INHIBITORY MECHANISM OF HUMAN NEUTROPHIL APOPTOSIS BY ANAPLASMA PHAGOCYTOPHILUM AND IDENTIFICATION OF NOVEL SURFACE PROTEINS OF A. PHAGOCYTOPHILUM AND EHRLICHIA CHAFFEENSIS

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

By

Yan Ge, M.S.

* * * * *

The Ohio State University 2007

Dissertation Committee:

Dr. Yasuko Rikihisa, Adviser Approved by Dr. Laura Rush

Dr. Young C. Lin ______Adviser Dr. William P. Lafuse Graduate Program in Veterinary Biosciences

ABSTRACT

The inhibition of neutrophil apoptosis plays a central role in human granulocytic anaplasmosis. Intracellular signaling pathways through which the obligatory intracellular bacterium Anaplasma phagocytophilum inhibits the spontaneous apoptosis of human peripheral blood neutrophils were investigated. In this paper, it was found that the decrease of bfl-1 (an anti-apoptotic bcl-2 family member) mRNA expression and activation of caspase 3 (the main executioner caspase) during spontaneous neutrophil apoptosis are inhibited by A. phagocytophilum infection. It was observed that most uninfected neutrophils lost mitochondrial membrane potential in contrast with high membrane potential in infected cells. This suggests that A. phagocytophilum inhibits the intrinsic pathway of the spontaneous neutrophil apoptosis by protecting the mitochondrial membrane integrity.

Next, we studied the molecular signaling of the extrinsic apoptotic pathway

(death receptor pathway) and the interaction between the intrinsic and extrinsic pathways in the inhibition of spontaneous human neutrophil apoptosis by A. phagocytophilum. Cell surface Fas clustering during spontaneous neutrophil apoptosis was significantly blocked by infection. In the extrinsic pathway of spontaneous neutrophil apoptosis as well as anti-Fas-induced neutrophil apoptosis, the activation of

ii caspase 8 (initiator caspase of extrinsic pathway) and pro-apoptotic Bid (linking the intrinsic and extrinsic pathways) were inhibited by A. phagocytophilum infection. These data together point to a novel mechanism induced by A. phagocytophilum involving both extrinsic and intrinsic pathways to ensure to delay the apoptosis of host neutrophils.

Including inhibiting neutrophil apoptosis, the surface of A. phagocytophilum plays a crucial role in subverting the hostile host cell environment. However, with the exception of P44 outer membrane protein family, the surface components of A. phagocytophilum are largely unknown. To globally identify the major surface proteins of A. phagocytophilum, a membrane-impermeable biotin reagent, Sulfo-NHS-SS-

Biotin, was used to label the intact bacteria. Among the major proteins captured by biotin-streptavidin affinity purification, were five A. phagocytophilum proteins i.e.,

OMP85, hypothetical protein APH_0404, hypothetical protein APH_0405, P44 family proteins and OMP-1A. APH_0404 and APH_0405 were highly conserved in the family

Anaplasmataceae. Except P44s, all of them were newly identified as major surface- exposed proteins.

Ehrlichia chaffeensis belongs to the same family of Anaplasmataceae as A. phagocytophilum. Despite the importance of surface proteins as a crucial interface for

E. chaffeensis-host interactions, the knowledge of them is limited. So far, only P28, gp47 and gp120 have been shown as surface-exposed proteins. To globally investigate surface proteins of E. chaffeensis, Sulfo-NHS-SS-Biotin-streptavidin affinity-captured surface proteins were subjected to Nano-LC/MS/MS analysis. Nineteen out of 22 OMP-

1/P28 family proteins were detected in E. chaffeensis cultured in human monocytic

iii leukemia THP-1 cells. For the first time, seventeen OMP-1/P28 family proteins were demonstrated to be expressed at the protein level. In addition, OMP85, hypothetical protein ECH_0525, gp47 and 11 other proteins were detected. The identification of E. chaffeensis surface-exposed proteins provides novel insights about E. chaffeensis surface and lays the foundation for rational studies on pathogen-host interaction and vaccine development.

iv

Dedicated to my deeply loved parents and husband who has been working far away in China during my several years’ PhD study but giving me full family support.

v

ACKNOWLEDGMENTS

I’d like to thank my adviser, Dr. Yasuko Rikihisa, for intellectual and scientific guidance, her patience in experimental discussion, conscientious efforts in revising paper which made this dissertation possible, and for providing me free space for research.

I am grateful to my committee members, Drs. Laura Rush, Young C. Lin and

William P. Lafuse for nice suggestion and encouragement on my study.

I wish to thank my co-workers from our laboratory, especially Dr. Kiyotaka

Yoshiie, Dr. Futoshi Kuribayashi and Tzunghuei Lai.

I also appreciate Kate Hayes for copy-editing the manuscripts.

This research is financially supported by grants from the National Health

Institute.

vi

VITA

October 12, 1974 ...... Born - Jiangsu Province, China

1991 - 1995 ...... B.S. Veterinary Public Health Yangzhou University, China

1995 - 1998 ...... M.S. Veterinary Microbiology and Immunology Nanjing Agricultural University, China

1998 - 2002 ...... Researcher Assistant Investigator since 2000 Institute of Veterinary Medicine and Animal Husbandry, Shanghai Academy of Agricultural Sciences, China

2002 - present...... Graduate Research Associate, The Ohio State University

PUBLICATIONS

Recent Research Publications

1. Yan Ge and Yasuko Rikihisa. Anaplasma phagocytophilum delays spontaneous human neutrophil apoptosis by modulation of multiple apoptotic pathways. Cellular Microbiology. 2006, 9: 1046-1416.

2. Yan Ge, Kiyotaka Yoshiie, Futoshi Kuribayashi, Mingqun Lin and Yasuko Rikihisa. Anaplasma phagocytophilum inhibits human neutrophil apoptosis via upregulation of bfl-1, maintenance of mitochondrial membrane potential and prevention of caspase 3 activation. Cellular Microbiology. 2005, 7: 29-38.

vii

FIELDS OF STUDY

Major Field: Veterinary Biosciences

viii

TABLE OF CONTENTS

Page Abstract...... ii

Dedication ...... v

Acknowledgments ...... vi

Vita ...... vii

List of Tables...... xi

List of Figures ...... xii

Chapters:

1. Introduction ...... 1

2. Anaplasma phagocytophilum inhibits human neutrophil apoptosis via upregulation of bfl-1, maintenance of mitochondrial membrane potential and prevention of caspase 3 ctivation ...... 25

2.1 Abstract ...... 25 2.2 Introduction ...... 26 2.3 Materials and Methods...... 28 2.4 Results...... 34 2.5 Discussion ...... 40

3. Anaplasma phagocytophilum delays spontaneous human neutrophil apoptosis by modulation of multiple apoptotic pathways ...... 52

3.1 Abstract ...... 52 3.2 Introduction ...... 53 3.3 Materials and Methods...... 56 3.4 Results...... 61 3.5 Discussion ...... 66

ix 4. Identification of novel surface proteins of Anaplasma phagocytophilum by affinity purification and proteomics ...... 83

4.1 Abstract ...... 83 4.2 Introduction ...... 84 4.3 Materials and Methods...... 87 4.4 Results...... 94 4.5 Discussion ...... 100

5. Surfaceome of Ehrlichia chaffeensis ...... 115

5.1 Abstract ...... 115 5.2 Introduction ...... 116 5.3 Materials and Methods...... 118 5.4 Results...... 124 5.5 Discussion ...... 130

Bibliography...... 146

x

LIST OF TABLES

Table Page

4.1 Surface-exposed proteins of A. phagocytophilum purified by biotinylation identified by Nano-LC/MS/MS ...... 104

5.1 Surface-exposed proteins of E. chaffeensis purified by biotinylation and identified by Nano-LC/MS/MS ...... 137

xi

LIST OF FIGURES

Figure Page

2.1 Inhibition of the reduction of bfl-1 transcript levels in human neutrophils by A. phagocytophilum infection. RNA was isolated from infected or uninfected neutrophils, and competitive-PCR was performed using primers against bfl-1text ...... 44

2.2 Inhibition of loss of mitochondrial staining with Mitotracker Red in human neutrophils by A. phagocytophilum infection ...... 47

2.3 Prevention of the loss of mitochondrial membrane potential in human neutrophils by the infection with A. phagocytophilum as determined by flow cytometry using JC-1 staining ...... 48

2.4 Inhibition of spontaneous apoptosis and caspase 3 enzyme activity in human neutrophils by A. phagocytophilum in a dose-dependent manner ...... 50

2.5 Prevention of generation of activated caspase 3 in human neutrophils by A. phagocytophilum infection ...... 51

3.1 Inhibition of Fas capping during spontaneous human neutrophil apoptosis by A. phagocytophilum infection ...... 71

3.2 Inhibition of activation of caspase 8 and cleavage of Bid in neutrophils by A. phagocytophilum infection ...... 73

3.3 Inhibition of anti-Fas IgM-induced human neutrophil apoptosis by A. phagocytophilum infection ...... 75

3.4 Inhibition of colocalization of Bax with mitochondria during spontaneous human neutrophil apoptosis by A. phagocytophilum infection ...... 77

xii 3.5 Inhibition of caspase 9 activation in spontaneous neutrophil apoptosis by A. phagocytophilum infection ...... 79

3.6 Inhibition of cleavage of XIAP in neutrophils infected with A. phagocytophilum ...... 80

3.7 Model for the inhibitory mechanism of human neutrophil apoptosis by A. phagocytophilum infection ...... 81

4.1 Biotin labeling of A. phagocytophilum surface proteins ...... 106

4.2 Streptavidin agarose affinity-purification of Sulfo-NHS-SS-Biotin- labeled A. phagocytophilum surface proteins ...... 107

4.3 The protein structure prediction based on PRED-TMBB for Asp62 and Asp55 of A. phagocytophilum ...... 108

4.4 Schematic diagram of the organization of genes encoding the APH_0404 (Asp62), APH_0405 (Asp55) and APH_0406 in A. phagocytophila HZ and the orthologus genes in A. marginale str. St. Maries, E. chaffeensis Arkansas, E.canis Jake and E. ruminantium Welgevonden ...... 111

4.5 Co-transcriptional analysis of Asp62 and Asp55 by RT-PCR ...... 112

4.6 Surface localization of A. phagocytophilum Asp62 and Asp55 by imunofluorescence assay ...... 113

4.7 Antigenicity of streptavidin agarose affinity-purification of Sulfo-NHS- SS-Biotin-labeled A. phagocytophilum surface proteins in human patients and an experimentally infected horse ...... 114

5.1 Biotin labeling of E. chaffeensis surface proteins ...... 139

5.2 Streptavidin agarose gel affinity purification of Sulfo-NHS-SS-Biotin- labeled E. chaffeensis surface proteins ...... 140

5.3 Diagram of protein expression of E. chaffeensis OMP-1/P28 multigene family identified by Nano-LC/MS/MS...... 141

5.4 2-D structure prediction of E. chaffeensis OMP-1A (A) and OMP-1N (B) with respect to the outer membrane lipid bilayer using Posterior Decoding method with the dynamic programming algorithm in PRED- TMBB; Surface localization of E. chaffeensis OMP-1A and OMP-1N by imunofluorescence assay (C) ...... 142

xiii 5.5 Western blot analysis of OMP-1B expression by E. chaffeensis cultured in THP-1 cells using anti-OMP-1B peptide rabbit serum...... 144

5.6 Western blot analysis of antigenicity of E. chaffeensis surface proteins purified by biotinylation method ...... 145

xiv

CHAPTER 1

INTRODUCTION

Taxonomy of family Anaplasmataceae

The organisms in the family Anaplasmataceae are a group of small Gram-

negative pleomorphic cocci. They are obligate intracellular bacteria, which replicate in

membrane-bound vacuoles (parasitophorous vacuoles) in the cytoplasm of eukaryotic

host cells, which are usually bone marrow or haematopoietic origin (Rikihisa, 1991).

Based on the alignments of 16S rRNA gene and groESL sequences, the family

Anaplasmataceae belongs to class α-Proteobacteria, order Rickettsiales (Dumler et al.,

2001). This family encompasses five genera, Anaplasma, Ehrlichia, Neorickettsia,

Aegyptianella and Wolbachia. Genus Anaplasma has seven species, A. phagocytophilum

(Bakken et al., 1994; Chen et al., 1994), A. marginale, A. centrale, A. bovis, A. ovis, A. equi and A. platys. Genus Ehrlichia includes the species of E. chaffeensis (Maeda et al.,

1987), E. ewingii, E. muris, E. canis, E. ovis and E. ruminantium. Genus Neorickettsia consists of N. risticii, N. helminthoeca and N. sennetsu. Genus Aegyptianella has only one species, A. pullorum, which is recently charcacterized, displaying closest phylogenic relationship to Anaplasma spp. (Rikihisa et al., journal of clinical microbiology, 2003).

Genus Wobachia has one named species, W. pipientis. Genus Wolbachia infects the ovaries of many species of arthropods and helminthes (Dumler et al., 2001).

1 Ecology of family Anaplasmataceae

Genera Anaplasma and Ehrlichia have been recently recognized as emerging

human pathogens in the United States as well as other countries (Demma et al., 2005).

The life cycle of genera Anaplasma and Ehrlichia are horizontally maintained alternatively between ticks and mammals via tick vectors (Rikihisa, 2003). These organisms are transstadially instead of transovarially transmitted by tick vectors. Tick infection occurs after bloodsucking of infectious animals. The humans can be accidentally infected through the bite of infected ticks. A. phagocytophilum are mainly transmitted by Ixodes ticks in North America. The mammalian reservoirs for A. phagocytophilum are wild rodents, e.g. white-footed mouse (Peromyscus leucopus) and dusky-footed wood rats (Neotoma fuscipes) in the eastern and western United States, respectively (Rikihisa, 2003). E. chaffeensis is usually found in the Lone Star tick,

Amblyomma americanum. White-tailed deer are considered to be the major reservoir of

E. chaffeensis.

The genus Neorickettsia is maintained through transovarial and transstadial passage in the trematode species, which is specific to each Neorickettsia species. Genus

Neorickettsia can be horizontally transmitted unidirectionally from trematodes to mammals, but not vice versa (Rikihisa, 2003). Mammals are required for the maintenance of trematode lifecycle for some Neorickettsia species, but the infection of mammals with

Neorickettsia itself is not required for maintenance of Neorickettsia (Rikihisa, 2003).

Human and other animal infection with Neorickettsia occur when they accidentally ingest infected trematodes in fish or aquatic insects. The genus Wobachia is an endosymbiont of

2 arthropods, nematodes and helminthes. It is transovarially transmitted in these invertebrate hosts. There is no evidence showing that Wobachia can directly infect vertebrates.

Cell structure and metabolism of A. phagocytophilum and E. chaffeensis

A. phagocytophilum and E. chaffeensis are small cocci (0.4 µm to 1.5 µm) and stained as dark blue to purple with Romanowsky stain. They often form mulberry-like microcolony, called morulae (in Latin). By transmission electron microscopy, these bacteria are generally round but sometime pleomorphic, especially in tissue culture

(Rikihisa, 1991). They replicate in membrane-bound vacuoles in the cytoplasm of eukaryotic host cells. Genera Anaplasma and Ehrlichia have thin bileaflets of outer and inner membranes. Morphologically, it seems that there are little amounts of peptidoglycan in the outer membrane (Rikihisa, 1991). A. phagocytophilum and E. chaffeensis have lost most genes required for the biosynthesis of peptidoglycan and all genes required for the biosynthesis of LPS (Hotopp et al., 2006; Lin and Rikihisa,

2003a). However, A. phagocytophilum and E. chaffeensis acquire unique capability to take up host cell cholesterol and incorporate cholesterol into bacterial membrane (Lin and

Rikihisa, 2003a). Cholesterol has become indispensable for their survival and successful establishment of infection (Lin and Rikihisa, 2003a). These bacteria have distinct ribosomes and a fine meshwork of DNA strands. Clumps of ribosomes are homogenously distributed in the cytoplasm instead of marginated beneath inner membrane (Rikihisa,

1991). The genome sizes of family Anaplasmataceae are small; for example, those of A. phagocytophilum and E. chaffeensis are 1.47 Mb and 1.17 Mb, respectively, which are about one fourth of Escherichia coli (Hotopp et al., 2006).

3 A. phagocytophilum and E. chaffeensis have the ability to synthesize all

nucleotides (Hotopp et al., 2006). This differs from Rickettsia prowazekii, which can not

make purines or pyrimidines and therefore, must rely on nucleotide translocases and

interconversion of the bases to obtain the full complement of nucleotides (Andersson et al., 1998). A. phagocytophilum and E. chaffeensis are able to synthesize most vitamins and cofactors (Hotopp et al., 2006). The presence of nucleotide, vitamin and cofactor biosynthetic pathway in A. phagocytophilum and E. chaffeensis suggests that they do not

need to compete with the host cell for essential vitamins and nucleotides (Hotopp et al.,

2006). However, the members of family Anaplasmataceae have a very limited ability to

synthesize amino acids and must rely on transporting them from host (Hotopp et al.,

2006). They can make glycine, glutamine, glutamate, and aspartate (Hotopp et al., 2006).

They can utilize glutamine and glutamate to generate adenosine triphosphate (ATP) as

genus Rickettsia, and they cannot utilize glucose-6-phophate or glucose (Rikihisa, 1991).

For the greatest metabolic activity for these bacteria, the optimal pH is at pH 7.2 to 8.0

(Rikihisa, 1991).

A. phagocytophilum and human granulocytic anaplasmosis

Human granulocytic anaplasmosis (HGA, formally human granulocytic

ehrlichiosis) was first recognized in 1990 from a patient who died with a severe febrile

disease two weeks after a tick bite in the United States (Dumler et al., 2005; Bakken et

al., 1994; Chen et al., 1994). In the final stage of infection, bacterial morulae were

observed in neutrophils (Dumler et al., 2005). In 1994, the causative agent of HGA, A.

phagocytophilum (formally the HGE agent) was identified using DNA sequencing (Chen

4 et al., 1994) in the United States and subsequently reported in Europe (Parola, 2004) and

Asia (Kawahara et al., 2006). A. phagocytophilum has become the predominant form of

anaplasmosis and among the most common tick-borne pathogens in the United States and

Europe (Parola, 2004; Massung and Slater, 2003). The clinical signs for HGA include

fever, headache, myalgia, malaise, absence of skin rash, leucopenia, thrombocytopenia,

increased amounts of C-reactive protein and abnormal activities of hepatic transaminase.

It can cause severe and potentially fatal disease in immunocompromised and elderly

people. Recent seroepidemiologic data suggest that as much as 15% to 36% of the

population has been infected in endemic areas (Dumler et al., 2005). Symptomatic infection often occurs in tick-endemic regions and varies in severity from mild, fever to death (Dumler et al., 2005). Half of symptomatic patients need hospitalization and approximately 5% to 7% patients require intensive care (Dumler et al., 2005).

Pathogenicity, but not infectivity, waned with mouse passage but could be resurrected by severe combined immunodeficiency disease (SCID) mouse passage, suggesting that reversible posttranscriptional modification of bacterial virulence factor or epigenetic modification (Hodzic et al., 1998).

To confirm the clinical diagnosis, the laboratory diagnostic methods of HGA,

polymerase chain reaction (PCR) and indirect fluorescent antibody test (IFA), are used

most frequently. 16S rRNA gene sequence PCR is good for detection of bacteria in the

acute phase of disease, whereas IFA is used for diagnosis during relatively late phase of

disease. Light microscopic examination of peripheral blood smears may reveal morulae in

the cytoplasm of neutrophils during the acute phase and thus provide immediate evidence

supporting the diagnosis (Bakken and Dumler, 2000). The success in finding morulae

5 varies directly with the experience of the microscopist and also with the duration of disease, since morulae tend to be detected less frequently after the first week of disease

(Bakken and Dumler, 2000). However, the absence of morulae in the peripheral blood

smear does not exclude the diagnosis of HGA for individuals who present with a non- specific febrile illness and have a history of recent tick exposure (Bakken and Dumler,

2000). Though the isolation of bacteria is ideal, it is too long to culture isolate relative to

the rapid development of disease, so that it is impractical in clinical laboratories

(Rikihisa, 2000). For antibiotic treatment, tetracycline and doxycycline are used to treat

the disease via intravenously or orally administration.

Pathogenesis of human granulocytic anaplasmosis

Inhibition of apoptosis

Apoptosis is one of the important defense mechanisms for phagocytes to kill

intracellular pathogens. The host cells induce apoptosis in the presence of several

pathogens, whereas some pathogens are known to inhibit host cell apoptosis (DeLeo,

2004). HGA is associated with the delay of apoptosis following the infection of

neutrophils with the obligatory intracellular bacterium, A. phagocytophilum, which has a

tropism for granulocytes. Yoshiie et al. previously reported that the time required for

50% of the infected neutrophils to show morphological apoptosis is 45.0 ± 9.8 h, which is

much longer than that of uninfected neutrophils, 12.2 ± 2.5 h (Yoshiie et al., 2000).

Therefore, the inhibition of neutrophil apoptosis by A. phagocytophilum provides the

bacteria with sufficient time to complete the developmental cycle in neutrophils. A recent

study corroborates this finding by showing that in vivo infection with an A.

6 phagocytophilum sheep isolate also reduced peripheral blood neutrophil apoptosis (Scaife

et al., 2003). Some mechanisms by which A. phagocytophilum inhibits the spontaneous apoptosis of human neutrophils have been investigated (Yoshiie et al., 2000). Binding of protein components of A. phagocytophilum to neutrophils and subsequent transglutaminase-dependent cross-linking and/or internalization of the receptor and bacteria are required for anti-apoptotic signalling. This study (Yoshiie et al., 2000) shows that this process does not require bacterial carbohydrates, bacterial new protein synthesis, host cell protein kinase A, nuclear translocation of NF-κB or IL-1β which are known to

be involved in the inhibition of neutrophil's apoptosis by some other agents (Ward et al.,

1999; Parvathenani et al., 1998; Watson et al., 1998).

Receptor-mediated internalization and creation of a replicative compartment not fused

with lysosomes

It has been reported that the monoclonal antibodies against the P-selectin

glycoprotein ligand-1 (PSGL-1) inhibit the binding of A. phagocytophilum to human

neutrophils or HL-60 cell (Herron et al., 2000). Furthermore, besides PSGL, the

expression of α 1, 3-fucosyltransferese is also required for bacterial binding (Herron et al., 2000). Whereas A. phagocytophilum binding to murine neutrophils requires the expression of α 1, 3-fucosyltransferese, which constructs the glycan determinant sialyl

Lewis x (sLex), but not PSGL-1(Carlyon et al., 2003). This discrepancy could be

explained at least in part by the facts that the amino acid sequences of human and murine

PSGL-1s are different, and that A. phagocytophilum binds only to fucosylated glycans in

murine cells, but cooperatively binds to PSGL-1 and to sLex expressed on PSGL-1 or

other glycoproteins in human neutrophils (Yago et al., 2003). Recently, A.

7 phagocytophilum organisms that do not rely on sialic acid for cellular adhesion and entry have been enriched by maintaining them in severely undersialylated HL-60 cells. The selected bacteria, termed NCH-1A, also exhibit lessened dependencies on PSGL-1 and a

1, 3-fucose. Optimal adhesion and invasion by NCH-1A require interactions with the known determinants (sialic acid, PSGL-1 and a 1, 3-fucose), but none of them is absolutely necessary. NCH-1A binding to sLex -modified PSGL-1 requires recognition of the known determinants in the same manner as other A. phagocytophilum strains. These data suggest that A. phagocytophilum expresses a separate adhesin from those targeting sialic acid, a 1,3-fucose and the N-terminal region of PSGL-1. It has been proposed that

NCH-1A upregulates expression of this adhesin (Reneer et al., 2006).

A variety of pathogens use caveolae-lipid raft mediated endocytosis to enter host cells, which may play a role in bypassing phagolysosmal pathways (Lafont and van der

Goot, 2005). The entry and intracellular infection of A. phagocytophilum involves caveolae and glycosylphosphatidylinositol-anchored proteins (GPI) (Lin and Rikihisa,

2003b). Instead, clathrin, clathrin-coated vesicles are one of major means that plasma membrane proteins and lipids are endocytosed (Kirchhausen, 2000), is not associated with the internalization of A. phagocytophilum (Lin and Rikihisa, 2003b). Caveolin-1 and tyrosine-phosphorylated proteins, including phospholipase C (PLC)-γ2 are colocalized with both early and late replicative inclusion of A. phagocytophilum (Lin et al., 2002).

The internalization of A. phagocytophilum requires the transglutaminase activity of host cell (Lin et al., 2002). Recently, it has been reported that the expression of PLC-β1, TG3-

like and Tec protein kinase genes are transcriptionally upregulated in A.

phagocytophilum-infected HL-60 cells (de la Fuente et al., 2005). These indicate that A.

8 phagocytophilum actively modulate the host cell proteins essential for their entry and

creation of surviving inclusion niche at both the transcriptional and post-translational

levels.

After internalization, it resides inside harsh niche of phagocytic host cells,

implying its capability to subvert phagolysosome biogenesis of host cells. A number of

endocytic markers including several early endosomal markers, such as a cytoplasmic

small GTPase, Rab5, vacuolar-type H+ ATPase and early endosomal antigen 1, have been

found excluded from A. phagocytophilum inclusion (Mott et al., 1999). The inclusion

compartments are negative for lysosomal marker proteins, such as lysosomal membrane-

associated protein and CD63, or acid phosphatase activity (Mott et al., 1999). The CI-

M6PRc, which delivers newly synthesized, soluble lysosomal enzymes to the

prelysosomes from the trans-Golgi network, is not colocalized with A. phagocytophilum

inclusion (Mott et al., 1999). Endogenously synthesized sphingomyelin in Golgi is not incorporated into inclusion. Brefeldin A, a fungal metabolite inhibiting antegrade (but not retrograde) transport of the Golgi complex, did not affect the growth of A.

phagocytophilum in HL-60 cells (Mott et al., 1999). A. phagocytophilum inclusion avoids

lysosomal fusion through excluding itself from host cell endocytic and exocytic vesicular

traffic pathways (Rikihisa, 2006).

Lysosomes, however, are found close to A. phagocytophilum inclusions under

electron microscope (Mott et al., 1999). For intracellular pathogens living inside of

membrane-bound compartments, there are a limited number of strategies which can be

used to get nutrients from host cells. These strategies depend, in part, on the intracellular

niche where pathogens live. Toxoplasma gondii selectively recruit endo-lysosomes to the

9 PV define an unanticipated process allowing the parasite intimate and concentrated

access to a diverse range of low molecular weight components produced by the endo-

lysosomal system (Coppens et al., 2006). Therefore, A. phagocytophilum may exploit

similar mechanism as T. gondii to acquire nutrients from host cells.

Interference of assembly of reduced nicotinamide adenine dinucleotide phosphate

(NADPH) oxidase, inhibition of ROS production

Neutrophils are one of principal phagocytes serving as first line of defense against

pathogens, which function in part through a powerful oxygen-dependent defense system

- that generates reactive oxygen speciess (ROS), such as superoxide radicals (O2 ), hydrogen peroxide and hydroxyl radicals, which are highly reactive oxidizing agents that destroy microbes. After receptors, such as Toll-like receptors (TLRs), Fc and complement 3 (C3) receptors, G protein-coupled receptors or cytokine receptors, recognize microbes, neutrophils are activated. Upon activation, neutrophils have increased oxygen consumption known as respiratory burst, during which oxygen is converted by phagocyte oxidase into ROS, which will kill ingested microbes.

The function of phagocyte oxidase is to reduce molecular oxygen into ROS with the reduced form of NADPH as an electron donor. So, this phagocyte oxidase is called

NADPH oxidase. NADPH oxidase is a complex enzyme composed of multiple subunits,

phox phox phox phox phox such as p21 , gp91 (a membrane bound cytochrome b558), p47 , p67 , p40 and a small GTPase Rac (cytosolic proteins). In resting neutrophils, these subunits are dissociated and inactive. When neutrophils are activated, cytosolic components of

NADPH oxidase translocate to the membrane and assemble with membrane bound subunits into active form of NADPH oxidase, which are in the phagolysosomal

10 membrane (mainly) and/or plasma membrane , allowing exertion of the lethal effects of

- O2 and its derivatives on ingested bacteria or extracellular in close proximity. A.

phagocytophilum interferes with the assembly of the NADPH oxidase subunits in the

inclusion membrane (Mott and Rikihisa, 2000), and subsequently block the activation of

NADPH oxidase induced by phorbol myristic acetate, N-formyl-methionyl-

leucylphenylalanine or Escherichia coli (Wang et al., 2004a; Mott et al., 2002; Mott and

Rikihisa, 2000). However, the generation of ROS by pre-activated or primed neutrophils

upon addition of exogenous stimuli can not be overridden by A. phagocytophilum (Mott

and Rikihisa, 2000). A. phagocytophilum inhibits the generation of ROS specifically in

neutrophils instead of monocytes (Mott and Rikihisa, 2000). Later on, it has been

confirmed by a series of study that A. phagocytophilum does not induce the activation of

NADPH oxidase in human or murine neutrophils (Borjesson et al., 2005; Carlyon et al.,

2004; JW and Mueller, 2004; Wang et al., 2004a; Mott et al., 2002).

A periodate oxidation-sensitive A. phagocytophilum component, but not in

bacterial protein or viability, is required for the inhibition of neutrophil ROS generation,

and bacterium-cell contact is also essential for the inhibition of ROS generation (Mott et

al., 2002; Mott and Rikihisa, 2000). These data suggests that A. phagocytophilum surface

carbohydrate moiety is responsible for the inhibition of neutrophil ROS generation. In

human neutrophils and HL-60 cells, A. phagocytophilum decreases the protein levels of

p22phox, but not other components of NADPH oxidase (Mott et al., 2002), suggest that the rapid destabilization of the gp91phox and p22phox complex may be involved in the inhibition of superoxide generation. Based on recent microarray data analyses using human neutrophils infected with A. phagocytophilum, there are no significant changes in

11 the transcription of the NADPH oxidase components (Borjesson et al., 2005). A. phagocytophilum has an iron-containing homolog of cytoplasmic SOD, which scavenges intracellular superoxide (Ohashi et al., 2002). A. phagocytophilum mechanically released

from infected host cells was reported to partially scavenge exogenous superoxide, which,

unlike hydrogen peroxide, can not readily diffuse through an intact bacterial inner

membrane (Carlyon et al., 2004).

Cell surface proteins of A. phagocytophilum

The family Anaplasmataceae members have diverse outer membrane proteins

(OMPs). Many of these membrane proteins are members of Pfam PF01617 and constitute

the OMP-1/MSP2/P44 superfamily (Hotopp et al., 2006). The genera Anaplasma and

Ehrlichia have a large expansion of this OMP-1/MSP2/P44 protein family. The A.

phagocytophilum genome has three omp-1, one msp2, two msp2 homologs, one msp4,

and 113 p44 loci belonging to the OMP-1/MSP2/P44 protein family (Hotopp et al.,

2006). msp2 in A. phagocytophilum is distinct from that in A. marginale (Lin et al.,

2004). p44 genes are the largest expansion of this protein family in A. phagocytophilum.

The p44s consist of a central hypervariable region of approximately 280 bp and

conserved flanking sequences from 100 to 500 bp. So far, P44-18 is the only A.

phagocytophilum protein that has been shown as being surface-exposed by

immunoelectron (Kim and Rikihisa, 1998) and immunofluorescence microscopy (Wang

et al., 2006). These polymorphic major outer membrane proteins can be differentially

expressed upon different signals, e.g. temperature and host species (Zhi et al., 2002). For

example, the transcription of the tick salivary gland-specific p44 genes of A.

phagocytophilum is up-regulated in culture at low temperature of 24 °C compared to

12 37°C (Zhi et al., 2002). One of the p44s is dominantly expressed by A. phagocytophilum in mammals at an early stage of tick transmission (Zhi et al., 2002). Experimental evidence of within-host p44 antigenic variation was also provided, suggesting that the rapid and synchronized switch of expression is an intrinsic property of p44s reinitiated after transmission to naïve mammalian hosts and shaped upon exposure to immune plasma (Wang et al., 2004a). These antigenic variations may benefit bacteria to adapt different host environments, such as ticks and mammals, and it may also help bacteria to avoid host immune response. For the particular mammal dominant p44 expression, a unique mRNA splicing mechanism has been proposed (Lin et al., 2003), which occurs through the RecF pathway (Lin et al., 2006). Recently, P44 has been shown to have porin function (Haibin Huang, 2006).

Immunopathogenicity

HGA is systemic febrile disease with abnormal leukocyte counts and liver enzyme activity. There is no endo- or exo-toxin activities ever detected (Rikihisa et al.,

2003). The mismatch between low bacterial load and histophathologic changes with

HGA suggests that pro-inflammatory cytokines is responsible for the tissue damage.

Instead of loss of expression, delayed expression of pro-inflammatory genes, e.g. TNF-α,

IL-1β and IL-6 are upregulated in human neutrophils infected with A. phagocytophilum

(Borjesson et al., 2005). These cytokines can also be induced in monocytes by A. phagocytophilum (Kim and Rikihisa, 2002). Although IL-1β generation is detected, there is no detectable activation of p38 MAPK and NF-κB in human neutrophils after A. phagocytophilum infection (Kim and Rikihisa, 2002). There are no significantly altered expressions of TLRs or myeloid differentiation primary response gene (Myd88) in A.

13 phagocytophilum-infected neutrophils (Borjesson et al., 2005). The activation of innate

immune responses through TLR2 or TLR4 does not contribute to the control of A.

phagocytophilum infection (Choi et al., 2004). At the mean time, there is no definitive

TLR signaling pathway described to be involved in A. phagocytophilum infection of

neutrophils. The protein amount of the hematopoietic system-specific transcription factor

PU.1, which acts specifically at the stage of promyelocyte differentiation into neutrophils

and monocytes, is reduced in A. phagocytophilum-infected HL-60 cells (Thomas et al.,

2005).

There is a discrepancy of the role of IFN-γ between in vivo and in vitro A.

phagocytophilum infection. A murine model shows a role of IFN-γ in histopathology and

restriction of infection, which is confirmed in IFN-γ knockout mice (Martin et al., 2001).

However, there is a defect in IFN-γ signaling in in vitro HL- 60 cell culture, which impairs the binding of phosphorylated Stat1 to the promoter of IFN regulatory factor-1, although it does not inhibit Stat1 phosphorylation (Thomas et al., 2005). The discrepancy of the role of IFN-γ may be due to the responses from different type of cells other than bacterial host neutrophils in vivo. During progressive levels of infection, it has been shown that A. phagocytophilum infection does not change the transcription of TfR gene or the activity of iron-responsive factor in human myelocytic leukemia THP-1 cells

(Barnewall et al., 1999). A microarray analysis found that upregulation of transferrin receptor (TfR) and downregulation of Tf in NB4 cells after 4 h of incubation with A.

phagocytophilum (Reneer et al., 2006; Pedra et al., 2005). Ferritin heavy chain mRNA

expression is increased and ferritin protein amount is decreased in A. phagocytophilum-

14 infected HL-60 cells (Carlyon et al., 2005). Taken together, these data suggest that A. phagocytophilum infection has an effect on IFN- γ signaling and alters host cell iron metabolism.

E. chaffeensis

E. chaffeensis and human monocytic ehrlichiosis

E. chaffeensis is the etiologic agent of human monocytic ehrlichiosis (HME).

HME is first described in the United States in 1987 by Maeda et al., in a patient who had clinical symptoms resembling those of patients with Rocky Mountain spotted fever (Maeda et al., 1987). In 1990, Dawson et al. first isolated the organism, E. chaffeensis Arkansas strain, a new monocytotropic ehrlichial species (Dawson et al.,

1991). E. chaffeensis infects specifically host monocyte/macrophage. Although the symptoms of HME and HGA are very similar, they can be distinguished by diagnostic

PCR based on 16SrRNA genes or IFA against specific bacterial antigens.

Internalization and intracellular trafficking of E. chaffeensis

Like A. phagocytophilum, the entry and establishment of infection of E. chaffeensis also involves caveolae, host GPI-anchored proteins and the incorporation of

2+ cholesterol into bacterial membrane. An increase in host cytosolic free calcium ([Ca( )]i)

is essential for E. chaffeensis entry into monocytes (Lin et al., 2002). The following

sequential signaling events were described to be induced by E. chaffeensis infection: host cell protein cross-linking by transglutaminase, tyrosine phosphorylation, PLC-γ2

2+ activation, IP3 production, and an increase in [Ca( )]i (Lin et al., 2002).

15 After internalization of host cell, E. chaffeensis also has the ability to exclude its

membrane-bound compartment from fusion with lysosomes. Despite sharing some

signaling features, E. chaffeensis and A. phagocytophilum inclusions are distinct from

each other, and they never co-localize with in same compartments even after coinfection

of the same HL-60 cells (Mott et al., 1999). Replicative E. chaffeensis inclusions are

weakly acidic, early endosomal compartments. They are weakly co-localized with

vacuolar-type H+-ATPase, and IFA labeling show the inclusion positive for the TfR, early endosomal antigen 1, and Rab5 (Mott et al., 1999).

Iron acquisition by E. chaffeensis is dependent on the labile iron pool in the host

cytosol. When monocytes are treated with deferoxamine, an intracellular iron chelator,

the infection of E. chaffeensis is completely blocked (Barnewall and Rikihisa, 1994).

Since E. chaffeensis inclusions accumulate TfR, the bacteria may directly acquire iron

from cytosolic transferrin. Interferon-γ (IFN- γ) has a protective role in mice infected E.

chaffeensis (Bitsaktsis et al., 2004; Ismail et al., 2004). The anti-ehrlichia activity

induced in human monocytes by exogenous IFN- γ is mediated by the limitation of

available cytosolic iron instead of generation of ROS or nitric oxide (Bakken et al.,

1994). E. chaffeensis infection up-regulates host cell TfR mRNA expression (Barnewall

et al., 1999). Both internalization and continuous proliferation of ehrlichial organisms or

the synthesis of ehrlichial proteins are required for the up-regulation of TfR mRNA

(Barnewall et al., 1999). The activation of iron-responsive protein 1 (IRP-1) to the iron-

responsive element and subsequent stabilization of TfR mRNA comprise the mechanism

of TfR mRNA up-regulation by E. chaffeensis (Barnewall et al., 1999).

16 E. chaffeensis surface proteins

It has been showed that E. chaffeensis OMP-1/MSP2/P44 family encodes

approximately 30-kDa major outer membrane proteins (OMPs) (Ohashi et al., 2001).

Upstream from secA and downstream of hypothetical transcriptional regulator,

22 paralogs of the omp gene family are tandemly arranged except for one or two genes with opposite orientations in a 27-kb locus in the E. chaffeensis genome. These genes

have signal peptide and are likely to be secreted across the cytoplasmic membrane by

SecA (Ohashi et al., 2001). The locus consists of three highly repetitive regions with four

non-repetitive intervening regions. These outer membrane proteins are also differentially

expressed in ticks and experimentally infected dogs (Unver et al., 2002). Recently, the

differential expression of this family in ticks and mammals has been confirmed and

further studied using proteomic approaches (Singu et al., 2006). The differentially

expressed proteins are also post-translationally modified by phosphorylation and

glycosylation to generate multiple expressed forms (Singu et al., 2006). And, the host

cell-specific protein expression is not influenced by growth temperatures and is

reversible. Host cell-specific protein expression coupled with posttranslational

modifications may be an evolutionary mechanism for E. chaffeensis’s adaptation to a

dual-host life cycle and its persistence (Singu et al., 2006).

Objectives of this study are:

Objective 1

Since neutrophils typically undergo apoptosis 6 – 12 h after their release from the

bone marrow (Akgul et al., 2001; Edwards, 1994), modulation of the apoptotic time

course by pathogens has been implicated as a critical factor in the severity and duration

17 of systemic and local infections (Kobayashi et al., 2003; Laskay et al., 2003). HGA is associated with the delay of apoptosis following the infection of neutrophils with the obligatory intracellular bacterium, A. phagocytophilum, which has a tropism for granulocytes. The mechanisms by which A. phagocytophilum inhibits the spontaneous apoptosis of human neutrophils have been investigated (Yoshiie et al., 2000). Binding of protein components of A. phagocytophilum to neutrophils and subsequent transglutaminase-dependent cross-linking and/or internalization of the receptor and bacteria are required for anti-apoptotic signalling. This study (Yoshiie et al., 2000) shows that this process does not require bacterial carbohydrates, bacterial new protein synthesis, host cell protein kinase A, nuclear translocation of NF-κB or IL-1β which are known to be involved in the inhibition of neutrophil's apoptosis by some other agents (Ward et al.,

1999; Parvathenani et al., 1998; Watson et al., 1998). Despite these data, intracellular signalling pathways leading to the inhibition of human neutrophil apoptosis by A. phagocytophilum remain largely uncharacterized.

Since the mitochondria-mediated apoptotic pathway is the major pathway in neutrophil spontaneous apoptosis (Maianski, NA. 2004), we hypothesize that it is involved in the inhibition of neutrophil spontaneous apoptosis by A. phagocytophilum.

Caspases, a family of cysteine proteases, play a critical role during apoptosis in many nucleated cells (Nicholson, 1999; Salvesen and Dixit, 1997). The activation of caspases is controlled by the bcl-2 family of intracellular proteins, which consists of anti-apoptotic and pro-apoptotic proteins (Cory and Adams, 2002) that integrate diverse survival and death signals to determine whether apoptosis occurs (Borner, 2003). A major function of the bcl-2 family is either to maintain (anti-apoptotic bcl-2 members) or to increase (pro-

18 apoptotic bcl-2 members) mitochondrial membrane permeability, which determines the

release of mitochondrial apoptotic factors, and subsequent activation of downstream

effector caspases. Thus, the goal of the present study was to investigate whether bcl-2

family members, caspases and mitochondria play a role in the inhibition of human

neutrophil apoptosis by A. phagocytophilum.

The results have shown that A. phagocytophilum inhibits the mitochondria-

mediated pathway of the spontaneous neutrophil apoptosis by protecting the

mitochondrial membrane integrity. The decrease of Bfl-1, an antiapoptotic Bcl-2 family

member, and activation of the major executioner caspase, caspase 3, during spontaneous

neutrophil apoptosis are inhibited by A. phagocytophilum infection. The present data extend on previous reports by demonstrating that A. phagocytophilum blocks this primary

spontaneous apoptotic pathway in human neutrophils and provide a greater understanding

of the pathogenesis of HGA and the regulation of apoptosis in neutrophils.

Objective 2

Signalling pathways leading to apoptosis are generally classified into two

categories: the extrinsic pathway (the death receptor pathway) and the intrinsic pathway

(the mitochondrial pathway), both of which have the same outcome: the activation of a

characteristic caspase cascade (Sprick and Walczak, 2004). In the extrinsic pathway,

stimulation of death receptors [e.g. Fas, tumour necrosis factor (TNF)-α receptor or TNF-

related apoptosis-inducing ligand (TRAIL) receptor] with their ligands (e.g. Fas ligand,

TNF-α or TRAIL) allows the formation of a death-inducing signalling complex (DISC),

which contains an adaptor protein [e.g. Fas-associated death domain (FADD)] and an

19 initiator caspase (mainly caspase 8). The autocleavage of caspase 8 in the DISC produces

active caspase 8 that initiates downstream apoptotic signalling (Donepudi et al., 2003). In the intrinsic pathway, when cells receive apoptotic stimuli, the mitochondrial outer membrane is permeabilized by pro-apoptotic Bcl-2 family members. This results in the release of apoptotic factors and loss of mitochondrial inner membrane potential (∆Ψm)

(Antonsson, 2004). Both the extrinsic and intrinsic pathways may converge at the level of mitochondria (Sprick and Walczak, 2004). During spontaneous neutrophil apoptosis, without ligation of external ligands, Fas clustering occurs through an unknown mechanism (Scheel-Toellner et al., 2004; Simon, 2003; Daigle and Simon, 2001).

Despite the progress in our understanding of spontaneous neutrophil apoptosis, mechanisms by which it is initiated and regulated are largely uncharacterized.

We have demonstrated that A. phagocytophilum inhibits the intrinsic pathway of the spontaneous neutrophil apoptosis by protecting the mitochondrial membrane integrity

(Ge et al., 2005). However, whether and how death receptor-mediated spontaneous human neutrophil apoptosis pathway is inhibited by A. phagocytophilum is still unknown.

In the present study, we investigated the signalling events in the extrinsic pathway and the cross-talk between the intrinsic and extrinsic pathways in the inhibition of spontaneous human neutrophil apoptosis by A. phagocytophilum infection. Our observation that agonistic anti-Fas antibody-induced neutrophil apoptosis is prevented by

A. phagocytophilum was also reported but not further investigated by Borjesson et al.

(Borjesson et al., 2005). Therefore, the present study included the signalling events in the

agonistic Fas antibody-induced human neutrophil apoptosis to contrast with spontaneous

human neutrophil apoptosis.

20 Objective 3

The surface of A. phagocytophilum provides an important interface for A. phagocytophilum-host interactions including adherence to and internalization of host cells(Wang et al., 2006), inhibition of neutrophil apoptosis (Ge and Rikihisa, 2006;

Borjesson et al., 2005; Ge et al., 2005; Yoshiie et al., 2000), inhibition of reactive oxygen species (ROS) production (Mott and Rikihisa, 2000) and scavenge of exogenous superoxide (Carlyon et al., 2004), conducting antigenic variation to avoid host immune response (Lin and Rikihisa, 2005; Wang et al., 2004a; Zhi et al., 2002), sensing the bacterial environment (Tzung-Huei Lai and Yasuko Rikihisa, unpublished), exchanging nutrients and metabolites with the host cytoplasm (Haibin Huang, 2006). Uniquely among Gram-negative bacteria, A. phagocytophilum has lost all genes required for the biosynthesis of LPS and most genes required for the biosynthesis of peptidoglycan

(Rikihisa, 2006; Lin and Rikihisa, 2003a). There is no pilus or capsule on the surface of organisms in the family Anaplasmataceae (Rikihisa, 1991), suggesting that outer membrane proteins play a crucial role in bacterial interaction with host cells. A. phagocytophilum outer membrane proteins, hence, have become the central focus as potential drug targets and as candidates for differential diagnostic antigens and novel vaccines.

The outer membrane proteins of A. phagocytophilum have not been systematically characterized. . P44 (Msp2) family is the most studied outer membrane protein family of

A. phagocytophilum. The gene expression of p44 paralogs are complicated by a unique gene conversion mechanism involving RecF pathway (Lin et al., 2006; Wang et al.,

2004b; Zhi et al., 2002). Compared to well-studied p44 mRNA expression, the proteins

21 of P44 paralogs are less defined, only P44-18 protein of A. phagocytophilum has been shown as surface-exposed (Kim and Rikihisa, 1998). A. phagocytophilum genome sequencing data provide a wealth of new genetic information (Hotopp et al., 2006).

However, with the exception of P44-18, there is no experimental evidence characterizing any other A. phagocytophilum surface-exposed proteins. Furthermore, almost half of predicted open reading frames (ORFs) of A. phagocytophilum encodes conserved or novel hypothetical proteins hitherto never characterized in any bacteria (Hotopp et al.,

2006), some of which may be surface proteins. Therefore, it has become imperative to take a new approach, including proteomics, to generate more complete picture on the expression and function of A. phagocytophilum surface proteins.

Cell surface biotinylation has emerged as an important tool for studying the cell surface protein expression. Sulfosuccinimidyl-2-[biotinamido]ethyl-1,3-dithiopropionate

(Sulfo-NHS-SS-Biotin) is a thiol-cleavable amine-reactive biotinylation reagent. The specificity of Sulfo-NHS biotin reagents for cell surface labeling has been demonstrated in the identification of eukaryotic membrane proteins (Scheurer et al., 2005) as well as bacterial surface proteins such as Helicobacter pylori (Sabarth et al., 2002) and

Streptococcus pyrogenes (Cole et al., 2005).

The purpose of this paper is to isolate the surface proteins of A. phagocytophilum via biotin-surface labeling of bacteria and streptavidin affinity purification of labeled proteins and then, to identify the proteins by proteomic analysis.

22 Objective 4

The entry and establishment of E. chaffeensis infection involve host caveolae,

glycosylphosphatidylinositol (GPI)-anchored proteins and the incorporation of

cholesterol into bacterial membrane (Lin and Rikihisa, 2003a). After internalization by

host monocytes, E. chaffeensis has the ability to subvert the hostile environment by residing in an early endosome-like compartment, which does not fuse with lysosomes

(Mott et al., 1999). These events are ehrlichial surface-related. However, the corresponding bacterial surface components have not been characterized yet.

Humoral immune response plays an important role in the protection of host against ehrlichial infection. Passive immunization of immunocompetent (Sun et al., 1997;

Kaylor et al., 1991) or immunocompromised (Li et al., 2002; Winslow et al., 2000) animals provides effective protection against Neorickettsia risticii, Anaplasma phagocytophilum or E. chaffeensis infection. Conversely, anti-E. chaffeensis antibody

bound to the E. chaffeensis surface induced potent proinflammatory cytokine mRNA

expression in human monocytes, which may contribute to the pathogenesis of HME (Lee

and Rikihisa, 1997). Assuming that the monoclonal antibodies are representative of the

polyclonal response, mouse antibody response against OMP-1g (P28) accounts for 40%

of the total antibodies against E. chaffeensis (Li et al., 2001). This suggests that ehrlichial

outer membrane proteins are dominant immunogens to stimulate host humoral immune

response. Western blot analysis has revealed several E. chaffeensis immunoreactive

proteins with molecular mass of approximately 74, 70, 64, 47, 31 and 29 kDa (Rikihisa et

al., 1994). However, their identities are still not well known.

23 Studies on bacteria from the family Anaplasmataceae have shown a critical role

of bacterial outer membrane proteins in the stimulation of host immune response and

protection of the host from infection. Immunization with recombinant P28 (one of the

major outer membrane OMP-1/P28 family members) protected mice from E. chaffeensis

challenge (Ohashi et al., 1998). Immunization of calves with A. marginale outer membranes induces a stronger protection efficacy against challenge compared to individual major surface protein (MSP), e.g. MSP-1 and MSP-2 (Abbott et al., 2005;

Brown et al., 1998; Tebele et al., 1991; Palmer et al., 1989; Palmer et al., 1986). Along with this line, efforts have been made to identify the global composition of A. marginale outer membrane immunogens (Lopez et al., 2005).

Despite the importance of E. chaffeensis surface proteins as a crucial interface for pathogen-host interactions as mentioned above, the knowledge of E. chaffeensis surface proteins is limited. In addition to P28 (Ohashi et al., 1998), two other E. chaffeensis proteins, gp47 (Doyle et al., 2006) and gp120 (Popov et al., 2000), have been identified as being surface-exposed by immunoelectron microscopy. There has been no systematic investigation on the surface proteins of E. chaffeensis. Therefore, this paper focuses on the characterization of E. chaffeensis major surface proteins via surface biotinylation using cleavable Sulfo-NHS-SS-Biotin labeling. The identification of E. chaffeensis surface-exposed proteins provides novel insights about E. chaffeensis surface and lays the foundation for rational studies on pathogen-host interaction and vaccine development.

24

CHAPTER 2

ANAPLASMA PHAGOCYTOPHILUM INHIBITS HUMAN NEUTROPHIL APOPTOSIS VIA UPREGULATION OF bfl-1, MAINTENANCE OF MITOCHONDRIAL MEMBRANE POTENTIAL AND PREVENTION OF CASPASE 3 ACTIVATION

2.1 Abstract

The inhibition of neutrophil apoptosis plays a central role in human granulocytic

anaplasmosis. Intracellular signaling pathways through which the obligatory intracellular

bacterium Anaplasma phagocytophilum inhibits the spontaneous apoptosis of human peripheral blood neutrophils were investigated. bfl-1 mRNA levels in uninfected neutrophils after 12 h in culture were reduced to approximately 5 – 25% of 0 h levels, but remained high in infected neutrophils. The eukaryotic RNA synthesis inhibitor, actinomycin D, prevented the maintenance of bfl-1 mRNA levels by A. phagocytophilum.

Differences in mcl-1, bax, bcl-w, bad, or bak mRNA levels in infected versus uninfected

neutrophils were not remarkable. By using mitochondrial fluorescent dyes, Mitotracker

Red and JC-1, it was found that most uninfected neutrophils lost mitochondrial

membrane potential after 10 to 12 h incubation, whereas A. phagocytophilum-infected

neutrophils maintained high membrane potential. Caspase 3 activity and the degree of

apoptosis were lower in dose-dependent manner in A. phagocytophilum-infected

neutrophils at 16 h post infection, as compared to uninfected neutrophils. Anti-active

25 caspase 3 antibody labeling showed less positively stained population in infected neutrophils compared to those in uninfected neutrophils after 12 h incubation. These results suggest that A. phagocytophilum inhibits human neutrophil apoptosis via transcriptional upregulation of bfl-1 and inhibition of mitochondria-mediated activation of caspase 3.

2.2 Introduction

Neutrophils are the primary mediators of inflammation and response to infection.

Since neutrophils typically undergo apoptosis 6 – 12 h after their release from the bone marrow (Akgul et al., 2001; Edwards, 1994), modulation of the apoptotic time course by pathogens has been implicated as a critical factor in the severity and duration of systemic and local infections (Kobayashi et al., 2003; Laskay et al., 2003). Human granulocytic anaplasmosis (HGA, formerly human granulocytic ehrlichiosis) is an acute febrile systemic disease of hematologic and liver enzyme abnormalities (Bakken et al., 1994;

Chen et al., 1994). The disease is associated with the delay of apoptosis following the infection of neutrophils with the obligatory intracellular bacterium, Anaplasma phagocytophilum, which has a tropism for granulocytes. We previously reported that the time required for 50% of the infected neutrophils to show morphological apoptosis is

45.0 ± 9.8 h, which is much longer than that of uninfected neutrophils, 12.2 ± 2.5 h

(Yoshiie et al., 2000). Therefore, the inhibition of neutrophil apoptosis by A. phagocytophilum provides the bacteria with sufficient time to complete the developmental cycle in neutrophils. A recent study corroborates this finding by showing that in vivo infection with an A. phagocytophilum sheep isolate also reduced peripheral blood neutrophil apoptosis (Scaife et al., 2003).

26 The mechanisms by which A. phagocytophilum inhibits the spontaneous apoptosis

of human neutrophils have been investigated (Yoshiie et al., 2000). Binding of protein

components of A. phagocytophilum to neutrophils and subsequent transglutaminase- dependent cross-linking and/or internalization of the receptor and bacteria are required for anti-apoptotic signaling. However, this study (Yoshiie et al., 2000) shows that this process does not require bacterial carbohydrates, bacterial new protein synthesis, host cell protein kinase A, nuclear translocation of NF-κB, or IL-1β which are known to be involved in the inhibition of neutrophil’s apoptosis by some other agents (Ward et al.,

1999; Parvathenani et al., 1998; Watson et al., 1998). Despite these data, intracellular

signaling pathways leading to the inhibition of human neutrophil apoptosis by A.

phagocytophilum remain largely uncharacterized.

Caspases，a family of cysteine proteases，play a critical role during apoptosis in

many nucleated cells (Nicholson, 1999; Salvesen and Dixit, 1997). They are classified

into initiator caspases and effector caspases. Caspase 3 is the main effector caspase in

many apoptotic processes, and it cleaves various cellular protein substrates, including

those involved in cytoskeletal assembly, cell-cycle regulation and DNA replication and

repair (Slee et al., 2001; Stroh and Schulze-Osthoff, 1998). Caspases are produced as

proenzymes and converted to active forms in response to pro-apoptotic signals. The

activation of caspases is controlled by the bcl-2 family of intracellular proteins, which

consists of anti-apoptotic and pro-apoptotic proteins (Cory and Adams, 2002) that

integrate diverse survival and death signals to determine whether apoptosis occurs

(Borner, 2003). A major function of the bcl-2 family is to either maintain (anti-apoptotic

bcl-2 members) or increase (pro-apoptotic bcl-2 members) mitochondrial membrane 27 permeability, which determines the release of mitochondrial apoptotic factors, such as

cytochrome c, Smac and HtrA2 (Tsujimoto, 2003), and subsequent activation of downstream effector caspases.

Thus, the goal of the present study was to investigate whether bcl-2 family

members, caspases, and mitochondria play a role in the inhibition of human neutrophil

apoptosis by A. phagocytophilum.

2.3 Materials and Methods

A. phagocytophilum and cell culture

The A. phagocytophilum HZ strain (Rikihisa et al., 1997) was propagated in HL-

60 cells in RPMI 1640 medium (Invitrogen, Carlsbad, CA) supplemented with 5% fetal

bovine serum (US Bio-Technologies, Parkerford, PA) and 2 mM L-glutamine (Invitrogen)

in a humidified 5% CO2 – 95% air atmosphere at 37°C. To check the infectivity, cells

were cytocentrifuged on a slide in Cytospin 4 (Thermoshandon, Pittsburgh, PA) and

examined by Diff-Quik staining (Baxter Scientific Products, Obetz, OH). When over

90% of the cells were infected, cells were collected and centrifuged at 500 × g for 10

min. The pellet was resuspended in RPMI 1640 at 2 × 106 cells ml-1 in 5 ml and sonicated

under setting 2 at 20 kHz for 7 s using an ultrasonic processor (model W-380; Heat

Systems, Farmingdale, NY). After centrifugation at 500 × g for 5 min, the supernatant

was collected. The supernatant was then centrifuged at 10,000 × g for 10 min, and the

pellet, containing the host cell-free viable A. phagocytophilum, was immediately used to

infect human peripheral blood neutrophils. The number of purified organisms was

estimated as previously described (Yoshiie et al., 2000).

28 Isolation of human peripheral blood neutrophils

Human peripheral blood neutrophils were isolated from the buffy coats of multiple healthy donors according to the Institutional Review Board approved protocol.

The buffy coat was overlaid on double layers of Histopaque 1077 and 1119 (Sigma

Diagnostics, Inc. St. Louis, MO) and centrifuged at 700 × g for 30 min. The neutrophil- rich layer at the interface of Histopaque 1077 and 1119 was collected and washed twice with phosphate-buffered saline (PBS, pH 7.4). The contaminated red blood cells were lysed by incubation with H2O for 20 s at room temperature. More than 95% of cells obtained were neutrophils, as confirmed by examining cells stained with Diff-Quik, and more than 98% of the cells were viable, as determined by the Trypan blue dye exclusion test.

Morphological assessment of apoptosis

Apoptotic cells were scored on Diff-Quick stained slides by morphological characteristics as previously described (Yoshiie et al., 2000). Briefly, the criteria used for apoptotic neutrophils were condensed nuclei and loss of connections between nuclei lobules. A minimum of 200 cells were counted per slide, and the relative percentage of apoptosis was calculated as a ratio of the number of apoptotic cells to the total number of cells. All samples were analyzed in triplicate.

RNA isolation and cDNA synthesis

1 × 107 neutrophils were stabilized in RNAlater (Qiagen, Valencia, CA) and total

RNA was extracted using an RNeasy Mini RNA extraction kit (Qiagen) according to the manufacturer’s instructions. The concentration and purity of the RNA were determined by measuring the A260 and the A260/A280 ratio with a GeneQuant II RNA and DNA

29 calculator (Pharmacia Biotech Inc., Piscataway, NJ). Two micrograms of the extracted

RNA was treated with 2 units of DNase I (Invitrogen) at room temperature for 15 min.

DNase I was then inactivated by the addition of 2 µl of 25 mM EDTA and subsequent

heating at 65°C for 10 min. The DNase I-treated RNA was reverse transcribed in a 30 µl

reaction mixture containing 1 × reaction buffer (50 mM Tris-HCl pH 8.3, 75 mM KCl, 3

mM MgCl2), 10 mM DTT, 1.5 µM oligo (dT) primer, 60 units of RNaseOUT (Invitrogen)

and 0.5 mM of dNTP mixture. Next, 300 units of SuperScript II reverse transcriptase

(Invitrogen) was added to the reaction mixture after 2 min of heating at 42°C, and the

reaction was carried out at 42°C for 50 min and then terminated by incubating at 70°C for

15 min.

PCR and competitive-PCR

To determine the transcription levels of anti-apoptotic (bfl-1, mcl-1, and bcl-w)

and pro-apoptotic (bad, bax, and bak) genes in neutrophils, 2 µl of cDNA was amplified

in a 50 µl reaction mixture containing 1 × reaction buffer (20 mM Tris-HCl pH 8.3, 50

mM KCl, 1.5 mM MgCl2), 0.2 mM of dNTP mixture, 0.4 µM of forward and reverse

primers and 2 units of Taq DNA polymerase (Invitrogen) in a DNA thermal cycler

GeneAmp PCR System 9700 (Perkin-Elmer, Foster City, CA). Taq DNA polymerase was

added to the reaction mixture after heating at 95°C for 5 min. One PCR cycle consisted of

denaturation at 95°C for 60 s, annealing at 60°C for 60 s and extension at 72°C for 90 s.

PCR reactions were performed for 25 cycles for g3pdh, bfl-1 and mcl-1 and for 27 cycles

for bcl-w, bax, bad and bak. The final extension was carried out at 72°C for 10 min. Next,

10 µl of PCR products were electrophoresed in a 1.2% agarose gel containing 0.5 µg ml-1 of ethidium bromide. DNA size markers (1Kb Plus DNA Ladder; Invitrogen) were run in 30 parallel. The primers for bcl-w, bad, bak, bax and bfl-1 (bfl-F1 and bfl-R1) (Santos-

Beneit and Mollinedo, 2000) and for g3pdh (Kim and Rikihisa, 2000) were as described previously. Primers for mcl-1 (Genbank accession number: L08246) mcl-F

[CGGCGCCGCTTGAGGAGATG] and mcl-R [TTACGCCGTCGCTGAAAACA] were designed for this study.

For competitive-PCR, competitor plasmid (mimic) for bfl-1 gene with a 77 bp deletion was constructed by PCR. The four primers used were bfl-F1, bfl-R2

[CATTATGAACTCCGCAACAAAATATGAAATCTGGTTACAATTCTTCCCC], bfl-

F2 [GGCATCATTAACTGGGGAAGAATTGTAACCAGATTTCATATTTTGTTGC] and bfl-R1. Two PCR reactions were performed using primer pairs, bfl-F1 + bfl-R2; and bfl-F2 + bfl-R1, respectively. The amplicons from these two PCR reactions were used as the templates for the third PCR using the primer pair, bfl-F1 + bfl-R1. The amplicon from the third PCR (e.g., bfl-1 gene with the 77 bp deletion) was then cloned into the pCRII vector (Invitrogen). These PCRs were all carried out for 30 cycles under the same amplification conditions as described above. The mimic for g3pdh was described previously (Kim and Rikihisa, 2000). The optimum amounts of the mimics were experimentally determined and added to the reaction mixture as follows: 20 atmol for bfl-1 and 7.5 atmol for g3pdh. The competitive-PCR for bfl-1 and g3pdh were performed under the same conditions as the non-competitive-PCR. Amplified target and mimic

DNA bands were identified by their predicted sizes in the agarose gel. The amounts of the target and mimic amplicons were determined by a gel video system (Gel Print 2000;

BioPhotonics Corp., Ann Arbor, MI) and analyzed by image analysis software

31 ImageQuant (Molecular Dynamics, Sunnyvale, CA). The ratio of bfl-1 target to its mimic was calculated and normalized against the amount of g3pdh mRNA, which was determined by the ratio of g3pdh target to its own mimic.

Mitochondrial staining and fluorescence microscopy

After washing cells, Mitotracker Red 580 (Molecular Probes, Eugene, OR) was added at a concentration of 100 nM to neutrophils at 0 h, and neutrophils after 12 h with or without A. phagocytophilum infection at an MOI of 100 in the fresh culture medium and incubated at 37°C for 20 min. Then the cells were centrifuged and resuspended in the prewarmed medium. Cells were cytocentrifuged on a glass slide and analyzed by a Nikon

Eclipse E400 fluorescence microscope with xenon–mercury light source (Nikon

Instruments Inc., Melville, NY).

Flow cytometric analysis of mitochondrial membrane potential

Uninfected human neutrophils and A. phagocytophilum-infected neutrophils at an

MOI of 100 were cultured at 37°C for 10 h. After culture, cells were washed and incubated with JC-1 (Molecular Probes) at 10 µM for 10 min at 37°C and washed with

PBS. For low mitochondrial membrane potential control, FCCP (Sigma) was added to the freshly isolated neutrophils at 20 nM for 15 min at 37°C. Fluorescence emission was collected by Coulter Epics Elite flow cytometer (Beckman Coulter, Miami, FL) equipped with a 15 mW air-cooled argon ion laser (Cyonics, San Jose, CA) using a 525 nm band pass filter for JC-1 monomers (green fluorescence) and a 635 nm band pass filter for JC-1 aggregates (red fluorescence).

32 Caspase 3 activity assay

Caspase 3 activity was measured using the Caspase 3 Cellular Activity Assay Kit

Plus (BIOMOL, Plymouth Meeting, PA). Briefly, 1 × 107 cells of infected and uninfected neutrophils were seeded into two wells of a 6-well plate. After incubation at 37°C for 16 h, neutrophils were centrifuged at 500 × g for 10 min at 4°C. Lysis buffer containing 1

µM cytochalacin D (Sigma, St. Louis, MO) and 1:100 dilution protease inhibitor cocktail set III (Calbiochem, San Diego, CA) was added to the pellet and kept on ice for 10 min.

The mixture was subjected to five freeze-thaw cycles in a methanol/dry ice bath and centrifuged at 18,000 × g at 4°C for 20 min. Eighty microliters of the assay buffer included in the kit, 10 µl of the supernatant and 10 µl of Ac-DEVD-pNA colorimetric substrate were mixed in triplicate wells and optical density was measured at 405 nm using a microplate spectrophotometer SPECTRAmax PLUS (Molecular Devices,

Sunnyvale, CA). The protein concentration was determined by the BCA protein assay reagent (Pierce Chemical Co., Rockford, IL). Specific activity of caspase 3 was expressed as pmol min-1 mg-1 of protein, according to the manufacturer’s instructions.

Flow cytometric analysis of active caspase 3

For flow cytometric analysis, 1×106 cells were pelleted and washed in PBS. All subsequent steps were performed at room temperature. Cells were fixed in 3% paraformaldehyde for 15 min. After washing in PBS, cells were incubated with 0.5 µg rabbit anti-active caspase 3 IgG (BD Biosciences Pharmingen, San Diego, CA) in PGS buffer (0.2% gelatin and 0.3% saponin in PBS) for 1 h, using normal rabbit IgG (Jackson

ImmunoResearch, West Grove, PA) as the isotype control. After being washed in PGS, the cells were labeled with FITC-conjugated goat anti-rabbit IgG (Jackson 33 ImmunoResearch) at a dilution of 1:100 in PGS for 1 h. The cells were washed in PGS,

resuspended in PBS and analyzed by the same flow cytometer as described above

operating at 488 nm. Positive staining was defined as fluorescence intensity above that of

the isotype control.

2.4 Results

A. phagocytophilum infection inhibits the decrease in bfl-1 mRNA in human neutrophils

The bcl-2 family regulates the downstream caspase cascade via modulation of mitochondrial membrane permeability. bcl-2 family members are themselves regulated at

a transcriptional level (Cory and Adams, 2002). Thus, using reverse transcription (RT)-

PCR (Choi et al., 2004; Itoh et al., 2003; Villunger et al., 2003), we investigated the

relationship between bcl-2 family and the inhibition of neutrophil apoptosis by A.

phagocytophilum. Neutrophil infection levels are dependent on A. phagocytophilum

multiplicity of infection (MOI) and we previously reported that the inhibition of

neutrophil apoptosis is dependent on MOI (Yoshiie et al., 2000). Since individual

ehrlichiae is small in size and a small number of dispersed organisms in the cytoplasm in

the early stage of infection are difficult to recognize under light microscope, the actual

infection rate of neutrophils is impractical to determine prior to 24 h (Yoshiie et al.,

2000). By immuno-fluorescence microscopy, the infection rate of neutrophils with A.

phagocytophilum at an MOI of 100, was approximately 50% to 60% at 12 h post

infection (PI), and the average number of organisms in each infected cell was

approximately 2 (data not shown).

34 Compared to freshly isolated neutrophils, levels of bfl-1 mRNA in uninfected

neutrophils after 12 h in culture were reduced to approximately 5 – 25% of respective 0 h

levels (from 6 different human blood specimens), but bfl-1 mRNA levels were about 60 –

150% of respective 0 h levels (from 6 different human blood specimens) in neutrophils infected with A. phagocytophilum at an MOI of 100 after 12 h PI (representative results are shown in Fig. 1A-D). Levels of mcl-1 mRNA were slightly decreased in both uninfected and infected neutrophils after 12 h in culture (Fig. 1A). In contrast, bax mRNA levels were slightly increased in uninfected neutrophils after 12 h in culture and slightly decreased in infected neutrophils (Fig. 1A). The mRNA levels of bcl-w, bad, and bak were relatively low and nearly undetectable under our experimental conditions (data not shown).

The change in bfl-1 mRNA levels between infected and uninfected neutrophils was further characterized by competitive-RT-PCR. At 12 h PI, the decrease in bfl-1 mRNA levels was significantly attenuated at an MOI of 10 and with an even greater effect at an MOI of 100 (Fig. 1B). Furthermore, time course analysis revealed that bfl-1 mRNA levels in uninfected neutrophils were progressively reduced (Fig. 1C). In contrast, the level of bfl-1 mRNA in infected cells was only reduced by 50% at 4 h and was gradually increased thereafter until 12 h PI (Fig. 1C).

To analyze whether the A. phagocytophilum-mediated attenuation of bfl-1 mRNA reduction was dependent on the transcriptional activation of bfl-1 gene, cells were incubated with an inhibitor of eukaryotic RNA synthesis, actinomycin D (ActD; 0.2 µg ml-1). Levels of bfl-1 mRNA in infected neutrophils treated with ActD for 12 h were 35 –

40% (n = 2 independent experiments) of those seen without ActD treatment (Fig. 1D).

35 Moreover, ActD had no effect on bfl-1 mRNA levels in uninfected cells at 12 h (Fig. 1D).

These data suggest that A. phagocytophilum infection in human neutrophils results in the upregulation of bfl-1 mRNA expression via increased transcription.

Infection of human neutrophils by A. phagocytophilum attenuates the decrease in mitochondrial membrane potential

Human neutrophils have a mitochondrial network that has very little role in ATP synthesis but regulates spontaneous apoptosis via modulation of mitochondrial membrane integrity which can be detected by measuring the membrane potential (Maianski et al.,

2004; Fossati et al., 2003; Regula et al., 2003). Therefore, we determined mitochondrial membrane integrity by analyzing the mitochondrial membrane potential in A. phagocytophilum-infected neutrophils. Mitotracker Red, a cell-permeant mitochondrion- selective dye, is well concentrated by mitochondria of high membrane potential and retained during further processing, thus, ideal for florescence microscopy (Fossati et al.,

2003). Freshly isolated neutrophils showed strong granular red mitochondrial staining throughout the cytosol, indicating that fresh neutrophils have high mitochondrial membrane potential (Fig. 2A). After 12 h incubation, 53.7 ± 3.6% (n = 3) of the uninfected neutrophils lost mitochondrial membrane potential as shown by weak fluorescence staining (Fig. 2B). In contrast, only 15.7 ± 0.9% (n = 3) of the A. phagocytophilum-infected neutrophils at an MOI of 100 lost mitochondrial membrane potential at 12h PI (Fig. 2C). The significant difference in mitochondrial membrane potential by Mitotracker staining between A. phagocytophilum-infected and uninfected neutrophils suggests that A. phagocytophilum infection inhibits the spontaneous loss of 36 mitochondrial membrane potential in neutrophils in vitro. At 12 h, the number of bacteria associated with each neutrophil was very small as described above. Therefore, the contribution of bacteria to Mitotracker Red staining is negligible in infected neutrophils.

Mitochondrial membrane potential in human neutrophils infected with A. phagocytophilum was further assessed with JC-1 which has the excitation wave length

(488 nm) ideal for flow cytometry. JC-1 (5, 5’, 6, 6’- tetrachloro - 1, 1’, 3, 3’- tetraethylbenzimidazolylcarbocyanine iodide) is a fluorescent dye that incorporates into mitochondria and forms aggregates (red fluorescence, maximal emission at 590 nm) at high mitochondrial membrane potential, and monomers (green fluorescence, maximal emission at 527 nm) at low mitochondrial membrane potential (Reers et al., 1991). When cells were treated with JC-1 to assess mitochondrial membrane potential, mitochondria in freshly isolated neutrophils were demonstrated to have high membrane potential (i.e., low green fluorescence; Fig. 3A). As a positive control, treatment with carbonyl cyanide 4-

(trifluoromethoxy) phenylhydrazone (FCCP), a protonophore and uncoupler of oxidative phosphorylation in mitochondria (Buckler and Vaughan-Jones, 1998), resulted in the loss of mitochondrial membrane potential (i.e., high green fluorescence; Fig. 3B). After 10 h incubation, most of the uninfected human neutrophils showed a loss of mitochondrial membrane potential (Fig. 3C), whereas approximately 70% (65.3 – 71.7%, n = 2 independent experiments) of neutrophils infected with A. phagocytophilum at an MOI of

100 maintained high levels of mitochondrial membrane potential (Fig. 3D). Both data of mitochondrial staining based on fluorescence microscopy using Mitotracker Red and flow cytometric analysis using JC-1 indicated that A. phagocytophilum infection of human neutrophils decreases the spontaneous loss of mitochondrial membrane potential.

37

A. phagocytophilum infection inhibits human neutrophil spontaneous caspase 3 activation

Caspase 3 is the primary effector caspase involved in the apoptosis of human neutrophils (Khwaja and Tatton, 1999; Pongracz et al., 1999; Yamashita et al., 1999).

Morphological signs of apoptosis are present in ~ 60 – 80% of uninfected neutrophils at

16 h post-infection, whereas neutrophils infected with A. phagocytophilum show differing

degrees of apoptosis that were related to the MOI (Yoshiie et al., 2000). Therefore, we

examined the relationship between caspase 3 activity and MOI in neutrophils infected

with A. phagocytophilum at 16 h PI. Caspase 3 activities were always lower in

neutrophils infected with A. phagocytophilum at various ranges of MOI at 16 h PI, as

compared with uninfected neutrophils from the same blood specimen. However, caspase

3 activities in uninfected neutrophils from different individuals widely varied. For

example, caspase 3 specific activities in uninfected neutrophils from six different blood

donors included in Fig.1 varied from 116.7 to 700.6 pmol min-1 mg-1 of protein after 16 h

incubation. Therefore, we normalized caspase 3 specific activities among individual

blood specimens by using the ratio of the caspase 3 activity in A. phagocytophilum-

infected neutrophils to that in uninfected neutrophils from the same blood specimen. The

data obtained from six independent experiments were grouped into two groups according

to the MOI, i.e. 51 – 70 and 5 – 20. Levels of the inhibition of both caspase 3 activity and

percentage of apoptosis in neutrophils infected with A. phagocytophilum at 51 – 70 MOI

were significantly higher than those in neutrophils infected at 5 – 20 MOI (P < 0.05, n = 6

38 independent experiments, Fig. 4). Therefore, activation of caspase 3 was inhibited during the delayed apoptosis in human neutrophils by infection with A. phagocytophilum, which was dependent on MOI.

Active caspase 3 consists of a heterodimer of 17 and 12 kDa subunits which are derived from the 32 kDa proenzyme (Pelletier et al., 2004). Therefore, using specific antibody to active caspase 3, which does not react with procaspase 3, we determined the population of neutrophils having active caspase 3 during A. phagocytophilum infection.

By this method, we could consistently observe the difference between infected and uninfected neutrophils as early as 12 h incubation; no positively stained cells were present in freshly isolated uninfected neutrophils (Fig. 5A), ~ 33% (33.2 – 33.7%, n = 2 independent experiments) of uninfected neutrophils stained positively after 12 h incubation (Fig. 5B), and ~ 15% (14.2 – 15.4%, n = 2 independent experiments) of A. phagocytophilum-infected neutrophils stained positively at 12 h post infection (Fig. 5C).

Although we observed a greater difference between infected and uninfected neutrophils at 16 h incubation than that at 12 h, due to many apoptotic neutrophils in the uninfected control group, some cells were clumped, thus the flow cytometric analysis data at 16 h was less reliable and not included in the result. Overall, in agreement with the data of caspase 3 enzyme activity, the flow cytometric analysis data showed that A. phagocytophilum infection inhibited caspase 3 activation during spontaneous apoptosis of human neutrophils.

39 2.5 Discussion

The present study suggests that A. phagocytophilum infection inhibits neutrophil

spontaneous apoptosis via the upregulation of bfl-1 mRNA levels. These data provide a greater understanding of the pathogenesis of HGA and the regulation of apoptosis in neutrophils.

Recent reports have characterized the expression of the bcl-2 family genes in

human neutrophils. Human neutrophils express two anti-apoptotic genes, bfl-1 (Moulding

et al., 2001; Santos-Beneit and Mollinedo, 2000) and mcl-1 (Fulop et al., 2002; Moulding

et al., 2001; Leuenroth et al., 2000), but do not express the anti-apoptotic gene, bcl-2

(Kim et al., 2001; Moulding et al., 2001; Santos-Beneit and Mollinedo, 2000). The

present study is the first to describe the upregulation of bfl-1 in conjunction with the

active inhibition of apoptosis in human neutrophils, which is consistent with reports that

have demonstrated an anti-apoptotic role of Bfl-1 in a variety of other cell types. For

example, retrovirus-mediated transfer of bfl-1 cDNA to a human microvascular

endothelial cell line provides protection against TNF-α–induced cell death (Karsan et al.,

1996), and bfl-1 upregulation is associated with protection against TNF-α-induced

apoptosis in HeLa cells (Zong et al., 1999). Transfection of Jurkat cells with retroviral

vectors containing bfl-1 protects the cells from CD95- and Trail receptor-induced

apoptosis (Werner et al., 2002). Further, overexpression of the murine homologue of

human Bfl-1, A1, protects murine B lymphoma cells from anti-IgM-induced apoptosis

(Craxton et al., 2000), and peripheral blood neutrophils in A1-a-/- mice display accelerated apoptosis (Hamasaki et al., 1998).

40 In contrast, overexpression of pro-apoptotic members of the bcl-2 family, such as bad or bax, induce or accelerate apoptosis in cancer cells (Li et al., 2003; Xu et al., 2002;

Kelekar et al., 1997). Expression of bax is downregulated during GM-CSF-induced delay

of apoptosis in human neutrophils (Weinmann et al., 1999). However, in the present

study, A. phagocytophilum infection had a minimum effect on bax mRNA expression,

while the expression of bad and bak were relatively weak. This suggests that the A.

phagocytophilum-induced delay in human neutrophil apoptosis is not dependent on

changes in pro-apoptotic bcl-2 family gene expression.

The present study suggests that the anti-apoptotic influence of A.

phagocytophilum is mediated via the upregulation of bfl-1. Previous reports have shown

that Bfl-1 inhibits apoptosis via its interaction with several pro-apoptotic factors such as

Bid (Werner et al., 2002) and Bax (Zhang et al., 2000). Bid cleavage occurs during the

early stages of spontaneous apoptosis in neutrophils (Baumann et al., 2003). Bfl-1 tightly

and selectively sequesters truncated Bid (tBid), thereby preventing tBid from associating

with other pro-apoptotic factors, such as Bax or Bak, and preventing tBid-induced

cytochrome c release (Werner et al., 2002). Bax interacts with a voltage-dependent anion

channel (VDAC) on the mitochondrial outer membrane (Shimizu et al., 2000; Shimizu et

al., 1999) or induces channel formation on the mitochondrial outer membrane

(Antonsson et al., 2001; Saito et al., 2000; Wei et al., 2000), thereby increasing the

permeability of the mitochondrial outer membrane. Mitochondrial membrane potential

correlates with mitochondrial outer membrane permeability during apoptosis (Regula et

al., 2003). The maintenance of mitochondrial membrane potential in A.

phagocytophilum-infected neutrophils demonstrated in the present study reflects the

41 maintenance of mitochondrial integrity by A. phagocytophilum. Similar findings have been observed in response to other pathogens in other cell types, such as Neisseria meningitides-infected HeLa cells, which are protected from staurosporine-induced apoptosis (Massari et al., 2003). However, the present study is the first report to describe the stabilization of the mitochondrial membrane potential in neutrophils in response to bacterial infection.

Cytochrome c and Smac are mitochondrial apoptotic proteins that are released into the cytosol during the spontaneous apoptosis of human neutrophils (Altznauer et al.,

2004; Maianski et al., 2004; Baumann et al., 2003). Cytochrome c forms the apoptosome complex with apoptotic protease activating factor 1 (Apaf-1) and caspase 9, which leads to the activation of caspase 9 and thereby stimulates the downstream caspase cascade

(Saleh et al., 1999). In contrast, Smac indirectly activates caspases by inhibiting the family of inhibitors of apoptosis protein (IAP) (Du et al., 2000). Infection of human neutrophils with A. phagocytophilum results in the maintenance of mitochondrial integrity and thereby prevents mitochondria from releasing these apoptotic factors, leading to delay in caspase activation and apoptosis. Among the caspase family of proteins, caspase 3 is the critical effector caspase in the apoptosis of human neutrophils

(Daigle and Simon, 2001; Khwaja and Tatton, 1999; Pongracz et al., 1999) and is activated by a mitochondria-dependent pathway (Altznauer et al., 2004; Maianski et al.,

2004; Murphy et al., 2003; Maianski et al., 2002; Watson et al., 1999). The present data extend on previous reports by demonstrating that A. phagocytophilum blocks this primary spontaneous apoptotic pathway in human neutrophils.

42 Based on our results and the overall mechanism of action for Bcl-2 family

members in the apoptosis signaling pathway inferred from numerous literatures at

present, we propose the following model of A. phagocytophilum–induced inhibition of human neutrophil apoptosis: infection of human neutrophils with A. phagocytophilum

upregulates the expression of the anti-apoptotic Bfl-1, leading to the sequestration of tBid

and prevention of tBid and Bax association in the mitochondrial outer membrane. This

subsequently leads to the dissociation of Bax from the membrane, prevention of

mitochondrial apoptotic protein release and prevention of downstream caspase cascade

activation.

In B lymphocytes, bfl-1 transcription is upregulated via NF-κB (Lee et al., 1999).

Edelstein et al (Edelstein et al., 2003) proposed that NF-κB-dependent assembly of an

enhanceosome-like complex on the promoter region of bfl-1 plays a critical role in the

control of bfl-1 transcription. However, the inhibitory mechanism of neutrophil apoptosis

by A. phagocytophilum seems to be unique, because it is not dependent on NF-κB nuclear

translocation (Yoshiie et al., 2000). Further studies designed to identify direct

transcriptional activators for bfl-1 in the A. phagocytophilum-infected neutrophils would

be of considerable interest.

In conclusion, the findings from this study suggest that A. phagocytophilum

infection inhibits neutrophil apoptosis via increased bfl-1 transcription and maintenance

of bfl-1 transcript levels. Antagonism of this anti-apoptotic pathway could potentially

prevent bacterial proliferation in neutrophils and allow host immune systems to eradicate

the pathogen.

43 Fig. 2.1. Inhibition of the reduction of bfl-1 transcript levels in human neutrophils

by A. phagocytophilum infection. RNA was isolated from infected or uninfected

neutrophils, and competitive-PCR was performed using primers against bfl-1 (See

Experimental procedures section for experimental conditions and primer sequences).

(A) bfl-1 mRNA levels were reduced to approximately 5 – 25% (6 different blood

specimens) in uninfected neutrophils after 12 h in culture, but were high in neutrophils

infected with A. phagocytophilum at an MOI of 100. mcl-1 mRNA were slightly reduced,

but no significant difference was observed between infected and uninfected neutrophils.

In contrast, bax mRNA level were slightly increased in uninfected neutrophils after 12 h

in culture and slightly decreased in infected neutrophils. Lane 1, Marker (1Kb Plus DNA

Ladder, Invitrogen); lanes 2 and 3, uninfected neutrophils (PMN) at 0 h; lanes 4 and 5,

uninfected neutrophils at 12 h; lanes 6 and 7, neutrophils infected with A.

phagocytophilum (A. phago) at an MOI of 100 at 12 h. RT- indicates the absence of

reverse transcriptase.

(B) A. phagocytophilum at MOI of 10 and 100 prevented the reduction of bfl-1 mRNA

levels in neutrophils as determined by competitive-PCR. Lane 1, Marker (1Kb Plus DNA

Ladder, Invitrogen); lane 2, uninfected neutrophils at 0 h; lane 3, uninfected neutrophils

after 12 h incubation; lane 4, neutrophils infected with A. phagocytophilum at an MOI of

10 at 12 h; lane 5, neutrophils infected with A. phagocytophilum at an MOI of 100 at 12 h.

(C) Infection of human neutrophils with A. phagocytophilum prevented the reduction of

bfl-1 mRNA levels in neutrophils as determined by time course competitive-PCR. Lane 1,

Marker (1Kb Plus DNA Ladder, Invitrogen); lanes 2, 3, 5, 7, uninfected neutrophils at 0,

4, 8, 12 h, respectively; lanes 4, 6, 8, neutrophils infected with A. phagocytophilum at an

44 MOI of 100 at 4, 8, 12 h, respectively.

(D) Actinomycin D treatment abrogated the protective effect of A. phagocytophilum on reduction of bfl-1 mRNA levels in neutrophils as determined by competitive-PCR. Lane

1, Marker (1Kb Plus DNA Ladder, Invitrogen); lanes 2, 3, 5 uninfected neutrophils at 0,

12, 12 h, respectively; lanes 4, 6, neutrophils infected at an MOI of 100 with A. phagocytophilum at 12 h; lanes 5, 6 treated with 0.2 µg ml-1 actinomycin D (ActD).

The cDNA ratios of bfl-1, mcl-1, or bax to g3pdh for each sample are indicated below

each panel. The ratios for the uninfected neutrophils at 0 h, which are normalized against

g3pdh cDNA in the corresponding samples, are arbitrarily set as 1.0. Numbers on the

right indicate respective amplicon sizes. Data are presented representative of two to four

independent experiments.

45 Fig. 2.1

46

Fig. 2.2 Inhibition of loss of mitochondrial staining with Mitotracker Red in human neutrophils by A. phagocytophilum infection. (A) Freshly isolated neutrophils showed extensive mitochondrial staining with Mitotracker Red 580. (B) Weak mitochondrial staining in neutrophils at 12 h culture. (C) Mitochondria in neutrophils infected with A. phagocytophilum at an MOI of 100 at 12 h PI retained strong red fluorescence. The scale bar is 5 µm.

47 Fig. 2.3. Prevention of the loss of mitochondrial membrane potential in human

neutrophils by the infection with A. phagocytophilum as determined by flow

cytometry using JC-1 staining. Uninfected human neutrophils and A. phagocytophilum- infected neutrophils at an MOI of 100 were cultured at 37°C for 10 h. After culture, cells were incubated with JC-1 at 10 µM for 10 min at 37°C and washed with PBS. For low mitochondrial membrane potential control, FCCP was added to the freshly isolated neutrophils at 20 nM for 15 min at 37°C. Fluorescence emission was collected by a flow cytometer using a 525 nm band pass filter for JC-1 monomers (green fluorescence).

(A) Freshly isolated neutrophils showed high mitochondrial membrane potential. (B)

Freshly isolated neutrophils treated with FCCP at 37°C for 15 min showed loss of mitochondrial membrane potential. (C) Following 10 h of incubation, high mitochondrial membrane potential was maintained in approximately 10% of the uninfected human neutrophils. (D) After 10 h culture, high mitochondrial membrane potential was maintained in approximately 70% of A. phagocytophilum-infected cells. Data presented are representative of two independent experiments.

48 Fig. 2.3

49

1 5 -20 MOI 51 - 70 MOI 0.8

0.6 * 0.4 * blood specimen specimen blood 0.2 value in infected to relative uninfected neutrophils in each each in neutrophils uninfected 0 percentage of apoptosis caspase 3 activity

Fig. 2.4. Inhibition of spontaneous apoptosis and caspase 3 enzyme activity in

human neutrophils by A. phagocytophilum in a dose-dependent manner. Neutrophils

from six different individuals were infected with A. phagocytophilum at either of two

MOI ranges of 5 – 20 (n = 3) and 51 – 70 (n = 3) for 16 h in six independent experiments.

Caspase 3 activities in whole cell extracts were measured by colorimetric assay using the

caspase 3-specific substrate, Ac-DEVD-pNA. Caspase 3 specific activities in infected

neutrophils among individual blood specimens were normalized by dividing the caspase

3 activity in A. phagocytophilum-infected neutrophils by that in uninfected neutrophils

from the same blood specimen. Apoptotic cells were scored on Diff-Quick stained slides

based on morphological characteristics as previously described (Yoshiie et al., 2000). A minimum of 200 cells were counted per slide, and the percentage of apoptosis was calculated as the number of apoptotic cells/total number of cells. Values represent the means ± standard deviations (n = 3). * indicates significantly difference between two groups of MOI (P<0.05).

50

Fig. 2.5 Prevention of generation of activated caspase 3 in human neutrophils by A. phagocytophilum infection. Infected or uninfected neutrophils were incubated with rabbit anti-active caspase 3 or normal rabbit IgG, then with FITC-conjugated goat anti- rabbit IgG, and analyzed by the flow cytometer operating at 488 nm. Positive staining was defined as fluorescence intensity above that of the normal IgG control. (A) No positively stained cells were present in freshly isolated (0 h) uninfected neutrophils. (B)

33% of uninfected neutrophils stained positively at 12 h incubation. (C) 15% of A. phagocytophilum-infected neutrophils stained positively at 12 h PI. Data presented are representative of two independent experiments.

51

CHAPTER 3

ANAPLASMA PHAGOCYTOPHILUM DELAYS SPONTANEOUS HUMAN NEUTROPHIL APOPTOSIS BY MODULATION OF MULTIPLE APOPTOTIC PATHWAYS

3.1 Abstract

Anaplasma phagocytophilum infects human neutrophils and inhibits the intrinsic

pathway of spontaneous neutrophil apoptosis by protecting mitochondrial membrane

integrity. In the present study, we investigated the molecular signaling of the extrinsic

pathway and the interaction between the intrinsic and extrinsic pathways in the inhibition

of spontaneous human neutrophil apoptosis by A. phagocytophilum. Cell surface Fas

clustering during spontaneous neutrophil apoptosis was significantly blocked by A.

phagocytophilum infection. The cleavage of pro-caspase 8, caspase 8 activation and the

cleavage of Bid, which links the intrinsic and extrinsic pathways, in the extrinsic pathway

of spontaneous neutrophil apoptosis were inhibited by A. phagocytophilum infection.

Inhibition of this pathway was active since the cleavage of pro-caspase 8 and Bid in anti-

Fas-induced neutrophil apoptosis was also inhibited by A. phagocytophilum infection.

Likewise, A. phagocytophilum infection inhibited the pro-apoptotic Bax translocation to mitochondria, activation of caspase 9, the initiator caspase in the intrinsic pathway, and the degradation of a potent caspase inhibitor, X-chromosome-linked IAP (XIAP) protein,

52 during spontaneous neutrophil apoptosis. These data point to a novel mechanism induced by A. phagocytophilum involving both extrinsic and intrinsic pathways to ensure to delay

the apoptosis of host neutrophils.

3.2 Introduction

Neutrophils play a pivotal role in innate immunity. They typically undergo

apoptosis within 6 – 12 h after their release into the peripheral blood from bone marrow

(Akgul et al., 2001), which is important in the maintenance of homeostatic levels of

neutrophils and the resolution of inflammatory responses. However, this short life span of

neutrophils is altered by infection with several pathogens, which has been implicated as a

critical factor in the severity and duration of systemic and local infections and

inflammation (DeLeo, 2004; Kobayashi et al., 2003; Laskay et al., 2003). Human

granulocytic anaplasmosis (HGA, formerly human granulocytic ehrlichiosis) is an acute febrile systemic disease accompanied by hematologic and liver enzyme abnormalities.

The etiologic agent of HGA, Anaplasma phagocytophilum, is an obligatory intracellular

bacterium, which has a tropism for granulocytes. Yoshiie et al. discovered that A.

phagocytophilum delays the spontaneous apoptosis of human neutrophils in vitro

(Yoshiie et al., 2000). A. phagocytophilum has inhibitory effects on the apoptosis of

isolated peripheral blood neutrophils for up to 48 h culture and neutrophils in the

mixtures of peripheral blood leukocyte cultures for up to 96 h by morphological

observation (Yoshiie et al., 2000). The delayed neutrophil apoptosis provides the

bacteria with sufficient time to complete the developmental cycle in neutrophils (Yoshiie

et al., 2000). This anti-apoptotic phenomenon has been confirmed by several in vitro

53 studies on human neutrophls (Borjesson et al., 2005; Choi et al., 2005; Ge et al., 2005) as well as by an in vitro study on ovine neutrophils infected in vivo with a sheep isolate

(Scaife et al., 2003).

Signaling pathways leading to apoptosis are generally classified into two categories: the extrinsic pathway (the death receptor pathway) and the intrinsic pathway

(the mitochondrial pathway), both of which have the same outcome: the activation of a characteristic caspase cascade (Sprick and Walczak, 2004). In the extrinsic pathway, stimulation of death receptors (e.g. Fas, tumor necrosis factor [TNF]-α-receptor, or TNF- related apoptosis-inducing ligand [TRAIL]-receptor) with their ligands (e.g. Fas ligand,

TNF-α, or TRAIL) allows the formation of a death-inducing signaling complex (DISC), which contains an adaptor protein (e.g. Fas-associated death domain [FADD]) and an initiator caspase (mainly caspase 8). The autocleavage of caspase 8 in the DISC produces active caspase 8 that initiates downstream apoptotic signaling (Donepudi et al., 2003). In the intrinsic pathway, when cells receive apoptotic stimuli, the mitochondrial outer membrane is permeabilized by pro-apoptotic Bcl-2 family members. This results in the release of apoptotic factors and loss of mitochondrial inner membrane potential (∆Ψm)

(Antonsson, 2004). One of these released factors, cytochrome c, activates Apaf-1, which in turn recruits and activates caspase 9, an initiator caspase in the intrinsic pathway

(Green, 2005). The active caspase 9 triggers the downstream caspase activation cascade.

Both the extrinsic and intrinsic pathways may converge at the level of mitochondria

(Sprick and Walczak, 2004). In addition, the inhibitor of apoptosis protein (IAP) family

54 functions distal to mitochondria as inhibitors of caspases. X-chromosome-linked IAP

(XIAP) , with much higher affinity for caspases than other IAP proteins, is the most

potent caspase suppressor in the IAP family (Salvesen and Duckett, 2002).

Neutrophils are exceptional cells with a high rate of constitutive apoptosis (Akgul

et al., 2001), during which both mitochondria- and death receptor-mediated apoptotic

signaling are shown to be activated (Daigle and Simon, 2001). During spontaneous

neutrophil apoptosis, without ligation of external ligands, Fas clustering occurs through

an unknown mechanism (Scheel-Toellner et al., 2004; Simon, 2003; Daigle and Simon,

2001). XIAP is expressed in human neutrophils, however, the regulation of XIAP during

spontaneous neutrophil apoptosis is still controversial. Despite the progress in our

understanding of spontaneous neutrophil apoptosis, mechanisms by which it is initiated

and regulated are largely uncharacterized.

The mechanisms by which A. phagocytophilum inhibits the spontaneous apoptosis

of human neutrophils have been partially elucidated. It seems that the delay of

spontaneous neutrophil apoptosis is triggered by the surface protein molecules of A. phagocytophilum (Borjesson et al., 2005; Yoshiie et al., 2000). Binding of the bacterial protein molecules to neutrophils and host transglutaminase activity are required for A. phagocytophilum anti-apoptotic effect (Yoshiie et al., 2000). We have also demonstrated that A. phagocytophilum inhibits the intrinsic pathway of the spontaneous neutrophil apoptosis by protecting the mitochondrial membrane integrity (Ge et al., 2005). The decrease of Bfl-1, an anti-apoptotic Bcl-2 family member, and activation of the major executioner caspase, caspase 3, during spontaneous neutrophil apoptosis are inhibited by

A. phagocytophilum infection (Ge et al., 2005). The higher expression level of Bfl-1 in A.

55 phagocytophilum-infected cells compared to uninfected cells was confirmed in recent

studies using human neutrophils as well as the promyelocytic leukemia cell line, NB4

(Pedra et al., 2005; Sukumaran et al., 2005), so was the inhibition of caspase 3 activation

(Choi et al., 2005). However, whether and how death receptor-mediated spontaneous human neutrophil apoptosis pathway is inhibited by A. phagocytophilum is still unknown.

In the present study, we investigated the signaling events in the extrinsic pathway and the

cross-talk between the intrinsic and extrinsic pathways in the inhibition of spontaneous

human neutrophil apoptosis by A. phagocytophilum infection. Our observation that

agonistic anti-Fas antibody-induced neutrophil apoptosis is prevented by A.

phagocytophilum was also reported but not further investigated by Borjesson et al.

(Borjesson et al., 2005). Therefore, the present study included the signaling events in the

agonistic Fas antibody-induced human neutrophil apoptosis to contrast with spontaneous

human neutrophil apoptosis.

3.3 Materials and Methods

A. phagocytophilum and cell culture

The A. phagocytophilum HZ strain (Rikihisa et al., 1997) was propagated in HL-

60 cells in RPMI 1640 medium (Invitrogen, Carlsbad, CA) supplemented with 5% fetal

bovine serum (US Bio-Technologies, Parkerford, PA) and 2 mM L-glutamine (Invitrogen) in a humidified 5% CO2 – 95% air atmosphere at 37°C. To check the infectivity, cells

were cytocentrifuged on a slide using a Cytospin 4 (ThermoShandon, Pittsburgh, PA) and

examined by Diff-Quik staining (Baxter Scientific Products, Obetz, OH). When over

90% of the cells were infected, cells were collected and centrifuged at 500 × g for 10 min.

56 The pellet was resuspended in RPMI 1640 at 2 × 106 cells ml-1 in 5 ml and sonicated under setting 2 at 20 kHz for 7 s using an ultrasonic processor (model W-380; Heat

Systems, Farmingdale, NY). After centrifugation at 500 × g for 5 min, the supernatant

was collected. The supernatant was then centrifuged at 10,000 × g for 10 min, and the

pellet containing the host cell-free viable A. phagocytophilum was used immediately to

infect human peripheral blood neutrophils at a multiplicity of infection (MOI) of 100.

The number of purified organisms was estimated as previously described (Yoshiie et al.,

2000).

Isolation of human peripheral blood neutrophils and culture

Human peripheral blood neutrophils were isolated from the buffy coats of

multiple healthy donors according to the Institutional Review Board approved protocol.

Neutrophils were isolated by Dextran sedimentation as previously described (Fossati et al.,

2003; Edwards, 1996) with some modifications. The leukocytes were sedimented by

1.5% Dextran and were resuspended in phosphate-buffered saline (PBS, pH 7.4). This leukocyte suspension was then overlaid onto Histopaque 1077 (Sigma Diagnostics, Inc.

St. Louis, MO) and centrifuged at 700 × g for 20 min. The neutrophil pellet in the bottom was collected and the contaminated red blood cells were lysed by hypotonic treatment with H2O for 20 s at room temperature. More than 95% of cells obtained were neutrophils,

as confirmed by examining cells stained with Diff-Quik, and more than 98% of the cells were viable, as determined by the Trypan blue dye exclusion test. Neutrophils were cultured at 2 × 106 ml-1 in RPMI 1640 medium supplemented with 5% fetal bovine serum and 2 mM L-glutamine in a humidified 5% CO2 – 95% air atmosphere at 37°C. To

evaluate the effect of A. phagocytophilum on anti-Fas IgM-accelerated neutrophil

57 apoptosis, human neutrophils were infected with A. phagocytophilum at 37°C for 30 min

followed by incubation with mouse anti-Fas IgM antibody (clone CH-11; MBL

International, Woburn, MA) at 500 ng ml-1 (Borjesson et al., 2005). The cells were then

incubated for 6 h before being harvested.

Morphological assessment of apoptosis

Apoptotic cells were scored on Diff-Quik stained slides by morphological

characteristics as previously described (Ge et al., 2005). Briefly, the criteria used for

apoptotic neutrophils were condensed nuclei and loss of connections between nuclei

lobules. A minimum of 200 cells were counted per slide, and the relative percentage of

apoptosis was calculated as a ratio of the number of apoptotic cells to the total number of

cells. All samples were scored in triplicate.

Flow cytometric and confocal analyses of Fas

For flow cytometric and confocal microscopic analyses of cell surface Fas

expression, 1 × 106 cells were pelleted and washed in PBS. All subsequent steps were

performed at room temperature. Cells were fixed in 3% paraformaldehyde for 15 min.

After washing in PBS, cells were incubated with 1 µg mouse anti-Fas IgG1 (clone ZB4;

Upstate, Lake Placid, NY) or isotype control mouse IgG1 (clone MOPC 21, Sigma-

Aldrich, St. Louis, MO) in PG buffer (0.2% gelatin in PBS) for 1 h. After being washed in PG, the cells were labeled with Alexa Fluor 488 goat anti-mouse IgG (Invitrogen) at a dilution of 1:200 in PG for 1 h. The cells were washed in PG, resuspended in PBS and analyzed in the flow cytometer (BD FACSCalibur system, San Jose, CA) and Zeiss LSM

510 confocal laser scanning microscope (Carl Zeiss AG, Vertrieb Deutschland, Germany)

58 operated at the excitation wave length of 488 nm. For confocal microscopy, cells were scanned at 512 × 512 pixels and processed with the Zeiss confocal software. Z stack images were collected for three dimensional (3D) projection.

Immunofluorescence microscopy of Bax

For double fluorescence staining of intracellular Bax and mitochondria, paraformaldehyde-fixed neutrophils were incubated with rabbit anti-human Bax antibody

(BD Biosciences Pharmingen, San Diego, CA) at a 1:200 dilution and 0.5 µg mouse anti-

human Manganese Superoxide Dismutase (MnSOD) antibody (clone MnS-1; Alexis, San

Diego, CA) in PGS buffer (0.2% gelatin and 0.3% saponin in PBS) for 1 h . After being washed in PGS, the cells were labeled with Alexa Fluor 488 goat anti-rabbit IgG

(Invitrogen) and Alexa Fluor 555 goat anti-mouse IgG (Invitrogen) at a dilution of 1:200 in PGS for 1 h. The cells were washed in PGS, cytocentrifuged onto a glass slide and analyzed using a Nikon Eclipse E400 fluorescence microscope with a xenon–mercury light source (Nikon Instruments Inc., Melville, NY).

Western blotting analysis

Neutrophil protein samples were collected using trichloroacetic acid (TCA) precipitation as described previously (Scheel-Toellner et al., 2004) with some modifications. Briefly, neutrophils were washed with PBS containing protease inhibitor cocktail set III at a 1:100 dilution (Calbiochem, San Diego, CA). Proteins were precipitated with 10% TCA and centrifuged at 18,000 × g for 10 min at 4°C. The pellets

were then lysed in 9 M urea, mixed with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer (50 mM Tris.Cl, pH 6.8, 5% 2-

Mercaptoethanol, 2% SDS, 0.1% bromophenol blue, 10% glycerol) and boiled for 5 min.

59 The protein samples were subjected to 12% SDS-PAGE and transferred to

polyvinylidene fluoride (PVDF) membranes (Westran S; Schleicher & Schuell

BioScience, Keene, NH). The membranes were incubated with mouse anti-human

caspase 8 (clone 1C12; Cell Signaling, Beverly, MA), rabbit anti-human Bid (Cell

Signaling), rabbit anti-human Bax (BD Biosciences Pharmingen), rabbit anti-XIAP (Cell

Signaling) or mouse anti-human α-tubulin IgG (Cedarlane, Hornby, Ontario, Canada),

which was followed by incubation with horseradish peroxidase-conjugated goat anti-

mouse or anti-rabbit IgG (KPL, Gaithersburg, MA). Human α-tubulin served as loading

controls. The blots were developed using the enhanced chemiluminescence technique

(ECL; Amersham Biosciences, Piscataway, NJ).

Caspase 8 and 9 activity assay

Caspase 8 and 9 activities were measured using Caspase-Glo 8 and Caspase-Glo 9

luminescent assay kits (Promega, Madison, WI), according to the manufacturer’s

instructions, which was based on assessing the luminescent signal generated from caspase

cleavage of proluminogenic substrate. Briefly, 2.5 × 104 cells of A. phagoctyphilum- infected and uninfected neutrophils in 100 µl were seeded into a white-walled 96-well

Cliniplate (Thermo Electron Corporation, Marietta, OH) in triplicate wells. After incubation at 37°C for 16 h, neutrophils were cooled to room temperature. Equal volumes of Caspase-Glo reagent containing proluminogenic caspase 8 substrate carrying the

LETD sequence or caspase 9 substrate carrying the LEHD sequence were added to each well. After incubation for 40 min at room temperature, the luminescence of each sample was read by a microplate luminometer Gemini XS (Molecular Devices, Sunnyvale, CA).

The caspase activity was expressed as mean luminescence intensity per 2.5 × 104 cells. 60 Statistical analysis

Statistical analyses were performed using Student’s t-test. Data were presented as mean ± standard deviation (SD), and the numbers of independent experiments were indicated in each legend. A p value < 0.05 was considered statistically significant.

3.4 Results

Inhibition of death receptor clustering during spontaneous human neutrophil apoptosis by A. phagocytophilum

During spontaneous human neutrophil apoptosis, the death receptor, Fas (CD95), clusters and is activated, which is independent of ligation by the Fas ligand (Scheel-

Toellner et al., 2004). Therefore, we investigated the effect of A. phagocytophilum infection on death receptor activation during spontaneous human neutrophil apoptosis.

First, the effects of A. phagocytophilum infection on the distribution of Fas on cell surface were analyzed by confocal laser scanning microscopy. In freshly isolated uninfected neutrophils, dotted and even surface distribution of Fas was observed (Fig.

1A). However, at 16 h culture, Fas aggregated into a large cluster (cap) as described previously in apoptotic neutrophils (Scheel-Toellner et al., 2004). In A. phagocytophilum- infected neutrophils at 16 h post infection (PI), Fas remained dotted and evenly distributed, and there was significantly less cap formation compare to uninfected neutrophils at 16 h culture (Fig. 1A and B), which was similar to that in freshly isolated neutrophils. These data indicate that Fas clustering on cell surface during spontaneous human neutrophil apoptosis is blocked by A. phagocytophilum infection. Second, to determine the influence of A. phagocytophilum infection on the cell surface expression of

61 Fas, cell surface Fas was measured by flow cytometry. Infected neutrophils at 16 h PI,

which had slightly less cell surface Fas amount than uninfected neutrophils at 0 h,

showed substantially less surface Fas than uninfected neutrophils at 16 h culture (Fig. 1C).

Inhibition of caspase 8 activation and Bid cleavage in A. phagocytophilum-infected

human neutrophils

When Fas molecules cluster, they interact with the cytosolic adapter proteins,

FADD, which in turn recruit pro-caspase 8 and promote caspase 8 aggregation to induce their auto-cleavage and production of active caspase 8 in human neutrophils (Scheel-

Toellner et al., 2004). Therefore, to examine whether A. phagocytophilum blocks caspase

8 activation in human neutrophils, the cleavage of caspase 8 was analyzed by western blotting. At 0 h, minimum cleavage of pro-caspase 8 (55/57 kDa) was detected. Pro- caspase 8 was cleaved and weak intermediate 43/41 kDa products were consistently detected in uninfected neutrophils at 16 h culture (Fig. 2A). Whereas in A. phagocytophilum-infected neutrophils at 16 h PI, the cleaved products were at the basal level as freshly isolated neutrophils. Alpha-tubulin amounts were shown as protein loading control as previously described (Mott et al., 2002). Caspase 8 activity was

assayed further using a proluminogenic substrate. In A. phagocytophilum-infected neutrophils, caspase 8 activities at 8 and 16 h PI were significantly lower than in uninfected neutrophils at 8 and 16 h culture, respectively (Fig. 2 B). Therefore, as measured by the levels of caspase 8 activity and pro-caspase 8 cleavage, the activation of caspase 8 in spontaneous neutrophil apoptosis was inhibited by A. phagocytophilum infection.

62 In the death receptor-mediated apoptotic pathway, activated caspase 8 cleave the

pro-apoptotic Bcl-2 family member, Bid, 22 kDa protein to yield the 15 kDa truncated

Bid (tBid) (Li et al., 1998), which is shown to be produced in spontaneous apoptotic

neutrophils (Scheel-Toellner et al., 2004). tBid is demonstrated to promote apoptosis by

linking death receptor-mediated extrinsic to mitochondrial intrinsic apoptotic pathways in

TNF-α or Fas-activated various cancer cell lines including FL5.12 cells, Jurkat cells,

HeLa cells or MCF7 cells (Kuwana et al., 2005; Grinberg et al., 2002; Li et al., 1998). A

similar mechanism was proposed in neutrophils in a review article (Akgul and Edwards,

2003) but yet to be demonstrated. We found that in uninfected neutrophils, Bid was intact

at 0 h and cleaved to its 15 kDa active form, tBid, at 16 h culture (Fig. 2A). In contrast,

there was no detectable cleavage of Bid in A. phagocytophilum-infected neutrophils at 16

h PI, indicating that the cleavage of Bid during spontaneous neutrophil apoptosis was

prevented by A. phagocytophilum infection.

Inhibition of caspase 8 activation and Bid cleavage in agonistic anti-Fas IgM-induced

human neutrophil apoptosis by A. phagocytophilum infection

Human neutrophil apoptosis can be accelerated by treatment with an agonistic anti-Fas IgM antibody (Beinert et al., 2000). At 6 h culture, 51.8 ± 2.2% of uninfected neutrophils with anti-Fas IgM treatment was apoptotic (Fig. 3A). A. phagocytophilum

infection significantly inhibited anti-Fas IgM-induced apoptosis (Fig. 3A). This is in

agreement with the report of Borjesson et al. (Borjesson et al., 2005). In neutrophils

treated with anti-Fas IgM, the caspase 8 intermediate cleavage products (43/41 kDa) and

tBid were clearly detectable at 6 h culture and the cleavage of both pro-caspase 8 and Bid

was inhibited by A. phagocytophilum infection (Fig. 3B).

63 Inhibition of cytosolic Bax translocation to mitochondria during spontaneous human

neutrophil apoptosis by A. phagocytophilum infection

During spontaneous human neutrophil apoptosis, pro-apoptotic Bcl-2 family

member Bax is translocated and colocalized with mitochondria (Maianski et al., 2002).

Therefore, the amount and intracellular distribution of Bax were investigated in A. phagocytophilum-infected human neutrophils. There was no significant difference in total

Bax protein levels among uninfected neutrophils at 0 h or 16 h culture, or A.

phagocytophilum-infected neutrophils at 16 h PI (Fig. 4A). Mitochondria were labeled

with antibody specific to human Manganese Superoxide Dismutase (MnSOD) (Maianski

et al., 2004). When analyzed by double immunofluorescence staining, Bax as well as

mitochondria formed large aggregates and Bax colocalized with mitochondria in

uninfected neutrophils at 16 h culture (Fig. 4B). In contrast, there were less large

aggregates formed by Bax or mitochondria in A. phagocytophilum-infected neutrophils at

16 h PI. In addition, as in neutrophils at 0 h, Bax did not colocalize with mitochondria in

A. phagocytophilum-infected cells. The percentage of neutrophils with Bax translocation

to mitochondria was significantly lower in infected cells compared with uninfected

neutrophils at 16 h culture (Fig. 4C). These results indicate that Bax translocation to

mitochondria during spontaneous neutrophil apoptosis was inhibited by A.

phagocytophilum infection. At the same time, the amount of MnSOD and number of labeled mitochondria appeared to be considerably up-regulated in infected cells at 16 h PI compared with uninfected neutrophils at 0 h (Fig. 4B).

64 Inhibition of caspase 9 activation during spontaneous human neutrophil apoptosis by

A. phagocytophilum

When mitochondria are permeabilized, cytochrome c is released into the cytosol and interacts with Apaf-1 to recruit and activate caspase 9, an initiator caspase in the mitochondrial pathway, as reviewed by Green (Green, 2005). We found that caspase 9 was activated in spontaneous neutrophil apoptosis, and that its activities in A. phagocytophilum-infected neutrophils at both 8 and 16 h PI were significantly lower than in uninfected neutrophils at 8 and 16 h, respectively (Fig. 5), further supporting our previous conclusion that A. phagocytophilum infection inhibits the mitochondrial apoptotic pathway during spontaneous human neutrophil apoptosis (Ge et al., 2005).

Prevention of XIAP degradation in human neutrophils by A. phagocytophilum infection

As an inhibitor of apoptosis, XIAP interacts directly with caspases and inhibits the caspase-mediated apoptosis process in Bax-treated 293T cells (Deveraux et al., 1997).

However, the role of XIAP in spontaneous neutrophil apoptosis is unknown. To investigate the involvement of XIAP in A. phagocytophilum-delayed human neutrophil apoptosis, XIAP protein levels were analyzed by western blotting. Compared to freshly isolated neutrophils, the XIAP protein level was decreased in neutrophils after 12 h incubation, and the cleaved 30 kDa fragment was evident (Fig. 6). In contrast, in A. phagocytophilum-infected neutrophils, the XIAP protein level was similar to freshly isolated neutrophils and no cleaved fragment was detected. These results showed that

XIAP protein was degraded during spontaneous human neutrophil apoptosis, and that this degradation was blocked by A. phagocytophilum infection.

65 3.5 Discussion

The present data define signaling pathways inhibited by A. phagocytophilum

infection in spontaneous apoptosis in human neutrophils. In spontaneous neutrophil apoptosis, Fas receptor clustering occurs without ligation of Fas by Fas ligand (Scheel-

Toellner et al., 2004). Our results showed that A. phagocytophilum inhibited both spontaneous Fas clustering and subsequent activation of extrinsic pathways. Treatment of neutrophils with monodansylcadaverine (MDC), a transglutaminase inhibitor, abrogates the inhibition of neutrophils apoptosis by A. phagocytophilum (Yoshiie et al.,

2000), suggesting that A. phagocytophilum may induce cross-linking proteins by the activation of transglutaminase in neutrophils to prevent spontaneous Fas clustering. Our results also showed that A. phagocytophilum inhibited the extrinsic signaling cascade induced by anit-Fas IgM at the upstream of caspase 8 activation, supporting the active inhibition of the upstream extrinsic pathway by A. phagocytophilum.

The present study suggests that the pro-apoptotic Bcl-2 family members of Bax and Bid are regulated at a post-translational level by A. phagocytophilum infection in human neutrophils. This is in agreement with previous reports that mRNA amounts of

Bax and Bid during neutrophil apoptosis were unchanged by A. phagocytophilum infection (Borjesson et al., 2005; Ge et al., 2005). The active tBid is shown to directly activate the pro-apoptotic Bcl-2 family member, Bax, causing a conformational change in

Bax and subsequent Bax translocation to permeabilize the mitochondrial membrane in in vitro mitochondrial assay (Cartron et al., 2004). In spontaneous neutrophil apoptosis, Bid is cleaved to tBid (Scheel-Toellner et al., 2004), and Bax is activated to translocate from cytosol to mitochondria and form aggregates (Maianski et al., 2002). Our data showed

66 that the production of tBid and translocation of Bax from the cytosol to the mitochondria in spontaneous neutrophil apoptosis were inhibited by A. phagocytophilum infection.

MnSOD was identified as a mitochondrial marker in human neutrophils (Maianski et al.,

2004). We found that human MnSOD was up-regulated in infected neutrophils at 16 h PI compared with uninfected cells at 0 h. This upregulation is unlikely from bacterial SOD since A. phagocytophilum has only one type of SOD, FeSOD (Dunning-Hotopp et al.,

2006). The microarray data presented by Borjesson et al. (Borjesson et al., 2005) shows 2 to 3 fold up-regulation of MnSOD (SOD2) in human neutrophils infected with A. phagocytohilum at 9 h although this was not mentioned in the text. Therefore, it is possible that A. phagocytophilum infection up-regulates MnSOD at the transcriptional level or perhaps increases mitochondrial biogenesis. In addition, A. phagocytophilum prevents the decrease of anti-apoptotic Bcl-2 protein Bfl-1 expression in spontaneous neutrophil apoptosis (Ge et al., 2005). Bfl-1 is reported to block the mitochondrial apoptotic pathway by sequestering tBid and preventing it from activating Bax as demonstrated by in vitro mitochondria assay (Werner et al., 2002). Thus, A. phagocytophilum may inhibit translocation of Bax from cytosol to mitochondria by both inhibition of tBid production and sequestration of tBid by Bfl-1.

Besides blocking both of death-receptor and mitochondria-mediated apoptotic pathways during spontaneous human neutrophil apoptosis, A. phagocytophilum also

regulates neutrophil apoptosis at the level of IAP. The IAP family typically contains one

or more baculoviral IAP repeat (BIR) and a RING zinc finger domain. The BIR domain is able to directly bind caspases and inhibit caspase activity. The RING zinc finger domain has ubiquitination ligase activity, which becomes auto-ubiquitinated upon

67 receiving apoptotic stimuli, resulting in degradation of IAP proteins. The over-

expression of one of IAP proteins, XIAP, protects human embryonic kidney cells from

Fas-induced apoptosis (Deveraux and Reed, 1999). XIAP is degraded in oxidants-

accelerated neutrophil apoptosis (Gardai et al., 2004). However, Maianski et al.

(Maianski et al., 2004) proposed that the XIAP protein amount was stable during

spontaneous neutrophil apoptosis in in vitro culture. In contrast, the present study

demonstrated the 30 kDa degraded fragment of XIAP (Deveraux and Reed, 1999) and the corresponding decrease in the XIAP protein level in apoptotic neutrophils, indicating that

XIAP protein is degraded during spontaneous neutrophil apoptosis. Our result is in agreement with the observation of Kobayashi et al. that the amount of XIAP in cultured neutrophils is less than that in freshly isolated neutrophils (Kobayashi et al., 2002). The pro-apoptotic protein Smac or HtrA2 released from mitochondria are known to antagonize the inhibition of caspase activity by IAP proteins (Green, 2005). Since the loss of mitochondrial inner membrane potential, an indicator of permeabilization of the mitochondrial outer membrane, during spontaneous neutrophil apoptosis is prevented by

A. phagocytophilum infection (Ge et al., 2005), the release of antagonists of XIAP from mitochondria is expected to be inhibited by A. phagocytophilum infection as well.

Therefore, A. phagocytophilum appears to inhibit inactivation of XIAP in neutrophils at least by two mechanisms, i.e. inhibition of XIAP degradation and sequestration of XIAP

antagonists within mitochondria.

Taken together, the present results advance our understanding of the inhibitory

mechanism of spontaneous human neutrophil apoptosis by A. phagocytophilum infection.

The following model regarding the inhibition of human neutrophil apoptosis by A.

68 phagocytophilum infection is proposed based on the findings with A. phagocytophilum

and the known mechanisms of action of apoptotic factors inferred from numerous other related publications (cited in the present paper) (Fig. 7). The activation of death receptor,

Fas, subsequent activation of initiator caspase, caspase 8, and pro-apoptotic Bcl-2 family protein, Bid, during spontaneous human neutrophil apoptosis are inhibited by A. phagocytophilum infection. The death receptor-mediated direct activation of executioner caspases and its amplification through mitochondria via tBid are blocked by A. phagocytophilum infection. On the other hand, A. phagocytophilum infection prevents the decrease of anti-apoptotic Bfl-1 expression during spontaneous neutrophil apoptosis, which prevents pro-apoptotic tBid and Bax from permeabilizing mitochondria and thus, blocks the release of pro-apoptotic mitochondrial factors from initiating the activation of caspase cascade. In addition, the degradation of the caspase inhibitor, XIAP, in spontaneous neutrophil apoptosis is prevented by A. phagocytophilum infection.

Therefore, A. phagocytophilum inhibits both death receptor-mediated and mitochondrial apoptotic pathways in spontaneous neutrophil apoptosis through multiple mechanisms, which are converged at the key apoptotic check point, mitochondria, delaying amplification of the apoptotic cascade.

A. phagocytophilum is the first bacterium that has been demonstrated to have the capability to inhibit both extrinsic and intrinsic apoptotic pathways in neutrophils at multiple levels. Especially, we are not aware of any bacterium that inhibits spontaneous

Fas clustering. Elucidation of the inhibitory mechanism of neutrophil apoptosis by A. phagocytophilum may help to unravel the unique initiation mechanism of extrinsic pathway in spontaneous neutrophil apoptosis. Now, we have delineated in part the

69 molecular participants in the inhibitory pathways of delayed spontaneous neutrophil apoptosis by A. phagocytophilum. This will set a stage for identifying A. phagocytophilum molecules and cognate neutrophil receptor molecules that lead to this global inhibition. In the future, vaccines or drugs designed to target A. phagocytophilum anti-apoptotic signaling in neutrophils may prevent this organism from proliferating in its host cells and eventually control the disease of HGA.

70 Fig. 3.1 Inhibition of Fas capping during spontaneous human neutrophil apoptosis

by A. phagocytophilum infection. Infected (A. phago) or uninfected neutrophils (PMN)

were fixed in paraformaldehyde, incubated with mouse anti-Fas (IgG1, clone ZB4) or

mouse IgG isotype control (IgG1, clone MOPC 21) and then stained with Alexa Fluor

488 goat anti-mouse IgG.

(A) Laser scanning confocal microscopy. a and d. Freshly isolated neutrophils showed

dotted Fas staining. b and e. Capping of Fas on neutrophil surface was seen as strong

green fluorescence aggregates at 16 h culture. c and f. Fas from infected neutrophils at 16

h PI retained scattered green fluorescence distribution. The scale bar is 5 µm. a-c. 3D

image (rendered from Z stack of 20 sections); d-f. Single image in one plane.

(B) The percentage of neutrophils with Fas capping. The percentage of neutrophils

having Fas capping at 16 h culture was significantly greater in uninfected neutrophils

(PMN) than in A. phagocytophilum-infected neutrophils (A. phago) at 16 h PI or in uninfected neutrophils at 0 h. Values represent the means ± standard deviations (n = 3). * indicates significant difference compared to infected neutrophils at 16 h PI or freshly isolated neutrophils at 0 h (P < 0.05). Data presented are representative of three independent experiments, each with neutrophils from different donors and different batches of freshly isolated host cell-free A. phagocytophilum.

(C) Flow cytometry. The intensity of cell surface Fas fluorescence labeling of neutrophils was less in infected neutrophils (A. phago) at 16 h PI than in uninfected neutrophils

(PMN) at 0 h, which were much less than in uninfected neutrophils at 16 h. The values of mean channel fluorescence intensity are shown in parentheses beside each histogram.

Data presented are representative of two independent experiments.

71 Fig. 3.1

72 Fig. 3.2 Inhibition of activation of caspase 8 and cleavage of Bid in neutrophils by A.

phagocytophilum infection. (A) Inhibition of the cleavage of pro-caspase 8 and Bid in

spontaneous neutrophil apoptosis by A. phagocytophilum. Cleavage of pro-caspase 8 was

detected as the appearance of intermediate cleavage products p43/p41 in uninfected neutrophils at 16 h culture. In A. phagocytophilum-infected neutrophils (A. phago), the

cleavage of pro-caspase 8 was inhibited. Bid was cleaved to 15 kDa active fragment,

truncated Bid (tBid), in uninfected neutrophils after 16 h incubation. In A.

phagocytophilum-infected neutrophils (A. phago) at 16 h PI, there was no detectable cleavage of Bid. Compared to the Caspase 8 (p57/55) panel, the exposure time for the cleaved caspase 8 (p43/p41) was increased to reveal the weak immunoreactive cleaved fragments. The relative density ratios of pro-caspase 8 (p57) and Bid to α-tubulin for each

sample are indicated below the panel. The ratios for the uninfected neutrophils at 0 h,

which are normalized against α-tubulin amounts in the corresponding samples, are

arbitrarily set as 1.0. Alpha-tubulin amounts were shown as loading control. (B) Caspase

8 activities were measured by luminescent assay using caspase 8 specific proluminogenic substrate containing the LETD sequence. Values of relative luminescence unit (RLU)

represent the means ± standard deviations (n = 3). * indicates significant difference

compared to uninfected neutrophils at 8 and 16 h, respectively (P < 0.05). Data are

representatives of three independent experiments.

73 Fig. 3.2

74 Fig. 3.3 Inhibition of anti-Fas IgM-induced human neutrophil apoptosis by A. phagocytophilum infection. (A) Human neutrophil apoptosis was accelerated by treatment with agonistic anti-Fas IgM antibody (clone CH-11). Values represent mean ± standard deviations (n = 3). Apoptotic cells at 6 h culture were scored on Diff-Quik- stained slides based on morphological characteristics as previously described (Ge, et al.,

2005). A minimum of 200 cells was counted per slide, and the percentage of apoptosis was determined. Values represent mean ± standard deviations (n = 3). * indicates significant difference compared to uninfected neutrophils with or without anti-Fas IgM treatment, respectively (P < 0.05). (B) In anti-Fas IgM-treated uninfected neutrophils at 6 h culture, the cleaved caspase 8 and Bid were clearly detected compare to untreated cells.

In A. phagocytophilum-infected neutrophils (A. phago), there was much less cleaved products of caspase 8 or Bid. Data presented are representative of three independent experiments.

75 Fig. 3.3

76 Fig. 3.4 Inhibition of colocalization of Bax with mitochondria during spontaneous human neutrophil apoptosis by A. phagocytophilum infection. (A) Expression of Bax in uninfected neutrophils and A. phagocytophilum-infected neutrophils (A. phago) at 16 h culture was analyzed by western blotting. Alpha-tubulin amounts served as loading control. (B) Cytoplasmic distribution of mitochondria (human MnSOD, red fluorescence) and Bax (green fluorescence) was examined by double immunofluorescence labeling. At 0 h, mitochondria were seen as dotted pattern and Bax was distributed diffusely throughout the cytoplasm. Bax was not colocalized with mitochondria in freshly isolated neutrophils (PMN). In uninfected neutrophils at 16 h culture, both mitochondria and Bax formed aggregates and were colocalized. In A. phagocytophilum-infected neutrophils (A. phago), Bax was distributed diffusely and mitochondria remained dotted in the cytoplasm, which was similar to that seen in freshly isolated 0 h neutrophils. The scale bar is 5 µm. (C) Numbers of neutrophils having Bax colocalized with mitochondria were scored. The percentage of neutrophils having Bax colocalized with mitochondria at 16 h culture was greater in uninfected neutrophils

(PMN) than in A. phagocytophilum-infected neutrophils (A. phago) at 16 h PI or uninfected neutrophils at 0 h. Values represent the means ± standard deviations (n = 3). * indicates significant difference compared to infected neutrophils at 16 h or uninfected neutrophils at 0 h (P < 0.05). Results are representative of three independent experiments.

77 Fig. 3.4

78

Fig. 3.5 Inhibition of caspase 9 activation in spontaneous neutrophil apoptosis by A. phagocytophilum infection. Caspase 9 activity was measured by luminescence assay using caspase 9 specific proluminogenic substrate containing the LEHD sequence.

Values of relative luminescence unit (RLU) represent the means ± standard deviations (n

= 3). * indicates significant difference compared to uninfected neutrophils at 8 and 16 h, respectively (P < 0.05). Data are representative of three independent experiments.

79

Fig. 3.6 Inhibition of cleavage of XIAP in neutrophils infected with A. phagocytophilum. After 12 h incubation, XIAP in uninfected neutrophils was cleaved and the 30 kDa fragment was produced. Whereas in A. phagocytophilum-infected neutrophils (A. phago) at 12 h PI, there was no detectable cleavage of XIAP, similar to that in freshly isolated neutrophils by western blotting analysis. Compared with the XIAP panel, the exposure time for the cleaved XIAP panel was increased to reveal the weak immunoreactive cleaved fragment. Alpha-tubulin amounts served as loading control. The relative density ratios of XIAP to α-tubulin for each sample are indicated below the panel.

The density ratio for the uninfected neutrophils at 0 h, which is normalized against α- tubulin amount in the corresponding sample, is arbitrarily set as 1.0. Data are representative of three independent experiments.

80 Fig. 3.7 Model for the inhibitory mechanism of human neutrophil apoptosis by A. phagocytophilum infection. A. phagocytophilum infection of human neutrophils inhibits the capping of death receptor, Fas, during spontaneous neutrophil apoptosis. The downstream of death receptor-mediated spontaneous neutrophil apoptosis, e.g. the activation of caspase 8 and cleavage of pro-apoptotic BH3 domain-only Bid, is blocked by A. phagocytophilum infection. At the same time, A. phagocytophilum infection prevents the decrease of anti-apoptotic Bfl-1 expression, which interferes with Bid and prevents it from interacting with and activating Bax to generate a pore-forming oligomer in the outer membrane of mitochondria, a key apoptotic check point. Consequently, the loss of mitochondrial inner membrane potential (∆Ψm) is blocked by A. phagocytophilum infection. Therefore, A. phagocytophilum infection initiates multiple signals to modulate the spontaneous neutrophil apoptotic machinery, which converge at the level of mitochondria, inhibiting the release of pro-apoptotic factors and the downstream activation of caspase cascade. A. phagocytophilum infection also inhibits the degradation of XIAP, a direct inhibitor of caspases. The up or down arrows inside of the text boxes indicate the modulation by A. phagocytophilum infection.

81 Fig. 3.7

82

CHAPTER 4

IDENTIFICATION OF NOVEL SURFACE PROTEINS OF ANAPLASMA PHAGOCYTOPHILUM BY AFFINITY PURIFICATION AND PROTEOMICS

4.1 Abstract

Anaplasma phagocytophilum is the etiologic agent of human granulocytic

anaplasmosis, which is one of the major tick-borne zoonoses in the United States. The

surface of A. phagocytophilum plays a crucial role in subverting the hostile host cell

environment. However, except for P44/Msp2 outer membrane protein family, the surface

components of A. phagocytophilum are largely unknown. To globally identify the major

surface proteins of A. phagocytophilum, a membrane-impermeable, cleavable biotin

reagent, Sulfo-NHS-SS-Biotin, was used to label the intact bacteria. The biotinylated

bacterial surface proteins were isolated by streptavidin agarose affinity purification. The

purified proteins were separated by electrophoresis and then analyzed by Nano-

LC/MS/MS. Among the major proteins captured by affinity purification, were five A. phagocytophilum proteins, i.e., OMP85, hypothetical protein APH_0404 (designated here as Asp62), hypothetical protein APH_0405 (designated here as Asp55), P44 family proteins and OMP-1A. The genes of Asp62 and Asp55 were co-transcribed and highly conserved in the family Anaplasmataceae. The surface-exposure of Asp62 and Asp55

83 were further verified by immunofluorescence microscopy. With the exception of P44-18,

all of them were newly discovered major surface-exposed proteins, and as such will

facilitate understanding the interaction between A. phagocytophilum and the host. These

proteins may serve as targets for development of chemotherapy, diagnostics and vaccines.

4.2 Introduction

Human granulocytic anaplasmosis (HGA, formerly human granulocytic

ehrlichiosis) has recently been recognized as a zoonotic disease of public health

importance and became one of the most common tick-borne zoonoses in the United

States and Europe (Alberti et al., 2005; Demma et al., 2005). HGA is an acute febrile systemic illness accompanied by hematologic and liver enzyme abnormalities. It can cause severe and potentially fatal illness especially in immunocompromised and elderly

people (Dumler et al., 2005). The etiologic agent of HGA, Anaplasma phagocytophilum,

is a Gram-negative, obligatory intracellular bacterium, which has been initially known as having tropism for granulocytes (Chen et al., 1994) and recently known to infect endothelial cells, too (Herron et al., 2005).

The surface of A. phagocytophilum provides an important interface for A. phagocytophilum-host interactions including adherence to and internalization of host cells (Wang et al., 2006), inhibition of neutrophil apoptosis (Ge and Rikihisa, 2006;

Borjesson et al., 2005; Ge et al., 2005; Yoshiie et al., 2000), inhibition of reactive oxygen species (ROS) production (Mott and Rikihisa, 2000) and scavenge of exogenous superoxide(Carlyon et al., 2004), exhibiting antigenic variation to avoid host immune

response (Lin and Rikihisa, 2005; Wang et al., 2004; Zhi et al., 2002), sensing the

84 bacterial environment (Tzung-Huei Lai and Yasuko Rikihisa, unpublished) and

exchanging nutrients and metabolites with the host cytoplasm (Haibin Huang, 2006).

Uniquely among Gram-negative bacteria, A. phagocytophilum has lost all genes required

for the biosynthesis of LPS and most genes required for the biosynthesis of peptidoglycan

(Rikihisa, 2006; Lin and Rikihisa, 2003). There is no pilus or capsule on the surface of organisms in the family Anaplasmataceae (Rikihisa, 1991), suggesting that outer

membrane proteins play a crucial role in bacterial interaction with host cells. A.

phagocytophilum outer membrane proteins, hence, have become the central focus as potential drug targets and as candidates for differential diagnostic antigens and novel vaccines.

The outer membrane proteins of A. phagocytophilum have not been systematically characterized. P44/Msp2 family is the most studied outer membrane protein family of A.

phagocytophilum. Each P44 consists of a central hypervariable region and conserved

flanking sequences (Lin et al., 2003; Zhi et al., 1999), and 113 paralogs were found in the

genome (Hotopp et al., 2006). The gene expression of p44 paralogs are enmeshed with a

unique gene conversion mechanism involving RecF pathway (Lin et al., 2006; Wang et al., 2004; Zhi et al., 2002). Compared to the well-studied p44 mRNA expression, the proteins of P44 paralogs are less defined, and only P44-18 protein has been shown as surface-exposed (Kim and Rikihisa, 1998). Recently, A. phagocytophilum genome sequencing data provide a wealth of new genetic information (Hotopp et al., 2006).

However, in addition to P44-18, there is no experimental evidence demonstrating any other A. phagocytophilum surface-exposed proteins. Furthermore, almost half of the predicted open reading frames (ORFs) of A. phagocytophilum encode conserved or novel

85 hypothetical proteins hitherto never characterized in any bacteria (Hotopp et al., 2006), some of which may be surface proteins. Therefore, it has become imperative to take a new approach, including proteomics, to generate more complete picture on the expression and function of A. phagocytophilum surface proteins.

Cell surface biotinylation has emerged as an important tool for studying cell surface protein. Sulfosuccinimidyl-2-[biotinamido]ethyl-1,3-dithiopropionate (Sulfo-

NHS-SS-Biotin) is a thiol-cleavable amine-reactive biotinylation reagent. The N- hydroxysulfosuccinimide (NHS) ester group on this reagent reacts with primary amines on a protein and form a stable conjugate. It is hydrophilic, making it membrane impermeable and thus, appropriate for surface protein labeling. The utility of Sulfo-NHS biotin reagents for cell surface labeling has been demonstrated in the identification of eukaryotic membrane proteins (Scheurer et al., 2005) as well as bacterial surface proteins such as Helicobacter pylori (Sabarth et al., 2002) and Streptococcus pyrogenes (Cole et al., 2005).

In this paper, to isolate the surface proteins of A. phagocytophilum, bacteria were surface-labeled by Sulfo-NHS biotin reagents, and the biotinylated proteins were captured by streptavidin affinity purification. The purified proteins were analyzed by proteomics. The data discovered novel surface proteins of A. phagocytophilum such as hypothetical proteins APH_0404 (here named Anaplasma surface protein 62 kDa, Asp62)

and APH_0405 (here named Anaplasma surface protein 55 kDa, Asp55).

86 4.3 Materials and methods

A. phagocytophilum and cell culture

The A. phagocytophilum HZ strain (Rikihisa et al., 1997) was propagated in HL-

60 cells in RPMI 1640 medium (Invitrogen, Carlsbad, CA) supplemented with 5% fetal

bovine serum (US Bio-Technologies, Parkerford, PA) and 2 mM L-glutamine (Invitrogen) in a humidified 5% CO2 – 95% air atmosphere at 37°C. No antibiotic was used

throughout the culture. The degree of bacterial infection in host cells was assessed by Diff-

Quik staining (Baxter Scientific Products, Obetz, OH) of cytocentrifuged preparations.

When over 90% of the cells were infected, cells were collected and centrifuged at 500 × g

for 10 min. The cell pellet was resuspended in RPMI 1640 medium at 2 × 106 cells/ml and homogenized using a 40 ml size, B type Dounce grinder (Kontes Glass, Vineland,

NJ). The homogenized suspension was subjected to centrifugation at 500 × g for 5 min, and the supernatant was collected and further purified through 2.7 µm size, 25 mm GD/X, glass microfiber syringe filter (Whatman, Florham Park, NJ). The filtrate was then centrifuged at 10,000 × g for 10 min. The pellet containing the host cell-free viable A. phagocytophilum was used immediately for biotinylation. The number of purified organisms was estimated as previously described (Yoshiie et al., 2000).

Bacterial surface biotinylation

The biotinylation of A. phagocytophilum with sulfosuccinimidobiotin (Sulfo-

NHS-Biotin) (Pierce, Rockford, IL) or Sulfo-NHS-SS-Biotin (Pierce) was performed according to the manufacturer’s instruction. Freshly purified host cell-free bacteria (2 ×

109) were washed three times in phosphate-buffered saline (PBS; 137 mM NaCl, 4.7 mM

KCl, 9.32 mM Na2HPO4 and 0.68 mM NaH2PO4, pH 8.0) containing 1 mM MgCl2

87 (PBS2+) by centrifugation at 8,000 × g for 3 min at 4°C. Bacterial pellets were

resuspended in 1 ml PBS2+ containing 1 mg Sulfo-NHS-Biotin or Sulfo-NHS-SS-Biotin.

The biotinylation reaction was performed at 4°C for 30 min. Free biotin was quenched

by washing in 500 mM glycine PBS three times. Bacterial lysates were obtained by brief sonication of biotin-labeled bacteria in radioimmunoprecipitation (RIPA) buffer (25 mM

Tris-HCl, pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS]) containing a 1:100 dilution of protease inhibitor cocktail set II

(Calbiochem, San Diego, CA). Lysates were incubated on ice for 30 min with occasional gentle vortexing. For preparing the Sulfo-NHS-SS-Biotin labeled bacterial lysates, additional oxidized glutathione (100 µM) was added to the RIPA buffer to protect the disulfide bonds in Sulfo-NHS-SS-Biotin(Scheurer et al., 2005). The biotinylated bacterial lysates were cleared by centrifugation at 16,000 × g for 10 min at 4°C. Glycerol was added to the supernatant at a final concentration of 10%. The biotinylated bacterial lysates were then stored at -80°C.

Streptavidin affinity purification of biotinylated proteins

The biotinylated A. phagocytophilum proteins were purified as previously

described (Roesli et al., 2006; Scheurer et al., 2005) with some modifications. Briefly,

300 µl of streptavidin agarose gel (Pierce) were washed three times with wash buffer A

(25 mM Tris-HCl, pH 7.6, 0.15 M NaCl, 0.5% NP-40, 0.5% sodium deoxycholate, 0.05%

SDS) and then mixed with the Sulfo-NHS-SS-Biotin-labeled bacterial lysate. The

bacterial lysates and streptavidin gel were incubated on ice for 2 h. Then, the mixture

was centrifuged at 500 × g for 1min, and the supernatant was discarded. The gel slurry

was transferred to an Ultrafree-MC centrifugal filter device (Durapore PVDF 5.0 µm, 88 Millipore). Unbound proteins were washed away with buffer B-1 (25 mM Tris-HCl, pH

7.6, 0.65 M NaCl, 0.1% NP-40) twice, followed by once with buffer B-2 (25 mM Tris-

HCl, pH 7.6, 1.15 M NaCl, 0.1% NP-40) and once with Tris-HCl buffer (25 mM Tris-

HCl, pH 7.6, 0.15 M NaCl) at 200 × g for 15 s. The captured bacterial proteins were

eluted from streptavidin agarose with 5% 2-mercaptoethanol PBS at 30°C for 30 min.

Elution in 5% 2-mercaptoethanol PBS was repeated three times. The eluate was pooled and proteins were precipitated in 10% TCA on ice as described previously (Ge and

Rikihisa, 2006). The precipitates were pelleted by centrifugation at 18,000 × g for 10 min. The protein pellet was washed once in cold acetone and air-dried. The pellet was

then dissolved in SDS-polyacrylamide gel electrophoresis (PAGE) sample buffer (50 mM

Tris-HCl, pH 6.8, 5% 2-mercaptoethanol, 2% SDS, 0.1% bromophenol blue, 10%

glycerol), boiled for 5 min and stored at - 80°C.

Proteomic analysis

The streptavidin agarose affinity-purified proteins were separated by 10% SDS-

PAGE. Seven bands of relatively abundant proteins were identified by the Mass

Spectrometry & Proteomics Facility (Campus Chemical Instrument Center, The Ohio

State University). Briefly, bands were excised from the gel and digested with sequencing

grade trypsin (Promega, Madison, WI) or chymotrypsin (Roche, Indianapolis, IN) using

the Montage In-Gel Digestion Kit (Millipore) following the manufacturer’s

recommended protocols. Capillary-liquid chromatography-nanospray tandem mass

spectrometry (Nano-LC/MS/MS) was performed on a Thermo Finnigan LTQ mass spectrometer equipped with a nanospray source operated in positive ion mode. The LC system was an UltiMate™ Plus system (LC-Packings A Dionex Co., Sunnyvale, CA) 89 with a Famos autosampler and Switchos column switcher. Five µl of each digested sample was first injected on to the trapping column (LC-Packings A Dionex Co), and washed with 50 mM acetic acid. The injector port was switched to inject, and the peptides were eluted off of the trap onto the column. A 5 cm × 75 µm ID ProteoPep II

C18 column (New Objective, Inc. Woburn, MA) packed directly in the nanospray tip was

used for chromatographic separations. Peptides were eluted directly off the column into the LTQ system using a gradient of 2-80% acetonitrile over 50 min, with a flow rate of

300 nl/min. The scan sequence of the mass spectrometer was programmed for a full scan, a zoom scan to determine the charge of the peptide and a MS/MS scan of the most abundant peak in the spectrum. Sequence information from the MS/MS data was processed using Mascot Distiller to form a peaklist (.mgf file) and using the MASCOT

MS/MS search engine and Turbo SEQUEST algorithm in BioWorks 3.1 Software.

In silico analysis of proteins Asp62 and Asp55

Amino acid sequences were analyzed by the Protean from DNASTAR software

(DNASTAR Inc., Madison, WI). Transmembrane β strands and their topology with respect to outer membrane lipid biolayer were predicted at the web server of PRED-

TMBB (http://bioinformatics.biol.uoa.gr/PRED-TMBB) (Bagos et al., 2004). BLAST

search of amino acid sequence homology was performed at the web server of National

Center for Biotechnology (NCBI) (http://www.ncbi.nlm.nih.gov) using non-redundant database. The gene annotation of A. phagocytophilum HZ and E. chaffeensis Arkansas is from the genome sequencing data (Hotopp et al., 2006).

90 RNA isolation and RT-PCR

Total RNA was extracted from A. phagocytophilum-infected HL-60 cells (5 × 106)

using an RNeasy Mini RNA extraction kit (Qiagen, Valencia, CA) according to the

manufacturer’s instructions. The concentration and purity of the RNA were determined

by measuring the A260 and the A260/A280 ratio with a GeneQuant II RNA and DNA

calculator (Pharmacia Biotech Inc., Piscataway, NJ). Five µg of the extracted RNA was

treated with 1 units of DNase I (Amplification grade) (Invitrogen) at 25°C for 10 min.

DNase I was then inactivated by the addition of 1 µl of 25 mM EDTA and subsequent

heating at 65°C for 10 min. The DNase I-treated RNA was added to a 30 µl reaction mixture containing 1 × reaction buffer (50 mM Tris-HCl pH 8.3, 75 mM KCl, 3 mM

MgCl2), 10 mM DTT, 375 ng random primers (Invitrogen), 60 units of RNaseOUT

(Invitrogen) and 0.5 mM each of dNTP mixture. After addition of 300 units of

SuperScript III reverse transcriptase (Invitrogen), the reaction mixture was incubated for

5 min at 25°C, and then, the reverse transcription reaction was carried out at 50°C for 50 min and terminated by incubating at 70°C for 15 min.

To examine the transcription of Asp62 and Asp55 genes and the intergenic region

in A. phagocytophilum, 1 µl of cDNA was amplified in a 25 µl reaction mixture

containing 1 × reaction buffer (20 mM Tris-HCl pH 8.3, 50 mM KCl, 1.5 mM MgCl2),

0.2 mM of dNTP mixture, 0.4 µM of forward and reverse primers and 1 units of Taq

DNA polymerase (Invitrogen) in a DNA thermal cycler GeneAmp PCR System 9700

(Perkin-Elmer, Foster City, CA). After heating at 94°C for 5 min, one PCR cycle consisted of denaturation at 94°C for 60 s, annealing at 55°C for 60 s and extension at

72°C for 90 s. PCR reactions were performed for 29 cycles. The final extension was 91 carried out at 72°C for 10 min. Next, PCR products (10 µl) were electrophoresed in a

1.2% agarose gel containing 0.5 µg/ml of ethidium bromide. DNA size markers (1Kb

Plus DNA Ladder; Invitrogen) were run in parallel. Based on genome sequencing data

(Hotopp et al., 2006), primers for Asp62-F1 (position, 1336-1356 nt;

5’CGCAATGATGCTAGGAACGTT 3’), Asp62-R1 (position, 1532-1512 nt; 5’

AGCACGCAGCGCATACTCTCC 3’), Asp55-F1 (positon, 67-87 nt;

GGAGAGCGTGCGTCGGTAACG 3’) and Asp55-R1 (position, 407-387 nt; 5’

ATACCAGGCGCACCATGAAAC 3’) were designed for this study. Primer pairs used

were Asp62-F1 and Asp62-R1 for Asp62, Asp55-F1 and Asp55-R1 for Asp55 and Asp62-

F1 and Asp55-R1 for amplifying the co-transcribed mRNA of Asp62 and Asp55.

Surface localization of Asp62 and Asp55 by immunofluorescence microscopy

Two relatively highly antigenic and hydrophilic peptide fragments were chosen from Asp62 and Asp55 amino acid sequences based on the analysis by Protean from

DNASTAR software. The 19-mer peptide, CRYNTRDVYHRDVGYKDHG, corresponding to the sequence from Asp62 C-terminus (534-552 aa) was synthesized and conjugated to keyhole limpet hemocyanin (KLH), and rabbit antibody was developed by

Proteintech Group, Inc. (Chicago, IL). The 15-mer peptide, CHEYKSTESSGFVLKE, underlined sequence corresponding to the 14 amino acids from Asp55 C-terminus (501-

515 aa) was synthesized and conjugated to KLH, and rabbit antibody was made by Sigma

Genosys (St. Louis, MO). According to BLAST search for short, nearly exact matched sequences in NCBI non-redundant database, these two peptide sequences had little to no homology to any other known proteins (E>25) and thus, were unique to each of Asp62 and Asp55 proteins.

92 For immunofluorescence microscopic analysis of A. phagocytophilum Asp62 and

Asp55 localization, paraformaldehyde-fixed bacteria were used as described previously

(Ge and Rikihisa, 2006; Wang et al., 2006). Briefly, host cell-free A. phagocytophilum

was pelleted and washed in PBS (137 mM NaCl, 2.68 mM KCl, 10 mM Na2HPO4 and

1.76 mM KH2PO4, pH7.4). All subsequent steps were performed at room temperature.

Bacteria were fixed in 2% paraformaldehyde for 45 min followed by quenching in PBS

containing 0.1 M glycine. After being washed in PBS, bacteria were incubated with 1:

100 diluted rabbit antisera against Asp62, Asp55 peptide, or rabbit pre-immune serum in

PG buffer (0.2% gelatin in PBS) for 1 h. After being washed in PG, the bacteria were

labeled with Alexa Fluor 488 goat anti-rabbit IgG (Invitrogen) at a dilution of 1:100 in

PG for 1 h. The bacteria were washed in PG, resuspended in PBS and observed using a

Nikon Eclipse E400 fluorescence microscope with a xenon-mercury light source (Nikon

Instruments, Melville, NY).

Western blotting analysis

The protein samples were subjected to 10% SDS-PAGE and transferred to polyvinylidene fluoride (PVDF) membranes (Westran S; Schleicher & Schuell

BioScience, Keene, NH). The membranes were incubated with anti-A. phagocytophilum

sera from experimentally infected horse (EQ005), collected on day 31 after infected tick

attachment (Wang et al., 2004), or human HGA patient NY31 (Lin et al., 2002) at a dilution of 1:1,000 at 4°C overnight, which was followed by incubation with horseradish peroxidase (HRP)-conjugated 1:1,000 diluted goat anti-horse or anti-human IgG (KPL,

Gaithersburg, MA) at room temperature for 1 h. For biotin blotting, the membrane was

93 incubated with 1:1,000 diluted HRP-conjugated streptavidin (Invitrogen) at room temperature for 1 h. The blots were developed by using the enhanced chemiluminescence

(ECL) kit (Pierce).

4.4 Results

Detection of biotinylated proteins

Since the Sulfo-NHS-Biotin or Sulfo-NHS-SS-Biotin labeling reagent is charged

by the sodium sulfoxide group on the succinimidyl ring, it is water soluble and

membrane-impermeable, which makes the reagent suitable for labeling cell surface

proteins via reaction with primary amines (Cole et al., 2005). Sulfo-NHS-SS-Biotin

enables isolation of biotin-labeled cell surface proteins by streptavidin affinity

chromatography via cleavage with a reducing agent (Scheurer et al., 2005). This feature,

nonetheless, prohibits the use of reducing agents in the SDS-PAGE sample buffer when

observing sulfo-NHS-SS-Biotin-labeled proteins via biotin blotting, which makes

molecular size pattern discrimination inaccurate. Therefore, we also labeled host cell-

free A. phagocytophilum with Sulfo-NHS-Biotin to show the approximate molecular

sizes of biotinylated proteins in the presence of reducing agents. The biotinylated proteins

were blotted onto the membrane and probed with HRP-conjugated streptavidin. As

shown in Fig. 1 Lane 2, Sulfo-NHS-Biotin labeling revealed A. phagocytophilum surface

proteins (approximate molecular mass of 30, 36, 40, 46, 55, 65, 75, 85, 105, 120, and 170

kDa). To preserve the disulfide bonds in Sulfo-NHS-SS-Biotin, there was no reducing

agent added to the sample buffer of Sulfo-NHS-SS-Biotin-labeled proteins (Fig. 1). .

94 Streptavidin affinity purification of biotinylated A. phagocytophilum surface proteins

To identify biotinylated A. phagocytophilum surface proteins, Sulfo-NHS-SS-

Biotin-labeled host cell-free bacteria were solubilized in RIPA buffer. The biotinylated proteins were purified by streptavidin affinity gel chromatography. The disulfide bonds in Sulfo-NHS-SS-Biotin were cleaved with the reducing agent to elute the streptavidin affinity-captured proteins. The captured proteins were separated by SDS-PAGE. As shown in Fig. 2, with GelCode blue protein staining in the SDS-PAGE gel (with the reducing agent in the sample buffer), there were seven bands of relatively abundant proteins corresponding to molecular masses of approximately 28, 40, 45, 55, 65, 85 and

105 kDa. The labeling efficiency was slightly different between Sulfo-NHS-Biotin and

Sulfo-NHS-SS-Biotin due to spacer arm length (Braschi and Wilson, 2006), and the detection sensitivity was also different due to the employment of different methods to detect biotin (Fig. 1) and protein (Fig. 2). Nonetheless, the overall molecular sizes of biotinylated proteins with Sulfo-NHS-Biotin (Fig. 1, lane 2) approximately matched with affinity-purified Sulfo-NHS-SS-Biotin-labeled proteins after cleavage of S-S bond (Fig.

2).

Nano-LC/MS/MS

The seven bands of relatively abundant proteins (labeled as bands 1-7 in Fig. 2) were subjected to proteomic analysis. Although the 120 kDa protein was strongly labeled with biotin (Fig.1), it was less abundant when analyzed by SDS-PAGE (Fig. 2) and not sufficient for proteomic detection. Table 1 summarized totally 16 A. phagocytophilum proteins identified by Nano-LC/MS/MS. Half of them were integral membrane proteins with a cleavable signal peptide and multi-pass transmembrane β strands when analyzed

95 by PRED-TMBB (The number of β strands was not shown in Table 1). The remainings

were either single-pass transmembrane proteins or might be peripheral membrane

proteins. Band 1 was a hypothetical protein APH_0441 (YP_505044). Band 2 mainly was

an outer membrane protein OMP85 (YP_505741). Bands 3 and 4 mainly were

hypothetical proteins APH_0404 (YP_505009) (Asp62) and APH_0405 (YP_505010)

(Asp55), respectively. Band 5 consisted of several P44 family proteins, i.e., P44-18ES

(YP_505752), P44-2b (YP_505759) and P44-59 (AAQ16676). Band 6 was human β actin (not shown in Table 1). Band 7 was OMP-1A, a protein predicted to belong to

OMP-1/P28 outer membrane protein family of Ehrlichia species (Hotopp et al.,

2006)(Ohashi et al., 2001). In addition to these major proteins separated by 10% SDS-

PAGE, several proteins with relatively low abundance were also detected by Nano-

LC/MS/MS (Table 1), e.g., translation enlongation factor G (YP_505591), pentapeptide

repeat family protein (YP_505709) and type IV secretion protein VirB8 (YP_505898).

In silico analysis of Asp62 and Asp55 proteins

PRED-TMBB is a web server capable of predicting the transmembrane strands and the topology of β-barrel outer membrane proteins of Gram-negative bacteria, which is based on a Hidden Markov Model and trained with non-homologous outer membrane proteins with structures known at atomic resolution according to the Conditional

Maximum Likelihood criterion (Bagos et al., 2004). Predicted by posterior decoding method using the dynamic programming algorithm in PRED-TMBB, there were 22 transmembrane β strands in each of Asp62 (Fig. 3 A and B) and Asp55 (Fig. 3 C and D) proteins, and these β strands were connected by 11 long loops on bacterial external side

and 10 short turns on the periplasmic space. The discrimination scores of Asp62 and

96 Asp55 for β barrel proteins were 2.956 and 2.905, respectively. Scores lower than the threshold value of 2.965 is considered the analysis significant (Bagos et al., 2004), suggesting that they are β-barrel outer membrane proteins.

Physical map of ortholog genes of Asp62 and Asp55 is shown in Fig. 4. Asp62 and Asp55 were paralogs with an E value of 3e-24. One more A. phagocytophilum paralog, hypothetical proteins APH_0406, was found downstream of Asp55, and another gene, aph_0407, was predicted to encode for a short protein of 65 amino acids, which was almost identical to a stretch of amino acids in Asp62. Based on BLAST search in NCBI database, the homologous proteins of A. phagocytophilum Asp62 and Asp55 were found in A. marginale str. St. Maries, E. chaffeensis Arkansas, E. chaffeensis Sapulpa (not shown), E. canis Jake, E. ruminantium Gardel (not shown) and E. ruminantium

Welgevonden. Three tandem orthologs exist in A. marginale str. St. Maries, E. chaffeensis Arkansas and E.canis Jake. Whereas, E. ruminantium Welgevonden keeps two of them. Additionally, Wolbachia endosymbiont strain TRS of Brugia malayi has one ortholog, hypothetical protein Wbm0010 with an E value as low as 1e-26 to Asp55,

Wolbachia endosymbiont of Drosophila melanogaster one ortholog, hypothetical protein

WD0745 with the E value of 2e-25 and Rickettsia bellii OSU 85-389 one ortholog of hypothetical protein RbelO_01000075 with the E value of 8e-11 (not shown in Fig. 4). All of these organisms are obligatory intracellular bacteria that belong to α-proteobacteria, order Rickettsiales.

97 Co-transcription of Asp62 and Asp55

The stop codon of Asp62 and start codon of Asp55 are separated by only 65 nucleotide (nt) intergenic region (Hotopp et al., 2006), leading to the hypothesis that they are co-transcribed. To test whether they are co-transcribed experimentally, RT-PCR of

A. phagocytophilum Asp62, Asp55 and the common transcript of Asp62 and Asp55 were performed with the absence of reverse transcriptase (RT -) as negative control. As shown in Fig. 5, Asp62, Asp55 and the common transcript of Asp62 and Asp55 including 65 nt intergenic region were all transcribed by A. phagocytophilum cultured in HL-60 cells at

37°C. The transcripts for RT- control were undetectable. These results indicate that the genes of Asp62 and Asp55 paralogs are organized within one operon.

Surface localization and antigenicity of Asp62 and Asp55

To verify the localization of Asp62 and Asp55 on A. phagocytophilum surface, the host cell-free A. phagocytophilum was paraformaldehyde-fixed first to prevent antibody permeabilization ( Ge and Rikihisa, 2006; Wang et al., 2006) and incubated with rabbit antiserum against Asp62 or Asp55 C-terminal peptide. The results confirmed not only the proteomic identification of Asp 62 and Asp55, but also surface exposure of

C-terminal peptides as determined by in silico analysis. As shown in Fig. 6A and B, both the antisera against Asp62 and Asp55 labeled the surface of individual bacteria of various sizes with a mottled ring-like staining pattern. Pre-immune rabbit serum did not label A. phagocytophilum (Fig. 6C).

Streptavidin affinity-captured A. phagocytophilum proteins were subjected to

Western blot analysis with sera from experimentally infected horse and human patient

(Fig. 7A and B). Experimentally infected horse serum, collected on day 31 after infected

98 tick attachment, recognized several bands, including those corresponding to OMP85,

Asp62, P44s and a protein with high molecular weight of approximately 120 kDa.

Human patient serum recognized several protein bands, corresponding to the 120 kDa protein, Asp62, Asp55 and P44s. Sera of normal horse and human from HGA-

nonendemic region (Ohio and Japan, respectively) did not react with A. phagocytophilum

proteins (not shown in Fig. 7).

4.5 Discussion

The present work has discovered novel surface-exposed proteins of A.

phagocytophilum using affinity purification and proteomic methods. In addition to P44s, the hypothetical proteins of APH_0404 and APH_0405 (Asp62 and Asp55) are the other two abundantly expressed A. phagocytophilum surface proteins. In this paper, the identification of known P44s, OMP-1A outer membrane protein families and a conserved outer membrane protein OMP85 validates the surface bintinylation experimental approach performed here. This is the first time that this method has been used to investigate the surface proteins of an obligate intracellular bacterium.

For the first time, Asp62 and Asp55 proteins have been experimentally shown being surface-exposed and expressed at the protein level. They are highly conserved in the family Anaplasmataceae, suggesting that they play an essential role in these bacterial survival. These related orthologs are all annotated as hypothetical proteins, and their

functions have not been elucidated yet. Both Asp62 and Asp55 have been predicted as β-

99 barrel outer membrane proteins with a secondary structure of 22 transmembrane β strands by posterior decoding method in PRED-TMBB web server, suggesting that Asp62 and

Asp55 function as outer membrane transporters (Bagos et al., 2004). Twenty two stranded β-barrel structure has been revealed in some bacterial outer membrane siderophore receptors by crystal structural data, i.e., FepA (Buchanan et al., 1999), FhuA

(Pawelek et al., 2006) and FecA from Escherichia coli (Ferguson et al., 2002) and FpvA

(Cobessi et al., 2005b) and FptA (Cobessi et al., 2005a) from Pseudomonas aeruginosa, which act as transporters to take up iron. When ligands bind to these normally closed transporters, they exhibit conformational changes that activate them to be open, hence their designation “ligand-gated porin” (Jiang et al., 1997). Therefore, it would be interesting to functionally characterize the newly uncovered A. phagocytophilum surface proteins Asp62 and Asp55, which is underway in our lab.

Many members of p44s are functional pseudogenes. In other words, although lacking the translational start site, they can be expressed as a full-length P44 protein (44- kDa) after RecF-dependent recombination into the p44-expression locus (Lin et al.,

2006). So far, 65 different p44s had been shown to be transcribed, and P44-2 and P44-18 have been shown to be translated (Felek et al., 2004; Lin et al., 2003; Zhi et al., 2002;

Zhi et al., 1999). Nevertheless, this is the first demonstration of surface exposure of P44-

2 and P44-59 proteins.

The predicted molecular mass of hypothetical protein APH_0441 was 65,911 Da, much less than the molecular mass of approximately 105 kDa, deduced from the actual migration distance by 10% SDS-PAGE (Fig. 2). This could be due to post-translational modification such as glycosylation proposed for gp47 protein of E. chaffeensis (Doyle et

100 al., 2006). OMP85 is a conserved outer membrane protein in Gram-negative bacteria

(Gentle et al., 2005) and is a central component of the apparatus for outer membrane

protein assembly (Robert et al., 2006; Voulhoux et al., 2003). It has been shown as outer

membrane protein of Neisseria gonorrhoeae (Manning et al., 1998), but never been

experimentally shown to be surface-exposed, or expressed at the protein or mRNA level

by A. phagocytophilum. Type IV secretion protein VirB8 is a core component of the type

IV secretion system apparatus. A clustered distribution of VirB8 over the bacterial surface has been demonstrated in Agrobacterium (Kumar et al., 2000), and it has been recently proposed to function as the assembly factor to target the type IV apparatus to the

cell pole (Judd et al., 2005).

Heat shock protein GroEL and translation elongation factor (EF)-G used to be

believed as cytoplasmic proteins. They may have been released from autolysed bacteria

and bind to the intact A. phagocytophilum within the host inclusions as a part of a normal

turnover process. Another possibility is that some bacterial outer membrane may be

damaged during isolating host cell-free bacteria before biotin-labeling. However, by

immunofluorescence or immunoelectron microscopy, GroEL has been shown on the

surface of Helicobacter pylori (Huesca et al., 1996), Legionella pneumophila (Garduno et

al., 1998), Haemophilus ducreyi (Frisk et al., 1998), Clostridium difficile (Hennequin et

al., 2001) and translation elongation factor u (EF-Tu) on the surface of Mycoplasma

pneumoniae (Dallo et al., 2002). EF-Tu has also been reported to be cell surface-

associated by proteomic analysis of the A. marginale outer membrane fraction (Lopez et

al., 2005). It is still unknown the mechanism how they are secreted to bacterial surface.

Mycoplasma pneumoniae (Dallo et al., 2002) cell surface EF-Tu functions as the

101 component of fibronectin binding. Haemophilus ducreyi GroEL has been reported to

mediate the binding of this bacterium to host carbohydrate receptors (Pantzar et al., 2006).

Therefore, A. phagocytophilum GroEL and EF-G may be surface-exposed and play a role

in the interaction between host and bacteria.

One obvious protein band captured by streptavidin agarose affinity purification

was host cell actin, which is one of the most abundant cytoskeleton proteins of eukaryotic

cells. This may be due to the binding of host cell actin to bacterial surface proteins during

the isolation of host cell-free bacteria or functional association. Actin association with

isolated obligate intracellular bacterium, rickettsia was previously reported to confer

intracellular motility to rickettsia (Martinez et al., 2005).

Since surface proteins of A. phagocytophilum are exposed to the host immune

system, they may represent major antigens. The most abundant and also only studied

immunodominant surface proteins of A. phagocytophilum are P44/Msp2 family proteins.

The central hypervariable region of P44 molecules is exposed on the bacterial surface and

has been proposed to be involved in antigenic variation and immune evasion (Wang et al.,

2004; Lin et al., 2002; Zhi et al., 1999; Kim and Rikihisa, 1998). Antibodies against P44s have been detected in human patients as well as in experimentally infected animals ( Lin

et al., 2002; Tajima et al., 2000; Kim and Rikihisa, 1998; JW et al., 1997; Dumler et al.,

1995). Two anti-Msp2 (P44) monoclonal antibodies (MAbs) and a recombinant Msp2

weakly block A. phagocytophilum adhesion and infection of HL-60 cells (Park et al.,

2003). However, the two P44 MAbs, 5C11 and 3E65, almost completely block the infection of the A. phagocytophilum population that predominantly expressed P44-18 in

102 HL-60 cells (Wang et al., 2006). Passive immunization of mice with these two MAbs

partially protects mice from challenge (Kim and Rikihisa, 1998). These studies suggest that antibodies against specific epitopes of P44s are involved in immuneprotection.

Asp62 and Asp55 have been shown to be immunogenic in a human patient in this paper. The experimentally infected horse serum recognized the protein band

corresponding to Asp62 protein. For P44s, the horse serum showed a stronger immune

reaction with the corresponding protein bands than the human patient serum. The

discrepancy of immune responses between human patient and experimentally infected

horse could be largely due to the facts that the serum from human patient was collected from acute phase, and that the horse serum from convalescent phase. Therefore, the

newly uncovered proteins extend the range of known surface-exposed proteins to

rationally select antigen candidates for vaccine development as well as sero-diagnosis in

the acute phase of HGA. Interestingly, both human and horse sera recognized an A. phagocytophilum protein of approximately 120 kDa. The protein was detected by

Western blot analysis but not by SDS-PAGE or Nano-LC/MS/MS. This may be due to its abundance below the detection threshold of the method used here. Nevertheless, it is an appropriate antigen for diagnosis since it is highly immunogenic in the early and late phases of A. phagocytophilum infection. In the future, the identity of this protein deserves more investigation.

In conclusion, the method of biotinylation of A. phagocytophilum has been developed to systematically identify the obligate intracellular bacterial surface proteins that are promising targets for future study on interaction between this bacterium and host, the diagnosis and control of this important tick-borne human pathogen.

103

Annotation or Predicted Nano- GenBank Subcellular Predicted Protein Signal Gene locus tag molecular LC/MS/MS accession localization function peptide mass sequence number sequence coverage by (Yes/No ammino acid position) (%) hypothetical protein 65911 4 YP_505044 Cytoplasmic Unknown; putative Yes (35) APH_0441 membrane lipoprotein outer membrane 85709 30 YP_505741 OuterMembrane Cell envelope Yes (21) protein, OMP85 family biogenesis translation elongation 76234 32 YP_505591 Cytoplasmic GTPase, power No factor G translation chaperone protein 70019 16 YP_504953 Cytoplasmic Chaperone No DnaK ppiC/parvulin rotamase 67535 15 YP_505198 Cytoplasmic Accelerate the folding No family protein of proteins hypothetical protein 63987 42 YP_505009 OuterMembrane Unknown Yes (19) APH_0404 (Asp62) 60 kDa chaperonin 57282 35 YP_504857 Cytoplasmic Chaperone No (GroEL) pentapeptide repeat 61870 5 YP_505709 Unknown Uncharacterized No family protein secreted protein (YP_198411) containing pentapeptide repeats [Wolbachia endosymbiont strain TRS of Brugia malayi] hypothetical protein 57595 13 YP_505010 OuterMembrane Unknown Yes (22) APH_0405 (Asp55)

Continued

Table 1. Surface-exposed proteins of A. phagocytophilum analyzed by Nano-

LC/MS/MS. The program SignalP 3.0 (http://www.cbs.dtu.dk/services/SignalP) was used to predict the presence of N-terminal signal peptides (Bendtsen et al., 2004).

Putative lipoprotein was predicted by LipoP 1.0. Subcellular localization of proteins was analyzed by PSORTb version 2.0.4 (http://www.psort.org/psortb) (Gardy et al., 2005).

Grey highlight marks the boundary between protein bands of different molecular sizes.

N/A represents not available.

104 Table 4.1 continued

P44-18ES, P44 outer 46027 38 YP_505752 Unknown Porin Yes (20) membrane protein expression locus with P44-18 P44-2b N/A 39 YP_505759 P44 paralog P44-59 N/A 10 AAQ16676 P44 paralog; pseudogene major outer membrane 31845 41 YP_505858 OuterMembrane Unknown Yes (19) protein OMP-1A type IV secretion 27582 16 YP_505898 Unknown Transportation of No protein VirB8-1 macromolecules across bacterial inner and outer membranes thiol:disulfide 27842 8 YP_504749 Unknown Protein folding and Yes (21) oxidoreductase stabilization co-chaperone GrpE 22761 8 YP_504670 Cytoplasmic Chaperone No

105

Fig. 4.1 Biotin labeling of A. phagocytophilum surface proteins. Biotinylated A. phagocytophilum was lysed in RIPA buffer, separated by 10% SDS-PAGE and blotted with HRP-conjugated streptavidin. Lane 1, Sulfo-NHS-SS-Biotin labeled A. phagocytophilum (in the absence of reducing agents in SDS-PAGE sample buffer). Lane

2, Sulfo-NHS-Biotin labeled A. phagocytophilum. Precision Plus prestained protein standards (Bio-Rad).

106

Fig. 4.2 Streptavidin agarose affinity purification of Sulfo-NHS-SS-Biotin-labeled A. phagocytophilum surface proteins. Lane 1, Sulfo-NHS-SS-Biotin-labeled A. phagocytophilum surface proteins were separated by 10% SDS-PAGE with a reducing agent in sample buffer and stained with GelCode blue. Bands 1-7 were subjected to

Nano-LC/MS/MS analysis. Marker, Precision Plus prestained protein standards (Bio-

Rad).

107 Fig. 4.3 The 2-D structure prediction of A. phagocytophilum Asp62 and Asp55 with

respect to the outer membrane lipid bilayer using Posterior Decoding method in

PRED-TMBB. (A) Plot of the posterior probabilities for the transmembrane β strands,

along the sequence of Asp62. Red color represents for transmembrane region, green for

bacterial inside and blue for bacterial outside. (B) Graphical representation of the

predicted topology with respect to the outer membrane lipid bilayer of Asp62. (C) Plot of

the posterior probabilities for the transmembrane β strands, along the sequence of Asp55.

(D) Graphical representation of the predicted topology with respect to the outer membrane lipid bilayer of Asp55.

108 Fig. 4.3

A Asp62

B

Asp62

Continued

109 Fig. 4.3 continued

C Asp55

D

Asp55

110

Fig. 4.4 Schematic diagram of the organization of genes encoding the APH_0404

(Asp62), APH_0405 (Asp55) and APH_0406 in A. phagocytophila HZ and the orthologus genes in A. marginale str. St. Maries, E. chaffeensis Arkansas, E.canis

Jake and E. ruminantium Welgevonden. ORFs are represented as open arrows indicating their orientations. The alignment indicates orthologs with dashed-lines from both ends of each ORF. The number of amino acid residues for each ORF is shown. The

E value cut off is e-22.

111

Fig. 4.5 Co-transcriptional analysis of Asp62 and Asp55 by RT-PCR. Total RNA was isolated from A. phagocytophilum-infected HL-60 cells. M, Marker (1Kb Plus DNA

Ladder, Invitrogen); Lane 1, the co-transcript of Asp62 and Asp55 including 65 nt intergenic region; lane 3, Asp62; lane 5, Asp55; lanes 2, 4 and 6, RT- of the co-transcript of Asp62 and Asp55, Asp62 and Asp55, respectively. RT- indicates the absence of reverse transcriptase. The amplicon sizes were in agreement with predicted amplicon length, i.e.,

197 bp for Asp62, 340 bp for Asp55 and 795 bp for the co-transcript of Asp62 and Asp55.

112

Fig. 4.6 Surface localization of A. phagocytophilum Asp62 and Asp55 by imunofluorescence assay. Host cell-free A. phagocytophilum were fixed in paraformaldehyde, incubated with rabbit serum against Asp62 C-terminal (534-552 aa) or

Asp55 C-terminal peptide (501-515 aa), stained with Alexa Fluor 488 goat anti-rabbit

IgG and visualized by fluorescence microscopy. (A) Mottled Ring-like bacterial surface staining of Asp62. (B) Mottled Ring-like bacterial surface staining of Asp55. (C).

Bacteria surface-stained with rabbit pre-immune serum. Scale bar, 1 µm.

113

Fig. 4.7 Antigenicity of streptavidin agarose affinity purification of Sulfo-NHS-SS-

Biotin-labeled A. phagocytophilum surface proteins in human patient and experimentally infected horse. Streptavidin affinity-purified Sulfo-NHS-SS-Biotin- labeled A. phagocytophilum surface proteins incubated with a human patient serum NY31

(Lane 1) or an experimentally infected horse EQ005 serum (Lane 2) by western blot analysis. Marker, Precision Plus prestained protein standards (Bio-Rad).

114

CHAPTER 5

Surfaceome of Ehrlichia chaffeensis

5.1 Abstract

The surface proteins of Ehrlichia chaffeensis provide an important interface for pathogen-host interactions. To globally investigate the surface proteins of E. chaffeensis, membrane-impermeable, cleavable Sulfo-NHS-SS-Biotin was used to label intact bacteria. The biotinylated bacterial surface proteins were isolated by streptavidin agarose affinity-purification. The affinity-captured proteins were separated by electrophoresis, and five relatively abundant protein bands were subjected to Nano-LC/MS/MS analysis.

Nineteen out of 22 OMP-1/P28 family proteins, including P28 (which previously was shown to be surface exposed), were detected in E. chaffeensis cultured in human monocytic leukemia THP-1 cells. For the first time, with the exception of P28 and P28-1,

17 OMP-1/P28 family proteins were demonstrated to be expressed at the protein level.

The surface exposure of OMP-1A and OMP-1N was verified by immunofluorescence microscopy. OMP-1B was undetectable either by surface biotinylation or by Western blotting of the whole bacterial lysates, suggesting that it is not expressed by E. chaffeensis cultured in THP-1 cells. Additional E. chaffeensis surface proteins detected were OMP85, hypothetical protein ECH_0525 (here, named Esp73), immunodominant

115 surface protein gp47 and 11 other proteins. Western blot analysis revealed that four

major bands of purified E. chaffeensis surface proteins were immunogenic in

experimentally infected dogs or human ehrlichiosis patients. The identification of E.

chaffeensis surface-exposed proteins provides novel insights about the E. chaffeensis

surface and lays the foundation for rational studies on pathogen-host interactions and

vaccine development.

5.2 Introduction

Human monocytic ehrlichiosis (HME) is an emerging tick-borne zoonosis in the

United States (Demma et al., 2005). It is an acute febrile systemic disease that can cause

severe and potentially fatal disease especially in immunocompromised and elderly people

(Paddock et al., 1997)(Demma et al., 2005). The etiologic agent of HME is Ehrlichia chaffeensis, belonging to the family Anaplasmataceae. In North America, the major vector of E. chaffeensis is the Lone Star tick, Amblyomma americanum, and the white- tailed deer is considered to be the major reservoir of E. chaffeensis as well as

Amblyomma americanum (Rikihisa, 2003). E. chaffeensis is a Gram-negative, obligatory intracellular bacterium, which has tropism for monocytes/macrophages. The entry and proliferation of E. chaffeensis involve host caveolae, glycosylphosphatidylinositol (GPI)- anchored proteins and the incorporation of cholesterol into the bacterial membrane (Lin and Rikihisa, 2003). After internalization by host monocytes, E. chaffeensis has the ability to subvert the hostile environment by residing in an early endosome-like

116 compartment, which does not fuse with lysosomes (Mott et al., 1999). These events are ehrlichial surface-related. However, the corresponding bacterial surface components have not been characterized.

The humoral immune response plays an important role in the protection of the host against ehrlichial infection. Passive immunization of immunocompetent (Kaylor et

al., 1991)(Sun et al., 1997) or immunocompromised (Winslow et al., 2000)(Li et al.,

2002) animals provides effective protection against ehrlichial agents such as

Neorickettsia risticii, Anaplasma phagocytophilum and E. chaffeensis infection.

Conversely, anti-E. chaffeensis antibodies bound to the E. chaffeensis surface induce

potent proinflammatory cytokine mRNA expression in human monocytes, which may

contribute to the pathogenesis of HME (Lee and Rikihisa, 1997). Assuming that the

monoclonal antibodies are representative of the polyclonal response, the mouse antibody

response against OMP-1g (P28) accounts for 40% of the total antibodies against E.

chaffeensis (Li et al., 2001). This suggests that ehrlichial outer membrane proteins are

the dominant immunogens that stimulate the host humoral immune response. Western

blot analysis revealed several E. chaffeensis immunoreactive proteins with molecular

masses of approximately 74, 70, 64, 47, 31 and 29 kDa (Rikihisa et al., 1994). However,

their identities of these proteins are not well known.

Studies on bacteria from the family Anaplasmataceae have revealed a critical role

for the bacterial outer membrane proteins in the stimulation of the host immune response

and protection of the host from infection. Immunization with recombinant P28 (one of

the major outer membrane OMP-1/P28 family members) protected mice from E.

chaffeensis challenge (Ohashi et al., 1998). Immunization of calves with Anaplasma

117 marginale outer membrane proteins induced stronger protection against challenge compared to individual major surface proteins (MSP), e.g., MSP-1 and MSP-2 (Brown et al., 1998)(Tebele et al., 1991)(Palmer et al., 1986)(Palmer et al., 1989)(Abbott et al.,

2005). Along with this line, efforts have been made to identify the global composition of

A. marginale outer membrane immunogens (Lopez et al., 2005).

Despite the importance of E. chaffeensis surface proteins as a crucial interface for pathogen-host interactions as mentioned above, the knowledge of E. chaffeensis surface proteins is limited. In addition to P28 (Ohashi et al., 1998), two other E. chaffeensis proteins, gp47 (Doyle et al., 2006) and gp120 (Popov et al., 2000), have been identified as being surface-exposed by immunoelectron microscopy. There has been no systematic investigation of the surface proteins of E. chaffeensis. Therefore, this paper focuses on the characterization of E. chaffeensis major surface proteins via surface biotinylation using cleavable sulfosuccinimidyl-2-[biotinamido]ethyl-1,3-dithiopropionate (Sulfo-

NHS-SS-Biotin) labeling (Scheurer et al., 2005)(Cole et al., 2005), streptavidin affinity purification of biotinylated proteins, idenfication of these proteins by proteomic analysis and investigation of the immunogenicity of these surface proteins.

5.3 Materials and Methods

E. chaffeensis and cell culture

The E. chaffeensis Arkansas strain (Dawson et al., 1991) was propagated in THP-

1 cells, a human monocytic leukemia cell line, in RPMI 1640 medium (Invitrogen,

Carlsbad, CA) supplemented with 10% fetal bovine serum (US Bio-Technologies,

Parkerford, PA) and 2 mM L-glutamine (Invitrogen) in a humidified 5% CO2 – 95% air

118 atmosphere at 37°C. No antibiotic was used throughout the study. The degree of bacterial

infection in host cells was assessed by Diff-Quik staining (Baxter Scientific Products, Obetz,

OH) of cytocentrifuged preparations. When over 90% of the cells were infected, cells were collected and centrifuged at 500 × g for 10 min. The cell pellet was resuspended in

RPMI 1640 medium at 2 × 106 cells/ml and homogenized using a 40 ml size, B type

Dounce grinder (Kontes Glass, Vineland, NJ). The homogenized suspension was

subjected to centrifugation at 500 × g for 5 min, and the supernatant was collected and

further purified consecutively through 5 µm, 25 mm GD/X, glass microfiber syringe filter

(Whatman, Florham Park, NJ) and Millex-AA Filter 0.8 µm (Millipore, Billerica, MA).

The filtrate was then centrifuged at 10,000 × g for 10 min. The pellet containing the host

cell-free viable E. chaffeensis was used immediately for biotinylation. The number of

purified organisms was estimated as previously described (Yoshiie et al., 2000)

Cell surface biotinylation

The biotinylation of E. chaffeensis with sulfosuccinimidobiotin (Sulfo-NHS-

Biotin) (Pierce, Rockford, IL) or Sulfo-NHS-SS-Biotin (Pierce) was performed according

to the manufacturer’s instruction. Purified host cell-free bacteria (2 × 109) were washed

three times in phosphate-buffered saline (PBS; 137 mM NaCl, 4.7 mM KCl, 9.32 mM

2+ Na2HPO4 and 0.68 mM NaH2PO4, pH 8.0) containing 1 mM MgCl2 (PBS ) by

centrifugation at 8,000 × g for 3 min at 4°C. Bacterial pellets were resuspended in 1 ml

PBS2+ containing 1 mg Sulfo-NHS-Biotin or Sulfo-NHS-SS-Biotin. The biotinylation

reaction was performed at 4°C for 30 min. Free biotin was quenched by washing in 500

mM glycine PBS three times. Bacterial lysate was obtained by brief sonication of biotin-

labeled bacteria in radioimmunoprecipitation (RIPA) buffer (25 mM Tris-HCl, pH 7.6, 119 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate

[SDS]) containing a 1:100 dilution of protease inhibitor cocktail set II (Calbiochem, San

Diego, CA). Lysates were incubated on ice for 30 min with occasional gentle vortexing.

For preparing the Sulfo-NHS-SS-Biotin-labeled bacterial lysates, additional oxidized

glutathione (100 µM) was added to the RIPA buffer to protect the disulfide bonds in

Sulfo-NHS-SS-Biotin (Scheurer et al., 2005). The biotinylated bacterial lysates were cleared by centrifugation at 16,000 × g for 10 min at 4°C. Glycerol was added to the

supernatant at a final concentration of 10%. The biotinylated bacterial lysates were then stored at -80°C.

Streptavidin affinity purification of biotinylated proteins

The biotinylated proteins were purified as previously described (Scheurer et al.,

2005)(Roesli, 2006) with some modifications. Briefly, 300 µl of streptavidin agarose gel

(Pierce) were washed three times with wash buffer A (25 mM Tris-HCl, pH 7.6, 0.15 M

NaCl, 0.5% NP-40, 0.5% sodium deoxycholate, 0.05% SDS) and then mixed with the

Sulfo-NHS-SS-Biotin-labeled bacterial lysate. The bacterial lysates and streptavidin gel were incubated on ice for 2 h. Then, the mixture was centrifuged at 500 × g for 1min, and the supernatant was discarded. The gel slurry was transferred to an Ultrafree-MC centrifugal filter device (Durapore PVDF 5.0 µm, Millipore). Unbound proteins were washed away with buffer B-1 (25 mM Tris-HCl, pH 7.6, 0.65 M NaCl, 0.1% NP-40) twice, followed by once with buffer B-2 (25 mM Tris-HCl, pH 7.6, 1.15 M NaCl, 0.1%

NP-40) and once with Tris-HCl buffer (25 mM Tris-HCl, pH 7.6, 0.15 M NaCl) at 200 × g for 15 s. The captured bacterial proteins were eluted from streptavidin agarose with

5% 2-mercaptoethanol PBS at 30°C for 30 min. Elution in 5% 2-mercaptoethanol PBS 120 was repeated three times. The eluates were pooled and proteins were precipitated in 10%

TCA on ice as described previously (Ge and Rikihisa, 2006). The precipitates were

pelleted by centrifugation at 18,000 × g for 10 min. The protein pellet was washed once

in cold acetone and air-dried. The pellet was then dissolved in SDS-polyacrylamide gel

electrophoresis (PAGE) sample buffer (50 mM Tris-HCl, pH 6.8, 5% 2-mercaptoethanol,

2% SDS, 0.1% bromophenol blue, 10% glycerol), boiled for 5 min and stored at - 80°C.

Proteomic analysis

The streptavidin agarose affinity-purified proteins were separated by 10% SDS-

PAGE. The proteins in five relatively abundant bands were identified by the Mass

Spectrometry & Proteomics Facility (Campus Chemical Instrument Center, The Ohio

State University). Briefly, bands were excised from the gel and digested with sequencing

grade trypsin (Promega, Madison, WI) or chymotrypsin (Roche, Indianapolis, IN) using

the Montage In-Gel Digestion Kit (Millipore) following the manufacturer’s

recommended protocols. Capillary-liquid chromatography-nanospray tandem mass

spectrometry (Nano-LC/MS/MS) was performed on a Thermo Finnigan LTQ mass

spectrometer equipped with a nanospray source operated in positive ion mode. The LC

system was an UltiMate™ Plus system (LC-Packings A Dionex Co., Sunnyvale, CA)

with a Famos autosampler and Switchos column switcher. Five µl of each digested sample was first injected on to the trapping column (LC-Packings A Dionex Co), and washed with 50 mM acetic acid. The injector port was switched to inject, and the peptides were eluted off of the trap onto the column. A 5 cm × 75 µm ID ProteoPep II

C18 column (New Objective, Inc. Woburn, MA) packed directly in the nanospray tip was

used for chromatographic separations. Peptides were eluted directly off the column into

121 the LTQ system using a gradient of 2-80% acetonitrile over 50 min, with a flow rate of

300 nl/min. The scan sequence of the mass spectrometer was programmed for a full scan, a zoom scan to determine the charge of the peptide and a MS/MS scan of the most abundant peak in the spectrum. Sequence information from the MS/MS data was processed using Mascot Distiller to form a peaklist (.mgf file) and using the MASCOT

MS/MS search engine and Turbo SEQUEST algorithm in BioWorks 3.1 Software.

2-D structure prediction and surface localization of OMP-1A and OMP-1N by immunofluorescence microscopy

Transmembrane β strands of all OMP-1/P28 family proteins and their topology with respect to the outer membrane lipid bilayer were predicted using the Posterior

Decoding method at the web server of PRED-TMBB

(http://bioinformatics.biol.uoa.gr/PRED-TMBB) (Bagos et al., 2004).

Two relatively highly antigenic and hydrophilic peptide fragments were chosen from OMP-1A and OMP-1N amino acid sequences based on Protean analysis in

DNASTAR software (DNASTAR Inc., Madison, WI). The 15-mer peptide,

CDLKDGFFEPKAEDT, underlined sequence corresponding to 14 amino acids of OMP-

1A (154-167aa) and the 15-mer peptide, CEISGGSNNPANNKY, underlined sequence corresponding to OMP-1N (142-155aa), were synthesized and conjugated to keyhole limpet hemocyanin (KLH), and rabbit antibodies were developed by Sigma Genosys (St.

Louis, MO). According to BLAST search results for short, nearly exact matched sequences in the NCBI non-redundant database, each of these two peptide sequences is unique and with little to no homology to any other known proteins (E>10).

122 For immunofluorescence microscopic analysis of the surface localization of E. chaffeensis OMP-1A and OMP-1N, organisms were fixed by the paraformaldehyde method (Wang et al., 2006)(Ge and Rikihisa, 2006). Briefly, host cell-free bacteria were pelleted and washed in PBS. All subsequent steps were performed at room temperature.

Bacteria were fixed in 2% paraformaldehyde for 45 min. After quenching in PBS (137 mM NaCl, 2.68 mM KCl, 10 mM Na2HPO4 and 1.76 mM KH2PO4, pH7.4) containing

0.1 M glycine and subsequent washing in PBS, bacteria were incubated with a 1:100 dilution of rabbit antiserum against OMP-1A, rabbit antiserum against OMP-1N or rabbit pre-immune serum in PG buffer (0.2% gelatin in PBS) for 1 h. After being washed in PG, the bacteria were labeled with Alexa Fluor 488 goat anti-rabbit IgG (Invitrogen) at a dilution of 1:100 in PG for 1 h. The bacteria were washed in PG, resuspended in PBS and observed using a Nikon Eclipse E400 fluorescence microscope with a xenon-mercury light source (Nikon Instruments, Melville, NY).

Western blotting analysis

The protein samples were subjected to 10% SDS-PAGE and transferred to polyvinylidine fluoride (PVDF) membranes (Westran S; Schleicher & Schuell

BioScience, Keene, NH). As a positive control, recombinant E. chaffeensis OMP-1B

(full-length without signal peptide cloned into pET29a between NdelI [5'] and XhoI [3']) was used. The membranes were incubated with primary antibodies at a dilution of 1:200 at 4°C overnight, which was followed by incubation with horseradish peroxidase (HRP)- conjugated goat anti-rabbit IgG (H+L), anti-dog IgG (γ) or anti-human IgA, IgM and IgG

(KPL, Gaithersburg, MA) at a 1:1,000 dilution at room temperature for 1 h. The primary antibodies used were rabbit antiserum against OMP-1B peptide (CEAPINGNTSITKKV,

123 OMP-1B sequence underlined), human ehrlichiosis patient sera (Rikihisa et al., 1994), sera from experimentally E. chaffeensis-infected specific-pathogen-free (SPF) dogs or experimentally Ehrlichia ewingii-infected SPF dogs. For the biotin blot, the membrane was incubated with 1:1,000 diluted HRP-conjugated streptavidin (Invitrogen) at room temperature for 1 h. The blot was then washed in TBST (15 mM NaCl, 5 mM Tris-HCl, pH 7.4, 0.02% Tween-20) four times for 10 min each and developed using the enhanced chemiluminescence (ECL) technique (Pierce).

5.4 Results

Detection of biotinylated proteins

Since the Sulfo-NHS-Biotin or Sulfo-NHS-SS-Biotin labeling reagent is charged by the sodium sulfoxide group on the succinimidyl ring, it is water soluble and membrane-impermeable, which makes the reagent suitable for labeling cell surface proteins via reaction with primary amines (Cole et al., 2005). Sulfo-NHS-SS-Biotin enables isolation of biotin-labeled cell surface proteins by streptavidin affinity chromatography via cleavage with a reducing agent (Scheurer et al., 2005). This feature, nontheless, prohibits the use of reducing agents in the SDS-PAGE sample buffer when observing sulfo-NHS-SS-Biotin-labeled proteins via biotin blotting, which makes molecular size pattern discrimination inaccurate. Therefore, we also labeled host cell- free E. chaffeensis with Sulfo-NHS-Biotin to determine the approximate molecular sizes of biotinylated proteins in the presence of reducing agent. The biotinylated proteins were blotted onto the membrane and probed with HRP-conjugated streptavidin. As shown in

Fig. 1 Lane 1, Sulfo-NHS-Biotin labeling revealed E. chaffeensis surface proteins

124 (approximate molecular mass of 22, 37, 40, 48, 55, 60, 74, 80, 90, 120 and 200 kDa). To

preserve the disulfide bonds in Sulfo-NHS-SS-Biotin, there was no reducing agent added

to the sample buffer of Sulfo-NHS-SS-Biotin-labeled proteins (Fig. 1). Similar surface

labeled proteins were detected with Sulfo-NHS-SS-Biotin as shown in Lane 2 (Fig. 1).

Streptavidin affinity purification of biotinylated E. chaffeensis surface proteins

To identify biotinylated E. chaffeensis surface proteins, Sulfo-NHS-SS-Biotin- labeled host cell-free bacteria were solubilized in RIPA buffer. The biotinylated proteins

were purified by streptavidin affinity gel chromatography. The disulfide bond in Sulfo-

NHS-SS-Biotin was cleaved with the reducing agent to elute the streptavidin affinity-

captured proteins. The captured proteins were seperated by SDS-PAGE. As shown in

Fig. 2, with GelCode blue protein staining in the SDS-PAGE (with the reducing agent in

the sample buffer), there are five bands of relatively abundant proteins corresponding to

molecular masses of approximately 28, 40, 50, 60 and 80 kDa. The labeling efficiency

was slightly different between Sulfo-NHS-Biotin and Sulfo-NHS-SS-Biotin due to spacer

arm length (Braschi and Wilson, 2006), and the detection sensitivity was also different

due to employment of different methods to detect biotin (Fig. 1) and protein (Fig. 2).

Nonetheless, the overall molecular sizes of biotinylated proteins with Sulfo-NHS-Biotin

(Fig. 1, lane 1) approximately matched with affinity-purified Sulfo-NHS-SS-Biotin-

labeled proteins after cleavage of S-S bond (Fig. 2).

Nano-LC/MS/MS

The five bands of streptavidin affinity-captured proteins separated by SDS-PAGE

were subjected to proteomic analysis. Band 1 consisted mainly of the OMP-1/P28

protein family. With the exception of OMP-1T (P28-4), OMP-1B (P28-14) and P28-2

125 (P28-21), 19 out of 22 E. chaffeensis OMP-1/P28 family proteins were detected (Table 1 and Fig. 3). Among these 19 OMP-1/P28 family proteins, except P28 (YP_507927) and

P28-1 (YP_507928) (Ohashi et al., 1998)(Singu et al., 2006), the remaining 17 OMP-

1/P28 proteins have not been directly proven to be expressed at the protein level. Except for P28, surface-exposure of other OMP-1/P28 proteins has not been shown previously.

Furthermore, the transmembrane β strands of all OMP-1/P28 proteins and their topology with respect to the outer membrane lipid bilayer have not been reported. We examined the 2-D structure of these 22 proteins predicted by the posterior decoding method using the dynamic programming algorithm in PRED-TMBB (Bagos et al., 2004). PRED-

TMBB is a web server capable of predicting the transmembrane strands and topology of

β-barrel outer membrane proteins of Gram-negative bacteria, which is based on a Hidden

Markov Model. It is trained using non-homologous outer membrane proteins with structures known at atomic resolution according to the Conditional Maximum Likelihood criterion (Bagos et al., 2004). All OMP-1/P28 proteins have a predicted 2-D structure of amphipathic and antiparallel multi-pass transmembrane β-strands (the structures of OMP-

1A and OMP-1N are shown in Fig. 4A and B).

The conjugal transfer protein VirB9-1 (YP_506875), a type IV secretion system subunit, was detected from band 1. The VirB9-1 ortholog (YP_505897) has been shown as surface-exposed on A. phagocytophilum (Niu et al., 2006), but previously was not observed on the surface of E. chaffeensis.

A protein from the serine protease DO/DeqQ family (YP_507837) was detected from band 3. This is the first time the surface-exposure of this family of proteins was shown on E. chaffeensis. The identification of the immunodominant surface protein gp47

126 (AAZ40202) is in agreement with the recent proposal of this protein as surface-exposed

based on immunoelectron microscopy (Doyle et al., 2006). This protein migrated as

band 3, with a molecular mass of approximately 50 kDa (Fig. 2), which was much

different from its predicted molecular mass of 33,040 Da. The major protein in band 4

was the 60 kDa chaperonin (GroEL) (YP_507185). Band 5 contained OMP85

(YP_507856), which is conserved in Gram-negative bacteria (Voulhoux et al., 2003), but

has not been shown to be expressed or surface-exposed on E. chaffeensis. For the first time, the hypothetical protein ECH_0525 (YP_507340), here named Ehrlichia surface protein 73 kDa Esp73, was revealed to be surface-exposed. Esp73 is an ortholog of

Anaplasma phagocytophilum hypothetical proteins APH_0404 and APH_0405, newly

uncovered surface-exposed proteins (Yan Ge and Yasuko Rikihisa, unpublished). Band

2, corresponding to approximately 40 kDa, was host cell β actin (NP_001092) (not shown

in Table 1). Some E. chaffeensis proteins of relative low abundance also were identified

in this study, such as translation elongation factor G (EF-G) (YP_507748), ATP synthesis

F1 β subunit (YP_507384), dihydrodipicolinate reductase (YP_507114), disulfide oxidoreductase (YP_507114) and the antioxidant ApC/Tsa family (YP_507536).

Surface localization of OMP-1A and OMP-1N

To verify the localization of OMP-1A and OMP-1N on the surface of E.

chaffeensis, rabbit polyclonal antibodies against immunogenic 15-mer peptides specific

to OMP-1A and OMP-1N, respectively, but not cross-reactive to any other proteins of E.

chaffeensis (by BLAST search of GenBank data base) were developed. Host cell-free E.

chaffeensis was prefixed with paraformaldehyde to prevent antibody permeabilization,

incubated with either anti-OMP-1A or OMP-1N peptide sera and examined by

127 immunofluorescence microscopy. As shown in Fig. 4C, the OMP-1A peptide antiserum

labeled the surface of individual bacteria in the typical ring-like surface staining pattern

as observed with A. phagocytophilum surface protein P44 (Wang et al., 2006). The

OMP-1N peptide antiserum also labeled the surface of individual E. chaffeensis bacteria

(Fig. 4C). Pre-immune rabbit serum did not label E. chaffeensis (Fig. 4C). The data confirmed the surface localization of OMP-1A and OMP-1N as identified by surface biotinylation method in this paper (Table 1). In the predicted 2-D structure, the epitope for anti-OMP-1A (154-167aa) (Fig. 4A) or anti-OMP-1N (142-155aa) antibodies was located within one of the extracellular loops (Fig. 4B). Thus, immunofluorescence labeling results for OMP-1A and OMP-1N using rabbit antiserum against each peptide experimentally validated the prediction.

Expression of OMP-1B by E. chaffeensis cultured in THP-1 cells

Among the E. chaffeensis OMP-1/P28 family of proteins, OMP-1B was not detected by surface biotinylation (Fig. 3). To determine whether OMP-1B protein was expressed by E. chaffeensis cultured in THP-1 cells, Western blot analysis of OMP-1B using rabbit anti-OMP-1B peptide serum was performed. In Lane 1 (Fig. 5), there was no detectable OMP-1B band around 28 kDa in E. chaffeensis-infected THP-1 cells. The pET29a vector and recombinant pET29a-omp-1B, which were expressed in BL21(DE3) competent cells, served as negative and positive controls (lanes 2 and 3), respectively.

This result suggests that OMP-1B is not expressed by E. chaffeensis when cultured in

THP-1 cells at 37 ºC.

128 Immunogenicity of E. chaffeensis surface-proteins

As shown in Fig. 6, although there was some variation in banding pattern across human and dog sera, the human ehrlichiosis patient serum (Lane 1) provided by the

Centers for Disease Control and Prevention (CDC, Atlanta, GA) (Rikihisa et al., 1994),

Dog # 3918815 anti-E. chaffeensis Arkansas serum (Lane 2) and Dog #1425 anti-E. chaffeensis St. Vincent serum (Paddock et al., 1997) (Lane 3) all strongly reacted with E. chaffeensis surface proteins with a molecular mass of approximately 28 kDa, corresponding to band 1 (Fig. 2), mainly from the OMP-1/P28 family (Table 1). Besides the 28 kDa proteins, the human ehrlichiosis patient serum (Lane 1) weakly reacted with the 60 kDa proteins, corresponding to band 4 (Fig. 2), containing mainly chaperonin

GroEL (Table 1). Dog # 3918815 anti-E. chaffeensis Arkansas serum (Lane 2) still reacted with 75 kDa proteins corresponding to band 5 (Fig. 2), containing mainly Esp73 and translation elongation factor G (Table 1). Dog #1425 anti-E. chaffeensis St. Vincent serum reacted strongly with approximately 55 kDa proteins, corresponding to band 3, containing mainly serine protease and aminopeptidase (Table 1), as well as 120 kDa, 85 kDa and 60 kDa proteins. In contrast, the two human ehrlichiosis patient sera, OK1

(Lane 4) and OK2 (Lane 5) from the Department of Health, Oklahoma City, OK

(Rikihisa et al., 1994), did not recognize 28 kDa OMP-1/P28 family proteins instead they mainly reacted with the 75, 60 or 47 kDa proteins (Fig. 6), which is consistent with the previous result using whole E. chaffeensis as antigens (purified from infected DH82 cells) (Rikihisa et al., 1994). Similarly, the sera from experimentally Ehrlichia ewingii- infected SPF dog # 2119 (Lane 6) and dog # 2185 (Lane 7) (Qingming Xiong, Weichao

Bao and Yasuko Rikihisa, unpublished) cross-reacted with 75 and 47 kDa proteins and

129 had weak (Lane 6) or undetectable (Lane 7) cross-reaction with 28 kDa proteins. Sera from normal dog or human from HME-non-endemic region (Japan) did not react with E. chaffeensis proteins (not shown in Fig. 6).

5.5 Discussion

Using a surface biotinylation method, the present study uncovered many E.

chaffeensis proteins that had not been shown previously or predicted as surface-exposed

(Table 1). Detection of two known surface-exposed proteins, P28 (P28-19) and gp47 by

surface biotinylation attests to the utility of this method. OMP-1/P28 family proteins are

the most abundantly expressed E. chaffeensis surface proteins. For the first time, 18 out

of 19 E. chaffeensis OMP-1/P28 family proteins have been shown to be surface-exposed,

and 17 have been detected directly at the protein level.

OMP-1/P28 outer membrane protein family members are the most studied E.

chaffeensis outer membrane proteins. They are encoded by a pleomorphic multigene

family composed of 22 paralogous genes clustered in a 27-kb gene locus of the E.

chaffeensis genome (Ohashi et al., 2001). The gene organization and the genomic locus

of E. chaffeensis omp-1/p28 gene cluster are conserved among Ehrlichia species, such as

E. canis (Ohashi et al., 2001) and E. ruminantium (Bekker et al., 2005). The mRNA

expression of omp-1/p28 family genes has been reported by several research groups. For

example, Reddy et al. reported that p28 (p28-19) is transcriptionally expressed among

130 five tested E. chaffeensis omp-1 paralogs (cultured in canine macrophage DH82 cell line)

(Reddy et al., 1998). Yu et al. showed that six of ten tested p28 (omp-1) genes were

actively transcribed in E. chaffeensis growing in DH82 cells (Yu et al., 2000). Long et al.

reported that 16 of the 22 omp-1/p28 p28 (omp-1) multigene family members were

transcribed by E. chaffeensis in DH82 cells (Long et al., 2002). Ohashi et al. have shown

that all 22 paralogs in the E. canis omp/p30s cluster were transcriptionally active in

DH82 cells (Ohashi et al., 2001). Unver et al. detected the transcripts of 16 of 22 E.

chaffeensis omp-1/p28 paralogs in the blood monocytes of infected dogs (Unver et al.,

2002).

In contrast, few studies have been undertaken concerning the protein expression

of OMP-1/P28 family members. Using enzyme-linked immunosorbent assay (ELISA)

with synthetic peptides of P28 OMPs (OMP-1/P28 paralogs) as antigens, Zhang et al. has

reported that all 22 P28 OMPs are expressed in infected dogs (Zhang et al., 2004). Using

N-terminal amino acid sequencing, Ohashi et al. detected P28 (P28-19) expression by E.

chaffeensis in DH82 cells (Ohashi et al., 1998). Recently, by proteomic analysis, two

OMP-1/P28 family members, P28-Omp19 (P28) and P28-Omp20 (P28-1), were

identified in E. chaffeensis cultured in DH82 cells (Singu et al., 2006). In the present

paper, 19 E. chaffeensis OMP-1/P28 proteins (cultured in THP-1 cells) including P28 and

P28-1 were directly identified by proteomics. These 19 OMP-1/P28 family proteins co-

exist in E. chaffeensis cultured in THP-1 cells. It is possible that each single organism

can express all 19 of these proteins, since at least two thirds of E. canis omp/p30s are co-

transcribed in DH82 cells (Ohashi et al., 2001). Alternatively, each individual organism expresses a few of them but in the given bacterial population, all 19 OMP-1/P28 family

131 proteins are expressed. Western blot data of OMP-1B (P28-14) suggests that it is not

expressed by E. chaffeensis cultured in THP-1 cells at 37ºC. Of note, OMP-1B is the only omp-1/p28 paralog transcribed by E. chaffeensis in three developmental stages of

Amblyomma americanum ticks before or after E. chaffeensis transmission to naïve dogs

(Unver, A. 2002). P28-Omp14 (OMP-1B) protein was also the only OMP-1/P28 paralog detected by proteomics in E. chaffeensis cultured in a tick cell line, ISE6 (Singu et al.,

2006). In the present paper, besides OMP-1B (P28-14), OMP-1T (P28-4) and P28-2

(P28-21) were the other two OMP-1/P28 family members not detected by surface biotinylation, suggesting that these two proteins are either not expressed at a detectable level or not surface-exposed by E. chaffeensis cultured in THP-1 cells at 37ºC. Therefore, the current data favor the notion that the E. chaffeensis OMP-1/P28 multigene family is differentially expressed in mammalian and tick hosts, which may be important for ehrlichial adaptation to different host environments (Singu et al., 2006; Unver et al.,

2002). Recently, Zhang et al. proposed that the expression of P28 (P28-19) was up-

regulated in E. chaffeensis reticulate cells and down-regulated in dense-cored cells

(Zhang et al., 2006). In future studies, it would be of interest to determine the global

pattern of surface-exposed proteins during the developmental cycle of E. chaffeensis and

the mechanism and functions of the differential expression by E. chaffeensis cultured in

different host cells or in different ehrlichial developmental stages.

The isolated outer membrane fraction of E. chaffeensis exhibits typical porin

activity (Haibin Huang and Yasuko Rikihisa, unpublished data). The OMP-1/P28 family

may be responsible for this activity since all of the proteins in this family have a

predicted 2-D structure of multi-pass β strands with some porin features (Nikaido, 2003)

132 such as amphipathic and antiparallel β-strands, an abundance of polar residues on the

external surface, a C-terminal phenylalanine and transmembrane strands connected by

short turns in the periplasmic side and long loops on the external side. Among 22 OMP-

1/P28 family, 12 (OMP-1A, P, V, W, X, Y, Z, B, C, E, F, and P28) have phenylalanine at

the C-terminus. The porin function of OMP-1/P28 family has not been confirmed.

Esp73 is an ortholog of the newly identified A. phagocytophilum outer membrane

proteins APH_0404 and APH_0405, which have predicted secondary structures of 22

transmembrane β strands and are very likely to function as transporters (Yan Ge and

Yasuko Rikihisa, unpublished). Except for the genus Neorickettsia, the orthologs have been found in genera of Anaplasma, Ehrlichia and Wolbachia in family

Anaplasmataceae by BLAST search of the GenBank database with an E value below e-15,

suggesting that they play an important role in the survival of these bacteria (Yan Ge and

Yasuko Rikihisa, unpublished). The theoretical molecular mass of the immunodominant

surface protein gp47 is different from the migration distance in 10% SDS-PAGE gel

(Table 1 and Fig. 2). It has been proposed that this may be caused by glycosylation

modification although there is no direct evidence that in fact, this protein is glycosylated

(Doyle et al., 2006). gp47 does not have signal peptide (predicted by SignalP 3.0).

Therefore, how this protein is secreted to the bacterial surface remains to be investigated.

The E. chaffeensis serine protease (YP_507837) with a definitive N-terminal signal

peptide sequence predicted by SignalP 3.0 was identified as a surface-exposed protein in

the present study. There are reports of serine proteases existing on the surface of other

bacteria. For example, serine protease HtrA (high temperature requirement) has been

described as a surface-exposed protease on Staphylococcus aureus (Rigoulay et al., 2005).

133 This protein is involved in the virulence of many pathogens via its role in thermal

stability, resistance to oxidative stress and bacterial survival (Cortes et al., 2002). In the

future, it would be interesting to characterize the function of this protein.

Some proteins, which were considered as bacterial cytoplasmic, periplasmic, or

inner membrane proteins, were identified by the present study, for example, chaperone

protein DnaK, translation elongation factor G, GroEL, adenosine triphosphate (ATP)

synthase F1, cytosol aminopeptidase and dihydrodipicolinate reductase. These proteins

may have been released from spontaneously lysed bacteria and bound to the intact

bacteria within the host inclusions. Another possibility is that some bacterial outer

membrane may be damaged during isolating cell-free bacteria before biotin-labeling.

However, the surface localization and some surface-related function of these proteins

have been reported in other bacteria. For example, by immunofluorescence or

immunoelectron microscopy, GroEL has been shown on the surface of Helicobacter

pylori (Huesca et al., 1996), Legionella pneumophila (Garduno et al., 1998),

Haemophilus ducreyi (Frisk et al., 1998) and Clostridium difficile (Hennequin et al.,

2001). DnaK was detected on the surface of Helicobacter pylori (Huesca et al., 1996), and translation elongation factor u (EF-Tu) was detected on the surface of Mycoplasma pneumoniae (Dallo et al., 2002). Evidence for specific secretion rather than autolysis has been reported in the release of some H. pylori proteins including GroEL (Vanet and

Labigne, 1998). H. ducreyi (Pantzar et al., 2006) GroEL has been reported to mediate the binding of this bacterium to host carbohydrate receptors. EF-Tu and cytosol aminopeptidase have been found in the outer membrane fraction of A. marginale (Lopez et al., 2005). Recently, ATP synthase alpha and beta have been identified from Brucella

134 abortus cell envelope using 2-D electrophoresis with MALDI-TOF MS and LC-MS/MS

(Connolly et al., 2006). Therefore, we can not deny the possibility that these proteins

with well-known functions in cytoplasmic, periplasmic or inner membrane are present on

the surface of E. chaffeensis and play unexpected roles in E. chaffeensis-host interaction.

One obvious host protein band captured by streptavidin affinity purification was

β-actin, which is one of the most abundant cytoskeleton proteins of eukaryotic cells. This

may be due to the binding of host cell actin to bacterial surface proteins during the

isolation of host cell-free bacteria or via a functional association. Actin association with

the isolated obligate intracellular bacterium, rickettsia, was previously reported to confer

intracellular motility to rickettsia (Martinez et al., 2005).

Since surface proteins of E. chaffeensis are exposed to the host immune system,

they may represent important antigens. P28 has been shown to be immunoreactive with

HME human patient serum (Ohashi et al., 1998); gp47 (Doyle et al., 2006) and gp120

(Popov et al., 2000) being immunogenic in E. chaffeensis-infected dogs. The Western

blot banding pattern of E. chaffeensis infections in dogs and humans has revealed several

E. chaffeensis immunoreactive proteins, e.g., 74, 70, 64, 47, 31 and 29 kDa (Rikihisa et al., 1994). In the present study, the four major E. chaffeensis surface protein bands with molecular mass of approximately 75, 60, 55 and 28 kDa have been shown either to be immunoreactive with E. chaffeensis-infected human patient sera or with E. chaffeensis- infected dog sera, or to cross-react with E. ewingii-infected dog sera or with Ehrlichia spp.-infected human sera. Therefore, the newly identified surface-exposed proteins have advanced the knowledge of E. chaffeensis immunogens, which extends the range of candidates for diagnostic antigens and vaccine development. Consistent with a previous

135 study (Rikihisa et al., 1994), human ehrlichiosis patient sera from Oklahoma recognized high molecular mass proteins instead of 28 kDa OMP-1/P28 family proteins, which is similar to dog anti-E. ewingii sera or human E. ewingii patient sera (Buller et al., 1999).

Therefore, the case of E. ewingii-infection of humans may have existed in Oklahoma in

1994, earlier than the first report in Missouri in 1999 (Buller et al., 1999). E. chaffeensis- infected and E. canis-infected dog sera cross-reacted with E. canis OMP/P30 proteins and

E. chaffeensis OMP-1/P28 proteins, respectively (Rikihisa et al., 1994). On the contrary, dog anti-E. ewingii sera poorly recognized E. chaffeensis OMP-1/P28 or E. canis

OMP/P30 proteins (Rikihisa et al., 1994). Hence, the major outer membrane protein E. chaffeensis OMP-1/P28 and E. canis OMP/P30 proteins can serve as diagnostic antigens to distinguish E. chaffeensis and E. canis from E. ewingii, which has not been culture- isolated yet. Dog anti-E. chaffeensis sera recognized the 120-kDa protein band, which may correspond to gp120 (Popov et al., 2000). However, this protein is not abundant enough in biotinylation-purified E. chaffeensis surface proteins to allow proteomic identification in the present study.

In conclusion, the identification of E. chaffeensis major surface proteins purified by the biotinylation method will greatly advance the knowledge of E. chaffeensis surface components and set the foundation for future studies on the ehrlichia-host interaction.

The antigenic analysis of E. chaffeensis major surface proteins provides a firm rationale for development of diagnostics and vaccines. Future studies will be aimed at characterizing the biological functions of the newly identified E. chaffeensis surface proteins.

136

Annotation or Predicted Nano- GenBank Subcellular Predicted Protein Signal Gene locus tag molecular LC/MS/MS accession localization function peptide mass sequence number sequence coverage by (Yes/No ammino acid position) (%) outer membrane OuterMembrane Outer membrane protein, OMP85 family 87546 11 YP_507856 protein assembly Yes (18) hypothetical protein OuterMembrane ECH_0525 (Esp73) 74244 12 YP_507340 Unknown Yes (23) chaperone protein Unknown DnaK 69080 45 YP_507287 Chaperone No translation elongation Cytoplasmic factor G 76707 10 YP_507748 Protein synthesis No chaperonin, 60 kd Cytoplasmic (GroEL) 58850 76 YP_507185 Chaperonin No ATP synthase F1, beta Unknown subunit 54933 18 YP_507384 Energy production No serine protease, Unknown Protein folding, DO/DeqQ family 50740 51 YP_507837 degradation Yes (22) cytosol aminopeptidase 54708 39 YP_507189 Unknown Metabolism No dehydrogenase, Cytoplasmic isocitrate/isopropylmala Amino acid te family 56858 9 YP_507898 biosynthesis No immunodominant Unknown surface protein gp47 33040 12 AAZ40202 Unknown No major outer membrane Unknown protein P28 30552 39 YP_507927 Unknown Yes (25) major outer membrane CytoplasmicMe protein P28-1 30145 59 YP_507928 mbrane Yes (23) major outer membrane CytoplasmicMe protein OMP-1N 30671 24 YP_507905 mbrane Yes (25) major outer membrane CytoplasmicMe protein OMP-1P 32744 18 YP_507908 mbrane No major outer membrane Unknown protein OMP-1E 30647 48 YP_507924 Yes (25) major outer membrane CytoplasmicMe protein OMP-1S 32045 30 YP_507916 mbrane Yes (25)

Continued

Table 5.1 Surface-exposed proteins of E. chaffeensis affinity-purified by biotinylation method and identified by Nano-LC/MS/MS. Subcellular localization of proteins was analyzed by PSORTb version 2.0.4 (http://www.psort.org/psortb) (J.L. Gardy, 2005,

Bioinformatics, 21(5):617). The program SignalP 3.0

(http://www.cbs.dtu.dk/services/SignalP) was used to predict the presence of N-terminal signal peptides (Bendtsen JD, J. Mol. Biol., 340:783-795, 2004). Grey highlight marks the boundary between protein bands of different molecular sizes. 137

Table 5.1 continued major outer membrane CytoplasmicMe protein OMP-1V 30921 16 YP_507911 mbrane Yes (26) major outer membrane CytoplasmicMe protein OMP-1M 33293 6 YP_507903 mbrane Yes (38) major outer membrane Unknown protein OMP-1A 33113 32 YP_507919 Yes (30) major outer membrane Unknown protein OMP-1H 33177 41 YP_507917 Yes (26) major outer membrane CytoplasmicMe protein OMP-1Q 33584 34 YP_507907 mbrane Yes (22) major outer membrane CytoplasmicMe protein OMP-1U 33766 20 YP_507910 mbrane Yes (28) major outer membrane CytoplasmicMe protein OMP-1W 31870 13 YP_507913 mbrane Yes (27) major outer membrane Unknown protein OMP-1X 30628 17 YP_507914 No major outer membrane Unknown protein OMP-1Y 31806 25 YP_507915 Yes (24) major outer membrane OuterMembrane protein OMP-1Z 33252 19 YP_507918 Yes (26) Unknown Transportation of macromolecules type IV secretion across bacterial inner system protein VirB9-1 31690 16 YP_506875 and outer membranes Yes (24) Unknown Biosynthesis of L- dihydrodipicolinate lysine and bacterial reductase 29614 48 YP_507261 cell walls No disulfide Unknown Protein folding and oxidoreductase 27606 39 YP_507114 associated processing Yes (21) antioxidant, AhpC/Tsa Cytoplasmic Reduction of family 23575 28 YP_507536 peroxides No

138

Fig. 5.1 Biotin labeling of E. chaffeensis surface proteins. Biotinylated E. chaffeensis was lysed in RIPA buffer, separated by 10% SDS-PAGE and blotted with HRP- conjugated streptavidin. Lane 1, Sulfo-NHS-Biotin-labeled E. chaffeensis. Lane 2,

Sulfo-NHS-SS-Biotin-labeled E. chaffeensis (in the absence of reducing agent in SDS-

PAGE sample buffer). Marker, Precision Plus prestained protein standards (Bio-Rad).

139

Fig. 5.2 Streptavidin agarose gel affinity purification of Sulfo-NHS-SS-Biotin- labeled E. chaffeensis surface proteins. Sulfo-NHS-SS-Biotin-labeled E. chaffeensis surface proteins were separated by 10% SDS-PAGE and stained with GelCode blue.

Bands 1-5 were subjected to Nano-LC/MS/MS analysis. Marker, Precision Plus prestained protein standards (Bio-Rad).

140

Fig. 5.3 Diagram of protein expression of E. chaffeensis OMP-1/P28 multigene family identified by Nano-LC/MS/MS. The grey box indicates that the protein is surface-exposed. The open box indicates that the protein was undetectable. OMP-1/P28 family members were designated as OMP-1M, N, Q, P, T, U, V, W, X, Y, S, H, Z, A, B,

C, D, E, F, P28, P28-1 and P28-2 in Ohashi et al.’s study {Ohashi, 1998 #20}{Ohashi,

2001 #12}. Another set of names for OMP-1/P28 family members came from Yu et al.’s report {Yu, 2000 #35} i.e., P28-1’, P28-1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,

17, 18, 19, 20 and 21.

141 Fig. 5.4 2-D structure prediction of E. chaffeensis OMP-1A (A) and OMP-1N (B)

with respect to the outer membrane lipid bilayer using Posterior Decoding method with the dynamic programming algorithm in PRED-TMBB. Multi-pass transmembrane β strands were predicted in both OMP-1A and OMP-1N. The discrimination scores of OMP-1A and OMP-1N were 2.935 and 2.947, respectively.

Scores lower than the threshold value of 2.965 is considered as significant prediction of

β-barrel outer membrane proteins (Bagos, Liakopoulos et al. 2004).

C. Surface localization of E. chaffeensis OMP-1A and OMP-1N by imunofluorescence assay. Host cell-free E. chaffeensis were fixed in paraformaldehyde, incubated with rabbit sera against OMP-1A peptide (154-167aa) or OMP-1N peptide

(142-155aa) stained with Alexa Fluor 488 goat anti-rabbit IgG and visualized by fluorescence microscopy. (a) Ring-like bacterial surface staining of OMP-1A. (b) Ring- like bacterial surface staining of OMP-1N. (c) Representative of ehrlichia surface-stained with rabbit pre-immune serum. Scale bar, 1 µm.

142 Fig. 5.4 A

B

C C

143

Fig. 5.5 Western blotting analysis of OMP-1B expression by E. chaffeensis cultured in THP-1 cells using anti-OMP-1B peptide rabbit serum. The recombinant pET29a- omp-1B or pET29a vector was expressed in E.coli BL21(DE3) competent cells

(Novagen). Lane 1, E. chaffeensis cultured in THP-1 cells, no band detected around 28 kDa. Lane 2, negative control, pET29a vector expressed in E.coli BL21(DE3). Lane 3, positive control, recombinant pET29a-omp-1B expressed in E.coli BL21(DE3). Marker,

Precision Plus prestained Protein standards (Bio-Rad).

144

Fig. 5.6 Western blotting analysis of antigenicity of E. chaffeensis surface proteins purified by biotinylation method. E. chaffeensis surface proteins affinity-purified by biotinylation method were separated by 10% SDS-PAGE, transferred to PVDF membrane and immunoblotted with human patient sera or experimentally-infected dog sera. Lane 1, Human ehrlichiosis patient serum (CDC). Lane 2, Dog # 3918815 anti-E. chaffeensis Arkansas. Lane 3, Dog #1425 anti-E. chaffeensis St. Vincent serum. Lanes 4 and 5, Human ehrlichiosis patient sera (OK1 and OK2, respectively). Lanes 6 and 7, sera from experimentally E. ewingii-infected SPF dog # 2185 and # 2119, respectively.

Marker, Precision Plus prestained protein standards (Bio-Rad).

145

BIBLIOGRAPHY

Abbott, J.R., Palmer, G.H., Kegerreis, K.A., Hetrick, P.F., Howard, C.J., Hope, J.C. and Brown, W.C. (2005) Rapid and long-term disappearance of CD4+ T lymphocyte responses specific for Anaplasma marginale major surface protein-2 (MSP2) in MSP2 vaccinates following challenge with live A. marginale. J Immunol. 174: 6702-6715.

Akgul, C., Moulding, D.A. and Edwards, S.W. (2001) Molecular control of neutrophil apoptosis. FEBS Lett. 487: 318-322.

Alberti, A., Addis, M.F., Sparagano, O., Zobba, R., Chessa, B., Cubeddu, T., et al (2005) Anaplasma phagocytophilum, Sardinia, Italy. Emerg Infect Dis. 11: 1322-1324.

Andersson, S.G., Zomorodipour, A., Andersson, J.O., Sicheritz-Ponten, T., Alsmark, U.C., Podowski, R.M., et al (1998) The genome sequence of Rickettsia prowazekii and the origin of mitochondria. Nature. 396: 133-140.

Antonsson, B. (2004) Mitochondria and the Bcl-2 family proteins in apoptosis signaling pathways. Mol Cell Biochem. 256-257: 141-155.

Bagos, P.G., Liakopoulos, T.D., Spyropoulos, I.C. and Hamodrakas, S.J. (2004) PRED- TMBB: a web server for predicting the topology of beta-barrel outer membrane proteins. Nucleic Acids Res. 32: W400-404.

Bakken, J.S. and Dumler, J.S. (2000) Human granulocytic ehrlichiosis. Clin Infect Dis. 31: 554-560.

Bakken, J.S., Dumler, J.S., Chen, S.M., Eckman, M.R., Van Etta, L.L. and Walker, D.H. (1994) Human granulocytic ehrlichiosis in the upper Midwest United States. A new species emerging? Jama. 272: 212-218.

Barnewall, R.E. and Rikihisa, Y. (1994) Abrogation of gamma interferon-induced inhibition of Ehrlichia chaffeensis infection in human monocytes with iron-transferrin. Infect Immun. 62: 4804-4810.

Barnewall, R.E., Ohashi, N. and Rikihisa, Y. (1999) Ehrlichia chaffeensis and E. sennetsu, but not the human granulocytic ehrlichiosis agent, colocalize with transferrin receptor and up-regulate transferrin receptor mRNA by activating iron-responsive protein 1. Infect Immun. 67: 2258-2265. 146 Bitsaktsis, C., Huntington, J. and Winslow, G. (2004) Production of IFN-gamma by CD4 T cells is essential for resolving ehrlichia infection. J Immunol. 172: 6894-6901.

Borjesson, D.L., Kobayashi, S.D., Whitney, A.R., Voyich, J.M., Argue, C.M. and Deleo, F.R. (2005) Insights into pathogen immune evasion mechanisms: Anaplasma phagocytophilum fails to induce an apoptosis differentiation program in human neutrophils. J Immunol. 174: 6364-6372.

Borner, C. (2003) The Bcl-2 protein family: sensors and checkpoints for life-or-death decisions. Mol Immunol. 39: 615-647.

Braschi, S. and Wilson, R.A. (2006) Proteins exposed at the adult schistosome surface revealed by biotinylation. Mol Cell Proteomics. 5: 347-356.

Brown, W.C., Shkap, V., Zhu, D., McGuire, T.C., Tuo, W., McElwain, T.F. and Palmer, G.H. (1998) CD4(+) T-lymphocyte and immunoglobulin G2 responses in calves immunized with Anaplasma marginale outer membranes and protected against homologous challenge. Infect Immun. 66: 5406-5413.

Buchanan, S.K., Smith, B.S., Venkatramani, L., Xia, D., Esser, L., Palnitkar, M., et al (1999) Crystal structure of the outer membrane active transporter FepA from Escherichia coli. Nat Struct Biol. 6: 56-63.

Carlyon, J.A., Ryan, D., Archer, K. and Fikrig, E. (2005) Effects of Anaplasma phagocytophilum on host cell ferritin mRNA and protein levels. Infect Immun. 73: 7629- 7636.

Carlyon, J.A., Abdel-Latif, D., Pypaert, M., Lacy, P. and Fikrig, E. (2004) Anaplasma phagocytophilum utilizes multiple host evasion mechanisms to thwart NADPH oxidase- mediated killing during neutrophil infection. Infect Immun. 72: 4772-4783.

Carlyon, J.A., Akkoyunlu, M., Xia, L., Yago, T., Wang, T., Cummings, R.D., et al (2003) Murine neutrophils require alpha1,3-fucosylation but not PSGL-1 for productive infection with Anaplasma phagocytophilum. Blood. 102: 3387-3395.

Chen, S.M., Dumler, J.S., Bakken, J.S. and Walker, D.H. (1994) Identification of a granulocytotropic Ehrlichia species as the etiologic agent of human disease. J Clin Microbiol. 32: 589-595.

Cobessi, D., Celia, H. and Pattus, F. (2005a) Crystal structure at high resolution of ferric- pyochelin and its membrane receptor FptA from Pseudomonas aeruginosa. J Mol Biol. 352: 893-904.

147 Cobessi, D., Celia, H., Folschweiller, N., Schalk, I.J., Abdallah, M.A. and Pattus, F. (2005b) The crystal structure of the pyoverdine outer membrane receptor FpvA from Pseudomonas aeruginosa at 3.6 angstroms resolution. J Mol Biol. 347: 121-134.

Cole, J.N., Ramirez, R.D., Currie, B.J., Cordwell, S.J., Djordjevic, S.P. and Walker, M.J. (2005) Surface analyses and immune reactivities of major cell wall-associated proteins of group a streptococcus. Infect Immun. 73: 3137-3146.

Coppens, I., Dunn, J.D., Romano, J.D., Pypaert, M., Zhang, H., Boothroyd, J.C. and Joiner, K.A. (2006) Toxoplasma gondii sequesters lysosomes from mammalian hosts in the vacuolar space. Cell. 125: 261-274.

Cory, S. and Adams, J.M. (2002) The Bcl2 family: regulators of the cellular life-or-death switch. Nat Rev Cancer. 2: 647-656.

Daigle, I. and Simon, H.U. (2001) Critical role for caspases 3 and 8 in neutrophil but not eosinophil apoptosis. Int Arch Allergy Immunol. 126: 147-156.

Dallo, S.F., Kannan, T.R., Blaylock, M.W. and Baseman, J.B. (2002) Elongation factor Tu and E1 beta subunit of pyruvate dehydrogenase complex act as fibronectin binding proteins in Mycoplasma pneumoniae. Mol Microbiol. 46: 1041-1051.

Dawson, J.E., Anderson, B.E., Fishbein, D.B., Sanchez, J.L., Goldsmith, C.S., Wilson, K.H. and Duntley, C.W. (1991) Isolation and characterization of an Ehrlichia sp. from a patient diagnosed with human ehrlichiosis. J Clin Microbiol. 29: 2741-2745.

de la Fuente, J., Ayoubi, P., Blouin, E.F., Almazan, C., Naranjo, V. and Kocan, K.M. (2005) Gene expression profiling of human promyelocytic cells in response to infection with Anaplasma phagocytophilum. Cell Microbiol. 7: 549-559.

DeLeo, F.R. (2004) Modulation of phagocyte apoptosis by bacterial pathogens. Apoptosis. 9: 399-413.

Demma, L.J., Holman, R.C., McQuiston, J.H., Krebs, J.W. and Swerdlow, D.L. (2005) Epidemiology of human ehrlichiosis and anaplasmosis in the United States, 2001-2002. Am J Trop Med Hyg. 73: 400-409.

Donepudi, M., Mac Sweeney, A., Briand, C. and Grutter, M.G. (2003) Insights into the regulatory mechanism for caspase-8 activation. Mol Cell. 11: 543-549.

Doyle, C.K., Nethery, K.A., Popov, V.L. and McBride, J.W. (2006) Differentially expressed and secreted major immunoreactive protein orthologs of Ehrlichia canis and E. chaffeensis elicit early antibody responses to epitopes on glycosylated tandem repeats. Infect Immun. 74: 711-720.

148 Dumler, J.S., Asanovich, K.M., Bakken, J.S., Richter, P., Kimsey, R. and Madigan, J.E. (1995) Serologic cross-reactions among Ehrlichia equi, Ehrlichia phagocytophila, and human granulocytic Ehrlichia. J Clin Microbiol. 33: 1098-1103.

Dumler, J.S., Barbet, A.F., Bekker, C.P., Dasch, G.A., Palmer, G.H., Ray, S.C., et al (2001) Reorganization of genera in the families Rickettsiaceae and Anaplasmataceae in the order Rickettsiales: unification of some species of Ehrlichia with Anaplasma, Cowdria with Ehrlichia and Ehrlichia with Neorickettsia, descriptions of six new species combinations and designation of Ehrlichia equi and 'HGE agent' as subjective synonyms of Ehrlichia phagocytophila. Int J Syst Evol Microbiol. 51: 2145-2165.

Dumler, J.S., Choi, K.S., Garcia-Garcia, J.C., Barat, N.S., Scorpio, D.G., Garyu, J.W., et al (2005) Human granulocytic anaplasmosis and Anaplasma phagocytophilum. Emerg Infect Dis. 11: 1828-1834.

Edwards, S.W. (1994) Biochemistry and Pysiology of the Neutrophil.

Felek, S., Telford, S., 3rd, Falco, R.C. and Rikihisa, Y. (2004) Sequence analysis of p44 homologs expressed by Anaplasma phagocytophilum in infected ticks feeding on naive hosts and in mice infected by tick attachment. Infect Immun. 72: 659-666.

Ferguson, A.D., Chakraborty, R., Smith, B.S., Esser, L., van der Helm, D. and Deisenhofer, J. (2002) Structural basis of gating by the outer membrane transporter FecA. Science. 295: 1715-1719.

Frisk, A., Ison, C.A. and Lagergard, T. (1998) GroEL heat shock protein of Haemophilus ducreyi: association with cell surface and capacity to bind to eukaryotic cells. Infect Immun. 66: 1252-1257. Garduno, R.A., Faulkner, G., Trevors, M.A., Vats, N. and Hoffman, P.S. (1998) Immunolocalization of Hsp60 in Legionella pneumophila. J Bacteriol. 180: 505-513.

Ge, Y. and Rikihisa, Y. (2006) Anaplasma phagocytophilum delays spontaneous human neutrophil apoptosis by modulation of multiple apoptotic pathways. Cell Microbiol. 8: 1406-1416.

Ge, Y., Yoshiie, K., Kuribayashi, F., Lin, M. and Rikihisa, Y. (2005) Anaplasma phagocytophilum inhibits human neutrophil apoptosis via upregulation of bfl-1, maintenance of mitochondrial membrane potential and prevention of caspase 3 activation. Cell Microbiol. 7: 29-38.

Gentle, I.E., Burri, L. and Lithgow, T. (2005) Molecular architecture and function of the Omp85 family of proteins. Mol Microbiol. 58: 1216-1225.

149 Haibin Huang, X.W., Takane Kikuchi, Yumi Kumagai, and Yasuko Rikihisa (2006) Porin Activity of Anaplasma phagocytophilum Outer Membrane Fraction and Purified P44. Journal of Bacteriology.

Hennequin, C., Porcheray, F., Waligora-Dupriet, A., Collignon, A., Barc, M., Bourlioux, P. and Karjalainen, T. (2001) GroEL (Hsp60) of Clostridium difficile is involved in cell adherence. Microbiology. 147: 87-96. Herron, M.J., Nelson, C.M., Larson, J., Snapp, K.R., Kansas, G.S. and Goodman, J.L. (2000) Intracellular parasitism by the human granulocytic ehrlichiosis bacterium through the P-selectin ligand, PSGL-1. Science. 288: 1653-1656.

Hotopp, J.C., Lin, M., Madupu, R., Crabtree, J., Angiuoli, S.V., Eisen, J., et al (2006) Comparative genomics of emerging human ehrlichiosis agents. PLoS Genet. 2: e21.

Ismail, N., Soong, L., McBride, J.W., Valbuena, G., Olano, J.P., Feng, H.M. and Walker, D.H. (2004) Overproduction of TNF-alpha by CD8+ type 1 cells and down-regulation of IFN-gamma production by CD4+ Th1 cells contribute to toxic shock-like syndrome in an animal model of fatal monocytotropic ehrlichiosis. J Immunol. 172: 1786-1800.

Jiang, X., Payne, M.A., Cao, Z., Foster, S.B., Feix, J.B., Newton, S.M. and Klebba, P.E. (1997) Ligand-specific opening of a gated-porin channel in the outer membrane of living bacteria. Science. 276: 1261-1264.

Judd, P.K., Kumar, R.B. and Das, A. (2005) Spatial location and requirements for the assembly of the Agrobacterium tumefaciens type IV secretion apparatus. Proc Natl Acad Sci U S A. 102: 11498-11503.

JW, I.J. and Mueller, A.C. (2004) Neutrophil NADPH oxidase is reduced at the Anaplasma phagocytophilum phagosome. Infect Immun. 72: 5392-5401.

Kawahara, M., Rikihisa, Y., Lin, Q., Isogai, E., Tahara, K., Itagaki, A., et al (2006) Novel genetic variants of Anaplasma phagocytophilum, Anaplasma bovis, Anaplasma centrale, and a novel Ehrlichia sp. in wild deer and ticks on two major islands in Japan. Appl Environ Microbiol. 72: 1102-1109.

Kaylor, P.S., Crawford, T.B., McElwain, T.F. and Palmer, G.H. (1991) Passive transfer of antibody to Ehrlichia risticii protects mice from ehrlichiosis. Infect Immun. 59: 2058- 2062.

Kim, H.Y. and Rikihisa, Y. (1998) Characterization of monoclonal antibodies to the 44- kilodalton major outer membrane protein of the human granulocytic ehrlichiosis agent. J Clin Microbiol. 36: 3278-3284.

Kim, H.Y. and Rikihisa, Y. (2002) Roles of p38 mitogen-activated protein kinase, NF- kappaB, and protein kinase C in proinflammatory cytokine mRNA expression by human

150 peripheral blood leukocytes, monocytes, and neutrophils in response to Anaplasma phagocytophila. Infect Immun. 70: 4132-4141.

Kirchhausen, T. (2000) Clathrin. Annu Rev Biochem. 69: 699-727.

Kobayashi, S.D., Braughton, K.R., Whitney, A.R., Voyich, J.M., Schwan, T.G., Musser, J.M. and DeLeo, F.R. (2003) Bacterial pathogens modulate an apoptosis differentiation program in human neutrophils. Proc Natl Acad Sci U S A. 100: 10948-10953.

Kumar, R.B., Xie, Y.H. and Das, A. (2000) Subcellular localization of the Agrobacterium tumefaciens T-DNA transport pore proteins: VirB8 is essential for the assembly of the transport pore. Mol Microbiol. 36: 608-617.

Lafont, F. and van der Goot, F.G. (2005) Bacterial invasion via lipid rafts. Cell Microbiol. 7: 613-620.

Laskay, T., van Zandbergen, G. and Solbach, W. (2003) Neutrophil granulocytes--Trojan horses for Leishmania major and other intracellular microbes? Trends Microbiol. 11: 210-214.

Lee, E.H. and Rikihisa, Y. (1997) Anti-Ehrlichia chaffeensis antibody complexed with E. chaffeensis induces potent proinflammatory cytokine mRNA expression in human monocytes through sustained reduction of IkappaB-alpha and activation of NF-kappaB. Infect Immun. 65: 2890-2897.

Li, J.S., Chu, F., Reilly, A. and Winslow, G.M. (2002) Antibodies highly effective in SCID mice during infection by the intracellular bacterium Ehrlichia chaffeensis are of picomolar affinity and exhibit preferential epitope and isotype utilization. J Immunol. 169: 1419-1425.

Li, J.S., Yager, E., Reilly, M., Freeman, C., Reddy, G.R., Reilly, A.A., et al (2001) Outer membrane protein-specific monoclonal antibodies protect SCID mice from fatal infection by the obligate intracellular bacterial pathogen Ehrlichia chaffeensis. J Immunol. 166: 1855-1862.

Lin, M. and Rikihisa, Y. (2003a) Obligatory intracellular parasitism by Ehrlichia chaffeensis and Anaplasma phagocytophilum involves caveolae and glycosylphosphatidylinositol-anchored proteins. Cell Microbiol. 5: 809-820.

Lin, M. and Rikihisa, Y. (2003b) Ehrlichia chaffeensis and Anaplasma phagocytophilum lack genes for lipid A biosynthesis and incorporate cholesterol for their survival. Infect Immun. 71: 5324-5331.

Lin, M., Zhu, M.X. and Rikihisa, Y. (2002) Rapid activation of protein tyrosine kinase and phospholipase C-gamma2 and increase in cytosolic free calcium are required by

151 Ehrlichia chaffeensis for internalization and growth in THP-1 cells. Infect Immun. 70: 889-898.

Lin, Q., Zhang, C. and Rikihisa, Y. (2006) Analysis of involvement of the RecF pathway in p44 recombination in Anaplasma phagocytophilum and in Escherichia coli by using a plasmid carrying the p44 expression and p44 donor loci. Infect Immun. 74: 2052-2062.

Lin, Q. and Rikihisa, Y. (2005) Establishment of cloned Anaplasma phagocytophilum and analysis of p44 gene conversion within an infected horse and infected SCID mice. Infect Immun. 73: 5106-5114.

Lin, Q., Rikihisa, Y., Ohashi, N. and Zhi, N. (2003) Mechanisms of variable p44 expression by Anaplasma phagocytophilum. Infect Immun. 71: 5650-5661.

Lin, Q., Rikihisa, Y., Felek, S., Wang, X., Massung, R.F. and Woldehiwet, Z. (2004) Anaplasma phagocytophilum has a functional msp2 gene that is distinct from p44. Infect Immun. 72: 3883-3889.

Lopez, J.E., Siems, W.F., Palmer, G.H., Brayton, K.A., McGuire, T.C., Norimine, J. and Brown, W.C. (2005) Identification of novel antigenic proteins in a complex Anaplasma marginale outer membrane immunogen by mass spectrometry and genomic mapping. Infect Immun. 73: 8109-8118.

Maeda, K., Markowitz, N., Hawley, R.C., Ristic, M., Cox, D. and McDade, J.E. (1987) Human infection with Ehrlichia canis, a leukocytic rickettsia. N Engl J Med. 316: 853- 856.

Manning, D.S., Reschke, D.K. and Judd, R.C. (1998) Omp85 proteins of Neisseria gonorrhoeae and Neisseria meningitidis are similar to Haemophilus influenzae D-15-Ag and Pasteurella multocida Oma87. Microb Pathog. 25: 11-21.

Martin, M.E., Caspersen, K. and Dumler, J.S. (2001) Immunopathology and ehrlichial propagation are regulated by interferon-gamma and interleukin-10 in a murine model of human granulocytic ehrlichiosis. Am J Pathol. 158: 1881-1888.

Massung, R.F. and Slater, K.G. (2003) Comparison of PCR assays for detection of the agent of human granulocytic ehrlichiosis, Anaplasma phagocytophilum. J Clin Microbiol. 41: 717-722.

Mott, J. and Rikihisa, Y. (2000) Human granulocytic ehrlichiosis agent inhibits superoxide anion generation by human neutrophils. Infect Immun. 68: 6697-6703.

Mott, J., Barnewall, R.E. and Rikihisa, Y. (1999) Human granulocytic ehrlichiosis agent and Ehrlichia chaffeensis reside in different cytoplasmic compartments in HL-60 cells. Infect Immun. 67: 1368-1378.

152 Mott, J., Rikihisa, Y. and Tsunawaki, S. (2002) Effects of Anaplasma phagocytophila on NADPH oxidase components in human neutrophils and HL-60 cells. Infect Immun. 70: 1359-1366.

Nicholson, D.W. (1999) Caspase structure, proteolytic substrates, and function during apoptotic cell death. Cell Death Differ. 6: 1028-1042.

Ohashi, N., Rikihisa, Y. and Unver, A. (2001) Analysis of transcriptionally active gene clusters of major outer membrane protein multigene family in Ehrlichia canis and E. chaffeensis. Infect Immun. 69: 2083-2091.

Ohashi, N., Zhi, N., Zhang, Y. and Rikihisa, Y. (1998) Immunodominant major outer membrane proteins of Ehrlichia chaffeensis are encoded by a polymorphic multigene family. Infect Immun. 66: 132-139.

Ohashi, N., Zhi, N., Lin, Q. and Rikihisa, Y. (2002) Characterization and transcriptional analysis of gene clusters for a type IV secretion machinery in human granulocytic and monocytic ehrlichiosis agents. Infect Immun. 70: 2128-2138.

Palmer, G.H., Barbet, A.F., Davis, W.C. and McGuire, T.C. (1986) Immunization with an isolate-common surface protein protects cattle against anaplasmosis. Science. 231: 1299- 1302.

Palmer, G.H., Barbet, A.F., Cantor, G.H. and McGuire, T.C. (1989) Immunization of cattle with the MSP-1 surface protein complex induces protection against a structurally variant Anaplasma marginale isolate. Infect Immun. 57: 3666-3669.

Pantzar, M., Teneberg, S. and Lagergard, T. (2006) Binding of Haemophilus ducreyi to carbohydrate receptors is mediated by the 58.5-kDa GroEL heat shock protein. Microbes Infect. 8: 2452-2458.

Park, J., Choi, K.S. and Dumler, J.S. (2003) Major surface protein 2 of Anaplasma phagocytophilum facilitates adherence to granulocytes. Infect Immun. 71: 4018-4025.

Parola, P. (2004) Tick-borne rickettsial diseases: emerging risks in Europe. Comp Immunol Microbiol Infect Dis. 27: 297-304.

Parvathenani, L.K., Buescher, E.S., Chacon-Cruz, E. and Beebe, S.J. (1998) Type I cAMP-dependent protein kinase delays apoptosis in human neutrophils at a site upstream of caspase-3. J Biol Chem. 273: 6736-6743.

Pawelek, P.D., Croteau, N., Ng-Thow-Hing, C., Khursigara, C.M., Moiseeva, N., Allaire, M. and Coulton, J.W. (2006) Structure of TonB in complex with FhuA, E. coli outer membrane receptor. Science. 312: 1399-1402.

153 Pedra, J.H., Sukumaran, B., Carlyon, J.A., Berliner, N. and Fikrig, E. (2005) Modulation of NB4 promyelocytic leukemic cell machinery by Anaplasma phagocytophilum. Genomics. 86: 365-377.

Popov, V.L., Yu, X. and Walker, D.H. (2000) The 120 kDa outer membrane protein of Ehrlichia chaffeensis: preferential expression on dense-core cells and gene expression in Escherichia coli associated with attachment and entry. Microb Pathog. 28: 71-80.

Reneer, D.V., Kearns, S.A., Yago, T., Sims, J., Cummings, R.D., McEver, R.P. and Carlyon, J.A. (2006) Characterization of a sialic acid- and P-selectin glycoprotein ligand- 1-independent adhesin activity in the granulocytotropic bacterium Anaplasma phagocytophilum. Cell Microbiol.

Rikihisa, Y. (1991) The tribe Ehrlichieae and ehrlichial diseases. Clin Microbiol Rev. 4: 286-308.

Rikihisa, Y., Zhi, N., Wormser, G.P., Wen, B., Horowitz, H.W. and Hechemy, K.E. (1997) Ultrastructural and antigenic characterization of a granulocytic ehrlichiosis agent directly isolated and stably cultivated from a patient in New York state. J Infect Dis. 175: 210-213.

Rikihisa, Y. (2000) Diagnosis of emerging ehrlichial diseases of dogs, horses, and humans. J Vet Intern Med. 14: 250-251.

Rikihisa, Y. (2003) Mechanisms to create a safe haven by members of the family Anaplasmataceae. Ann N Y Acad Sci. 990: 548-555.

Rikihisa, Y. (2006) Ehrlichia subversion of host innate responses. Curr Opin Microbiol. 9: 95-101.

Rikihisa, Y., Ewing, S.A. and Fox, J.C. (1994) Western immunoblot analysis of Ehrlichia chaffeensis, E. canis, or E. ewingii infections in dogs and humans. J Clin Microbiol. 32: 2107-2112.

Rikihisa, Y., Zhang, C. and Christensen, B.M. (2003) Molecular characterization of Aegyptianella pullorum (Rickettsiales, Anaplasmataceae). J Clin Microbiol. 41: 5294- 5297.

Robert, V., Volokhina, E.B., Senf, F., Bos, M.P., Van Gelder, P. and Tommassen, J. (2006) Assembly factor Omp85 recognizes its outer membrane protein substrates by a species-specific C-terminal motif. PLoS Biol. 4: e377.

Roesli, C., Neri, D. and Rybak, J.N. (2006) In vivo protein biotinylation and sample preparation for the proteomic identification of organ- and disease-specific antigens accessible from the vasculature. Nature Protocols. 1: 192-199.

154 Sabarth, N., Lamer, S., Zimny-Arndt, U., Jungblut, P.R., Meyer, T.F. and Bumann, D. (2002) Identification of surface proteins of Helicobacter pylori by selective biotinylation, affinity purification, and two-dimensional gel electrophoresis. J Biol Chem. 277: 27896- 27902.

Salvesen, G.S. and Dixit, V.M. (1997) Caspases: intracellular signaling by proteolysis. Cell. 91: 443-446.

Scheel-Toellner, D., Wang, K., Craddock, R., Webb, P.R., McGettrick, H.M., Assi, L.K., et al (2004) Reactive oxygen species limit neutrophil life span by activating death receptor signaling. Blood. 104: 2557-2564.

Scheurer, S.B., Rybak, J.N., Roesli, C., Brunisholz, R.A., Potthast, F., Schlapbach, R., et al (2005) Identification and relative quantification of membrane proteins by surface biotinylation and two-dimensional peptide mapping. Proteomics. 5: 2718-2728.

Simon, H.U. (2003) Neutrophil apoptosis pathways and their modifications in inflammation. Immunol Rev. 193: 101-110.

Singu, V., Peddireddi, L., Sirigireddy, K.R., Cheng, C., Munderloh, U. and Ganta, R.R. (2006) Unique macrophage and tick cell-specific protein expression from the p28/p30- outer membrane protein multigene locus in Ehrlichia chaffeensis and Ehrlichia canis. Cell Microbiol. 8: 1475-1487.

Sprick, M.R. and Walczak, H. (2004) The interplay between the Bcl-2 family and death receptor-mediated apoptosis. Biochim Biophys Acta. 1644: 125-132.

Sun, W., JW, I.J., Telford, S.R., 3rd, Hodzic, E., Zhang, Y., Barthold, S.W. and Fikrig, E. (1997) Immunization against the agent of human granulocytic ehrlichiosis in a murine model. J Clin Invest. 100: 3014-3018.

Tebele, N., McGuire, T.C. and Palmer, G.H. (1991) Induction of protective immunity by using Anaplasma marginale initial body membranes. Infect Immun. 59: 3199-3204.

Thomas, V., Samanta, S., Wu, C., Berliner, N. and Fikrig, E. (2005) Anaplasma phagocytophilum modulates gp91phox gene expression through altered interferon regulatory factor 1 and PU.1 levels and binding of CCAAT displacement protein. Infect Immun. 73: 208-218.

Unver, A., Rikihisa, Y., Stich, R.W., Ohashi, N. and Felek, S. (2002) The omp-1 major outer membrane multigene family of Ehrlichia chaffeensis is differentially expressed in canine and tick hosts. Infect Immun. 70: 4701-4704.

155 Voulhoux, R., Bos, M.P., Geurtsen, J., Mols, M. and Tommassen, J. (2003) Role of a highly conserved bacterial protein in outer membrane protein assembly. Science. 299: 262-265.

Wang, T., Akkoyunlu, M., Banerjee, R. and Fikrig, E. (2004) Interferon-gamma deficiency reveals that 129Sv mice are inherently more susceptible to Anaplasma phagocytophilum than C57BL/6 mice. FEMS Immunol Med Microbiol. 42: 299-305.

Wang, X., Kikuchi, T. and Rikihisa, Y. (2006) Two monoclonal antibodies with defined epitopes of P44 major surface proteins neutralize Anaplasma phagocytophilum by distinct mechanisms. Infect Immun. 74: 1873-1882.

Ward, C., Chilvers, E.R., Lawson, M.F., Pryde, J.G., Fujihara, S., Farrow, S.N., et al (1999) NF-kappaB activation is a critical regulator of human granulocyte apoptosis in vitro. J Biol Chem. 274: 4309-4318.

Watson, R.W., Rotstein, O.D., Parodo, J., Bitar, R., Marshall, J.C., William, R. and Watson, G. (1998) The IL-1 beta-converting enzyme (caspase-1) inhibits apoptosis of inflammatory neutrophils through activation of IL-1 beta. J Immunol. 161: 957-962.

Winslow, G.M., Yager, E., Shilo, K., Volk, E., Reilly, A. and Chu, F.K. (2000) Antibody-mediated elimination of the obligate intracellular bacterial pathogen Ehrlichia chaffeensis during active infection. Infect Immun. 68: 2187-2195.

Yago, T., Leppanen, A., Carlyon, J.A., Akkoyunlu, M., Karmakar, S., Fikrig, E., et al (2003) Structurally distinct requirements for binding of P-selectin glycoprotein ligand-1 and sialyl Lewis x to Anaplasma phagocytophilum and P-selectin. J Biol Chem. 278: 37987-37997.

Yoshiie, K., Kim, H.Y., Mott, J. and Rikihisa, Y. (2000) Intracellular infection by the human granulocytic ehrlichiosis agent inhibits human neutrophil apoptosis. Infect Immun. 68: 1125-1133.

Zhi, N., Ohashi, N. and Rikihisa, Y. (1999) Multiple p44 genes encoding major outer membrane proteins are expressed in the human granulocytic ehrlichiosis agent. J Biol Chem. 274: 17828-17836.

Zhi, N., Ohashi, N., Tajima, T., Mott, J., Stich, R.W., Grover, D., et al (2002) Transcript heterogeneity of the p44 multigene family in a human granulocytic ehrlichiosis agent transmitted by ticks. Infect Immun. 70: 1175-1184.

156