of October 21, 2018.This information is current as

Program in Human Neutrophils Fails to Induce an Apoptosis Differentiation

Anaplasma phagocytophilumMechanisms: Insights into Pathogen Immune Evasion

Jovanka M. Voyich, Cynthia M. Argue and Frank R. DeLeoDori L. Borjesson, Scott D. Kobayashi, Adeline R. Whitney,

http://www.jimmunol.org/content/174/10/6364doi: 10.4049/jimmunol.174.10.6364

2005; 174:6364-6372; ;J Immunol 

MaterialSupplementary http://www.jimmunol.org/content/suppl/2005/05/03/174.10.6364.DC1

Referenceshttp://www.jimmunol.org/content/174/10/6364.full#ref-list-1

, 20 of which you can access for free at: cites 28 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 2005 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on October 21, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on October 21, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Insights into Pathogen Immune Evasion Mechanisms:Anaplasma phagocytophilum Fails to Induce an ApoptosisDifferentiation Program in Human Neutrophils1

Dori L. Borjesson,2* Scott D. Kobayashi,† Adeline R. Whitney,† Jovanka M. Voyich,†

Cynthia M. Argue,* and Frank R. DeLeo†

Polymorphonuclear leukocytes (PMNs or neutrophils) are essential to human innate host defense. However, some bacterial patho-gens circumvent destruction by PMNs and thereby cause disease. Anaplasma phagocytophilum, the agent of human granulocyticanaplasmosis, survives within PMNs in part by altering normal host cell processes, such as production of reactive oxygen species(ROS) and apoptosis. To investigate the molecular basis of A. phagocytophilum survival within neutrophils, we used Affymetrixmicroarrays to measure global changes in human PMN gene expression following infection with A. phagocytophilum. Notably, A.phagocytophilum uptake induced fewer perturbations in host cell gene regulation compared with phagocytosis of Staphylococcusaureus. Although ingestion of A. phagocytophilum did not elicit significant PMN ROS, proinflammatory genes were graduallyup-regulated, indicating delayed PMN activation rather than loss of proinflammatory capacity normally observed during phago-cytosis-induced apoptosis. Importantly, ingestion of A. phagocytophilum failed to trigger the neutrophil apoptosis differentiationprogram that typically follows phagocytosis and ROS production. Heat-killed A. phagocytophilum caused some similar initialalterations in neutrophil gene expression and function, which included delaying normal PMN apoptosis and blocking Fas-inducedprogrammed cell death. However, at 24 h, down-regulation of PMN gene transcription may be more reliant on active infection.Taken together, these findings suggest two separate antiapoptotic processes may work concomitantly to promote bacterial sur-vival: 1) uptake of A. phagocytophilum fails to trigger the apoptosis differentiation program usually induced by bacteria, and 2)a protein or molecule on the pathogen surface can mediate an early delay in spontaneous neutrophil apoptosis. The Journal ofImmunology, 2005, 174: 6364–6372.

P olymorphonuclear leukocytes (PMNs3 or neutrophils) area first line of defense in the human innate immune re-sponse to bacterial pathogens. Most ingested bacteria are

killed by the combined effects of PMN reactive oxygen species(ROS) and cytotoxic granule components (1). However, somepathogens have evolved means to circumvent killing by PMNs andcause disease. The mechanisms used by pathogens to evade de-struction by the innate immune system are incompletelycharacterized.

Anaplasma phagocytophilum, the agent of human granulocyticanaplasmosis, is an obligate intracellular bacterium known to sur-vive within PMNs (2, 3). A. phagocytophilum enters neutrophilsprimarily through a receptor-mediated endocytic pathway (4), andthus may ultimately reside in a modified endosomal compartment(5). Studies evaluating functional alterations in A. phagocytophi-

lum-infected PMNs suggest that the pathogen delays PMN apo-ptosis (6, 7), minimizes proinflammatory cytokine release (8, 9),and inhibits and/or fails to activate ROS production (10–13).Mechanisms underlying the ability of A. phagocytophilum to in-hibit production of PMN ROS are unclear (12–15).

Recent studies indicate bacterial phagocytosis induces an apo-ptosis differentiation program in human PMNs that most likelycontributes to the resolution of infections (16, 17). The apoptosisdifferentiation program is characterized by changes in a commonset of genes triggered by uptake of bacteria or other particles thataccelerate normal PMN apoptosis (16, 17). In contrast, A. phago-cytophilum delays normal PMN apoptosis to facilitate productiveintracellular infection (6, 7). Inasmuch as induction of PMN apo-ptosis is regulated at the level of transcription (16–18), a compre-hensive view of gene expression patterns in A. phagocytophilum-infected PMNs is critical to understanding processes that permitthe pathogen to reside within PMNs. To that end, we studiedglobal gene expression in human PMNs during infection with A.phagocytophilum. Our results suggest failure of A. phagocytophi-lum to trigger the PMN apoptosis differentiation program and abil-ity of the organism to inhibit spontaneous neutrophil apoptosis aredistinct processes that contribute to pathogen survival.

Materials and MethodsMaterials

Dextran T-500 and Ficoll-PaquePLUS (1.077 g/L) were purchased fromAmersham Biosciences. Sterile water ((Irrigation, United States Pharma-copoeia (USP)) and 0.9% sodium chloride (Irrigation, USP) were obtainedfrom Baxter Healthcare. PE-conjugated mAb specific for IL-1R antagonist(IL-1RN) (clone AS17) and isotype control mAb were purchased from BDBiosciences. RPMI 1640 medium was purchased from Invitrogen Life

*Department of Veterinary Population Medicine, College of Veterinary Medicine,University of Minnesota, St. Paul, MN 55108; and †Laboratory of Human BacterialPathogenesis, Rocky Mountain Laboratories, National Institute of Allergy and Infec-tious Diseases, Hamilton, MT 59840

Received for publication August 19, 2004. Accepted for publication March 2, 2005.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.1 This work was supported in part by Grant AI-51529 (to D.L.B.) from the NationalInstitutes of Health.2 Address correspondence and reprint requests to Dr. Dori L. Borjesson, Departmentof Veterinary Population Medicine, College of Veterinary Medicine, University ofMinnesota, 1365 Gortner Avenue, St. Paul, MN 55108. E-mail address:borje002@umn.edu3 Abbreviations used in this paper: PMN, polymorphonuclear leukocyte; BAX, B cellline 2-associated X-protein; IL-1RN, IL-1R antagonist; ROS, reactive oxygen spe-cies; RPMI/H, RPMI 1640 medium buffered with 10 mM HEPES.

The Journal of Immunology

Copyright © 2005 by The American Association of Immunologists, Inc. 0022-1767/05/$02.00

by guest on October 21, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Technologies. Unless indicated, all other reagents were obtained fromSigma-Aldrich.

Neutrophil isolation

Human PMNs were isolated from heparinized venous blood of healthyindividuals with a widely used method (18). All studies with human bloodwere performed in accordance with protocols approved by the InstitutionalReview Board for Human Subjects at the University of Minnesota and theNational Institute of Allergy and Infectious Diseases. Blood was mixed 1:1with 0.9% sodium chloride (Baxter Healthcare) containing 3.0% DextranT-500 (Amersham) for 20 min at room temperature to sediment eryth-rocytes. The leukocyte-enriched supernatant was centrifuged at 670 � g for10 min and resuspended in 35 ml of 0.9% sodium chloride. Ficoll-PaquePLUS (10 ml) was pipetted carefully beneath the cell suspension andthen centrifuged at 380 � g for 25 min to separate PMNs from PBMCs.PBMCs were removed by aspiration, sides of the tubes were swabbed, anderythrocytes were lysed with water (Irrigation USP; Baxter Healthcare) for15–30 s, followed by immediate mixing with 1.7% sodium chloride.Purified PMNs were centrifuged at 380 � g, resuspended in RPMI 1640medium buffered with 10 mM HEPES, pH 7.4 (RPMI/H), and enumeratedby microscopy. Purity of PMN preparations and cell viability were rou-tinely assessed by flow cytometry (FACSCalibur; BD Biosciences) ormicroscopy. The entire procedure was performed at room temperature, andcell preparations contained �99% PMNs. Reagents used for neutrophilpurification typically contained �25.0 pg/ml endotoxin (Limulus amebo-cyte lysate assay; Fisher Scientific).

Growth of bacteria

A. phagocytophilum was isolated from a patient with human granulocyticanaplasmosis in Nantucket, Massachusetts (NCH-1 isolate) (19). Patientblood (100 �l) was inoculated into SCID mice, as described (20). One dropof mouse blood, collected �3 wk after infection, was inoculated into HL60cells (5 � 105 cells/ml), which were cultured in IMDM containing 20%heat-inactivated FBS at 37°C with 5% CO2, as described (20). Infected anduninfected HL60 cells were counted and lysed using gentle sonication (3quick pulses at power 6, Model 100 Ultrasonic Dismembrator; BransonUltrasonics). Host cell debris from each was removed by centrifugation(2200 rpm for 5 min). The supernatant containing cell-free A. phagocyto-philum or that from uninfected HL60 cells was centrifuged at 10,000 � gfor 10 min to pellet bacteria. Pellet material containing A. phagocytophilumor that from uninfected HL60 lysate (negative control) was resuspended inRPMI/H and immediately inoculated into cultures of human PMNs. Al-ternatively, A. phagocytophilum (and control lysate) was heated for 10 minat 95°C and then inoculated into neutrophil cultures. Heat treatment ofGram-positive bacteria at 95°C for 5 min has been shown to effectively killall viable pathogens (16). These conditions (except that we heated for 10min) also effectively killed A. phagocytophilum, as evidenced by their in-ability to infect or propagate within HL60 cells compared with freshlyisolated, unheated A. phagocytophilum. All negative control samples (un-stimulated PMNs) were inoculated with clarified HL60 lysate from unin-fected cells that was prepared as described above. Thus, identical methodswere used to prepare A. phagocytophilum and negative control HL60lysates.

Staphylococcus aureus strain COL was cultured at 37°C to mid-expo-nential phase of growth (OD600 � 0.75) in tryptic soy broth, as described(16). Bacteria were washed once with 1 ml of PBS and opsonized withfresh 50% human serum for 30 min at 37°C, where indicated. OpsonizedS. aureus were washed twice with PBS and chilled on ice before use.

Neutrophil uptake of A. phagocytophilum

Uptake of bacteria by human PMNs was measured with a previously de-scribed method (16, 18), but with modifications. Following isolation,PMNs were resuspended in RPMI/H at a concentration of 106 cells/ml.Glass coverslips (12 mm round; presoaked in nitric acid and flamed withethanol) were placed into the wells of a 24-well tissue culture plate. Cov-erslips were coated with 100% autologous human serum at 37°C for 1 hand then washed twice with PBS. PMNs (3 � 105 in 300 �l) were addedto each well and allowed to settle at room temperature for 15 min. Plateswere then chilled on ice. Bacteria isolated from infected HL60 cells (ratioof 1 infected HL60 cell:2 PMNs, �5–20 A. phagocytophilum per PMN)were resuspended in 1 ml of PBS containing 7.5 �g/ml Alexa Fluor 488(Molecular Probes), and incubated for 15 min at room temperature. Stainedbacteria were washed twice with 1 ml of PBS to remove unbound AlexaFluor 488 and resuspended in RPMI/H. A total of 100 �l of Alexa Fluor488-labeled A. phagocytophilum was added to each assay well, and plateswere centrifuged at 380 � g for 8 min at 4°C to synchronize ingestion ofbacteria by PMNs. Culture plates were incubated at 37°C in a humidified

incubator with 5% CO2 for up to 24 h. At the indicated times, medium wasaspirated and cells were washed with 500 �l of cold PBS and fixed with4% paraformaldehyde on ice for 30 min. Fixative was aspirated, and cellswere washed with 500 �l of PBS. Uningested bacteria were then counter-stained with anti-Alexa Fluor 488 Ab conjugated to Alexa Fluor 594 in 500�l of PBS (7.5 �g/ml final Ab concentration) (Molecular Probes) for 15min. Coverslips/Cells were washed twice in 500 �l of PBS and mountedonto microscope slides with mounting medium (Aqua Polymount; Poly-sciences). One hundred PMNs from random fields of view were evaluatedat each time point, and ingested bacteria were scored using a fluorescencemicroscope (Nikon E800 microscope; Nikon). Bacteria stained with AlexaFlour 488 (green only) were scored as ingested, and those stained withAlexa Fluor 488 and Alexa Fluor 594 were counted as uningested/surfacebound. The experiment was repeated using PMNs from three separatedonors.

Neutrophil ROS production

PMN ROS production was measured using a published fluorometricmethod (16, 21), but with modifications. Neutrophils were incubated with25 �M 2�,7�-dihydrodichlorofluorescein diacetate (Molecular Probes) for30–45 min at room temperature in RPMI/H. Subsequently, PMNs (106),bacteria (�1 � 107 S. aureus or �0.5–4 � 107 live or heat-killed A.phagocytophilum), or PMA (1–5 �g/ml) were combined in wells of a 96-well microtiter plate at 4°C. The plates were centrifuged for 5 min at 380 �g and immediately transferred to a microplate fluorometer (SpectamaxGemini; Molecular Devices). ROS production was measured for up to 120min at 37°C with excitation and emission wavelengths of 485 and 538 nm,respectively. Vmax was calculated as the highest rate of ROS productionwithin a 10-min time period using Softmax Pro version 3.1.2 (MolecularDevices) (18).

Neutrophil apoptosis

Neutrophil apoptosis was assessed by analysis of nuclear morphology (16)or with a modified TUNEL assay (Apo-BrdU; BD Biosciences), as de-scribed (15). Briefly, PMNs (2 � 106) were plated directly into 24-wellplates precoated with normal human serum and removed by aspiration atthe desired time point. Following interaction with live or heat-killed A.phagocytophilum, determination of PMN apoptosis by light microscopywas performed on cells stained with a modified Wright-Giemsa. Neutrophilapoptosis was scored by morphological assessment of nuclei in �250PMNs per condition for each experiment. Alternatively, apoptosis was de-termined by DNA fragmentation in neutrophils following exposure to bac-teria (described above) with a modified TUNEL assay, as described by themanufacturer (BD Biosciences). Samples (10,000 events each) were ana-lyzed using a FACSCalibur flow cytometer (BD Biosciences). In someexperiments, 500 ng/ml anti-Fas mAb (clone CH-11; Upstate Biotechnol-ogy) was added to PMN cultures 30 min after A. phagocytophilum.

PMN gene expression and microarray analysis

PMN phagocytosis experiments were performed, as described previously(18). In brief, PMNs (107) were combined on ice with live S. aureus (108)or with live or heat-killed A. phagocytophilum (bacteria isolated from 5 �106 infected HL60 cells for a ratio of 1 infected HL60 cell:2 PMNs, �5–20A. phagocytophilum/PMN) in wells of a 12-well tissue culture plate (pre-coated with 20% autologous normal human serum). Unstimulated controlassays received either buffer (for S. aureus comparisons) or clarified HL60lysate prepared as described above (for A. phagocytophilum comparisons).Plates were centrifuged at 350 � g for 8 min at 4°C to synchronize uptakeof bacteria and then incubated at 37°C with 5% CO2 for up to 24 h. Se-lection of the times at which gene expression was measured in A. phago-cytophilum-infected neutrophils (1.5, 3, 6, 9, 12, and 24 h) was based inpart on previous experience with PMN gene expression and timing fortriggering of the apoptosis differentiation program (typically between 3 and6 h after PMN-pathogen interaction) (16, 18, 22-25). Inasmuch as apopto-sis was not triggered by A. phagocytophilum at early time points (1.5–6 h)and was significantly repressed by 24 h, these time points were of interest.The 9-h time point was selected for analysis of S. aureus-induced changesin gene expression because the apoptosis differentiation program is wellunderway at that time and because it was the time at which the greatestnumber of genes was differentially expressed by A. phagocytophilum (Fig.2) (16, 18).

At each time point, culture medium was aspirated from the plate andPMNs were lysed with RLT buffer (Qiagen) supplemented with 2-ME.Purification of PMN RNA and preparation of labeled cRNA target wereperformed, as described (16, 18, 22, 24, 25). Labeling of samples, hybrid-ization of cRNA with HU133A oligonucleotide arrays (Affymetrix), andGeneChip scanning (GeneArray 2500; Affymetrix) were performed with

6365The Journal of Immunology

by guest on October 21, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

standard Affymetrix protocols (see www.affymetrix.com). Experimentswere performed with PMNs from three healthy donors using a separateHU133A GeneChip (Affymetrix) for each donor. For experiments in Fig.2, data were analyzed with Microarray Suite version 5.0 (Affymetrix) andGeneSpring expression analysis software version 6.0 (Silicon Genetics), asdescribed previously (16, 18, 22, 24, 25). For microarray experiments pre-sented in Figs. 3 and 4, data were analyzed with GeneChip OperatingSoftware version 1.2 (Affymetrix), which replaced Microarray Suite ver-sion 5.0, and GeneSpring software version 7.0 (Silicon Genetics). Further-more, Affymetrix microarrays for the experiments presented in Fig. 2 werescanned on a GeneArray 2500 scanner (Affymetrix), whereas those in Figs.3 and 4 were scanned on a new GeneChip Scanner 3000 (Affymetrix).Changes in gene expression were determined by comparing neutrophiltranscript levels from samples stimulated with A. phagocytophilum or S.aureus with those from control, unstimulated cells. Briefly, genes identifiedas differentially expressed were called “Present” by GeneChip OperatingSoftware version 1.2 (Affymetrix) in two of three individuals tested. Theaverage fold change in gene expression was at least 1.5-fold for A. phago-cytophilum-stimulated PMNs and/or 2.0-fold for those stimulated with S.aureus. Changes in PMN gene expression were also statistically significantin at least one of the pathogen treatments (at the level of p � 0.05, unpairedStudent’s t test). A complete set of microarray results is published as sup-plemental Table I (first set of microarray experiments encompassing a 24-htime course after PMN uptake, plus S. aureus data at 9 h) and supplementalTable II (second set of microarray experiments that compare live and heat-killed A. phagocytophilum at 9 and 24 h), supplemental material.4 Raw andnormalized microarray data are also posted online in the Gene ExpressionOmnibus (�www.ncbi.nim.nih.gov/geo�; accession no. GSE2405).

TaqMan real-time RT-PCR confirmation of microarray data

Phagocytosis experiments and RNA preparation for TaqMan analysis wereperformed with conditions identical with those used for microarray exper-iments (see above). DNA was removed from RNA samples by treatmentwith RNase-free DNase, as suggested by the manufacturer (Qiagen). RNAfrom three individuals was analyzed in duplicate wells with an ABI 7700thermocycler (Applied Biosystems), as described (18). Primers and probesets were designed with Primer Express version 1.5a and manufactured byApplied Biosystems. RNA was normalized using the Applied Biosystemspredeveloped assay reagent for human ribosomal RNA.

Flow cytometric evaluation of intracellular IL-1RN production

Following uptake of A. phagocytophilum (described above), intracellularproduction of IL-1RN by human PMNs was measured with mAb AS17conjugated to PE (BD Biosciences) using a previously described method(22). Briefly, staining for intracellular IL-1RN was performed 14 h fol-lowing initial exposure to A. phagocytophilum in the presence of GolgiPlug(brefeldin A), added 6 h before measurements, as suggested by the man-ufacturer (BD Biosciences). The ability to detect intracellular IL-1RN isdue to the capacity of brefeldin A (or GolgiPlug) to block endoplasmicreticulum-to-Golgi trafficking. The length of incubation with brefeldin Adictates in part the percentage of PMNs in which IL-1RN can be detected.The percentage of A. phagocytophilum-treated PMNs staining positive forintracellular IL-1RN was routinely 0–0.25% at 9 h, and this percentageincreased significantly over time to 6–8% by 14 h. This significant in-crease is due to accumulating cytokine in the endoplasmic reticulum aftertreatment with GolgiPlug. Thus, it is likely that all PMNs constitutivelyproduce IL-1RN, albeit not to detectable levels. Samples were analyzedwith a FACSCalibur flow cytometer (BD Biosciences), and a single gatewas used to exclude autofluorescent cells and debris.

Statistical analysis

Statistics were performed with Student’s t test (Microsoft Excel 2002; Mi-crosoft) or one-way ANOVA with a Tukey’s or Bonferroni’s posttest formultiple comparisons (GraphPad Prism version 4.0 for Windows;GraphPad).

ResultsIngestion kinetics, ROS production, and apoptosis of A.phagocytophilum-infected human PMNs

To evaluate ingestion kinetics of A. phagocytophilum by humanPMNs in our assay system, pathogen uptake was visualized and

quantified by fluorescence microscopy (Fig. 1, A and B). Com-pared with phagocytosis typically observed for other bacteria (16),there was relatively slow neutrophil uptake of A. phagocytophilum.However, most of the neutrophils (75.3 � 7.8%) contained in-gested bacteria 6 h after initiation of neutrophil-A. phagocytophi-lum interaction (Fig. 1B). In addition, PMNs generally contained

4 The online version of this article contains supplemental material.

FIGURE 1. Interaction of A. phagocytophilum with human PMNs. A,Ingestion of A. phagocytophilum by human PMNs. Neutrophils were in-cubated with �5–20 A. phagocytophilum/PMN for the indicated times, andpercentage of ingestion (number of PMNs containing internalized A.phagocytophilum) was determined by fluorescence microscopy. B, Micro-graphs illustrating ingestion of A. phagocytophilum by human PMNs.Green, intracellular bacteria. Red or red-green (yellow), extracellular bac-teria. C, A. phagocytophilum fails to trigger significant PMN ROS produc-tion. op S.a., Serum-opsonized S. aureus (�10 S. aureus/PMN); A.p. 0.5(�5–20 A. phagocyophilum/PMN); A.p. 1 (�10–40 A. phagocyophilum/PMN): �A.p., heat-killed A. phagocytophilum. The Vmax for PMN ROSproduction was 20 min for S. aureus-activated cells and usually 70 min forthose exposed to A. phagocytophilum. D, A. phagocytophilum delays PMNapoptosis. Neutrophils were incubated with live (left panel) or heat-killed(right panel) A. phagocytophilum (�5–20 bacteria per PMN) for the indi-cated times, and apoptosis was determined by morphological assessment ofnuclei. Results for A, C, and D are the mean � SEM of three to six separateexperiments.

6366 NEUTROPHIL RESPONSE TO A. phagocytophilum

by guest on October 21, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

high numbers of minimally degraded A. phagocytophilum (Fig.1B). The relatively slow PMN uptake of A. phagocytophilum isconsistent with a recent report by Carlyon et al. (13).

We next evaluated generation of intracellular ROS by humanPMNs during ingestion of A. phagocytophilum to determinewhether pathogen uptake activates neutrophils in our assay system(Fig. 1C). As with previous studies (10, 11, 13), neither live norheat-killed pathogens elicited significant PMN ROS productioncompared with either PMA or S. aureus (Fig. 1C). These findingsare consistent with an earlier report that demonstrated that inhibi-tion of neutrophil superoxide generation is not dependent on viableor structurally intact bacteria (10).

Bacterial phagocytosis normally induces neutrophil apoptosis(16, 23). In contrast, uptake of live or heat-killed A. phagocyto-philum failed to induce neutrophil apoptosis (Fig. 1D). Spontane-ous neutrophil apoptosis was also delayed significantly by live orheat-killed organisms 24 h after initial A. phagocytophilum-PMN

interaction (PMN apoptosis was inhibited 51.3 � 6.1% by livebacteria and 52.8 � 9.7% by heat-killed organisms compared withspontaneous apoptosis noted in control PMNs) (Fig. 1D). The abil-ity of live A. phagocytophilum to delay neutrophil apoptosis di-minished by 48 h (57.7 � 2.9% of the cells were apoptotic afterinfection with live organisms vs 70.1 � 10.5% for uninfectedcells; data not shown), indicating some of the processes underlyingthe delay in neutrophil apoptosis occur early during infection.These findings are consistent with reports by Yoshiie et al. andScaife et al. (6, 7) that suggest pathogen protein synthesis and/orintracellular proliferation are not required for the observed anti-apoptotic effect. The apoptosis differentiation program is typicallytriggered within hours of bacterial uptake (16, 23). Inasmuch asthere was a significant delay in neutrophil apoptosis 24 h afterinfection (Fig. 1), we examined transcript levels in neutrophils attime points up to 24 h.

FIGURE 2. Global changes inPMN gene expression following inges-tion of A. phagocytophilum. A, Neutro-phil gene expression was measuredwith human oligonucleotide microar-rays following ingestion of live A.phagocytophilum (�5–20 bacteria perPMN). Data were analyzed with Mi-croarray Suite version 5.0 (Affymetrix)and GeneSpring software version 6.0(Silicon Genetics, Redwood City, CA),as described (16, 18, 22, 24, 25). Dif-ferentially expressed genes were cate-gorized based on known or putativefunction and enumerated. CC&GF,chemokines, cytokines, and growth fac-tors; CAS&HD, cell adhesion and hostdefense; A&CF, apoptosis and cell fate;T&DB, transcription and DNA binding;M&VT, metabolism and vesicle traf-ficking; R&ST, receptors and signaltransduction; UNK, gene encoding pro-teins with unknown function. B, Meanfold induction or repression of specifichuman neutrophil genes following up-take of live A. phagocytophilum (A.p.,�5–20 bacteria per PMN) or S. aureus(S.a., �10 bacteria per PMN) at the in-dicated time(s). Scale bar indicates foldincrease or decrease in gene expression.In some instances, gene names are fol-lowed by the common name of the en-coded protein. Where more than oneAffymetrix probe set representedchanges for a single gene, the numberof probe sets is indicated in superscript.Results presented are the mean of threeexperiments (two or three blood donorswith assays done on separate days). GF,growth factors. Olive shading indicateschanges in gene expression generallysimilar between S. aureus and A.phagocytophilum.

6367The Journal of Immunology

by guest on October 21, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

A. phagocytophilum induces global changes in PMN geneexpression

To gain insight into the molecular processes that permit survival ofA. phagocytophilum within PMNs, we screened 14,500 humangenes for changes in expression following ingestion of A. phago-cytophilum (Fig. 2, and supplemental Table I).4 Compared withphagocytosis of S. aureus (Fig. 2) or other bacterial pathogens(16), ingestion of A. phagocytophilum induced far fewer changesin PMN gene expression (Figs. 2, A and B, and 3A). For example,2,040 human genes were differentially regulated 9 h after phago-cytosis of S. aureus by PMNs (Fig. 3A). By comparison, only 722genes (35.4% of those altered by ingestion of S. aureus) wereinduced or repressed by A. phagocytophilum at the same time point(Fig. 3A).

Previous work demonstrated that marked changes in PMN geneexpression occur by 3–6 h after phagocytosis, and accompany in-duction of apoptosis and down-regulation of proinflammatory ca-pacity (16, 22). The S. aureus strain COL used for comparison inthis study has been shown to elicit this typical response (16). Thus,24 h following interaction with S. aureus, neutrophils have under-gone apoptosis or primary necrosis and are no longer viable (ourunpublished observations). By comparison, neutrophils containingingested A. phagocytophilum 24 h after uptake were viable and hadthe greatest number of differentially regulated genes (1245) com-pared with any of the other times measured (Figs. 2A and 3, A andB). This response included genes involved in the acute inflamma-tory response, such as those encoding TNF-�, IL-1�, IL-1�, IL-6,CXCL1, CXCL2, CXCL3, CCL3, CCL4, CCL20, CD54, IL-1RN,IL-1R1, and orosomucoid (Figs. 2A and 4). The patterns of generegulation were also strikingly different between S aureus and A.phagocytophilum at 9 h postinfection (Fig. 2B). Notably, a signif-icant number of genes that regulate apoptosis and signal transduc-tion and encode structural proteins were down-regulated afterphagocytosis of S. aureus, but remained unchanged following in-gestion of A. phagocytophilum (Fig. 2B). Furthermore, manychanges in PMN gene expression following ingestion of A. phago-cytophilum were at variance with those in S. aureus-activated cells,or the time at which the change in expression occurred was de-layed (Figs. 2A and 4).

Live and heat-killed pathogen-induced alterations in PMN generegulation

Live and nonviable (heat-killed) A. phagocytophilum had similarcapacities to delay neutrophil apoptosis, and each failed to elicitgeneration of PMN ROS (Fig. 1, C and D). To test the hypothesisthat PMN gene regulation following uptake of live or heat-killedorganisms is concomitantly similar, we performed a second set ofmicroarray experiments to directly compare the ability of live andheat-killed A. phagocytophilum to modulate human neutrophilgene expression (Figs. 3 and 4, and supplemental Table II).4 Com-pared with live pathogens at 9 or 24 h after uptake, interaction withheat-killed A. phagocytophilum resulted in fewer alterations inPMN gene expression (Fig. 3). For example, 722 neutrophil geneswere differentially regulated by live A. phagocytophilum 9 h afteringestion compared with 409 genes (56.7% of those altered by livepathogens) induced or repressed by heat-killed bacteria (Fig. 3A).At 9 h following ingestion of A. phagocytophilum, there was aproportional number of up- and down-regulated genes. At 24 h,although the number of genes up-regulated by live and heat-killedA. phagocytophilum was generally similar (shared 72% of up-reg-ulated genes) (Fig. 3B), there were far fewer genes down-regulatedby heat-killed bacteria (only 8.3% of those down-regulated by livebacteria) (Fig. 3C). These results suggest that down-regulation of

PMN genes late during interaction with A. phagocytophilum maybe reliant on active intracellular infection, a finding that meritsfurther investigation (Fig. 3, B and C). In contrast, there was strik-ing similarity in the overall patterns of gene regulation withinfunctional categories between live and heat-killed A. phagocyto-philum (Fig. 4). Notably, there were large tracts of genes in allfunctional categories remaining unchanged after ingestion of liveor dead A. phagocytophilum that were differentially regulated fol-lowing phagocytosis of S. aureus (Fig. 4). Thus, our data may

FIGURE 3. Changes in PMN gene expression following uptake of liveand heat-killed A. phagocytophilum. A, Following bacteria-neutrophil in-teraction (�10 S. aureus per PMN, or �5–20 live or heat-killed A. phago-cytophilum per PMN), neutrophil gene expression was measured with oli-gonucleotide microarrays. Numbers of differentially expressed genes at 9 hafter bacteria-neutrophil interaction are compared between S. aureus, liveand heat-killed A. phagocytophilum (B). Numbers of differentially ex-pressed genes 24 h after ingestion of live or heat-killed A. phagocytophilum(C). Differentially expressed genes were categorized based on known orputative function and enumerated, as described in the legend of Fig. 2.Experiments were performed with PMNs from three healthy donors usinga separate HU133A GeneChip for each donor. Data were analyzed withGeneChip Operating Software version 1.2 (Affymetrix) and GeneSpringsoftware version 7.0 (Silicon Genetics, Redwood City, CA), as described inMaterials and Methods. Red bars, up-regulated genes; blue bars, down-regulated genes.

6368 NEUTROPHIL RESPONSE TO A. phagocytophilum

by guest on October 21, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

suggest that the majority of differences in PMN gene expressionbetween live and heat-killed A. phagocytophilum are primarily inthe time frame and magnitude of change.

Expression of genes encoding NADPH oxidase components

Previous studies in promyelocytic HL60 cells suggest that the abil-ity of A. phagocytophilum to inhibit ROS generation is linked todown-regulation of NADPH oxidase components (12–15). How-ever, this mechanism has not been demonstrated in human neu-trophils. To that end, we analyzed all Affymetrix probe sets en-coding NADPH oxidase proteins in our microarray data set (TableI). Genes encoding p47phox (NCF1), p40phox (NCF4), and Rap1A(RAP1A), regulators of NADPH oxidase, were down-regulatedfollowing phagocytosis of S. aureus (Fig. 2B). These genes, alongwith p67phox (NCF2), RAC2, and p22phox (CYBA), remained un-changed or were up-regulated after ingestion of A. phagocytophi-lum (Figs. 2B and 4, and Table I). Moreover, none of the genesencoding NADPH oxidase components, including CYBB(gp91phox) and CYBA (p22phox), were significantly down-regulatedafter A. phagocytophilum infection (Table I). These results clearlyindicate that failure of A. phagocytophilum to elicit production ofROS by PMNs, or to block ROS production by a second stimulus,is not due to repression of genes encoding NADPH oxidase pro-teins, as previously suggested (12–15).

Regulation of inflammatory response mediators after A.phagocytophilum infection

There were at least two general patterns of proinflammatory geneexpression apparent in A. phagocytophilum-infected PMNs (Fig.2). First, a group of genes encoding proinflammatory molecules,such as IL-1RN, TNF-�, IL-6, IL-1�, IL-1�, CCL20, IL-12�,CCL3, CCL4, CXCL1, CXCL2, CD16, IL-13RA1, and CD18,was similarly up- or down-regulated by uptake of S. aureus and A.phagocytophilum (Figs. 2 and 4). However, induction or repressionof many of these genes by A. phagocytophilum was blunted ordelayed compared with that by S. aureus (Fig. 2). Second, therewere genes for which expression was repressed following phago-cytosis of S. aureus, but transcript levels were up-regulated orremained unchanged after ingestion of A. phagocytophilum. Thiscategory included genes encoding lymphotoxin B, TGF-�1, IL-16,IL-17R, CD32, CD47, CD114 (G-CSF receptor), IL-1R2, MyD88,TLR4, leukocyte Ig-like receptors, STAT1, STAT3, STAT5B,STAT6, and formyl peptide receptor 1 (Fig. 2). These findingswere confirmed by TaqMan real-time RT-PCR analysis, whichwas performed for six genes representative of the entire microarraydata set and included three proinflammatory molecules (Fig. 5A).In addition, we used flow cytometry to confirm that IL-1RN pro-tein accumulated following ingestion of A. phagocytophilum (Fig.5B). The finding that A. phagocytophilum-infected PMNs retainedproinflammatory capacity is consistent with the observation thatthe pathogen failed to induce neutrophil apoptosis.

A. phagocytophilum fails to induce an apoptosis differentiationprogram in human neutrophils

A. phagocytophilum did not induce rapid PMN apoptosis that usu-ally follows phagocytosis and ROS production (Fig. 1). As such,the expression of genes involved in apoptosis was examined

categorization of genes, olive shading, superscript numbers, and parenthet-ical annotation are described in the legend of Fig. 2. Results are presentedas the mean fold induction or repression of genes from three experiments(three blood donors with assays done on separate days).

FIGURE 4. Global changes in PMN gene expression following inges-tion of live (A.p.) or heat-killed (� A.p.) A. phagocytophilum or live S.aureus (S.a.) at the indicated times. Neutrophil gene expression data wereanalyzed with GeneChip Operating Software version 1.2 (Affymetrix),which replaced Microarray Suite version 5.0 and GeneSpring software ver-sion 7.0 (Silicon Genetics, Redwood City, CA). Affymetrix microarrayswere scanned on a new GeneChip Scanner 3000 (Affymetrix). Functional

6369The Journal of Immunology

by guest on October 21, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

closely. Previous work has shown that many genes encoding ap-optosis regulators are typically induced or repressed within 3–6 hafter PMN phagocytosis (18). Notably, two such apoptosis regu-lators, B cell line 2-associated X-protein (BAX) and nuclear or-phan receptor TR3 (NR4A1), remained unchanged following up-take of A. phagocytophilum (18) (supplemental Tables I and II).4

Recently, we used PMNs from patients with chronic granuloma-tous disease to demonstrate that ROS modulate expression of BAXand are an important component of the PMN apoptosis differen-tiation program induced by phagocytosis (24). Thus, our findingthat A. phagocytophilum-infected PMNs do not elicit significantROS production (Fig. 1C) most likely explains why expression ofBAX remained unchanged (data not shown). Genes encoding ap-optotic protease-activating factor 1, BAX-inhibitor 1 (testis en-hanced gene transcript), B cell chronic lymphocytic leukemia/lym-phoma 10, caspase 3, caspase 8, glucocorticoid receptor (NRC31),programmed cell death proteins 6 and 10, myeloid cell leukemiasequence 1, p21-activated kinase, JUN, tumor up-regulatedcaspase recruitment domain-containing antagonist of caspase 9,TNF superfamily member 10, TNF receptor superfamily, member10B, and TNF receptor superfamily, member 1A-associated viadeath domain (TNFR-associated death domain protein), were in-duced or repressed after phagocytosis of S. aureus (Figs. 2 and 4).By contrast, these genes remained unchanged following uptake ofA. phagocytophilum or were at variance with changes in S. aureus(Figs. 2 and 4). Importantly, several antiapoptosis genes, includingBIRC2, BIRC3, CFLAR, TNFAIP8, and TNIP2, were up-regulatedin PMNs after uptake of live or dead A. phagocytophilum (Fig. 4).These data are consistent with the ability of live or heat-killed A.phagocytophilum to delay normal neutrophil apoptosis (Fig. 1D).Taken together, the results indicate that A. phagocytophilum failedto induce the apoptosis differentiation program in human PMNs.Importantly, the data provide an underlying molecular basis for theinability of A. phagocytophilum to trigger PMN apoptosis.

A. phagocytophilum inhibits Fas-induced PMN apoptosis

Although our results explain the inability of A. phagocytophilum toinduce PMN apoptosis, it is not known whether the pathogen ac-tively blocks neutrophil death engaged by external stimuli. There-fore, we tested the ability of live and heat-killed A. phagocytophi-lum to inhibit Fas-induced PMN apoptosis (Fig. 5C). Consistentwith results in Fig. 1D, viable and nonviable A. phagocytophiluminhibited Fas-induced cell death to similar degrees (inhibition ofFas-induced PMN apoptosis was 69.3 � 8.5% with live organisms( p � 0.001) and 68.3 � 9.5% with heat-killed organisms ( p �0.01)) (Fig. 5C, left panel). These findings indicate that a pre-existing surface structure or protein promotes neutrophil survivalduring A. phagocytophilum infection, presumably to facilitategrowth and replication of the pathogen.

DiscussionThere is limited understanding of how bacterial pathogens evadehuman innate host defense to cause disease. Inasmuch as neutro-phils are the most prominent innate immune effector cell in hu-mans, the ability of microorganisms to subvert PMN killing mech-anisms is essential for pathogenesis. A. phagocytophilum fails totrigger production of PMN ROS and delays neutrophil apoptosis,thereby promoting pathogen survival within granulocytes (6, 7,10–15). Although progress has been made, the molecular basis forA. phagocytophilum pathogenesis is not known.

To that end, we investigated global changes in human neutrophilgene expression following uptake of A. phagocytophilum andgained new insight into pathogen survival mechanisms. First, wefound that neutrophils infected with A. phagocytophilum graduallyup-regulated proinflammatory capacity, which peaked at 24 hpostinfection, the latest time point studied. In contrast, there areimmediate (within 30 min) and dramatic increases in neutrophil

Table I. Expression of NADPH oxidase components in human neutrophils following ingestion of A. phagocytophiluma

Gene Name Live/HK Unigene Encoded Protein

Fold change

1.5 h 3 h 6 h 9 h 12 h 24 h

CYBB1 Live A.p. Hs.88974 gp91phox nc nc nc nc nc 1.8CYBB2 Live A.p. Hs.88974 gp91phox nc nc 1.6 nc nc 1.6CYBB1 HK A.p. Hs.88974 gp91phox nc 1.8CYBB2 HK A.p. Hs.88974 gp91phox nc 2.0CYBA1 Live A.p. Hs.68877 p22phox nc nc nc nc nc 2.5CYBA1 HK A.p. Hs.68877 p22phox nc 1.8*NCF11 Live A.p. Hs.458275 p47phox nc nc nc nc nc 2.0NCF12 Live A.p. Hs.448231 p47phox nc nc nc nc nc 2.8NCF11 HK A.p. Hs.1583 p47phox nc 2.1*NCF12 HK A.p. Hs.1583 p47phox nc 2.0*NCF21 Live A.p. Hs.949 p67phox nc nc 1.5 nc nc 1.9NCF21 HK A.p. Hs.949 p67phox nc 1.7*NCF41 Live A.p. Hs.196352 p40 phox nc nc nc nc nc ncNCF42 Live A.p. Hs.196352 p40 phox nc nc nc nc nc 1.9NCF41 HK A.p. Hs.196352 p40 phox 1.7 2.1*NCF42 HK A.p. Hs.196352 p40 phox nc 1.9RAC21 Live A.p. Hs.301175 Rac2 nc nc nc nc nc ncRAC22 Live A.p. Hs.301175 Rac2 nc nc 1.8 nc nc 1.9RAC21 HK A.p. Hs.301175 Rac2 1.5 2.5*RAC22 HK A.p. Hs.301175 Rac2 nc 1.9*RAP1A1 Live A.p. Hs.865 Rap1A nc 1.5 1.7 nc nc ncRAP1A1 HK A.p. Hs.865 Rap1A nc 2.3*

a Live or heat-killed (HK) A. phagocytophilum was incubated with human PMNs for the indicated times, and expression of genes encoding NADPH oxidase proteins wasmeasured with Affymetrix microarrays (HU133A), as described in Materials and Methods. Data are the mean fold change in genes from two or three individuals. The gene datafor infection with live organisms is derived from the first set of microarray experiments. nc (no change), Gene failed to meet criteria for being differentially expressed (seeMaterials and Methods). HK, heat-killed A. phagocytophilum. *, p � 0.05 vs unstimulated (clarified HL60 lysate-treated) PMNs. Superscript numbers indicate paired probe sets.

6370 NEUTROPHIL RESPONSE TO A. phagocytophilum

by guest on October 21, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

proinflammatory capacity following receptor-mediated phagocyto-sis or ingestion of other bacterial pathogens, which (with receptor-mediated phagocytosis) diminish dramatically by 24 h (16, 18, 22).Our observation that there is a delay in proinflammatory gene up-regulation may help bridge previous findings indicating an absenceof PMN proinflammatory cytokine production (8, 9) with concur-rent neutrophil activation and inflammatory disease (26, 27) duringA. phagocytophilum infection. The delay in PMN proinflammatoryresponses following infection with A. phagocytophilum might pro-mote intracellular survival by limiting early neutrophil signalingand recruitment/activation of other immune cells. Furthermore, re-tention of proinflammatory capacity in A. phagocytophilum-in-fected cells is consistent with the observation that the pathogenfails to induce PMN apoptosis.

Second, our data clearly demonstrate that genes encoding com-ponents of NADPH oxidase, including gp91phox and Rac2, remainunchanged or are up-regulated in human neutrophils during A.phagocytophilum infection (Table I). These results are at variancewith previous studies that suggest the inability of PMNs to produceROS production following uptake of A. phagocytophilum is due todecreased gp91phox and/or Rac2 transcript levels (12, 14). How-ever, those studies were limited in scope and/or used HL60 cellsrather than human PMNs (12, 14). HL60 cells are an immortalizedcell line with significant differences in functional capacity com-pared with human neutrophils. Thus, the relevance of HL60 infec-tion models to human disease is unclear. The inability of A. phago-cytophilum to trigger generation of ROS may be due in part touptake by host cell endocytosis rather than phagocytosis (4, 5, 28).In addition, recent data support alterations in NADPH oxidase as-sembly and superoxide scavenging as the basis for failure of thepathogen to directly elicit significant ROS production, findingsmost compatible with our current data (13).

Third, our results demonstrate that failure of A. phagocytophi-lum to induce the neutrophil apoptosis differentiation program inpart underlies intracellular survival and represents a novel mech-anism for pathogen immune evasion. As pathogen uptake also failsto elicit PMN ROS production, it may be that inability to inducethe apoptosis differentiation program is secondary to the absenceof ROS production. This proposed mechanism is clearly distinctfrom that underlying the ability of A. phagocytophilum to delayspontaneous neutrophil apoptosis or block Fas-induced cell death(Figs. 1D and 5C). It is possible that changes in expression of agene or group of genes, such as induction of those encoding an-tiapoptosis molecules, contribute to these latter phenomena (Fig. 4and Table I).

Interestingly, live and heat-killed bacteria modulated early neu-trophil processes (inhibition of apoptosis and failure to elicit ROSgeneration) to similar degrees. Moreover, patterns of PMN geneexpression induced by heat-killed bacteria were in general similarto those induced by viable A. phagocytophilum, but generally dis-similar to the patterns induced by S. aureus (Figs. 2 and 4). Thisobservation suggests that there is an A. phagocytophilum-specificresponse, very different from that in bacteria studied to date (e.g.,S. aureus), which most likely involves in part a bacterial surfacestructure or protein. The observation that a bacterial surface struc-ture is sufficient to alter PMN function has been reported by others(6, 10). At 9 h, over one-half of the genes differentially regulatedby live A. phagocytophilum were similarly altered by heat-killedpathogen. At 24 h, up-regulated genes were, again, similar. Con-versely, interaction with heat-killed pathogen by 24 h resulted infar fewer down-regulated genes compared with uptake of viableorganisms. Although the number of PMN genes down-regulatedby live or dead A. phagocytophilum at 24 h represents a minorpercentage of the overall gene expression data at that time point(28.5%), these differences may suggest that living organisms arenecessary for ongoing functional alterations in PMNs infected withA. phagocytophilum. Alternatively, this phenomenon may simplybe due to bacterial replication and the associated increase in bac-terial burden in viable infection compared with PMNs exposed tononviable organisms (29). Finally, heat killing or bacterial fixation,in and of itself, may alter the efficiency with which a surface struc-ture or protein can modulate neutrophil function. In the end, theobservation that there are differences between live and heat-killedA. phagocytophilum in numbers of down-regulated genes at 24 hstill fails to explain the similar capacity of each to block apoptosiswithin that same time frame (Fig. 1). Thus, the mechanisms bywhich PMN functions are altered by long-term (�24-h) infectionwith A. phagocytophilum warrant additional study.

FIGURE 5. Confirmation of microarray data and inhibition of Fas-in-duced neutrophil apoptosis. A, TaqMan analysis of PMN gene expression.There was strong correlation (88.9%) between TaqMan and microarraydata. Results are the mean � SEM fold change in PMN transcript levelsfrom three individuals. �, p � 0.05 vs unstimulated cells. B, Left panel,Intracellular accumulation of neutrophil IL-1RN 14 h following ingestionof A. phagocytophilum. Right panel, Mean � SEM of five experiments. C,A. phagocytophilum inhibits neutrophil apoptosis triggered by anti-FasmAb. Neutrophils were preincubated with live or heat-killed A. phagocy-tophilum (�5–20 bacteria per PMN), and then cultured with 500 ng/mlanti-Fas mAb, clone CH-11 for 6 h. �A.p. or �lysate, A. phagocytophilumor clarified lysate from uninfected HL60 cells that was heated at 95°C for10 min. Apoptosis was measured with a modified TUNEL assay using aflow cytometer. Results are the mean � SEM of three experiments.

6371The Journal of Immunology

by guest on October 21, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Based on our current studies and other recent reports, the abilityof bacteria to alter normal PMN apoptosis is most likely a com-ponent of pathogenesis (23). Importantly, our study suggests thatfailure of A. phagocytophilum to induce the neutrophil apoptosisdifferentiation program and concomitant ability to delay/blockPMN apoptosis represent two separate mechanisms used by A.phagocytophilum to prolong pathogen survival. Prolonged patho-gen survival permits ongoing PMN exposure to a host of patho-genic mechanisms. The combined strategies implemented by A.phagocytophilum to prolong neutrophil survival may thus repre-sent a general mechanism used by obligate intracellular pathogensto evade human innate host defense, thereby contributing to dis-ease. Notably, our studies provide a global view of the host-patho-gen interface important for our understanding of human diseasescaused by intracellular bacteria.

AcknowledgmentsWe thank Kelly O’Donnell for technical assistance.

DisclosuresThe authors have no financial conflict of interest.

References1. Jones, S., F. Lindberg, and E. Brown. 2003. Phagocytosis. In Fundamental Im-

munology. W. E. Paul, ed. Lippincott Williams & Wilkins, Philadelphia, p. 997–1020.

2. Chen, S. M., J. S. Dumler, J. S. Bakken, and D. H. Walker. 1994. Identificationof a granulocytotropic Ehrlichia species as the etiologic agent of human disease.J. Clin. Microbiol. 32: 589–595.

3. Dumler, J. S., A. F. Barbet, C. P. Bekker, G. A. Dasch, G. H. Palmer, S. C. Ray,Y. Rikihisa, and F. R. Rurangirwa. 2001. Reorganization of genera in the familiesRickettsiaceae and Anaplasmataceae in the order Rickettsiales: unification ofsome species of Ehrlichia with Anaplasma, Cowdria with Ehrlichia and Ehrli-chia with Neorickettsia, descriptions of six new species combinations and des-ignation of Ehrlichia equi and ‘HGE agent’ as subjective synonyms of Ehrlichiaphagocytophila. Int. J. Syst. Evol. Microbiol. 51: 2145–2165.

4. Herron, M. J., C. M. Nelson, J. Larson, K. R. Snapp, G. S. Kansas, andJ. L. Goodman. 2000. Intracellular parasitism by the human granulocytic ehrli-chiosis bacterium through the P-selectin ligand, PSGL-1. Science 288:1653–1656.

5. Webster, P., J. W. IJdo, L. M. Chicoine, and E. Fikrig. 1998. The agent of humangranulocytic ehrlichiosis resides in an endosomal compartment. J. Clin. Invest.101: 1932–1941.

6. Yoshiie, K., H. Y. Kim, J. Mott, and Y. Rikihisa. 2000. Intracellular infection bythe human granulocytic ehrlichiosis agent inhibits human neutrophil apoptosis.Infect. Immun. 68: 1125–1133.

7. Scaife, H., Z. Woldehiwet, C. A. Hart, and S. W. Edwards. 2003. Anaplasmaphagocytophilum reduces neutrophil apoptosis in vivo. Infect. Immun. 71:1995–2001.

8. Klein, M. B., S. Hu, C. C. Chao, and J. L. Goodman. 2000. The agent of humangranulocytic ehrlichiosis induces the production of myelosuppressing chemo-kines without induction of proinflammatory cytokines. J. Infect. Dis. 182:200–205.

9. Kim, H. Y., and Y. Rikihisa. 2000. Expression of interleukin-1�, tumor necrosisfactor �, and interleukin-6 in human peripheral blood leukocytes exposed tohuman granulocytic ehrlichiosis agent or recombinant major surface protein P44.Infect. Immun. 68: 3394–3402.

10. Mott, J., and Y. Rikihisa. 2000. Human granulocytic ehrlichiosis agent inhibitssuperoxide anion generation by human neutrophils. Infect. Immun. 68:6697–6703.

11. Wang, T., S. E. Malawista, U. Pal, M. Grey, J. Meek, M. Akkoyunlu, V. Thomas,and E. Fikrig. 2002. Superoxide anion production during Anaplasma phagocy-tophila infection. J. Infect. Dis. 186: 274–280.

12. Carlyon, J. A., W. T. Chan, J. Galan, D. Roos, and E. Fikrig. 2002. Repressionof rac2 mRNA expression by Anaplasma phagocytophila is essential to the in-hibition of superoxide production and bacterial proliferation. J. Immunol. 169:7009–7018.

13. Carlyon, J. A., D. Abdel-Latif, M. Pypaert, P. Lacy, and E. Fikrig. 2004.Anaplasma phagocytophilum utilizes multiple host evasion mechanisms to thwartNADPH oxidase-mediated killing during neutrophil infection. Infect. Immun. 72:4772–4783.

14. Banerjee, R., J. Anguita, D. Roos, and E. Fikrig. 2000. Cutting edge: infection bythe agent of human granulocytic ehrlichiosis prevents the respiratory burst bydown-regulating gp91phox. J. Immunol. 164: 3946–3949.

15. Mott, J., Y. Rikihisa, and S. Tsunawaki. 2002. Effects of Anaplasma phagocy-tophila on NADPH oxidase components in human neutrophils and HL-60 cells.Infect. Immun. 70: 1359–1366.

16. Kobayashi, S. D., K. R. Braughton, A. R. Whitney, J. M. Voyich, T. G. Schwan,J. M. Musser, and F. R. DeLeo. 2003. Bacterial pathogens modulate an apoptosisdifferentiation program in human neutrophils. Proc. Natl. Acad. Sci. USA 100:10948–10953.

17. Kobayashi, S. D., and F. R. DeLeo. 2004. An apoptosis differentiation pro-gramme in human polymorphonuclear leukocytes. Biochem. Soc. Trans. 32:474–476.

18. Kobayashi, S. D., J. M. Voyich, C. L. Buhl, R. M. Stahl, and F. R. DeLeo. 2002.Global changes in gene expression by human polymorphonuclear leukocytes dur-ing receptor-mediated phagocytosis: cell fate is regulated at the level of geneexpression. Proc. Natl. Acad. Sci. USA 99: 6901–6906.

19. Telford, S. R., III, J. E. Dawson, P. Katavolos, C. K. Warner, C. P. Kolbert, andD. H. Persing. 1996. Perpetuation of the agent of human granulocytic ehrlichiosisin a deer tick-rodent cycle. Proc. Natl. Acad. Sci. USA 93: 6209–6214.

20. Hodzic, E., J. W. Ijdo, S. Feng, P. Katavolos, W. Sun, C. H. Maretzki, D. Fish,E. Fikrig, S. R. Telford III, and S. W. Barthold. 1998. Granulocytic ehrlichiosisin the laboratory mouse. J. Infect. Dis. 177: 737–745.

21. DeLeo, F. R., L. A. Allen, M. Apicella, and W. M. Nauseef. 1999. NADPHoxidase activation and assembly during phagocytosis. J. Immunol. 163:6732–6740.

22. Kobayashi, S. D., J. M. Voyich, K. R. Braughton, and F. R. DeLeo. 2003. Down-regulation of proinflammatory capacity during apoptosis in human polymorpho-nuclear leukocytes. J. Immunol. 170: 3357–3368.

23. DeLeo, F. R. 2004. Modulation of phagocyte apoptosis by bacterial pathogens.Apoptosis 9: 399–413.

24. Kobayashi, S. D., J. M. Voyich, K. R. Braughton, A. R. Whitney, W. M. Nauseef,H. L. Malech, and F. R. DeLeo. 2004. Gene expression profiling provides insightinto the pathophysiology of chronic granulomatous disease. J. Immunol. 172:636–643.

25. Kobayashi, S. D., J. M. Voyich, G. A. Somerville, K. R. Braughton,H. L. Malech, J. M. Musser, and F. R. DeLeo. 2003. An apoptosis-differentiationprogram in human polymorphonuclear leukocytes facilitates resolution of inflam-mation. J. Leukocyte Biol. 73: 315–322.

26. Borjesson, D. L., S. I. Simon, E. Hodzic, H. E. DeCock, C. M. Ballantyne, andS. W. Barthold. 2003. Roles of neutrophil �2 integrins in kinetics of bacteremia,extravasation, and tick acquisition of Anaplasma phagocytophila in mice. Blood101: 3257–3264.

27. Choi, K. S., D. J. Grab, and J. S. Dumler. 2004. Anaplasma phagocytophiluminfection induces protracted neutrophil degranulation. Infect. Immun. 72:3680–3683.

28. Mott, J., R. E. Barnewall, and Y. Rikihisa. 1999. Human granulocytic ehrlichiosisagent and Ehrlichia chaffeensis reside in different cytoplasmic compartments inHL-60 cells. Infect. Immun. 67: 1368–1378.

29. Branger, S., J. M. Rolain, and D. Raoult. 2004. Evaluation of antibiotic suscep-tibilities of Ehrlichia canis, Ehrlichia chaffeensis, and Anaplasma phagocytophi-lum by real-time PCR. Antimicrob. Agents Chemother. 48: 4822–4828.

6372 NEUTROPHIL RESPONSE TO A. phagocytophilum

by guest on October 21, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from