is crying tears of joy in heaven right now. Jag älskar dig, Poppop!

“Vi kan inte göra stora saker, bara små saker med stor kärlek” Mother Teresa

“Being a graduate student is like becoming all of the Seven Dwarves. In the beginning you’re Dopey and Bashful. In the middle you are usually Sneezy, Sleepy, and Grumpy. But at the end, they call you Doc, and then you’re Happy.” Unknown

LIST OF PAPERS

This thesis is based on the following papers, which are referred to in the text by their Roman numerals:

I. McGill, S., R. Regnery, and K. Karem. (1998) Characterization of human immunoglobulin (Ig) isotype and IgG subclass response to Bartonella henselae infection. Infect Immun, 66(12):5915-20.

II. Holmberg, M., S. McGill, C. Ehrenborg, L. Wesslén, E. Hjelm, J. Darelid, L. Blad, L. Engstrand, R. Regnery, and G. Friman. (1999) Evaluation of human seroreactivity to Bartonella species in Sweden. J Clin Microbiol, 37(5):1381-4.

III. McGill, S., L. Wesslén, E. Hjelm, M. Holmberg, C. Rolf, and G. Friman. (2001) Serological and epidemiological analysis of the prevalence of Bartonella spp. antibodies in Swedish elite orienteers 1992-93. Scand J Infect Dis, 33(6):423-8.

IV. Hjelm, E., S. McGill, and G. Blomqvist. (2002) Prevalence of antibodies to Bartonella henselae, B. elizabethae, and B. quintana in Swedish domestic cats. Scand J Infect Dis, 34(3):192-6.

V. McGill, S., J. Rajs, E. Hjelm, O. Lindquist, and G. Friman. (2003) A study on forensic samples of Bartonella spp. antibodies in Swedish intravenous heroin addicts. APMIS, 111:507-13.

VI. McGill, S., L. Wesslén, E. Hjelm, M. Holmberg, M. Auvinen, K. Berggren, B. Grandin-Jarl, U. Johnson, S. Wikström, and G. Friman. (2005) Bartonella spp. seroprevalence in healthy Swedish blood donors. Scand J Infect Dis, 37(10):723-30.

Reprints were made with permission from the publishers.

CONTENTS

INTRODUCTION ...... 11 Historical overview ...... 11 Bacteriology and epidemiology...... 13 Bartonella spp. genomes ...... 15 Bartonellae and their spectrum of clinical manifestations...... 16 B. bacilliformis ...... 17 B. quintana...... 17 B. henselae...... 18 CSD...... 18 B. elizabethae...... 20 Other Bartonellae ...... 20 Laboratory diagnosis of Bartonella infection ...... 22 Serological cross-reactivity ...... 24 Bartonella seroprevalence studies...... 27 Seroprevalence in various human subpopulations...... 27 Seroprevalence in cats ...... 29 Bartonella in Sweden ...... 31 Swedish orienteering investigation...... 31 Human and animal studies in Sweden ...... 32 AIMS OF THE PRESENT STUDY ...... 36 MATERIALS AND METHODS...... 37 Cultivation of Bartonella spp. antigens (I-VI) ...... 37 Study samples (I-VI) ...... 37 Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS- PAGE) (I) ...... 38 Western blot analysis (I) ...... 38 IFA (II-VI)...... 39 PCR and DNA sequencing (II)...... 39 Post-mortem and forensic analysis (II, V)...... 40 Collection of epidemiological data (III,VI)...... 40 Statistical analysis (III-VI) ...... 40 RESULTS AND DISCUSSION...... 42 Analysis of human humoral immune response to Bartonella spp. infection by Western blot (I) ...... 42 Total immunoglobulin (Ig)G reactivity to B. henselae whole-cell and Vero cell co-cultivated antigen (I) ...... 42 IgG-specific fragment and subclass reactivity to B. henselae antigen.44 Immunoglobulin isotype reactivity to B. henselae antigen...... 45 Seroreactivity to Bartonella spp. among various cohorts of the Swedish human and animal population (II-VI) ...... 46 Blood donors (II, III, VI) ...... 46 Patients (II) ...... 48 Elite orienteers (III) ...... 49 Intravenous drug users (IVDU) (V)...... 51 Cats (IV) ...... 53 Demographic and epidemiologic findings (III, IV, V, VI)...... 55 Future prospects and the significance of high B. elizabethae seroprevalence...... 56 GENERAL CONCLUSIONS...... 59 ACKNOWLEDGEMENTS...... 60 REFERENCES ...... 62

ABBREVIATIONS

ARVC Arrhythmogenic right ventricular cardiomyopathy BA Bacillary angiomatosis Bc Bartonella clarridgeiae Be Bartonella elizabethae Bg Bartonella grahamii Bh, Bh(H-1) Bartonella henselae (Houston-1) Bh(M) Bartonella henselae (Marseille) BMI Body mass index Bq Bartonella quintana Bv Bartonella vinsonii cANCA Anti-neutrophil cytoplasmic antibody test CDC Centers for Disease Control and Prevention CI Confidence interval CSD Cat scratch disease CT Computerized tomography; Cutoff titer DNA Deoxyribonucleic acid EIA Enzyme immunoassay (synonymous with ELISA) ELISA Enzyme-linked immunosorbent assay (synonymous with EIA) FITC Fluorescein isothiocynate GEI Genomic island GMT Geometric mean titer HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HIV Human immunodeficiency virus IFA Indirect fluorescent-antibody/immunofluorescence assay Ig Immunoglobulin IVD Intravenous drug kDa Kilodalton M Multivariate logistic regression OR Odds ratio PBS/T Phosphate-buffered saline/ plus Tween-20 POS Parinaud’s oculoglandular syndrome PCR Polymerase chain reaction SDS-PAGE Sodium dodecyl sulfate-polyacrylamide-gel electrophoresis SUCD Sudden unexpected cardiac death Spp/Sp Species, plural/ Species, singular U Univariate logistic regression

INTRODUCTION

Historical overview The genus Bartonella has undoubtedly achieved a level of undisputed clinical importance as both an emerging and re-emerging human pathogen in a relatively short expanse of time. Prior to 1993, the genus Bartonella was composed of a single species, B. bacilliformis, the causative agent of Oroya fever, or Carrion’s disease, first described in South America in the late 1800’s. Subsequent to its isolation and identification in 1909, B. bacilliformis remained the sole member of the genus Bartonella for nearly a century. As a consequence of various factors, highlighted by improved detection and diagnostic capabilities, as well as reclassification of several bacterial genera on the basis of phylogenetic relatedness using 16S rRNA gene sequences, the genus has undergone a substantial expansion. In 1993, the genus was initially broadened through unification with the genus Rochalimaea (39), which was comprised of the species R. quintana (formerly Rickettsia quintana), R. henselae, R. vinsonii, and R. elizabethae, and again in 1995 through a subsequent unification with the genus Grahamella (29), which contained the species G. grahamii, G. taylorii, G. talpae, and G. doshiae. Following these amalgamations, all of the aforementioned species were reclassified as Bartonella and the genus was removed from the order Rickettsiales (29). The genus Bartonella has continued to grow in recent years with the isolation and characterization of numerous additional species, including multiple strains recovered from a broad range of rodents and other wild and domestic mammals in various regions of the world (10, 11, 27, 28, 31, 36, 38, 47, 51, 73, 108, 120, 124, 125, 131, 137, 143, 153, 165, 166, 168, 183, 191, 214, 220, 241, 248). Currently, the genus has been broadened to include over 20 named species, more than half of which are recognized as human pathogens: B. bacilliformis, B. quintana, B. henselae, B. elizabethae, B. clarridgeiae, B. koehlerae, B. vinsonii subspecies berkhoffii and arupensis, B. rochalimaea, B. washoensis, B. grahamii, B. tamiae, B. doshiae, and B. alsatica. See Table 1 for a description of the clinical manifestations in humans and animals and the vectors and reservoirs associated with each of the currently recognized species of Bartonella.

11 TABLE 1. Currently recognized Bartonella species and their associated clinical diseases in humans and animals, arthropod vectors and mammalian reservoirs Species Clinical disease Vector Reservoir B. alsatica Endocarditis, Unknown Wild rabbits lymphadenitis B. australis Unknown Kangaroo B. bacilliformis Carrion’s disease Sand fly Humans (Bartonellosis) [Oroya fever, verruga peruana], internal vascular proliferative lesions B. quintana Trench fever, “urban Human body Humans trench fever,” bacillary louse angiomatosis, fever and bacteremia, endocarditis (humans and dogs), lytic bone lesions, CNS manifestations B. henselae Cat scratch disease, Cat flea, tick? Domestic and wild (Houston-1) bacillary angiomatosis, fly? cats bacillary peliosis, peliosis hepatis (humans and dogs), granulomatous hepatitis (dogs), fever and bacteremia, osteomyelitis, endocarditis, encephalopathy, CNS manifestations, uveitis (humans, cats?) B. henselae Cat scratch disease Cat flea? Domestic cats (Marseille) B. elizabethae Endocarditis Flea? Rodents B. grahamii Neuroretinitis Unknown Bank voles, rodents B. taylorii Flea? Field mice, voles B. doshiae Cat scratch disease Flea? Field voles B. vinsonii subsp. Flea? Field voles vinsonii B. vinsonii subsp. Fever and bacteremia, Tick? Mice arupensis endocarditis, CNS manifestations B. vinsonii subsp. Endocarditis (humans and Tick?, flea? Dogs, wild canids berkhoffii dogs), neurologic disease, myocarditis (dogs), granulomatous lymphadenitis (dogs), uveitis (dogs) B. clarridgeiae Cat scratch disease, Unknown Domestic cats, dogs endocarditis (humans and dogs), lymphocytic

12 hepatitis (dogs) B. koehlerae Cat scratch disease, Flea? Domestic cats endocarditis B. talpae Flea? Shrew moles, rodents B. peromysci Flea? Deer mice

B. tribocorum Unknown Rattus norvegicus B. weissii (/bovis) Tick? Domestic cats, ruminants? B. birtlesii Unknown Field mice B. schoenbuchensis Fly? Roe deer B. tamiae Fever Mite? B. washoensis Fever, myocarditis, Unknown Ground squirrels endocarditis (dogs) B. rochalimaea Fever and bacteremia, Flea? Dogs, foxes, splenomegaly, cutaneous raccoons, coyotes lesions B. rattimasiliensis Unknown Unknown Candidatus B. Associated with Sheep ked? Sheep melophagi pericarditis and dermotologic lesions

Bacteriology and epidemiology Bacteria within the genus Bartonella are classified as Gram-negative, aerobic, pleiomorphic bacilli belonging to the alpha-2 subgroup of the phylum Proteobacteria. Organisms are urease, oxidase, and catalase negative and are nonreactive in carbohydrate utilization tests (252). Further, as members of the genus are characteristically slow-growing and fastidious, axenic culture of the bacteria is typically rendered difficult to achieve. Notably, Bartonella species characteristically require highly specific conditions for optimal growth, including a CO2-rich environment, erythrocyte-enriched blood agar growth media, and extended incubation periods which typically exceed those of the majority of other bacteria (40). Primary colony formation can take up to 5 weeks to become evident; however with successive culture passaging, the incubation time required for colony generation has been found to decrease significantly to as few as 3-4 days (252). Multiple successive passaging also demonstrates alterations in the morphology of the bacteria, from rough, “cauliflower-like” autoaggregates deeply embedded in the medium initially, to smooth and less adherent cells after subculturing (252). This exhibited phase variation, involving alterations in the expression of immunodominant outer-membrane proteins, may potentially play a role in the pathogenesis of Bartonella spp., enabling successful evasion of the host immune response and subsequent

13 colonization or persistence of the pathogens in host cells (310). Polar flagellae, expressed exclusively by B. bacilliformis, B. clarridgeiae, B. rochalimaea and B. schoenbuchensis, have also been shown to play a role in the virulence and invasiveness of the organisms, mediating motility and contributing to deformations in erythrocyte membranes in the case of B. bacilliformis in particular (19, 267). To date, no other Bartonella spp. have been found to possess flagellae. Bartonella organisms have been shown to be susceptible to various antibiotics in vitro, with the highest susceptibility to ketolides such as telithromycin, macrolides, particularly clarithromycin, doxycyline and rifampin, as well as gentamicin, tetracycline, minocycline, azithromycin, clarithromycin, amoxicillin, cefotaxime, and deftriaxone (16, 77, 204). However, due to the intracellular propensity of the organism, as Bartonella spp. are facultative intracellular pathogens, clinical use of these antimicrobial agents in vivo has proven less efficacious, with frequent cases of relapse and treatment failure (300). Further, the hemotropic nature of Bartonella organisms and their ability to persist in the erythrocytes of their hosts contribute to their infectivity via transmission by blood-sucking arthropods. In fact, Bartonella is the only genus of bacteria capable of parasitizing and persisting in mammalian reservoir host erythrocytes and endothelial cells and inducing pathological angiogenesis and vasoproliferative growth in the vascular bed (219). Transmission via arthropod vectors and the unique pathogen-host cell interaction of colonization in mammalian host reservoirs within endothelial and red blood cells resulting in long-lasting hemotropic infection are commonalities shared among Bartonella species. Various hematophagous arthropods have been implicated in Bartonella spp. transmission, primarily fleas, ticks, flies, and human body lice (See Table 1). Transmission from an animal reservoir to human may also occur directly by way of skin trauma (i.e. scratch or bite), epitomized by the transmission of B. henselae by a cat scratch, resulting in cat scratch disease (CSD, discussed below). Bartonella spp. have been isolated from an extensive range of animal species, most commonly from rodents, ruminants and carnivores, yet also from ungulates, lagomorphs, insectivores, and bats (288). Notably, each single Bartonella species is typically capable of infecting only a limited range of reservoir mammalian hosts, exemplifying host specificity characteristic of the genus. To this end, humans are the only known reservoir hosts for both B. bacilliformis and B. quintana; likewise, cats have been shown to serve as the only known mammalian reservoirs for B. henselae and B. koehlerae. The underlying factors contributing to the predilection for a given species of Bartonella to exhibit preferential infectivity in specific host species are not sufficiently understood, yet have become further elucidated through recent comparative genomic analyses of several pathogenic Bartonella species (discussed below). Typically, Bartonella spp. induce prolonged, persistent, and asymptomatic bacteremia in their respective mammalian reservoirs,

14 although overt clinical manifestations rarely occur in such natural hosts; dogs, in particular, serving as natural reservoir hosts for Bartonella spp. have been shown to occasionally exhibit clinical manifestations mimicking the severity of clinical entities observed in incidental human hosts (38). Interestingly, induction of disease in incidental hosts may occur in the absence of intra-erythrocytic invasion, as evidenced by the intraerythrocytic presence of B. quintana in humans (natural host) in contrast to the absence of erythrocytic infection of B. henselae in humans (incidental host). Given the multitude of domestic and wild animal reservoirs and arthropod vectors in addition to relatively high seroprevalence of antibodies to Bartonella spp. among various populations of humans and animals (discussed in further detail below), exposure to Bartonella spp. is likely more extensive than is commonly acknowledged (38).

Bartonella spp. genomes Recent advancements in the genomic analyses of various Bartonella spp. have begun to elucidate the genetic factors contributing to the pathogenicity and host specificity characteristic of members of the genus Bartonella. To date, the complete genome sequences of 4 Bartonella spp. have been deciphered: feline-specific B. henselae (1.9Mb) (3), human-specific B. quintana (1.6Mb) (3), rat-specific B. tribocorum (2.6Mb) (262), and rodent- specific B. grahamii (2.3Mb) (22). Additionally, the gene content of the feline-specific B. koehlerae has been estimated (1.7-1.8Mb) using comparative genomic hybridization with the known B. quintana genome (184), and the genome of human-specific B. bacilliformis (1.4Mb) has been completely sequenced, though has not been reported in detail to date (unpublished data). With the exception of B. grahamii, which contains an additional 28 Kb plasmid, all currently known Bartonella genomes consist of single, circular chromosomes with backbones that are colinear and homologous to a large extent. A core genome, consisting of over 900 genes, is common to all species sequenced to date. However, variations in the genomic structure of each species outside of this core appear to relate to differences in the ecology and host specificity of the organisms (93). For example, B. quintana exhibits a very narrow and strict vector-host ecological relationship, utilizing only the human as a reservoir host and body louse as a vector, whereas B. henselae is more diverse, thriving in both cats and humans and utilizing vectors of a broader host potential. Genome comparisons of the two pathogens reveal a large degree of homology, yet distinct differences include the deletion in B. quintana of a prophage region and 3 genomic islands (GEIs) coding for filamentous hemagglutinin that are present in B. henselae, and a generalized genome reduction in B. quintana (3). The vector-host specialization characteristic of B. quintana has been

15 associated with a degradation of extrachromosomal DNA and phage genes leading to inflexibility in the accommodation of change, notably in the expansion of environmental niches (3). Likewise, sequences in the prophage and GEI regions present in B. henselae are partially absent in the genome of B. koehlerae, also considered a specialist with the cat as a limited host, as well. The partial retention of these regions, however, has led to the hypothesis that B. koehlerae is a genomic intermediate of B. henselae and B. quintana (184). In contrast to the genomic reductions apparent in B. quintana and B. koehlerae, the genomic sequence of B. tribocorum reveals a significant genome expansion, resulting from gene duplication and accessory gene regions, including host-adaptability genes encoding type IV (VirB, Trw) and Type V (adhesins, autotransporters) secretion systems, which appear to play limited roles in the host adaptation and specificity of not only B. tribocorum, but B. quintana, B. henselae, and B. grahamii as well (22, 262). The genome sequences of B. tribocorum and B. grahamii, both rodent-specific pathogens, represent the two largest genomes to date and contain the greatest abundance of genes for putative host-cell interaction systems (22). B. grahamii, for example, contains 16 highly variable GEIs containing many gene clusters, several of which are also present in B. henselae GEIs to a lesser extent, encoding secretion systems (22). The impact of genome sizes and interspecies heterogeneity on pathogenicity and inclination for host specificity will likely be further defined through continued genome sequencing of additional Bartonella spp.

Bartonellae and their spectrum of clinical manifestations The spectrum of human disease and pathologic syndromes associated with these pathogens is remarkably diverse (see Table 2) and has continued to broaden since the identification of the first member of the genus Bartonella in the early 1900’s. Bartonellae are considered emerging infectious agents; the array of clinical syndromes found to be associated with Bartonella spp. is broadening at nearly an equivalent pace as the number of species within the genus. Pathologies attributable to Bartonella spp. range from vasoproliferative lesions to granulomatous inflammation and may have a relatively benign and self-limited evolution or potentially even a severe, life- threatening one. Largely attributable to recent advancements in molecular biology techniques and increased accessibility thereof, in addition to redefined phylogenetic classification systems, Bartonella spp. have been implicated as the zoonotic causative agents of a wide range of clinical pathologies in both immunocompetent and immunocompromised patients, with manifestations including culture-negative endocarditis, CSD, chronic

16 bacteremia, bacillary angiomatosis (BA), chronic lymphadenopathy, bacillary peliosis hepatitis, neuroretinitis, meningoencephalitis, osteomyelitis, granulomatous hepatitis, and recurrent fever. The impact of Bartonella infections has thus become profound.

B. bacilliformis The sole member of the genus Bartonella for decades, B. bacilliformis is the causative agent of Carrion’s disease, or classical bartonellosis, which was responsible for the historic outbreaks of Oroya fever in the late 19th century in the Andes Mountains and is still endemic to regions of Peru, Ecuador, and Columbia (50, 190, 263). The biphasic nature of this disease demonstrates the two pathogenic features unique to the genus Bartonella; the first phase, Oroya fever, involves invasion of human erythrocytes and results in an often fatal acute hemolytic anemia; the second phase, verruga peruana, involves infection and prominent proliferation of endothelial cells, resulting in characteristic skin eruptions of papules and nodules. The disease was named in 1885 after a medical student, Daniel Carrión, who inoculated himself with the agent to prove that Oroya fever and verruga peruana were manifestations of the same infection; he subsequently died of the illness (109). Bartonella bacilliformis is transmitted by the sandfly (Lutzomya verrucarum). In endemic regions, 45-77% of the human population may harbor antibodies to the organism (50, 164), 10-15% of which may be asymptomatic carriers, serving as the disease reservoirs (289).

B. quintana B. quintana was also associated with a historic endemic, trench fever, which afflicted over one million soldiers in World War I and a smaller number in World War II. Trench fever is characterized by a recurrent, cycling fever. Historically recognized as the etiologic agent of trench fever, B. quintana has also recently re-emerged as a cause of chronic bacteremia and endocarditis, termed “urban trench fever,” primarily among alcoholics and homeless men (42, 44, 78, 79, 88, 247, 278), as well as BA and bacillary peliosis in immunocompromised individuals (20, 111, 159). BA is a vasoproliferative disorder characterized by cutaneous manifestations of vascular tumors of the skin and subcutaneous tissues, including single or multiple discolored cutaneous nodules, hyperpigmented lichenoid plaques, and subcutaneous nodules with the potential to ulcerate (65, 152, 157). The lesions are often indistinguishable from those observed in the Oroya fever stage of Carrion’s disease (228). Bacillary peliosis, often occurring in conjunction with BA, is also characterized by angiogenesis and involves dissemination of infection to the reticuloendothelial organs, the most common of which is the liver (peliosis hepatis) and spleen. Bartonella

17 quintana is transmitted by the human body louse (Pediculus humanus corporis) and, like B. bacilliformis, utilizes the human as its sole natural reservoir.

B. henselae First identified in 1990 as an etiologic agent of recurrent fever, BA and its parenchymal variant among a cohort of human immunodeficiency virus (HIV)-infected individuals (252, 256, 274, 293), B. henselae (Houston-1 strain) is regarded as one of the most highly pathogenic species among all of the Bartonellae. The domestic cat (Felis domesticus), asymptomatic during the course of bacteremia, serves as the primary vertebrate reservoir for human B. henselae infection (158, 161), though limited reports of CSD associated with dog contact have been forthcoming in recent years (147, 194, 285). Nearly 50% of cats in some regions, domestic and feral combined, have been shown to harbor B. henselae at a given time (203). Experimental studies have substantiated the cat flea (Ctenocephalides felis) to be a vector responsible for feline horizontal transmission of the organism (56, 307); rarely, arthropod vectors such as flies and ticks may additionally be implicated in transmission of the disease to humans (60, 307). Incidences of CSD, in particular, have been found to directly correlate with increasing climate temperature gradients, which directly correlate as well with increasing numbers of potential arthropod vectors, such as fleas (138). The extensive spectrum of granulomatous and vasculoproliferative diseases stemming from this emerging pathogen, manifested in both immunocompetent and immunocompromised patients, exemplify the diversity of histopathologic lesions for which the organisms can be attributed. Implicated in the etiology of BA, bacillary peliosis, endocarditis, fever of unknown origin, and persistent bacteremia with recurrent fever (132, 140, 247, 254, 293), B. henselae is perhaps most notably recognized for its role as the etiologic agent of CSD.

CSD Cat scratch disease (CSD), a common infection afflicting primarily children, has a worldwide distribution and is the most common zoonotic human infection caused by Bartonella spp. Documentation of the first report of symptoms consistent with CSD can be traced back to the late 19th century in association with the description of Parinaud’s oculoglandular syndrome (POS) and its link to animal contact in 1889 (235). A direct association between cats and the disease was subsequently described by Debre et al. (72) six decades later. Having remained elusive for decades, the definitive etiologic agent of CSD was only in recent years identified as Rochalimae henselae, now classified as B. henselae, by Regnery et al. in 1992 (252). Though B. henselae Houston-1 strain constitutes the primary etiologic agent

18 of CSD, B. henselae Marseille strain, B. clarridgeiae, B. koehlerae, and B. doshiae have also been identified as agents of the disease in limited cases (162, 176, 196, 300). Initially, CSD presents clinically as an erythematous and nonpruritic papule on the skin at the site of inoculation, followed by tender regional, ipsilateral lymphadenopathy proximal to the inoculation site, typically affecting the axillary, cervical, submandibular, inguinal, pre- auricular, and femoral lymph nodes (45, 193, 197). Other less prevalent symptoms include malaise, fever, anorexia/emesis, headache, splenomegaly, pharyngitis, exanthema, conjunctivitis, arthralgia, and dermatologic manifestations, such as macular or papulovesicular rash, urticaria, erythema annulare, and erythema nodosum (144, 193, 197). Interestingly, despite its designation as “cat scratch fever,” only approximately 10-30% of CSD patients actually develop a low-grade fever as a manifestation of infection (45, 197). Typically, the course of CSD is benign and self-limited, with the majority of cases resolving spontaneously after a course of two to eight weeks in immunocompetent patients. Severe immunosuppression, however, particularly secondary to HIV and its advanced stage of acquired immunodeficiency syndrome (AIDS), as well as organ transplantation, is associated with disseminated, often tissue-invasive forms of CSD which have the potential to be life-threatening (32). Manifestations in such patients include BA, bacillary peliosis, and persistent or relapsing fever in association with bacteremia. In the immunocompromised patient with severe manifestations and complications of CSD, a prolonged regimen of antibiotics is recommended (76, 300). Atypical sequelae of CSD manifest in 5-25% of immunocompetent patients, with ocular involvement representing the most prevalent presentation in such cases (144, 300). Documented as the first description of CSD decades ago, POS is in fact the most common form of atypical infection in immunocompetent individuals. Occurring in approximately 6% of all CSD cases (118), POS is a form of conjunctivitis characterized by a granulomatous nodular lesion of the palpebral conjunctiva at the site of inoculation with associated ipsilateral preauricular lymphadenopathy. Neuroretinitis, also an optical manifestation of atypical CSD, manifests as painless acute onset of unilateral visual field loss in association with optic nerve edema and stellate exudative macular lesions (251). Symptoms of POS and neuroretinitis typically resolve with full recovery of vision within several months with or without antiobiotic treatment (187, 195). Atypical CSD sequelae can also manifest as neurological involvement, accounting for 1-7% of CSD cases (8, 217); encephalopathy is the most common such manifestation and presents as acute onset of severe headache, confusion, seizure, coma, or other focal neurologic deficit weeks following the initial presentation of lymphadenopathy (269). Symptoms, however, are typically self-limiting and resolve without permanent neurological sequelae in the absence of antibiotic treatment (300).

19 B. elizabethae Regarded as a human pathogen among the Bartonellae, B. elizabethae has been isolated from a human only once, in 1992, from the blood of an immunocompetent patient with endocarditis (70). Since that time, no other reports of B. elizabethae isolation from humans have been documented, with the exception of a recent study documenting the isolation of two genotypes of Bartonella with a 97.9% and 95.4% rate of similarity to B. elizabethae from febrile patients in Thailand (170). Interestingly, these patients were among a cohort of which 71% and 36% reported rat exposure and dog or cat ownership, respectively (170). Various recent studies primarily during the past decade have obtained isolates either identical or closely related to the organism from non-human vertebrates, however, namely rodents of the genus Rattus and other small mammals (10, 12, 30, 71, 89, 113, 123, 135, 137, 151, 222, 283, 301, 304, 306) and dogs (216), as well as mouse and rat fleas (71, 198), in various countries around the world, including the U.S., China, Thailand, Israel, Taiwan, Spain, Portugal, Indonesia, Bangladesh, Japan, and Afghanistan. A serological analysis of trapped Norway rats (Rattus norvegicus) in the U.S. (84) demonstrated that 39% of such rats had detectable antibodies against B. elizabethae. Serological studies in the absence of isolation have indicated a potential link between B. elizabethae and other clinical manifestations, including Leber’s idiopathic stellate neuroretinitis (230) and myocarditis (130, 298). Additionally, an increased rate of seropositivity to B. elizabethae of between 33% and 52% has been demonstrated among cohorts of homeless individuals (87, 276) and urban intravenous drug users (63, 64). Limited by the paucity of confirmed clinical cases of B. elizabethae infection, the significance of the increased seroreactivity to the antigen among certain subgroups remains to be firmly established and is the focus of further discussion below.

Other Bartonellae Several Bartonella species have been implicated as etiologic agents of endocarditis in addition to B. quintana, B. henselae, and B. elizabethae as mentioned above, namely B. vinsonii subsp. berkhoffii, B. vinsonii susp. arupensis, B. clarridgeiae, and B. koehlerae; however, B. henselae and B. quintana remain the most prevalent Bartonella etiologic agents of endocarditis (9, 70, 100, 176, 261). B. grahamii has been implicated in neuroretinitis (148). B. rochalimaea, B. vinsonii susp. arupensis, B. washoensis, and B. tamiae have all been associated with fever in the host (95, 167, 169, 176, 294). Additionally, B. washoensis has been implicated as an agent of myocarditis (167). Finally, B. rochalimiaea has been involved with cases of bacteremia, splenomegaly, and cutaneous lesions (95).

20 TABLE 2. Clinical entities associated with the pathogenic Bartonella species Clinical Disease Bb Bc Bd Be Bg BhH BhM Bk Bq Br Bt Bva Bvb Bw Bartonellosis X Trench fever X Cat scratch disease X X X X X Bacillary angiomatosis X X Bacillary peliosis hepatitis X X Bacteremia X X X X X X Fever X X X X X X Endocarditis X X X X X X X Myocarditis X Neurologic disorders X X Neuroretinitis X Bb, B. bacilliformis; Bc, B. clarridgeiae; Bd, B. doshiae; Be, B. elizabethae; Bg, B. grahamii; BhH, B. henselae (Houston-1 isolate); BhM, B. henselae (Marseille isolate); Bk, B. koehlerae; Bq, B. quintana; Br, B. rochalimaea; Bt, B. tamiae; Bva, B. vinsonii subsp. arupensis; Bvb, B. vinsonii subsp. berkhoffii; Bw, B. washoensis

Laboratory diagnosis of Bartonella infection As there are no clinical features of Bartonella infections that are pathognomonic among a wide range of potential differential diagnoses, laboratory testing is required for a definitive diagnosis of Bartonella infection. Historically, the original criteria for clinical diagnosis of CSD dictated that 3 out of 4 given conditions be fulfilled: (i) history of contact with a cat with concomitant presence of a scratch or primary dermal or eye lesion, (ii) a positive skin test, utilizing pus derived from infected lymph nodes as antigen, (iii) regional lymphadenopathy in the absence of positive tests for other differential diagnoses of lymphadenopathy, and (iv) characteristic histopathologic changes in a lymph node biopsy specimen (292). Given the potential for transmission of infectious agents, however, in addition to the current availability of advanced and standardized diagnostic assays, the skin test is no longer utilized. Bacterial culturing of blood or tissue samples is impractical and unreliable due to the fastidious nature of the organisms and the highly specific conditions they require for optimal growth, as described above. The prolonged incubation period and fastidious growth requirements characteristic of Bartonellae are also associated with primary media contamination and low isolation sensitivities (6). It has been found that successful isolation or detection of Bartonella spp. organisms from the blood or tissues of infected patients is rare in the absence of systemic disease; in the case of B. henselae infection, for example, the organisms can generally be isolated from immunocompromised patients with endocarditis, BA, peliosis hepatis, and relapsing fever, yet are rarely isolated from specimens of immunocompetent patients with CSD, as much of the lymphadenopathy and other pathologic manifestations of infection in CSD are considered secondary to a cell-mediated immune response against antigenic components of the pathogen rather than to actual bacterial invasion and proliferation (6, 112). Further, as patient specimens are often collected subsequent to the administration of antimicrobial agents, the success rate for bacterial growth remains low (173). For these reasons, Bartonella spp. are rarely cultivated in most microbiology laboratories for diagnostic purposes. PCR techniques are a highly sensitive and specific method for definitive species identification in the diagnosis of Bartonella spp. infections. The genes and genetic loci most commonly targeted in diagnostic PCR assays, as well as phylogenetic analyses, include the 16S rRNA-encoding gene, the citrate synthase gene (gltA), the 16S/23S intergenic spacer region (ITS), the cell division protein-encoding gene (ftsZ), and the 60 kDa heat-shock protein-encoding gene (groEL) (133). Advancements in PCR methodology using real-time PCR or quantitative PCR in the place of conventional PCR have led to decreased risk of carryover contamination and increased quantitative capability, speed, simplicity, and reproducibility (99).

22 Serological testing is currently the most frequently utilized means of clinically diagnosing infection with Bartonella spp., largely on account of accessibility and practicality. Developed by Regnery et al. (253) as a highly sensitive and specific serologic diagnostic assay for the detection of Bartonella spp. in the diagnosis of CSD, the indirect fluorescent-antibody (IFA), which uses Vero cell co-cultivated antigen to reduce the autoadherent nature of the organism, has proven the most commonly employed diagnostic modality to date for the laboratory diagnosis of Bartonella infections (4, 25, 69, 246, 253, 307, 309), though the enzyme-linked immunosorbent assay (ELISA), also referred to as enzyme immunoassay (EIA), is used routinely as well (14, 25, 114, 185). High immunoglobulin (Ig) G titers of >1:64 are suggestive of recent infection, and a four-fold or greater rise in IgG titer in acute and convalescent serum samples is confirmatory of infection (65). A positive titer to IgM is indicative of acute infection, though the window of IgM production is minimal. Whereas titers of >1:64-256 are generally considered seropositive in cases of Bartonella spp. infection, a cutoff titer value of >1:800 is suggested in the clinical serodiagnosis of Bartonella endocarditis (107). Though these serological methodologies are efficient, widespread, and practical, they are hindered by limitations as well, namely differences in sensitivities found in different laboratories, which routinely utilize differing methods of antigen preparation, and numerous reports of cross-reactivity (see below). Additionally, in limited cases, individuals infected with Bartonella spp. may fail to mount a detectable antibody response (189), and, conversely, some infected individuals have been found to remain seropositive for years (69). Modifications to improve the efficacies of serologic detection methods are pending a more comprehensive delineation of the factors influencing the pathogenesis of infection due to Bartonella spp. and the evoked human immune response, such that more refined antigens, such as purified organisms or recombinant antigens, specific to the various species at hand be substituted in order to enhance the sensitivities and specificities of the assays. Various attempts at identifying more specific antigens for incorporation into diagnostic serologic assays have been made via Western blot studies. Facilitated by the identification of species-specific epitopes, Western immunoblot has been used to differentiate Bartonella spp. infectious endocarditis from other Bartonella spp. infections as well as to serologically differentiate Bartonella spp. pathogens at the species level; 10-, 20-, 30-, 45-, 65-, and 83-kDa proteins of B. quintana were found to be more highly recognized by patients with B. quintana endocarditis than by patients with B. henselae endocarditis (134, 192). On account of the cross-reactive potential inherent in other serological assays as mentioned above, various additional investigations have sought to identify species-specific antigenic targets for inclusion in diagnostic assays for other Bartonella infections as well in order to improve their specificity; species-specific immunoreactive epitopes

23 identified thus far include 8-, 17-, 23-, 80-, 83-, 116-, 208-, 208.5-, 209-kDa B. henselae (Houston-1) antigens (5, 182, 185, 210). Further immunoreactive proteins of 41-, 21-, 32-, and 26-kDa have been identified as species-specific epitopes for B. bacilliformis, B. henselae (Marseille), B. quintana, and B. elizabethae, respectively (182). Anti-Bartonella adhesion A (BadA) antibodies have additionally been targeted for use in immunoblot assays for the laboratory diagnosis of B. henselae infections, with a reported sensitivity and specificity of 74% using IFA-positive sera (290). For the diagnosis of chronic bartonellosis in association with B. bacilliformis infection, a diagnostic immunoblot utilizing a sonicated antigen preparation has been developed with a reported 94% sensitivity and 100% specificity (192). Other diagnostic means of more limited efficacies have been employed as well for the identification of Bartonella spp. Microscopic evaluation of infected tissue using Warthin-Starry stain, immunohistochemistry, and transmission electron microscopy has been utilized to provide a histologic diagnosis of Bartonella spp. infection, particularly in the cases of endocarditis, BA, and CSD (65, 111, 181, 250); staining with murine monoclonal antibodies to detect the intracellular, polymorphic organisms directly has also been used either solely or in conjunction with Warthin- Starry staining to visualize the pathogens (136, 259).

Serological cross-reactivity Immunological cross-reactivity as a result of conserved epitope recognition across species, highlighted by the high degree of homologous overlap of Bartonella spp. genome backbones as discussed above, is a frequent occurrence in diagnostic serologic assays for a wide spectrum of bacterial genera. The first documentation of a serological cross-reaction between a Bartonella species and a heterologous pathogen was published in 1976, when Coxiella burnetii sera was found to cross-react with Rochalimaea (later Bartonella) quintana antigen in the diagnosis of trench fever using a passive hemagglutination test (66). Since that time, numerous reports have documented cross-reactivity in serologic assays between Bartonella and other heterologous agents, including Gram-positive and –negative bacteria as well as viruses, namely Afipia felis, Anaplasma phagocytophila, Brucella spp., Borrelia spp., Chlamydia pneumoniae, C. psittaci, C. trachomatis, C. burnetii, cytomegalovirus, Ebstein-Barr virus, Ehrlichia canis, Orientia tsutsugamushi, Rickettsia conorii, R. prowazekii, R. rickettsii, and R. typhi (57, 78, 90-92, 115, 117, 128, 129, 154-156, 172, 179, 186, 192, 205-208, 210, 234, 242, 247, 253, 265, 273, 275, 280, 309, 311). Cross-reactions among the different species of the genus Bartonella have been the most extensively studied and documented cross-reactions in the

24 literature to date. Numerous studies analyzing the serologic response to Bartonella spp. in the diagnosis of CSD and endocarditis have reported markedly high concordance rates of seropositivity to more than one Bartonella spp. antigen. Elevated titers to both B. henselae and B. quintana in the serologic analyses of suspected or confirmed CSD patient sera with concordance rates as high as 94-100% on IFA is a frequent finding (69, 83, 208, 242, 247, 265, 280, 291). Studies employing serologic assays other than the IFA, such as EIA and immunoelectrophoresis, have also reported significant cross-reactivity for these antigens (90-92). Additionally, immunoblot studies have evidenced cross-reactivity between B. henselae, B. vinsonii, B. quintana and B. elizabethae (134, 205). Due to the cross-reactivity seen within the species of Bartonella, several Bartonella spp. antigens are typically included in the antigenic panel of serologic assays, particularly the IFA, in the detection and diagnosis of infections for which more than one of several Bartonella spp. are potential etiologic agents, such as CSD or endocarditis. In general, both B. henselae and B. quintana antigens are included in the assay, with the frequent inclusion of B. elizabethae as well; infrequently, other Bartonella agents are additionally tested, including B. clarridgeiae, B. grahamii, and B. henselae (Marseille strain). Table 3 provides a listing of the numerous heterologous pathogenic agents which have been shown to cross-react with Bartonella spp. antigen or antibodies in various serological assays.

TABLE 3. Studies reporting on the occurrence of heterologous cross-reactivity with Bartonella spp. Author (year) Year Antigen Cross-Reactive Antibodies Assay Cooper et al. 1976 R. quintana C. burnetii PHA1 Hollingdale et al. 1978 R. quintana R. tsutsugamushi, R. prowazekii, EIA2, CIE3 R. typhi, R. rickettsii Hollingdale et al. 1980 B. quintana R. tsutsugamushi CIE Koehler et al. 1988 B. henselae A. felis IPS4 Knobloch et al. (a) 1988 B. bacilliformis C. psittaci ELISA5 (lipopolysaccha- ride) Knobloch et al. (b) 1988 B. bacilliformis C. psittaci ELISA LeBoit et al. 1988 B. henselae A. felis IPS Slater et al. 1990 EBV, CMV B. henselae IFA6 Reed et al. 1992 B. quintana B. henselae IHC7 Regnery et al. (b) 1992 B. henselae Brucella spp., Borrelia spp. IFA Slater et al. (b) 1992 S. aureus B. henselae EIA Engbaek et al. (a) 1994 B. henselae B. quintana, A. felis RLIE8 B. quintana B. henselae, A. felis RLIE A. felis B. henselae, B. quintana RLIE Engbaek et al. (b) 1994 B. quintana B. henselae, A. felis EIA B. henselae B. quintana EIA B. abortus B. henselae, B. quintana EIA

25 Maurin et al. 1994 B. quintana B. vinsonii, B. henselae WI9 Drancourt et al. 1995 C. pneumoniae B. quintana IFA, WI C. trachomatis, B. quintana IFA C. psittaci Szelc-Kelly et al. 1995 A. felis B. henselae ELISA Zbinden et al. 1995 B. henselae EBV (IgM) IFA, WI La Scola et al. 1996 B. henselae C. burnetii IFA B. quintana C. burnetii IFA C. burnetii B. henselae IFA Raoult et al. 1996 C. pneumoniae, B. henselae, B. quintana IFA C. trachomatis, B. henselae, B. elizabethae IFA C. psittaci B. henselae (Houston) IFA B. quintana B. henselae (Marseille) Engbaek et al. 1997 B. henselae B. quintana, A. felis EIA Maurin et al. 1997 C. pneumoniae B. quintana IFA, WI B. quintana C. pneumoniae WI Pappalardo et al. 1997 E. canis B. vinsonii (berkhoffii) IFA McGill et al. 1998 B. henselae R. rickettsii, R. prowazekii, F. WI tularensis, T. pallidum, Chlamydia spp., M. pneumoniae, E. chaffeensis, E. coli, O. tsutsugamushi Sander et al. 1998 B. henselae EBV, C. pneumoniae IFA Graham et al. 2000 C. burnetii B. henselae, B. quintana EIA Mallqui et al. 2000 B. bacilliformis Brucella spp., C. psittaci, C. WI (sonicated) burnetii B. bacilliformis WI (glycine- Brucella spp., C. psittaci, C. extracted) burnetii Maurin et al. 2000 C. trachomatis B. henselae, B. quintana IFA Chomel et al. 2001 B. clarridgeiae E.(Anaplasma) phagocytophila IFA Maurin et al. 2002 B. henselae B. quintana IFA B. quintana B. henselae IFA B. henselae, B. CMV, C. burnetii, R. conorii IFA quintana Gilmore et al. 2003 B. vinsonii F. tularensis, C. burnetii, R. typhi, WI (berkhoffii) B. quintana, B. vinsonii (vinsonii), (sucB gene) B. henselae (Houston), B. elizabethae Houpikian et al. 2003 B. henselae, B. B. henselae, B. quintana WI quintana, B. (endocarditis antisera) elizabethae, B. vinsonii subsp. berkhoffii Litwin et al. 2004 B. henselae B. melitensis, M. pneumoniae, F. WI (sucB gene) tularensis, C. burnetii, R. typhi Znazen et al. 2005 C. pneumoniae, B. henselae, B. quintana IFA C. psittaci, C. trachomatis Rahimian et al. 2006 C. burnetii B. quintana IFA B. quintana B. henselae IFA 1PHA, passive hemagglutination test; 2EIA, enzyme immunoassay; 3CIE, counterimmunoelectrophoresis; 4IPS, immunoperoxidase staining; 5ELISA, enzyme-linked immunosorbent assay; 6IFA, indirect fluorescent-antibody assay; 7IHC, immunohistochemistry; 8RLIE, rocket-line immunoelectrophoresis; 9WI, Western immunoblot

26 Bartonella seroprevalence studies Seroprevalence in various human subpopulations As the spectrum of clinical manifestations with which Bartonella species are etiologically associated continues to unfold, seroepidemiological studies surveying the degree of exposure to different Bartonella spp. within various human subpopulations have aided the understanding of risk factors for infection and the susceptibilities of various individuals to disease. A high degree of variation in prevalence of Bartonella spp. antibodies in select populations has been found in different studies conducted around the world. The seroprevalence of antibodies against B. henselae and B. quintana among blood donors and otherwise healthy individuals, for example, has been shown to range from 0.6% to 61.6% and 0% to 57%, respectively (2, 7, 33, 41, 49, 61, 68, 87, 102, 110, 119, 122, 130, 139, 171, 177, 200, 202, 213, 218, 232, 239, 265, 266, 268, 270, 279, 281, 305, 308) (See Table 4 for a comprehensive listing of the Bartonella spp. seroprevalence reported among blood donors and healthy adults and children world-wide). Among homeless individuals, the prevalence of antibodies against B. quintana varies widely by country as well, with reports of 1.8-54% (41, 43, 106, 119, 260), 57% (270), 48.3% (54), 4.2% (87), and 9-20% (139, 276) seroprevalence in France, Japan, Poland, Sweden, and the U.S., respectively. In Sweden and the U.S., sera from the homeless population was tested against additional Bartonella antigens as well, with reported seroprevalence of 29.2%, 52.1%, and 2.1% to B. henselae, B. elizabethae, and B. grahamii antigens, respectively, in Sweden (87) and 3.5% and 12.5% to B. henselae and B. elizabethae, respectively, in the U.S. (276). Among HIV-positive individuals, prevalence of antibodies against Bartonella spp. have been reported to be relatively high, with seroprevalence of 38.4% (177) and 40.8% (175) in Brazil, 14% (245), 17.3% (34) and 22.3% (240) in Spain, and 17% (160) in the U.S. High levels of B. elizabethae antibodies have been reported among intravenous drug users (IVDU) in both the U.S. and Sweden, with seroprevalences of 33% and 46% in two different studies in the U.S. (63, 64) and 39% in Sweden (212). Reported seroprevalence for other Bartonella spp. antigens among this cohort include the following: B. henselae, 11% (63), 10% (63), and 14% (212); B. quintana, 10% (63, 64), and 3% (212); B. grahamii, 3% (212); and total Bartonella spp., 14% (245). Investigations targeting veterinary-associated individuals have shown B. henselae seroprevalence of 65% in Austria (145), 10.3% in Chile (102), 10.9-15% in Japan (150, 171), 45% in Poland (54), 1.7% in Taiwan (52), and 7.1% in the U.S. (227). Finally, B. henselae antibodies have been shown to be highly prevalent among cat owners in various distributions around the world, with 26%, 10.3%, 53.3%, and 28.9% seroprevalence reported in the

27 countries of Austria (272), Chile (102), Poland (54), and Spain (33), respectively.

TABLE 4. Seroprevalence of Bartonella spp. among blood donors and other healthy adults and children in various countries Ref Country Assay N Study population Bartonella spp. Total tested sero- positive (%) 68 Brazil IFA1 437 Blood donors B. henselae 13.7 B. quintana 12.8 177 Brazil IFA 125 Blood donors B. henselae 34.4 61 British IFA 142 Healthy Adults B. henselae 36.8 Columbia 102 Chile IFA 181 Healthy children B. henselae 13.3 279 China IFA 351 Blood donors B. henselae 14.8 49 China IFA 579 Healthy Bartonella spp. 10.9 individuals 232 Croatia IFA 100 Blood donors B. henselae 50 Healthy children 268 Denmark IFA 159 Blood donors B. henselae 0.6 B. quintana 0 B. elizabethae 0 41 France IFA 250 Blood donors B. quintana 0 119 France IFA 53 Blood donors B. quintana 2 265 Germany IFA 270 Healthy adults B. henselae 30 266 Germany IFA 116 Healthy children B. henselae 13 7 Greece IFA 63 Healthy children Bartonella spp. 15.9 281 Greece IFA 500 Blood donors B. henselae 19.8 B. quintana 15 218 Greece IFA 491 Blood donors B. henselae 21.6 202 Italy IFA 508 Healthy children B. henselae 61.6 171 Japan IP2 155 Healthy students B. henselae 4.5 270 Japan IFA 200 Blood donors B. quintana 57 2 Jordan IFA 482 Healthy children B. henselae 11 B. quintana 4.1 308 New IFA 140 Blood donors B. henselae 5 Zealand 164 Peru WB3 544 Healthy B. bacilliformis 77.5 individuals 50 Peru IFA 690 Children and B. bacilliformis 45 adults in endemic area4 33 Spain IFA 85 Blood donors B. henselae 5.9 110 Spain IFA 146 Healthy B. henselae, 24.7 individuals B. quintana 239 Spain IFA 218 Healthy B. henselae 8.7 individuals 130 Sweden IFA 100 Blood donors B. henselae 1

28 B. quintana 0 B. elizabethae 4 211 Sweden IFA 322 Blood donors B. henselae 1.6 B. quintana 0.3 B. elizabethae 6.8 213 Sweden IFA 498 Blood donors B. henselae(H)5 1.2 B. henselae (M)6 1.8 B. quintana 0.2 B. elizabethae 14.1 B. grahamii 2.6 B. vinsonii6 0 87 Sweden IFA 61 Blood donors B. henselae 1.6 B. quintana 1.6 B. elizabethae 14.8 B. grahamii 4.9 200 Thailand IFA 163 Blood donors B. henselae 5.5 305 Turkey IFA 800 Blood donors B. henselae 6 139 U.S.A. IFA 199 Blood donors B. quintana 2 122 U.K. IFA 200 Blood donors B. henselae 1.5 B. quintana 1.5 1IFA, indirect fluorescent-antibody assay; 2IP, immunoperoxidase test; 3WB, Western blot; 4subjects derived from an endemic area of Carrion’s disease; 5B. henselae (H), Houston-1 strain; 6B. henselae (M), Marseille strain; 6B. vinsonii subsp. vinsonii

Seroprevalence in cats As the domestic cat serves as a natural reservoir for some Bartonella spp., namely B. henselae, B. clarridgeiae, B. koehlerae, and B. weissii, and is responsible for direct transmission of disease to humans, a multitude of seroprevalence studies over the past 2 decades have sought to investigate the degree of exposure of cats to Bartonella spp. in a worldwide distribution. The results of such serological studies indicate that cats harbor a significant degree of Bartonella spp. antibodies, primarily to B. henselae, with total seroprevalence of up to 93% (229). See Table 5 for a complete listing of the results of the numerous feline seroprevalence studies conducted in various regions of the world. Observed trends in the results of the studies below indicate that feral cats typically exhibit a higher prevalence of antibodies against B. henselae than domestic cats, presumably due to increased exposure to arthropod vectors. There is also a temperature gradient such that higher B. henselae seroprevalences correlate with warmer regions of the world, as evidenced in the studies in Chile (85.6%), Spain (71.4%), Hawaii (73.7%), compared to the cooler regions of Sweden (1%), Alaska (5%), and Norway (1%).

29 TABLE 5. Feline seroprevalence of Bartonella spp. in various cat populations of a world-wide distribution Ref. Country Assay No. of Type of cat (N) Bartonella spp. Pos. Total cats tested (%) Pos. (%) 67 Brazil IFA1 40 Veterinary2 B. henselae 47.5 47.5 180 Canada IFA 242 Farmed (242) B. henselae 17.8 NS3 B. clarridgeiae 18.6 1 01 Chile IFA 187 Domestic (NS) B. henselae 72.5 85.6 Feral (NS) 90.4 58 Denmark IFA 93 Domestic (44) B. henselae 44.2 45.7 Feral (49) 46.9 121 Germany IFA 713 Veterinary B. henselae 6.2 15 (713) (H)4 B. henselae 5.6 (M)5 B. quintana 14.3 96 Italy IFA 540 Feral (540) B. henselae 38 38 97 Italy IFA 1581 Feral (1416) B. henselae 39 39.5 Domestic 43.5 (165) 236 Italy IFA 79 Veterinary (79) B. henselae 21.5 21.5 286 Japan IFA 199 Domestic B. henselae 15.1 15.1 (199) 199 Japan IFA 471 Domestic B. henselae 9.1 9.1 (471) 201 Japan IFA 1,447 Domestic B. henselae 8.8 8.8 (1,447) 1 Jordan IFA 153 Domestic B. henselae 32 32 (153) B. quintana 1.5 26 Netherlands EIA 163 Sheltered (113) B. henselae 50 51.8 Domestic (50) 56 21 Norway IFA 100 Veterinary B. henselae 0 1 EIA6 (100) 1 238 Poland IFA 36 Domestic (36) B. henselae 86 86 18 Scotland EIA 78 Domestic (61) Bartonella spp. 14.3 15.4 Sheltered (17) 20.0 225 Singapore IFA 80 Feral (80) B. henselae 47.5 47.5 277 Spain IFA 168 NS B. henselae 71.4 71.4 127 Sweden IFA 292 Veterinary B. henselae 1 25 B. quintana 0 B. elizabethae 25 116 Switzerland IFA 728 Veterinary B. henselae 8.3 8.3 52 Taiwan IFA 131 Domestic (30) B. henselae 26.7 23.7 Farmed (37) 0 Sheltered (64) 40.0 48 Turkey IFA 256 Domestic B. henselae 13.0 18.8 (154) 27.5 Feral (102)

30 15 U.K. EIA 148 Domestic (69) B. henselae 40.6 41.2 Feral (79) 41.8 53 U.S.A. IFA 1370 Veterinary B. henselae 27.4 29.4 (NS) B. quintana 21.8 Sheltered (NS) 75 U.S.A. IFA 57 Veterinary (57) B. henselae 73.7 73.7 (Hawaii) 142 U.S.A. IFA 628 Diagnostic B. henselae 27.9 Southeast laboratory- 54.6 Hawaii associated 47.4 California (628) 40.0 Pacific NW 34.3 S C7 Plains 36.7 Alaska 5.0 RM/GP8 3.7 Midwest 6.7 105 U.S.A. IFA 159 Cattery (159) B. henselae 35.8 35.8 229 U.S.A. IFA 176 Domestic (76) B. henselae 75 88.2 Feral (100) 93 46 U.S.A. IFA 170 Domestic (20) B. henselae 30.0 24.1 Feral (36) 38.9 Sheltered (114) 18.4 81 U.S.A. MAT9 210 Domestic B. henselae 25.7 25.7 (210) 17 U.S.A. EIA 458 Domestic B. henselae 19.0 19.0 (458) 80 West Indies MAT 176 Domestic (75) B. henselae 50.6 51.7 Feral (101) 52.4 1IFA, indirect fluorescent-antibody assay; 2Veterinary, domestic + feral cats; 3NS, not specified; 4B. henselae(H), Houston-1 strain; 5B. henselae(M), Marseille strain; 6EIA, enzyme immunoassay or enzyme-linked immunosorbent assay (ELISA); 7S C Plains, South Central Plains, U.S.A.; 8RM/GP, Rocky Mountain-Great Plains Region, U.S.A.; 9MAT, modified agglutination test

Bartonella in Sweden Swedish orienteering investigation During the 14-year period between 1979 and 1992, an accumulation of cases of sudden unexpected cardiac death (SUCD) occurred among Swedish orienteers, the majority of whom were nationally ranked as elite competitors in the exclusively outdoor sport pursued by 2% of the entire Swedish population. That a total of 16 cases of SUCD (corresponding to an incidence of 10-100/100,000 person years) occurred in this relatively short expanse of time was indicative of an increased death rate among the group of ranked male orienteers <35 y when compared to an expected SUCD incidence of 0.5-1/100,000 among young athletes (209, 298). At autopsy, the clinical

31 diagnosis for cause of death in the majority (75%) of cases was histopathological myocarditis, with concomitant arrhythmogenic right ventricular cardiomyopathy (ARVC) present in 25% of the myocarditis cases (178, 297). Given that myocarditis is predominantly associated with a microbial etiology (146, 303), an investigation was commenced in 1992 to probe any potential link between pathogenic microbial agents and the deaths in the orienteers. Subsequently, it was reported that Chlamydia pneumonia genetic material was detected in myocardial and lung tissues by PCR from one of the SUCD cases, in addition to serological evidence of C. pneumonia in all available sera (296). Later studies using healthy, age- and gender- matched controls, however, failed to detect a difference in seroprevalence of C. pneumonia antibodies in the orienteers and blood donor control sera (126, 297). As Bartonellae were known to cross-react with Chlamydial antigens on serologic assays and included several novel zoonotic agents with significant pathogenic potential, the focus of the investigation shifted to investigate the potential association of Bartonella spp. with the cases of SUCD. In 2001, Wesslén et al. (298) reported the finding of Bartonella spp. DNA in the myocardial tissues of 4 SUCD cases and in the lung of a fifth case, while myocardial tissues from controls (cases of accidental deaths) were negative. Sequences were similar to B. quintana in 2 cases and identical to B. henselae in 3 cases, leading to the hypothesis that Bartonella- induced silent subacute myocarditis with the potential involvement of ARVC contributed to the increased SUCD among the Swedish orienteers (298). The finding of 2 different species of Bartonella in these cases as potential etiologic agents of myocarditis is paralleled by the etiologic association between Bartonella spp. and endocarditis, for which multiple Bartonella spp., including B. henselae, B. quintana, and B. elizabethae, are implicated as etiologic agents. This potential causal link between Bartonella spp. and myocarditis and/or ARVC does not, however, preclude the possibility that the isolation of Bartonella spp. in the aforementioned cases was an incidental finding. As the majority of case sera also evidenced antibodies to Bartonella spp., it became of interest to conduct a large-scale serological investigation of all elite orienteers in Sweden for antibodies to Bartonella spp.

Human and animal studies in Sweden Bartonella infections are recognized as prominent zoonoses in the United States, with CSD alone afflicting approximately 24,000 persons annually (138). However, prior to the initiation of the studies included in this thesis, the existence of Bartonella spp. infection and seroprevalence of Bartonella spp. antibodies among various cohorts of the Swedish population were unknown. The first evidence of Bartonella spp. infection in Sweden was

32 reported by Holmberg et al. in 1999 (130) with the PCR detection of B. quintana-like DNA in myocardial and liver tissue of a Swedish patient with fatal myocarditis in accordance with serological evidence of antibodies to pathogenic Bartonella spp. in several patient sera. Shortly thereafter, PCR detection of Bartonella spp. was likewise confirmed in several additional cases of myocarditis associated with SUCD in the sport of orienteering, as described above (298). Such corroborative data led to the suggestion of an association between Bartonella infection and fatal myocarditis. Subsequent human studies in Sweden have documented PCR-confirmed Bartonella spp. endocarditis (88), B. henselae isolated from patients with atypical manifestations of infection (85), Bartonella spp. isolated from pathologic schlerotic aortic valves (226), as well as Bartonella spp. seropositivity in various cohorts of the Swedish population, including blood donors (87, 130, 213), homeless persons (87), intravenous drug users (212), orienteers (211), and patients with infectious endocarditis (295) and cat bites (299). In addition to human studies, various animal investigations have recently reported the isolation and sequencing of various Bartonella spp. strains, including B. grahamii, from small sylvatic mammals (23, 86, 131) and seroprevalence to Bartonella spp. in cats (94, 127) in Sweden. See Table 6 for a current, comprehensive listing of the clinical, serological, microbiological, forensic, field, and histological studies conducted on human and animal samples in Sweden since the initiation of Bartonella testing in 1999. Serological testing for Bartonella spp. antibodies has currently become routine in the diagnostic laboratories in Uppsala, Sweden.

TABLE 6. Investigative human and animal studies of Bartonella in Sweden Authors Type Subjects (N) Controls Assays Significant Findings (year) of (N) study1 Holmberg C, S, Human Human IFA2, First documentation of et al. M patients blood PCR3 Bartonella spp. in Sweden; (1999) (126) donors 8.3% vs 4% Bartonella spp. (100) seroprevalence in subjects and controls, respectively; Bartonella isolated from 2 patients, including fatal myocarditis case Ehrenborg C, M Human NT4 IFA, B. henselae ftsZ gene et al. patients (7) PCR amplified in Swedish patients (2000) with atypical manifestations of infection McGill S, E Elite Human IFA High Bartonella spp. et al. orienteers blood seroprevalence in Swedish (2001) (1136) donors elite orienteers (31% vs 6.8% (322) in controls)

33 Wesslén S, M Deceased Trauma IFA, Bartonella spp. isolated from et al. orienteers autopsy PCR myocardial (4/5) and lung (2001) (SUCD5 cases (1/5) tissue obtained from 5 cases) (7) (6) cases of SUCD among orienteers in Sweden; seropositivity to Bartonella spp. in 2 additional cases of ARVC death in orienteers Hjelm S, E Domestic NT IFA 25%, 1% and 0% et cats (292) seroprevalence to B. al.(2002) elizabethae, B. henselae and B. quintana, respectively, in Swedish cats Holmberg FD, Sylvatic NT PCR Six novel Bartonella spp. et al. M small genotypes isolated from 6 (2003) mammals species of sylvatic small (236) mammals in Sweden; the most frequent genotype, B. grahamii, isolated from A. flavicollis and M. musculus Engvall M Domestic NT PCR Two of 91 pet cats were found et al. cats (91) to be bacteremic with B. (2003) henselae, indicating a low overall prevalence McGill F IVDU6 at Non- IFA High rate of seropositivity et al. autopsy (59) IVDU at (39%; primarily B. (2003) autopsy elizabethae) to Bartonella spp. (44) among Swedish intravenous heroin addicts at autopsy Ehrenborg B. grahamii NT IRS- Different IRS-PCR patterns et al. isolates from PCR7, obtained from all B. grahamii (2003) 4 rodent spp. culture isolates; all strains clustered (31) into 2 distinct groups Werner S IE8 patients NT IFA Bartonella spp. serologically et al. (329) evidenced in 4.5% of IE (2003) episodes (yet with titers <1:8009) La Scola M Ixodes NT PCR, No evidence of Bartonella et al. ricinis ticks culture spp. demonstrated in any of (2004) (167) the 167 ticks tested Nilsson F, M, Autopsy Autopsy PCR, Bartonella spp. DNA et al. H pathological healthy IHC10 demonstrated in 2.4% of the (2005) aortic valves aortic pathological valves studied (84) valves (15) McGill S, E Blood NT IFA Risk factors for Bartonella et al. donors (500) infection in Sweden identified, (2005) including working outdoors and contact with cats

34 Ehrenborg C, S, Homeless Blood PCR, Total seroprevalence of 62.5% et al. M, E patients (50) donors IFA to Bartonella spp. (primarily (2008) (50) B. elizabethae) in homeless subjects vs. 18% in controls; no evidence of trench fever Westling S Cat bite Non-cat IFA Total Bartonella spp. et al. patients (71) bite seroreactivity of 39% (34% to (2008) patients B. henselae) in patients with (117) infected cat bites (vs. 3.4% in controls) Ehrenborg C, S, Endocarditis NT IFA, Positive PCR and serology et al. M, H patient (1) PCR findings for Bartonella spp. in (2009) endocarditis patient; first confirmed Swedish case of endocarditis 1Type of study: C, clinical; S, serological; M, microbiological; E, epidemiological; FD, field; F, forensic;H, histological; 2IFA, indirect fluorescent-antibody assay; 3PCR, polymerase chain reaction; 4NT, not tested; 5SUCD, sudden unexpected cardiac death; 6IVDU, intravenous drug user; 7IRS-PCR, infrequent restriction sequence polymerase chain reaction; 8IE, infectious endocarditis; 9Cut-off levels for diagnosis of Bartonella endocarditis are recommended at >1:800 (107); 10IHC, immunohistochemistry

35 AIMS OF THE PRESENT STUDY

The specific aims of the present investigations were as follows:

Expand the characterization of Bartonella spp. immunoreactivity by 1) identifying seroreactive and serodominant epitopes of B. henselae antigen using a series of Western immunoblot analyses (I) and 2) dissecting the induced humoral immune response to Bartonella spp. antigen, specifically B. henselae, at the immunoglobulin (Ig) isotype and IgG-specific fragment and subclass levels (I)

Evaluate the prevalence of antibodies against pathogenic strains of Bartonella within various cohorts of the Swedish human and animal population, including patients (II), blood donors (II, III, VI), elite orienteers (III), intravenous drug users (V), and cats (IV)

Conduct statistical analyses to identify epidemiological and demographical factors (IV, VI), as well as correlative co-morbidities (V) associated with Bartonella infection

Assess the level of Bartonella spp. intragenic cross-reactivity, as well as intergenic cross-reactivity with members of other genera in serological assays, namely Western blot and indirect fluorescent- antibody assay (IFA)

36 MATERIALS AND METHODS

Cultivation of Bartonella spp. antigens (I-VI) The following bacterial strains were used in the present studies: B. elizabethae (F9251, ATCC 49927) (II-VI), B. grahamii (ATCC 700132) (V,VI), B. henselae (Houston-1 isolate, ATCC 49882) (I-VI), B. henselae (Marseille isolate, URL-LY8) (V,VI), B. quintana (OK-90-268) (I-VI), and B. vinsonii subsp. vinsonii (ATCC VR-152) (V,VI). All Bartonella spp. antigens used in the studies were cultivated with the assistance of Dr. R. Regnery or provided by him directly. The bacterial strains were cultivated on heart infusion agar supplemented with 5% defibrinated rabbit blood (BBL, Cockeysville, Md.). For bacterial whole-cell antigen, plates inoculated with Bartonella spp. were incubated for 3 to 5 days at 32°C in the presence of 5% CO2. Following incubation, bacterial cells were harvested and suspended in brain heart infusion media. The cells were then pelleted and resuspended in phosphate-buffered saline (PBS). Number of colony forming units (CFU) of the harvested Bartonella antigen was determined using titration on blood agar plates. For cell culture-derived antigen, Bartonella spp. cells were co-cultivated with antibiotic-free Vero cell monolayers using a previously established protocol (253). All antigens were inactivated by gamma irradiation (5 x 105 rad) and stored at -70°C prior to further use.

Study samples (I-VI) See Table 7 for a complete listing of the cases and controls used in the present studies and corresponding clinical and epidemiological data (I-VI).

37 TABLE 7. Cases and controls studied Paper Type Clinical data Sample type Year/s No. of samples Age range collected (M,F) (mean) I Case CSD1 Human serum 1994-95 54 (ND2) ND Control No CSD Human serum 1994-95 15 (ND) ND Control B. quintana + Human serum 1994-95 4 (ND) ND Control Stock antisera3 Human serum ND 8 (ND) ND II Case SDT4 Human serum 1994-97 126 (61,51) 1-82 (34) Control Healthy Human serum 1992-93 100 (64,36) 20-59 (39) Case Myocarditis Human heart 1994 1 (NA5) NA tissue, serum Control No heart Human heart NA 6 (NA) NA disease tissue III Case Elite orienteers Human serum 1992-93 1136 (766,370) 16-39 (6) Control Healthy; non- Human serum 1992-93 100 (64,36) 20-59 (39) orienteer Control Healthy; non- Human serum 1992-93 222 (110,112) 18-32 (24) orienteer IV Case SDT Cat serum 1998 292 (NA) NA Control B. henselae7 + Cat serum 1998 1 (NA) NA V Case Heroin IVDU8 Human serum 1987-92 59 (50,9) 22-43 (32) at autopsy Control Non-IVDU at Human serum 1987-92 44 (37,7) 20-55 (41) autopsy VI Case Healthy Human serum 1999 498 (314,184) 18-73 (40) 1CSD, cat-scratch disease; 2ND, not determined; 3Human antisera against the following bacterial strains: Rickettsia rickettsii (spotted fever group), Chlamydia group, Treponema pallidum, Orientia tsutsugamushi, Fransciscella tularensis, Ehrlichia chaffeensis, Mycoplasma pneumoniae, and Escherichia coli; 4SDT, samples submitted for diagnostic testing; 5NA, not available; 674% of the orienteers were <25 years of age; 7serum sample from a cat immunized with B. henselae; 8IVDU, intravenous drug user

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (I) Cell proteins (~7.5mg/ml) of whole-cell agar-grown Bartonella spp. antigen were concentrated by centrifugation and then lysed in a solution of 1X Tris- glycine SDS sample buffer (Novex, San Diego, Calif.) and 10% beta- mercaptoethanol at 100°C for 10 min. Directly thereafter, the heated suspension underwent electrophoresis in a 4-20% polyacrylamide Tris- glycine minigel (Novex) at 120V for 2h. A prestained broad-range molecular weight protein marker (Bio-Rad) was used as a standard.

Western blot analysis (I) Following SDS-PAGE, proteins were transferred and fixed to a nitrocellulose membrane (Bio-Rad) using the protocol described in Paper I.

38 Antibodies from the reaction of test and control sera that were bound to proteins on membranes were detected by incubation with a 1:5,000 dilution of horseradish peroxidase-labeled anti-human immunoglobulin (Ig) (Kirkegaard & Perry, Gaithersburg, Md.) with specificity for one of the following: IgG (heavy and light chains), IgG1, IgG2, IgG3, IgG4, IgM, IgE, IgA (secretory), and IgA (alpha). Antigens were detected with a 3,3’, 5,5’ tetramethyl benzidine (TMB) membrane substrate developer (Kirkegaard & Perry).

IFA (II-VI) Aliquots of crude Bartonella spp. whole cell antigen co-cultivated with Vero cells as described above were applied to multiple-welled Teflon-coated microscope slides (Novakemi AB, Sweden), air-dried, fixed in acetone and stored at -70°C until used further. The IFA assay used in the current studies followed a previously established protocol (253) with slight modifications. A 1:120 dilution of commercial fluorescein isothiocyanate (FITC)- conjugated rabbit anti-human IgG (Dakopatts, Denmark) was used as conjugate. Bartonella-specific immunofluorescence was subjectively scored by microscope analysis on a scale of 0-3+, with a rating of >2+ at a dilution of 1:64 rendered indicative of seropositivity for all Bartonella spp. antigens in each study. The cutoff titer (CT) 1:64 has been previously documented for B. elizabethae, B. henselae (Houston-1 isolate), and B. quintana (63, 253). However, as no CT values had been previously established for B. grahamii, B. henselae (Marseille isolate), or B. vinsonii subsp. vinsonii due to an absence of a clinical correlate, this 1:64 CT, though arbitrary, was used for those antigens as well in the present investigations. Titers were reported as reciprocal values of serum endpoint dilutions.

PCR and DNA sequencing (II) PCR amplifications were performed targeting the gltA gene, as the conserved nature of this particular gene allows for optimal differentiation of Bartonella spp. and exhibits narrow standard deviation with DNA-DNA homology results (30). The oligonucleotide primers selected were those complementary to the three Bartonella species tested in the study: B. henselae, B. quintana, and B. elizabethae. A set of primers amplifying 685 bp of the gene was constructed utilizing Oligo version 4.0 for Macintosh (National Biosciences, Inc., Plymouth, Minn.). A seminested amplification protocol was followed, with conditions and parameters listed in Paper II. The first PCR amplification used primers BHCS212.p and BHCS897.n,

39 followed by a second amplification using 1 µl from the first reaction as template. Two separate seminested PCR reactions were performed using either the primers BHCS212.p and BHCS613.n or the primers BHCS510.p and BHCS897.n. Sequencing products were generated from the four aforementioned oligonucleotides using a DNA sequencing kit (Dye terminator cycle sequencing kit; Applied Biosystems Inc., Foster City, Calif.). Sequencing products were subsequently separated using a 310 automated sequencer (Applied Biosystems).

Post-mortem and forensic analysis (II, V) (II) Myocardial tissue and serum samples from a patient with fatal myocarditis were obtained at autopsy (See Table 7). DNA was extracted from the myocardial tissue for further analysis using the QiaAmp Tissue Kit (Qiagen Inc., Stanford, Calif), with an additional final ethanol precipitation. (V) Bodies were stored in the mortuary at +4-8C. Prior to autopsy, blood samples were analyzed for HIV antibodies as previously recommended (244). Myocardial tissue was obtained from sections of the anterior wall, interatrial and intraventricular septa, the left anterior and posterior ventricular wall, and the left anterior and posterior papillary muscles according to a previously established protocol (243). Pooled bilateral femoral vein blood and bladder urine were collected for toxicological investigation.

Collection of epidemiological data (III,VI) Participants were requested to provide information in the form of a voluntary questionnaire regarding the following demographical and epidemiological information: age, gender, history of blood transfusion, pet contacts and history of bites/ scratches, living and working milieu, frequency of outdoor activity in fields and forest, involvement in hunting and farming activities with specification of the animals involved, level of physical activity and participation in sports, and travel abroad.

Statistical analysis (III-VI) Data analysis consisted of comparison of independent proportions using the 2-test for testing observed frequencies (III-VI), calculating odds ratios with 95% confidence intervals (III), and using univariate ( 2-test and t-test) and multiple logistic regression (M) analyses to determine the impact of different independent risk factors using SPSS 10.0 software with the assistance of Dr.

40 J. Bring of Statisticon AB Statistics and Research (VI). Both forward selection and backward elimination were used as model selection methods, with the stepwise criteria set at p <0.10 (VI).

41 RESULTS AND DISCUSSION

Analysis of human humoral immune response to Bartonella spp. infection by Western blot (I) Total immunoglobulin (Ig)G reactivity to B. henselae whole-cell and Vero cell co-cultivated antigen (I) In an effort to elucidate serologic parameters of Bartonella spp. infection and identify potential species-specific epitopes of B. henselae antigens, a total of 54 human serum specimens obtained from patients with IFA- confirmed cat scratch disease (CSD) and 15 control sera were reacted by Western blot against two different preparations of B. henselae antigens, a whole-cell agar-grown preparation and a Vero cell culture-derived antigen preparation. Current serological diagnostic assays, namely the IFA and EIA, are based primarily on circulating levels of IgG antibodies in the sera; therefore, total IgG reactivity against B. henselae antigens by immublot was tested initially. Results indicate that human humoral responses to B. henselae infection are variable from patient to patient, with varying numbers of antigenic proteins recognized as well as varying corresponding relative strengths of such reactions on immunoblot; this variability, however, does not appear to directly correlate with variations in IFA titers. High-titer serum, for example, does not necessarily correspond to stronger or more numerous bands on Western blot relative to low-titer serum. Such data preliminarily suggests that IFA is favorable over Western blot analysis in the determination of IgG titer values for diagnostic assessment. As seen in Figure 1 below, seroreactions of CSD sera to both B. henselae agar-grown and Vero cell co-cultivated antigen yielded heterogeneous IgG activity, with an average of approximately 10-15 B. henselae recognized epitopes per serum sample; reactive proteins for both antigen preparations were found to be predominantly in the range of 17-150 kDa, a finding which corroborates the findings reported in other studies of B. henselae immunoblot analyses (5, 92, 138, 185, 188, 205), with a particular concentration of bands in the 50- to 150-kDa region of the membrane. However, reaction against previously documented immunodominant proteins in other studies, namely at 17-, 48-, 69-, 97-, 116-, and 200-kDa, yielded inconsistent reactivity in our trials and occurred in our negative control samples as well, such as the multiple bands at approximately 200-kDa and

42 the 116-kDa proteins described by Litwin et al. (185). In contrast, one B. henselae protein 83-kDa in size, designated Bh83, proved to be an immunodominant epitope in the reactions of all seropositive CSD samples tested in our study and, notably, was not recognized by any of the negative control sera (Figure 1). That Bh83 was immunoreactive by only IFA- confirmed B.henselae-infected sera to both whole-cell agar-grown antigen as well as Vero-cell cocultivated antigen substantiates the implication that this particular epitope of B. henselae is a conserved antigen that is universally expressed during infection (International Patent No. PCT/US99/27344; CDC Ref. No I-002-99/0). As such, in addition to the finding that Bh83 was not recognized by other serum derived from infections caused by heterogeneous bacterial pathogens (see below “Cross Reactivity”), the future incorporation of this epitope into current standard diagnostic tests may hold potential for the development of a more specific serologic assay for the clinical diagnosis of B. henselae infections.

FIGURE 1. (a) Western blot analysis of serum IgG (heavy and light chains) in CSD-positive sera (lanes 3, 4, 5, 7, 8, and 9) and negative controls (lanes 1, 2, 6, and 10) reacting with SDS-PAGE-separated antigenic proteins of B. henselae whole-cell antigen co-cultivated with Vero cells. (b) Western blot analysis of serum IgG (heavy and light chains) in CSD-positive sera (lanes 2-8) and negative controls (lanes 1 and 9) reacting with SDS-PAGE-separated proteins of agar-grown B. henselae antigen. Position of Bh83, recognized exclusively by all IFA-postive sera tested, indicated by an arrow in each panel. Molecular weight standards (kDa) listed at right.

43 IgG-specific fragment and subclass reactivity to B. henselae antigen As immunoblot analysis indicated a strong heterogeneic response of serum total IgG (heavy and light chains) to B. henselae antigen with a specific reaction to Bh83, further dissections of the specific fragments and subclasses of IgG responsible for such reactions were carried out. Whereas each of three IgG-specific fragments, IgG (F[ab]), IgG (F[ab’]2) and IgG (Fc), yielded indistinguishable immunoblot patterns from those of total IgG (heavy and light chains) when reacted against total B. henselae antigen (see Table 8), immunoblot banding patterns among the four subclasses of IgG, IgG1-4, were discriminative and provided insight into the distribution of IgG subclass antibodies induced in the humoral immune response of CSD patients. Given that the four different subclasses comprising human IgG are antigenically distinct with subtle amino acid variations conferring differing biological functions, assessment of the distribution and selective expression of IgG subclasses in association with B. henselae infection may provide clues to the pathogenicity of the organism as well as the type of immune response induced during infection. For example, IgG1 and IgG3 antibodies are generally targeted at protein antigenic components, whereas IgG2 and IgG4 have an affinity for polysaccharide antigens (233). Further, IgG subclasses differ in their induction of the complement cascade (IgG3> IgG1> IgG2); IgG4 is not able to initiate the complement cascade at all. Differences in affinities for the Fc receptors of phagocytic cells also mediate bacterial clearance, with IgG1 and IgG3 binding with the greatest affinity, followed by IgG4 and IgG2 with intermediate and low affinities, respectively (233). Western blot assays specific for IgG2, IgG3, and IgG4 yielded little to no seroreactivity; a prominent reaction against B. henselae antigen among the IgG subclasses was limited to IgG1, indicating a primary role in the immune response induction during CSD infection. In fact, the immunoblot banding patterns of anti-IgG1 conjugate were nearly indistinguishable from the banding patterns observed when using anti-total IgG (heavy and light chains) conjugate. The predominant IgG1 activity in sera and concomitant lack of IgG2 and IgG4 production suggests that the humoral immune response elicited during B. henselae infection is targeted against bacterial protein domains rather than carbohydrate surface antigens of the pathogen. Mediation of antibody-dependent cytotoxicity and enhanced phagocytosis via Fc receptor attachment presiding over complement activation is further implied. The lack of significant IgG3 activity in the reactions of CSD sera and B. henselae antigen further suggests that the role of complement activation, if induced, is assumed by IgG1. An in-vitro study, in fact, reported a lack of antibody-mediated complement fixation in Bartonella- specific sera, supporting the absence of an IgG3-mediated response in CSD infection (258). However, being the most abundant IgG subclass in human

44 sera, it is plausible that the intrinsic levels of IgG1 in human sera are sufficient to mount a complement activation response and commence the cascade leading to bactericidal activity in vivo. It could further be postulated that preexisting IgG1 antibody levels in the sera may play a role as a determinant of clinical disease severity, duration, and symptomatology in CSD; this supposition is reinforced by the atypical severe disease associated with immunocompromised individuals. The aforementioned roles of IgG1 in the immune response, however, do not preclude the potential role played by other arms of the human immune response, such as production of specific secretory IgA at the mucosal level or cellular-mediated immune mechanisms.

Immunoglobulin isotype reactivity to B. henselae antigen In addition to dissecting the subclasses of IgG elicited during CSD infection and identifying reactive B. henselae antigens, the involvement of immunoglobulin isotypes in the induced immune response was assessed as well. Western blot analysis was employed to detect differences in the levels of IgA, IgG, IgM, and IgE antibodies present in 25 CSD-positive sera following a mounted immune response upon infection with B. henselae. Whereas IgE and IgM antibodies were essentially absent in our samples, with the exception of 2 faint bands in the reactions of 2 CSD sera with anti- IgM conjugate, B. henselae-specific IgA (secretory and alpha chains) antibody activity was evidenced in 15 of the 25 positive samples tested, with multiband recognition of B. henselae whole-cell antigen of a heterogeneous nature similar to that seen with total IgG sereactivity. Included among the spectrum of B. henselae proteins recognized in the reactions of IgA and B. henselae whole-cell antigen was Bh83. As circulating IgM antibodies have been detected in the sera of CSD-infected individuals via IFA yet are noted to be short-lived with a minimal window of production, the lack of IgM in our assays may simply represent a convalescent rather than acute status of our infected serum samples. Further, the presence of secretory IgA in our patient samples may be indicative of a secretory component of B. henselae immunity. See Table 8 for a breakdown of the distribution of human immunoglobulins present in CSD-infected sera and their specific reactivities to total B. henselae antigen and Bh83. Continued advancements in the knowledge regarding human immune characterization of CSD are essential in our efforts to decipher the complex nature of B. henselae infections.

45 TABLE 8. Immunoglobulin isotype and IgG-specific fragment and subclass seroreactivity to B. henselae antigen and Bh83epitope as assessed by Western blot Reactivity against: Antibody type B. henselaea Bh83 Total IgG (heavy and light chains) + + IgG (F[ab]) + + IgG (F[ab’]2) + + IgG (Fc) + + IgG1 + + IgG2 - - IgG3 - - IgG4 - - IgAb + + IgM - - IgE - - aReactivity to B. henselae antigen with multiple bands; bIgA alpha and secretory chain- specific activity

Seroreactivity to Bartonella spp. among various cohorts of the Swedish human and animal population (II-VI) Blood donors (II, III, VI) Various studies over the past two decades have documented seroprevalence to Bartonella spp. in healthy adult and adolescent and child populations, including blood donors, in countries spanning Europe (See Table 4). Prior to the initiation of the present studies, however, the seroprevalence and mere existence of Bartonella spp. in Sweden had not been evaluated. In an endeavor to assess the level of endemicity of antibodies to Bartonella spp. in the general population in Sweden and establish a baseline for comparison with other subsets of the population, a series of blood donor serum samples were serologically tested for reactivity against various Bartonella spp. antigens. Initially, a collection of 100 blood donor serum samples were analyzed for evidence of reactivity to 3 pathogenic species of Bartonella: B. henselae, B. quintana, and B. elizabethae (Paper II). An overall seropositivity rate of 4% in this preliminary study indicated a low baseline exposure to Bartonella spp. in healthy individuals in Sweden. Seroreactivity to B. elizabethae antigen was evident in all 4 of the positive samples, with one sample reacting concomitantly against B. henselae. None of the blood donors evidenced antibodies to B. quintana. An additional 222 blood donor serum samples were added to the initial 100 samples for a total of 322 blood donors in Paper III. In this sample set, 6.8% (22/322) of the blood donors manifested antibodies to at least one of the 3 Bartonella species tested with

46 seroprevalences of 6.8%, 1.6%, and 0.3% to B. elizabethae, B. henselae, and B. quintana, respectively (See Table 10b). Seroreactivity to B. elizabethae was the most prevalent finding, as all of the 22 seropositive samples included reactivity to this agent. One sample (0.3%) was seropositive to all 3 antigens, including B. quintana, with four additional sera reactive against both B. elizabethae and B. henselae. Geometric mean titers (GMT) for all antigens were low, with values of 75 for B. elizabethae, 84 for B. henselae, and 64 for B. quintana. As the blood donor sample groups in Papers II and III were obtained locally in the towns of Uppsala and Gävle, Sweden, respectively, a larger scale study was commenced in order to encompass a more diverse geographical area and enroll a larger number of individuals (Paper VI). Serological analysis of the 498 serum samples collected from the 5 regions of Boden, Jönköping, Lund, Skövde, and Uppsala, revealed a total Bartonella spp. seropositivity rate of 16.1%. Immunogobulin G antibody seroprevalences to the six Bartonella antigens tested, B. elizabethae, B. grahamii, B. henselae (Houston-1), B. henselae (Marseille), B. quintana, B. vinsonii subsp. vinsonii, were 13.9%, 2.4%, 1.2%, 2.0%, 0.2% and 0.0% respectively. Refer to Table 10a for a summary of the reactivities for each antigen as well as cross-reactions with different antigen combinations. Among the seropositive samples, GMT values for each Bartonella antigen were as follows (reported as reciprocal values): B. elizabethae, 127.0; B. grahamii, 121.0; B. henselae (Houston-1), 90.5; B. henselae (Marseille), 73.5; and B. quintana, 64.0. None of the blood donor sera reacted to B. vinsonii subsp. vinsonii. This study not only spanned a wider geographical distribution and included a larger sample size, but incorporated a greater number of Bartonella spp. antigens to which the sera were tested against as well. Seroprevalence to three of the Bartonella antigens, B. grahamii, B. henselae (Marseille), and B. vinsonii, subsp. vinsonii, had never been evaluated previously in blood donors to our knowledge and, in fact, currently remain undocumented in blood donors outside of Sweden. Consistent with the findings of the initial blood donor assays, seroreactivity to B. elizabethae was the most prominent finding, with low seroreactivities to B. henselae and B. quintana. A similar pattern of high B. elizabethae seroprevalence compared to low seroprevalences to B. henselae and B. quintana was found in the U.S. as well (63, 64), and is discussed in further detail below. Likewise, a recent study in Sweden (87) analyzing the serological response of 61 blood donors to Bartonella spp. antigen as a control reported nearly equivalent findings, with seroprovalences of 14.8%, 4.9%, 1.6%, and 1.6% to B. elizabethae, B. grahamii, B. henselae, and B. quintana, respectively, as compared to our values of 13.9%, 2.4%, 1.2%, and 0.2% against the same antigens. B. grahamii had the second highest rate of seroreactivity following B. elizabethae, with 10% of the seropositive samples reacting solely to this agent. No reports outside of Sweden have documented seroreactivity to B.

47 grahamii in blood donors or otherwise healthy individuals. Interestingly, B. grahamii was recently isolated as the most frequent species infecting rodents in central Sweden (131). The finding of a low seroprevalence of 1.2% to B. henselae in the current study is lower than many of the findings in “warmer” countries in Europe, with seroprevalences ranging from 5.9%-25%, yet is consistent with the reported seroprevalences in countries of closer proximity with similar climates, such as the 1.5% seroprevalence in the U.K. and 0.6% seroprevalence in Denmark (122, 268). Similarly, the nearly negligible B. quintana seroprevalence found among blood donors in Sweden is mirrored by the comparably low seroprevalences of 0% and 0-2% reported in Denmark and France, respectively (41, 119, 268).

Patients (II) At the onset of the investigations into the existence of Bartonella spp. infection in Sweden, only two species of Bartonella had been previously isolated from patients in Europe, B. henselae and B. quintana (24, 42, 78, 141, 247). Serological evidence of Bartonella spp. antibodies had also been documented Germany, Spain, France, Italy, and Denmark (33, 35, 41, 74, 249, 265). In order to assess the occurrence of Bartonella infection in the Swedish patient population, a subset of 126 serum samples collected from 109 patients were serologically analyzed by IFA for antibodies to B. henselae, B. quintana and B. elizabethae. In total, 11% (14/126) of the patient serum samples were seropositive to one or more of the Bartonella spp. tested, corresponding to 8.3% (9/109) of the total patient group as multiple serum samples were available for several of the patients. The seropositivity rate in this patient cohort was significantly higher than that found in the blood donor control group (p < 0.02). As in the blood donors, antibodies against B. elizabethae were the most prevalent finding, often in the absence of concomitant seroreactivity to B. henselae or B. quintana. For further discussion regarding the increased seropositivity to this antigen, see below “Significance of high B. elizabethae seroprevalence.” Specified clinical histories for the seropositive patients included lymphadenopathy (2/9), granulomatous hepatitis (3/9), Wegener’s granulomatosis (2/9), and arm paralysis secondary to a cat scratch (1/9). Data available on the clinical histories of the seronegative patients revealed that 26% had lymphadenopathy, 7% had hepatitis, 5% had symptoms affecting the central nervous system, 3% had Wegener’s granulomatosis, 3% had heart involvement, and 1% had fever of unknown origin. Indicating a recent and/or active infection, 3 of the 5 seropositive patients for whom paired or multiple serum samples were available demonstrated a fourfold or greater increase in IFA titer from the acute to convalescent phase of infection to at least one of 3 Bartonella spp. tested. See Table 9 for a listing of patient characteristics and results of serology and PCR in these

48 patients. Tissue samples were available for analysis from two or these patients (Patients 3 and 4). In both cases, sequence analysis of the PCR product identified a strain of Bartonella with 99.7% similarity to B. quintana and 85% similarity to B. elizabethae. The successful amplification of Bartonella spp. from these two patients substantiated for the first time the presence of Bartonella spp. infection in Sweden and in a country in Europe other than France, Finland, or the Netherlands. Additionally, the amplification of Bartonella spp. from Patient 4 represented one of the first associations of myocarditis as a clinical manifestation of Bartonella infection; only one report had been previously documented of a patient, an immunocompetent adult, with B. henselae endocarditis who manifested associative signs of myocarditis (132). At present, myocarditis is now recognized as a documented clinical manifestation of Bartonella spp. infection in not only humans, but cats and dogs as well (see below “Elite orienteers” for further discussion). Of interest as well was the finding of seroreactivity to Bartonella spp. among two patients with documented Wegener’s granulomatosis, a disorder for which the etiology currently remains unknown. Further studies are warranted to investigate a potential link between this disease and infection with Bartonella spp.

TABLE 9. Four patient case reports with four-fold increases in Bartonella spp. IFA titers and/or PCR positivity to Bartonella. Patient Age Diagnosis Date of IFA titer1 to: PCR product Animal no. serum B. h2 B. e3 B. q4 contact sample 1 18 Wegener’s 03/10/95 128 128 256 NP5 Cat granuloma 03/28/95 128 512 256 05/17/95 <64 128 64 01/09/97 <64 64 <64 05/12/97 <64 <64 <64 05/18/98 <64 <64 <64 2 17 Submandi- 05/05/97 128 <64 64 NP Cat bular ulcer 06/16/97 128 128 128 3 73 Hepatic 06/18/97 <64 <64 <64 B. quintana- Mice, granuloma 10/09/97 <64 128 <64 like6 rats 05/10/98 <64 512 <64 4 59 Myocarditis 04/03/94 <64 2,048 128 B. quintana- like 1 Titers reported as reciprocal values of serum IFA endpoint dilutions; 2B. h, B. henselae; 3B.e, B. elizabethae; 4B. q, B. quintana; 5NP, not performed; 6B. quintana-like, 99.7% sequence alignment similarity to B. quintana Fuller strain and 85% similarity to B. elizabethae

Elite orienteers (III) The finding of an increased occurrence of sudden unexpected cardiac death (SUCD) among elite orienteers in Sweden and the subsequent finding of

49 Bartonella spp., both B. henselae and B. quintana, in the myocardial and lung tissues of several orienteers who succumbed to SUCD (130, 298) led to a large-scale screening of ranked elite orienteer sera for serological evidence of antibodies to Bartonella spp. Results of this screening indicated a significantly high Bartonella seroprevalence (p<0.001) among this cohort in Sweden, with 355 of the 1136 (31%) orienteer serum samples assayed in the study manifesting IgG antibodies to at least 1 of the 3 Bartonella agents tested: B. elizabethae, B. henselae, and B. quintana. Total Bartonella spp. seroreactivity in the blood donor control group, as discussed above, was 6.8%. Reciprocal titers of the orienteer sera ranged from 64-512, with GMT values of 103, 92, and 80 for the aforementioned antigens, respectively. The Bartonella species eliciting the greatest seroreactivity, as in the preceding studies, was B. elizabethae; the majority (88%) of the 355 seropositive orienteer sera assayed manifested IgG antibodies solely against this antigen. See below for further discussion of this finding. Although seroprevalences of antibodies to B. henselae and B. quintana were substantially lower (3.0% and 1.4%, respectively), they were still slightly higher than those found in the blood donor sample group, for which obtained seroprevalences were 1.6% and 0.3%, respectively, for these two antigens. In the serologic analyses of suspected or confirmed cat-scratch disease (CSD) patient sera, cross-reactions between B. henselae and B. quintana are reported in up to 94-100% of cases (69, 83, 208, 242, 247, 265, 280, 291); however, no reactions solely between these two antigens occurred in this study, though 6/1136 (0.5%) samples reacted to all 3 antigens concomitantly (discussed further below). In the investigation into the SUCD of the elite orienteers during the years 1992-1993, histopathological findings at autopsy implicated myocarditis and associated fatty infiltration mimicking arrhythmogenic right ventricular cardiomyopathy (ARVC) as the cause of death in the majority of cases (178, 298). As mentioned above, a Bartonella spp. with 99.7% similarity to B. quintana was found in a patient with myocarditis (130). This patient, in fact, was a former elite orienteer. Subsequently, Bartonella spp. were found in cardiac and various other tissue samples of five elite orienteers succumbing to SUCD, all of whom were diagnosed at autopsy with a form of cardiac dysfunction: myocarditis (4/5) or pericarditis (1/5). B. henselae was isolated from three of the cases, and a B. quintana-like spp. was isolated from the remaining two. Since these findings, in addition to a non-orienteer case of Bartonella- associated ARVC (85), numerous reports have surfaced documenting additional associations between Bartonella spp., primarily B. henselae and B. quintana, and cases of myocarditis as well as ARVC in not only humans (103, 215, 221, 237, 271), but dogs (37, 282), cats (163, 287), and mice (62), as well. Among the elite orienteers, the intense and rigorous daily training regimens characteristic of the sport in which the athletes would often compete even when ill may have induced a state of depressed immune

50 function (104), which is correlative with an exacerbated course of Bartonella infection in which even latent, silent, and subacute infections can become severe in immunocompromised individuals (257). The corresponding finding of an increased seroprevalence of antibodies to Bartonella spp. in the cohort of elite orienteers is suggestive of an increased exposure among this group. As Bartonella infections are zoonoses associated with transmission via primarily rodents, arthropod vectors, cats, and several other mammals, the close contact with nature characteristic of the sport of orienteering is a potential source of zoonotic/ vector-borne transmission of the agents. This intimate contact with nature unique to orienteering may account for the fact that the increase in SUCD cases has been limited to elite orienteering and not found in association with any other endurance sports in Sweden (297). Recent studies have in fact isolated Bartonella spp., predominantly B. grahamii, from wild rodents (131), as well as B. henselae from Swedish cats (231), yet failed to isolate Bartonella from Ixodes ricinis ticks (174). As elite orienteers frequently compete in other countries, foreign acquisition is a possibility as well. The potential mode of transmission to orienteers and the existence of a sylvatic transmission cycle remains uncertain; further studies are needed in order to elucidate the exact source of exposure.

Intravenous drug users (IVDU) (V) On account of the behavioral, environmental, and immunological characteristics associated with intravenous drug use, IVDUs are at an increased risk for acquisition of blood-borne, non-blood-borne, and zoonotic pathogens. As a predominance of antibodies to B. elizabethae among various subsets of the human population in Sweden was found in prior studies, and a paralled high rate of reactivity to B. elizabethae was found in two different studies analyzing urban IVDUs in the United States in Baltimore, Maryland (63) and New York City (NYC) (64), a study was commenced in Sweden to investigate the degree of seroreactivity to Bartonella spp. among Swedish IVDUs. Serum samples obtained at autopsy from a total of 59 drug users whose death was attributed directly to the administration of heroin were tested by IFA for antibodies against 6 strains of Bartonella, 5 of which are recognized as pathogenic. An overall Bartonella spp. seropositivity rate of 39% was found among the forensic samples, with resulting seroprevalence to each specific antigen as follows: B. elizabethae, 39% (23/59); B. grahamii, 3% (2/59); B. henselae (Houston-1), 14% (8/59); B. quintana, 3% (2/59); no seroreactivity against B. henselae (Marseille) or B. vinsonii subsp. vinsonii was evident. This seroreactivity rate was considerably higher than that found among a group of 44 non- IVDU autopsy sera (21%). More than half of the seropositive IVDU sera (14/23; 23.7% of all 59 sera) reacted to only one Bartonella antigen, B. elizabethae in all cases. In fact, all of the positive reactions in both the

51 IVDU and control groups included seroreactivity against B. elizabethae antigen (See Table 10a for a complete distribution of antigen reactivities). Comparably high prevalences of antibodies against B. elizabethae, 33% and 46%, were found in inner-city IVDUs in the U.S. studies in Baltimore and NYC, respectively, with titers ranging up to 1:512 (63, 64). Overall seroreactivities to Bartonella spp. in these two studies were considerably elevated as well, with nearly half of the IVDU sera (47.5%) in the NYC study reacting with one or more Bartonella antigens and 37% of the Baltimore sera exhibiting anti-Bartonella spp. antibodies (63, 64). The obtained seroprevalences of B. elizabethae antibodies among IVDUs in our Swedish cases and those in the U.S. are significantly high when compared to background seroprevalences among blood donors of 15.1% and 6.8-14.8% in the U.S. (JL Cooper, written communication, via Comer 1996 [63]) and Sweden (87, 211, 213), respectively. To our knowledge, no other studies have included B. elizabethae in the analyses of Bartonella seroprevalence among IVDUs. The high reactivity to B. elizabethae antigen is discussed further below. In addition to B. elizabethae, elevated seroreactivity among Swedish IVDUs was found against B. henselae (Houston-1) and B. quintana antigens as well. With background seroprevalences of antibodies against these two agents in Swedish blood donors reported as 1.0-1.6% and 0-0.3% for B. henselae and B. quintana (87, 211, 213), respectively, the finding of 14% and 3% seroprevalences for these two antigens, respectively, represented significantly elevated seroprevalence levels among this cohort of individuals. Studies in several other countries have reported similar elevated rates of seroreactivity to B. henselae and B. quintana among IVDUs. In Spain, total seroreactivity to a combination of these two antigens was found to be 14% (245); in Slovenia, seroprevalence to B. henselae and B. quintana was reported as 37.5-60% and 27.5-47.5%, respectively, compared to 25.0% and 22.5% seroprevalence, respectively, among control groups (312); and in the two studies in the U.S., high seroprevalence for B. henselae was detected in Baltimore (11%) and NYC (10%), as well as for B. quintana, with rates of 10% in Baltimore and 3% in NYC (63, 64). In contrast, a recent study in Poland reported a 0% seropositivity rate to both B. henselae and B. quintana antigens among homeless IVDUs (54). A recent study in Sweden of homeless individuals, 46% of whom were IVDUs, found a total seroprevalence rate to any Bartonella spp. of 46%, with 29.2% of the case sera harboring antibodies to B. henselae and 4.2% to B. quintana, compared to a 1.6% seropositivity rate among controls to both antigens (87). In addition, more than half (52.1%) of the homeless individuals in Sweden in this particular study manifested antibodies to B. elizabethae. As nearly half of the homeless individuals in this study were IVDUs, suggesting that the homeless and IVDU populations overlap to a significant degree, and comparable seroreactivities to Bartonella spp. were obtained, the supposition

52 that these two groups are exposed to similar environmental factors and thereby exposed to similar potential vectors of transmission of the Bartonella agents is suggested.

Cats (IV) The cat has been shown to serve as a natural reservoir for harboring several species of Bartonella, directly transmitting the organisms to humans. As initial studies demonstrated an increased seroprevalence to Bartonella spp., particularly B. elizabethae, among certain cohorts of the Swedish population, the prevalence of IgG antibodies to pathogenic Bartonella spp. associated with human infections, namely B. henselae, B. quintana, and B. elizabethae, was investigated in the cat population in Sweden. A total of 292 cat sera were analyzed by means of IFA for evidence of antibodies against the aforementioned 3 species of Bartonella. Results indicated a 25%, 1% and 0% sereprevalence rate for B. elizabethae, B. henselae, and B. quintana, respectively. Positive antibody titers ranged between 64 and 512 for B. elizabethae, whereas titers for B. henselae seropositive cat sera (3/292) were no greater than 64 and included cross-reactions with B. elizabethae (See Table 10a). The finding of a high seroprevalence of B. elizabethae antibodies in cats is consistent with the paralleled findings of high seroprevalences of antibodies seen among human populations in Sweden in previous investigations. No other studies have reported on seroprevalence of B. elizabethae in cats, providing no means for comparison with results from other cat populations. Studies in the human population, as discussed above, have shown seroprevalences in populations outside of Sweden ranging from 15% to 46% among blood donors, patient samples submitted for diagnostic testing, and intravenous drug users (63, 64, 69). Seroprevalences in the human population in Sweden have been comparable, with a range of 4% to 39% in equivalent population groups. The seroprevalence of B. henselae antibodies among cats, however, has been extensively studied in various different regions of the world (See Table 5). The finding of a 1% seroprevalence to B. henselae in our study in Sweden was considerably lower than seroprevalences found in other regions of the world, which range up to 93% (229) (See Table 5). However, trends in varying seroprevalences have indicated that B. henselae seropositivity tends to correlate directly with climate fluctuations, as evidenced in the high seroprevalences found in warm regions such as Spain (71.4%), Chile (85.6%) and Hawaii (73.7%), compared to the lower seroprevalences found in cooler regions such as Norway (0-1%), Alaska (5%), and the Rocky Mountain/Great Plains region of the U.S. (3.7%), which has climates reaching Arctic levels due to its high elevation (21, 101, 142, 277). Thus, the predominantly cool climate of Sweden is likely a contributing factor to the low seroprevalence of B. henselae antibodies in Swedish cats, corroborating the findings as well of the

53 0-1% B. henselae seroprevalence among cats in Sweden’s neighboring country Norway (21). Additionally, B. henselae seroprevalence has been found to be associated with cat fleas, which act as potential vectors of transmission of the pathogens (56, 105, 223, 284). As the levels of cat fleas in Sweden are relatively low, the diminished source of potential vector transmission from cat to cat may contribute to the low seroprevalence of B. henselae among the feline population in Sweden as well.

Table 10a. Summary and comparison of serologic cross-reactions and specific reactivity to B. henselae, B. quintana, and B. elizabethae among patients, elite orienteers, cats and IVD users in Sweden Antigen Patients (II) % Elite % Cats % IVD % Orienteers (IV) Users (III) (V) Any1 14/126 11.1 355/1136 31.3 73/292 25.0 23/59 39.0 Bh2, total 34/1136 3.0 3/292 1.0 8/59 13.6 B e3, total 350/1136 30.8 73/292 25.0 23/59 39.0 B q4, total 16/1136 1.4 0/292 0.0 2/59 3.4 B h, only 2/1136 0.2 0/292 0.0 0/59 0.0 B e, only 312/1136 27.5 70/292 24.0 14/59 23.7 Bq, only 3/1136 0.3 0/292 0.0 0/59 0.0 Bh + Be 25/1136 2.2 3/292 1.0 7/59 11.9 Bh + Bq 0/1136 0.0 0/292 0.0 0/59 0.0 Be + Bq 6/1136 0.5 0/292 0.0 1/59 1.7 All 3 spp. 7/1136 0.6 0/292 0.0 1/59 1.7 No spp. 112/126 88.9 781/1136 68.8 219/292 75.0 36/59 61.0 1Any one of B. henselae, B. elizabethae, or B. quintana antigen; 2Bh, B. henselae; 3Be, B. elizabethae; 4Bq, B. quintana

Table 10b. Summary and comparison of serologic cross-reactions and specific reactivity to B. henselae, B. quintana, and B. elizabethae among Swedish blood donors and control sera Antigen Blood % Blood % Non- % Blood % donors donors IVDU donors (II) (III) (V) (VI) Any1 4/100 4.0 22/322 6.8 9/44 20.5 80/498 16.1 Bh2, total 1/100 1.0 5/322 1.6 1/44 2.3 6/498 1.2 Be3, total 4/100 4.0 22/322 6.8 9/44 20.5 69/498 13.9 Bq4, total 0/100 0.0 1/322 0.3 1/44 2.3 1/498 0.2 Bh, only 0/100 0.0 0/322 0.0 0/44 0.0 0/498 0.0 Be, only 3/100 3.0 16/322 5.0 8/44 18.2 55/498 11.0 Bq, only 0/100 0.0 0/322 0.0 0/44 0.0 0/498 0.0 Bh + Be 1/100 1.0 5/322 1.6 0/44 0.0 4/498 0.8 Bh + Bq 0/100 0.0 0/322 0.0 0/44 0.0 0/498 0.0 Be + Bq 0/100 0.0 0/322 0.0 0/44 0.0 0/498 0.0 All 3 spp. 0/100 0.0 1/322 0.3 1/44 2.3 1/498 0.2 No spp. 96/100 96.0 300/322 93.2 35/44 79.5 418/498 83.9 1Any one of B. henselae, B. elizabethae, or B. quintana antigen; 2Bh, B. henselae; 3Be, B. elizabethae; 4Bq, B. quintana

54 Demographic and epidemiologic findings (III, IV, V, VI) Various data were analyzed in the current studies regarding demographic and epidemiologic variables in an attempt to identify risk factors for infection with Bartonella spp. in Sweden among different cohorts of the population. No significant associations between gender or reported clinical symptoms and Bartonella spp. seropositivity were found among any of the human or cat populations. However, although no significant differences in antibody prevalence were found with regard to age among the human population studies, a significant association between Bartonella spp. seropositivity and increasing age was found in the cat population; older cats (> 2 y old) were found to harbor higher levels of antibodies to Bartonella spp., specifically B. elizabethae, than younger cats, with seroprevalences of 32% and 19%, respectively. Previous studies have reported an inverse relationship with regard to B. henselae, in that younger cats less than 1 year of age have been found to exhibit a greater degree of B. henselae bacteremia than older cats (1, 55). No comparative studies involving B. elizabethae are currently available, and the premise of this relationship is unclear. Among the cohort of orienteers, no association with current pet ownership and Bartonella spp. seropositivity was found. However, among blood donors throughout the country of Sweden, a significant association was demonstrated between seropositivity to B. elizabethae antigen at a CT titer of 1:128 and current or past cat contact. The difference was not significant at a CT of 1:64. As discussed above, 25% of cats among a study population in Sweden were found to harbor antibodies to B. elizabethae (Paper IV). Though a significant association between cats and B. henselae has been established for many years, with cats serving as a natural reservoir for the pathogens and responsible for transmission of the agents to humans, B. elizabethae has not to date been isolated from either cats or cat fleas. See below for additional discussion. Significant risk factors for B. elizabethae seropositivity also included working outdoors and frequent exposure to the wilderness, including hunting moose. Though this correlation may initially be deemed unsurprising due to the abundance of potential sources of Bartonella spp. found in nature, B. elizabethae has yet to be isolated from any human or animal in Sweden. Other Bartonellae, however, have been isolated from animals in Sweden in a limited number of studies, including B. grahamii, B. birtlesii, and four other novel genotypes from small woodland mammals (131), as well as B. henselae from cats (231). Future studies are needed to identify a source of B. elizabethae or cross-reacting antigen (discussed below) in Sweden to elucidate the findings of high levels of antibodies to this antigen and its potential connection to the outdoors. As opportunistic pathogens among immunocompromised individuals, particularly those with HIV and AIDS, Bartonella spp. typically present with exacerbated clinical manifestations in such individuals (59, 252, 255, 302).

55 Among the 59 IVDUs analyzed in our study, 41% (24/59) were HIV- positive. However, no significant differences were found in the seropositivies of the HIV-positive and HIV-negative subjects in our study. Interestingly, in the parallel study in the U.S. by Comer et al. (63), HIV seronegative status was found to be associated with Bartonella seropositivity. Further, though several species of Bartonella are known to serve as etiologic agents of endocarditis, no evidence of endocarditis was observed in any of our seropositive subjects.

Future prospects and the significance of high B. elizabethae seroprevalence The results of the current studies have indicated a relatively and universally high rate of seroreactivity to one Bartonella antigen in particular among cohorts of both human and cat populations- B. elizabethae. The findings of high prevalences of antibodies against this agent are particularly interesting. As mentioned previously, B. elizabethae has only been isolated on one occasion two decades ago from a human with endocarditis, in addition to the isolation of two genotypes with a high degree of similarity to B. elizabethae from two febrile patients in Thailand (170). Largely on account of its minimal association with human disease, knowledge regarding its pathogenicity, transmission, and endemicity is lacking and remains to be elucidated. In fact, nearly half of the publications available to date regarding B. elizabethae were published following completion of the current studies and 90% have been published merely in the past 15 years. It is known, however, that small rodents serve as non-human vertebrate reservoirs for the pathogen, as B. elizabethae or phylogenically similar organisms have been reportedly isolated from rodents, primarily of the genus Rattus, and other small mammals (10, 12, 30, 71, 89, 113, 123, 135, 137, 151, 222, 283, 301, 304, 306). Additionally, Norway rats (Rattus norvegicus) have been shown to harbor antibodies against B. elizabethae at a seropositivity rate of up to 39% (84). In the recent Holmberg et al. (131) investigation, Bartonellae were in fact recovered from small mammals in Sweden, with B. grahamii being the most prevalent. Notably, however, no species of the genus Rattus were included in the study. As rats are known to harbor B. elizabethae and serve as reservoir hosts for the species, future additional studies on the potential presence of B. elizabethae in rats, particularly of the genus Rattus, in Sweden are warranted. Likewise, as B. elizabethae has also been detected in dogs and dog ticks in recent years in several different regions of the world, including Thailand (13), North Africa (149), and the U.S. (82, 216, 284), future studies investigating a potential link between B. elizabethae or a closely related genotype and dogs and their ectoparasites in Sweden are also

56 warranted in an endeavor to identify a source of B. elizabethae or cross- reacting antigen in Sweden and elucidate the findings of high levels of antibodies to this antigen. The finding of a high prevalence of B. grahamii in addition to a novel genotype genetically similar to and clustering with B. grahamii in rodents in Sweden is of interest in light of the phylogenetic relationship between B. grahamii and B. elizabethae and the propensity for closely related Bartonella spp. to cross-react on serologic assays, as discussed above. On account of the inherent serological cross-reactivity among Bartonella spp. in the IFA test used in our studies, the observed seropositivity to B. elizabethae antigen may represent a previous or current infection with another Bartonella spp. or antigen of unknown origin. Comparisons of phylogenetic information resulting from 16S rDNA, gltA- and groEL-based genotypic analyses consistently result in clusters formed by B. elizabethae, B. grahamii, B. tribocorum, and several unnamed species, primarily obtained from the genus Rattus (131, 133). Bartonella henselae and B. quintana also consistently cluster together in such parsimony trees and have been found to exhibit a significant degree of cross-reactivity by IFA of as great as 94-100% (69, 83, 208, 242, 247, 265, 280, 291). As B. grahamii antigen was only utilized in two studies and was found to cross-react with B. elizabethae antigen in only 17% (Paper V) and 3% (Paper VI) of cases, it is improbable that cross- reactivity with B. grahamii is responsible for the seropositivity of B. elizabethae by IFA in our studies; however, it is possible that an antigenic variant of B. grahamii in Sweden exists which cross-reacts with B. elizabethae by IFA. Given the phylogenetic proximity of B. elizabethae and B. tribocorum, both isolated from wild rats, in addition to the various unnamed species with which they cluster phylogenetically, the potential cross-reactivity of B. elizabethae with these agents by IFA should be investigated in future studies as well. Subsequent incorporation of cross- adsorption assays utilizing such species of closest phylogenetic relation to B. elizabethae may determine a potential source of cross-reactivity. Cross- adsorption assays were not utilized in the current studies as no apparent sources of cross-reactivity were evident; however, future studies may employ a broad spectrum analysis of the reactivity of B. elizabethae-positive sera against an array of heterologous antigens in addition to closely related Bartonella spp. in an endeavour to identify a potential source of cross- reactivity to account for the high degree of seroreactivity to B. elizabethae antigen in our studies. That two patients with positive titers to B. elizabethae were found to be PCR positive for B. quintana, despite low or even absent IFA titers to the agent, led to an initial speculation that an antigenic variant of B. quintana may alternatively exist in Sweden, sharing antigenic epitopes with the B. elizabethae strain used in the IFA. As with B. grahamii, cross-reactivity with B. quintana in our studies was minimal or absent, as the majority of

57 seropositive samples demonstrated antibodies solely against B. elizabethae (See Tables 10a and 10b). Other studies reporting on B. elizabethae seropositivities have also reported little serological cross-reactivity between this antigen and B. henselae and/or B. quintana (63, 64, 69, 230). The lack of seroreactivity to B. quintana in the presence of PCR amplification in the aforementioned patient cases may be the result of genetic heterogeneity affecting the surface antigens reacting in the serologic assay, as in the study by Sander et al. (264), which demonstrated the existence of B. henselae variants in different geographic regions in Germany. Further, the isolation of B. quintana or B. henselae from the above patients with concordant seropositivities to B. elizabethae does not preclude the possibility of a co- infection or prior asymptomatic infection with B. elizabethae or cross- reacting agent, in that the observed antibody positivity could be related to the persistence of convalescent antibodies. Alternatively, an “asymptomatic antibody carrier status” may have been established in our B. elizabethae seropositive subjects, as has been demonstrated in various studies in which family members of patients with CSD have a significantly higher level of antibodies against Bartonella spp. than controls (75, 224, 307). Though the route of transmission of Bartonella spp. to the subjects in our studies is undefined, it appears evident that exposure to the wilderness and settings inhabited by wild animals and their ectoparasites, commonplace to Swedes in general, and particularly in the case of orienteers, IVDUs, who exhibit frequent sociologic overlap with the homeless in Sweden (87), cats, and blood donors of self-reported high exposure to the wilderness, is likely a significant contributing factor. In addition to the pool of potential rodent reservoirs, B. elizabethae has been isolated from mouse and rat fleas (71, 198), as well as dog ticks, as mentioned above, thereby substantiating their potential as vectors of the pathogen. Future studies investigating the presence of Bartonella spp. in rodent fleas and dog ticks, in addition to the potential reservoirs as mentioned above, may provide clues to the means of exposure and transmission of this agent to animal and human subjects in Sweden, leading to an increased level of antibodies as evidenced in the current studies.

58 GENERAL CONCLUSIONS

Prominent human immunoglobulin (Ig) isotype-specific and IgG subclass-specific reactivity in the human humoral immune response to B. henselae infection appears to be predominated by IgG1 and IgA (both secretory and alpha chains), as assessed by Western blot analysis.

An 83-kDa immunodominant protein of B. henselae (Bh83) appears to be immunoreactive exclusively with Bartonella antigen and a conserved epitope during human infection.

Bartonella spp. infection and seropositivity is present in Sweden in both the patient and general population, including Bartonella spp. isolated from Swedish patients with fatal myocarditis.

The cat population in Sweden demonstrates a relatively high prevalence of antibodies directed against B. elizabethae antigen, with a concordant low seroreactivity to B. henselae antigen.

The highest currently demonstrated rate of Bartonella spp. seroreactivity in Sweden is to B. elizabethae, with relatively low or no rates of seroreactivity to B. henselae, B. quintana, B. grahamii, and B. vinsonii.

A high prevalence of antibodies to Bartonella organisms, particularly B. elizabethae, among Swedish elite orienteers suggests a widespread exposure to infection within this cohort.

Forensic cases of intravenous heroin users in inner-city Stockholm demonstrate a high seropositivity rate to Bartonella spp., comparable to similar findings in the U.S., with no evidenced correlation to HIV status.

Potential risk factors for Bartonella spp. infection in Sweden include working outdoors, hunting moose, activity in the wilderness, cat contact, and travel to Eastern Europe.

59 ACKNOWLEDGEMENTS

This whole experience has been one of a lifetime and I am grateful to countless individuals for helping me along the way. The story begins with Russell Regnery, truly an amazing person- an inspirational teacher, a pioneer in Bartonella research, and a dear friend. Russ was my supervisor at the CDC and taught me everything I needed to know to form a strong foundation in Bartonella serology. I enjoyed not only each day we worked together at the CDC, whether it be reading diagnostic slides, days out in the field capturing rodents suited up in our Biohazard Level 4 gear, or cultivating antigen, but also our time spent together in Uppsala when he would come for support and continuing collaboration. He shared my love for Sweden and Uppsala in particular; a true Viking at heart, really. Russ is one of the most helpful, generous, and pleasant people I know. I’ll never forget, also, the rest of the gang at the CDC- “the kiddie corp,” as Ted affectionately called us (Feda [aka “Roo”], Shady, Kim, and Eric), Kevin, Ted, Joe, John, Andy, Barbara, Jim, William, Rob, Jane, Chris, Luis… What a fun group we had! I cherish every memory with each of them and consider them not only co-workers, but lifelong friends. Special recognition is extended to Kevin Karem, my secondary supervisor at the CDC, for not only expanding my vocabulary by teaching me necessary phrases such as “that dog’ll hunt”, “waffling”, and “I’m gonna open up a can of whoopass (or maybe that was Kim!), but also for all of his assistance and guidance with our countless attempts to isolate and sequence Bh83. Then some visitors arrived at the CDC- two Swedish scientists named Calle Påhlson and Martin Holmberg. Little did I know that my life was about to change. As chief serologist for Bartonella diagnostics at the time, I had the opportunity to collaborate with these researchers and unknowingly embark on a journey that would mark the beginnings of Bartonella research in Sweden. I am forever indebted to these two individuals, as well as Göran Friman, my soon-to-be supervisor, for offering me the opportunity of a lifetime to move to Sweden to help establish a Bartonella serology program there at UAS. When they told me the news that I was invited to Uppsala to work with them, I literally jumped for joy, even higher than when my high school crush called to invite me to the prom! From the bottom of my heart, I am grateful to my initial team, Göran Friman, Eva Hjelm, Lars Wesslén, Christian Ehrenborg, and Martin Holmberg for all that they did for me. Not only did they provide me with everything I needed in the lab and offer guidance and assistance throughout the years, but they helped introduce me to Swedish culture, invited me to delicious home-cooked dinners, provided me with a few bikes to get around, arranged my living quarters, and just made me feel welcome and at home in Uppsala, my home away from home. I am particularly indebted to Martin Holmberg, my current supervisor, as well as Björn Olsen, for

60 helping to make this thesis a reality. Birgitta Sembrant was often behind the scenes ensuring that things were carried out, and was always so sweet and helpful. Lars Engstrand was like my pseudo-supervisor! Together we wrote a book chapter on Bartonella, played tennis, and collaborated on Bartonella monoclonal research “on the side.” He also indirectly helped to name my first daughter, Elsa! (one of his doctorand students had a kitten named Elsa). So even she thanks him for that! Working in the “Baktlab” would not have been the same without my fellow co-workers, the “Baktlab Babes:” Sandra, Maria K, Maria H, Helene, Sara, Ingegerd, and Caroline! I have so many fond memories, including our Luciatåg, particulary the one when I got to be Lucia and ended up looking like a unicorn from the melted wax that unknowingly dripped on my head and Sandra then had to iron out of my hair (!), our fredag fikas, our pipette tip contests, and our trips to Halmstad and Stockholm. I cannot say enough about my dear friend, Sandra (Hjalmarsson) Rodin (the fact that I dedicated this entire thesis to her, though, perhaps says it all!). Maria (Sjölund) Karlsson is another dear friend who even lived with me in Atlanta for several months who I cannot say enough about. Maria Held introduced me to the sport of orienteering and, among many other things, helped to teach me the intricacies of the Swedish language (i.e. differentiating between “Olle,” “Ola” and “Ulla” and that the Swedish word “finner” is in fact the Finnish airline!). I would also like to thank Christian Ehrenborg for his PCR expertise, Olga Svensson for her assistance with providing samples for our studies, Lars Wesslén and Martin Storm for their computer savyness, Johan Bring and Kirk Easley for statistical consultation, Marius Domeika for his friendship, and Olle Lindquist for his interesting conversations, taking me on numerous trips to SMI to work on Paper V, and allowing me to spend time with his friend Lennart Nilsson, a true Swedish icon (who I should in turn thank as well, especially for the “hushållspengar” he gave me as a gift!). I would like to express my gratitude especially to Martin Holmberg at UAS, Rob Massung at the CDC, and Nick Johnson at the Univ. of Pittsburgh for their in- depth review of this thesis. During my years in Uppsala, I made so many wonderful friends outside of the lab as well, especially my roommates, Anna and Elin, Melle-Pelle, Johannes (“Your Highness”), Emil, Anna S, Johan, Jakob, Anna E, Anna-Karin, Ulf, “QT1”, Bosse, Thomas, Peter, Marten, and Robert. Finally, I would not be the person I am today without my dear family. I am grateful for my parents, who pushed me my entire life to do always my best (and who came to the rescue with LOTS of babysitting during the writing and preparing of this thesis!), my parents-in-law for their wonderful support and help, my brothers Tommy, Brian, and Keith (each one of them my best friend), my beloved husband, Sam, for his unconditional love, my wonderful and beautiful 3 little ones: Elsa, Aiden, and Atticus (and their little sister on the way in the spring!)… You are my sunshine, my only sunshine, you make me happy when skies are gray, you never know dears how much I love you!!, and my belated grandfather, Sven, who would be so proud of me being here. He taught me my first Swedish words when I was young, a prayer he would say when he was young… Gud som haver barnen kär, Se till mig som liten är Vart jag mig I världen vänder Stor min lycka in Guds händer Lyckan kommer, lyckan går Den Gud älskar, lyckan får.

61 REFERENCES

[1] Al-Majali AM. Seroprevalence of and risk factors for Bartonella henselae and Bartonella quintana infections among pet cats in Jordan. Prev Vet Med 2004a;64(1):63-71. [2] Al-Majali AM, Al-Qudah KM. Seroprevalence of Bartonella henselae and Bartonella quintana infections in children from Central and Northern Jordan. Saudi Med J 2004b;25(11):1664-9. [3] Alsmark CM, Frank AC, Karlberg EO, Legault BA, Ardell DH, Canbäck B, Eriksson AS, Näslund AK, Handley SA, Huvet M, La Scola B, Holmberg M, Andersson SG. The louse-borne human pathogen Bartonella quintana is a genomic derivative of the zoonotic agent Bartonella henselae. Proc Natl Acad Sci U S A 2004;101(26):9716-21. [4] Amerein MP, De Breil D, Jaulhac B, Meyer P, Monteil H, Piémont Y. Diagnostic value of the indirect immunofluorescence assay in cat scratch disease with Bartonella henselae and Afipia felis antigens. Clin Diagn Lab Immunol 1996;3:200-4. [5] Anderson B, Lu E, Jones D, Regnery R. Characterization of a 17-kilodalton antigen of Bartonella henselae reactive with sera from patients with cat scratch disease. J Clin Microbiol 1995;33:2358-65. [6] Anderson BE, Neuman MA. Bartonella spp. as emerging human pathogens. Clin Microbiol Rev 1997; 10:203-19. [7] Antoniou M, Economou I, Wang X, Psaroulaki A, Spyridaki I, Papadopoulos B, Christidou A, Tsafantakis E, Tselentis Y. Fourteen-year seroepidemiological study of zoonoses in a Greek village. Am J Trop Med Hyg 2002;66(1):80-5. [8] Armengol CE, Hendley JO. Cat-scratch disease encephalopathy: a cause of status epilepticus in school-aged children. J Paediatr 1999;134:635-8. [9] Avidor B, Graidy M, Efrat G, Leibowitz C, Shapira G, Schattner A, Zimhony O, Giladi M. Bartonella koehlerae, a new cat-associated agent of culture- negative human endocarditis. J Clin Microbiol 2004;42:3462-8. [10] Bai Y, Montgomery SP, Sheff KW, Cowdhury MA, Breiman RF, Kabeya H, Kosoy MY. Bartonella strains in small mammals from Dhaka, Bangladesh, related to Bartonella in America and Europe. Am J Trop Med Hyg 2007;77(3):567-70. [11] Bai Y, Kosoy M, Martin A, Ray C, Sheff K, Chalcraft L, Collinge SK. Characterization of Bartonella strains isolated from black-tailed prairie dogs (Cynomys ludovicianus). Vector Borne Zoonotic Dis 2008;8(1):1-5. [12] Bai Y, Kosoy MY, Lerdthusnee K, Peruski LF, Richardson JH. Prevalence and genetic heterogeneity of Bartonella strains cultured from rodents from 17 provinces in Thailand. Am J Trop Med Hyg 2009;81(5):811-6. [13] Bai Y, Kosoy MY, Boonmar S, Sawatwong P, Sangmaneedet S, Peruski LF. Enrichment culture and molecular identification of diverse Bartonella species in stray dogs. Vet Microbiol 2010;146(3-4):314-9.

62 [14] Barka NE, Hadfield T, Patnaik M, Schwartzman WA, Peter JB. EIA for detection of Rochalimaea henselae-reactive IgG, IgM, and IgA antibodies in patients with suspected cat-scratch disease. J Infect Dis 1993;167:1503-4. [15] Barnes A, Bell SC, Isherwood DR, Bennett M, Carter SD. Evidence of Bartonella henselae infection in cats and dogs in the United Kingdom. Vet Rec 2000;147(24):673-7. [16] Bass JW, Freitas BC, Freitas AD, Sisler CL, Chan DS, Vincent JM, Person DA, Claybaugh JR, Wittler RR, Weisse ME, Regnery RL, Slater LN. Prospective randomized double blind placebo-controlled evaluation of azithromycin for treatment of cat-scratch disease. Pediatr Infect Dis J 1998;17(6):447-52. [17] Bayliss DB, Steiner JM, Sucholdolski JS, Radecki SV, Brewer MM, Morris AK, Lappin MR. Serum feline pancreatic lipase immunoreactivity concentration and seroprevalences of antibodies against Toxoplasma gondii and Bartonella species in client-owned cats. J Feline Med Surg 2009;11(8):663-7. [18] Bennett AD, Gunn-Moore DA, Brewer M, Lappin MR. Prevalence of Bartonella species, haemoplasmas and Toxoplasma gondii in cats in Scotland. J Feline Med Surg 2011;13(8):553-7. [19] Benson LA, Kar S, McLaughlin G, Ihler GM. Entry of Bartonella bacilliformis into erythrocytes. Infect Immun 1986;54(2):347-53. [20] Berger TG, Koehler JE. Bacillary angiomatosis. AIDS Clin Rev 1993- 1994:43-60. [21] Bergh K, Bevanger L, Hanssen I, Løseth K. Low prevalence of Bartonella henselae infections in Norwegian domestic and feral cats. APMIS 2002;110(4):309-14. [22] Berglund EC, Frank AC, Calteau A, Pettersson OV, Granberg F, Eriksson A- S, Näslund K, Holmberg M, Lindroos H, Andersson SG. Run-off replication of host-adaptability genes is associated with gene transfer agents in the genome of mouse-infecting Bartonella grahamii. PLoS Genet 2009;5(7):e1000546. [23] Berglund EC, Ehrenborg C, Vinnere Pettersson O, Granberg F, Näslund K, Holmberg M, Andersson SG. Genome dynamics of Bartonella grahamii in micro-populations of woodland rodents. BMC Genomics 2010; 11:152. [24] Bergmans AM, Schellekens JF, van Embden JD, Schouls LM. Predominance of two Bartonella henselae variants among cat-scratch disease patients in the Netherlands. J Clin Microbiol 1996;34(2):254-60. [25] Bergmans AM, Peeters MF, Schellekens JF, Vos MC, Sabbe LJ, Ossewaarde JM, Verbakel H, Hooft HJ, Schouls LM. Pitfalls and fallacies of cat scratch disease serology: evaluation of Bartonella henselae-based indirect fluorescence assay and enzyme-linked immunoassay. J Clin Microbiol 1997a;35(8):1931-7. [26] Bergmans AM, de Jong CM, van Amerongen G, Schot CS, Schouls LM. Prevalence of Bartonella species in domestic cats in The Netherlands. J Clin Microbiol 1997b;35(9):2256-61. [27] Bermond D, Heller R, Barrat F, Delacour G, Dehio C, Alliot A, Monteil H, Chomel B, Boulouis HJ, Piémont Y. Bartonella birtlesii sp. nov., isolated from small mammals (Apodemus spp.). Int J Syst Evol Microbiol 2000;50(Pt 6):1973-9. [28] Bermond D, Boulois HJ, Heller R, Van Laere G, Monteil H, Chomel BB, Sander A, Dehio C, Piémont Y. Bartonella bovis sp. nov. and Bartonella

63 capreoli sp. nov., isolated from European ruminants. Int J Syst Evol Microbiol 2002;52:383-90. [29] Birtles RJ, Harrison TG, Saunders NA, Molyneux DH. Proposals to unify the genera Grahamella and Bartonella, with descriptions of Bartonella talpae comb. nov., Bartonella peromysci comb. nov., and three new species, Bartonella grahamii sp. nov., Bartonella taylorii sp. nov., and Bartonella doshiae sp. nov. Int J Syst Bacteriol 1995;45:1-8. [30] Birtles RJ, Raoult D. Comparison of partial citrate synthase gene (gltA) sequences for phylogenetic analysis of Bartonella species. Int J Syst Bacteriol 1996;46(4):891-7. [31] Birtles RJ, Canales J, Ventosilla P, Alvarez E, Guerra H, Llanos-Cuentas A, Raoult D, Doshi N, Harrison TG. Survey of Bartonella species infecting intradomicillary animals in the Huayllacallan Valley, Ancash, Peru, a region endemic for human bartonellosis. Am J Trop Med Hyg 1999;60(5):799-805. [32] Black JR, Herrington DA, Hadfield TL, Wear DJ, Margileth AM, Shigekawa B. Life-threatening cat scratch disease in an immunocompromised host. Arch Intern Med 1986;146:394-6. [33] Blanco Ramos JR, Oteo Revuelta JA, Martínez de Artola V, Ramalle Gómara E, García Pineda A, Ibarra Cucalón V. [Seroepidemiology of Bartonella henselae infection in a risk group]. Rev Clin Esp 1998;198(12):805-9. [34] Blanco JR, Oteo JA, Martínez V, Ramalle E, García A, Ibarra V, Rosel L. [Seroepidemiology of Bartonella henselae infection in HIV-infected patients]. Enferm Infec Microbiol Clin 1999;17(9):434-8. [35] Blomgren M, Hardt-Madsen M. [Cat-scratch disease- an overlooked disease in Denmark?]. Ugeskr Laeger 1997;159(19):2876-7. [36] Bown KJ, Bennet M, Begon M. Flea-borne Bartonella grahamii and Bartonella taylorii in bank voles. Emerg Infect Dis 2004;11:684-7. [37] Breitschwerdt EB, Atkins CE, Brown TT, Kordick DL, Snyder PS. Bartonella vinsonii subsp. berkhoffii and related members of the alpha subdivision of the Proteobacteria in dogs with cardiac arrhythmias, endocarditis, or myocarditis. J Clin Microbiol 1999;37(11):3618-26. [38] Breitschwerdt EB, Kordick DL. Bartonella infection in animals: carriership, reservoir potential, pathogenicity, and zoonotic potential for human infection. Clin Microbiol Rev 2000;13(3):428-38. [39] Brenner DJ, O’Connor S, Winkler HH, Steigerwalt AG. Proposals to unify the genera Bartonella and Rochalimaea, with descriptions of Bartonella quintana comb. nov., Bartonella vinsonii comb. nov. and to remove the family Bartonellaceae from the order Rickettsiales. Int J Syst Bacteriol 1993;43:777-86. [40] Brenner SA, Rooney JA, Manzewitsch P, Regnery RL. Isolation of Bartonella (Rochalimaea) henselae: effects of methods of blood collection and handling. J Clin Microbiol 1997;35(3):544-7. [41] Brouqui P, Houpikian P, Dupont HT, Toubiana P, Obadia Y, Lafay V, Raoult D. Survey of the seroprevalence of Bartonella quintana in homeless people. Clin Infect Dis 1996;23(4):756-9. [42] Brouqui P, La Scola B, Roux V, Raoult D. Chronic Bartonella quintana bacteremia in homeless patients. N Engl J Med 1999;340:184-9. [43] Brouqui P, Stein A, Dupont HT, Gallian P, Badiaga S, Rolain JM, Mege JL, La Scola B, Berbis P, Raoult D. Ectoparasitism and vector-borne diseases in 930 homeless people from Marseilles. Medicine (Baltimore) 2005;84(1):61- 8.

64 [44] Bruneel F, D’Estanque J, Fournier PE, Arlet G, Thuong M, Wolff M, Bedos JP, Lariven S, Regnier B. Isolated right-sided Bartonella quintana endocarditis in an immunocompetent adult. Scand J Infect Dis 1998;30:424- 5. [45] Carithers HA. Cat scratch disease: an overview based on a study of 1,200 patients. Am J Dis Child 1985;139:1124-33. [46] Case JB, Chomel B, Nicholson W, Foley JE. Serological survey of vector- borne zoonotic pathogens in pet cats and cats from animal shelters and feral colonies. J Feline Med Surg 2006;8(2):111-7. [47] Castle KT, Kosoy M, Lerdthusnee, K, Phelan L, Bai Y, Gage KL, Leepitakrat W, Monkanna T, Khlaimanee N, Chandranoi K, Jones JW, Coleman RE. Prevalence and diversity of Bartonella in rodents of northern Thailand: a comparison with Bartonella in rodents from southern China. Am J Trop Med Hyg 2004;70:429-33. [48] Celebi B, Kilic S, Aydin N, Tarhan G, Carhan A, Babur C. Investigation of Bartonella henselae in cats in Ankara, Turkey. Zoonoses Public Health 2009;56(4):169-75. [49] Chai CL, Lu QY, Sun JM, Jiang LP, Ling F, Zhang LJ, Zheng SG, Zhang H, Ge JH. [Sero-epidemiologic investigation on tick-borne diseases of humans and domestic animals in Zhejiang province]. Zhonghua Liu Xing Bing Xue Za Zhi 2010;31(10):1144-1147. [50] Chamberlin J, Laughlin LW, Romero S, Solorzano N, Gordon S, Andre RG, Pachas P, Friedman H, Ponce C, Watts D. Epidemiology of endemic Bartonella bacilliformis: a prospective cohort study in a Peruvian mountain valley community. J Infect Dis 2002;186:983-90. [51] Chang CC, Chomel BB, Kasten RW, Heller RM, Ueno H, Yamamoto K, Bleich VC, Pierce BM, Gonzales BJ, Swift PK, Boyce WM, Jang SS, Boulouis HJ, Piémont Y, Rossolini GM, Riccio ML, Cornaglia G, Pagani L, Lagatolla C, Selan L, Fontana R. Bartonella spp. isolated from wild and domestic ruminants in North America. Emerg Infect Dis 2000;6(3): 306-11. [52] Chang CC, Lee CC, Maruyama S, Lin JW, Pan MJ. Cat-scratch disease in veterinary-associated populations and in its cat reservoir in Taiwan. Vet Res 2006;37(4):565-77. [53] Childs JE, Olson JG, Wolf A, Cohen N, Fakile Y, Rooney JA, Bacellar F, Regnery RL. Prevalence of antibodies to Rochalimaea species (cat-scratch disease agent) in cats. Vet Rec 1995;136(20):519-20. [54] Chmielewski T, Podsiad y E, Tylewska-Wierzbanowska S. Presence of Bartonella spp. in various human populations. Pol J Microbiol 2007;56(1):33-8. [55] Chomel BB, Abbot RC, Kasten RW, Floyd-Hawkins KA, Kass PH, Glaser CA, Pedersen NC, Koehler JE. Bartonella henselae prevalence in domestic cats in California: risk factors and association between bacteremia and antibody titers. J Clin Microbiol 1995;33:2445-50. [56] Chomel BB, Kasten RW, Floyd-Hawkins K, Chi B, Yamamoto K, Roberts- Wilson J, Gurfield AN, Abbott RC, Pedersen NC, Koehler JE. Experimental transmission of Bartonella henselae by the cat flea. J Clin Microbiol 1996;34:1952-6. [57] Chomel BB, Mac Donald KA, Kasten RW, Chang CC, Wey AC, Foley JE, Thomas WP, Kittleson MD. Aortic valve endocarditis in a dog due to Bartonella clarridgeiae. J Clin Microbiol 2001;39(10):3548-54.

65 [58] Chomel BB, Boulouis HJ, Petersen H, Kasten RW, Yamamoto K, Chang CC, Gandoin C, Bouillin C, Hew CM. Prevalence of Bartonella infection in domestic cats in Denmark. Vet Res 2002;33(2):205-13. [59] Chomel BB, Boulouis HJ, Breitschwerdt EB. Cat scratch disease and other zoonotic Bartonella infections. J Am Vet Med Assoc 2004;224(8):1270-9. [60] Chung CY, Kasten RW, Paff SM, Van Horn BA, Vayssier-Taussat M, Boulouis HJ, Chomel BB. Bartonella spp. DNA associated with biting flies from California. Emerg Infect Dis 2004;10:1311–1313. [61] Cimolai N, Benoit L, Hill A, Lyons C. Bartonella henselae infection in British Columbia: evidence for an endemic disease among humans. Can J Microbiol 2000;46(10):908-12. [62] Colton L, Zeidner N, Lynch T, Kosoy MY. Human isolates of Bartonella tamiae induce pathology in experimentally inoculated immunocompetent mice. BMC Infect Dis 2010;10:229. [63] Comer JA, Flynn D, Regnery RL, Vlahov D, Childs JE. Antibodies to Bartonella species in inner-city intravenous drug users in Baltimore, Md. Arch Intern Med 1996;156:2491-5. [64] Comer JA, Diaz T, Vlahov D, Monterroso E, Childs JE. Evidence of rodent- associated Bartonella and Rickettsia infections among intravenous drug users form Central and East Harlem, New York City. Am J Trop Med Hyg 2001;65:855-60. [65] Conrad DA. Treatment of cat-scratch disease. Curr Opin Pediatr 2001;13(1):56-9. [66] Cooper MD, Hollingdale MR, Vinson JW, Costa J. A passive hemagglutination test for diagnosis of trench fever due to Rochalimaea quintana. J Infect Dis 1976;134:605-9. [67] Crissiuma A, Favacho A, Gershony L, Mendes-de-Almeida F, Gomes R, Mares-Guia A, Rozental T, Barreira J, Lemos E, Labarthe N. Prevalence of Bartonella species DNA and antibodies in cats (Felis catus) submitted to a spay/neuter program in Rio de Janeiro, Brazil. J Feline Med Surg 2011;13(2):149-51. [68] da Costa PS, Brigatte ME, Greco DB. Antibodies to Rickettsia rickettsii, Rickettsia typhi, Coxiella burnetii, Bartonella henselae, Bartonella quintana, and Ehrlichia chaffeensis among healthy population in Minas Gerais, Brazil. Mem Inst Oswaldo Cruz 2005;100(8):853-9. [69] Dalton MJ, Robinson LE, Cooper J, Regnery RL, Olson JG, Childs JE. Use of Bartonella antigens for serologic diagnosis of cat-scratch disease at a national referral center. Arch Intern Med 1995;155:1670-6. [70] Daly JS, Worthington MG, Brenner DJ, Moss CW, Hollis DG, Weyant AG, Steigerwalt AG, Weaver RE, Daneshvar MI, O’Connor SP. Rochalimaea elizabethae sp. nov. isolated from a patient with endocarditis. J Clin Microbiol 1993;31:872-81. [71] De Sousa R, Edouard-Fournier P, Santos-Silva M, Amaro F, Bacellar F, Raoult D. Molecular detection of Rickettsia felis, Rickettsia typhi and two genotypes closely related to Bartonella elizabethae. Am J Trop Med Hyg 2006;75(4):727-31. [72] Debre R, Lamy M, Jammet M, Costil L, Mozzconacci P. La maladie des griffes de chat. Semaine Des Hopitaux Paris 1950;26:1895-901. [73] Dehio D, Lanz D, Pohl R, Behrens P, Bermond D, Piémont Y, Pelz K, Sander A. Bartonella schoenbuchii sp. nov., isolated from the blood of wild roe deer. Int J Syst Evol Microbiol 2001;51(Pt 4):1557-65.

66 [74] Del Prete R, Fumarola D, Fumarola L, Basile V, Mosca A, Miragliotta G. Prevalence of antibodies to Bartonella henselae in patients with suspected cat scratch disease (CSD) in Italy. Eur J Epidemiol 1999;15(6):583-7. [75] Demers DM, Bass JW, Vincent JM, Person DA, Noyes DK, Staege CM, Samlaska CP, Lockwood NH, Regnery RL, Anderson BE. Cat-scratch disease in Hawaii: etiology and seroepidemiology. J Pediatr 1995;127(1):23- 6. [76] Dennis CA. Treatment of cat-scratch disease. Curr Opin Pediatr 2001;13(1):56-9. [77] Dörbecker C, Sander A, Oberle K, Schülin-Casonato T. In vitro susceptibility of Bartonella species to 17 antimicrobial compounds: comparison of Etest and agar dilution. J Antimicrob Chemother 2006;58(4):784-8. [78] Drancourt M, Mainardi JL, Brouqui P, Vandenesch F, Carta A, Lehnert F, Etienne J, Goldstein F, Acar J, Raoult R. Bartonella (Rochalimaea) quintana endocarditis in three homeless men. N Engl J Med 1995;332:419-23. [79] Dreier J, Vollmer T, Freytag CC, Bäumer D, Körfer R, Kleesiek K. Culture- negative infectious endocarditis caused by Bartonella spp.: 2 case reports and a review of the literature. Diagn Microbiol Infect Dis 2008;61(4):476-83. [80] Dubey JP, Bhatia CR, Lappin MR, Ferreira LR, Thorn A, Kwok OC. Seroprevalence of Toxoplasma gondii and Bartonella spp. antibodies in cats from Pennsylvania. J Parasitol 2009a;95(3):578-80. [81] Dubey JP, Lappin MR, Kwok OC, Mofya S, Chikweto A, Baffa A, Doherty D, Shakeri J, Macpherson CN, Sharma RN. Seroprevalence of Toxoplasma gondii and concurrent Bartonella spp., feline immunodeficiency virus, and feline leukemia virus infections in cats from Grenada, West Indies. J Parasitol 2009b;95(5):1129-33. [82] Duncan AW, Marr HS, Birkenheuer AJ, Maggi RG, Williams LE, Correa MT, Breitschwerdt EB. Bartonella DNA in the blood and lymph nodes of Golden Retrievers with lymphoma and in healthy controls. J Vet Intern Med 2008;22(1):89-95. [83] Dupon M, Delarclause AMS, Brouqui P, Drancourt M, Raoult D, Demascarel A, Lacut JY. Evaluation of serological response to Bartonella henselae, Bartonella quintana and Afipia felis antigens in 64 patients with suspected cat scratch disease. Scand J Infect Dis 1996;28:361-6. [84] Easterbrook JD, Kaplan JB, Vanasco NB, Reeves WK, Purcell RH, Kosoy MY, Glass GE, Watson J, Klein SL. A survey of zoonotic pathogens carried by Norway rats in Baltimore, Maryland, USA. Epidemiol Infect 2007; 135(7):1192-9. [85] Ehrenborg C, Wesslén L, Jakobson A, Friman G, Holmberg M. Sequence variation of the ftsZ gene of Bartonella henselae isolates and clinical samples. J Clin Microbiol 2000;38(2):682-7. [86] Ehrenborg C, Handley S, Ellis B, Mills J, Holmberg M. Bartonella grahamii infecting rodents display high genetic diversity over short geographic distances. Ann N Y Acad Sci 2003;990:233-5. [87] Ehrenborg C, Byström R, Hjelm E, Friman G, Holmberg M. High Bartonella spp. seroprevalence in a Swedish homeless population but no evidence of trench fever. Scand J Infect Dis 2008;40(3):208-15. [88] Ehrenborg C, Hagberg S, Alden J, Makitalo S, Myrdal G, Larsson E, Hjelm E, Friman G. First known case of Bartonella quintana endocarditis in Sweden. Scand J Infect Dis 2009;41(1):73-5.

67 [89] Ellis B, Regnery R, Beati L, Bacellar F, Rood M, Glass G, Marston E, Ksiazek TG, Jones D, Childs JE. Rats of the genus Rattus are reservoir hosts for pathogenic Bartonella species: an Old World origin for a New World disease. J Infect Dis 1999;180:220-4. [90] Engbaek K, Koch C. Immunoelectrophoretic characterization and cross- reactivity of Rochalimaea henselae, Rochalimaea quintana, and Afipia felis. APMIS 1994a;102:931-42. [91] Engbaek K, Koch C. Antibody response in rabbits infected with Rochalimaea henselae, Rochalimaea quintana, and Afipia felis. APMIS 1994b;102:943-49. [92] Engbaek K, Uttenthal LO, Koch C. Immunopurified extracellular Bartonella henselae antigen for detecting specific antibodies by enzyme immunoassay. APMIS 1997;105:941-50. [93] Engel P, Dehio C. Genomics of host-restricted pathogens of the genus bartonella. Genome Dyn 2009;6:158-69. [94] Engvall EO, Brändström B, Fermér C, Blomqvist G, Englund L. Prevalence of Bartonella henselae in young, healthy cats in Sweden. Vet Rec 2003;52(12):366-9. [95] Eremeeva ME, Gerns HL, Lydy SL, Goo JS, Ryan ET, Mathew SS, Ferraro MJ, Holden JM, Nicholson WL, Dasch GA, Koehler JE. Bacteremia, fever, and splenomegaly caused by a newly recognized Bartonella species. N Engl J Med 2007;356:2381-7. [96] Fabbi M, Vicari N, Tranquillo M, Pozzi C, Prati P, De Meneghi D, Bertoletti I, Lauzi S, Guiso P, Genchi C. [Prevalence of Bartonella henselae in stray and domestic cats in different Italian areas: evaluation of the potential risk of transmission of Bartonella to humans]. Parassitologia 2004a;46(1-2):127-9. [97] Fabbi M, De Giuli L, Tranquillo M, Bragoni R, Casiraghi M, Genchi C. Prevalence of Bartonella henselae in Italian stray cats: evaluation of serology to assess the risk of transmission of Bartonella to humans. J Clin Microbiol 2004b;42(1):264-8. [98] Fenimore A, Varanat M, Maggi R, Schultheiss P, Breitschwerdt E, Lappin MR. Bartonella spp. DNA in cardiac tissues from dogs in Colorado and Wyoming. J Vet Intern Med 2011;25(3):613-6. [99] Fenollar F, Raoult D. Molecular genetic methods for the diagnosis of fastidious microorganisms. APMIS 2004;112:785-807. [100] Fenollar F, Sire S, Raoult D. Bartonella vinsonii subsp. arupensis as an agent of blood culture-negative endocarditis in a human. J Clin Microbiol 2005;43(2):945-7. [101] Ferrés M, Abarca K, Godoy P, García P, Palavecino E, Méndez G, Valdés A, Ernst S, Thibaut J, Koberg J, Chanqueo L, Vial PA. [Presence of Bartonella henselae in cats: natural reservoir quantification and human exposition risk of this zoonosis in Chile]. Rev Med Chil 2005;133(12):1465-71. [102] Ferrés G M, Abarca V K, Prado D P, Montecinos P L, Navarrete C M, Vial C PA. [Prevalence of Bartonella henselae antibodies in Chilean children, adolescents and veterinary workers]. Rev Med Chil 2006;134(7):863-7. [103] Fisher A, van der Loo B, Schär G, Zbinden R, Duru F, Brunckhorst C, Rousson V, Delacrétaz E, Stuber T, Oechslin E, Follath F, Jenni R. Serological evidence for the association of Bartonella henselae infection with arrhythmogenic right ventricular cardiomyopathy. Clin Cardiol 2008;31(10):469–471. [104] Fitzgerald L. Overtraining increases the susceptibility to infection. Int J Sports Med 1991; 12(Suppl 1):S5-S8.

68 [105] Foley JE, Chomel B, Kikuchi Y, Yamamoto K, Pedersen NC. Seroprevalence of Bartonella henselae in cattery cats: association with cattery hygiene and flea infestation. Vet Q 1998;20(1):1-5. [106] Foucault C, Barrau K, Brouqui P, Raoult D. Bartonella quintana bacteremia among homeless people. Clin Infect Dis 2002;35(6):684-9. [107] Fournier PE, Mainardi JL, Raoult D. Value of microimmunofluorescence for diagnosis and follow-up of Bartonella endocarditis. Clin Diagn Lab Immunol 2002;9(4):795-801. [108] Fournier PE, Taylor C, Rolain JM, Barrassi L, Smith G, Raoult D. Bartonella australis sp. nov. from kangaroos, Australia. Emerg Infect Dis 2007;13(12):1961-2. [109] Garcia-Caceres U, Garcia FU. Bartonellosis. An immunodepressive disease and the life of Daniel Alcides Carrión. Am J Clin Pathol 1991;95(4 Suppl 1):S58-66. [110] García-García JA, Baquerizo R, Vargas J, Mira JA, Merchante N, Macías J, Pineda JA. [Prevalence of serum antibodies against Bartonella spp. in a healthy population from the south area of the Seville province]. Rev Clin Esp 2005;205(11):541-4. [111] Gasquet S, Maurin M, Brouqui P, Lepidi H, Raoult D. Bacillary angiomatosis in immunocompromised patients. AIDS 1998;12(14):1793- 803. [112] Gerber MA, Rapacz P, Kalter SS, Ballow M. Cell-mediated immunity in cat-scratch disease. J Allergy Clin Immunol 1986;78(5 Pt 1):887-90. [113] Gil H, García-Esteban C, Barandika JF, Peig J, Toledo A, Escudero R, Jado I, Rodríguez-Vargas M, García-Amil C, Lobo B, Roales P, Rodríguez- Moreno I, Olmeda AS, García-Pérez AL, Anda P. Variability of Bartonella genotypes among small mammals in Spain. Appl Environ Microbiol 2010;76(24):8062-70. [114] Giladi M, Kletter Y, Avidor B, Metzkor-Cotter E, Varon M, Golan Y, Weinberg M, Riklis I, Ephros M, Slater L. Enzyme immunoassay for the diagnosis of cat-scratch disease defined by polymerase chain reaction. Clin Infect Dis 2001;33:1852-8. [115] Gilmore RD, Carpio AM, Kosoy MY, Gage KL. Molecular characterization of the sucB gene encoding the immunogenic dihydrolipoamide succinyltransferase protein of Bartonella vinsonii subsp. berkhoffii and Bartonella quintana. Infect Immun 2003;71(8):4818-22. [116] Glaus T, Hofmann-Lehmann R, Greene C, Glaus B, Wolfensberger C, Lutz H. Seroprevalence of Bartonella henselae infection and correlation with disease status in cats in Switzerland. J Clin Microbiol 1997; 35(11):2883-5. [117] Graham J, Baden L, Tsiodras S, Karchmer A. Q fever endocarditis associated with extensive serological cross-reactivity. Clin Infect Dis 2000;30:609-10. [118] Grando D, Sullivan LJ, Flexman JP, Watson MW, Andrew JH. Bartonella henselae associated with Parinaud’s Oculoglandular Syndrome. Clin Infect Dis 1999;28:1156-8. [119] Guibal F, de La Salmoniere P, Rybojad M, Hadjrabia S, Dehen L, Arlet G. High seroprevalence to Bartonella quintana in homeless patients with cutaneous parasitic infestations in downtown Paris. J Am Acad Dermatol 2001;44:219-23. [120] Gundi VA, Davoust B, Khamis A, Boni M, Raoult D, La Scola B. Isolation of Bartonella rattimassiliensis sp. nov. and Bartonella phoceensis sp. nov. from European Rattus norvegicus. J Clin Microbiol 2004;42(8):3816-8.

69 [121] Haimerl M, Tenter AM, Simon K, Rommel M, Hilger J, Autenrieth IB. Seroprevalence of Bartonella henselae in cats in Germany. J Med Microbiol 1999;48(9):849-56. [122] Harrison TG, Doshi N. Serological evidence of Bartonella spp. infection in the UK. Epidemiol Infect 1999;123(2):233-40. [123] Harrus S, Bar-Gal GK, Golan A, Elazari-Volcani R, Kosoy MY, Morick D, Avidor B, Baneth G. Isolation and genetic characterization of a Bartonella strain closely related to Bartonella tribocorum and Bartonella elizabethae in Israeli commensal rats. Am J Trop Med Hyg 2009;81(1):55-8. [124] Heller R, Kubina M, Mariet P, Riegel P, Delacour G, Dehio C, Lamarque F, Kasten R, Boulouis HJ, Monteil H, Chomel B, Piémont Y. Bartonella alsatica sp. nov., a new Bartonella species isolated from the blood of wild rats. Int J Syst Bacteriol 1998a;48:1333-9. [125] Heller R, Riegel P, Hansmann Y, Delacour G, Bermond D, Dehio C, Lamarque F, Monteil H, Chomel B, Piémont Y. Bartonella tribocorum sp. nov., a new Bartonella species isolated from the blood of wild rats. Int J Syst Bacteriol 1998b;48(Pt 4):1333-9. [126] Hjelm E, Wesslén L, Gnarpe H, Gnarpe J, Nyström-Rosander C, Rolf C, Friman G. Antibodies to Chlamydia pneumoniae in young Swedish orienteers. Scand J Infect Dis 2001;33(8):589-92. [127] Hjelm, E., S. McGill, and G. Blomqvist. Prevalence of antibodies to Bartonella henselae, B. elizabethae, and B. quintana in Swedish domestic cats. Scand J Infect Dis 2002;34(3):192-6. [128] Hollingdale MR, Herrmann JE, Vinson JW. Enzyme immunoassay of antibody to Rochalimaea quintana: diagnosis of trench fever and serologic cross-reactions among other rickettsiae. J Infect Dis 1978;137:578-82. [129] Hollingdale MR, Vinson JW, Herrmann JE. Immunochemical and biological properties of the outer membrane-associated lipopolysaccharide and protein of Rochalimaea quintana. J Infect Dis 1980;141:672-8. [130] Holmberg M, McGill S, Ehrenborg C, Wesslén L, Hjelm E, Darelid J, Blad L, Engstrand L, Regnery R, Friman G. Evaluation of human seroreactivity to Bartonella species in Sweden. J Clin Microbiol 1999;37:1381-4. [131] Holmberg M, Mills JN, McGill S, Benjamin G, Ellis BA. Bartonella infection in sylvatic small mammals of central Sweden. Epidemiol Infect 2003;130:149-57. [132] Holmes AH, Greenough TC, Balady GJ, Regnery RL, Anderson BE, O’Keane JC, Fonger JD, McCrone EL. Bartonella henselae endocarditis in an immunocompetent adult. Clin Infect Dis 1995;21:1004-7. [133] Houpikian P, Raoult D. Molecular phylogeny of the genus Bartonella: what is the current knowledge? FEMS Microbiol Lett 2001;200(1)1-7. [134] Houpikian P, Raoult D. Western immunoblotting for Bartonella endocarditis. Clin Diagn Lab Immunol 2003;10(1):95-102. [135] Hsieh JW, Tung KC, Chen WC, Lin JW, Chien LJ, Hsu YM, Wang HC, Chomel BB, Chang CC. Epidemiology of Bartonella infection in rodents and shrews in Taiwan. Zoonoses Public Health 2010;57(6):439-46. [136] Huang J, Dai L, Lei S, Liao DY, Wang XQ, Luo TY, Chen Y, Hang ZB, Li GD, Dong DD, Xu G, Gu ZC, Hao JL, Hua P, He L, Duan FL. [Application of Warthin-Starry stain, immunohistochemistry and transmission electron microscopy in diagnosis of cat scratch disease]. Zhonghua Bing Li Xue Za Zhi 2010;39(4):225-9. [137] Inoue K, Maruyama S, Kabeya H, Yamada N, Ohashi N, Sato Y, Yukawa M, Masuzawa T, Kawamori F, Kadosaka T, Takada N, Fujita H, Kawabata H.

70 Prevalence and genetic diversity of Bartonella species isolated from wild rodents in Japan. Appl Environ Microbiol 2008;74(16):5086-92. [138] Jackson LA, Perkins BA, Wenger JD. Cat scratch disease in the United States: an analysis of three national databases. Am J Public Health 1993;83:1707-11. [139] Jackson LA, Spach DH, Kippen DA, Sugg NK, Regnery RL, Sayers MH, Stamm WE. Seroprevalence to Bartonella quintana among patients at a community clinic in downtown Seattle. J Infect Dis 1996;173(4):1023-6. [140] Jacobs RF, Schutze GE. Bartonella henselae as a cause of prolonged fever and fever of unknown origin in children. Clin Infect Dis 1998;26:80-4. [141] Jalava J, Kotilainen P, Nikkari S, Skurnik M, Vänttinen E, Lehtonen OP, Eerola E, Toivanen P. Use of the polymerase chain reaction and DNA sequencing for detection of Bartonella quintana in the aortic valve of a patient with culture-negative infective endocarditis. Clin Infect Dis 1995;21(4):891-6. [142] Jameson P, Greene C, Regnery R, Dryden M, Marks A, Brown J, Cooper J, Glaus B, Greene R. Prevalence of Bartonella henselae antibodies in pet cats throughout regions of North America. J Infect Dis 1995;172(4):1145-9. [143] Jardine C, Appleyard G, Kosoy MY, McColl D, Chirino-Trejo M, Wobeser G, Leighton FA. Rodent-associated Bartonella in Saskatchewan, Canada. Vector Borne Zoonotic Dis 2005;5(4):402-9. [144] Jerris RC, Regnery RL. Will the real agent of cat-scratch disease please stand up? Annu Rev Microbiol 1996;50:707-25. [145] Juncker-Voss M, Prosl H, Lussy H, Enzenberg U, Auer H, Lassnig H, Müller M, Nowotny N. [Screening for antibodies against zoonotic agents among employees of the Zoological Garden of Vienna, Schönbrunn, Austria]. Berl Munch Tierarztl Wochenschr 2004;117(9-10):404-9. [146] Karjalainen J, Heikkilä J. Incidence of three presentations of acute myocarditis in young men in military service: A 20-year experience. Eur Heart J 1999;20:1120-5. [147] Keret D, Giladi M, Kletter Y, Weintroub S. Cat-scratch disease osteomyelitis from a dog scratch. J Bone Joint Surg 1998;80:766-7. [148] Kerkhoff FT, Bergmans AM, van Der Zee A, Rothova A. Demonstration of Bartonella grahamii DNA in ocular fluids of a patient with neuroretinitis. J Clin Microbiol 1999;37:4034-8. [149] Kernif T, Aissi M, Doumandji SE, Chomel BB, Raoult D, Bitam I. Molecular evidence of Bartonella infection in domestic dogs from Algeria, North Africa, by polymerase chain reaction (PCR). Am J Trop med Hyg 2010;83(2):298-300. [150] Kikuchi E, Maruyama S, Sakai T, Tanaka S, Yamaguchi F, Hagiwara T, Katsube Y, Mikami T. Serological investigation of Bartonella henselae infections in clinically cat-scratch disease-suspected patients, patients with cardiovascular disease, and healthy veterinary students in Japan. Microbiol Immunol 2002;46(5):313-6. [151] Kim CM, Kim JY, Yi YH, Lee MJ, Cho MR, Shah DH, Klein TA, Kim HC, Song JW, Chong ST, O'Guinn ML, Lee JS, Lee IY, Park JH, Chae JS. Detection of Bartonella species from ticks, mites and small mammals in Korea. J Vet Sci 2005;6(4):327-34. [152] Klotz SA, Ianas V, Elliott SP. Cat-scratch disease. Am Fam Physician 2011;83(2):152-5.

71 [153] Knap N, Duh D, Birtles R, Trilar T, Petrovec M, Avsic-Zupanc T. Molecular detection of Bartonella species infecting rodents in Slovenia. FEMS Immunol Med Microbiol 2007;50(1):45-50. [154] Knobloch J. Analysis and preparation of Bartonella bacilliformis antigens. Am J Trop Med Hyg 1988a;39:173-8. [155] Knobloch J, Bialek R, Muller G, Asmus P. Common surface epitope of Bartonella bacilliformis and Chlamydia psittaci. Am J Trop Med Hyg 1988b;39(5):427-33. [156] Koehler JE, LeBoit PE, Egbert BM, Berger TG. Cutaneous vascular lesions and disseminated cat scratch disease in patients with the acquired immunodeficiency syndrome (AIDS) and AIDS-related complex. Ann Intern Med 1988;109:449-55. [157] Koehler JE, Tappero JW. Bacillary angiomatosis and bacillary peliosis in patients with human immunodeficiency virus. Clin Infect Dis 1993;17(4):612-24. [158] Koehler JE, Glasser CA, Tappero JW. Rochalimaea henselae infection: a new zoonosis with the domestic cat as reservoir. JAMA 1994;271:531-5. [159] Koehler JE. Bartonella-associated infections in HIV-infected patients. AIDS Clin Care 1995;7(12):97-102. [160] Koehler JE, Sanchez MA, Tye S, Garrido-Rowland CS, Chen FM, Maurer T, Cooper JL, Olson JG, Reingold AL, Hadley WK, Regnery RR, Tappero JW. Prevalence of Bartonella infection among human immunodeficiency virus- infected patients with fever. Clin Infect Dis 2003;37(4):559-66. [161] Kordick DL, Wilson KH, Sexton DJ, Hadfield TL, Berkhoff HA, Breitschwerdt EB. Prolonged Bartonella bacteremia in cats associated with cat-scratch disease patients. J Clin Microbiol 1995;33(12):3245-51. [162] Kordick DL, Hiljard EJ, Hadfield TL, Wilson KH, Steigerwalt AG, Brenner DJ, Breitschwerdt EB. Bartonella clarridgeiae, a newly recognized zoonotic pathogen causing inoculation papules, fever, and lymphadenopathy (cat scratch disease). J Clin Microbiol 1997;35:1813-8. [163] Kordick DL, Brown TT, Shin K, Breitschwerdt EB. Clinical and pathologic evaluation of chronic Bartonella henselae or Bartonella clarridgeiae infection in cats. J Clin Microbiol 1999;37(5):1536-47. [164] Kosek M, Lavarello R, Gilman RH, Delgado J, Maguiña C, Verástegui M, Lescano AG, Mallqui V, Kosek JC, Recavarren S, Cabrera L. Natural history of infection with Bartonella bacilliformis in a nonendemic population. J Infect Dis 2000;182(3):865-72. [165] Kosoy MY, Regnery RL, Tzianabos T, Marston EL, Jones DC, Green D, Maupin GO, Olson JG, Childs JE. Distribution, diversity, and host specificity of Bartonella in rodents from the Southeastern United States. Am J Trop Med Hyg 1997;57:578-88. [166] Kosoy MY, Regnery RL, Kosaya OI, Jones DC, Marston EL, Childs JE. Isolation of Bartonella spp. from embryos and neonates of naturally infected rodents. J Wildl Dis 1998;34(2):305-9. [167] Kosoy M, Murrau M, Gilmore RD Jr, Bai Y, Gage KL. Bartonella strains from ground squirrels are identical to Bartonella washoensis isolated from a human patient. J Clin Microbiol 2003;41(2):645-50. [168] Kosoy M, Mandel E, Green D, Marston E, Childs J. Prospective studies of Bartonella of rodents. Part I. Demographic and temporal patterns in population dynamics. Vector Borne Zoonotic Dis 2004;4(4):285-95. [169] Kosoy M, Morway C, Sheff KW, Bai Y, Colborn J, Chalcraft L, Dowell SF, Peruski LF, Maloney SA, Baggett H, Sutthirattana S, Sidhirat A, Maruyama

72 S, Kabeya H, Chomel BB, Kasten R, Popov V, Robinson J, Kruglov A, Petersen LR. Bartonella tamiae sp. nov., a newly recognized pathogen isolated from three human patients from Thailand. J Clin Microbiol 2008;46:772-5. [170] Kosoy M, Bai Y, Sheff K, Morway C, Baggett H, Maloney SA, Boonmar S, Bhengsri S, Dowell SF, Sitdhirasdr A, Lerdthusnee K, Richardson J, Peruski LF. Identification of Bartonella infections in febrile human patients from Thailand and their potential animal reservoirs. Am J Trop Med Hyg 2010;82(6):1140-5. [171] Kumasaka K, Arashima Y, Yanai M, Hosokawa N, Kawano K. Survey of veterinary professionals for antibodies to Bartonella henselae in Japan. Rinsho Byori 2001;49(9):906-10. [172] La Scola B, Raoult D. Serological cross-reactions between Bartonella quintana, Bartonella henselae, and Coxiella burnetii. J Clin Microbiol 1996;34(9):2270-4. [173] La Scola B, Raoult D. Culture of Bartonella quintana and Bartonella henselae from human samples: a 5-year experience (1993 to 1998). J Clin Microbiol 1999;37:1899-1905. [174] La Scola B, Holmberg M, Raoult D. Lack of Bartonella spp. in 167 Ixodes ricinus ticks collected in central Sweden. Scand J Infect Dis 2004;36(4):305- 6. [175] Lamas C, Rozental T, Favacho A, Bóia MN, Maduro R, Oliveira AP, Lemos ERS. Bartonella spp. infections in HIV positive individuals in Rio de Janeiro, Brazil. Int J Infect Dis 2006;10(Suppl. 1):S176. [176] Lamas D, Curi A, Boia MN, Lemos ERS. Human bartonellosis: seroepidemiological and clinical features with an emphasis on data from Brazil- a review. Mem Inst Oswaldo Cruz 2008;103(3):221-35. [177] Lamas CC, Mares-Guia MA, Rozental T, Moreira N, Favacho AR, Barreira J, Guterres A, Bóia MN, de Lemos ER. Bartonella spp. infection in HIV positive individuals, their pets and ectoparasites in Rio de Janeiro, Brazil: serological and molecular study. Acta Trop 2010;115(1-2):137-41. [178] Larsson E, Wesslén L, Lindquist O, Baandrup U, Eriksson L, Olsen E, Rolf C, Friman G. Sudden unexpected cardiac deaths among young Swedish orienteers- morphological changes in hearts and other organs. APMIS 1999;107(3):325-36. [179] LeBoit PE, Berger TG, Egbert BM, Beckstead JH, Benedict Yen TS, et al. Epithelioid haemangioma-like vascular proliferation in AIDS: manifestation of cat-scratch disease bacillus infection? Lancet 1988;1:960-3. [180] Leighton FA, Artsob HA, Chu MC, Olson JG. A serological survey of rural dogs and cats on the southwestern Canadian prairie for zoonotic pathogens. Can J Public Health 2001;92(1):67-71. [181] Lepidi H, Fournier PE, Raoult D. Quantitative analysis of valvular lesions during Bartonella endocarditis. Am J Clin Pathol 2000;114(6):880-9. [182] Liang Z, Raoult D. Differentiation of Bartonella species by a microimmunofluorescence assay, sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and Western immunoblotting. Clin Diagn Lab Immunol 2000;7(4):617-24. [183] Lin JW, Chen CY, Chen WC, Chomel BB, Chang CC. Isolation of Bartonella species from rodents in Taiwan including a strain closely related to ‘Bartonella rochalimaea’ from Rattus norvegicus. J Med Microbiol 2008;57(Pt 12):1496-501.

73 [184] Lindroos HL, Mira A, Repsilber D, Vinnere O, Näslund K, Dehio M, Dehio C, Andersson SG. Characterization of the genome composition of Bartonella koehlerae by microarray comparative genomic hybridization profiling. J Bacteriol 2005;187(17):6155-65. [185] Litwin CM, Martins TB, Hill HR. Immunologic response to Bartonella henselae as determined by enzyme immunoassay and Western blot analysis. Am J Clin Pathol 1997;108:202-9. [186] Litwin CM, Johnson JM, Martins TB. The Bartonella henselae sucB gene encodes a dihydrolipoamide succinyltransferase protein reactive with sera from patients with cat-scratch disease. J Med Microbiol 2004;53(Pt 12):1221-7. [187] Lombardo J. Cat-scratch neuroretinitis. J Am Optom Assoc 1999;70:525- 30. [188] Loutit JS. Bartonella infections. Curr Clin Top Infect Dis 1997;17:269-90. [189] Lucey D, Dolan MJ, Moss CW, Garcia M, Hollis DG,Wegner S, Morgan G, Almeida R, Leong D, Greisen KS, et al. Relapsing illness due to Rochalimaea henselae in immunocompetent hosts: implication for therapy and new epidemiological associations. Clin Infect Dis 1992;14:683-8. [190] Maguina C, Gotuzzo E. Bartonellosis. New and old. Infect Dis Clin North Am 2000;14(1):1-22. [191] Maillard R, Riegel P, Barrat F, Bouillin C, Thibault D, Gandoin C, Halos L, Demanche C, Alliot A, Guillot J, Piémont Y, Boulouis HJ, Vayssier-Taussat M. Bartonella chomelii sp. nov., isolated from French domestic cattle (Bos taurus). Int J Syst Evol Microbiol 2004;54(Pt 1):215-20. [192] Mallqui V, Speelmon E, Verastegui M, Maguina-Vargas C, Pinell-Salles P, Lavarello R, Delgado J, Kosek M, Romero S, Arana Y, Gilman RH. Sonicated diagnostic immunoblot for bartonellosis. Clin Diagn Lab Immunol 2000;7(1):1-5. [193] Margileth AM. Dermotologic manifestations and update of cat scratch disease. Pediatr Dermatol 1988;5:1-9. [194] Margileth AM. Cat scratch disease update: etiology, diagnosis and treatment of 1900 patients. 1992a. Session 80, Symp Am Soc Microbiol, Anaheim, Calif. [195] Margileth AM. Antibiotic therapy for cat-scratch disease: clinical study of therapeutic outcome in 268 patients and a review of the literature. Pediatr Infect Dis J 1992b;11:474-8. [196] Margileth A, Baehren DF. Chest wall abscess due to cat-scratch disease (CSD) in an adult with antibodies to Bartonella clarridgeiae: case report and review of the thoracopulmonary manifestations of CSD. Clin Infect Dis 1998;27:353-7. [197] Margileth AM. Recent advances in diagnosis and treatment of cat scratch disease. Curr Infect Dis Rep 2000;2:141-6. [198] Marié JL, Fournier PE, Rolain JM, Briolant S, Davoust B, Raoult D. Molecular detection of Bartonella quintana, B. elizabethae, B. koehlerae, B. doshiae, B. taylorii, and Rickettsia felis in rodent fleas collected in Kabul, Afghanistan. Am J Trop Med Hyg 2006;74(3):436-9. [199] Maruyama S, Hiraga S, Yokoyama E, Naoi M, Tsuruoka Y, Ogura Y, Tamura K, Namba S, Kameyama Y, Nakamura S, Katsube Y. Seroprevalence of Bartonella henselae and Toxoplasma gondii infections among pet cats in Kanagawa and Saitama Prefectures. J Vet Med Sci 1998;60(9):997-1000.

74 [200] Maruyama S, Boonmar S, Morita Y, Sakai T, Tanaka S, Yamaguchi F, Kabeya H, Katsube Y. Seroprevalence of Bartonella henselae and Toxoplasma gondii among healthy individuals in Thailand. J Vet Med Sci 2000;62(6):635-7. [201] Maruyama S, Kabeya H, Nakao R, Tanaka S, Sakai T, Xuan X, Katsube Y, Mikami T. Seroprevalence of Bartonella henselae, Toxoplasma gondii, FIV and FeLV infections in domestic cats in Japan. Microbiol Immunol 2003;47(2):147-53. [202] Massei F, Messina F, Gori L, Macchia P, Maggiore G. High prevalence of antibodies to Bartonella henselae among Italian children without evidence of cat scratch disease. Clin Infect Dis 2004;38(1):145-8. [203] Massei F, Gori L, Macchia P, Maggiore G. The expanded spectrum of bartonellosis in children. Infect Dis Clin North Am 2005;19(3):691-711. [204] Maurin M, Raoult D. Antimicrobial susceptibility of Rochalimaea quintana, Rochalimaea vinsonii, and the newly recognized Rochalimaea henselae. J Antimicrobial Chemother 1993;32:587-94. [205] Maurin M, Roux V, Stein A, Ferrier F, Viraben R, Raoult D. Isolation and characterization by immunofluorescence, sodium dodecyl sulfate- polyacrylamide gel electrophoresis, Western blot, restriction fragment length polymorphism-PCR, 16S rRNA gene sequencing, and pulsed-field gel electrophoresis of Rochalimaea quintana from a patient with bacillary angiomatosis. J Clin Microbiol 1994;32:1166-71. [206] Maurin M, Eb F, Etienne J, Raoult D. Serological cross-reactions between Bartonella and Chlamydia species: implications for diagnosis. J Clin Microbiol 1997;35(9):2283-7. [207] Maurin M, Raoult D. Isolation in endothelial cell cultures of Chlamydia trachomatis LGV (Serovar L2) from a lymph node of a patient with suspected cat scratch disease. J Clin Microbiol 2000;38(6):2062-4. [208] Maurin M, Rolain JM, Raoult D. Comparison of in-house and commercial slides for detection of immunoglobulin G and M by immunofluorescence against Bartonella henselae and Bartonella quintana. Clin Diagn Lab Immunol 2002;9:1004-9. [209] McCaffrey FM, Braconier DS, Strong WB. Sudden cardiac death in young athletes. Am J Dis Child 1991;145:177-83. [210] McGill S, Regnery R, Karem K. 1998. Characterization of human immunoglobulin (Ig) isotype and IgG subclass response to Bartonella henselae infection. Infect Immun 1998;66(12):5915-20. [211] McGill S, Wesslén L, Hjelm E, Holmberg M, Rolf C, Friman G. Serological and epidemiological analysis of the prevalence of Bartonella spp. antibodies in Swedish elite orienteers 1992-93. Scand J Infect Dis 2001;33(6):423-8. [212] McGill S, Rajs J, Hjelm E, Lindquist O, Friman G. A study on forensic samples of Bartonella spp. antibodies in Swedish intravenous heroin addicts. APMIS 2003;111:507-13. [213] McGill, S, Wesslén L, Hjelm E, Holmberg M, Auvinen M, Berggren K, Grandin-Jarl B, Johnson U, Wikström S, Friman G. Bartonella spp. seroprevalence in healthy Swedish blood donors. Scand J Infect Dis 2005;37(10):723-30. [214] Mediannikov O, Ivanov L, Zdanovskaya N, Vysochina N, Fournier PE, Tarasevich I, Raoult D. Molecular screening of Bartonella species in rodents from the Russian Far East. Ann N Y Acad Sci 2005;1063:308-11.

75 [215] Meininger GR, Nadasdy T, Hruban RH, Bollinger RC, Baughman KL, Hare JM. Chronic active myocarditis following acute Bartonella henselae infection (cat scratch disease). Am J Surg Pathol 2001;25(9):1211-4. [216] Mexas AM, Hancock SL, Breitschwerdt EB. Bartonella henselae and Bartonella elizabethae as potential canine pathogens. J Clin Microbiol 2002;40(12):4670-4. [217] Midani S, Ayoub EM, Anderson B. Cat-scratch disease. Adv Paediatr 1996;43:397-423. [218] Minadakis G, Chochlakis D, Kokkini S, Gikas A, Tselentis Y, Psaroulaki A. Seroprevalence of Bartonella henselae antibodies in blood donors in Crete. Scand J Infect Dis 2008;40(10):846-7. [219] Minnick MF, Battisti JM. Pestilence, persistence and pathogenicity: infection strategies of Bartonella. Future Microbiol 2009;4(6):743-58. [220] Molia S, Chomel BB, Kasten RW, Leutenegger CM, Steele BR, Marker L, Martenson JS, Keet DF, Bengis RG, Peterson RP, Munson L, O’Brien SJ. Prevalence of Bartonella infection in wild African lions (Panthera leo) and cheetahs (Acinonyx jubatus). Vet Microbiol 2004;100(1-2):31-41. [221] Montcriol A, Benard F, Fenollar F, Ribeiri A, Bonnet M, Collart F, Guidon C. Fatal myocarditis-associated Bartonella quintana endocarditis: a case report. J Med Case Reports 2009;3:7325. [222] Morick D, Baneth G, Avidor B, Kosoy MY, Mumcuoglu KY, Mintz D, Eyal O, Goethe R, Mietze A, Shpigel N, Harrus S. Detection of Bartonella spp. in wild rodents in Israel using HRM real-time PCR. Vet Microbiol 2009;139(3- 4):293-7. [223] Mosbacher M, Elliott SP, Shehab Z, Pinnas JL, Klotz JH, Klotz SA. Cat scratch disease and arthropod vectors: more to it than a scratch? J Am Board Fam Med 2010;23(5):685-6. [224] Nadal D, Zbinden R. Serology to Bartonella (Rochalimaea) henselae may replace traditional diagnostic criteria for cat-scratch disease. Eur J Pediatr 1995;154(11):906-8. [225] Nasirudeen AM, Thong ML. Prevalence of Bartonella henselae immunoglobulin G antibodies in Singaporean cats. Pediatr Infect Dis J 1999;18(3):276-8. [226] Nilsson K, Liu A, Påhlson C, Lindquist O. Demonstration of intracellular microorganisms (Rickettsia spp., Chlamydia pneumoniae, Bartonella spp.) in pathological human aortic valves by PCR. J Infect 2005;50(1):46-52. [227] Noah DL, Kramer CM, Verbsky MP, Rooney JA, Smith KA, Childs JE. Survey of veterinary professionals and other veterinary conference attendees for antibodies to Bartonella henselae and B. quintana. J Am Vet Med Assoc 1997;210(3):342-4. [228] Nosal JM. Bacillary angiomatosis, cat-scratch disease, and bartonellosis: what’s the connection? Int J Dermatol 1997;36:405-11. [229] Nutter FB, Dubey JP, Levine JF, Breitschwerdt EB, Ford RB, Stoskopf MK. Seroprevalences of antibodies against Bartonella henselae and Toxoplasma gondii and fecal shedding of Cryptosporidium spp., Giardia spp. and Toxocara cati in feral and pet domestic cats. J Am Vet Med Assoc 2004;225(9):1394-8. [230] O’Halloran HS, Draud K, Minix M, Rivard AK, Pearson PA. Leber’s neuroretinitis in a patient with serologic evidence of Bartonella elizabethae. Retina 1998;18(3):276-8.

76 [231] Olsson-Engvall E, Englund L, Fasth C, Brändström C, Blomqvist G. B. henselae in Swedish cats. American Society for Rickettsiology/Bartonella as an Emerging Pathogen Group Joint Conference, Big Sky, MT, 2001. [232] Pandak N, Dakovi -Rode O, Cabraja I, Kristof Z, Kotarac S. Prevalence of Bartonella henselae antibodies in children and blood donors in Croatia. Infection 2009;37(2):166-7. [233] Papadea C, Check IJ. Human immunoglobulin G and immunoglobulin G subclasses: biochemical, genetic, and clinical aspects. Crit Rev Clin Lab Sci 1989;27:27-58. [234] Pappalardo BL, Correa MT, York CC, Peat CY, Breitschwerdt EB. Epidemiologic evaluation of the risk factors associated with exposure and seroreactivity to Bartonella vinsonii in dogs. Am J Vet Res. 1997;58:467-71. [235] Parinaud H. Conjoncitivité infectieuse transmise par les animaux. Annales D Oculistique 1889;101:252-3. [236] Pinna Parpaglia ML, Masu G, Masala G, Porcu R, Zobba R, Pintori G, Cocco R. Seroprevalence of Bartonella henseae in dogs and cats in Sassari. Vet Res Commun 2007;31 Suppl 1:317-20. [237] Pipili C, Katsogridakis K, Cholongitas E. Myocarditis due to Bartonella henselae. South Med J 2008; 101(11):1186. [238] Podsiad y E, Soko owska E, Tylewska-Wierzbanowska S. [Seroprevalence of Bartonella henselae and Bartonella quintana infections in Poland in 1998- 2001]. Przegl Epidemiol 2002;56(3):399-407. [239] Pons I, Sanfeliu I, Cardeñosa N, Nogueras MM, Font B, Segura F. Serological evidence of Bartonella henselae infection in healthy people in Catalonia, Spain. Epidemiol Infect 2008a;136(12):1712-6. [240] Pons I, Sanfeliu I, Nogueras MM, Sala M, Cervantes M, Amengual MJ, Segura F. Seroprevalence of Bartonella spp. infection in HIV patients in Catalonia, Spain. BMC Infect Dis 2008b;8:58. [241] Pretorius AM, Beati L, Birtles RJ. Diversity of bartonellae associated with small mammals inhabiting Free State province, South Africa. Int J Syst Evol Microbiol 2004;54 (Pt 6):1959-67. [242] Rahimian J, Raoult D, Yi-Wei Tang, Hanna BA. Bartonella quintana endocarditis with positive serology for Coxiella burnetii. J Infect 2006;53(3):e151-3. [243] Rajs J, Falconer B. Cardiac lesions in intravenous drug addicts. Forensic Sci Int 1979;13:193-209. [244] Rajs J, Fugelstad A. Suicide related to human immunodeficiency virus infection in Stockholm. Acta Psychiatr Scand 1992;85:234-9. [245] Ramos AJ, Vargas J, Fernández-Rivera J, Macías J, Mira JA, Pineda JA. [Prevalence of Bartonella spp. seropositivity in human immunodeficiency virus-infected and non-infected intravenous drug abusers]. Med Clin (Barc) 2002;119(15):565-7. [246] Raoult D, Dupont HT, Enea-Mutillod M. Positive predictive value of Rochalimaea henselae antibodies in the diagnosis of cat-scratch disease. Clin Infect Dis 1994;19:335. [247] Raoult D, Fournier PE, Drancourt M, Marrie TJ, Etienne J, Cosserat J, Cacoub P, Poinsignon Y, Leclercq P, Sefton AM. Diagnosis of 22 new cases of Bartonella endocarditis. Ann Inern Med 1996;125:646-52. [248] Raoult D, La Scola B, Kelly PJ, Davoust B, Gomez J. Bartonella bovis in cattle in Africa. Vet Microbiol 2005;105:155-6.

77 [249] Rath PM, von Recklinghausen G, Ansorg R. Seroprevalence of immunoglobulin G antibodies to Bartonella henselae in cat owners. Eur J Clin Microbiol Infect Dis 1997;16(4):326-7. [250] Reed JA, Brigati DJ, Flynn SD, McNutt SN, Min K, Welch DF, Slater LN. Immunocytochemical identification of Rochalimaea henselae in bacillary (epithelioid) angiomatosis, parenchymal bacillary peliosis, and persistent fever with bacteremia. Am J Surg Pathol 1992;16:6540-7. [251] Reed JB, Scales DK, Wong MT, Lattuada CP, Jr., Dolan MJ, Schwab IR. Bartonella henselae neuroretinitis in cat scratch disease. Diagnosis, management, and sequelae. Ophthalmology 1998;105:459-66. [252] Regnery RL, Anderson BE, Clarridge JE, Rodriguez-Barradas MC, Jones DC, Carr JH. Characterization of a novel Rochalimaea species, R. henselae sp. nov., isolated from blood of a febrile, human immunodeficiency virus- positive patient. J Clin Microbiol 1992a;30:265–274. [253] Regnery RL, Olson JG, Perkins BA, Bibb W. Serological response to Rochalimaea henselae antigen in suspected cat scratch disease. Lancet 1992b;339:1443-5. [254] Regnery R, Tappero J. Unraveling mysteries associated with cat-scratch disease, bacillary angiomatosis, and related syndromes. Emerg Infect Dis 1995a;1:16-21. [255] Regnery RL, Childs JE, Koehler JE. Infections associated with Bartonella species in persons infected with human immunodeficiency virus. Clin Infect Dis 1995b;21(suppl 1):S94-8. [256] Relman DA, Loutit JS, Schmidt TM, Falkow S, Tompkins LS. The agent of bacillary angiomatosis. An approach to the identification of uncultured pathogens. N Engl J Med 1990;323(23):1573-80. [257] Resto-Ruiz S, Burgess A, Anderson BE. The role of the host immune response in pathogenesis of Bartonella henselae. DNA Cell Biol 2003;22(6):431-40. [258] Rodriguez-Barradas MC, Bandres JC, Hamill RJ, Trial J, Clarridge JE, Baughn RE, Rossen RD. In vitro evaluation of the role of humoral immunity against Bartonella henselae. Infect Immun 1995;63(6):2367-70. [259] Rolain JM, Locatelli C, Chabanne L, Davoust B, Raoult D. Prevalence of Bartonella clarridgeiae and Bartonella henselae in domestic cats from France and detection of the organisms in erythrocytes by immunofluorescence. Clin Diagn Lab Immunol 2004;11(2):423-5. [260] Rotily M, Obadia Y, Tissot-Dupont H, Cavailler P, Raoult D. [The epidemiology of trench fever: a pilot study in homeless people in Marseilles]. Sante 1996;6(5):275-8. [261] Roux V, Eykyn SJ, Wyllie S, Raoult D. Bartonella vinsonii subsp. berkhoffii as an agent of afebrile blood culture negative endocarditis in a human. J Clin Microbiol 2000;38:1698-700. [262] Saenz HL, Engel P, Stoeckli MC, Lanz C, Raddatz G, Vayssier-Taussat M, Birtles R, Schuster SC, Dehio C. Genomic analysis of Bartonella identifies type IV secretion systems as host adaptability factors. Nat Genet 2007;39(12):1469-76. [263] Sanchez E, Perez E, Benocilla P, Merello J, Revilla L, Guevara Z, Birtles R, Raoult D. Epidemiology of Bartonella bacilliformis in new endemic areas in the jungle of the Amazonas. Int J Infect Dis 2006;10 (Suppl.1):13. [264] Sander A, Bühler C, Pelz K, von Cramm E, Bredt W. Detection and identification of two Bartonella henselae variants in domestic cats in Germany. J Clin Microbiol 1997;35(3):584-7.

78 [265] Sander A, Posselt M, Oberle K, Bredt W. Seroprevalence of antibodies to Bartonella henselae in patients with cat scratch disease and in healthy controls: evaluation and comparison of two commercial tests. Clin Diagn Lab Immunol 1998;5(4):486-90. [266] Sander A, Berner R, Ruess M. Serodiagnosis of cat scratch disease: response to Bartonella henselae in children and a review of diagnostic methods. Eur J Clin Microbiol Infect Dis 2001;20:392-401. [267] Scherer DC, DeBuron-Connors I, Minnick MF. Characterization of Bartonella bacilliformis flagella and effect of antiflagellin antibodies on invasion of human erythrocytes. Infect Immun 1993;61(12):4962-71. [268] Schiellerup P, Dyhr T, Rolain JM, Christensen M, Damsgaard R, Ethelberg S, Fisker N, Frost Andersen N, Raoult D, Krogfelt KA. Low seroprevalence of Bartonella species in Danish elite orienteers. Scand J Infect Dis 2004;36(8):604-6. [269] Seals JE, Oken HA. Cat scratch encephalopathy. Md Med J 1999; 48:176-8. [270] Seki N, Sasaki T, Sawabe K, Sasaki T, Matsuoka M, Arakawa Y, Marui E, Kobayashi M. Epidemiological studies on Bartonella quintana infections among homeless people in Tokyo, Japan. Jpn J Infect Dis 2006;59(1):31-5. [271] Shah SS, McGowan JP. Rickettsial, ehrlichial and Bartonella infections of the myocardium and pericardium. Front Biosci 2003;8:e197-201. [272] Skerget M, Wenisch C, Daxboeck F, Krause R, Haberl R, Stuenzner D. Cat or dog ownership and seroprevalence of ehrlichiosis, Q fever, and cat-scratch disease. Emerg Infect Dis 2003;9(10):1337-40. [273] Slater LN, Welch DF, Hensel D, Coody DW. A newly recognized pathogen as a cause of fever and bacteraemia. N Engl J Med 1990;323:1587-93. [274] Slater LN, Welch DF, Min KW. Rochalimaea henselae causes bacillary angiomatosis and peliosis hepatis. Arch Intern Med 1992a;152:602-6. [275] Slater LN, Coody DW, Woolridge LK, Welch DF. Murine antibody responses distinguish Rochalimaea henselae from Rochalimaea quintana. J Clin Microbiol 1992b;30(7):1722-7. [276] Smith HM, Reporter R, Rood MP, Linscott AJ, Mascola LM, Hogrefe W, Purcell RH. Prevalence study of antibody to rat borne pathogens and other agents among patients using a free clinic in downtown Los Angeles. J Infect Dis 2002;186(11):1673-6. [277] Solano-Gallego L, Hegarty B, Espada Y, Llull J, Breitschwerdt E. Serological and molecular evidence of exposure to arthropod-borne organisms in cats from northeastern Spain. Vet Microbiol 2006;118(3- 4):274-7. [278] Spach DH, Kanter AS, Daniels NA, Nowowiejski DJ, Larson AM, Schmidt RA, Swaminathan B, Brenner DJ. Bartonella (Rochalimaea) species as a cause of apparent ‘culture-negative’ endocarditis. Clin Infect Dis 1995;20:1044-7. [279] Sun J, Fu G, Lin J, Song X, Lu L, Liu Q. Seroprevalence of Bartonella in Eastern China and analysis of risk factors. BMC Infect Dis 2010;10:121. [280] Szelc-Kelly CM, Goral S, Perez-Perez GI, Perkins BA, Regnery RL, Edwards KM. Serologic responses to Bartonella and Afipia antigens in patients with cat scratch disease. Pediatrics 1995;96:1137-42. [281] Tea A, Alexiou-Daniel S, Arvanitidou M, Diza E, Antoniadis A. Occurrence of Bartonella quintana in a healthy Greek population. Am J Trop Med Hyg 2003;68(5):554-6.

79 [282] Trafny DJ, Oyama MA, Wormser C, Reynolds CA, Singletary GE, Peddle GD. Cardiac troponin-I concentrations in dogs with bradyarrhythmias before and after artificial pacing. J Vet Cardiol 2010;12(3):183-90. [283] Tsai YL, Chuang ST, Chang CC, Kass PH, Chomel BB. Bartonella species in small mammals and their ectoparasites in Taiwan. Am J Trop Med Hyg 2010;83(4):917-23. [284] Tsai YL, Lin CC, Chomel BB, Chuang ST, Tsai KH, Wu WJ, Huang CG, Yu JC, Sung MH, Kass PH, Chang CC. Bartonella infection in shelter cats and dogs and their ectoparasites. Vector Borne Zoonotic Dis 2011;11(8):1023- 30. [285] Tsukahara M, Tsuneoka H, Lino H, Ohno K, Murano I. Bartonella henselae infection from a dog. Lancet 1998;352:1682. [286] Ueno H, Muramatsu Y, Chomel BB, Hohdatsu T, Koyama H, Morita C. Seroepidemiological survey of Bartonella (Rochalimaea) henselae in domestic cats in Japan. Microbiol Immunol 1995;39(5):339-41. [287] Varanat M, Broadhurst J, Linder KE, Maggi RG, Breitschwerdt EB. Identification of Bartonella henselae in 2 cats with pyogranulomatous myocarditis and diaphragmatic myositis. Vet Pathol 2011. [288] Vayssier-Taussat M, Le Rhun D, Bonnet S, Cotté V. Insights in Bartonella host specificity. Ann N Y Acad Sci 2009;1166:127-32. [289] Velho PE, Cintra ML, Uthida-Tanaka AM, de Moraes AM, Mariotto A. What do we (not) know about the human bartonelloses? Braz J Infect Dis 2003;7(1):1-6. [290] Wagner CL, Riess T, Linke D, Eberhardt C, Schäfer A, Reutter S, Maggi R, Kempf VAJ. Use of Bartonella adhesion A (BadA) immunoblotting in the serodiagnosis of Bartonella henselae infections. Int J Med Microbiol 2008;298(7-8):579-90. [291] Waldvogel K, Regnery RL, Anderson BE, Caduff R, Caduff J, Nadal D. Disseminated cat-scratch disease: detection of Rochalimaea henselae in affected tissue. Eur J Pediatr 1994;153:23-7. [292] Warwick WJ. The cat-scratch syndrome, many diseases or one disease? Prog Med Virol 1967;9:256-301. [293] Welch DF, Pickett DA, Slater LN, Steigerwalt AG, Brenner DJ. Rochalimaea henselae sp. nov., a cause of septicemia, bacillary angiomatosis, and parenchymal bacillary peliosis. J Clin Microbiol 1992;30:275-80. [294] Welch DF, Carroll KC, Hofmeister EK, Persing DH, Robinson DA, Steigerwalt AG, Brenner DJ. Isolation of a new subspecies, Bartonella vinsonii subsp. arupensis, from a cattle rancher: identity with isolates found in conjunction with Borrelia burgdorferi and Babesia microti among naturally infected mice. J Clin Microbiol 1999;37(8): 2598-601. [295] Werner M, Fournier PE, Andersson R, Hogevik H, Raoult D. Bartonella and Coxiella antibodies in 334 prospectively studied episodes of infective endocarditis in Sweden. Scand J Infect Dis 2003;35(10):724-7. [296] Wesslén L, Påhlson C, Friman G, Fohlman J, Lindquist O, Johansson C. Myocarditis caused by Chlamydia pneumonia (TWAR) and sudden unexpected death in a Swedish elite orienteer. Lancet 1992;340:427-8. [297] Wesslén L, Påhlson C, Lindquist O, Hjelm E, Gnarpe J, Larsson E, Baandrup U, Eriksson L, Fohlman J, Engstrand L, Linglöf T, Nyström-Rosander C, Gnarpe H, Magnius L, Rolf C, Friman G. An increase in sudden unexpected cardiac deaths among young Swedish orienteers during 1979-1992. Eur Heart J 1996;17:902-10.

80 [298] Wesslén L, Ehrenborg C, Holmberg M, McGill S, Hjelm E, Lindquist O, Henriksen E, Rolf C, Larsson E, Friman G. Subacute Bartonella infection in Swedish orienteers succumbing to sudden unexpected cardiac death or having malignant arrhythmias. Scand J Infect Dis 2001;33:429-38. [299] Westling K, Farra A, Jorup C, Nordenberg A, Settergren B, Hjelm E. Bartonella henselae antibodies after cat bite. Emerg Infect Dis 2008;14(12):1943-4. [300] Windsor J. Cat-scratch disease: epidemiology, aetiology and treatment. Br J Biomed Sci 2001;58:101-10. [301] Winoto IL, Goethert H, Ibrahim IN, Yuniherlina I, Stoops C, Susanti I, Kania W, Maguire JD, Bangs MJ, Telford SR 3rd, Wongsrichanalai C. Bartonella species in rodents and shrews in the greater Jakarta area. Southeast Asian J Trop Med Public Health 2005;36(6):1523-9. [302] Wong MT, Dolan MJ, Lattuada CP Jr., Regnery RL, Garcia ML, Mokulis EC, LaBarre RA, Ascher DP, Delmar JA, Kelly JW, et al. Neuroretinitis, aseptic meningitis, and lymphadenitis associated with Bartonella (Rochalimaea) henselae infection in immunocompetent patients and patients infected with human immunodeficiency virus type 1. Clin Infect Dis 1995:21(2);352-60. [303] Woodruff JF. Viral myocarditis. Am J Pathol 1980;101:424-79. [304] Ye X, Li GW, Yao ML, Luo W, Su LQ. [Study on the prevalence and genotypes of Bartonella species in rodent hosts from Fujian coastal regions]. Zhonghua Liu Xing Bing Xue Za Zhi 2009;30(10):989-92. [305] Yilmaz C, Ergin C, Kaleli I. [Investigation of Bartonella henselae seroprevalence and related risk factors in blood donors admitted to Pamukkale University Blood Center]. Mikrobiyol Bul 2009;43(3):391-401. [306] Ying B, Kosoy MY, Maupin GO, Tsuchiya KR, Gage KL. Genetic and ecologic characteristics of Bartonella communities in rodents in southern China. Am J Trop Med Hyg 2002;66(5):622-7. [307] Zangwill KM, Hamilton DH, Perkins BA, Regnery RL, Plikaytis BD, Hadler JL, Cartter ML, Wenger JD. Cat scratch disease in Connecticut. Epidemiology, risk factors, and evaluation of a new diagnostic test. N Engl J Med 1993;329(1):8-13. [308] Zarkovic A, McMurray C, Deva N, Ghosh S, Whitley D, Guest S. Seropositivity rates for Bartonella henselae, Toxocara canis and Toxoplasma gondii in New Zealand blood donors. Clin Experiment Ophthalmol 2007;35(2):131-4. [309] Zbinden R, Hochli M, Nadal D. Intracellular location of Bartonella henselae cocultivated with Vero cells and used for an indirect fluorescent-antibody test. Clin Diagn Lab Immunol 1995;2:693-5. [310] Zhang P, Chomel BB, Schau MK, Goo JS, Droz S, Kelminson KL, George SS, Lerche NW, Koehler JE. A family of variably expressed outer- membrane proteins (Vomp) mediates adhesion and autoaggregation in Bartonella quintana. Proc Natl Acad Sci USA 2004;101(37):13630-5. [311] Znazen A, Rolain JM, Hammami N, Kammoun S, Hammami A, Raoult D. High prevalence of Bartonella quintana endocarditis in Sfax, Tunisia. Am J Trop Med Hyg 2005;72(5):503-7. [312] Zupan S, Poljak M, Avsic-Zupanc T. Prevalence of Bartonella infections in Slovenian intravenous drug users. Ann N Y Acad Sci 2003;990:414-8.

81