MAMMALIAN HOST CELL RESERVOIRS DURING ANAPLASMA INFECTION

By

HEATHER L. WAMSLEY

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2009

1

© 2009 Heather L. Wamsley

2

TABLE OF CONTENTS

page

LIST OF TABLES...........................................................................................................................5 

LIST OF FIGURES .........................................................................................................................6 

ABSTRACT.....................................................................................................................................8 

CHAPTER

1 BACKGROUND AND SIGNIFICANCE..............................................................................10 

Overview and Long-Term Objectives ....................................................................................10 Anaplasma Background..........................................................................................................10 

Anaplasma phagocytophilum and Anaplasma marginale as Significant Pathogens within Anaplasmataceae ..............................................................................................10 

Anaplasma Life Cycle versus the Natural Disease Course of Anaplasmosis .................12 Anaplasma phagocytophilum ...................................................................................12 Anaplasma marginale ..............................................................................................14 

Central Hypothesis..................................................................................................................15 

2 IN SITU DETECTION OF ANAPLASMA BY DNA TARGET-PRIMED ROLLING-CIRCLE AMPLIFICATION OF A PADLOCK PROBE AND INTRACELLULAR COLOCALIZATION WITH IMMUNOFLUORESCENTLY LABELED HOST CELL VON WILLEBRAND FACTOR ...........................................................................................17 

Introduction.............................................................................................................................17 Materials and Methods ...........................................................................................................17 

Cultivation of Anaplasma spp. ........................................................................................17 In Situ Rolling-Circle Amplification of Padlock Probes.................................................18 Padlock Probe Design......................................................................................................21 Indirect Immunofluorescent Staining of von Willebrand Factor in Cultured

Endothelial cells...........................................................................................................22 Results.....................................................................................................................................23 

In Situ Detection of Anaplasma spp. by DNA Target-Primed Rolling-Circle Amplification of Padlock Probes .................................................................................23 

In situ Intracellular Colocalization of the A. phagocytophilum Rolling-Circle Amplification Product and von Willebrand Factor Immunofluorescence...................24 

Discussion...............................................................................................................................25 

3 EXPERIMENTAL INOCULATION OF DOGS WITH ANAPLASMA PHAGOCYTOPHILUM, MOLECULAR EVIDENCE OF PERSISTENT INFECTION FOLLOWING DOXYCYCLINE THERAPY, AND INVESTIGATION OF ENDOTHELIAL CELLS AS SOURCE OF INFECTIOUS INOCULUM AND A REPOSITORY OF CHRONIC INFECTION IN DOGS .......................................................33 

3

Introduction.............................................................................................................................33 Materials and Methods ...........................................................................................................34 

Animals, Inocula, and Monitoring for Development of Infection...................................34 Serologic Assays for A. phagocytophilum Infection .......................................................36 PCR Detection of A. phagocytophilum DNA in Blood and Tissues ...............................37 Immunosuppression, Antibiotic Treatment, and Euthanasia...........................................38 In Situ DNA Target-Primed Rolling-Circle Amplification of a Padlock Probe for

Detection of Anaplasma phagocytophilum ..................................................................39 Results.....................................................................................................................................39 

Physical Exam, Hematopathology, and Clinical Chemistry ...........................................39 Serologic Evidence of Infection ......................................................................................40 Molecular Evidence of Infection .....................................................................................40 

NY18-infected dogs .................................................................................................40 Canine isolate-infected dogs ....................................................................................41 

In Situ Detection of A. phagocytophilum ........................................................................42 Discussion...............................................................................................................................42 

4 EXPERIMENTAL INFECTION OF CATTLE WITH ANAPLASMA MARGINALE, DISTRIBUTION OF ORGANISMS IN BOVINE TISSUES, AND INVESTIGATION OF ENDOTHELIAL CELLS AS A NIDUS OF ACUTE AND CHRONIC INFECTION ...........................................................................................................................53 

Introduction.............................................................................................................................53 Materials and Methods ...........................................................................................................54 

Animals, Inoculum, Monitoring Infection and Parasitemia, and Euthanasia..................54 Tissue Sample Collection and Processing.......................................................................55 PCR Detection of A. marginale DNA in Blood and Tissues ..........................................56 Microscopic Detection of A. marginale Organisms in Tissues.......................................57 

Dual indirect immunofluorescent staining of A. marginale and endothelial cell antigens .................................................................................................................57 

Electron microscopy.................................................................................................58 In Situ DNA target-primed rolling-circle amplification of a padlock probe for

detection of Anaplasma marginale .......................................................................59 Statistical Analysis ..........................................................................................................60 

Results.....................................................................................................................................60 Seroconversion, Parasitemia, and Anemia ......................................................................60 Tissue Distribution of A. marginale ................................................................................60 Microscopic Examination for A. marginale in Tissues ...................................................61 

Discussion...............................................................................................................................63 

5 CONCLUSIONS AND FUTURE RESEARCH ....................................................................78 

LIST OF REFERENCES...............................................................................................................81 

BIOGRAPHICAL SKETCH .........................................................................................................92 

4

LIST OF TABLES

Table page 3-1 Treatments administered during two infection trials using different A.

phagocytophilum isolates...................................................................................................45 

3-2 Comparison of A. phagocytophilum SNAP®4Dx® seroconversion between the two canine experimental inoculation studies (NY18 isolate vs. canine isolate).......................45 

3-3 Detection of A. phagocytophilum DNA by nested- and real-time PCR in the blood of the NY18 isolate-infected dogs pre-infection through day 340 post-infection..................46 

5

LIST OF FIGURES

Figure page 2-1 Padlock probes and fluorescent oligonucleotides. .............................................................29 

2-2 Detection of A. phagocytophilum using in situ DNA target-primed rolling-circle amplification of a padlock probe. ......................................................................................30 

2-3 Detection of A. marginale using in situ DNA target-primed rolling-circle amplification of a padlock probe. ......................................................................................31 

2-4 Detection of A. phagocytophilum using in situ DNA target-primed rolling-circle amplification of a padlock probe and concurrent indirect immunofluorescent staining of von Willebrand Factor. ..................................................................................................32 

3-1 PCR primers and fluoresceinated oligonucleotide probe...................................................47 

3-2 Body temperature of one of the A. phagocytophilum canine isolate-infected dogs during the infection trial.....................................................................................................48 

3-3 Immunoblot of polyclonal sera against A. phagocytophilum NY18 isolate demonstrates immuonoreactivity (seropositivity) in the canine isolate-infected dogs......49 

3-4 Southern blot of the A. phagocytophilum msp2-based nested-PCR products from the NY18 isolate-infected dogs................................................................................................50 

3-5 Southern blots of the A. phagocytophilum msp2-based nested-PCR products from the canine isolate-infected dogs that received doxycycline from day 155 to 183 post-inoculation..........................................................................................................................51 

3-6 In situ DNA target-primed rolling-circle amplification of a padlock probe in canine isolate-infected Dog 1 kidney day 233 post-infection. ......................................................52 

4-1 Steer from A. marginale tick-feeding transmission trial....................................................66 

4-2 Competitive ELISA results for steers during the A. marginale tick-feeding transmission trial................................................................................................................67 

4-3 Packed cell volume and microscopic parasitemia for steers during the A. marginale tick-feeding transmission trial............................................................................................68 

4-4 Real-time quantitative PCR parasitemia detected as opag2 copies per microliter of peripheral blood of steers during the A. marginale tick-feeding transmission trial...........69 

4-5 Real-time quantitative PCR opag2 copies per milligram of tissue in the infected steer (5237) that was euthanized during the first peak in parasitemia day 41 post-infection. ...70 

6

4-6 Real-time quantitative PCR opag2 copies per milligram of tissue in the infected steer (3102) that was euthanized during the first trough in parasitemia day 64 post-infection. ............................................................................................................................71 

4-7 Dual indirect immunofluorescent staining of A. marginale MSP5 (green) and von Willebrand factor (red) in uninfected (steer 5180) and infected (steer 5237) tick-bite site dermal punch biopsies day 8 post-infection. ...............................................................72 

4-8 Toluidine blue-stained and transmission electron microscopy tick-bite site dermal punch biopsies day 6 post-infection...................................................................................73 

4-9 Toluidine blue-stained and transmission electron microscopy lung from the steer (3102) that was euthanized during the first trough in parasitemia day 64 post-infection. ............................................................................................................................74 

4-10 In situ DNA target-primed rolling-circle amplification of a padlock probe using sonicated fish sperm in the oligonucleotide hybridization solutions in A. marginale-infected (steer 3102) tick-bite site dermis day 6 post-infection.........................................75 

4-11 In situ DNA target-primed rolling-circle amplification of a padlock probe using sonicated fish sperm in the oligonucleotide hybridization solutions in lung from the A. marginale-infected (steer 5237) that was euthanized during the first peak in parasitemia day 41 post-infection. .....................................................................................76 

4-12 In situ DNA target-primed rolling-circle amplification of a nonspecific padlock probe using sheared calf thymus DNA in the oligonucleotide hybridization solutions in A. marginale-infected tick-bite site dermis....................................................................77 

7

ABSTRACT OF DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE

REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

MAMMALIAN HOST CELL RESERVOIRS DURING ANAPLASMA INFECTION

By

Heather L. Wamsley

May 2009 Chair: Anthony F. Barbet Major: Veterinary Medical Sciences

Endothelial cell culture and preliminary immunofluorescent staining of Anaplasma-

infected tissues suggest that endothelial cells may be an in vivo nidus of mammalian infection.

To investigate endothelial cells and other potential sites of Anaplasma infection in mammalian

tissues, a sensitive and specific, in situ rolling-circle amplification technique to detect localized

Anaplasma gene sequences was developed. Via the technique described here and von

Willebrand factor immunofluorescence, A. phagocytophilum and A. marginale were successfully

localized in situ within cultured mammalian cells. This is the first application of this in situ

method for detection of a microorganism and forms the foundation for applications of this

technique to detect, localize, and analyze Anaplasma nucleotide sequences in the tissues of

infected hosts and in cell cultures.

Three Anaplasma-infection trials using immunocompetent dogs and cattle were performed

to investigate different aspects of endothelial cells as they relate to Anaplasma life cycles and to

further describe clinical aspects of Anaplasma pathogenesis. Four Beagles were inoculated with

A. phagocytophilum from different sources, allowed to develop chronic infection, and treated

with doxycycline. Regardless of isolate or duration of doxycycline treatment, A.

phagocytophilum DNA remained detectable for several months in blood and tissues, though

8

organisms were not identified microscopically. This is the first infection of dogs using cultured

endothelial cells as the source of inoculum, the first demonstration of molecular evidence of

chronic, persistent infection in blood and tissues of subclinical dogs despite doxycycline

treatment, and the first investigation of endothelial cells as a potential in vivo source of A.

phagocytophilum during chronic canine infection using in situ rolling-circle amplification. Two

steers were inoculated with A. marginale by tick-feeding transmission and were euthanized at

different points within the parasitemic cycle. The tissue distribution of A. marginale during peak

and trough parasitemia was described using real-time PCR, though organisms were not identified

in tissues microscopically. This is the first survey of A. marginale tissue distribution after tick-

transmission and the first investigation in immunocompetent cattle of endothelial cells as a

potential in vivo source of A. marginale in tick-bite sites and distant tissues using three

techniques (immunofluorescence, electron microscopy, in situ rolling-circle amplification).

9

CHAPTER 1 BACKGROUND AND SIGNIFICANCE

Overview and Long-Term Objectives

The order Rickettsiales is comprised of many important zoonotic pathogens, including the

agent of human granulocytotropic anaplasmosis, Anaplasma phagocytophilum, the agent of

Rocky Mountain spotted fever, Rickettsia rickettsii, and the agent of human monocytotropic

ehrlichiosis, Ehrlichia chaffeensis (33). In vivo, endothelial cell infection in mammals has been

suggested or demonstrated for many pathogens within this order, such as, A. phagocytophilum,

A. marginale, R. rickettsii, R. typhi, R. conori, R. prowazekii, E. ruminantium (24, 35, 54, 94,

100, 112). The importance of endothelial cells during the pathogenesis of these rickettsial and

ehrlichial diseases and their interactions with common co-infecting tick-borne organisms, such as

the agent of Lyme disease, Borrelia burgdorferi, is beginning to be recognized (24, 54, 81, 88,

100, 112).

The long-term objectives of this work are to develop a more complete understanding of the

• contribution of mammalian endothelial cells to the mechanisms leading to the establishment and persistence of Anaplasma infection in mammals

• underlying host cell-specific adaptations by Anaplasma during different life cycle stages.

This knowledge will guide future investigations aimed at development of effective treatment and

control strategies for these ehrlichial pathogens of humans and animals.

Anaplasma Background

Anaplasma phagocytophilum and Anaplasma marginale as Significant Pathogens within Anaplasmataceae

Since initially reported in 1994 (25), the number of human A. phagocytophilum infections

reported to the CDC has substantially increased, rendering it the third most common tick-borne

infection of humans in the United States (8, 34). A. phagocytophilum infections are likely under-

10

reported since subclinical infections or mild, cold-like disease are probable given the relatively

high seroprevalence among humans residing in endemic regions (7, 8, 34). When A.

phagocytophilum infections are diagnosed, over half of the individuals with this diagnosis

require potentially costly hospitalization, and 7% require admission to the intensive care unit

(34). When diagnosed early, most patients exhibit rapid clinical improvement after doxycycline

treatment; however, protracted disease course or death is possible among those who are heavily

infected, elderly, or otherwise immunocompromised (8, 25, 34).

Globally, A. marginale is the most prevalent tick-transmitted pathogen of cattle and other

ruminants. It is the etiologic agent of bovine anaplasmosis, an economically significant,

arthropod-borne hemolytic disease of cattle that engenders substantial morbidity and mortality in

United States’ bovine livestock, with estimated annual losses exceeding 300 million dollars

(1986 U.S. dollars) (63).

In addition to being an important pathogen of ruminant livestock, A. marginale is the type

species of the family Anaplasmataceae, which encompasses several human and veterinary

pathogens with zoonotic potential (33). Previous studies of A. marginale infections of cattle

have revealed important features of the biology of these organisms including the nature and

length of persistent infections (39, 91), the immunologic and molecular basis of persistence

(10,18, 44, 45, 80), and transmission dynamics in the mammalian host and tick vectors (62, 92,

93, 103, 104).

Anaplasmataceae are obligate intracellular parasites of eukaryotic hosts. Anaplasma are

all maintained in a cycle between mammalian cells of myeloid lineage and tick epithelial cells

(33). Beyond similar morphologic and life cycle features, there is molecular biological

homology between A. marginale and A. phagocytophilum. Based upon housekeeping gene

11

sequence, there is 96.1% or greater similarity between all of the members of Anaplasmataceae;

and Anaplasma share several common surface protein antigens (33). Phylogeographic clades of

A. marginale and A. phagocytophilum can be established based upon comparison of the genetic

sequence for one of the surface proteins, major surface protein 4 (MSP4), which is common to

both organisms (28, 29). These two organisms and other members of the family

Anaplasmataceae employ host cell-specific differential expression of surface antigens (11, 15,

16, 46, 56, 58, 74, 75, 83, 87, 106, 109, 110, 119), likely representing an adaptation to improve

cell-specific infectivity or mammalian host immune evasion. A. marginale and A.

phagocytophilum exhibit similar operon structure and promoter regulation for the

immunodominant antigen common to both organisms (MSP2 and MSP2 (p44), respectively) (9,

11, 12). Both organisms display a similar, plasmid-independent adaptation to evade the

mammalian host immune system via sequential generation of antigenic variants by segmental

gene conversion using a single chromosomal MSP2 expression site and a system of functional

chromosomal pseudogenes (9, 11, 18, 45, 71, 72, 73, 114, 120); this facilitates persistent

mammalian infection, an essential component of the Anaplasma life cycle.

Anaplasma Life Cycle versus the Natural Disease Course of Anaplasmosis

Anaplasma phagocytophilum

High seroprevalence in the canine (78, 98), equine (22, 77), and human (7,8) populations

suggests that subclinical A. phagocytophilum infection or mild disease due to infection is

common in endemic regions where ixodid tick density is high. Chronic A. phagocytophilum

infection has been reported in dogs (36, 37), cats (67), rodents (55, 108), horses (42, 89, 97),

lambs (107), and sheep (41) and is suspected in some human cases (30, 31, 95, 96). Persistent

infection by Ehrlichia and other Anaplasma species is commonly observed in animals and

humans (5, 26, 27, 32, 39, 52, 59, 82, 91, 97, 99, 118).

12

Doxycycline is considered the optimal treatment choice for most cases of A.

phagocytophilum infection (1, 6, 34, 50, 79), and rapid clinical improvement within one to two

days of treatment is typically observed in most uncomplicated cases (1, 6, 21, 34, 36, 50, 67).

However, doxycycline is bacteriostatic against A. phagocytophilum (79). Controlled studies

investigating the optimal duration of treatment or the dosage regime required to completely clear

viable A. phagocytophilum organisms from infected individuals, rather than to ameliorate clinical

signs and potentially induce a carrier state of infection, have not been reported. This represents a

gap in the current understanding of this zoonotic pathogen. Further, there is evidence that A.

phagocytophilum deoxyribonucleic acid (DNA) in domestic cats (67), E. chaffeensis DNA in

dogs (21), and E. canis DNA and viable organisms in dogs (51, 57, 102, 116) persist in blood or

organs despite doxycycline administration at a dose and duration generally considered effective

against Ehrlichiae. It is unknown whether or not A. phagocytophilum DNA can be detected in

tissues of infected humans after treatment with doxycycline since tissues have not been

specifically examined for this purpose in cases of human granulocytotropic anaplasmosis (34).

A. phagocytophilum-infected neutrophils have delayed apoptosis and prolonged half-life

compared to uninfected neutrophils, which normally have a half-life of 10 to 12 hours (23, 47,

69, 101, 117). The observation that A. phagocytophilum establishes chronic infections in

animals on the order of several months (37, 41, 89) is incongruous with the theory that terminally

differentiated circulating granulocytes, which have a relatively brief, finite half-life, are the only

site of infection in mammals. Also, though it is theoretically possible that A. phagocytophilum

may transfer between terminally differentiated granulocytes in the peripheral blood (34), it is

questionable whether these peripheral blood cells are the sole source of microorganisms required

for sustained infection.

13

Based upon initial in vitro cultivation of A. phagocytophilum in undifferentiated bone

marrow hematopoietic precursors (CD34+, HLA-DR+), it was suggested that bone marrow

progenitors may serve as a continuous source of microorganisms during chronic infection (60).

However, that study and subsequent studies suggest that A. phagocytophilum preferentially

infects cells with mature myeloid or neutrophil-like differentiation rather than less differentiated

precursor cells (14, 60, 61). It is currently unknown whether A. phagocytophilum-infected

granulocytes that are identified in peripheral blood are infected as immature precursors within

the bone marrow or if they are infected while in peripheral circulation (14).

More recent reports indicate that endothelial cells may serve as a nidus of infection in

mammals and as a source of organisms to infect granulocytes circulating in peripheral blood. A.

phagocytophilum can be continuously cultivated in fetal primate endothelial cells (81). And,

immunofluorescence has been used to colocalize an endothelial cell antigen and an A.

phagocytophilum surface protein within the cardiac and hepatic microvasculature of mice with

severe combined immunodeficiency (SCID) after 7 weeks of infection. However,

photomicrographs were not published (54).

Anaplasma marginale

The only recognized mammalian life cycle stage of A. marginale is within terminally

differentiated, circulating mature erythrocytes (33), which have a half-life of approximately 130

days (64). However, bovine anaplasmosis is characterized by obligatory, persistent infection of

ruminant hosts associated with chronic, sub-microscopic, cyclic parasitemias due to antigenic

variation of the surface proteins, MSP2 and MSP3 (3, 9, 10, 17, 18, 19, 38, 39, 43, 44, 59, 80, 90,

91). There are disparities between this mammalian life cycle model and the natural disease

course of anaplasmosis. While, A. marginale organisms may transfer between terminally

differentiated erythrocytes (48, 111), it is equivocal whether these peripheral blood cells are the

14

sole source of microorganisms required for sustained and proliferative infection. The finite half-

life of mature erythrocytes suggests that other, secondarily infected mammalian cells may be

required to produce the chronic disease course and cyclic parasitemias of bovine anaplasmosis.

Electron microscopy has revealed several intact initial bodies within A. marginale-infected

erythrocytes co-cultured with endothelial cells, whereas, only degenerating initial bodies were

observed in erythrocytes cultured in the absence of endothelial cells (111). Recent experiments

have documented that A. marginale can be successfully propagated within primate endothelial

cells in long term cell culture (81). Preliminary observation of immunofluorescently stained

tissue has suggested the presence of A. marginale within renal endothelial cells of a

splenectomized steer, though negative control tissues were not examined (24).

Central Hypothesis

The life cycle of A. phagocytophilum and related ehrlichial organisms is complex,

involving peripheral blood cells and epithelial cells of mammalian and arthropod hosts,

respectively (33). However, steps in the establishment and persistence of A. phagocytophilum

and A. marginale infection within mammalian hosts are incompletely characterized. The current

paradigm indicates that within mammalian hosts, Anaplasma infection only occurs within

mature, terminally-differentiated blood cells (33). These well-differentiated blood cells that are

infected by Anaplasma have a finite half-life (53); this creates an incongruity between the current

life cycle paradigm and the natural disease course of anaplasmosis, which is characterized by

persistent infection of humans and animals (30, 36, 37, 41, 42, 55, 59, 67, 89, 91, 95, 96, 97, 107,

108). It is questionable whether or not terminally differentiated peripheral blood cells are the

sole source of microorganisms required for sustained, if not proliferative, infection. This

disparity between the present understanding of Anaplasma life cycles and the observed natural

course of anaplasmosis represents a gap in current knowledge.

15

The central hypothesis of this work is that endothelial cells serve as a nidus of A.

phagocytophilum and A. marginale infection in dogs and cattle, respectively. The central

hypothesis is founded on the following four observations. First, A. phagocytophilum and A.

marginale establish chronic infection in mammalian hosts (19, 36, 37, 39, 41, 42, 55, 59, 89, 91,

97, 107, 108). Second, A. phagocytophilum and A. marginale can attach in vitro to bovine and

primate endothelial cells, invade, and replicate (81). Third, preliminary observations of

immunofluorescently stained post-mortem tissues suggest that A. phagocytophilum and A.

marginale may invade endothelial cells in SCID mice (54) and a splenectomized steer (24).

Fourth, an endothelial cell life cycle stage has been confirmed for another pathogen in the family

Anaplasmataceae, E. ruminantium, which exhibits a vegetative growth stage within endothelial

cells (94). The work described here seeks to further define the life cycle of Anaplasma through

investigations designed to complete the following:

• discover whether or not A. phagocytophilum persists within canine endothelial cells during chronic infection

• determine whether or not A. marginale invades bovine endothelial cells at tick attachment sites during initial infection and whether or not A. marginale persists within bovine endothelial cells during cyclic parasitemia (peak parasitemia and during a trough).

By developing a more thorough understanding of Anaplasma life cycles, this work will lay

the foundation for future comparative studies of host cell-specific adaptations in

Anaplasmataceae and guide the development of effective treatment and control strategies for

both the zoonotic pathogen, A. phagocytophilum, and the significant ruminant pathogen, A.

marginale.

16

CHAPTER 2 IN SITU DETECTION OF ANAPLASMA BY DNA TARGET-PRIMED ROLLING-CIRCLE

AMPLIFICATION OF A PADLOCK PROBE AND INTRACELLULAR COLOCALIZATION WITH IMMUNOFLUORESCENTLY LABELED HOST CELL VON WILLEBRAND

FACTOR

Introduction

The life cycle of Anaplasma spp. involves mammalian peripheral blood cells and arthropod

epithelial cells (33). Steps in the tick feeding-associated establishment and persistence of

Anaplasma infection within mammalian hosts are incompletely characterized. Recent in vitro

and preliminary in vivo immunofluorescence studies suggest that endothelial cells may be a

nidus of Anaplasma infection in mammals and a source of organisms to infect circulating blood

cells (24, 54, 81). The purpose of this investigation was to develop a specific and sensitive

technique for in situ detection of Anaplasma within tissues of infected hosts with special

attention to mammalian endothelial cells. In the future, this method could be used to determine

the in vivo cellular localization of potentially cryptic infection nidi and to provide nucleotide

sequence information in situ.

Materials and Methods

Cultivation of Anaplasma spp.

The NY18 isolate of A. phagocytophilum or the Virginia isolate of A. marginale was

cultivated in fetal rhesus monkey (Macaca mulatta) RF/6A endothelial cells (American Type

Culture Collection, ATCC CRL-1780™, Manassas, VA), and the HZ isolate of A.

phagocytophilum was cultivated in human (Homo sapiens) HL-60 myeloblastic leukemia cells

(American Type Culture Collection, ATCC CCL-240™, Manassas, VA) as described (49, 81)

(DMEM medium for the RF/6A endothelial cells (HyClone, Logan, UT) or RPMI-1640 medium

for the HL-60 myeloblastic leukemia cells (HyClone, Logan, UT) supplemented with 10% heat-

inactivated fetal bovine serum (HyClone, Logan, UT), 2 mM Gibco L-Glutamine-200 mM

17

(100X) liquid (final concentration 4 mM) (Invitrogen, Carlsbad, CA), 0.25% NaHCO3 (Sigma,

St. Louis, MO), 25 mM HEPES (Sigma, St. Louis, MO), [pH 7.5], 37°C, 5% CO2). Uninfected

RF/6A endothelial cells and uninfected HL-60 myeloblastic leukemia cells were similarly

maintained. When at least 80% of the cells were A. phagocytophilum-infected or at least 20% of

the cells were A. marginale-infected as determined by light microscopy, cell suspensions were

diluted and cytocentrifuged on to Bond-Rite glass microscope slides (Richard-Allan Scientific,

Kalamazoo, MI). To form a cell suspension, the RF/6A endothelial cell culture monolayers were

detached from the culture flask using 0.25% trypsin (HyClone, Logan, UT). The HL-60

myeloblastic leukemia cell line is a nonadherent cell line, which grows as a cell suspension;

therefore, treatment with trypsin was not necessary.

In Situ Rolling-Circle Amplification of Padlock Probes

In situ DNA target-primed rolling-circle amplification of padlock probes was performed

with modifications of a previously described technique (68). All reactions were performed on

microscope slides without coverslips. The final volume of all reactions was 40 µL. All heated

reactions were performed in a 16/16 dual block slide chamber mounted on a DNA Engine (PTC-

200) Peltier Thermal Cycler (Bio-Rad Laboratories, Hercules, CA). Cytospin culture material on

the microscope slides was uniformly treated as follows.

The slides were washed twice in 1X phosphate-buffered saline (PBS) ([pH 7.4], 2 min),

fixed in 70% denatured ethanol (20 min), and subsequently washed twice in 1X PBS (2 min).

The cells were permeabilized in HCl ([pH 3.6], 37°C, RF/6A endothelial cells: 3 min; HL-60

myeloblastic leukemia cells: 1.5 min). Afterward, the slides were washed three times in 1X

PBS (2 min).

The bacterial genome was made irreversibly linear by endonuclease digestion (A.

phagocytophilum: AfeI (New England BioLabs, Ipswich, MA); A. marginale: ZraI (New

18

England BioLabs, Ipswich, MA), 0.5 U/µL in 1X supplied enzyme buffer plus 0.2 µg/µL bovine

serum albumin (New England BioLabs, Ipswich, MA), 37°C, 30 min) followed by a brief rinse

in buffer A (0.1 M Tris-HCl [pH 7.5], 0.15 M NaCl, and 0.05% Tween-20). The genomic DNA

target was made single-stranded by 5’ to 3’ exonucleolysis (Lambda Exonuclease (New England

BioLabs, Ipswich, MA), 0.2 U/µL in 1X supplied enzyme buffer plus 0.2 µg/µL bovine serum

albumin and 10% glycerol (Sigma, St. Louis, MO), 37°C, 15 min) followed by a brief rinse in

buffer A.

The genomic DNA target was detected by hybridization to a circularizable, linear,

oligonucleotide padlock probe (Figure 2-1) (either an A. phagocytophilum-specific probe, an A.

marginale-specific probe, or a nonspecific probe, 0.10 µM in 2X SSC [pH 7.0], 20% formamide

(Fisher Scientific, Pittsburg, PA), and 0.5 µg/µL sonicated fish sperm (DNA MB-grade, Roche

Applied Science, Indianapolis, IN), 37°C, 15 min). Subsequently, the slides were washed in

prewarmed buffer B (2X SSC [pH 7.0] and 0.05% Tween-20, 37°C, 5 min) and rinsed briefly in

buffer A.

The A. phagocytophilum-specific probe or the A. marginale-specific padlock probe, which

should hybridize as a nicked circle to its complementary genomic DNA target, was then

irreversibly locked into place by enzymatic formation of a phosphodiester bond between the

juxtaposed 5' phosphate and 3' hydroxyl termini of the padlock probe (T4 ligase (New England

BioLabs, Ipswich, MA), 100 U/µL in 1X supplied enzyme buffer plus 0.2 µg/µL bovine serum

albumin, 16°C, 15 min). The nonspecific padlock probe should remain linear and be washed

away during this and subsequent steps since this probe should fail to form the appropriate

conformation required for ligase recognition and activity. Afterward, the slides were washed in

19

prewarmed buffer B (37°C, 5 min), rinsed once in buffer A, and dehydrated in graded denatured

ethanols (75%, 85%, 100%; 3 min each).

After exonucleolysis of any remaining 3’ single-stranded genomic DNA, the genomic

DNA target was used to prime isothermal, in situ rolling-circle amplification of the specifically

bound padlock probe (phi29 DNA Polymerase (New England BioLabs, Ipswich, MA), 1.0 U/µL

in 1X supplied enzyme buffer plus 0.25 mM of each dNTP (Deoxynucleotide Solution Mix, New

England BioLabs, Ipswich, MA), 0.2 µg/µL bovine serum albumin, and 10% glycerol, 37°C, 30

min). During the 30 min DNA polymerization, the slides were removed from the slide chamber

every 10 min to gently agitate the reaction mixture. The slides were then rinsed briefly in buffer

A before the single-stranded, loosely coiled, concatameric amplification product was hybridized

to a fluorescently labeled, linear oligonucleotide tag (oligonucleotide AB1252 (green) or

oligonucleotide AB1279 (red), 0.25 µM in 2X SSC, 20% formamide, and 0.5 µg/µL sonicated

fish sperm, 37°C, 15 min).

Afterward, the slides were rinsed briefly in buffer A, dehydrated in graded denatured

ethanols (75%, 85%, 100%; 3 min each), and either immunofluorescently stained for von

Willebrand factor or immediately air-dried and coverslip-mounted using VECTASHIELD

HardSet Mounting Medium with 1.5 µg/mL DAPI (4',6-diamidino-2-phenylindole) (Vector

Laboratories, Burlingame, CA). The slides were examined using a Leica DMI 3000B inverted

microscope fitted for epifluorescence and equipped with a digital camera (Micropublisher 3.3

RTV, QImaging Corporation, Surrey, BC Canada) and then stored at 4°C. Digital images were

collected using QCapture Pro 5.1.1.14 (QImaging Corporation, Surrey, BC Canada) and

uniformly processed using SPOT Advanced Windows Version 4.0.9 (Diagnostic Instruments,

20

Sterling Heights, MI) and Adobe Photoshop Elements 2.0 (Adobe Systems Incorporated, San

Jose, CA).

Padlock Probe Design

Three circularizable, linear, oligonucleotide padlock probes were used (Figure 2-1) (MWG

Biotech, High Point, NC). The A. phagocytophilum-specific probe (AB1251) was designed to

include an A. phagocytophilum genomic DNA target-specific-sequence whose cognate is a

conserved 5’ region of msp2 (p44) that is present in the expression site of three geographic

isolates (HZ, United States (35), Norway (13), and Sweden (13) and 81 different pseudogenes of

the HZ isolate (35). Of these target sequences within the A. phagocytophilum genome (HZ

isolate), there are 29 potential targets (the expression site and 28 pseudogenes) that are

associated with a 3’ AfeI endonuclease recognition site within 100 base pairs (bp) of the genomic

target and could, therefore, be detected by the technique described here.

The A. marginale-specific probe (AB1270) was designed to include an A. marginale

genomic DNA target-specific-sequence whose cognate is orfY that is repeated 8 times in the

genome (20). Of these target sequences within the A. marginale genome (St. Maries isolate),

there are 7 potential targets that are associated with a 3’ ZraI endonuclease recognition site

within 1 bp of the genomic target and could, therefore, be detected by the technique described

here. The use of repetitive sequences (A. phagocytophilum: msp2 (p44); A. marginale: orfY) as

genomic DNA targets for padlock probe hybridization was expected to increase sensitivity

beyond that provided by rolling-circle amplification alone.

The target-specific-sequence of the padlock probe is within the 5’ and 3’ arms of the probe,

which are joined by an intervening linker region. The rolling-circle amplification product is

detected by a fluorescently labeled oligonucleotide tag (AB1252 or AB1279) that is the

complement of the padlock probe linker region amplification product. The linker region of the

21

nonspecific padlock probe (AB1253) was identical to the A. phagocytophilum-specific and A.

marginale-specific padlock probes. The 5’ and 3’ arms of the nonspecific probe contained the

same nucleotide composition as the A. phagocytophilum-specific probe; however, the sequence

of the nucleotides was randomized.

Indirect Immunofluorescent Staining of von Willebrand Factor in Cultured Endothelial cells

After the final graded alcohol dehydration of the in situ rolling-circle amplification

procedure, slides that had been reacted with the A. phagocytophilum-specific padlock probe or

the nonspecific padlock probe were immunofluorescently stained for von Willebrand factor. All

reactions were performed on microscope slides without coverslips; the final volume of all

reactions was 150 µL. The cytospin culture material was blocked with normal rabbit serum

(X0902 (Dako, Carpinteria, CA) 5% in 1X PBS [pH 7.4], 30 min) and subsequently washed once

in 1X PBS (5 min). The cytospin culture material was then incubated with either antibody-free

diluent or rabbit polyclonal anti-human von Willebrand factor antibody (N1505 (Dako,

Carpinteria, CA), 1 to 2 dilution of the proprietary antibody solution in 0.05 M Tris-HCl [pH 7.5]

and 1% bovine serum albumin fraction V ((Fisher Scientific, Pittsburg, PA), 30 min). The slides

were washed twice in 1X PBS (5 min) prior to secondary antibody labeling. The cytospin

culture material was subsequently incubated with a highly cross-adsorbed goat polyclonal anti-

rabbit-IgG antibody-Alexa Fluor 568 conjugate (A11036 (Invitrogen Molecular Probes,

Carlsbad, CA), 2 µg/mL in 0.05 M Tris-HCl [pH 7.5] and 1% bovine serum albumin fraction V,

30 min). Afterward, the slides were washed twice in 1X PBS (5 min), air-dried, coverslip-

mounted using VECTASHIELD HardSet Mounting Medium with DAPI, and examined as above

described.

22

Results

In Situ Detection of Anaplasma spp. by DNA Target-Primed Rolling-Circle Amplification of Padlock Probes

During persistent, latent infections that can be caused by Anaplasma spp., it would be

valuable to sensitively and specifically detect organisms in infected tissues using an isothermal

DNA amplification technique that has the potential to provide information about organism

genotype and host-cellular localization. To determine whether such a technique could be used to

detect A. phagocytophilum and A. marginale, in situ DNA target-primed rolling-circle

amplification of padlock probes was developed based upon previous in situ genotyping of human

mitochondrial DNA using rolling-circle amplification of padlock probes (68).

Cytospin preparations of uninfected or A. phagocytophilum HZ-infected human

myeloblastic leukemia cell cultures and a padlock probe, either a nonspecific probe or an A.

phagocytophilum-specific probe, were used for in situ DNA target-primed rolling-circle

amplification. When A. phagocytophilum HZ-infected culture cytospins were microscopically

examined after in situ DNA target-primed rolling-circle amplification of an A. phagocytophilum-

specific padlock probe, numerous aggregates of stippled, green or red fluorescence, which

represented the fluorescently labeled, localized amplification product, were frequently identified

within intact cultured myeloblastic leukemia cells. The intracellular location of fluorescence

correlated well with microscopic observations of A. phagocytophilum morulae (mulberry-like

aggregates of bacteria) within intact myeloblastic leukemia cells in a Wright’s-Giemsa-stained

infected culture cytospin (Figure 2-2 A, B, and D). Similar fluorescence was not observed when

either an uninfected culture cytospin (not shown) or a nonspecific padlock probe was used

(Figure 2-2 C and E). The results were confirmed in five independent experiments.

23

Cytospin preparations of uninfected or A. marginale Virginia-infected fetal rhesus monkey

endothelial cell cultures and a padlock probe, either a nonspecific probe or an A. marginale-

specific probe, were used for in situ DNA target-primed rolling-circle amplification. When A.

marginale Virginia-infected culture cytospins were microscopically examined after in situ DNA

target-primed rolling-circle amplification of an A. marginale-specific padlock probe, round

aggregates of stippled, green or red fluorescence, which represented the fluorescently labeled,

localized amplification product, were frequently identified within intact cultured endothelial

cells. The intracellular location of fluorescence correlated well with microscopic observations of

A. marginale morulae within intact endothelial cells in a Wright’s-Giemsa-stained infected

culture cytospin (Figure 2-3 A, B, and D). Similar fluorescence was not observed when either an

uninfected culture cytospin (not shown) or a nonspecific padlock probe was used (Figure 2-3 C

and E). The results were confirmed in three independent experiments.

In situ Intracellular Colocalization of the A. phagocytophilum Rolling-Circle Amplification Product and von Willebrand Factor Immunofluorescence

Since A. phagocytophilum can be continuously cultivated in endothelial cells (81) and

preliminary immunofluorescent staining of SCID mouse tissues suggests endothelial cells may

also be infected in vivo (54), there has been heightened interested in more conclusively

determining whether endothelial cells are an in vivo nidus of A. phagocytophilum in naturally or

experimentally infected, immunocompetent mammals. To that end, in situ A. phagocytophilum

DNA target-primed rolling-circle amplification of a padlock probe was combined with indirect

immunofluorescent staining of von Willebrand factor, which is present within Weibel-Palade

bodies of endothelial cells (115). Cytospin preparations of uninfected or A. phagocytophilum

NY18-infected fetal rhesus monkey endothelial cell cultures, two padlock probes (a nonspecific

probe or an A. phagocytophilum-specific probe), and two antibody staining variations (secondary

24

antibody only or primary and secondary antibody) were examined using these combined

techniques.

When A. phagocytophilum NY18-infected culture cytospins were microscopically

examined after in situ DNA target-primed rolling-circle amplification of an A. phagocytophilum-

specific padlock probe, focal aggregates of stippled, green fluorescence, which represented the

fluorescently labeled, localized amplification product, were often identified perinuclearly within

intact cultured endothelial cells. The intracellular location of green fluorescence correlated well

with microscopic observations of A. phagocytophilum morulae within intact endothelial cells in a

Wright’s-Giemsa-stained infected culture cytospin (Figure 2-4 A and B). Similar green

fluorescence was not observed when either an uninfected culture cytospin (not shown) or a

nonspecific padlock probe was used (Figure 2-4 C). The results were confirmed in five

independent experiments.

When A. phagocytophilum-infected culture cytospins were examined microscopically after

combined in situ DNA target-primed rolling-circle amplification of an A. phagocytophilum-

specific padlock probe and indirect immunofluorescent staining of von Willebrand factor, focal

aggregates of stippled, green fluorescence were identified perinuclearly juxtaposed with focal

areas of red fluorescence, which represented the fluorescently labeled von Willebrand factor

within Weibel-Palade bodies (Figure 2-4 D). Similar green and red fluorescence were not

observed when the combined procedures were performed using a nonspecific padlock probe and

only the fluorescently labeled secondary antibody (Figure 2-4 E).

Discussion

The in situ DNA target-primed rolling-circle amplification of a padlock probe technique

adapted here for Anaplasma detection was first described as a method to distinguish single

nucleotide polymorphisms in human mitochondrial DNA (68). We considered it likely that this

25

technique could be used for in situ detection of intracellular microorganisms. Here it is

demonstrated that it is possible to detect and localize Anaplasma spp. within intact cultured

mammalian cells using in situ DNA target-primed rolling-circle amplification of padlock probes

and that this technique can be combined with immunofluorescent staining to identify A.

phagocytophilum and an endothelial cell antigen within a single cultured endothelial cell. These

observations suggest that this technique could be applied to natural Anaplasma spp. isolates in

tissues obtained from naturally or experimentally infected mammalian or arthropod hosts to

determine the cellular localization of potentially cryptic infections in vivo and to provide

nucleotide sequence information in situ.

The specificity of this technique arises from use of a padlock probe, which depends upon

specific hybridization to the genomic DNA target sequence. The use of ligase allows

interrogation of the hybridization quality between the target-specific-sequence of the padlock

probe and the genomic target. If the quality of the hybridization is not optimal, ligation of the

padlock probe to form a closed, partially double-stranded circle involving the genomic target will

not occur. The use of ligase in this procedure is the basis for distinction of genomic single

nucleotide polymorphisms based on the quality of hybridization with the 3’ terminus of the

padlock probe (65, 66, 68, 84, 85, 86). The exponential amplification of the bound padlock

probe is one basis for the sensitivity of this technique (68, 86). The amplification product

remains bound to the target sequence and is subsequently detected by the addition of a second

fluorescently labeled oligonucleotide tag that specifically binds repeated linker sequence within

the concatameric amplification product. The fact that the amplification product remains tightly

bound as a 3’ extension of the genomic target limits background fluorescence, which also

26

contributes to the sensitivity and specificity of this technique and is another improvement over

other in situ nucleotide detection techniques (68, 85).

An additional benefit of the in situ DNA amplification technique applied here over other

methods of in situ DNA detection is that the reactions are isothermal, which preserves tissue

architecture and allows subsequent immunostaining of host cell antigens, such as the endothelial

cell antigen, von Willebrand factor. This is of particular import given the recently heightened

interest in endothelial cells as a potential natural mammalian infection nidus in light of the ability

to continuously cultivate A. phagocytophilum and A. marginale in fetal primate endothelial cells

(81) and preliminary immunofluorescent colocalization of A. phagocytophilum or A. marginale

with endothelial cell antigens in vivo (24, 54).

Transformation of A. phagocytophilum rendering it able to express green fluorescent

protein and subsequent experimental infection of mice with A. phagocytophilum transformants

have been recently described (40). Such transformants may facilitate temporal tracking of

organism tissue distribution during experimental infection, including host cellular binding, entry,

and intracellular development; however, their use is dependent upon an experimental model of

mammalian infection and the ability to maintain stable transformants in long-term culture and

during the full course of acute and chronic infection. The fluorescence microscopy-based

Anaplasma spp. detection method described here does not depend upon a genetic transformant

that may be unstable or attenuated and could be applied to naturally or experimentally infected

mammalian or arthropod host tissues to detect potentially cryptic sites of infection. Additionally,

with knowledge of variable Anaplasma genomic DNA sequences that may potentially be present

in host tissues during an infection, multiple padlock probes that bind to differentially labeled

27

fluorescent oligonucleotide tags could be designed and used in a single isothermal reaction to

provide information about Anaplasma nucleotide sequence in situ.

In summary, the observations presented here show that in situ DNA target-primed rolling-

circle amplification of padlock probes can be used to detect Anaplasma spp. within intact

cultured mammalian cells. This demonstrates that the amplification technique can be used for in

situ detection of an intracellular microorganism. Also, this technique can be combined with

immunofluorescent staining in order to identify a host cell antigen and the fluorescently labeled

rolling-circle amplification product within a single cell. This work forms the foundation for

future applications of this technique to detect, localize, and analyze Anaplasma nucleotide

sequences in the tissues of infected hosts and in cell cultures. There is also the potential to apply

the technique described here to investigate prospective cryptic host-cellular localization of other

microorganisms and further elucidate the etiopathogenesis of disease associated with infection

by other obligate or facultative intracellular microorganisms.

28

Figure 2-1. Padlock probes and fluorescent oligonucleotides. Three padlock probes were used, AB1251, A. phagocytophilum-specific, AB1270, A. marginale-specific, or AB1253, nonspecific. The genomic DNA target-specific-sequences of AB1251 and AB1270 are within the underlined 5’ and 3’ arms of the padlock probes. When the single-stranded genomic complement of the probe’s target-specific-sequence is detected, the probe hybridizes as a nicked circle, which is subsequently locked in place by ligase as a partially double helical, closed circle. The circularized padlock probe is the template for in situ DNA target-primed rolling-circle amplification. The 5’ and 3’ arms of the nonspecific probe, AB1253, contained the same nucleotide composition as AB1251; however, the sequence of the nucleotides was randomized within the underlined regions. The 5’ and 3’ arms of the padlock probes were joined by an identical intervening linker region. The amplification product of the italicized portion of the probe linker region is the complement of the fluorescently-labeled oligonucleotide tags, AB1252 or AB1279 (also known as Lin 33 [68]). All oligonucleotides contained a 5’ modification designated as P, which stands for phosphate, FAM, which stands for phosphoramidite coupled fluorescein, or Alexa 555, which stands for Alexa Fluor® 555 (Invitrogen, Carlsbad, CA).

29

Figure 2-2. Detection of A. phagocytophilum using in situ DNA target-primed rolling-circle amplification of a padlock probe. A) Wright-Giemsa-stained cytospin preparation 63X objective, bright field, B) in situ rolling-circle amplification product within cytospin preparation 63X objective, differential interference contrast with blue and green emissions, C) negative control for reaction shown in B in situ rolling-circle amplification product within cytospin preparation 63X objective, differential interference contrast with blue and green emissions, D) in situ rolling-circle amplification product within cytospin preparation 63X objective, differential interference contrast with blue and red emissions, E) negative control for reaction shown in D in situ rolling-circle amplification product within cytospin preparation 63X objective, differential interference contrast with blue and red emissions. Numerous intracellular morulae were observed within heavily A. phagocytophilum HZ-infected, intact human myeloblasts in cytospin preparations (A). Morulae were also detected as the stippled, green fluorescent product or red fluorescent product of in situ DNA target-primed rolling-circle amplification of an A. phagocytophilum-specific padlock probe (B and D). Similar fluorescence was not detected when a nonspecific padlock probe was used in the otherwise identical procedure (C and E).

30

Figure 2-3. Detection of A. marginale using in situ DNA target-primed rolling-circle amplification of a padlock probe. A) Wright-Giemsa-stained cytospin preparation 63X objective, bright field, B) in situ rolling-circle amplification product within cytospin preparation 63X objective, differential interference contrast with blue and green emissions, C) negative control for reaction shown in B in situ rolling-circle amplification product within cytospin preparation 63X objective, differential interference contrast with blue and green emissions, D) in situ rolling-circle amplification product within cytospin preparation 63X objective, differential interference contrast with blue and red emissions, E) negative control for reaction shown in D in situ rolling-circle amplification product within cytospin preparation 63X objective, differential interference contrast with blue and red emissions. Intracellular morulae were observed within A. marginale-infected, intact fetal rhesus monkey endothelial cells in cytospin preparations (A). Morulae were also detected as the stippled, green fluorescent product or red fluorescent product of in situ DNA target-primed rolling-circle amplification of an A. marginale-specific padlock probe (B and D). Similar fluorescence was not detected when a nonspecific padlock probe was used in the otherwise identical procedure (C and E).

31

Figure 2-4. Detection of A. phagocytophilum using in situ DNA target-primed rolling-circle

amplification of a padlock probe and concurrent indirect immunofluorescent staining of von Willebrand Factor. A) Wright-Giemsa-stained cytospin preparation 100X objective, bright field, B) in situ rolling-circle amplification product within cytospin preparation 63X objective, differential interference contrast with blue and green emissions, C) negative control for reaction shown in B in situ rolling-circle amplification product within cytospin preparation 40X objective, phase contrast with blue and green emissions, D) in situ rolling-circle amplification product and indirect immunofluorescent labeling of von Willebrand factor within cytospin preparation 40X objective, blue, red, and green emissions, E) negative control for reaction shown in D in situ rolling-circle amplification product and secondary antibody labeling only within cytospin preparation 40X objective, blue, red, and green emissions. A few perinuclear morulae were observed within intact A. phagocytophilum NY18-infected fetal rhesus monkey endothelial cells in cytospin preparations (A). Morulae were also detected as the stippled, green fluorescent product of in situ DNA target-primed rolling-circle amplification of an A. phagocytophilum-specific padlock probe (B). A similar green fluorescent product was not detected when a nonspecific padlock probe was used in the otherwise identical procedure (C). A. phagocytophilum-infected fetal rhesus monkey endothelial cells in cytospin preparations were subjected to combined in situ DNA target-primed rolling-circle amplification of an A. phagocytophilum-specific padlock probe and indirect immunofluorescent staining of von Willebrand factor. Focal aggregates of stippled, green fluorescence were identified perinuclearly and juxtaposed with focal areas of red fluorescence, which represented the fluorescently labeled von Willebrand factor (D). Similar green and red fluorescence were not observed when the combined procedure was performed using a nonspecific padlock probe and only the fluorescently labeled secondary antibody (E).

32

CHAPTER 3 EXPERIMENTAL INOCULATION OF DOGS WITH ANAPLASMA PHAGOCYTOPHILUM, MOLECULAR EVIDENCE OF PERSISTENT INFECTION FOLLOWING DOXYCYCLINE

THERAPY, AND INVESTIGATION OF ENDOTHELIAL CELLS AS SOURCE OF INFECTIOUS INOCULUM AND A REPOSITORY OF CHRONIC INFECTION IN DOGS

Introduction

A. phagocytophilum is the third most common tick-borne infection of humans in the

United States (8, 34). When diagnosed early, most patients exhibit rapid clinical improvement

within one to two days of treatment; however, protracted disease course or death is possible

among those who are heavily infected, elderly, or otherwise immunocompromised (8, 25, 34).

Doxycycline, which is bacteriostatic against A. phagocytophilum, is considered the optimal

treatment choice for most cases of A. phagocytophilum infection (1, 6, 34, 50, 79). Though, it is

uncertain whether currently recommended treatment regimes merely alleviate clinical signs and

potentially induce a carrier state of infection since controlled studies investigating the optimal

duration of treatment or the dosage required to completely clear viable A. phagocytophilum

organisms from infected individuals have not been reported.

Chronic A. phagocytophilum infection has been described in several animals: dogs (36,

37), cats (67), rodents (55, 108), horses (42, 89, 97), lambs (107), and sheep (41). Chronic

infection is suspected in some human cases (30, 31, 95, 96). Also, there is molecular evidence

for persistent infection by A. phagocytophilum and related organisms after treatment with

doxycycline. A. phagocytophilum DNA in domestic cats (67), E. chaffeensis DNA in dogs (21),

and E. canis DNA and viable organisms in dogs (51, 57, 102, 116) have been detected in blood

or organs despite doxycycline administration at a dose and duration generally considered

effective against Ehrlichiae.

Previously described experimental infection of dogs with A. phagocytophilum have

depended upon use of infected blood from a naturally occurring clinical case or cultivation of

33

organisms in human (Homo sapiens) HL-60 myeloblastic leukemia cells (American Type

Culture Collection, ATCC CCL-240™, Manassas, VA) or autologous neutrophils (70, 76, 105).

These culture systems rely upon mammalian cells of myeloid origin which are known to harbor

A. phagocytophilum infection in vivo. Determining whether or not endothelial cells can serve as

a source of viable inoculum capable of establishing infection in dogs represents an initial step in

the investigation of whether or not endothelial cells may be a cell that is involved in the

mammalian life cycle stages of this microorganism. Also, should endothelial cell-derived

inoculum prove capable of establishing mammalian infection, this would have important

implications upon future vaccine development. As opposed to previously used in vitro cell lines,

which are loosely adherent or non-adherent, endothelial cells are a tightly adherent cell line. The

tightly adherent nature of cultured endothelial cells facilitates their use in genetic manipulation

of Anaplasma and clonal selection by plaque purification (81).

The purposes of this investigation were to determine if experimental inoculation of

different A. phagocytophilum isolates could establish chronic canine infection, to observe

whether or not infection persisted in spite of antimicrobial treatment, and to discover whether or

not cultured endothelial cells could be a source of infectious inoculum in dogs and whether or

not infection of endothelial cells occurred in vivo in chronically infected dogs.

Materials and Methods

Animals, Inocula, and Monitoring for Development of Infection

Four adult, intact male, Sprague Dawley Beagles, confirmed by polymerase chain reaction

(PCR) or serology (immunofluorescent antibody testing or SNAP®4Dx®, IDEXX, Westbrook,

ME) to be negative for common arthropod-borne diseases (A. phagocytophilum, A. platys, E.

canis, E. chaffeensis, B. burgdorferi, Babesia gibsoni, Bartonella henselae, Dirofilaria immitis)

were inoculated intravenously with A. phagocytophilum from one of two sources. Two dogs

34

were inoculated with a human isolate, NY18 strain of A. phagocytophilum, cultivated in fetal

rhesus monkey endothelial cells; the other two dogs were inoculated with a canine isolate via

injection of stabilate prepared from parasitemic blood of a naturally infected dog from

Minnesota.

The NY18 isolate of A. phagocytophilum was cultivated in fetal rhesus monkey (Macaca

mulatta) RF/6A endothelial cells (American Type Culture Collection, ATCC CRL-1780™,

Manassas, VA) (81) (DMEM medium (HyClone, Logan, UT) supplemented with 10% heat-

inactivated fetal bovine serum (HyClone, Logan, UT), 0.25% NaHCO3 (Sigma, St. Louis, MO),

25 mM HEPES (Sigma, St. Louis, MO), [pH 7.5], 37°C, 5% CO2). Two 25 mL infected-tissue

culture flasks were propagated (one for each dog). When at least 35% of the cells in each flask

were A. phagocytophilum-infected as determined by light microscopic examination of cytospin

culture material on microscope slides, the entire cultures were harvested by scraping the bottom

of the flasks using a sterile cell scraper (Fisher Scientific, Pittsburg, PA) to form a cell

suspension within the culture medium. The cell suspension from each flask was transferred into

individual tubes and was centrifuged at 300 x g for 10 min at 25°C. The supernatant medium

was removed from each tube and used to inoculate other ongoing cell cultures. Each of the

resultant cellular pellets was individually washed in 5 mL of fresh, sterile 1X Hanks Balanced

Salt Solution (HBSS) (20-021-CV, Mediatech, Inc., Manassas, VA) and centrifuged at 300 x g

for 10 min at 25°C twice. After the second wash, each cellular pellet was resuspended in 5 mL

of fresh HBSS giving a final concentration of 2.5 x 105 cells/mL. Four milliliters of each

infected cell suspension (containing both intracellular and free endothelial cell-derived A.

phagocytophilum) was used to intravenously inoculate a dog (2 dogs total, 4 mL per dog); 1 mL

of each infected cell suspension was stored at -80°C. Four milliliters of parasitemic blood from a

35

naturally infected dog was collected in ethylenediaminetetraacetic acid (EDTA) and a stabilate

was prepared by adding dimethyl sulfoxide (DMSO) to a final concentration of 10%. Two

milliliters of infected blood stabilate was used to intravenously inoculate 2 different dogs (1 mL

per dog).

After inoculation, the dogs were examined and phlebotomized at regular intervals for the

duration of the trial. Blood was collected into both EDTA-containing and plain sterile tubes,

from which serum was harvested. Rectal temperature, gait, and diarthrodial joints were

monitored for abnormalities consistent with anaplasmosis, such as pyrexia and polyarthropathy.

Complete blood cell counts were monitored for the development of thrombocytopenia or

leukopenia, and serum biochemistry profiles were monitored for elevations in hepatocellular

leakage enzyme activities. Wright’s Giemsa-stained (Harleco, EM Science, Gibbstown, NJ)

peripheral blood films were examined by light microscopy for the occurrence of parasitemia as

evidenced by the presence of basophilic, stippled, circular granulocytic inclusions consistent

with intragranulocytic A. phagocytophilum morulae.

Serologic Assays for A. phagocytophilum Infection

To document the occurrence of serum antibodies directed against A. phagocytophilum,

polyclonal sera were tested using a commercially available sandwich ELISA based upon the

MSP2 (p44) protein (SNAP®4Dx®, IDEXX, Westbrook, ME) and were reacted with

immunoblots of sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)

gradient gels containing separated, denatured proteins of whole A. phagocytophilum derived

from RF/6A endothelial cell culture (81), as described (2). Pre- and post-infection polyclonal

sera from the four A. phagocytophilum-inoculated dogs and from one PCR-positive, naturally A.

phagocytophilum-infected human were used. The canine sera were diluted 1:2000; the human

sera were diluted 1:5000. Horseradish peroxidase-conjugated, species-specific secondary

36

antibodies directed against canine or human IgG were diluted 1:75,000 (Sigma, St. Louis, MO).

Chemiluminescence was detected using SuperSignal West Dura Extended Duration Substrate

(Pierce, Rockford, IL).

PCR Detection of A. phagocytophilum DNA in Blood and Tissues

DNA was extracted from anticoagulated, EDTA-whole blood and plain-frozen post-

mortem tissue samples using the QIAamp DNA Mini Kit (Qiagen, Valencia, CA) in accord with

the manufacturer instructions, which indicated the following amounts of biologic material be

used for DNA extraction: 200 µL of anticoagulated whole blood, 25 mg of tissues other than

spleen, or 10 mg of spleen. Control DNA samples, including positive control DNA extracted

from A. phagocytophilum grown in fetal rhesus (Macaca mulatta) RF/6A endothelial cells (81)

and negative control DNA extracted from an uninfected dog were similarly extracted.

Extracted DNA was used as the template for nested-PCR amplification of a conserved

region of msp2. Increased sensitivity for detection of A. phagocytophilum beyond that which is

expected from nested-PCR alone was derived from use of synthetic oligonucleotide primers

(Figure 3-1) (MWG Biotech, High Point, NC) that were designed to detect nearly all genomic

copies of this conserved region within msp2, including copies that are present in the single

expression site and in most of the approximately 100 pseudogenes.

Fifty microliter DNA amplification reactions were run in a GeneAmp PCR System 9600

(Perkin Elmer, Waltham, MA) using 0.05 U/µL AmpliTaq® DNA Polymerase (Applied

Biosystems Inc, Foster City, CA) in 1X supplied enzyme buffer plus 0.2 mM of each dNTP

(Applied Biosystems Inc, Foster City, CA) and 0.2 µM of each primer (forward and reverse) for

the reaction (primary or nested) (Figure 3-1). The following thermocycler conditions were used:

hot start; 3 cycles of 97°C for 15 sec denaturation, 50°C for 30 sec annealing, 72°C for 30 sec

polymerization; 47 cycles of 94°C for 15 sec denaturation, 50°C for 30 sec annealing, 72°C for

37

30 sec polymerization; 72°C for 3 min final extension; hold at 4°C. Two and a half microliters

of the primary PCR product was used as the DNA template in the subsequent nested-PCR. Use

of the nested primers was expected to yield a 464 bp final amplification product. The identity of

the amplification products was confirmed by fluoresceinated-DNA oligonucleotide (Figure 3-1,

AB1214) probing of southern blotted PCR products that were separated on a 2% agarose gel, as

described (9).

Real-time PCR detection of two A. phagocytophilum genes, msp2 and the heat shock

protein, groEL, in anticoagulated, EDTA-whole blood samples collected from the A.

phagocytophilum (NY18)-infected dogs was performed using a test offered by a commercial

laboratory (IDEXX, Westbrook, ME).

Immunosuppression, Antibiotic Treatment, and Euthanasia

In order to see if post-infection parasitemia or clinical signs could be enhanced, all four

dogs were administered an immunosuppressive dose of prednisone for approximately two weeks.

After the period of immunosuppression, all four dogs were treated with doxycycline to determine

if parasitemia or clinical signs could be abrogated. After the antibiotic treatment, all four dogs

were again administered an immunosuppressive dose of prednisone for approximately two

weeks. The NY18 isolate-infected dogs were euthanized 11.5 months post-infection; the canine

isolate-infected dogs were euthanized 7.8 months post-infection (Table 3-1).

The efficacy of two dosage regimes was investigated. The NY18 isolate-infected dogs

were treated with doxycycline 10 mg/kg orally once daily for two weeks (21); the two canine

isolate-infected dogs were treated with doxycycline at the currently recommended dose (E. B.

Breitschwerdt, personal communication) of 10 mg/kg orally twice daily for four weeks (21).

38

In Situ DNA Target-Primed Rolling-Circle Amplification of a Padlock Probe for Detection of Anaplasma phagocytophilum

Post-mortem tissues from the canine isolate-infected dogs were cryopreserved

immediately after collection. Tissue samples were initially stored and transported in sterile

tissue culture medium (HyClone, Logan, UT) on ice. Subsequently, samples were trimmed and

cryoembedded in Neg-50 medium (Thermo Scientific Richard-Allan, Pittsburg, PA) by freezing

in 2-methylbutane on liquid nitrogen. Frozen tissue blocks were stored at -80°C until

cryosectioned using a Leica cryostat (Leica Microsystems, Bannockburn, IL). Four micrometer

tissue sections were prepared at -20°C and immediately fixed in acetone on ice for 5 min.

Sections were stored at -80°C until examined using in situ DNA target-primed rolling-circle

amplification of padlock probes for detection of A. phagocytophilum modified from the

described (113) to use sheared calf thymus DNA (R&D Systems, Minneapolis, MN) as the DNA

carrier in the oligonucleotide hybridization reactions. Sheared calf thymus DNA was used at 0.5

µg/µL in the padlock probe hybridization solution and was used at 6.0 µg/µL in the fluorescently

labeled oligonucleotide hybridization solution. Cytospins of A. phagocytophilum HZ isolate

cultivated in human (Homo sapiens) HL-60 myeloblastic leukemia cells (American Type Culture

Collection, ATCC CCL-240™, Manassas, VA), as described (49), were also run with an A.

phagocytophilum-specific padlock probe as a reagent positive control.

Results

Physical Exam, Hematopathology, and Clinical Chemistry

Hematologic abnormalities, clinical chemistry abnormalities, or circulating

intragranulocytic morulae were not observed during the two canine infection trials. The physical

exam findings were unremarkable, except that one of the canine isolate-infected dogs was

intermittently pyrexic (body temperature >102.5°F) (Figure 3-2). There was an average of 26.1

39

days (median 26.0 days) between instances or episodes (more than one consecutive

measurement) of body temperatures >102.5°F. The minimum number of days between febrile

episodes was 8, and the maximum was 49 days.

Serologic Evidence of Infection

In the NY18-infected dogs, seroconversion to A. phagocytophilum-positive status was first

detected about one week after infection using the MSP2 (p44) antigen-based IDEXX

SNAP®4Dx® (Table 3-2). Subsequent samples that were collected on day 11 through day 316

post-infection remained seropositive (25 samples per dog tested). In the canine isolate-infected

dogs, seroconversion to A. phagocytophilum-positive status was not detected until day 51 (dog 1)

or day 59 (dog 2) after infection. Subsequent samples that were collected on day 71 through day

210 post-infection were seronegative by this assay (27 samples per dog tested).

The seropositive status of the canine isolate-infected dogs was confirmed based upon

immuonoreactivity detected when 37 day post-infection polyclonal sera was reacted with an

immunoblot containing separated, denatured proteins of A. phagocytophilum (NY18) (Figure 3-

3). Serum from this day was selected based on knowledge of the temporally associated changes

in immuonoreactivity detected by competitive-enzyme-linked immunosorbent assay (cELISA) in

the NY18-infected dogs (4).

Molecular Evidence of Infection

NY18-infected dogs

In the NY18-infected dogs, A. phagocytophilum DNA was readily detected by at least one

of three assays (nested-PCR amplification of msp2 or real-time PCR amplification of msp2 or

groEL) for about 1 month after initial infection (Table 3-3). Subsequently, A. phagocytophilum

DNA was rarely detected between day 36 and day 230 post-infection (3 out of 42 total tested

samples were positive); PCR detection improved after immunosuppression. By day 15 post-

40

immunosuppression (day 245 post-infection), both dogs were PCR-positive using nested- or real-

time, msp2-based PCR.

Doxycycline treatment (10 mg/kg by mouth once daily for 2 weeks) did not eliminate

molecular evidence of A. phagocytophilum infection in the NY18-infected dogs. A.

phagocytophilum DNA was still detectable in samples collected 5.4 weeks after treatment with

doxycycline had ended (day 299 post-infection). Eight weeks after doxycycline therapy

concluded, immunosuppression was performed a second time (days 316 through 325 post-

infection). Subsequently, samples from the re-immunosuppressed dogs, which were collected

for 11 weeks after doxycycline therapy, remained PCR positive.

After the predicted time for prednisone to be eliminated had passed, the NY18-infected

dogs were euthanized on day 344 post-infection (11.9 weeks after the last doxycycline dose, 2.7

weeks after the last prednisone dose). Heart, lung, liver, spleen, and bone marrow were

subjected to nested-PCR; the heart and spleen of Dog 1 were nested-PCR-positive (Figure 3-4).

Canine isolate-infected dogs

In the canine isolate-infected dogs, A. phagocytophilum DNA was detected by nested-PCR

amplification of msp2 intermittently during the initial period of infection, prior to

immunosuppression (Dog 1: days 16, 20, 39, and 72 post-infection; Dog 2: days 2, 48, and 62

post-infection). During the first period of immunosuppression (day 143 through 156 post-

infection) (Table 3-1), both canine isolate-infected dogs were nested-PCR-positive within 6 days

of initiating prednisone administration (Dog 1: days 149 and 153 post-infection; Dog 2: days

149 and 151 post-infection).

Doxycycline treatment (10 mg/kg by mouth twice daily for 4 weeks) did not eliminate

molecular evidence of A. phagocytophilum infection in the canine isolate-infected dogs. Prior to

the second period of immunosuppression, A. phagocytophilum DNA was detected by nested-

41

PCR in one of the dogs 3 days after doxycycline administration ended (Dog 1, day 187 post-

infection).

One week after concluding doxycycline therapy, immunosuppression was performed a

second time (day 191 through 204 post-infection). Subsequently, one blood sample (Dog 2, day

213 post-infection) and one post-mortem tissue sample (Dog 1 right kidney, day 233 post-

infection) were nested-PCR-positive (Figure 3-5).

In Situ Detection of A. phagocytophilum

When tissues from the canine isolate-infected dogs (Dog 1: right kidney (Figure 3-6) and

liver; Dog 2: lung) were microscopically examined after in situ DNA target-primed rolling-

circle amplification, multifocal areas of autofluorescence were identified regardless of whether

the A.phagocytophilum-specific padlock probe or the nonspecific padlock probe was used.

Fluorescently labeled, in situ rolling-circle amplification products were not identified in any of

the tissues examined.

Discussion

Regardless of the two isolates tested (human or canine) or the two doxycycline dosages

tested, A. phagocytophilum DNA remained detectable for several months in the peripheral blood

and some post-mortem tissues (heart, spleen, kidney) from these four dogs. This is the first

molecular evidence of chronic, persistent A. phagocytophilum infection in blood and tissues of

subclinical dogs despite doxycycline treatment using the currently recommended dosage. This

has critical implications since treatment with doxycycline may not eliminate A. phagocytophilum

human or canine isolate infection from dogs. Dogs may become life-long A. phagocytophilum-

carriers, at risk for recrudescence with concurrent illness or immunosuppression (30, 95). Also,

chronic carriers may act as an environmental source of organisms to ixodid ticks during

42

acquisition-feeding. Future trials investigating different doxycycline dosages with precise

quantitation of plasma drug concentration may be warranted.

There were isolate-dependent differences between the two A. phagocytophilum-infection

trials. The NY18-infected dogs exhibited greater immuonoreactivity in the IDEXX SNAP®4Dx®

test and on immunoblots compared to the canine isolate-infected dogs. In the case of the

immunoblots, this difference may have been due to the antigen used in the immunoblot, A.

phagocytophilum NY18. It is expected that the NY18-infected dogs would demonstrate a more

robust immune response against this antigen than the canine isolate-infected dogs. However, this

explanation cannot be used to explain the difference in IDEXX SNAP®4Dx® immuonoreactivity

since the test uses a highly conserved region of MSP2 (p44) as the antigen. During the course of

the trials, DNA was also more frequently detected in the blood of the NY18-infected dogs than

in the blood of the canine isolate-infected dogs. In addition to a potential inherent variation

between the isolates (e.g., degree of host-adaptation), the difference in IDEXX SNAP®4Dx®

immuonoreactivity and DNA detection between the two infection trials may reflect a difference

in the dose of organisms in the inoculum, 4 mL of tissue culture containing approximately

350,000 infected cells at high multiplicity of infection vs. 1 mL of blood stabilate from a

parasitemic dog. It is also of note that one of the canine isolate-infected dogs exhibited cyclical

periodicity in pyrexia episodes. This may have correlated with the emergence of new MSP2

variants (11), but this possibility was not investigated in these dogs.

The A. phagocytophilum NY18-infection trial reported here represents the first infection of

dogs using cultured endothelial cells as the source of inoculum. This is an important proof of

concept as progress is made toward potential vaccines based upon genetically modified A.

43

phagocytophilum (40) cultured in endothelial cells, which are a more tractable source of

organisms compared to other in vitro systems (81).

The A. phagocytophilum canine isolate-infection trial reported here is the first investigation

of endothelial cells as a potential in vivo source of A. phagocytophilum during chronic canine

infection. In SCID mice, dual indirect immunofluorescent colocalization of an endothelial cell

antigen and an A. phagocytophilum surface protein within the cardiac and hepatic

microvasculature after 7 weeks of infection has been described; however, photomicrographs and

rigorous controls were not published in the report (54). Unequivocal evidence of in vivo

endothelial cell infection in examined tissues (kidney, liver, lung) from chronically infected dogs

was not identified using the in situ rolling-circle amplification technique presented here. There

are a few potential explanations for why this might be so. First, endothelial cells may, in fact,

not be involved in the in vivo A. phagocytophilum life cycle. Second, endothelial cells may be

involved in the in vivo A. phagocytophilum life cycle, but at a level that is below the detection

limit of the in situ rolling-circle amplification technique described here. Third, since positive

control tissue sections were not available (cytospins of culture material were used for this

purpose), the in situ rolling-circle technique described here may not have been properly

optimized for detection of prokaryotic cells in frozen canine tissue sections. Future examination

of A. phagocytophilum infection of endothelial cells in vivo using other techniques (e.g.,

monoclonal-based immunofluorescence, fluorescence in situ hybridization (FISH), or in situ

PCR) an